diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,62820 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 8966,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval_loss": 2.3125102519989014,
+      "eval_runtime": 636.0298,
+      "eval_samples_per_second": 99.225,
+      "eval_steps_per_second": 1.552,
+      "step": 0
+    },
+    {
+      "epoch": 0.00022306491188935982,
+      "grad_norm": 447.7063293457031,
+      "learning_rate": 0.0,
+      "loss": 2.6996,
+      "step": 1
+    },
+    {
+      "epoch": 0.00044612982377871963,
+      "grad_norm": 410.13433837890625,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 2.5839,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006691947356680794,
+      "grad_norm": 514.7770385742188,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.538,
+      "step": 3
+    },
+    {
+      "epoch": 0.0008922596475574393,
+      "grad_norm": 73.62993621826172,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 2.5648,
+      "step": 4
+    },
+    {
+      "epoch": 0.001115324559446799,
+      "grad_norm": 35.405982971191406,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 2.5318,
+      "step": 5
+    },
+    {
+      "epoch": 0.0013383894713361588,
+      "grad_norm": 31.43743133544922,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.4902,
+      "step": 6
+    },
+    {
+      "epoch": 0.0015614543832255187,
+      "grad_norm": 32.07968521118164,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 2.4386,
+      "step": 7
+    },
+    {
+      "epoch": 0.0017845192951148785,
+      "grad_norm": 19.430936813354492,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 2.0901,
+      "step": 8
+    },
+    {
+      "epoch": 0.002007584207004238,
+      "grad_norm": 20.800960540771484,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 2.1737,
+      "step": 9
+    },
+    {
+      "epoch": 0.002230649118893598,
+      "grad_norm": 20.56642723083496,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 2.1823,
+      "step": 10
+    },
+    {
+      "epoch": 0.002453714030782958,
+      "grad_norm": 20.692852020263672,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.1278,
+      "step": 11
+    },
+    {
+      "epoch": 0.0026767789426723177,
+      "grad_norm": 14.481375694274902,
+      "learning_rate": 2.2e-06,
+      "loss": 1.9787,
+      "step": 12
+    },
+    {
+      "epoch": 0.0028998438545616775,
+      "grad_norm": 13.366188049316406,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.9635,
+      "step": 13
+    },
+    {
+      "epoch": 0.0031229087664510374,
+      "grad_norm": 14.910027503967285,
+      "learning_rate": 2.6e-06,
+      "loss": 1.8838,
+      "step": 14
+    },
+    {
+      "epoch": 0.0033459736783403972,
+      "grad_norm": 13.10051441192627,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.8825,
+      "step": 15
+    },
+    {
+      "epoch": 0.003569038590229757,
+      "grad_norm": 14.625020027160645,
+      "learning_rate": 3e-06,
+      "loss": 1.6607,
+      "step": 16
+    },
+    {
+      "epoch": 0.0037921035021191165,
+      "grad_norm": 9.08655834197998,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.5911,
+      "step": 17
+    },
+    {
+      "epoch": 0.004015168414008476,
+      "grad_norm": 8.972038269042969,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 1.5538,
+      "step": 18
+    },
+    {
+      "epoch": 0.004238233325897837,
+      "grad_norm": 7.949400901794434,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.5478,
+      "step": 19
+    },
+    {
+      "epoch": 0.004461298237787196,
+      "grad_norm": 7.048157215118408,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 1.3381,
+      "step": 20
+    },
+    {
+      "epoch": 0.004684363149676556,
+      "grad_norm": 7.008754730224609,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.3275,
+      "step": 21
+    },
+    {
+      "epoch": 0.004907428061565916,
+      "grad_norm": 6.20928955078125,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.2469,
+      "step": 22
+    },
+    {
+      "epoch": 0.005130492973455275,
+      "grad_norm": 4.8463029861450195,
+      "learning_rate": 4.4e-06,
+      "loss": 1.1167,
+      "step": 23
+    },
+    {
+      "epoch": 0.005353557885344635,
+      "grad_norm": 20.94325065612793,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 1.1391,
+      "step": 24
+    },
+    {
+      "epoch": 0.005576622797233995,
+      "grad_norm": 3.783233404159546,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.0692,
+      "step": 25
+    },
+    {
+      "epoch": 0.005799687709123355,
+      "grad_norm": 3.780400276184082,
+      "learning_rate": 5e-06,
+      "loss": 1.0715,
+      "step": 26
+    },
+    {
+      "epoch": 0.0060227526210127145,
+      "grad_norm": 3.059216260910034,
+      "learning_rate": 5.2e-06,
+      "loss": 1.0057,
+      "step": 27
+    },
+    {
+      "epoch": 0.006245817532902075,
+      "grad_norm": 2.7420809268951416,
+      "learning_rate": 5.400000000000001e-06,
+      "loss": 0.9421,
+      "step": 28
+    },
+    {
+      "epoch": 0.006468882444791434,
+      "grad_norm": 2.8878915309906006,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 0.9625,
+      "step": 29
+    },
+    {
+      "epoch": 0.0066919473566807944,
+      "grad_norm": 2.9568190574645996,
+      "learning_rate": 5.8e-06,
+      "loss": 0.924,
+      "step": 30
+    },
+    {
+      "epoch": 0.006915012268570154,
+      "grad_norm": 2.239628314971924,
+      "learning_rate": 6e-06,
+      "loss": 0.8886,
+      "step": 31
+    },
+    {
+      "epoch": 0.007138077180459514,
+      "grad_norm": 2.2312073707580566,
+      "learning_rate": 6.200000000000001e-06,
+      "loss": 0.9105,
+      "step": 32
+    },
+    {
+      "epoch": 0.0073611420923488735,
+      "grad_norm": 1.7348324060440063,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 0.8685,
+      "step": 33
+    },
+    {
+      "epoch": 0.007584207004238233,
+      "grad_norm": 1.3611754179000854,
+      "learning_rate": 6.600000000000001e-06,
+      "loss": 0.8088,
+      "step": 34
+    },
+    {
+      "epoch": 0.007807271916127593,
+      "grad_norm": 1.2034807205200195,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 0.8101,
+      "step": 35
+    },
+    {
+      "epoch": 0.008030336828016953,
+      "grad_norm": 1.3573353290557861,
+      "learning_rate": 7e-06,
+      "loss": 0.8048,
+      "step": 36
+    },
+    {
+      "epoch": 0.008253401739906312,
+      "grad_norm": 1.4660226106643677,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 0.7663,
+      "step": 37
+    },
+    {
+      "epoch": 0.008476466651795673,
+      "grad_norm": 1.4917473793029785,
+      "learning_rate": 7.4e-06,
+      "loss": 0.7847,
+      "step": 38
+    },
+    {
+      "epoch": 0.008699531563685033,
+      "grad_norm": 0.9017640352249146,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 0.7472,
+      "step": 39
+    },
+    {
+      "epoch": 0.008922596475574392,
+      "grad_norm": 0.9448667764663696,
+      "learning_rate": 7.800000000000002e-06,
+      "loss": 0.7384,
+      "step": 40
+    },
+    {
+      "epoch": 0.009145661387463751,
+      "grad_norm": 1.0112793445587158,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 0.735,
+      "step": 41
+    },
+    {
+      "epoch": 0.009368726299353113,
+      "grad_norm": 0.8147673606872559,
+      "learning_rate": 8.2e-06,
+      "loss": 0.7463,
+      "step": 42
+    },
+    {
+      "epoch": 0.009591791211242472,
+      "grad_norm": 0.8109986782073975,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.7275,
+      "step": 43
+    },
+    {
+      "epoch": 0.009814856123131831,
+      "grad_norm": 0.6843619346618652,
+      "learning_rate": 8.6e-06,
+      "loss": 0.7413,
+      "step": 44
+    },
+    {
+      "epoch": 0.01003792103502119,
+      "grad_norm": 0.5661330223083496,
+      "learning_rate": 8.8e-06,
+      "loss": 0.7226,
+      "step": 45
+    },
+    {
+      "epoch": 0.01026098594691055,
+      "grad_norm": 0.8250797986984253,
+      "learning_rate": 9e-06,
+      "loss": 0.7099,
+      "step": 46
+    },
+    {
+      "epoch": 0.010484050858799911,
+      "grad_norm": 0.5656530857086182,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 0.7145,
+      "step": 47
+    },
+    {
+      "epoch": 0.01070711577068927,
+      "grad_norm": 0.7327658534049988,
+      "learning_rate": 9.4e-06,
+      "loss": 0.716,
+      "step": 48
+    },
+    {
+      "epoch": 0.01093018068257863,
+      "grad_norm": 0.5881125330924988,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 0.6747,
+      "step": 49
+    },
+    {
+      "epoch": 0.01115324559446799,
+      "grad_norm": 0.5523539185523987,
+      "learning_rate": 9.800000000000001e-06,
+      "loss": 0.6521,
+      "step": 50
+    },
+    {
+      "epoch": 0.01137631050635735,
+      "grad_norm": 0.5446374416351318,
+      "learning_rate": 1e-05,
+      "loss": 0.6752,
+      "step": 51
+    },
+    {
+      "epoch": 0.01159937541824671,
+      "grad_norm": 0.475178599357605,
+      "learning_rate": 1.02e-05,
+      "loss": 0.7108,
+      "step": 52
+    },
+    {
+      "epoch": 0.01182244033013607,
+      "grad_norm": 0.7163774967193604,
+      "learning_rate": 1.04e-05,
+      "loss": 0.6717,
+      "step": 53
+    },
+    {
+      "epoch": 0.012045505242025429,
+      "grad_norm": 0.4185207784175873,
+      "learning_rate": 1.0600000000000002e-05,
+      "loss": 0.6645,
+      "step": 54
+    },
+    {
+      "epoch": 0.012268570153914788,
+      "grad_norm": 0.43035364151000977,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 0.6814,
+      "step": 55
+    },
+    {
+      "epoch": 0.01249163506580415,
+      "grad_norm": 0.37698522210121155,
+      "learning_rate": 1.1000000000000001e-05,
+      "loss": 0.7271,
+      "step": 56
+    },
+    {
+      "epoch": 0.012714699977693509,
+      "grad_norm": 0.41325655579566956,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 0.6718,
+      "step": 57
+    },
+    {
+      "epoch": 0.012937764889582868,
+      "grad_norm": 0.4455505907535553,
+      "learning_rate": 1.14e-05,
+      "loss": 0.6461,
+      "step": 58
+    },
+    {
+      "epoch": 0.013160829801472228,
+      "grad_norm": 0.39668968319892883,
+      "learning_rate": 1.16e-05,
+      "loss": 0.6744,
+      "step": 59
+    },
+    {
+      "epoch": 0.013383894713361589,
+      "grad_norm": 0.43382027745246887,
+      "learning_rate": 1.18e-05,
+      "loss": 0.686,
+      "step": 60
+    },
+    {
+      "epoch": 0.013606959625250948,
+      "grad_norm": 0.3178131878376007,
+      "learning_rate": 1.2e-05,
+      "loss": 0.6461,
+      "step": 61
+    },
+    {
+      "epoch": 0.013830024537140308,
+      "grad_norm": 0.36169150471687317,
+      "learning_rate": 1.22e-05,
+      "loss": 0.6521,
+      "step": 62
+    },
+    {
+      "epoch": 0.014053089449029667,
+      "grad_norm": 0.3932245671749115,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 0.7118,
+      "step": 63
+    },
+    {
+      "epoch": 0.014276154360919028,
+      "grad_norm": 0.3924449384212494,
+      "learning_rate": 1.2600000000000001e-05,
+      "loss": 0.6723,
+      "step": 64
+    },
+    {
+      "epoch": 0.014499219272808388,
+      "grad_norm": 0.32354405522346497,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 0.6589,
+      "step": 65
+    },
+    {
+      "epoch": 0.014722284184697747,
+      "grad_norm": 0.3272797465324402,
+      "learning_rate": 1.3000000000000001e-05,
+      "loss": 0.6413,
+      "step": 66
+    },
+    {
+      "epoch": 0.014945349096587107,
+      "grad_norm": 0.36967551708221436,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 0.6432,
+      "step": 67
+    },
+    {
+      "epoch": 0.015168414008476466,
+      "grad_norm": 0.2998528480529785,
+      "learning_rate": 1.3400000000000002e-05,
+      "loss": 0.6543,
+      "step": 68
+    },
+    {
+      "epoch": 0.015391478920365827,
+      "grad_norm": 0.30827096104621887,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 0.6715,
+      "step": 69
+    },
+    {
+      "epoch": 0.015614543832255186,
+      "grad_norm": 0.3031168580055237,
+      "learning_rate": 1.38e-05,
+      "loss": 0.6504,
+      "step": 70
+    },
+    {
+      "epoch": 0.015837608744144548,
+      "grad_norm": 0.3063329756259918,
+      "learning_rate": 1.4e-05,
+      "loss": 0.6214,
+      "step": 71
+    },
+    {
+      "epoch": 0.016060673656033905,
+      "grad_norm": 0.2707573175430298,
+      "learning_rate": 1.4200000000000001e-05,
+      "loss": 0.6347,
+      "step": 72
+    },
+    {
+      "epoch": 0.016283738567923266,
+      "grad_norm": 0.3051629066467285,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 0.6318,
+      "step": 73
+    },
+    {
+      "epoch": 0.016506803479812624,
+      "grad_norm": 0.3260219991207123,
+      "learning_rate": 1.46e-05,
+      "loss": 0.6219,
+      "step": 74
+    },
+    {
+      "epoch": 0.016729868391701985,
+      "grad_norm": 0.2733941376209259,
+      "learning_rate": 1.48e-05,
+      "loss": 0.6384,
+      "step": 75
+    },
+    {
+      "epoch": 0.016952933303591346,
+      "grad_norm": 0.35662564635276794,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.622,
+      "step": 76
+    },
+    {
+      "epoch": 0.017175998215480704,
+      "grad_norm": 0.31423047184944153,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 0.6344,
+      "step": 77
+    },
+    {
+      "epoch": 0.017399063127370065,
+      "grad_norm": 0.3120841681957245,
+      "learning_rate": 1.54e-05,
+      "loss": 0.6566,
+      "step": 78
+    },
+    {
+      "epoch": 0.017622128039259423,
+      "grad_norm": 0.30338573455810547,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 0.6537,
+      "step": 79
+    },
+    {
+      "epoch": 0.017845192951148784,
+      "grad_norm": 0.3202762007713318,
+      "learning_rate": 1.58e-05,
+      "loss": 0.6705,
+      "step": 80
+    },
+    {
+      "epoch": 0.018068257863038145,
+      "grad_norm": 0.3163682520389557,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 0.6175,
+      "step": 81
+    },
+    {
+      "epoch": 0.018291322774927503,
+      "grad_norm": 0.2831353545188904,
+      "learning_rate": 1.62e-05,
+      "loss": 0.6301,
+      "step": 82
+    },
+    {
+      "epoch": 0.018514387686816864,
+      "grad_norm": 0.267448365688324,
+      "learning_rate": 1.64e-05,
+      "loss": 0.6224,
+      "step": 83
+    },
+    {
+      "epoch": 0.018737452598706225,
+      "grad_norm": 0.2999647259712219,
+      "learning_rate": 1.66e-05,
+      "loss": 0.6294,
+      "step": 84
+    },
+    {
+      "epoch": 0.018960517510595583,
+      "grad_norm": 0.27424749732017517,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 0.5998,
+      "step": 85
+    },
+    {
+      "epoch": 0.019183582422484944,
+      "grad_norm": 0.2582665979862213,
+      "learning_rate": 1.7e-05,
+      "loss": 0.6043,
+      "step": 86
+    },
+    {
+      "epoch": 0.0194066473343743,
+      "grad_norm": 0.2448103129863739,
+      "learning_rate": 1.72e-05,
+      "loss": 0.631,
+      "step": 87
+    },
+    {
+      "epoch": 0.019629712246263663,
+      "grad_norm": 0.25323718786239624,
+      "learning_rate": 1.7400000000000003e-05,
+      "loss": 0.6248,
+      "step": 88
+    },
+    {
+      "epoch": 0.019852777158153024,
+      "grad_norm": 0.32291799783706665,
+      "learning_rate": 1.76e-05,
+      "loss": 0.5797,
+      "step": 89
+    },
+    {
+      "epoch": 0.02007584207004238,
+      "grad_norm": 0.3439197242259979,
+      "learning_rate": 1.7800000000000002e-05,
+      "loss": 0.6141,
+      "step": 90
+    },
+    {
+      "epoch": 0.020298906981931743,
+      "grad_norm": 0.30653074383735657,
+      "learning_rate": 1.8e-05,
+      "loss": 0.6226,
+      "step": 91
+    },
+    {
+      "epoch": 0.0205219718938211,
+      "grad_norm": 0.5994735956192017,
+      "learning_rate": 1.8200000000000002e-05,
+      "loss": 0.614,
+      "step": 92
+    },
+    {
+      "epoch": 0.02074503680571046,
+      "grad_norm": 0.30260932445526123,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 0.6224,
+      "step": 93
+    },
+    {
+      "epoch": 0.020968101717599823,
+      "grad_norm": 0.2914034426212311,
+      "learning_rate": 1.86e-05,
+      "loss": 0.6201,
+      "step": 94
+    },
+    {
+      "epoch": 0.02119116662948918,
+      "grad_norm": 0.28529268503189087,
+      "learning_rate": 1.88e-05,
+      "loss": 0.6046,
+      "step": 95
+    },
+    {
+      "epoch": 0.02141423154137854,
+      "grad_norm": 0.26552823185920715,
+      "learning_rate": 1.9e-05,
+      "loss": 0.6058,
+      "step": 96
+    },
+    {
+      "epoch": 0.021637296453267903,
+      "grad_norm": 0.4168465733528137,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 0.6289,
+      "step": 97
+    },
+    {
+      "epoch": 0.02186036136515726,
+      "grad_norm": 0.40763556957244873,
+      "learning_rate": 1.94e-05,
+      "loss": 0.6113,
+      "step": 98
+    },
+    {
+      "epoch": 0.02208342627704662,
+      "grad_norm": 0.29686489701271057,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 0.6288,
+      "step": 99
+    },
+    {
+      "epoch": 0.02230649118893598,
+      "grad_norm": 0.27713605761528015,
+      "learning_rate": 1.98e-05,
+      "loss": 0.6404,
+      "step": 100
+    },
+    {
+      "epoch": 0.02252955610082534,
+      "grad_norm": 0.2593631148338318,
+      "learning_rate": 2e-05,
+      "loss": 0.5652,
+      "step": 101
+    },
+    {
+      "epoch": 0.0227526210127147,
+      "grad_norm": 0.2575121819972992,
+      "learning_rate": 1.999999972306855e-05,
+      "loss": 0.6062,
+      "step": 102
+    },
+    {
+      "epoch": 0.02297568592460406,
+      "grad_norm": 0.24929626286029816,
+      "learning_rate": 1.999999889227421e-05,
+      "loss": 0.6336,
+      "step": 103
+    },
+    {
+      "epoch": 0.02319875083649342,
+      "grad_norm": 0.2693648040294647,
+      "learning_rate": 1.9999997507617033e-05,
+      "loss": 0.6611,
+      "step": 104
+    },
+    {
+      "epoch": 0.023421815748382778,
+      "grad_norm": 0.5212802290916443,
+      "learning_rate": 1.9999995569097088e-05,
+      "loss": 0.6162,
+      "step": 105
+    },
+    {
+      "epoch": 0.02364488066027214,
+      "grad_norm": 0.25279495120048523,
+      "learning_rate": 1.999999307671449e-05,
+      "loss": 0.6062,
+      "step": 106
+    },
+    {
+      "epoch": 0.0238679455721615,
+      "grad_norm": 0.2529296278953552,
+      "learning_rate": 1.999999003046937e-05,
+      "loss": 0.6164,
+      "step": 107
+    },
+    {
+      "epoch": 0.024091010484050858,
+      "grad_norm": 0.3054715096950531,
+      "learning_rate": 1.9999986430361896e-05,
+      "loss": 0.5984,
+      "step": 108
+    },
+    {
+      "epoch": 0.02431407539594022,
+      "grad_norm": 0.26101136207580566,
+      "learning_rate": 1.9999982276392274e-05,
+      "loss": 0.5809,
+      "step": 109
+    },
+    {
+      "epoch": 0.024537140307829577,
+      "grad_norm": 0.26034069061279297,
+      "learning_rate": 1.9999977568560734e-05,
+      "loss": 0.6026,
+      "step": 110
+    },
+    {
+      "epoch": 0.024760205219718938,
+      "grad_norm": 0.2570051848888397,
+      "learning_rate": 1.999997230686753e-05,
+      "loss": 0.6299,
+      "step": 111
+    },
+    {
+      "epoch": 0.0249832701316083,
+      "grad_norm": 0.2621450424194336,
+      "learning_rate": 1.999996649131296e-05,
+      "loss": 0.6141,
+      "step": 112
+    },
+    {
+      "epoch": 0.025206335043497657,
+      "grad_norm": 0.2425510734319687,
+      "learning_rate": 1.999996012189734e-05,
+      "loss": 0.6184,
+      "step": 113
+    },
+    {
+      "epoch": 0.025429399955387018,
+      "grad_norm": 0.28203898668289185,
+      "learning_rate": 1.999995319862103e-05,
+      "loss": 0.5922,
+      "step": 114
+    },
+    {
+      "epoch": 0.02565246486727638,
+      "grad_norm": 0.2976488173007965,
+      "learning_rate": 1.9999945721484407e-05,
+      "loss": 0.586,
+      "step": 115
+    },
+    {
+      "epoch": 0.025875529779165737,
+      "grad_norm": 0.27894967794418335,
+      "learning_rate": 1.999993769048789e-05,
+      "loss": 0.6099,
+      "step": 116
+    },
+    {
+      "epoch": 0.026098594691055098,
+      "grad_norm": 0.2924571931362152,
+      "learning_rate": 1.999992910563192e-05,
+      "loss": 0.5646,
+      "step": 117
+    },
+    {
+      "epoch": 0.026321659602944456,
+      "grad_norm": 0.2446456104516983,
+      "learning_rate": 1.9999919966916976e-05,
+      "loss": 0.5722,
+      "step": 118
+    },
+    {
+      "epoch": 0.026544724514833817,
+      "grad_norm": 0.23295633494853973,
+      "learning_rate": 1.9999910274343562e-05,
+      "loss": 0.5806,
+      "step": 119
+    },
+    {
+      "epoch": 0.026767789426723178,
+      "grad_norm": 0.24767954647541046,
+      "learning_rate": 1.999990002791221e-05,
+      "loss": 0.6003,
+      "step": 120
+    },
+    {
+      "epoch": 0.026990854338612535,
+      "grad_norm": 0.25390875339508057,
+      "learning_rate": 1.99998892276235e-05,
+      "loss": 0.5798,
+      "step": 121
+    },
+    {
+      "epoch": 0.027213919250501897,
+      "grad_norm": 0.22989511489868164,
+      "learning_rate": 1.999987787347802e-05,
+      "loss": 0.5873,
+      "step": 122
+    },
+    {
+      "epoch": 0.027436984162391254,
+      "grad_norm": 0.4493498206138611,
+      "learning_rate": 1.99998659654764e-05,
+      "loss": 0.5994,
+      "step": 123
+    },
+    {
+      "epoch": 0.027660049074280615,
+      "grad_norm": 0.2416534572839737,
+      "learning_rate": 1.99998535036193e-05,
+      "loss": 0.5787,
+      "step": 124
+    },
+    {
+      "epoch": 0.027883113986169977,
+      "grad_norm": 0.23413562774658203,
+      "learning_rate": 1.9999840487907414e-05,
+      "loss": 0.608,
+      "step": 125
+    },
+    {
+      "epoch": 0.028106178898059334,
+      "grad_norm": 0.28015846014022827,
+      "learning_rate": 1.9999826918341462e-05,
+      "loss": 0.6034,
+      "step": 126
+    },
+    {
+      "epoch": 0.028329243809948695,
+      "grad_norm": 0.2494441717863083,
+      "learning_rate": 1.999981279492219e-05,
+      "loss": 0.6256,
+      "step": 127
+    },
+    {
+      "epoch": 0.028552308721838057,
+      "grad_norm": 0.3094983398914337,
+      "learning_rate": 1.9999798117650385e-05,
+      "loss": 0.6351,
+      "step": 128
+    },
+    {
+      "epoch": 0.028775373633727414,
+      "grad_norm": 0.24032524228096008,
+      "learning_rate": 1.9999782886526863e-05,
+      "loss": 0.5703,
+      "step": 129
+    },
+    {
+      "epoch": 0.028998438545616775,
+      "grad_norm": 0.24950377643108368,
+      "learning_rate": 1.9999767101552458e-05,
+      "loss": 0.5952,
+      "step": 130
+    },
+    {
+      "epoch": 0.029221503457506133,
+      "grad_norm": 0.24877335131168365,
+      "learning_rate": 1.999975076272805e-05,
+      "loss": 0.6143,
+      "step": 131
+    },
+    {
+      "epoch": 0.029444568369395494,
+      "grad_norm": 0.2579665184020996,
+      "learning_rate": 1.999973387005455e-05,
+      "loss": 0.6063,
+      "step": 132
+    },
+    {
+      "epoch": 0.029667633281284855,
+      "grad_norm": 0.2793569564819336,
+      "learning_rate": 1.9999716423532884e-05,
+      "loss": 0.613,
+      "step": 133
+    },
+    {
+      "epoch": 0.029890698193174213,
+      "grad_norm": 0.23061081767082214,
+      "learning_rate": 1.999969842316402e-05,
+      "loss": 0.5918,
+      "step": 134
+    },
+    {
+      "epoch": 0.030113763105063574,
+      "grad_norm": 0.2353406399488449,
+      "learning_rate": 1.999967986894896e-05,
+      "loss": 0.585,
+      "step": 135
+    },
+    {
+      "epoch": 0.030336828016952932,
+      "grad_norm": 0.2254764884710312,
+      "learning_rate": 1.9999660760888722e-05,
+      "loss": 0.611,
+      "step": 136
+    },
+    {
+      "epoch": 0.030559892928842293,
+      "grad_norm": 0.24640028178691864,
+      "learning_rate": 1.9999641098984378e-05,
+      "loss": 0.5911,
+      "step": 137
+    },
+    {
+      "epoch": 0.030782957840731654,
+      "grad_norm": 0.22932305932044983,
+      "learning_rate": 1.9999620883237004e-05,
+      "loss": 0.5999,
+      "step": 138
+    },
+    {
+      "epoch": 0.031006022752621012,
+      "grad_norm": 0.2379560023546219,
+      "learning_rate": 1.999960011364773e-05,
+      "loss": 0.5987,
+      "step": 139
+    },
+    {
+      "epoch": 0.031229087664510373,
+      "grad_norm": 0.22063873708248138,
+      "learning_rate": 1.99995787902177e-05,
+      "loss": 0.5867,
+      "step": 140
+    },
+    {
+      "epoch": 0.03145215257639973,
+      "grad_norm": 0.23651117086410522,
+      "learning_rate": 1.9999556912948096e-05,
+      "loss": 0.5809,
+      "step": 141
+    },
+    {
+      "epoch": 0.031675217488289095,
+      "grad_norm": 0.2621421813964844,
+      "learning_rate": 1.9999534481840134e-05,
+      "loss": 0.5935,
+      "step": 142
+    },
+    {
+      "epoch": 0.03189828240017845,
+      "grad_norm": 0.24506062269210815,
+      "learning_rate": 1.9999511496895047e-05,
+      "loss": 0.5931,
+      "step": 143
+    },
+    {
+      "epoch": 0.03212134731206781,
+      "grad_norm": 0.2224111109972,
+      "learning_rate": 1.999948795811412e-05,
+      "loss": 0.6156,
+      "step": 144
+    },
+    {
+      "epoch": 0.03234441222395717,
+      "grad_norm": 0.23910044133663177,
+      "learning_rate": 1.9999463865498644e-05,
+      "loss": 0.5839,
+      "step": 145
+    },
+    {
+      "epoch": 0.03256747713584653,
+      "grad_norm": 0.2555723786354065,
+      "learning_rate": 1.9999439219049964e-05,
+      "loss": 0.6474,
+      "step": 146
+    },
+    {
+      "epoch": 0.03279054204773589,
+      "grad_norm": 0.25059348344802856,
+      "learning_rate": 1.9999414018769442e-05,
+      "loss": 0.5846,
+      "step": 147
+    },
+    {
+      "epoch": 0.03301360695962525,
+      "grad_norm": 0.22700609266757965,
+      "learning_rate": 1.9999388264658467e-05,
+      "loss": 0.5867,
+      "step": 148
+    },
+    {
+      "epoch": 0.03323667187151461,
+      "grad_norm": 0.25858354568481445,
+      "learning_rate": 1.9999361956718476e-05,
+      "loss": 0.5984,
+      "step": 149
+    },
+    {
+      "epoch": 0.03345973678340397,
+      "grad_norm": 0.2813456356525421,
+      "learning_rate": 1.9999335094950922e-05,
+      "loss": 0.5943,
+      "step": 150
+    },
+    {
+      "epoch": 0.03368280169529333,
+      "grad_norm": 0.288655549287796,
+      "learning_rate": 1.9999307679357293e-05,
+      "loss": 0.5975,
+      "step": 151
+    },
+    {
+      "epoch": 0.03390586660718269,
+      "grad_norm": 0.23721511662006378,
+      "learning_rate": 1.9999279709939102e-05,
+      "loss": 0.5637,
+      "step": 152
+    },
+    {
+      "epoch": 0.03412893151907205,
+      "grad_norm": 0.23953364789485931,
+      "learning_rate": 1.999925118669791e-05,
+      "loss": 0.5693,
+      "step": 153
+    },
+    {
+      "epoch": 0.03435199643096141,
+      "grad_norm": 0.23194071650505066,
+      "learning_rate": 1.9999222109635283e-05,
+      "loss": 0.5446,
+      "step": 154
+    },
+    {
+      "epoch": 0.03457506134285077,
+      "grad_norm": 0.2160234898328781,
+      "learning_rate": 1.999919247875284e-05,
+      "loss": 0.5812,
+      "step": 155
+    },
+    {
+      "epoch": 0.03479812625474013,
+      "grad_norm": 0.22934173047542572,
+      "learning_rate": 1.999916229405222e-05,
+      "loss": 0.6149,
+      "step": 156
+    },
+    {
+      "epoch": 0.03502119116662949,
+      "grad_norm": 0.24892570078372955,
+      "learning_rate": 1.999913155553509e-05,
+      "loss": 0.6053,
+      "step": 157
+    },
+    {
+      "epoch": 0.035244256078518846,
+      "grad_norm": 0.2235739678144455,
+      "learning_rate": 1.9999100263203165e-05,
+      "loss": 0.5799,
+      "step": 158
+    },
+    {
+      "epoch": 0.03546732099040821,
+      "grad_norm": 0.21852736175060272,
+      "learning_rate": 1.9999068417058168e-05,
+      "loss": 0.5793,
+      "step": 159
+    },
+    {
+      "epoch": 0.03569038590229757,
+      "grad_norm": 0.269131600856781,
+      "learning_rate": 1.9999036017101864e-05,
+      "loss": 0.5817,
+      "step": 160
+    },
+    {
+      "epoch": 0.035913450814186926,
+      "grad_norm": 0.22516575455665588,
+      "learning_rate": 1.999900306333605e-05,
+      "loss": 0.6,
+      "step": 161
+    },
+    {
+      "epoch": 0.03613651572607629,
+      "grad_norm": 0.24313850700855255,
+      "learning_rate": 1.999896955576255e-05,
+      "loss": 0.5738,
+      "step": 162
+    },
+    {
+      "epoch": 0.03635958063796565,
+      "grad_norm": 0.26203158497810364,
+      "learning_rate": 1.999893549438322e-05,
+      "loss": 0.5687,
+      "step": 163
+    },
+    {
+      "epoch": 0.036582645549855006,
+      "grad_norm": 0.23275348544120789,
+      "learning_rate": 1.9998900879199948e-05,
+      "loss": 0.5698,
+      "step": 164
+    },
+    {
+      "epoch": 0.03680571046174437,
+      "grad_norm": 0.2560306489467621,
+      "learning_rate": 1.9998865710214646e-05,
+      "loss": 0.5877,
+      "step": 165
+    },
+    {
+      "epoch": 0.03702877537363373,
+      "grad_norm": 0.2175760716199875,
+      "learning_rate": 1.999882998742927e-05,
+      "loss": 0.5797,
+      "step": 166
+    },
+    {
+      "epoch": 0.037251840285523086,
+      "grad_norm": 0.22381433844566345,
+      "learning_rate": 1.999879371084579e-05,
+      "loss": 0.5642,
+      "step": 167
+    },
+    {
+      "epoch": 0.03747490519741245,
+      "grad_norm": 0.2363629788160324,
+      "learning_rate": 1.9998756880466224e-05,
+      "loss": 0.5651,
+      "step": 168
+    },
+    {
+      "epoch": 0.03769797010930181,
+      "grad_norm": 0.22718748450279236,
+      "learning_rate": 1.9998719496292603e-05,
+      "loss": 0.5867,
+      "step": 169
+    },
+    {
+      "epoch": 0.037921035021191166,
+      "grad_norm": 0.23127657175064087,
+      "learning_rate": 1.9998681558327005e-05,
+      "loss": 0.5724,
+      "step": 170
+    },
+    {
+      "epoch": 0.03814409993308052,
+      "grad_norm": 0.23175480961799622,
+      "learning_rate": 1.9998643066571527e-05,
+      "loss": 0.5919,
+      "step": 171
+    },
+    {
+      "epoch": 0.03836716484496989,
+      "grad_norm": 0.22871741652488708,
+      "learning_rate": 1.9998604021028304e-05,
+      "loss": 0.5653,
+      "step": 172
+    },
+    {
+      "epoch": 0.038590229756859246,
+      "grad_norm": 0.21524189412593842,
+      "learning_rate": 1.999856442169949e-05,
+      "loss": 0.6123,
+      "step": 173
+    },
+    {
+      "epoch": 0.0388132946687486,
+      "grad_norm": 0.23523908853530884,
+      "learning_rate": 1.999852426858729e-05,
+      "loss": 0.5931,
+      "step": 174
+    },
+    {
+      "epoch": 0.03903635958063797,
+      "grad_norm": 0.23540736734867096,
+      "learning_rate": 1.9998483561693926e-05,
+      "loss": 0.5614,
+      "step": 175
+    },
+    {
+      "epoch": 0.039259424492527326,
+      "grad_norm": 0.22590956091880798,
+      "learning_rate": 1.999844230102164e-05,
+      "loss": 0.5917,
+      "step": 176
+    },
+    {
+      "epoch": 0.03948248940441668,
+      "grad_norm": 0.22635167837142944,
+      "learning_rate": 1.999840048657273e-05,
+      "loss": 0.5886,
+      "step": 177
+    },
+    {
+      "epoch": 0.03970555431630605,
+      "grad_norm": 0.24712443351745605,
+      "learning_rate": 1.9998358118349513e-05,
+      "loss": 0.5936,
+      "step": 178
+    },
+    {
+      "epoch": 0.039928619228195406,
+      "grad_norm": 0.266133576631546,
+      "learning_rate": 1.999831519635433e-05,
+      "loss": 0.5662,
+      "step": 179
+    },
+    {
+      "epoch": 0.04015168414008476,
+      "grad_norm": 0.2172316312789917,
+      "learning_rate": 1.9998271720589558e-05,
+      "loss": 0.5679,
+      "step": 180
+    },
+    {
+      "epoch": 0.04037474905197413,
+      "grad_norm": 0.3194250464439392,
+      "learning_rate": 1.999822769105761e-05,
+      "loss": 0.6105,
+      "step": 181
+    },
+    {
+      "epoch": 0.040597813963863486,
+      "grad_norm": 0.2604449391365051,
+      "learning_rate": 1.9998183107760915e-05,
+      "loss": 0.5942,
+      "step": 182
+    },
+    {
+      "epoch": 0.04082087887575284,
+      "grad_norm": 0.26620209217071533,
+      "learning_rate": 1.9998137970701955e-05,
+      "loss": 0.5859,
+      "step": 183
+    },
+    {
+      "epoch": 0.0410439437876422,
+      "grad_norm": 0.2345753312110901,
+      "learning_rate": 1.9998092279883215e-05,
+      "loss": 0.5939,
+      "step": 184
+    },
+    {
+      "epoch": 0.041267008699531565,
+      "grad_norm": 0.22213764488697052,
+      "learning_rate": 1.999804603530724e-05,
+      "loss": 0.5815,
+      "step": 185
+    },
+    {
+      "epoch": 0.04149007361142092,
+      "grad_norm": 0.23276378214359283,
+      "learning_rate": 1.9997999236976587e-05,
+      "loss": 0.5842,
+      "step": 186
+    },
+    {
+      "epoch": 0.04171313852331028,
+      "grad_norm": 0.22480995953083038,
+      "learning_rate": 1.9997951884893843e-05,
+      "loss": 0.5897,
+      "step": 187
+    },
+    {
+      "epoch": 0.041936203435199645,
+      "grad_norm": 0.22761783003807068,
+      "learning_rate": 1.9997903979061635e-05,
+      "loss": 0.5873,
+      "step": 188
+    },
+    {
+      "epoch": 0.042159268347089,
+      "grad_norm": 0.27698761224746704,
+      "learning_rate": 1.9997855519482614e-05,
+      "loss": 0.5933,
+      "step": 189
+    },
+    {
+      "epoch": 0.04238233325897836,
+      "grad_norm": 0.2002975046634674,
+      "learning_rate": 1.9997806506159466e-05,
+      "loss": 0.5698,
+      "step": 190
+    },
+    {
+      "epoch": 0.042605398170867725,
+      "grad_norm": 0.233673095703125,
+      "learning_rate": 1.9997756939094905e-05,
+      "loss": 0.5765,
+      "step": 191
+    },
+    {
+      "epoch": 0.04282846308275708,
+      "grad_norm": 0.26376616954803467,
+      "learning_rate": 1.999770681829168e-05,
+      "loss": 0.5741,
+      "step": 192
+    },
+    {
+      "epoch": 0.04305152799464644,
+      "grad_norm": 0.22886492311954498,
+      "learning_rate": 1.9997656143752556e-05,
+      "loss": 0.5709,
+      "step": 193
+    },
+    {
+      "epoch": 0.043274592906535805,
+      "grad_norm": 0.3004070520401001,
+      "learning_rate": 1.9997604915480352e-05,
+      "loss": 0.5514,
+      "step": 194
+    },
+    {
+      "epoch": 0.04349765781842516,
+      "grad_norm": 0.22209997475147247,
+      "learning_rate": 1.99975531334779e-05,
+      "loss": 0.6095,
+      "step": 195
+    },
+    {
+      "epoch": 0.04372072273031452,
+      "grad_norm": 0.20531293749809265,
+      "learning_rate": 1.9997500797748067e-05,
+      "loss": 0.5618,
+      "step": 196
+    },
+    {
+      "epoch": 0.04394378764220388,
+      "grad_norm": 0.20507480204105377,
+      "learning_rate": 1.9997447908293753e-05,
+      "loss": 0.5465,
+      "step": 197
+    },
+    {
+      "epoch": 0.04416685255409324,
+      "grad_norm": 0.22681620717048645,
+      "learning_rate": 1.999739446511789e-05,
+      "loss": 0.5546,
+      "step": 198
+    },
+    {
+      "epoch": 0.0443899174659826,
+      "grad_norm": 0.21772046387195587,
+      "learning_rate": 1.999734046822343e-05,
+      "loss": 0.5934,
+      "step": 199
+    },
+    {
+      "epoch": 0.04461298237787196,
+      "grad_norm": 0.22321215271949768,
+      "learning_rate": 1.9997285917613375e-05,
+      "loss": 0.579,
+      "step": 200
+    },
+    {
+      "epoch": 0.04483604728976132,
+      "grad_norm": 0.21166378259658813,
+      "learning_rate": 1.9997230813290737e-05,
+      "loss": 0.5978,
+      "step": 201
+    },
+    {
+      "epoch": 0.04505911220165068,
+      "grad_norm": 0.23722383379936218,
+      "learning_rate": 1.999717515525857e-05,
+      "loss": 0.5747,
+      "step": 202
+    },
+    {
+      "epoch": 0.04528217711354004,
+      "grad_norm": 0.36909613013267517,
+      "learning_rate": 1.9997118943519962e-05,
+      "loss": 0.5417,
+      "step": 203
+    },
+    {
+      "epoch": 0.0455052420254294,
+      "grad_norm": 0.23573465645313263,
+      "learning_rate": 1.9997062178078023e-05,
+      "loss": 0.5971,
+      "step": 204
+    },
+    {
+      "epoch": 0.04572830693731876,
+      "grad_norm": 0.25214946269989014,
+      "learning_rate": 1.9997004858935894e-05,
+      "loss": 0.5852,
+      "step": 205
+    },
+    {
+      "epoch": 0.04595137184920812,
+      "grad_norm": 0.20637395977973938,
+      "learning_rate": 1.9996946986096754e-05,
+      "loss": 0.5394,
+      "step": 206
+    },
+    {
+      "epoch": 0.046174436761097476,
+      "grad_norm": 0.21749068796634674,
+      "learning_rate": 1.9996888559563804e-05,
+      "loss": 0.5662,
+      "step": 207
+    },
+    {
+      "epoch": 0.04639750167298684,
+      "grad_norm": 0.2306276559829712,
+      "learning_rate": 1.9996829579340284e-05,
+      "loss": 0.6038,
+      "step": 208
+    },
+    {
+      "epoch": 0.0466205665848762,
+      "grad_norm": 0.22346408665180206,
+      "learning_rate": 1.999677004542946e-05,
+      "loss": 0.6024,
+      "step": 209
+    },
+    {
+      "epoch": 0.046843631496765556,
+      "grad_norm": 0.22103498876094818,
+      "learning_rate": 1.9996709957834627e-05,
+      "loss": 0.5805,
+      "step": 210
+    },
+    {
+      "epoch": 0.04706669640865492,
+      "grad_norm": 0.2120116651058197,
+      "learning_rate": 1.9996649316559118e-05,
+      "loss": 0.5712,
+      "step": 211
+    },
+    {
+      "epoch": 0.04728976132054428,
+      "grad_norm": 0.24006275832653046,
+      "learning_rate": 1.9996588121606286e-05,
+      "loss": 0.5487,
+      "step": 212
+    },
+    {
+      "epoch": 0.047512826232433636,
+      "grad_norm": 0.2095099836587906,
+      "learning_rate": 1.9996526372979522e-05,
+      "loss": 0.555,
+      "step": 213
+    },
+    {
+      "epoch": 0.047735891144323,
+      "grad_norm": 0.23283496499061584,
+      "learning_rate": 1.999646407068225e-05,
+      "loss": 0.6025,
+      "step": 214
+    },
+    {
+      "epoch": 0.04795895605621236,
+      "grad_norm": 0.2015821784734726,
+      "learning_rate": 1.9996401214717912e-05,
+      "loss": 0.5725,
+      "step": 215
+    },
+    {
+      "epoch": 0.048182020968101716,
+      "grad_norm": 0.2395429164171219,
+      "learning_rate": 1.999633780509e-05,
+      "loss": 0.5549,
+      "step": 216
+    },
+    {
+      "epoch": 0.04840508587999108,
+      "grad_norm": 0.2157035768032074,
+      "learning_rate": 1.9996273841802017e-05,
+      "loss": 0.5644,
+      "step": 217
+    },
+    {
+      "epoch": 0.04862815079188044,
+      "grad_norm": 0.2156548947095871,
+      "learning_rate": 1.9996209324857516e-05,
+      "loss": 0.5901,
+      "step": 218
+    },
+    {
+      "epoch": 0.048851215703769796,
+      "grad_norm": 0.2213698923587799,
+      "learning_rate": 1.999614425426006e-05,
+      "loss": 0.5463,
+      "step": 219
+    },
+    {
+      "epoch": 0.049074280615659154,
+      "grad_norm": 0.21965420246124268,
+      "learning_rate": 1.9996078630013253e-05,
+      "loss": 0.5741,
+      "step": 220
+    },
+    {
+      "epoch": 0.04929734552754852,
+      "grad_norm": 0.213240385055542,
+      "learning_rate": 1.999601245212074e-05,
+      "loss": 0.5698,
+      "step": 221
+    },
+    {
+      "epoch": 0.049520410439437876,
+      "grad_norm": 0.21078208088874817,
+      "learning_rate": 1.9995945720586177e-05,
+      "loss": 0.5923,
+      "step": 222
+    },
+    {
+      "epoch": 0.04974347535132723,
+      "grad_norm": 0.21629688143730164,
+      "learning_rate": 1.9995878435413264e-05,
+      "loss": 0.5438,
+      "step": 223
+    },
+    {
+      "epoch": 0.0499665402632166,
+      "grad_norm": 0.22872765362262726,
+      "learning_rate": 1.9995810596605725e-05,
+      "loss": 0.5543,
+      "step": 224
+    },
+    {
+      "epoch": 0.050189605175105956,
+      "grad_norm": 0.2012244164943695,
+      "learning_rate": 1.999574220416732e-05,
+      "loss": 0.557,
+      "step": 225
+    },
+    {
+      "epoch": 0.05041267008699531,
+      "grad_norm": 0.21912412345409393,
+      "learning_rate": 1.9995673258101837e-05,
+      "loss": 0.5816,
+      "step": 226
+    },
+    {
+      "epoch": 0.05063573499888468,
+      "grad_norm": 0.22358618676662445,
+      "learning_rate": 1.999560375841309e-05,
+      "loss": 0.6222,
+      "step": 227
+    },
+    {
+      "epoch": 0.050858799910774036,
+      "grad_norm": 0.22544872760772705,
+      "learning_rate": 1.9995533705104936e-05,
+      "loss": 0.5481,
+      "step": 228
+    },
+    {
+      "epoch": 0.05108186482266339,
+      "grad_norm": 0.20495697855949402,
+      "learning_rate": 1.999546309818125e-05,
+      "loss": 0.5371,
+      "step": 229
+    },
+    {
+      "epoch": 0.05130492973455276,
+      "grad_norm": 0.22267979383468628,
+      "learning_rate": 1.9995391937645944e-05,
+      "loss": 0.5916,
+      "step": 230
+    },
+    {
+      "epoch": 0.051527994646442116,
+      "grad_norm": 0.200364887714386,
+      "learning_rate": 1.9995320223502958e-05,
+      "loss": 0.541,
+      "step": 231
+    },
+    {
+      "epoch": 0.05175105955833147,
+      "grad_norm": 0.20090313255786896,
+      "learning_rate": 1.9995247955756267e-05,
+      "loss": 0.565,
+      "step": 232
+    },
+    {
+      "epoch": 0.05197412447022083,
+      "grad_norm": 5.10443639755249,
+      "learning_rate": 1.9995175134409868e-05,
+      "loss": 0.6056,
+      "step": 233
+    },
+    {
+      "epoch": 0.052197189382110196,
+      "grad_norm": 0.30669254064559937,
+      "learning_rate": 1.99951017594678e-05,
+      "loss": 0.589,
+      "step": 234
+    },
+    {
+      "epoch": 0.05242025429399955,
+      "grad_norm": 0.21791155636310577,
+      "learning_rate": 1.9995027830934125e-05,
+      "loss": 0.5438,
+      "step": 235
+    },
+    {
+      "epoch": 0.05264331920588891,
+      "grad_norm": 0.25966677069664,
+      "learning_rate": 1.9994953348812937e-05,
+      "loss": 0.582,
+      "step": 236
+    },
+    {
+      "epoch": 0.052866384117778276,
+      "grad_norm": 0.24935361742973328,
+      "learning_rate": 1.9994878313108362e-05,
+      "loss": 0.5934,
+      "step": 237
+    },
+    {
+      "epoch": 0.05308944902966763,
+      "grad_norm": 0.22276830673217773,
+      "learning_rate": 1.9994802723824557e-05,
+      "loss": 0.5845,
+      "step": 238
+    },
+    {
+      "epoch": 0.05331251394155699,
+      "grad_norm": 0.2349478304386139,
+      "learning_rate": 1.9994726580965704e-05,
+      "loss": 0.5716,
+      "step": 239
+    },
+    {
+      "epoch": 0.053535578853446356,
+      "grad_norm": 0.22332008183002472,
+      "learning_rate": 1.9994649884536026e-05,
+      "loss": 0.6009,
+      "step": 240
+    },
+    {
+      "epoch": 0.05375864376533571,
+      "grad_norm": 0.19660501182079315,
+      "learning_rate": 1.9994572634539767e-05,
+      "loss": 0.5574,
+      "step": 241
+    },
+    {
+      "epoch": 0.05398170867722507,
+      "grad_norm": 0.2085440754890442,
+      "learning_rate": 1.999449483098121e-05,
+      "loss": 0.5676,
+      "step": 242
+    },
+    {
+      "epoch": 0.054204773589114436,
+      "grad_norm": 0.21077819168567657,
+      "learning_rate": 1.999441647386466e-05,
+      "loss": 0.5507,
+      "step": 243
+    },
+    {
+      "epoch": 0.05442783850100379,
+      "grad_norm": 0.2927658259868622,
+      "learning_rate": 1.9994337563194457e-05,
+      "loss": 0.5834,
+      "step": 244
+    },
+    {
+      "epoch": 0.05465090341289315,
+      "grad_norm": 0.22845080494880676,
+      "learning_rate": 1.9994258098974974e-05,
+      "loss": 0.5524,
+      "step": 245
+    },
+    {
+      "epoch": 0.05487396832478251,
+      "grad_norm": 0.24501702189445496,
+      "learning_rate": 1.999417808121061e-05,
+      "loss": 0.5376,
+      "step": 246
+    },
+    {
+      "epoch": 0.05509703323667187,
+      "grad_norm": 0.19732032716274261,
+      "learning_rate": 1.99940975099058e-05,
+      "loss": 0.5697,
+      "step": 247
+    },
+    {
+      "epoch": 0.05532009814856123,
+      "grad_norm": 0.21217837929725647,
+      "learning_rate": 1.9994016385065005e-05,
+      "loss": 0.5941,
+      "step": 248
+    },
+    {
+      "epoch": 0.05554316306045059,
+      "grad_norm": 0.25333619117736816,
+      "learning_rate": 1.999393470669272e-05,
+      "loss": 0.5482,
+      "step": 249
+    },
+    {
+      "epoch": 0.05576622797233995,
+      "grad_norm": 0.21620452404022217,
+      "learning_rate": 1.9993852474793457e-05,
+      "loss": 0.5712,
+      "step": 250
+    },
+    {
+      "epoch": 0.05598929288422931,
+      "grad_norm": 0.20523187518119812,
+      "learning_rate": 1.9993769689371788e-05,
+      "loss": 0.5676,
+      "step": 251
+    },
+    {
+      "epoch": 0.05621235779611867,
+      "grad_norm": 0.20364375412464142,
+      "learning_rate": 1.999368635043229e-05,
+      "loss": 0.5764,
+      "step": 252
+    },
+    {
+      "epoch": 0.05643542270800803,
+      "grad_norm": 0.2179926186800003,
+      "learning_rate": 1.9993602457979574e-05,
+      "loss": 0.5583,
+      "step": 253
+    },
+    {
+      "epoch": 0.05665848761989739,
+      "grad_norm": 0.23844636976718903,
+      "learning_rate": 1.9993518012018297e-05,
+      "loss": 0.5768,
+      "step": 254
+    },
+    {
+      "epoch": 0.05688155253178675,
+      "grad_norm": 0.21450263261795044,
+      "learning_rate": 1.9993433012553128e-05,
+      "loss": 0.5814,
+      "step": 255
+    },
+    {
+      "epoch": 0.05710461744367611,
+      "grad_norm": 0.22201555967330933,
+      "learning_rate": 1.9993347459588777e-05,
+      "loss": 0.5565,
+      "step": 256
+    },
+    {
+      "epoch": 0.05732768235556547,
+      "grad_norm": 0.2182476669549942,
+      "learning_rate": 1.9993261353129988e-05,
+      "loss": 0.5818,
+      "step": 257
+    },
+    {
+      "epoch": 0.05755074726745483,
+      "grad_norm": 0.1971082240343094,
+      "learning_rate": 1.9993174693181517e-05,
+      "loss": 0.5686,
+      "step": 258
+    },
+    {
+      "epoch": 0.057773812179344186,
+      "grad_norm": 0.20535312592983246,
+      "learning_rate": 1.999308747974818e-05,
+      "loss": 0.5687,
+      "step": 259
+    },
+    {
+      "epoch": 0.05799687709123355,
+      "grad_norm": 0.218933567404747,
+      "learning_rate": 1.9992999712834794e-05,
+      "loss": 0.5769,
+      "step": 260
+    },
+    {
+      "epoch": 0.05821994200312291,
+      "grad_norm": 0.2132030874490738,
+      "learning_rate": 1.9992911392446227e-05,
+      "loss": 0.596,
+      "step": 261
+    },
+    {
+      "epoch": 0.058443006915012266,
+      "grad_norm": 0.19926026463508606,
+      "learning_rate": 1.999282251858737e-05,
+      "loss": 0.558,
+      "step": 262
+    },
+    {
+      "epoch": 0.05866607182690163,
+      "grad_norm": 0.2089562714099884,
+      "learning_rate": 1.9992733091263144e-05,
+      "loss": 0.582,
+      "step": 263
+    },
+    {
+      "epoch": 0.05888913673879099,
+      "grad_norm": 0.20722773671150208,
+      "learning_rate": 1.9992643110478504e-05,
+      "loss": 0.5758,
+      "step": 264
+    },
+    {
+      "epoch": 0.059112201650680346,
+      "grad_norm": 0.21631532907485962,
+      "learning_rate": 1.999255257623843e-05,
+      "loss": 0.5725,
+      "step": 265
+    },
+    {
+      "epoch": 0.05933526656256971,
+      "grad_norm": 0.21997545659542084,
+      "learning_rate": 1.999246148854794e-05,
+      "loss": 0.5915,
+      "step": 266
+    },
+    {
+      "epoch": 0.05955833147445907,
+      "grad_norm": 0.4292175769805908,
+      "learning_rate": 1.9992369847412076e-05,
+      "loss": 0.5891,
+      "step": 267
+    },
+    {
+      "epoch": 0.059781396386348426,
+      "grad_norm": 0.2055254876613617,
+      "learning_rate": 1.9992277652835918e-05,
+      "loss": 0.5219,
+      "step": 268
+    },
+    {
+      "epoch": 0.06000446129823779,
+      "grad_norm": 0.20326492190361023,
+      "learning_rate": 1.9992184904824566e-05,
+      "loss": 0.5462,
+      "step": 269
+    },
+    {
+      "epoch": 0.06022752621012715,
+      "grad_norm": 0.19965752959251404,
+      "learning_rate": 1.9992091603383164e-05,
+      "loss": 0.5688,
+      "step": 270
+    },
+    {
+      "epoch": 0.060450591122016506,
+      "grad_norm": 0.19592702388763428,
+      "learning_rate": 1.9991997748516872e-05,
+      "loss": 0.5556,
+      "step": 271
+    },
+    {
+      "epoch": 0.060673656033905864,
+      "grad_norm": 0.2036285698413849,
+      "learning_rate": 1.9991903340230898e-05,
+      "loss": 0.5452,
+      "step": 272
+    },
+    {
+      "epoch": 0.06089672094579523,
+      "grad_norm": 0.21509170532226562,
+      "learning_rate": 1.9991808378530465e-05,
+      "loss": 0.5682,
+      "step": 273
+    },
+    {
+      "epoch": 0.061119785857684586,
+      "grad_norm": 0.2015625238418579,
+      "learning_rate": 1.9991712863420832e-05,
+      "loss": 0.5846,
+      "step": 274
+    },
+    {
+      "epoch": 0.061342850769573944,
+      "grad_norm": 0.21707095205783844,
+      "learning_rate": 1.9991616794907286e-05,
+      "loss": 0.5666,
+      "step": 275
+    },
+    {
+      "epoch": 0.06156591568146331,
+      "grad_norm": 0.1988460123538971,
+      "learning_rate": 1.9991520172995158e-05,
+      "loss": 0.5875,
+      "step": 276
+    },
+    {
+      "epoch": 0.061788980593352666,
+      "grad_norm": 0.19290049374103546,
+      "learning_rate": 1.999142299768979e-05,
+      "loss": 0.5497,
+      "step": 277
+    },
+    {
+      "epoch": 0.062012045505242024,
+      "grad_norm": 0.2069414258003235,
+      "learning_rate": 1.9991325268996567e-05,
+      "loss": 0.5438,
+      "step": 278
+    },
+    {
+      "epoch": 0.06223511041713139,
+      "grad_norm": 0.22346992790699005,
+      "learning_rate": 1.9991226986920906e-05,
+      "loss": 0.5668,
+      "step": 279
+    },
+    {
+      "epoch": 0.062458175329020746,
+      "grad_norm": 0.21194760501384735,
+      "learning_rate": 1.9991128151468247e-05,
+      "loss": 0.5957,
+      "step": 280
+    },
+    {
+      "epoch": 0.06268124024091011,
+      "grad_norm": 0.209646537899971,
+      "learning_rate": 1.9991028762644063e-05,
+      "loss": 0.5748,
+      "step": 281
+    },
+    {
+      "epoch": 0.06290430515279946,
+      "grad_norm": 0.20566052198410034,
+      "learning_rate": 1.9990928820453858e-05,
+      "loss": 0.5506,
+      "step": 282
+    },
+    {
+      "epoch": 0.06312737006468883,
+      "grad_norm": 0.21374207735061646,
+      "learning_rate": 1.999082832490317e-05,
+      "loss": 0.5697,
+      "step": 283
+    },
+    {
+      "epoch": 0.06335043497657819,
+      "grad_norm": 0.21012777090072632,
+      "learning_rate": 1.999072727599757e-05,
+      "loss": 0.5112,
+      "step": 284
+    },
+    {
+      "epoch": 0.06357349988846754,
+      "grad_norm": 0.19954928755760193,
+      "learning_rate": 1.9990625673742644e-05,
+      "loss": 0.579,
+      "step": 285
+    },
+    {
+      "epoch": 0.0637965648003569,
+      "grad_norm": 0.20394836366176605,
+      "learning_rate": 1.9990523518144027e-05,
+      "loss": 0.5425,
+      "step": 286
+    },
+    {
+      "epoch": 0.06401962971224627,
+      "grad_norm": 0.20405294001102448,
+      "learning_rate": 1.9990420809207375e-05,
+      "loss": 0.5538,
+      "step": 287
+    },
+    {
+      "epoch": 0.06424269462413562,
+      "grad_norm": 0.24065051972866058,
+      "learning_rate": 1.9990317546938373e-05,
+      "loss": 0.5786,
+      "step": 288
+    },
+    {
+      "epoch": 0.06446575953602499,
+      "grad_norm": 0.20158305764198303,
+      "learning_rate": 1.9990213731342747e-05,
+      "loss": 0.5475,
+      "step": 289
+    },
+    {
+      "epoch": 0.06468882444791434,
+      "grad_norm": 0.29271385073661804,
+      "learning_rate": 1.9990109362426243e-05,
+      "loss": 0.5468,
+      "step": 290
+    },
+    {
+      "epoch": 0.0649118893598037,
+      "grad_norm": 0.2596181333065033,
+      "learning_rate": 1.999000444019464e-05,
+      "loss": 0.5632,
+      "step": 291
+    },
+    {
+      "epoch": 0.06513495427169307,
+      "grad_norm": 0.20416077971458435,
+      "learning_rate": 1.9989898964653753e-05,
+      "loss": 0.5755,
+      "step": 292
+    },
+    {
+      "epoch": 0.06535801918358242,
+      "grad_norm": 0.24204829335212708,
+      "learning_rate": 1.998979293580942e-05,
+      "loss": 0.5679,
+      "step": 293
+    },
+    {
+      "epoch": 0.06558108409547178,
+      "grad_norm": 0.20215743780136108,
+      "learning_rate": 1.9989686353667522e-05,
+      "loss": 0.5596,
+      "step": 294
+    },
+    {
+      "epoch": 0.06580414900736115,
+      "grad_norm": 0.24796119332313538,
+      "learning_rate": 1.998957921823395e-05,
+      "loss": 0.546,
+      "step": 295
+    },
+    {
+      "epoch": 0.0660272139192505,
+      "grad_norm": 0.22253303229808807,
+      "learning_rate": 1.9989471529514647e-05,
+      "loss": 0.5606,
+      "step": 296
+    },
+    {
+      "epoch": 0.06625027883113986,
+      "grad_norm": 0.19948336482048035,
+      "learning_rate": 1.9989363287515577e-05,
+      "loss": 0.596,
+      "step": 297
+    },
+    {
+      "epoch": 0.06647334374302923,
+      "grad_norm": 0.26588407158851624,
+      "learning_rate": 1.9989254492242727e-05,
+      "loss": 0.5347,
+      "step": 298
+    },
+    {
+      "epoch": 0.06669640865491858,
+      "grad_norm": 0.2680055797100067,
+      "learning_rate": 1.9989145143702132e-05,
+      "loss": 0.588,
+      "step": 299
+    },
+    {
+      "epoch": 0.06691947356680794,
+      "grad_norm": 0.2872098684310913,
+      "learning_rate": 1.9989035241899844e-05,
+      "loss": 0.5261,
+      "step": 300
+    },
+    {
+      "epoch": 0.0671425384786973,
+      "grad_norm": 0.19267728924751282,
+      "learning_rate": 1.998892478684195e-05,
+      "loss": 0.5807,
+      "step": 301
+    },
+    {
+      "epoch": 0.06736560339058666,
+      "grad_norm": 0.20785115659236908,
+      "learning_rate": 1.9988813778534568e-05,
+      "loss": 0.5972,
+      "step": 302
+    },
+    {
+      "epoch": 0.06758866830247602,
+      "grad_norm": 0.1926824450492859,
+      "learning_rate": 1.998870221698385e-05,
+      "loss": 0.5495,
+      "step": 303
+    },
+    {
+      "epoch": 0.06781173321436539,
+      "grad_norm": 0.21504642069339752,
+      "learning_rate": 1.9988590102195968e-05,
+      "loss": 0.5593,
+      "step": 304
+    },
+    {
+      "epoch": 0.06803479812625474,
+      "grad_norm": 0.22837965190410614,
+      "learning_rate": 1.9988477434177136e-05,
+      "loss": 0.5749,
+      "step": 305
+    },
+    {
+      "epoch": 0.0682578630381441,
+      "grad_norm": 0.27670159935951233,
+      "learning_rate": 1.9988364212933595e-05,
+      "loss": 0.5474,
+      "step": 306
+    },
+    {
+      "epoch": 0.06848092795003347,
+      "grad_norm": 0.1969708949327469,
+      "learning_rate": 1.9988250438471612e-05,
+      "loss": 0.573,
+      "step": 307
+    },
+    {
+      "epoch": 0.06870399286192282,
+      "grad_norm": 0.19591858983039856,
+      "learning_rate": 1.9988136110797494e-05,
+      "loss": 0.5677,
+      "step": 308
+    },
+    {
+      "epoch": 0.06892705777381218,
+      "grad_norm": 0.1888752281665802,
+      "learning_rate": 1.998802122991757e-05,
+      "loss": 0.5485,
+      "step": 309
+    },
+    {
+      "epoch": 0.06915012268570155,
+      "grad_norm": 0.20352420210838318,
+      "learning_rate": 1.9987905795838204e-05,
+      "loss": 0.5635,
+      "step": 310
+    },
+    {
+      "epoch": 0.0693731875975909,
+      "grad_norm": 0.19486026465892792,
+      "learning_rate": 1.9987789808565785e-05,
+      "loss": 0.5624,
+      "step": 311
+    },
+    {
+      "epoch": 0.06959625250948026,
+      "grad_norm": 0.20236323773860931,
+      "learning_rate": 1.9987673268106742e-05,
+      "loss": 0.5817,
+      "step": 312
+    },
+    {
+      "epoch": 0.06981931742136963,
+      "grad_norm": 0.2420659214258194,
+      "learning_rate": 1.998755617446753e-05,
+      "loss": 0.5586,
+      "step": 313
+    },
+    {
+      "epoch": 0.07004238233325898,
+      "grad_norm": 0.21799051761627197,
+      "learning_rate": 1.9987438527654633e-05,
+      "loss": 0.5918,
+      "step": 314
+    },
+    {
+      "epoch": 0.07026544724514834,
+      "grad_norm": 0.19442883133888245,
+      "learning_rate": 1.9987320327674566e-05,
+      "loss": 0.5621,
+      "step": 315
+    },
+    {
+      "epoch": 0.07048851215703769,
+      "grad_norm": 0.19582122564315796,
+      "learning_rate": 1.9987201574533876e-05,
+      "loss": 0.5432,
+      "step": 316
+    },
+    {
+      "epoch": 0.07071157706892706,
+      "grad_norm": 0.1984792798757553,
+      "learning_rate": 1.998708226823914e-05,
+      "loss": 0.5532,
+      "step": 317
+    },
+    {
+      "epoch": 0.07093464198081642,
+      "grad_norm": 0.2259417176246643,
+      "learning_rate": 1.9986962408796972e-05,
+      "loss": 0.5676,
+      "step": 318
+    },
+    {
+      "epoch": 0.07115770689270577,
+      "grad_norm": 0.2008512020111084,
+      "learning_rate": 1.9986841996213998e-05,
+      "loss": 0.5499,
+      "step": 319
+    },
+    {
+      "epoch": 0.07138077180459514,
+      "grad_norm": 0.19798852503299713,
+      "learning_rate": 1.99867210304969e-05,
+      "loss": 0.564,
+      "step": 320
+    },
+    {
+      "epoch": 0.0716038367164845,
+      "grad_norm": 0.21567919850349426,
+      "learning_rate": 1.998659951165237e-05,
+      "loss": 0.5701,
+      "step": 321
+    },
+    {
+      "epoch": 0.07182690162837385,
+      "grad_norm": 0.21192528307437897,
+      "learning_rate": 1.998647743968714e-05,
+      "loss": 0.595,
+      "step": 322
+    },
+    {
+      "epoch": 0.07204996654026322,
+      "grad_norm": 0.1853858083486557,
+      "learning_rate": 1.9986354814607974e-05,
+      "loss": 0.5546,
+      "step": 323
+    },
+    {
+      "epoch": 0.07227303145215258,
+      "grad_norm": 0.19545771181583405,
+      "learning_rate": 1.998623163642166e-05,
+      "loss": 0.5602,
+      "step": 324
+    },
+    {
+      "epoch": 0.07249609636404193,
+      "grad_norm": 0.20142099261283875,
+      "learning_rate": 1.998610790513502e-05,
+      "loss": 0.5589,
+      "step": 325
+    },
+    {
+      "epoch": 0.0727191612759313,
+      "grad_norm": 0.1894591748714447,
+      "learning_rate": 1.9985983620754914e-05,
+      "loss": 0.5855,
+      "step": 326
+    },
+    {
+      "epoch": 0.07294222618782066,
+      "grad_norm": 0.19726653397083282,
+      "learning_rate": 1.998585878328822e-05,
+      "loss": 0.559,
+      "step": 327
+    },
+    {
+      "epoch": 0.07316529109971001,
+      "grad_norm": 0.19552524387836456,
+      "learning_rate": 1.998573339274185e-05,
+      "loss": 0.5643,
+      "step": 328
+    },
+    {
+      "epoch": 0.07338835601159938,
+      "grad_norm": 0.2093891054391861,
+      "learning_rate": 1.9985607449122754e-05,
+      "loss": 0.5661,
+      "step": 329
+    },
+    {
+      "epoch": 0.07361142092348874,
+      "grad_norm": 0.2397080659866333,
+      "learning_rate": 1.9985480952437902e-05,
+      "loss": 0.565,
+      "step": 330
+    },
+    {
+      "epoch": 0.07383448583537809,
+      "grad_norm": 0.21570439636707306,
+      "learning_rate": 1.998535390269431e-05,
+      "loss": 0.5505,
+      "step": 331
+    },
+    {
+      "epoch": 0.07405755074726746,
+      "grad_norm": 0.18352028727531433,
+      "learning_rate": 1.9985226299899006e-05,
+      "loss": 0.5501,
+      "step": 332
+    },
+    {
+      "epoch": 0.07428061565915682,
+      "grad_norm": 0.20006653666496277,
+      "learning_rate": 1.9985098144059058e-05,
+      "loss": 0.5559,
+      "step": 333
+    },
+    {
+      "epoch": 0.07450368057104617,
+      "grad_norm": 0.22780755162239075,
+      "learning_rate": 1.998496943518157e-05,
+      "loss": 0.5886,
+      "step": 334
+    },
+    {
+      "epoch": 0.07472674548293554,
+      "grad_norm": 0.18735693395137787,
+      "learning_rate": 1.9984840173273662e-05,
+      "loss": 0.5714,
+      "step": 335
+    },
+    {
+      "epoch": 0.0749498103948249,
+      "grad_norm": 0.1882167011499405,
+      "learning_rate": 1.99847103583425e-05,
+      "loss": 0.5586,
+      "step": 336
+    },
+    {
+      "epoch": 0.07517287530671425,
+      "grad_norm": 0.20951317250728607,
+      "learning_rate": 1.9984579990395274e-05,
+      "loss": 0.5286,
+      "step": 337
+    },
+    {
+      "epoch": 0.07539594021860362,
+      "grad_norm": 0.2166478931903839,
+      "learning_rate": 1.9984449069439197e-05,
+      "loss": 0.5661,
+      "step": 338
+    },
+    {
+      "epoch": 0.07561900513049297,
+      "grad_norm": 0.191757470369339,
+      "learning_rate": 1.998431759548153e-05,
+      "loss": 0.5431,
+      "step": 339
+    },
+    {
+      "epoch": 0.07584207004238233,
+      "grad_norm": 0.2109043151140213,
+      "learning_rate": 1.998418556852955e-05,
+      "loss": 0.5642,
+      "step": 340
+    },
+    {
+      "epoch": 0.0760651349542717,
+      "grad_norm": 0.2524346709251404,
+      "learning_rate": 1.9984052988590573e-05,
+      "loss": 0.5505,
+      "step": 341
+    },
+    {
+      "epoch": 0.07628819986616105,
+      "grad_norm": 0.22443033754825592,
+      "learning_rate": 1.9983919855671937e-05,
+      "loss": 0.5454,
+      "step": 342
+    },
+    {
+      "epoch": 0.07651126477805041,
+      "grad_norm": 0.1956530660390854,
+      "learning_rate": 1.9983786169781017e-05,
+      "loss": 0.5546,
+      "step": 343
+    },
+    {
+      "epoch": 0.07673432968993978,
+      "grad_norm": 0.20710930228233337,
+      "learning_rate": 1.9983651930925217e-05,
+      "loss": 0.5554,
+      "step": 344
+    },
+    {
+      "epoch": 0.07695739460182913,
+      "grad_norm": 0.2150043398141861,
+      "learning_rate": 1.9983517139111978e-05,
+      "loss": 0.5434,
+      "step": 345
+    },
+    {
+      "epoch": 0.07718045951371849,
+      "grad_norm": 0.21276871860027313,
+      "learning_rate": 1.998338179434876e-05,
+      "loss": 0.5623,
+      "step": 346
+    },
+    {
+      "epoch": 0.07740352442560786,
+      "grad_norm": 0.20740285515785217,
+      "learning_rate": 1.998324589664306e-05,
+      "loss": 0.5804,
+      "step": 347
+    },
+    {
+      "epoch": 0.0776265893374972,
+      "grad_norm": 0.20988604426383972,
+      "learning_rate": 1.99831094460024e-05,
+      "loss": 0.5742,
+      "step": 348
+    },
+    {
+      "epoch": 0.07784965424938657,
+      "grad_norm": 0.19990280270576477,
+      "learning_rate": 1.9982972442434346e-05,
+      "loss": 0.5397,
+      "step": 349
+    },
+    {
+      "epoch": 0.07807271916127594,
+      "grad_norm": 0.295187383890152,
+      "learning_rate": 1.9982834885946482e-05,
+      "loss": 0.5713,
+      "step": 350
+    },
+    {
+      "epoch": 0.07829578407316529,
+      "grad_norm": 0.20176567137241364,
+      "learning_rate": 1.998269677654643e-05,
+      "loss": 0.5548,
+      "step": 351
+    },
+    {
+      "epoch": 0.07851884898505465,
+      "grad_norm": 0.20890064537525177,
+      "learning_rate": 1.9982558114241837e-05,
+      "loss": 0.5739,
+      "step": 352
+    },
+    {
+      "epoch": 0.07874191389694402,
+      "grad_norm": 0.2795037031173706,
+      "learning_rate": 1.998241889904038e-05,
+      "loss": 0.5343,
+      "step": 353
+    },
+    {
+      "epoch": 0.07896497880883337,
+      "grad_norm": 0.22098244726657867,
+      "learning_rate": 1.9982279130949775e-05,
+      "loss": 0.517,
+      "step": 354
+    },
+    {
+      "epoch": 0.07918804372072273,
+      "grad_norm": 0.20403707027435303,
+      "learning_rate": 1.998213880997776e-05,
+      "loss": 0.5599,
+      "step": 355
+    },
+    {
+      "epoch": 0.0794111086326121,
+      "grad_norm": 0.2201310694217682,
+      "learning_rate": 1.9981997936132107e-05,
+      "loss": 0.5348,
+      "step": 356
+    },
+    {
+      "epoch": 0.07963417354450145,
+      "grad_norm": 0.22818854451179504,
+      "learning_rate": 1.998185650942062e-05,
+      "loss": 0.551,
+      "step": 357
+    },
+    {
+      "epoch": 0.07985723845639081,
+      "grad_norm": 0.23051077127456665,
+      "learning_rate": 1.9981714529851127e-05,
+      "loss": 0.5926,
+      "step": 358
+    },
+    {
+      "epoch": 0.08008030336828018,
+      "grad_norm": 0.21485844254493713,
+      "learning_rate": 1.99815719974315e-05,
+      "loss": 0.5801,
+      "step": 359
+    },
+    {
+      "epoch": 0.08030336828016953,
+      "grad_norm": 0.1935802698135376,
+      "learning_rate": 1.998142891216963e-05,
+      "loss": 0.5411,
+      "step": 360
+    },
+    {
+      "epoch": 0.08052643319205889,
+      "grad_norm": 0.1971837282180786,
+      "learning_rate": 1.998128527407344e-05,
+      "loss": 0.5805,
+      "step": 361
+    },
+    {
+      "epoch": 0.08074949810394826,
+      "grad_norm": 0.20717322826385498,
+      "learning_rate": 1.9981141083150886e-05,
+      "loss": 0.5493,
+      "step": 362
+    },
+    {
+      "epoch": 0.0809725630158376,
+      "grad_norm": 0.2011585831642151,
+      "learning_rate": 1.9980996339409957e-05,
+      "loss": 0.5579,
+      "step": 363
+    },
+    {
+      "epoch": 0.08119562792772697,
+      "grad_norm": 0.19211354851722717,
+      "learning_rate": 1.9980851042858664e-05,
+      "loss": 0.5383,
+      "step": 364
+    },
+    {
+      "epoch": 0.08141869283961632,
+      "grad_norm": 0.21699954569339752,
+      "learning_rate": 1.998070519350506e-05,
+      "loss": 0.5524,
+      "step": 365
+    },
+    {
+      "epoch": 0.08164175775150569,
+      "grad_norm": 0.20210997760295868,
+      "learning_rate": 1.9980558791357222e-05,
+      "loss": 0.5738,
+      "step": 366
+    },
+    {
+      "epoch": 0.08186482266339505,
+      "grad_norm": 0.18855418264865875,
+      "learning_rate": 1.9980411836423256e-05,
+      "loss": 0.547,
+      "step": 367
+    },
+    {
+      "epoch": 0.0820878875752844,
+      "grad_norm": 0.21640720963478088,
+      "learning_rate": 1.9980264328711305e-05,
+      "loss": 0.5964,
+      "step": 368
+    },
+    {
+      "epoch": 0.08231095248717377,
+      "grad_norm": 0.2280052751302719,
+      "learning_rate": 1.9980116268229536e-05,
+      "loss": 0.561,
+      "step": 369
+    },
+    {
+      "epoch": 0.08253401739906313,
+      "grad_norm": 0.1860683411359787,
+      "learning_rate": 1.9979967654986155e-05,
+      "loss": 0.545,
+      "step": 370
+    },
+    {
+      "epoch": 0.08275708231095248,
+      "grad_norm": 0.19923582673072815,
+      "learning_rate": 1.9979818488989383e-05,
+      "loss": 0.5421,
+      "step": 371
+    },
+    {
+      "epoch": 0.08298014722284185,
+      "grad_norm": 0.23478665947914124,
+      "learning_rate": 1.997966877024749e-05,
+      "loss": 0.5652,
+      "step": 372
+    },
+    {
+      "epoch": 0.08320321213473121,
+      "grad_norm": 0.2126697599887848,
+      "learning_rate": 1.9979518498768768e-05,
+      "loss": 0.5599,
+      "step": 373
+    },
+    {
+      "epoch": 0.08342627704662056,
+      "grad_norm": 0.20854973793029785,
+      "learning_rate": 1.9979367674561535e-05,
+      "loss": 0.5425,
+      "step": 374
+    },
+    {
+      "epoch": 0.08364934195850993,
+      "grad_norm": 0.20865046977996826,
+      "learning_rate": 1.9979216297634148e-05,
+      "loss": 0.5529,
+      "step": 375
+    },
+    {
+      "epoch": 0.08387240687039929,
+      "grad_norm": 0.24369299411773682,
+      "learning_rate": 1.997906436799499e-05,
+      "loss": 0.535,
+      "step": 376
+    },
+    {
+      "epoch": 0.08409547178228864,
+      "grad_norm": 0.2064153105020523,
+      "learning_rate": 1.9978911885652475e-05,
+      "loss": 0.5488,
+      "step": 377
+    },
+    {
+      "epoch": 0.084318536694178,
+      "grad_norm": 0.21306705474853516,
+      "learning_rate": 1.997875885061505e-05,
+      "loss": 0.5627,
+      "step": 378
+    },
+    {
+      "epoch": 0.08454160160606737,
+      "grad_norm": 0.19575418531894684,
+      "learning_rate": 1.9978605262891196e-05,
+      "loss": 0.5565,
+      "step": 379
+    },
+    {
+      "epoch": 0.08476466651795672,
+      "grad_norm": 0.19842910766601562,
+      "learning_rate": 1.9978451122489412e-05,
+      "loss": 0.5709,
+      "step": 380
+    },
+    {
+      "epoch": 0.08498773142984609,
+      "grad_norm": 0.20483329892158508,
+      "learning_rate": 1.9978296429418237e-05,
+      "loss": 0.5773,
+      "step": 381
+    },
+    {
+      "epoch": 0.08521079634173545,
+      "grad_norm": 0.22734715044498444,
+      "learning_rate": 1.997814118368624e-05,
+      "loss": 0.5469,
+      "step": 382
+    },
+    {
+      "epoch": 0.0854338612536248,
+      "grad_norm": 0.19727206230163574,
+      "learning_rate": 1.997798538530202e-05,
+      "loss": 0.5541,
+      "step": 383
+    },
+    {
+      "epoch": 0.08565692616551417,
+      "grad_norm": 0.19417184591293335,
+      "learning_rate": 1.9977829034274205e-05,
+      "loss": 0.5513,
+      "step": 384
+    },
+    {
+      "epoch": 0.08587999107740353,
+      "grad_norm": 0.2037704885005951,
+      "learning_rate": 1.9977672130611454e-05,
+      "loss": 0.5567,
+      "step": 385
+    },
+    {
+      "epoch": 0.08610305598929288,
+      "grad_norm": 0.18121762573719025,
+      "learning_rate": 1.997751467432246e-05,
+      "loss": 0.5357,
+      "step": 386
+    },
+    {
+      "epoch": 0.08632612090118225,
+      "grad_norm": 0.27105584740638733,
+      "learning_rate": 1.997735666541594e-05,
+      "loss": 0.5654,
+      "step": 387
+    },
+    {
+      "epoch": 0.08654918581307161,
+      "grad_norm": 0.20133094489574432,
+      "learning_rate": 1.997719810390065e-05,
+      "loss": 0.5349,
+      "step": 388
+    },
+    {
+      "epoch": 0.08677225072496096,
+      "grad_norm": 0.18842002749443054,
+      "learning_rate": 1.997703898978537e-05,
+      "loss": 0.5352,
+      "step": 389
+    },
+    {
+      "epoch": 0.08699531563685033,
+      "grad_norm": 0.2452889233827591,
+      "learning_rate": 1.9976879323078913e-05,
+      "loss": 0.5899,
+      "step": 390
+    },
+    {
+      "epoch": 0.08721838054873968,
+      "grad_norm": 0.2256636619567871,
+      "learning_rate": 1.9976719103790118e-05,
+      "loss": 0.5115,
+      "step": 391
+    },
+    {
+      "epoch": 0.08744144546062904,
+      "grad_norm": 0.1972501426935196,
+      "learning_rate": 1.9976558331927868e-05,
+      "loss": 0.5523,
+      "step": 392
+    },
+    {
+      "epoch": 0.0876645103725184,
+      "grad_norm": 0.2026263177394867,
+      "learning_rate": 1.9976397007501062e-05,
+      "loss": 0.5758,
+      "step": 393
+    },
+    {
+      "epoch": 0.08788757528440776,
+      "grad_norm": 0.20572836697101593,
+      "learning_rate": 1.9976235130518632e-05,
+      "loss": 0.5429,
+      "step": 394
+    },
+    {
+      "epoch": 0.08811064019629712,
+      "grad_norm": 0.18851211667060852,
+      "learning_rate": 1.997607270098955e-05,
+      "loss": 0.5616,
+      "step": 395
+    },
+    {
+      "epoch": 0.08833370510818649,
+      "grad_norm": 0.20447294414043427,
+      "learning_rate": 1.9975909718922806e-05,
+      "loss": 0.5527,
+      "step": 396
+    },
+    {
+      "epoch": 0.08855677002007584,
+      "grad_norm": 0.19188085198402405,
+      "learning_rate": 1.997574618432744e-05,
+      "loss": 0.558,
+      "step": 397
+    },
+    {
+      "epoch": 0.0887798349319652,
+      "grad_norm": 0.19854480028152466,
+      "learning_rate": 1.997558209721249e-05,
+      "loss": 0.5756,
+      "step": 398
+    },
+    {
+      "epoch": 0.08900289984385457,
+      "grad_norm": 0.19832177460193634,
+      "learning_rate": 1.997541745758706e-05,
+      "loss": 0.569,
+      "step": 399
+    },
+    {
+      "epoch": 0.08922596475574392,
+      "grad_norm": 0.20165055990219116,
+      "learning_rate": 1.9975252265460265e-05,
+      "loss": 0.5376,
+      "step": 400
+    },
+    {
+      "epoch": 0.08944902966763328,
+      "grad_norm": 0.18997038900852203,
+      "learning_rate": 1.997508652084125e-05,
+      "loss": 0.5745,
+      "step": 401
+    },
+    {
+      "epoch": 0.08967209457952265,
+      "grad_norm": 0.19646713137626648,
+      "learning_rate": 1.9974920223739195e-05,
+      "loss": 0.5454,
+      "step": 402
+    },
+    {
+      "epoch": 0.089895159491412,
+      "grad_norm": 0.1976052224636078,
+      "learning_rate": 1.997475337416332e-05,
+      "loss": 0.537,
+      "step": 403
+    },
+    {
+      "epoch": 0.09011822440330136,
+      "grad_norm": 0.2045581340789795,
+      "learning_rate": 1.9974585972122857e-05,
+      "loss": 0.5405,
+      "step": 404
+    },
+    {
+      "epoch": 0.09034128931519073,
+      "grad_norm": 0.18631087243556976,
+      "learning_rate": 1.9974418017627076e-05,
+      "loss": 0.6034,
+      "step": 405
+    },
+    {
+      "epoch": 0.09056435422708008,
+      "grad_norm": 0.2022215723991394,
+      "learning_rate": 1.9974249510685285e-05,
+      "loss": 0.5682,
+      "step": 406
+    },
+    {
+      "epoch": 0.09078741913896944,
+      "grad_norm": 0.19569364190101624,
+      "learning_rate": 1.9974080451306816e-05,
+      "loss": 0.5739,
+      "step": 407
+    },
+    {
+      "epoch": 0.0910104840508588,
+      "grad_norm": 0.20746709406375885,
+      "learning_rate": 1.9973910839501035e-05,
+      "loss": 0.55,
+      "step": 408
+    },
+    {
+      "epoch": 0.09123354896274816,
+      "grad_norm": 0.21757768094539642,
+      "learning_rate": 1.997374067527733e-05,
+      "loss": 0.5404,
+      "step": 409
+    },
+    {
+      "epoch": 0.09145661387463752,
+      "grad_norm": 0.1881277859210968,
+      "learning_rate": 1.997356995864513e-05,
+      "loss": 0.5435,
+      "step": 410
+    },
+    {
+      "epoch": 0.09167967878652689,
+      "grad_norm": 0.1868787705898285,
+      "learning_rate": 1.9973398689613892e-05,
+      "loss": 0.5373,
+      "step": 411
+    },
+    {
+      "epoch": 0.09190274369841624,
+      "grad_norm": 0.1938547044992447,
+      "learning_rate": 1.9973226868193096e-05,
+      "loss": 0.5846,
+      "step": 412
+    },
+    {
+      "epoch": 0.0921258086103056,
+      "grad_norm": 0.19920659065246582,
+      "learning_rate": 1.9973054494392265e-05,
+      "loss": 0.5615,
+      "step": 413
+    },
+    {
+      "epoch": 0.09234887352219495,
+      "grad_norm": 0.18801911175251007,
+      "learning_rate": 1.997288156822094e-05,
+      "loss": 0.5675,
+      "step": 414
+    },
+    {
+      "epoch": 0.09257193843408432,
+      "grad_norm": 0.1990530639886856,
+      "learning_rate": 1.9972708089688705e-05,
+      "loss": 0.5626,
+      "step": 415
+    },
+    {
+      "epoch": 0.09279500334597368,
+      "grad_norm": 0.18555153906345367,
+      "learning_rate": 1.9972534058805163e-05,
+      "loss": 0.557,
+      "step": 416
+    },
+    {
+      "epoch": 0.09301806825786303,
+      "grad_norm": 0.20318473875522614,
+      "learning_rate": 1.9972359475579953e-05,
+      "loss": 0.5196,
+      "step": 417
+    },
+    {
+      "epoch": 0.0932411331697524,
+      "grad_norm": 0.20757220685482025,
+      "learning_rate": 1.997218434002275e-05,
+      "loss": 0.5454,
+      "step": 418
+    },
+    {
+      "epoch": 0.09346419808164176,
+      "grad_norm": 0.20588824152946472,
+      "learning_rate": 1.997200865214325e-05,
+      "loss": 0.5347,
+      "step": 419
+    },
+    {
+      "epoch": 0.09368726299353111,
+      "grad_norm": 0.20199880003929138,
+      "learning_rate": 1.9971832411951186e-05,
+      "loss": 0.5531,
+      "step": 420
+    },
+    {
+      "epoch": 0.09391032790542048,
+      "grad_norm": 0.19605985283851624,
+      "learning_rate": 1.9971655619456313e-05,
+      "loss": 0.5263,
+      "step": 421
+    },
+    {
+      "epoch": 0.09413339281730984,
+      "grad_norm": 0.19716261327266693,
+      "learning_rate": 1.997147827466843e-05,
+      "loss": 0.5722,
+      "step": 422
+    },
+    {
+      "epoch": 0.09435645772919919,
+      "grad_norm": 0.20400184392929077,
+      "learning_rate": 1.997130037759736e-05,
+      "loss": 0.5629,
+      "step": 423
+    },
+    {
+      "epoch": 0.09457952264108856,
+      "grad_norm": 0.203161358833313,
+      "learning_rate": 1.997112192825295e-05,
+      "loss": 0.5538,
+      "step": 424
+    },
+    {
+      "epoch": 0.09480258755297792,
+      "grad_norm": 0.20870056748390198,
+      "learning_rate": 1.997094292664509e-05,
+      "loss": 0.5548,
+      "step": 425
+    },
+    {
+      "epoch": 0.09502565246486727,
+      "grad_norm": 0.1856376975774765,
+      "learning_rate": 1.9970763372783687e-05,
+      "loss": 0.5193,
+      "step": 426
+    },
+    {
+      "epoch": 0.09524871737675664,
+      "grad_norm": 0.1954246610403061,
+      "learning_rate": 1.997058326667869e-05,
+      "loss": 0.5336,
+      "step": 427
+    },
+    {
+      "epoch": 0.095471782288646,
+      "grad_norm": 0.1996607482433319,
+      "learning_rate": 1.9970402608340076e-05,
+      "loss": 0.5708,
+      "step": 428
+    },
+    {
+      "epoch": 0.09569484720053535,
+      "grad_norm": 0.21878129243850708,
+      "learning_rate": 1.9970221397777848e-05,
+      "loss": 0.5794,
+      "step": 429
+    },
+    {
+      "epoch": 0.09591791211242472,
+      "grad_norm": 0.19142527878284454,
+      "learning_rate": 1.9970039635002044e-05,
+      "loss": 0.5221,
+      "step": 430
+    },
+    {
+      "epoch": 0.09614097702431408,
+      "grad_norm": 0.18871352076530457,
+      "learning_rate": 1.996985732002273e-05,
+      "loss": 0.5321,
+      "step": 431
+    },
+    {
+      "epoch": 0.09636404193620343,
+      "grad_norm": 0.1942790001630783,
+      "learning_rate": 1.996967445285001e-05,
+      "loss": 0.5269,
+      "step": 432
+    },
+    {
+      "epoch": 0.0965871068480928,
+      "grad_norm": 0.19900670647621155,
+      "learning_rate": 1.9969491033494e-05,
+      "loss": 0.6048,
+      "step": 433
+    },
+    {
+      "epoch": 0.09681017175998216,
+      "grad_norm": 0.193121075630188,
+      "learning_rate": 1.9969307061964873e-05,
+      "loss": 0.5613,
+      "step": 434
+    },
+    {
+      "epoch": 0.09703323667187151,
+      "grad_norm": 0.1859043687582016,
+      "learning_rate": 1.9969122538272807e-05,
+      "loss": 0.5657,
+      "step": 435
+    },
+    {
+      "epoch": 0.09725630158376088,
+      "grad_norm": 0.18597924709320068,
+      "learning_rate": 1.9968937462428028e-05,
+      "loss": 0.5336,
+      "step": 436
+    },
+    {
+      "epoch": 0.09747936649565024,
+      "grad_norm": 0.17796443402767181,
+      "learning_rate": 1.9968751834440783e-05,
+      "loss": 0.5611,
+      "step": 437
+    },
+    {
+      "epoch": 0.09770243140753959,
+      "grad_norm": 0.1958586871623993,
+      "learning_rate": 1.9968565654321356e-05,
+      "loss": 0.5377,
+      "step": 438
+    },
+    {
+      "epoch": 0.09792549631942896,
+      "grad_norm": 0.21395692229270935,
+      "learning_rate": 1.996837892208006e-05,
+      "loss": 0.5347,
+      "step": 439
+    },
+    {
+      "epoch": 0.09814856123131831,
+      "grad_norm": 0.19348031282424927,
+      "learning_rate": 1.9968191637727235e-05,
+      "loss": 0.5576,
+      "step": 440
+    },
+    {
+      "epoch": 0.09837162614320767,
+      "grad_norm": 0.19345401227474213,
+      "learning_rate": 1.9968003801273253e-05,
+      "loss": 0.5697,
+      "step": 441
+    },
+    {
+      "epoch": 0.09859469105509704,
+      "grad_norm": 0.18203531205654144,
+      "learning_rate": 1.9967815412728523e-05,
+      "loss": 0.5416,
+      "step": 442
+    },
+    {
+      "epoch": 0.09881775596698639,
+      "grad_norm": 0.20109198987483978,
+      "learning_rate": 1.9967626472103472e-05,
+      "loss": 0.5514,
+      "step": 443
+    },
+    {
+      "epoch": 0.09904082087887575,
+      "grad_norm": 0.18818293511867523,
+      "learning_rate": 1.996743697940857e-05,
+      "loss": 0.5535,
+      "step": 444
+    },
+    {
+      "epoch": 0.09926388579076512,
+      "grad_norm": 0.18718019127845764,
+      "learning_rate": 1.996724693465431e-05,
+      "loss": 0.5595,
+      "step": 445
+    },
+    {
+      "epoch": 0.09948695070265447,
+      "grad_norm": 0.18856103718280792,
+      "learning_rate": 1.9967056337851217e-05,
+      "loss": 0.5364,
+      "step": 446
+    },
+    {
+      "epoch": 0.09971001561454383,
+      "grad_norm": 0.18582871556282043,
+      "learning_rate": 1.996686518900985e-05,
+      "loss": 0.5632,
+      "step": 447
+    },
+    {
+      "epoch": 0.0999330805264332,
+      "grad_norm": 0.19166669249534607,
+      "learning_rate": 1.9966673488140794e-05,
+      "loss": 0.5727,
+      "step": 448
+    },
+    {
+      "epoch": 0.10015614543832255,
+      "grad_norm": 0.18631108105182648,
+      "learning_rate": 1.9966481235254667e-05,
+      "loss": 0.5472,
+      "step": 449
+    },
+    {
+      "epoch": 0.10037921035021191,
+      "grad_norm": 0.18970704078674316,
+      "learning_rate": 1.996628843036212e-05,
+      "loss": 0.5355,
+      "step": 450
+    },
+    {
+      "epoch": 0.10060227526210128,
+      "grad_norm": 0.18836888670921326,
+      "learning_rate": 1.9966095073473828e-05,
+      "loss": 0.5309,
+      "step": 451
+    },
+    {
+      "epoch": 0.10082534017399063,
+      "grad_norm": 0.1924007087945938,
+      "learning_rate": 1.99659011646005e-05,
+      "loss": 0.5213,
+      "step": 452
+    },
+    {
+      "epoch": 0.10104840508587999,
+      "grad_norm": 0.19710196554660797,
+      "learning_rate": 1.996570670375288e-05,
+      "loss": 0.5333,
+      "step": 453
+    },
+    {
+      "epoch": 0.10127146999776936,
+      "grad_norm": 0.1823819875717163,
+      "learning_rate": 1.9965511690941737e-05,
+      "loss": 0.5378,
+      "step": 454
+    },
+    {
+      "epoch": 0.1014945349096587,
+      "grad_norm": 0.21200565993785858,
+      "learning_rate": 1.9965316126177867e-05,
+      "loss": 0.5594,
+      "step": 455
+    },
+    {
+      "epoch": 0.10171759982154807,
+      "grad_norm": 0.18618617951869965,
+      "learning_rate": 1.9965120009472106e-05,
+      "loss": 0.5317,
+      "step": 456
+    },
+    {
+      "epoch": 0.10194066473343744,
+      "grad_norm": 0.19600717723369598,
+      "learning_rate": 1.996492334083532e-05,
+      "loss": 0.5612,
+      "step": 457
+    },
+    {
+      "epoch": 0.10216372964532679,
+      "grad_norm": 0.1988876461982727,
+      "learning_rate": 1.9964726120278394e-05,
+      "loss": 0.5709,
+      "step": 458
+    },
+    {
+      "epoch": 0.10238679455721615,
+      "grad_norm": 0.18938012421131134,
+      "learning_rate": 1.9964528347812255e-05,
+      "loss": 0.5413,
+      "step": 459
+    },
+    {
+      "epoch": 0.10260985946910552,
+      "grad_norm": 0.18799841403961182,
+      "learning_rate": 1.9964330023447854e-05,
+      "loss": 0.5534,
+      "step": 460
+    },
+    {
+      "epoch": 0.10283292438099487,
+      "grad_norm": 0.229729562997818,
+      "learning_rate": 1.9964131147196185e-05,
+      "loss": 0.5471,
+      "step": 461
+    },
+    {
+      "epoch": 0.10305598929288423,
+      "grad_norm": 0.19921550154685974,
+      "learning_rate": 1.9963931719068253e-05,
+      "loss": 0.5419,
+      "step": 462
+    },
+    {
+      "epoch": 0.1032790542047736,
+      "grad_norm": 0.19583772122859955,
+      "learning_rate": 1.9963731739075106e-05,
+      "loss": 0.5412,
+      "step": 463
+    },
+    {
+      "epoch": 0.10350211911666295,
+      "grad_norm": 0.19680215418338776,
+      "learning_rate": 1.996353120722782e-05,
+      "loss": 0.5477,
+      "step": 464
+    },
+    {
+      "epoch": 0.10372518402855231,
+      "grad_norm": 0.20173701643943787,
+      "learning_rate": 1.9963330123537507e-05,
+      "loss": 0.5523,
+      "step": 465
+    },
+    {
+      "epoch": 0.10394824894044166,
+      "grad_norm": 0.23238573968410492,
+      "learning_rate": 1.9963128488015294e-05,
+      "loss": 0.559,
+      "step": 466
+    },
+    {
+      "epoch": 0.10417131385233103,
+      "grad_norm": 0.22626003623008728,
+      "learning_rate": 1.996292630067236e-05,
+      "loss": 0.526,
+      "step": 467
+    },
+    {
+      "epoch": 0.10439437876422039,
+      "grad_norm": 0.21178299188613892,
+      "learning_rate": 1.9962723561519893e-05,
+      "loss": 0.5301,
+      "step": 468
+    },
+    {
+      "epoch": 0.10461744367610974,
+      "grad_norm": 0.1937963217496872,
+      "learning_rate": 1.996252027056913e-05,
+      "loss": 0.5449,
+      "step": 469
+    },
+    {
+      "epoch": 0.1048405085879991,
+      "grad_norm": 0.18095934391021729,
+      "learning_rate": 1.996231642783133e-05,
+      "loss": 0.5676,
+      "step": 470
+    },
+    {
+      "epoch": 0.10506357349988847,
+      "grad_norm": 0.2064565271139145,
+      "learning_rate": 1.9962112033317776e-05,
+      "loss": 0.5157,
+      "step": 471
+    },
+    {
+      "epoch": 0.10528663841177782,
+      "grad_norm": 0.1964641511440277,
+      "learning_rate": 1.9961907087039796e-05,
+      "loss": 0.6127,
+      "step": 472
+    },
+    {
+      "epoch": 0.10550970332366719,
+      "grad_norm": 0.17736569046974182,
+      "learning_rate": 1.996170158900874e-05,
+      "loss": 0.545,
+      "step": 473
+    },
+    {
+      "epoch": 0.10573276823555655,
+      "grad_norm": 0.1836784929037094,
+      "learning_rate": 1.9961495539235985e-05,
+      "loss": 0.5525,
+      "step": 474
+    },
+    {
+      "epoch": 0.1059558331474459,
+      "grad_norm": 0.2029949426651001,
+      "learning_rate": 1.996128893773295e-05,
+      "loss": 0.5558,
+      "step": 475
+    },
+    {
+      "epoch": 0.10617889805933527,
+      "grad_norm": 0.207282155752182,
+      "learning_rate": 1.9961081784511073e-05,
+      "loss": 0.5789,
+      "step": 476
+    },
+    {
+      "epoch": 0.10640196297122463,
+      "grad_norm": 0.20277360081672668,
+      "learning_rate": 1.9960874079581828e-05,
+      "loss": 0.5555,
+      "step": 477
+    },
+    {
+      "epoch": 0.10662502788311398,
+      "grad_norm": 0.21180874109268188,
+      "learning_rate": 1.996066582295672e-05,
+      "loss": 0.574,
+      "step": 478
+    },
+    {
+      "epoch": 0.10684809279500335,
+      "grad_norm": 0.22067199647426605,
+      "learning_rate": 1.996045701464729e-05,
+      "loss": 0.5641,
+      "step": 479
+    },
+    {
+      "epoch": 0.10707115770689271,
+      "grad_norm": 0.2735600471496582,
+      "learning_rate": 1.9960247654665088e-05,
+      "loss": 0.5473,
+      "step": 480
+    },
+    {
+      "epoch": 0.10729422261878206,
+      "grad_norm": 0.19656887650489807,
+      "learning_rate": 1.9960037743021723e-05,
+      "loss": 0.5614,
+      "step": 481
+    },
+    {
+      "epoch": 0.10751728753067143,
+      "grad_norm": 0.19436019659042358,
+      "learning_rate": 1.9959827279728815e-05,
+      "loss": 0.5554,
+      "step": 482
+    },
+    {
+      "epoch": 0.10774035244256079,
+      "grad_norm": 0.1874009072780609,
+      "learning_rate": 1.9959616264798022e-05,
+      "loss": 0.5723,
+      "step": 483
+    },
+    {
+      "epoch": 0.10796341735445014,
+      "grad_norm": 0.2946108877658844,
+      "learning_rate": 1.9959404698241037e-05,
+      "loss": 0.5572,
+      "step": 484
+    },
+    {
+      "epoch": 0.1081864822663395,
+      "grad_norm": 0.1900031566619873,
+      "learning_rate": 1.9959192580069567e-05,
+      "loss": 0.5629,
+      "step": 485
+    },
+    {
+      "epoch": 0.10840954717822887,
+      "grad_norm": 0.2330961972475052,
+      "learning_rate": 1.9958979910295367e-05,
+      "loss": 0.5725,
+      "step": 486
+    },
+    {
+      "epoch": 0.10863261209011822,
+      "grad_norm": 0.23831723630428314,
+      "learning_rate": 1.9958766688930215e-05,
+      "loss": 0.5549,
+      "step": 487
+    },
+    {
+      "epoch": 0.10885567700200759,
+      "grad_norm": 0.2095993012189865,
+      "learning_rate": 1.9958552915985923e-05,
+      "loss": 0.541,
+      "step": 488
+    },
+    {
+      "epoch": 0.10907874191389694,
+      "grad_norm": 0.18837721645832062,
+      "learning_rate": 1.9958338591474327e-05,
+      "loss": 0.5245,
+      "step": 489
+    },
+    {
+      "epoch": 0.1093018068257863,
+      "grad_norm": 0.19427752494812012,
+      "learning_rate": 1.99581237154073e-05,
+      "loss": 0.5442,
+      "step": 490
+    },
+    {
+      "epoch": 0.10952487173767567,
+      "grad_norm": 0.19493944942951202,
+      "learning_rate": 1.9957908287796743e-05,
+      "loss": 0.5559,
+      "step": 491
+    },
+    {
+      "epoch": 0.10974793664956502,
+      "grad_norm": 0.18494555354118347,
+      "learning_rate": 1.9957692308654586e-05,
+      "loss": 0.5427,
+      "step": 492
+    },
+    {
+      "epoch": 0.10997100156145438,
+      "grad_norm": 0.19978608191013336,
+      "learning_rate": 1.9957475777992794e-05,
+      "loss": 0.5391,
+      "step": 493
+    },
+    {
+      "epoch": 0.11019406647334375,
+      "grad_norm": 0.18339702486991882,
+      "learning_rate": 1.9957258695823358e-05,
+      "loss": 0.5454,
+      "step": 494
+    },
+    {
+      "epoch": 0.1104171313852331,
+      "grad_norm": 0.18613296747207642,
+      "learning_rate": 1.99570410621583e-05,
+      "loss": 0.5371,
+      "step": 495
+    },
+    {
+      "epoch": 0.11064019629712246,
+      "grad_norm": 0.19776608049869537,
+      "learning_rate": 1.9956822877009676e-05,
+      "loss": 0.5405,
+      "step": 496
+    },
+    {
+      "epoch": 0.11086326120901183,
+      "grad_norm": 0.1891903132200241,
+      "learning_rate": 1.9956604140389574e-05,
+      "loss": 0.5755,
+      "step": 497
+    },
+    {
+      "epoch": 0.11108632612090118,
+      "grad_norm": 0.20654836297035217,
+      "learning_rate": 1.9956384852310102e-05,
+      "loss": 0.5471,
+      "step": 498
+    },
+    {
+      "epoch": 0.11130939103279054,
+      "grad_norm": 0.18682821094989777,
+      "learning_rate": 1.995616501278341e-05,
+      "loss": 0.5479,
+      "step": 499
+    },
+    {
+      "epoch": 0.1115324559446799,
+      "grad_norm": 0.20868845283985138,
+      "learning_rate": 1.995594462182167e-05,
+      "loss": 0.5405,
+      "step": 500
+    },
+    {
+      "epoch": 0.11175552085656926,
+      "grad_norm": 0.20793192088603973,
+      "learning_rate": 1.9955723679437093e-05,
+      "loss": 0.5468,
+      "step": 501
+    },
+    {
+      "epoch": 0.11197858576845862,
+      "grad_norm": 0.18734110891819,
+      "learning_rate": 1.9955502185641915e-05,
+      "loss": 0.5336,
+      "step": 502
+    },
+    {
+      "epoch": 0.11220165068034799,
+      "grad_norm": 0.18563862144947052,
+      "learning_rate": 1.9955280140448404e-05,
+      "loss": 0.549,
+      "step": 503
+    },
+    {
+      "epoch": 0.11242471559223734,
+      "grad_norm": 0.19853924214839935,
+      "learning_rate": 1.9955057543868858e-05,
+      "loss": 0.5494,
+      "step": 504
+    },
+    {
+      "epoch": 0.1126477805041267,
+      "grad_norm": 0.2054789960384369,
+      "learning_rate": 1.9954834395915604e-05,
+      "loss": 0.6006,
+      "step": 505
+    },
+    {
+      "epoch": 0.11287084541601607,
+      "grad_norm": 0.1880260407924652,
+      "learning_rate": 1.9954610696601e-05,
+      "loss": 0.5664,
+      "step": 506
+    },
+    {
+      "epoch": 0.11309391032790542,
+      "grad_norm": 0.20359358191490173,
+      "learning_rate": 1.9954386445937444e-05,
+      "loss": 0.5387,
+      "step": 507
+    },
+    {
+      "epoch": 0.11331697523979478,
+      "grad_norm": 0.20446328818798065,
+      "learning_rate": 1.995416164393735e-05,
+      "loss": 0.6078,
+      "step": 508
+    },
+    {
+      "epoch": 0.11354004015168415,
+      "grad_norm": 0.20027638971805573,
+      "learning_rate": 1.9953936290613166e-05,
+      "loss": 0.4875,
+      "step": 509
+    },
+    {
+      "epoch": 0.1137631050635735,
+      "grad_norm": 0.19979824125766754,
+      "learning_rate": 1.9953710385977382e-05,
+      "loss": 0.5599,
+      "step": 510
+    },
+    {
+      "epoch": 0.11398616997546286,
+      "grad_norm": 0.17714561522006989,
+      "learning_rate": 1.9953483930042503e-05,
+      "loss": 0.536,
+      "step": 511
+    },
+    {
+      "epoch": 0.11420923488735223,
+      "grad_norm": 0.21002604067325592,
+      "learning_rate": 1.9953256922821075e-05,
+      "loss": 0.5388,
+      "step": 512
+    },
+    {
+      "epoch": 0.11443229979924158,
+      "grad_norm": 0.19017495214939117,
+      "learning_rate": 1.995302936432567e-05,
+      "loss": 0.5474,
+      "step": 513
+    },
+    {
+      "epoch": 0.11465536471113094,
+      "grad_norm": 0.1896909773349762,
+      "learning_rate": 1.995280125456889e-05,
+      "loss": 0.5577,
+      "step": 514
+    },
+    {
+      "epoch": 0.11487842962302029,
+      "grad_norm": 0.20575307309627533,
+      "learning_rate": 1.9952572593563375e-05,
+      "loss": 0.5204,
+      "step": 515
+    },
+    {
+      "epoch": 0.11510149453490966,
+      "grad_norm": 0.19565477967262268,
+      "learning_rate": 1.9952343381321785e-05,
+      "loss": 0.5971,
+      "step": 516
+    },
+    {
+      "epoch": 0.11532455944679902,
+      "grad_norm": 0.1984451860189438,
+      "learning_rate": 1.995211361785681e-05,
+      "loss": 0.5531,
+      "step": 517
+    },
+    {
+      "epoch": 0.11554762435868837,
+      "grad_norm": 0.18850469589233398,
+      "learning_rate": 1.9951883303181184e-05,
+      "loss": 0.5419,
+      "step": 518
+    },
+    {
+      "epoch": 0.11577068927057774,
+      "grad_norm": 0.22787964344024658,
+      "learning_rate": 1.9951652437307664e-05,
+      "loss": 0.5403,
+      "step": 519
+    },
+    {
+      "epoch": 0.1159937541824671,
+      "grad_norm": 0.29980340600013733,
+      "learning_rate": 1.995142102024903e-05,
+      "loss": 0.5365,
+      "step": 520
+    },
+    {
+      "epoch": 0.11621681909435645,
+      "grad_norm": 0.17774684727191925,
+      "learning_rate": 1.995118905201811e-05,
+      "loss": 0.545,
+      "step": 521
+    },
+    {
+      "epoch": 0.11643988400624582,
+      "grad_norm": 0.20445318520069122,
+      "learning_rate": 1.995095653262774e-05,
+      "loss": 0.5883,
+      "step": 522
+    },
+    {
+      "epoch": 0.11666294891813518,
+      "grad_norm": 0.1800977885723114,
+      "learning_rate": 1.9950723462090803e-05,
+      "loss": 0.5579,
+      "step": 523
+    },
+    {
+      "epoch": 0.11688601383002453,
+      "grad_norm": 0.18807321786880493,
+      "learning_rate": 1.9950489840420207e-05,
+      "loss": 0.5295,
+      "step": 524
+    },
+    {
+      "epoch": 0.1171090787419139,
+      "grad_norm": 0.17663832008838654,
+      "learning_rate": 1.9950255667628894e-05,
+      "loss": 0.5377,
+      "step": 525
+    },
+    {
+      "epoch": 0.11733214365380326,
+      "grad_norm": 0.18556755781173706,
+      "learning_rate": 1.9950020943729834e-05,
+      "loss": 0.5314,
+      "step": 526
+    },
+    {
+      "epoch": 0.11755520856569261,
+      "grad_norm": 0.18376345932483673,
+      "learning_rate": 1.994978566873602e-05,
+      "loss": 0.5383,
+      "step": 527
+    },
+    {
+      "epoch": 0.11777827347758198,
+      "grad_norm": 0.18771792948246002,
+      "learning_rate": 1.9949549842660495e-05,
+      "loss": 0.5273,
+      "step": 528
+    },
+    {
+      "epoch": 0.11800133838947134,
+      "grad_norm": 0.1931321620941162,
+      "learning_rate": 1.9949313465516312e-05,
+      "loss": 0.5692,
+      "step": 529
+    },
+    {
+      "epoch": 0.11822440330136069,
+      "grad_norm": 0.18806159496307373,
+      "learning_rate": 1.9949076537316566e-05,
+      "loss": 0.5375,
+      "step": 530
+    },
+    {
+      "epoch": 0.11844746821325006,
+      "grad_norm": 0.18746712803840637,
+      "learning_rate": 1.9948839058074383e-05,
+      "loss": 0.5418,
+      "step": 531
+    },
+    {
+      "epoch": 0.11867053312513942,
+      "grad_norm": 0.18869024515151978,
+      "learning_rate": 1.9948601027802908e-05,
+      "loss": 0.5059,
+      "step": 532
+    },
+    {
+      "epoch": 0.11889359803702877,
+      "grad_norm": 0.20482954382896423,
+      "learning_rate": 1.994836244651533e-05,
+      "loss": 0.5359,
+      "step": 533
+    },
+    {
+      "epoch": 0.11911666294891814,
+      "grad_norm": 0.17584972083568573,
+      "learning_rate": 1.9948123314224862e-05,
+      "loss": 0.5221,
+      "step": 534
+    },
+    {
+      "epoch": 0.1193397278608075,
+      "grad_norm": 0.1878448724746704,
+      "learning_rate": 1.994788363094475e-05,
+      "loss": 0.5358,
+      "step": 535
+    },
+    {
+      "epoch": 0.11956279277269685,
+      "grad_norm": 0.17549312114715576,
+      "learning_rate": 1.9947643396688266e-05,
+      "loss": 0.5391,
+      "step": 536
+    },
+    {
+      "epoch": 0.11978585768458622,
+      "grad_norm": 0.18577958643436432,
+      "learning_rate": 1.9947402611468714e-05,
+      "loss": 0.5606,
+      "step": 537
+    },
+    {
+      "epoch": 0.12000892259647558,
+      "grad_norm": 0.17562542855739594,
+      "learning_rate": 1.994716127529944e-05,
+      "loss": 0.5186,
+      "step": 538
+    },
+    {
+      "epoch": 0.12023198750836493,
+      "grad_norm": 0.18019866943359375,
+      "learning_rate": 1.9946919388193803e-05,
+      "loss": 0.524,
+      "step": 539
+    },
+    {
+      "epoch": 0.1204550524202543,
+      "grad_norm": 0.18850797414779663,
+      "learning_rate": 1.99466769501652e-05,
+      "loss": 0.5496,
+      "step": 540
+    },
+    {
+      "epoch": 0.12067811733214365,
+      "grad_norm": 0.1817145198583603,
+      "learning_rate": 1.9946433961227062e-05,
+      "loss": 0.5303,
+      "step": 541
+    },
+    {
+      "epoch": 0.12090118224403301,
+      "grad_norm": 0.2111491858959198,
+      "learning_rate": 1.9946190421392845e-05,
+      "loss": 0.5526,
+      "step": 542
+    },
+    {
+      "epoch": 0.12112424715592238,
+      "grad_norm": 0.294716477394104,
+      "learning_rate": 1.9945946330676036e-05,
+      "loss": 0.5172,
+      "step": 543
+    },
+    {
+      "epoch": 0.12134731206781173,
+      "grad_norm": 0.21202930808067322,
+      "learning_rate": 1.994570168909016e-05,
+      "loss": 0.5295,
+      "step": 544
+    },
+    {
+      "epoch": 0.12157037697970109,
+      "grad_norm": 0.18554560840129852,
+      "learning_rate": 1.9945456496648763e-05,
+      "loss": 0.5347,
+      "step": 545
+    },
+    {
+      "epoch": 0.12179344189159046,
+      "grad_norm": 0.18445701897144318,
+      "learning_rate": 1.9945210753365426e-05,
+      "loss": 0.5433,
+      "step": 546
+    },
+    {
+      "epoch": 0.12201650680347981,
+      "grad_norm": 0.2190207540988922,
+      "learning_rate": 1.9944964459253757e-05,
+      "loss": 0.5342,
+      "step": 547
+    },
+    {
+      "epoch": 0.12223957171536917,
+      "grad_norm": 0.19571848213672638,
+      "learning_rate": 1.99447176143274e-05,
+      "loss": 0.5675,
+      "step": 548
+    },
+    {
+      "epoch": 0.12246263662725854,
+      "grad_norm": 0.1853405386209488,
+      "learning_rate": 1.994447021860003e-05,
+      "loss": 0.5857,
+      "step": 549
+    },
+    {
+      "epoch": 0.12268570153914789,
+      "grad_norm": 0.19547483325004578,
+      "learning_rate": 1.9944222272085344e-05,
+      "loss": 0.5208,
+      "step": 550
+    },
+    {
+      "epoch": 0.12290876645103725,
+      "grad_norm": 0.20552557706832886,
+      "learning_rate": 1.994397377479708e-05,
+      "loss": 0.5518,
+      "step": 551
+    },
+    {
+      "epoch": 0.12313183136292662,
+      "grad_norm": 0.21979011595249176,
+      "learning_rate": 1.9943724726748996e-05,
+      "loss": 0.5658,
+      "step": 552
+    },
+    {
+      "epoch": 0.12335489627481597,
+      "grad_norm": 0.18478159606456757,
+      "learning_rate": 1.994347512795489e-05,
+      "loss": 0.5599,
+      "step": 553
+    },
+    {
+      "epoch": 0.12357796118670533,
+      "grad_norm": 0.20098286867141724,
+      "learning_rate": 1.9943224978428582e-05,
+      "loss": 0.5455,
+      "step": 554
+    },
+    {
+      "epoch": 0.1238010260985947,
+      "grad_norm": 0.2217499017715454,
+      "learning_rate": 1.994297427818393e-05,
+      "loss": 0.5512,
+      "step": 555
+    },
+    {
+      "epoch": 0.12402409101048405,
+      "grad_norm": 0.17720763385295868,
+      "learning_rate": 1.9942723027234817e-05,
+      "loss": 0.5295,
+      "step": 556
+    },
+    {
+      "epoch": 0.12424715592237341,
+      "grad_norm": 0.1762504130601883,
+      "learning_rate": 1.9942471225595162e-05,
+      "loss": 0.5361,
+      "step": 557
+    },
+    {
+      "epoch": 0.12447022083426278,
+      "grad_norm": 0.2265862226486206,
+      "learning_rate": 1.994221887327891e-05,
+      "loss": 0.5669,
+      "step": 558
+    },
+    {
+      "epoch": 0.12469328574615213,
+      "grad_norm": 0.19906531274318695,
+      "learning_rate": 1.994196597030004e-05,
+      "loss": 0.5302,
+      "step": 559
+    },
+    {
+      "epoch": 0.12491635065804149,
+      "grad_norm": 0.17008185386657715,
+      "learning_rate": 1.9941712516672553e-05,
+      "loss": 0.5333,
+      "step": 560
+    },
+    {
+      "epoch": 0.12513941556993086,
+      "grad_norm": 0.18816471099853516,
+      "learning_rate": 1.994145851241049e-05,
+      "loss": 0.5679,
+      "step": 561
+    },
+    {
+      "epoch": 0.12536248048182022,
+      "grad_norm": 0.19027023017406464,
+      "learning_rate": 1.9941203957527927e-05,
+      "loss": 0.5507,
+      "step": 562
+    },
+    {
+      "epoch": 0.12558554539370956,
+      "grad_norm": 0.17551641166210175,
+      "learning_rate": 1.994094885203895e-05,
+      "loss": 0.5311,
+      "step": 563
+    },
+    {
+      "epoch": 0.12580861030559892,
+      "grad_norm": 0.1766747236251831,
+      "learning_rate": 1.9940693195957696e-05,
+      "loss": 0.5279,
+      "step": 564
+    },
+    {
+      "epoch": 0.1260316752174883,
+      "grad_norm": 0.1826571673154831,
+      "learning_rate": 1.9940436989298322e-05,
+      "loss": 0.5195,
+      "step": 565
+    },
+    {
+      "epoch": 0.12625474012937765,
+      "grad_norm": 0.18142053484916687,
+      "learning_rate": 1.9940180232075025e-05,
+      "loss": 0.5678,
+      "step": 566
+    },
+    {
+      "epoch": 0.12647780504126702,
+      "grad_norm": 0.21495957672595978,
+      "learning_rate": 1.993992292430202e-05,
+      "loss": 0.5393,
+      "step": 567
+    },
+    {
+      "epoch": 0.12670086995315638,
+      "grad_norm": 0.18905828893184662,
+      "learning_rate": 1.9939665065993556e-05,
+      "loss": 0.5507,
+      "step": 568
+    },
+    {
+      "epoch": 0.12692393486504572,
+      "grad_norm": 0.19366420805454254,
+      "learning_rate": 1.9939406657163916e-05,
+      "loss": 0.5458,
+      "step": 569
+    },
+    {
+      "epoch": 0.12714699977693508,
+      "grad_norm": 0.20456644892692566,
+      "learning_rate": 1.9939147697827415e-05,
+      "loss": 0.5532,
+      "step": 570
+    },
+    {
+      "epoch": 0.12737006468882445,
+      "grad_norm": 0.20060132443904877,
+      "learning_rate": 1.9938888187998397e-05,
+      "loss": 0.5865,
+      "step": 571
+    },
+    {
+      "epoch": 0.1275931296007138,
+      "grad_norm": 0.18896770477294922,
+      "learning_rate": 1.9938628127691232e-05,
+      "loss": 0.5372,
+      "step": 572
+    },
+    {
+      "epoch": 0.12781619451260318,
+      "grad_norm": 0.1824081540107727,
+      "learning_rate": 1.9938367516920323e-05,
+      "loss": 0.5232,
+      "step": 573
+    },
+    {
+      "epoch": 0.12803925942449254,
+      "grad_norm": 0.4999215602874756,
+      "learning_rate": 1.993810635570011e-05,
+      "loss": 0.5323,
+      "step": 574
+    },
+    {
+      "epoch": 0.12826232433638188,
+      "grad_norm": 0.23097532987594604,
+      "learning_rate": 1.993784464404505e-05,
+      "loss": 0.544,
+      "step": 575
+    },
+    {
+      "epoch": 0.12848538924827124,
+      "grad_norm": 0.1975661814212799,
+      "learning_rate": 1.993758238196964e-05,
+      "loss": 0.5411,
+      "step": 576
+    },
+    {
+      "epoch": 0.1287084541601606,
+      "grad_norm": 0.18706879019737244,
+      "learning_rate": 1.9937319569488414e-05,
+      "loss": 0.5386,
+      "step": 577
+    },
+    {
+      "epoch": 0.12893151907204997,
+      "grad_norm": 0.19232505559921265,
+      "learning_rate": 1.993705620661592e-05,
+      "loss": 0.5327,
+      "step": 578
+    },
+    {
+      "epoch": 0.12915458398393934,
+      "grad_norm": 0.20714089274406433,
+      "learning_rate": 1.9936792293366744e-05,
+      "loss": 0.5381,
+      "step": 579
+    },
+    {
+      "epoch": 0.12937764889582867,
+      "grad_norm": 0.18375153839588165,
+      "learning_rate": 1.993652782975551e-05,
+      "loss": 0.5664,
+      "step": 580
+    },
+    {
+      "epoch": 0.12960071380771804,
+      "grad_norm": 0.19149671494960785,
+      "learning_rate": 1.993626281579686e-05,
+      "loss": 0.5265,
+      "step": 581
+    },
+    {
+      "epoch": 0.1298237787196074,
+      "grad_norm": 0.19362886250019073,
+      "learning_rate": 1.9935997251505473e-05,
+      "loss": 0.517,
+      "step": 582
+    },
+    {
+      "epoch": 0.13004684363149677,
+      "grad_norm": 0.19237902760505676,
+      "learning_rate": 1.993573113689606e-05,
+      "loss": 0.529,
+      "step": 583
+    },
+    {
+      "epoch": 0.13026990854338613,
+      "grad_norm": 0.2098451405763626,
+      "learning_rate": 1.9935464471983354e-05,
+      "loss": 0.536,
+      "step": 584
+    },
+    {
+      "epoch": 0.1304929734552755,
+      "grad_norm": 0.17832371592521667,
+      "learning_rate": 1.993519725678213e-05,
+      "loss": 0.528,
+      "step": 585
+    },
+    {
+      "epoch": 0.13071603836716483,
+      "grad_norm": 0.19082553684711456,
+      "learning_rate": 1.9934929491307194e-05,
+      "loss": 0.5208,
+      "step": 586
+    },
+    {
+      "epoch": 0.1309391032790542,
+      "grad_norm": 0.19741079211235046,
+      "learning_rate": 1.9934661175573363e-05,
+      "loss": 0.5605,
+      "step": 587
+    },
+    {
+      "epoch": 0.13116216819094356,
+      "grad_norm": 0.17371487617492676,
+      "learning_rate": 1.9934392309595504e-05,
+      "loss": 0.5462,
+      "step": 588
+    },
+    {
+      "epoch": 0.13138523310283293,
+      "grad_norm": 0.17552132904529572,
+      "learning_rate": 1.9934122893388512e-05,
+      "loss": 0.5343,
+      "step": 589
+    },
+    {
+      "epoch": 0.1316082980147223,
+      "grad_norm": 0.19698654115200043,
+      "learning_rate": 1.9933852926967305e-05,
+      "loss": 0.5711,
+      "step": 590
+    },
+    {
+      "epoch": 0.13183136292661166,
+      "grad_norm": 0.19045893847942352,
+      "learning_rate": 1.993358241034684e-05,
+      "loss": 0.5601,
+      "step": 591
+    },
+    {
+      "epoch": 0.132054427838501,
+      "grad_norm": 0.17538850009441376,
+      "learning_rate": 1.9933311343542094e-05,
+      "loss": 0.5403,
+      "step": 592
+    },
+    {
+      "epoch": 0.13227749275039036,
+      "grad_norm": 0.1708664894104004,
+      "learning_rate": 1.9933039726568078e-05,
+      "loss": 0.5346,
+      "step": 593
+    },
+    {
+      "epoch": 0.13250055766227972,
+      "grad_norm": 0.1795656532049179,
+      "learning_rate": 1.9932767559439844e-05,
+      "loss": 0.5516,
+      "step": 594
+    },
+    {
+      "epoch": 0.1327236225741691,
+      "grad_norm": 0.18116529285907745,
+      "learning_rate": 1.9932494842172465e-05,
+      "loss": 0.5311,
+      "step": 595
+    },
+    {
+      "epoch": 0.13294668748605845,
+      "grad_norm": 0.20366770029067993,
+      "learning_rate": 1.9932221574781043e-05,
+      "loss": 0.5546,
+      "step": 596
+    },
+    {
+      "epoch": 0.13316975239794782,
+      "grad_norm": 0.1805567592382431,
+      "learning_rate": 1.9931947757280713e-05,
+      "loss": 0.5754,
+      "step": 597
+    },
+    {
+      "epoch": 0.13339281730983715,
+      "grad_norm": 0.1815255880355835,
+      "learning_rate": 1.9931673389686642e-05,
+      "loss": 0.5172,
+      "step": 598
+    },
+    {
+      "epoch": 0.13361588222172652,
+      "grad_norm": 0.1993352770805359,
+      "learning_rate": 1.9931398472014024e-05,
+      "loss": 0.5483,
+      "step": 599
+    },
+    {
+      "epoch": 0.13383894713361588,
+      "grad_norm": 0.1812392920255661,
+      "learning_rate": 1.993112300427809e-05,
+      "loss": 0.5418,
+      "step": 600
+    },
+    {
+      "epoch": 0.13406201204550525,
+      "grad_norm": 0.23115472495555878,
+      "learning_rate": 1.9930846986494098e-05,
+      "loss": 0.5661,
+      "step": 601
+    },
+    {
+      "epoch": 0.1342850769573946,
+      "grad_norm": 0.1799275279045105,
+      "learning_rate": 1.9930570418677327e-05,
+      "loss": 0.5536,
+      "step": 602
+    },
+    {
+      "epoch": 0.13450814186928395,
+      "grad_norm": 0.19140706956386566,
+      "learning_rate": 1.9930293300843103e-05,
+      "loss": 0.5405,
+      "step": 603
+    },
+    {
+      "epoch": 0.1347312067811733,
+      "grad_norm": 0.20953412353992462,
+      "learning_rate": 1.993001563300677e-05,
+      "loss": 0.5215,
+      "step": 604
+    },
+    {
+      "epoch": 0.13495427169306268,
+      "grad_norm": 0.18991845846176147,
+      "learning_rate": 1.992973741518371e-05,
+      "loss": 0.5577,
+      "step": 605
+    },
+    {
+      "epoch": 0.13517733660495204,
+      "grad_norm": 0.18286292254924774,
+      "learning_rate": 1.9929458647389333e-05,
+      "loss": 0.5531,
+      "step": 606
+    },
+    {
+      "epoch": 0.1354004015168414,
+      "grad_norm": 0.1848326176404953,
+      "learning_rate": 1.9929179329639075e-05,
+      "loss": 0.5387,
+      "step": 607
+    },
+    {
+      "epoch": 0.13562346642873077,
+      "grad_norm": 0.17560309171676636,
+      "learning_rate": 1.9928899461948407e-05,
+      "loss": 0.5444,
+      "step": 608
+    },
+    {
+      "epoch": 0.1358465313406201,
+      "grad_norm": 0.18911704421043396,
+      "learning_rate": 1.9928619044332837e-05,
+      "loss": 0.506,
+      "step": 609
+    },
+    {
+      "epoch": 0.13606959625250947,
+      "grad_norm": 0.19542188942432404,
+      "learning_rate": 1.9928338076807888e-05,
+      "loss": 0.5222,
+      "step": 610
+    },
+    {
+      "epoch": 0.13629266116439884,
+      "grad_norm": 0.16921466588974,
+      "learning_rate": 1.9928056559389123e-05,
+      "loss": 0.5095,
+      "step": 611
+    },
+    {
+      "epoch": 0.1365157260762882,
+      "grad_norm": 0.19137442111968994,
+      "learning_rate": 1.9927774492092137e-05,
+      "loss": 0.5504,
+      "step": 612
+    },
+    {
+      "epoch": 0.13673879098817757,
+      "grad_norm": 0.2050313800573349,
+      "learning_rate": 1.9927491874932553e-05,
+      "loss": 0.5695,
+      "step": 613
+    },
+    {
+      "epoch": 0.13696185590006693,
+      "grad_norm": 0.18225258588790894,
+      "learning_rate": 1.992720870792602e-05,
+      "loss": 0.5286,
+      "step": 614
+    },
+    {
+      "epoch": 0.13718492081195627,
+      "grad_norm": 0.18813620507717133,
+      "learning_rate": 1.9926924991088226e-05,
+      "loss": 0.5362,
+      "step": 615
+    },
+    {
+      "epoch": 0.13740798572384563,
+      "grad_norm": 0.21695192158222198,
+      "learning_rate": 1.9926640724434882e-05,
+      "loss": 0.5548,
+      "step": 616
+    },
+    {
+      "epoch": 0.137631050635735,
+      "grad_norm": 0.1780216544866562,
+      "learning_rate": 1.9926355907981735e-05,
+      "loss": 0.547,
+      "step": 617
+    },
+    {
+      "epoch": 0.13785411554762436,
+      "grad_norm": 0.182960644364357,
+      "learning_rate": 1.9926070541744557e-05,
+      "loss": 0.524,
+      "step": 618
+    },
+    {
+      "epoch": 0.13807718045951373,
+      "grad_norm": 0.18580903112888336,
+      "learning_rate": 1.9925784625739157e-05,
+      "loss": 0.5515,
+      "step": 619
+    },
+    {
+      "epoch": 0.1383002453714031,
+      "grad_norm": 0.17631803452968597,
+      "learning_rate": 1.9925498159981368e-05,
+      "loss": 0.5269,
+      "step": 620
+    },
+    {
+      "epoch": 0.13852331028329243,
+      "grad_norm": 0.1774953305721283,
+      "learning_rate": 1.9925211144487057e-05,
+      "loss": 0.5279,
+      "step": 621
+    },
+    {
+      "epoch": 0.1387463751951818,
+      "grad_norm": 0.1902550458908081,
+      "learning_rate": 1.992492357927212e-05,
+      "loss": 0.5199,
+      "step": 622
+    },
+    {
+      "epoch": 0.13896944010707116,
+      "grad_norm": 0.18148526549339294,
+      "learning_rate": 1.9924635464352486e-05,
+      "loss": 0.5273,
+      "step": 623
+    },
+    {
+      "epoch": 0.13919250501896052,
+      "grad_norm": 0.18067635595798492,
+      "learning_rate": 1.9924346799744108e-05,
+      "loss": 0.5431,
+      "step": 624
+    },
+    {
+      "epoch": 0.1394155699308499,
+      "grad_norm": 0.1752486228942871,
+      "learning_rate": 1.992405758546298e-05,
+      "loss": 0.5012,
+      "step": 625
+    },
+    {
+      "epoch": 0.13963863484273925,
+      "grad_norm": 0.18866072595119476,
+      "learning_rate": 1.992376782152512e-05,
+      "loss": 0.5408,
+      "step": 626
+    },
+    {
+      "epoch": 0.1398616997546286,
+      "grad_norm": 0.21433256566524506,
+      "learning_rate": 1.9923477507946573e-05,
+      "loss": 0.5625,
+      "step": 627
+    },
+    {
+      "epoch": 0.14008476466651795,
+      "grad_norm": 0.17377375066280365,
+      "learning_rate": 1.9923186644743425e-05,
+      "loss": 0.516,
+      "step": 628
+    },
+    {
+      "epoch": 0.14030782957840732,
+      "grad_norm": 0.19306115806102753,
+      "learning_rate": 1.9922895231931775e-05,
+      "loss": 0.5397,
+      "step": 629
+    },
+    {
+      "epoch": 0.14053089449029668,
+      "grad_norm": 0.18144343793392181,
+      "learning_rate": 1.992260326952777e-05,
+      "loss": 0.5215,
+      "step": 630
+    },
+    {
+      "epoch": 0.14075395940218605,
+      "grad_norm": 0.19684460759162903,
+      "learning_rate": 1.9922310757547584e-05,
+      "loss": 0.5472,
+      "step": 631
+    },
+    {
+      "epoch": 0.14097702431407538,
+      "grad_norm": 0.18824423849582672,
+      "learning_rate": 1.9922017696007413e-05,
+      "loss": 0.5595,
+      "step": 632
+    },
+    {
+      "epoch": 0.14120008922596475,
+      "grad_norm": 0.18633319437503815,
+      "learning_rate": 1.992172408492349e-05,
+      "loss": 0.568,
+      "step": 633
+    },
+    {
+      "epoch": 0.1414231541378541,
+      "grad_norm": 0.18547959625720978,
+      "learning_rate": 1.9921429924312074e-05,
+      "loss": 0.5421,
+      "step": 634
+    },
+    {
+      "epoch": 0.14164621904974348,
+      "grad_norm": 0.16735389828681946,
+      "learning_rate": 1.9921135214189466e-05,
+      "loss": 0.499,
+      "step": 635
+    },
+    {
+      "epoch": 0.14186928396163284,
+      "grad_norm": 0.1912708729505539,
+      "learning_rate": 1.992083995457198e-05,
+      "loss": 0.5655,
+      "step": 636
+    },
+    {
+      "epoch": 0.1420923488735222,
+      "grad_norm": 0.19804726541042328,
+      "learning_rate": 1.9920544145475975e-05,
+      "loss": 0.5639,
+      "step": 637
+    },
+    {
+      "epoch": 0.14231541378541154,
+      "grad_norm": 0.17582711577415466,
+      "learning_rate": 1.992024778691783e-05,
+      "loss": 0.5577,
+      "step": 638
+    },
+    {
+      "epoch": 0.1425384786973009,
+      "grad_norm": 0.18506725132465363,
+      "learning_rate": 1.9919950878913962e-05,
+      "loss": 0.5287,
+      "step": 639
+    },
+    {
+      "epoch": 0.14276154360919027,
+      "grad_norm": 0.1914191097021103,
+      "learning_rate": 1.9919653421480816e-05,
+      "loss": 0.5125,
+      "step": 640
+    },
+    {
+      "epoch": 0.14298460852107964,
+      "grad_norm": 0.17707392573356628,
+      "learning_rate": 1.9919355414634864e-05,
+      "loss": 0.5351,
+      "step": 641
+    },
+    {
+      "epoch": 0.143207673432969,
+      "grad_norm": 0.1675226092338562,
+      "learning_rate": 1.9919056858392618e-05,
+      "loss": 0.5408,
+      "step": 642
+    },
+    {
+      "epoch": 0.14343073834485837,
+      "grad_norm": 0.18883782625198364,
+      "learning_rate": 1.9918757752770607e-05,
+      "loss": 0.5387,
+      "step": 643
+    },
+    {
+      "epoch": 0.1436538032567477,
+      "grad_norm": 0.18036174774169922,
+      "learning_rate": 1.99184580977854e-05,
+      "loss": 0.5444,
+      "step": 644
+    },
+    {
+      "epoch": 0.14387686816863707,
+      "grad_norm": 0.1817016750574112,
+      "learning_rate": 1.99181578934536e-05,
+      "loss": 0.527,
+      "step": 645
+    },
+    {
+      "epoch": 0.14409993308052643,
+      "grad_norm": 0.18899092078208923,
+      "learning_rate": 1.991785713979182e-05,
+      "loss": 0.5245,
+      "step": 646
+    },
+    {
+      "epoch": 0.1443229979924158,
+      "grad_norm": 0.18643809854984283,
+      "learning_rate": 1.991755583681673e-05,
+      "loss": 0.5247,
+      "step": 647
+    },
+    {
+      "epoch": 0.14454606290430516,
+      "grad_norm": 0.1903519332408905,
+      "learning_rate": 1.9917253984545014e-05,
+      "loss": 0.5497,
+      "step": 648
+    },
+    {
+      "epoch": 0.14476912781619453,
+      "grad_norm": 0.1695372760295868,
+      "learning_rate": 1.991695158299339e-05,
+      "loss": 0.5535,
+      "step": 649
+    },
+    {
+      "epoch": 0.14499219272808386,
+      "grad_norm": 0.19261977076530457,
+      "learning_rate": 1.9916648632178605e-05,
+      "loss": 0.5324,
+      "step": 650
+    },
+    {
+      "epoch": 0.14521525763997323,
+      "grad_norm": 0.1795564889907837,
+      "learning_rate": 1.9916345132117442e-05,
+      "loss": 0.5316,
+      "step": 651
+    },
+    {
+      "epoch": 0.1454383225518626,
+      "grad_norm": 0.27002382278442383,
+      "learning_rate": 1.9916041082826713e-05,
+      "loss": 0.5521,
+      "step": 652
+    },
+    {
+      "epoch": 0.14566138746375196,
+      "grad_norm": 0.17605867981910706,
+      "learning_rate": 1.9915736484323246e-05,
+      "loss": 0.5228,
+      "step": 653
+    },
+    {
+      "epoch": 0.14588445237564132,
+      "grad_norm": 0.1808420866727829,
+      "learning_rate": 1.9915431336623928e-05,
+      "loss": 0.5593,
+      "step": 654
+    },
+    {
+      "epoch": 0.14610751728753066,
+      "grad_norm": 0.189579576253891,
+      "learning_rate": 1.991512563974565e-05,
+      "loss": 0.5405,
+      "step": 655
+    },
+    {
+      "epoch": 0.14633058219942002,
+      "grad_norm": 0.1883007138967514,
+      "learning_rate": 1.9914819393705342e-05,
+      "loss": 0.5571,
+      "step": 656
+    },
+    {
+      "epoch": 0.1465536471113094,
+      "grad_norm": 0.18864606320858002,
+      "learning_rate": 1.9914512598519972e-05,
+      "loss": 0.519,
+      "step": 657
+    },
+    {
+      "epoch": 0.14677671202319875,
+      "grad_norm": 0.17059919238090515,
+      "learning_rate": 1.9914205254206527e-05,
+      "loss": 0.533,
+      "step": 658
+    },
+    {
+      "epoch": 0.14699977693508812,
+      "grad_norm": 0.4270278215408325,
+      "learning_rate": 1.9913897360782036e-05,
+      "loss": 0.5282,
+      "step": 659
+    },
+    {
+      "epoch": 0.14722284184697748,
+      "grad_norm": 0.18411633372306824,
+      "learning_rate": 1.9913588918263545e-05,
+      "loss": 0.5128,
+      "step": 660
+    },
+    {
+      "epoch": 0.14744590675886682,
+      "grad_norm": 0.2003640979528427,
+      "learning_rate": 1.9913279926668146e-05,
+      "loss": 0.5213,
+      "step": 661
+    },
+    {
+      "epoch": 0.14766897167075618,
+      "grad_norm": 0.18532606959342957,
+      "learning_rate": 1.9912970386012943e-05,
+      "loss": 0.498,
+      "step": 662
+    },
+    {
+      "epoch": 0.14789203658264555,
+      "grad_norm": 0.1674346923828125,
+      "learning_rate": 1.9912660296315083e-05,
+      "loss": 0.5383,
+      "step": 663
+    },
+    {
+      "epoch": 0.1481151014945349,
+      "grad_norm": 0.19120754301548004,
+      "learning_rate": 1.9912349657591748e-05,
+      "loss": 0.5301,
+      "step": 664
+    },
+    {
+      "epoch": 0.14833816640642428,
+      "grad_norm": 0.178116112947464,
+      "learning_rate": 1.9912038469860135e-05,
+      "loss": 0.5583,
+      "step": 665
+    },
+    {
+      "epoch": 0.14856123131831364,
+      "grad_norm": 0.1826133131980896,
+      "learning_rate": 1.9911726733137484e-05,
+      "loss": 0.5188,
+      "step": 666
+    },
+    {
+      "epoch": 0.14878429623020298,
+      "grad_norm": 0.18221695721149445,
+      "learning_rate": 1.991141444744106e-05,
+      "loss": 0.5236,
+      "step": 667
+    },
+    {
+      "epoch": 0.14900736114209234,
+      "grad_norm": 0.18043336272239685,
+      "learning_rate": 1.9911101612788157e-05,
+      "loss": 0.5201,
+      "step": 668
+    },
+    {
+      "epoch": 0.1492304260539817,
+      "grad_norm": 0.19176125526428223,
+      "learning_rate": 1.9910788229196104e-05,
+      "loss": 0.5501,
+      "step": 669
+    },
+    {
+      "epoch": 0.14945349096587107,
+      "grad_norm": 0.18294784426689148,
+      "learning_rate": 1.9910474296682256e-05,
+      "loss": 0.5304,
+      "step": 670
+    },
+    {
+      "epoch": 0.14967655587776044,
+      "grad_norm": 0.19021780788898468,
+      "learning_rate": 1.9910159815264e-05,
+      "loss": 0.5561,
+      "step": 671
+    },
+    {
+      "epoch": 0.1498996207896498,
+      "grad_norm": 0.18545284867286682,
+      "learning_rate": 1.9909844784958762e-05,
+      "loss": 0.536,
+      "step": 672
+    },
+    {
+      "epoch": 0.15012268570153914,
+      "grad_norm": 0.2078470140695572,
+      "learning_rate": 1.990952920578398e-05,
+      "loss": 0.513,
+      "step": 673
+    },
+    {
+      "epoch": 0.1503457506134285,
+      "grad_norm": 0.20093394815921783,
+      "learning_rate": 1.9909213077757138e-05,
+      "loss": 0.5406,
+      "step": 674
+    },
+    {
+      "epoch": 0.15056881552531787,
+      "grad_norm": 0.212370365858078,
+      "learning_rate": 1.9908896400895745e-05,
+      "loss": 0.5108,
+      "step": 675
+    },
+    {
+      "epoch": 0.15079188043720723,
+      "grad_norm": 0.1903519630432129,
+      "learning_rate": 1.990857917521734e-05,
+      "loss": 0.5163,
+      "step": 676
+    },
+    {
+      "epoch": 0.1510149453490966,
+      "grad_norm": 0.18388496339321136,
+      "learning_rate": 1.9908261400739494e-05,
+      "loss": 0.5361,
+      "step": 677
+    },
+    {
+      "epoch": 0.15123801026098593,
+      "grad_norm": 0.1817573606967926,
+      "learning_rate": 1.9907943077479802e-05,
+      "loss": 0.5582,
+      "step": 678
+    },
+    {
+      "epoch": 0.1514610751728753,
+      "grad_norm": 0.19696584343910217,
+      "learning_rate": 1.9907624205455903e-05,
+      "loss": 0.5326,
+      "step": 679
+    },
+    {
+      "epoch": 0.15168414008476466,
+      "grad_norm": 0.17569434642791748,
+      "learning_rate": 1.990730478468545e-05,
+      "loss": 0.5073,
+      "step": 680
+    },
+    {
+      "epoch": 0.15190720499665403,
+      "grad_norm": 0.1812848299741745,
+      "learning_rate": 1.9906984815186142e-05,
+      "loss": 0.539,
+      "step": 681
+    },
+    {
+      "epoch": 0.1521302699085434,
+      "grad_norm": 0.1694687008857727,
+      "learning_rate": 1.9906664296975696e-05,
+      "loss": 0.5286,
+      "step": 682
+    },
+    {
+      "epoch": 0.15235333482043276,
+      "grad_norm": 0.18631073832511902,
+      "learning_rate": 1.990634323007187e-05,
+      "loss": 0.5807,
+      "step": 683
+    },
+    {
+      "epoch": 0.1525763997323221,
+      "grad_norm": 0.18517658114433289,
+      "learning_rate": 1.9906021614492438e-05,
+      "loss": 0.5393,
+      "step": 684
+    },
+    {
+      "epoch": 0.15279946464421146,
+      "grad_norm": 0.19088414311408997,
+      "learning_rate": 1.990569945025522e-05,
+      "loss": 0.5583,
+      "step": 685
+    },
+    {
+      "epoch": 0.15302252955610082,
+      "grad_norm": 0.1954737901687622,
+      "learning_rate": 1.9905376737378056e-05,
+      "loss": 0.5348,
+      "step": 686
+    },
+    {
+      "epoch": 0.1532455944679902,
+      "grad_norm": 0.1780342310667038,
+      "learning_rate": 1.990505347587882e-05,
+      "loss": 0.5083,
+      "step": 687
+    },
+    {
+      "epoch": 0.15346865937987955,
+      "grad_norm": 0.18818046152591705,
+      "learning_rate": 1.9904729665775417e-05,
+      "loss": 0.51,
+      "step": 688
+    },
+    {
+      "epoch": 0.15369172429176892,
+      "grad_norm": 0.1797480434179306,
+      "learning_rate": 1.990440530708578e-05,
+      "loss": 0.5114,
+      "step": 689
+    },
+    {
+      "epoch": 0.15391478920365825,
+      "grad_norm": 0.18315070867538452,
+      "learning_rate": 1.9904080399827883e-05,
+      "loss": 0.5322,
+      "step": 690
+    },
+    {
+      "epoch": 0.15413785411554762,
+      "grad_norm": 0.2024601846933365,
+      "learning_rate": 1.990375494401971e-05,
+      "loss": 0.5198,
+      "step": 691
+    },
+    {
+      "epoch": 0.15436091902743698,
+      "grad_norm": 0.2151622325181961,
+      "learning_rate": 1.990342893967929e-05,
+      "loss": 0.5308,
+      "step": 692
+    },
+    {
+      "epoch": 0.15458398393932635,
+      "grad_norm": 0.1977054476737976,
+      "learning_rate": 1.990310238682468e-05,
+      "loss": 0.579,
+      "step": 693
+    },
+    {
+      "epoch": 0.1548070488512157,
+      "grad_norm": 0.1899496465921402,
+      "learning_rate": 1.990277528547397e-05,
+      "loss": 0.5435,
+      "step": 694
+    },
+    {
+      "epoch": 0.15503011376310508,
+      "grad_norm": 0.18425041437149048,
+      "learning_rate": 1.9902447635645273e-05,
+      "loss": 0.5582,
+      "step": 695
+    },
+    {
+      "epoch": 0.1552531786749944,
+      "grad_norm": 0.1868348866701126,
+      "learning_rate": 1.9902119437356737e-05,
+      "loss": 0.5208,
+      "step": 696
+    },
+    {
+      "epoch": 0.15547624358688378,
+      "grad_norm": 0.19318504631519318,
+      "learning_rate": 1.990179069062654e-05,
+      "loss": 0.581,
+      "step": 697
+    },
+    {
+      "epoch": 0.15569930849877314,
+      "grad_norm": 0.17516325414180756,
+      "learning_rate": 1.990146139547289e-05,
+      "loss": 0.522,
+      "step": 698
+    },
+    {
+      "epoch": 0.1559223734106625,
+      "grad_norm": 0.18644174933433533,
+      "learning_rate": 1.990113155191402e-05,
+      "loss": 0.5361,
+      "step": 699
+    },
+    {
+      "epoch": 0.15614543832255187,
+      "grad_norm": 0.204257071018219,
+      "learning_rate": 1.9900801159968207e-05,
+      "loss": 0.5216,
+      "step": 700
+    },
+    {
+      "epoch": 0.15636850323444124,
+      "grad_norm": 0.19151191413402557,
+      "learning_rate": 1.990047021965375e-05,
+      "loss": 0.5476,
+      "step": 701
+    },
+    {
+      "epoch": 0.15659156814633057,
+      "grad_norm": 0.18692530691623688,
+      "learning_rate": 1.9900138730988976e-05,
+      "loss": 0.5516,
+      "step": 702
+    },
+    {
+      "epoch": 0.15681463305821994,
+      "grad_norm": 0.19439953565597534,
+      "learning_rate": 1.9899806693992242e-05,
+      "loss": 0.5404,
+      "step": 703
+    },
+    {
+      "epoch": 0.1570376979701093,
+      "grad_norm": 0.1809748411178589,
+      "learning_rate": 1.989947410868194e-05,
+      "loss": 0.5571,
+      "step": 704
+    },
+    {
+      "epoch": 0.15726076288199867,
+      "grad_norm": 0.19498126208782196,
+      "learning_rate": 1.9899140975076495e-05,
+      "loss": 0.5121,
+      "step": 705
+    },
+    {
+      "epoch": 0.15748382779388803,
+      "grad_norm": 0.19838570058345795,
+      "learning_rate": 1.9898807293194352e-05,
+      "loss": 0.5535,
+      "step": 706
+    },
+    {
+      "epoch": 0.15770689270577737,
+      "grad_norm": 0.18156638741493225,
+      "learning_rate": 1.9898473063054e-05,
+      "loss": 0.5275,
+      "step": 707
+    },
+    {
+      "epoch": 0.15792995761766673,
+      "grad_norm": 0.1888674944639206,
+      "learning_rate": 1.989813828467394e-05,
+      "loss": 0.5274,
+      "step": 708
+    },
+    {
+      "epoch": 0.1581530225295561,
+      "grad_norm": 0.2016175538301468,
+      "learning_rate": 1.9897802958072722e-05,
+      "loss": 0.5057,
+      "step": 709
+    },
+    {
+      "epoch": 0.15837608744144546,
+      "grad_norm": 0.21129021048545837,
+      "learning_rate": 1.989746708326892e-05,
+      "loss": 0.5691,
+      "step": 710
+    },
+    {
+      "epoch": 0.15859915235333483,
+      "grad_norm": 0.20351482927799225,
+      "learning_rate": 1.9897130660281127e-05,
+      "loss": 0.5513,
+      "step": 711
+    },
+    {
+      "epoch": 0.1588222172652242,
+      "grad_norm": 0.2297467589378357,
+      "learning_rate": 1.9896793689127988e-05,
+      "loss": 0.5618,
+      "step": 712
+    },
+    {
+      "epoch": 0.15904528217711353,
+      "grad_norm": 0.222783625125885,
+      "learning_rate": 1.989645616982816e-05,
+      "loss": 0.5336,
+      "step": 713
+    },
+    {
+      "epoch": 0.1592683470890029,
+      "grad_norm": 0.1896630972623825,
+      "learning_rate": 1.9896118102400334e-05,
+      "loss": 0.5562,
+      "step": 714
+    },
+    {
+      "epoch": 0.15949141200089226,
+      "grad_norm": 0.188043013215065,
+      "learning_rate": 1.989577948686324e-05,
+      "loss": 0.497,
+      "step": 715
+    },
+    {
+      "epoch": 0.15971447691278162,
+      "grad_norm": 0.21135394275188446,
+      "learning_rate": 1.9895440323235635e-05,
+      "loss": 0.5714,
+      "step": 716
+    },
+    {
+      "epoch": 0.159937541824671,
+      "grad_norm": 0.1962571144104004,
+      "learning_rate": 1.98951006115363e-05,
+      "loss": 0.5686,
+      "step": 717
+    },
+    {
+      "epoch": 0.16016060673656035,
+      "grad_norm": 0.196267768740654,
+      "learning_rate": 1.9894760351784047e-05,
+      "loss": 0.557,
+      "step": 718
+    },
+    {
+      "epoch": 0.1603836716484497,
+      "grad_norm": 0.1771324872970581,
+      "learning_rate": 1.9894419543997724e-05,
+      "loss": 0.5492,
+      "step": 719
+    },
+    {
+      "epoch": 0.16060673656033905,
+      "grad_norm": 0.17055559158325195,
+      "learning_rate": 1.9894078188196213e-05,
+      "loss": 0.5311,
+      "step": 720
+    },
+    {
+      "epoch": 0.16082980147222842,
+      "grad_norm": 0.18929804861545563,
+      "learning_rate": 1.9893736284398414e-05,
+      "loss": 0.5388,
+      "step": 721
+    },
+    {
+      "epoch": 0.16105286638411778,
+      "grad_norm": 0.1930275857448578,
+      "learning_rate": 1.9893393832623266e-05,
+      "loss": 0.554,
+      "step": 722
+    },
+    {
+      "epoch": 0.16127593129600715,
+      "grad_norm": 0.1900511384010315,
+      "learning_rate": 1.9893050832889734e-05,
+      "loss": 0.5446,
+      "step": 723
+    },
+    {
+      "epoch": 0.1614989962078965,
+      "grad_norm": 0.17943377792835236,
+      "learning_rate": 1.9892707285216816e-05,
+      "loss": 0.541,
+      "step": 724
+    },
+    {
+      "epoch": 0.16172206111978585,
+      "grad_norm": 0.18837152421474457,
+      "learning_rate": 1.9892363189623546e-05,
+      "loss": 0.5424,
+      "step": 725
+    },
+    {
+      "epoch": 0.1619451260316752,
+      "grad_norm": 0.1987512707710266,
+      "learning_rate": 1.989201854612897e-05,
+      "loss": 0.5412,
+      "step": 726
+    },
+    {
+      "epoch": 0.16216819094356458,
+      "grad_norm": 0.1869790405035019,
+      "learning_rate": 1.9891673354752192e-05,
+      "loss": 0.5139,
+      "step": 727
+    },
+    {
+      "epoch": 0.16239125585545394,
+      "grad_norm": 0.17572949826717377,
+      "learning_rate": 1.9891327615512315e-05,
+      "loss": 0.5137,
+      "step": 728
+    },
+    {
+      "epoch": 0.1626143207673433,
+      "grad_norm": 0.17914965748786926,
+      "learning_rate": 1.9890981328428502e-05,
+      "loss": 0.5416,
+      "step": 729
+    },
+    {
+      "epoch": 0.16283738567923264,
+      "grad_norm": 0.17955482006072998,
+      "learning_rate": 1.989063449351992e-05,
+      "loss": 0.5307,
+      "step": 730
+    },
+    {
+      "epoch": 0.163060450591122,
+      "grad_norm": 0.17228074371814728,
+      "learning_rate": 1.9890287110805787e-05,
+      "loss": 0.5179,
+      "step": 731
+    },
+    {
+      "epoch": 0.16328351550301137,
+      "grad_norm": 0.19471096992492676,
+      "learning_rate": 1.9889939180305343e-05,
+      "loss": 0.5787,
+      "step": 732
+    },
+    {
+      "epoch": 0.16350658041490074,
+      "grad_norm": 0.19749003648757935,
+      "learning_rate": 1.9889590702037857e-05,
+      "loss": 0.5369,
+      "step": 733
+    },
+    {
+      "epoch": 0.1637296453267901,
+      "grad_norm": 0.17162089049816132,
+      "learning_rate": 1.9889241676022628e-05,
+      "loss": 0.5426,
+      "step": 734
+    },
+    {
+      "epoch": 0.16395271023867947,
+      "grad_norm": 0.18129977583885193,
+      "learning_rate": 1.988889210227899e-05,
+      "loss": 0.5297,
+      "step": 735
+    },
+    {
+      "epoch": 0.1641757751505688,
+      "grad_norm": 0.18625618517398834,
+      "learning_rate": 1.9888541980826307e-05,
+      "loss": 0.5169,
+      "step": 736
+    },
+    {
+      "epoch": 0.16439884006245817,
+      "grad_norm": 0.21191106736660004,
+      "learning_rate": 1.9888191311683966e-05,
+      "loss": 0.5322,
+      "step": 737
+    },
+    {
+      "epoch": 0.16462190497434753,
+      "grad_norm": 0.4791422486305237,
+      "learning_rate": 1.988784009487139e-05,
+      "loss": 0.5285,
+      "step": 738
+    },
+    {
+      "epoch": 0.1648449698862369,
+      "grad_norm": 0.18190808594226837,
+      "learning_rate": 1.9887488330408033e-05,
+      "loss": 0.5627,
+      "step": 739
+    },
+    {
+      "epoch": 0.16506803479812626,
+      "grad_norm": 0.17037086188793182,
+      "learning_rate": 1.9887136018313374e-05,
+      "loss": 0.5329,
+      "step": 740
+    },
+    {
+      "epoch": 0.16529109971001563,
+      "grad_norm": 0.18794186413288116,
+      "learning_rate": 1.9886783158606934e-05,
+      "loss": 0.5469,
+      "step": 741
+    },
+    {
+      "epoch": 0.16551416462190496,
+      "grad_norm": 0.19635480642318726,
+      "learning_rate": 1.9886429751308252e-05,
+      "loss": 0.5707,
+      "step": 742
+    },
+    {
+      "epoch": 0.16573722953379433,
+      "grad_norm": 0.1907324194908142,
+      "learning_rate": 1.9886075796436902e-05,
+      "loss": 0.5676,
+      "step": 743
+    },
+    {
+      "epoch": 0.1659602944456837,
+      "grad_norm": 0.1943301409482956,
+      "learning_rate": 1.9885721294012487e-05,
+      "loss": 0.5653,
+      "step": 744
+    },
+    {
+      "epoch": 0.16618335935757306,
+      "grad_norm": 0.19706295430660248,
+      "learning_rate": 1.9885366244054646e-05,
+      "loss": 0.5258,
+      "step": 745
+    },
+    {
+      "epoch": 0.16640642426946242,
+      "grad_norm": 0.2110588550567627,
+      "learning_rate": 1.9885010646583038e-05,
+      "loss": 0.5509,
+      "step": 746
+    },
+    {
+      "epoch": 0.1666294891813518,
+      "grad_norm": 0.18079644441604614,
+      "learning_rate": 1.988465450161736e-05,
+      "loss": 0.5299,
+      "step": 747
+    },
+    {
+      "epoch": 0.16685255409324112,
+      "grad_norm": 0.2338915318250656,
+      "learning_rate": 1.988429780917734e-05,
+      "loss": 0.5637,
+      "step": 748
+    },
+    {
+      "epoch": 0.1670756190051305,
+      "grad_norm": 0.18777087330818176,
+      "learning_rate": 1.9883940569282737e-05,
+      "loss": 0.5502,
+      "step": 749
+    },
+    {
+      "epoch": 0.16729868391701985,
+      "grad_norm": 0.17644049227237701,
+      "learning_rate": 1.988358278195333e-05,
+      "loss": 0.5477,
+      "step": 750
+    },
+    {
+      "epoch": 0.16752174882890922,
+      "grad_norm": 0.1912970095872879,
+      "learning_rate": 1.9883224447208936e-05,
+      "loss": 0.5447,
+      "step": 751
+    },
+    {
+      "epoch": 0.16774481374079858,
+      "grad_norm": 0.1746956706047058,
+      "learning_rate": 1.9882865565069408e-05,
+      "loss": 0.5425,
+      "step": 752
+    },
+    {
+      "epoch": 0.16796787865268792,
+      "grad_norm": 0.18779419362545013,
+      "learning_rate": 1.9882506135554614e-05,
+      "loss": 0.5329,
+      "step": 753
+    },
+    {
+      "epoch": 0.16819094356457728,
+      "grad_norm": 0.21110400557518005,
+      "learning_rate": 1.9882146158684473e-05,
+      "loss": 0.5421,
+      "step": 754
+    },
+    {
+      "epoch": 0.16841400847646665,
+      "grad_norm": 0.17395779490470886,
+      "learning_rate": 1.9881785634478915e-05,
+      "loss": 0.5321,
+      "step": 755
+    },
+    {
+      "epoch": 0.168637073388356,
+      "grad_norm": 0.1749841570854187,
+      "learning_rate": 1.988142456295791e-05,
+      "loss": 0.522,
+      "step": 756
+    },
+    {
+      "epoch": 0.16886013830024538,
+      "grad_norm": 0.17398926615715027,
+      "learning_rate": 1.988106294414145e-05,
+      "loss": 0.5563,
+      "step": 757
+    },
+    {
+      "epoch": 0.16908320321213474,
+      "grad_norm": 0.1813582479953766,
+      "learning_rate": 1.9880700778049575e-05,
+      "loss": 0.5242,
+      "step": 758
+    },
+    {
+      "epoch": 0.16930626812402408,
+      "grad_norm": 0.18003934621810913,
+      "learning_rate": 1.9880338064702337e-05,
+      "loss": 0.5468,
+      "step": 759
+    },
+    {
+      "epoch": 0.16952933303591344,
+      "grad_norm": 0.17160138487815857,
+      "learning_rate": 1.9879974804119827e-05,
+      "loss": 0.5321,
+      "step": 760
+    },
+    {
+      "epoch": 0.1697523979478028,
+      "grad_norm": 0.19145694375038147,
+      "learning_rate": 1.9879610996322168e-05,
+      "loss": 0.5365,
+      "step": 761
+    },
+    {
+      "epoch": 0.16997546285969217,
+      "grad_norm": 0.18664413690567017,
+      "learning_rate": 1.9879246641329505e-05,
+      "loss": 0.5278,
+      "step": 762
+    },
+    {
+      "epoch": 0.17019852777158154,
+      "grad_norm": 0.18776154518127441,
+      "learning_rate": 1.987888173916202e-05,
+      "loss": 0.5639,
+      "step": 763
+    },
+    {
+      "epoch": 0.1704215926834709,
+      "grad_norm": 0.17712879180908203,
+      "learning_rate": 1.9878516289839923e-05,
+      "loss": 0.5119,
+      "step": 764
+    },
+    {
+      "epoch": 0.17064465759536024,
+      "grad_norm": 0.1806483119726181,
+      "learning_rate": 1.9878150293383457e-05,
+      "loss": 0.5507,
+      "step": 765
+    },
+    {
+      "epoch": 0.1708677225072496,
+      "grad_norm": 0.18238821625709534,
+      "learning_rate": 1.9877783749812892e-05,
+      "loss": 0.5581,
+      "step": 766
+    },
+    {
+      "epoch": 0.17109078741913897,
+      "grad_norm": 0.17094913125038147,
+      "learning_rate": 1.9877416659148525e-05,
+      "loss": 0.5502,
+      "step": 767
+    },
+    {
+      "epoch": 0.17131385233102833,
+      "grad_norm": 0.18648453056812286,
+      "learning_rate": 1.9877049021410696e-05,
+      "loss": 0.5568,
+      "step": 768
+    },
+    {
+      "epoch": 0.1715369172429177,
+      "grad_norm": 0.20364215970039368,
+      "learning_rate": 1.9876680836619762e-05,
+      "loss": 0.5233,
+      "step": 769
+    },
+    {
+      "epoch": 0.17175998215480706,
+      "grad_norm": 0.17506776750087738,
+      "learning_rate": 1.9876312104796117e-05,
+      "loss": 0.5479,
+      "step": 770
+    },
+    {
+      "epoch": 0.1719830470666964,
+      "grad_norm": 0.17632229626178741,
+      "learning_rate": 1.9875942825960183e-05,
+      "loss": 0.5241,
+      "step": 771
+    },
+    {
+      "epoch": 0.17220611197858576,
+      "grad_norm": 0.18557725846767426,
+      "learning_rate": 1.9875573000132414e-05,
+      "loss": 0.5794,
+      "step": 772
+    },
+    {
+      "epoch": 0.17242917689047513,
+      "grad_norm": 0.1729423552751541,
+      "learning_rate": 1.987520262733329e-05,
+      "loss": 0.5247,
+      "step": 773
+    },
+    {
+      "epoch": 0.1726522418023645,
+      "grad_norm": 0.18643899261951447,
+      "learning_rate": 1.9874831707583328e-05,
+      "loss": 0.5305,
+      "step": 774
+    },
+    {
+      "epoch": 0.17287530671425386,
+      "grad_norm": 0.17992724478244781,
+      "learning_rate": 1.987446024090307e-05,
+      "loss": 0.5548,
+      "step": 775
+    },
+    {
+      "epoch": 0.17309837162614322,
+      "grad_norm": 0.19330288469791412,
+      "learning_rate": 1.9874088227313093e-05,
+      "loss": 0.5151,
+      "step": 776
+    },
+    {
+      "epoch": 0.17332143653803256,
+      "grad_norm": 0.18407322466373444,
+      "learning_rate": 1.9873715666834e-05,
+      "loss": 0.5262,
+      "step": 777
+    },
+    {
+      "epoch": 0.17354450144992192,
+      "grad_norm": 0.17531730234622955,
+      "learning_rate": 1.987334255948642e-05,
+      "loss": 0.51,
+      "step": 778
+    },
+    {
+      "epoch": 0.1737675663618113,
+      "grad_norm": 0.1791767179965973,
+      "learning_rate": 1.987296890529103e-05,
+      "loss": 0.5217,
+      "step": 779
+    },
+    {
+      "epoch": 0.17399063127370065,
+      "grad_norm": 0.17693190276622772,
+      "learning_rate": 1.9872594704268516e-05,
+      "loss": 0.5346,
+      "step": 780
+    },
+    {
+      "epoch": 0.17421369618559002,
+      "grad_norm": 0.17644071578979492,
+      "learning_rate": 1.9872219956439607e-05,
+      "loss": 0.5335,
+      "step": 781
+    },
+    {
+      "epoch": 0.17443676109747935,
+      "grad_norm": 0.17572622001171112,
+      "learning_rate": 1.987184466182506e-05,
+      "loss": 0.5302,
+      "step": 782
+    },
+    {
+      "epoch": 0.17465982600936872,
+      "grad_norm": 0.18111130595207214,
+      "learning_rate": 1.987146882044565e-05,
+      "loss": 0.5205,
+      "step": 783
+    },
+    {
+      "epoch": 0.17488289092125808,
+      "grad_norm": 0.17094580829143524,
+      "learning_rate": 1.987109243232221e-05,
+      "loss": 0.527,
+      "step": 784
+    },
+    {
+      "epoch": 0.17510595583314745,
+      "grad_norm": 0.1917412430047989,
+      "learning_rate": 1.9870715497475583e-05,
+      "loss": 0.5289,
+      "step": 785
+    },
+    {
+      "epoch": 0.1753290207450368,
+      "grad_norm": 0.19271767139434814,
+      "learning_rate": 1.9870338015926634e-05,
+      "loss": 0.5123,
+      "step": 786
+    },
+    {
+      "epoch": 0.17555208565692618,
+      "grad_norm": 0.19731736183166504,
+      "learning_rate": 1.9869959987696282e-05,
+      "loss": 0.543,
+      "step": 787
+    },
+    {
+      "epoch": 0.1757751505688155,
+      "grad_norm": 0.172173410654068,
+      "learning_rate": 1.9869581412805462e-05,
+      "loss": 0.5211,
+      "step": 788
+    },
+    {
+      "epoch": 0.17599821548070488,
+      "grad_norm": 0.1778416931629181,
+      "learning_rate": 1.9869202291275144e-05,
+      "loss": 0.5168,
+      "step": 789
+    },
+    {
+      "epoch": 0.17622128039259424,
+      "grad_norm": 0.1906319111585617,
+      "learning_rate": 1.986882262312632e-05,
+      "loss": 0.5502,
+      "step": 790
+    },
+    {
+      "epoch": 0.1764443453044836,
+      "grad_norm": 0.18600904941558838,
+      "learning_rate": 1.986844240838002e-05,
+      "loss": 0.5093,
+      "step": 791
+    },
+    {
+      "epoch": 0.17666741021637297,
+      "grad_norm": 0.17453651130199432,
+      "learning_rate": 1.986806164705731e-05,
+      "loss": 0.5316,
+      "step": 792
+    },
+    {
+      "epoch": 0.17689047512826234,
+      "grad_norm": 0.18647123873233795,
+      "learning_rate": 1.9867680339179268e-05,
+      "loss": 0.5293,
+      "step": 793
+    },
+    {
+      "epoch": 0.17711354004015167,
+      "grad_norm": 0.18260005116462708,
+      "learning_rate": 1.9867298484767022e-05,
+      "loss": 0.5429,
+      "step": 794
+    },
+    {
+      "epoch": 0.17733660495204104,
+      "grad_norm": 0.1722402721643448,
+      "learning_rate": 1.9866916083841715e-05,
+      "loss": 0.5211,
+      "step": 795
+    },
+    {
+      "epoch": 0.1775596698639304,
+      "grad_norm": 0.16579583287239075,
+      "learning_rate": 1.9866533136424537e-05,
+      "loss": 0.5173,
+      "step": 796
+    },
+    {
+      "epoch": 0.17778273477581977,
+      "grad_norm": 0.18849937617778778,
+      "learning_rate": 1.9866149642536683e-05,
+      "loss": 0.5482,
+      "step": 797
+    },
+    {
+      "epoch": 0.17800579968770913,
+      "grad_norm": 0.1786874532699585,
+      "learning_rate": 1.98657656021994e-05,
+      "loss": 0.5203,
+      "step": 798
+    },
+    {
+      "epoch": 0.1782288645995985,
+      "grad_norm": 0.1848941594362259,
+      "learning_rate": 1.986538101543397e-05,
+      "loss": 0.5248,
+      "step": 799
+    },
+    {
+      "epoch": 0.17845192951148783,
+      "grad_norm": 0.1833941787481308,
+      "learning_rate": 1.9864995882261674e-05,
+      "loss": 0.5192,
+      "step": 800
+    },
+    {
+      "epoch": 0.1786749944233772,
+      "grad_norm": 0.1813460886478424,
+      "learning_rate": 1.9864610202703858e-05,
+      "loss": 0.5307,
+      "step": 801
+    },
+    {
+      "epoch": 0.17889805933526656,
+      "grad_norm": 0.18154440820217133,
+      "learning_rate": 1.9864223976781876e-05,
+      "loss": 0.5468,
+      "step": 802
+    },
+    {
+      "epoch": 0.17912112424715593,
+      "grad_norm": 0.18413223326206207,
+      "learning_rate": 1.9863837204517124e-05,
+      "loss": 0.5376,
+      "step": 803
+    },
+    {
+      "epoch": 0.1793441891590453,
+      "grad_norm": 0.17483587563037872,
+      "learning_rate": 1.986344988593102e-05,
+      "loss": 0.5333,
+      "step": 804
+    },
+    {
+      "epoch": 0.17956725407093463,
+      "grad_norm": 0.17205438017845154,
+      "learning_rate": 1.9863062021045017e-05,
+      "loss": 0.4933,
+      "step": 805
+    },
+    {
+      "epoch": 0.179790318982824,
+      "grad_norm": 0.17578768730163574,
+      "learning_rate": 1.98626736098806e-05,
+      "loss": 0.5153,
+      "step": 806
+    },
+    {
+      "epoch": 0.18001338389471336,
+      "grad_norm": 0.1773952841758728,
+      "learning_rate": 1.9862284652459275e-05,
+      "loss": 0.5265,
+      "step": 807
+    },
+    {
+      "epoch": 0.18023644880660272,
+      "grad_norm": 0.17616188526153564,
+      "learning_rate": 1.9861895148802594e-05,
+      "loss": 0.5438,
+      "step": 808
+    },
+    {
+      "epoch": 0.1804595137184921,
+      "grad_norm": 0.1791033297777176,
+      "learning_rate": 1.9861505098932127e-05,
+      "loss": 0.5294,
+      "step": 809
+    },
+    {
+      "epoch": 0.18068257863038145,
+      "grad_norm": 0.18261539936065674,
+      "learning_rate": 1.986111450286947e-05,
+      "loss": 0.548,
+      "step": 810
+    },
+    {
+      "epoch": 0.1809056435422708,
+      "grad_norm": 0.1767009198665619,
+      "learning_rate": 1.986072336063627e-05,
+      "loss": 0.5278,
+      "step": 811
+    },
+    {
+      "epoch": 0.18112870845416015,
+      "grad_norm": 0.19362005591392517,
+      "learning_rate": 1.9860331672254182e-05,
+      "loss": 0.5206,
+      "step": 812
+    },
+    {
+      "epoch": 0.18135177336604952,
+      "grad_norm": 0.185451477766037,
+      "learning_rate": 1.98599394377449e-05,
+      "loss": 0.5513,
+      "step": 813
+    },
+    {
+      "epoch": 0.18157483827793888,
+      "grad_norm": 0.17266923189163208,
+      "learning_rate": 1.985954665713015e-05,
+      "loss": 0.5648,
+      "step": 814
+    },
+    {
+      "epoch": 0.18179790318982825,
+      "grad_norm": 0.170868918299675,
+      "learning_rate": 1.9859153330431692e-05,
+      "loss": 0.5343,
+      "step": 815
+    },
+    {
+      "epoch": 0.1820209681017176,
+      "grad_norm": 0.17769230902194977,
+      "learning_rate": 1.98587594576713e-05,
+      "loss": 0.5539,
+      "step": 816
+    },
+    {
+      "epoch": 0.18224403301360695,
+      "grad_norm": 0.16811443865299225,
+      "learning_rate": 1.9858365038870803e-05,
+      "loss": 0.5209,
+      "step": 817
+    },
+    {
+      "epoch": 0.1824670979254963,
+      "grad_norm": 0.1751745492219925,
+      "learning_rate": 1.985797007405203e-05,
+      "loss": 0.5509,
+      "step": 818
+    },
+    {
+      "epoch": 0.18269016283738568,
+      "grad_norm": 0.1694861203432083,
+      "learning_rate": 1.985757456323687e-05,
+      "loss": 0.5505,
+      "step": 819
+    },
+    {
+      "epoch": 0.18291322774927504,
+      "grad_norm": 0.1734897792339325,
+      "learning_rate": 1.985717850644722e-05,
+      "loss": 0.5022,
+      "step": 820
+    },
+    {
+      "epoch": 0.1831362926611644,
+      "grad_norm": 0.18104127049446106,
+      "learning_rate": 1.9856781903705026e-05,
+      "loss": 0.5434,
+      "step": 821
+    },
+    {
+      "epoch": 0.18335935757305377,
+      "grad_norm": 0.17157487571239471,
+      "learning_rate": 1.9856384755032245e-05,
+      "loss": 0.5356,
+      "step": 822
+    },
+    {
+      "epoch": 0.1835824224849431,
+      "grad_norm": 0.1781257838010788,
+      "learning_rate": 1.985598706045088e-05,
+      "loss": 0.522,
+      "step": 823
+    },
+    {
+      "epoch": 0.18380548739683247,
+      "grad_norm": 0.16910432279109955,
+      "learning_rate": 1.985558881998295e-05,
+      "loss": 0.5324,
+      "step": 824
+    },
+    {
+      "epoch": 0.18402855230872184,
+      "grad_norm": 0.180936798453331,
+      "learning_rate": 1.985519003365052e-05,
+      "loss": 0.5307,
+      "step": 825
+    },
+    {
+      "epoch": 0.1842516172206112,
+      "grad_norm": 0.1872517466545105,
+      "learning_rate": 1.9854790701475676e-05,
+      "loss": 0.5667,
+      "step": 826
+    },
+    {
+      "epoch": 0.18447468213250057,
+      "grad_norm": 0.16592343151569366,
+      "learning_rate": 1.985439082348053e-05,
+      "loss": 0.4938,
+      "step": 827
+    },
+    {
+      "epoch": 0.1846977470443899,
+      "grad_norm": 0.18184059858322144,
+      "learning_rate": 1.9853990399687237e-05,
+      "loss": 0.536,
+      "step": 828
+    },
+    {
+      "epoch": 0.18492081195627927,
+      "grad_norm": 0.1687706708908081,
+      "learning_rate": 1.985358943011797e-05,
+      "loss": 0.4902,
+      "step": 829
+    },
+    {
+      "epoch": 0.18514387686816863,
+      "grad_norm": 0.18062466382980347,
+      "learning_rate": 1.985318791479494e-05,
+      "loss": 0.5235,
+      "step": 830
+    },
+    {
+      "epoch": 0.185366941780058,
+      "grad_norm": 0.1865355372428894,
+      "learning_rate": 1.985278585374038e-05,
+      "loss": 0.5439,
+      "step": 831
+    },
+    {
+      "epoch": 0.18559000669194736,
+      "grad_norm": 0.21087662875652313,
+      "learning_rate": 1.985238324697657e-05,
+      "loss": 0.5338,
+      "step": 832
+    },
+    {
+      "epoch": 0.18581307160383673,
+      "grad_norm": 0.18593360483646393,
+      "learning_rate": 1.9851980094525795e-05,
+      "loss": 0.5139,
+      "step": 833
+    },
+    {
+      "epoch": 0.18603613651572606,
+      "grad_norm": 0.17105735838413239,
+      "learning_rate": 1.9851576396410395e-05,
+      "loss": 0.5385,
+      "step": 834
+    },
+    {
+      "epoch": 0.18625920142761543,
+      "grad_norm": 0.1764969825744629,
+      "learning_rate": 1.9851172152652722e-05,
+      "loss": 0.5381,
+      "step": 835
+    },
+    {
+      "epoch": 0.1864822663395048,
+      "grad_norm": 0.211869478225708,
+      "learning_rate": 1.985076736327517e-05,
+      "loss": 0.512,
+      "step": 836
+    },
+    {
+      "epoch": 0.18670533125139416,
+      "grad_norm": 0.17822885513305664,
+      "learning_rate": 1.9850362028300162e-05,
+      "loss": 0.5394,
+      "step": 837
+    },
+    {
+      "epoch": 0.18692839616328352,
+      "grad_norm": 0.17853499948978424,
+      "learning_rate": 1.9849956147750137e-05,
+      "loss": 0.5273,
+      "step": 838
+    },
+    {
+      "epoch": 0.1871514610751729,
+      "grad_norm": 0.1717289835214615,
+      "learning_rate": 1.9849549721647586e-05,
+      "loss": 0.5126,
+      "step": 839
+    },
+    {
+      "epoch": 0.18737452598706222,
+      "grad_norm": 0.167042076587677,
+      "learning_rate": 1.9849142750015014e-05,
+      "loss": 0.5154,
+      "step": 840
+    },
+    {
+      "epoch": 0.1875975908989516,
+      "grad_norm": 0.1893247812986374,
+      "learning_rate": 1.9848735232874966e-05,
+      "loss": 0.5205,
+      "step": 841
+    },
+    {
+      "epoch": 0.18782065581084095,
+      "grad_norm": 0.18898561596870422,
+      "learning_rate": 1.984832717025001e-05,
+      "loss": 0.5513,
+      "step": 842
+    },
+    {
+      "epoch": 0.18804372072273032,
+      "grad_norm": 0.18172813951969147,
+      "learning_rate": 1.984791856216274e-05,
+      "loss": 0.5056,
+      "step": 843
+    },
+    {
+      "epoch": 0.18826678563461968,
+      "grad_norm": 0.1696682870388031,
+      "learning_rate": 1.98475094086358e-05,
+      "loss": 0.5313,
+      "step": 844
+    },
+    {
+      "epoch": 0.18848985054650905,
+      "grad_norm": 0.1731012910604477,
+      "learning_rate": 1.9847099709691843e-05,
+      "loss": 0.5108,
+      "step": 845
+    },
+    {
+      "epoch": 0.18871291545839838,
+      "grad_norm": 0.18047916889190674,
+      "learning_rate": 1.9846689465353563e-05,
+      "loss": 0.5108,
+      "step": 846
+    },
+    {
+      "epoch": 0.18893598037028775,
+      "grad_norm": 0.17469045519828796,
+      "learning_rate": 1.9846278675643684e-05,
+      "loss": 0.5273,
+      "step": 847
+    },
+    {
+      "epoch": 0.1891590452821771,
+      "grad_norm": 0.18713539838790894,
+      "learning_rate": 1.9845867340584957e-05,
+      "loss": 0.5587,
+      "step": 848
+    },
+    {
+      "epoch": 0.18938211019406648,
+      "grad_norm": 0.2040695995092392,
+      "learning_rate": 1.984545546020016e-05,
+      "loss": 0.5651,
+      "step": 849
+    },
+    {
+      "epoch": 0.18960517510595584,
+      "grad_norm": 0.24084708094596863,
+      "learning_rate": 1.984504303451211e-05,
+      "loss": 0.5186,
+      "step": 850
+    },
+    {
+      "epoch": 0.1898282400178452,
+      "grad_norm": 0.18628886342048645,
+      "learning_rate": 1.9844630063543655e-05,
+      "loss": 0.5091,
+      "step": 851
+    },
+    {
+      "epoch": 0.19005130492973454,
+      "grad_norm": 0.18985594809055328,
+      "learning_rate": 1.9844216547317656e-05,
+      "loss": 0.5457,
+      "step": 852
+    },
+    {
+      "epoch": 0.1902743698416239,
+      "grad_norm": 0.1807604879140854,
+      "learning_rate": 1.9843802485857028e-05,
+      "loss": 0.5137,
+      "step": 853
+    },
+    {
+      "epoch": 0.19049743475351327,
+      "grad_norm": 0.1864684522151947,
+      "learning_rate": 1.984338787918469e-05,
+      "loss": 0.5232,
+      "step": 854
+    },
+    {
+      "epoch": 0.19072049966540264,
+      "grad_norm": 0.17398470640182495,
+      "learning_rate": 1.984297272732362e-05,
+      "loss": 0.5207,
+      "step": 855
+    },
+    {
+      "epoch": 0.190943564577292,
+      "grad_norm": 0.21071763336658478,
+      "learning_rate": 1.9842557030296804e-05,
+      "loss": 0.5032,
+      "step": 856
+    },
+    {
+      "epoch": 0.19116662948918134,
+      "grad_norm": 0.17372627556324005,
+      "learning_rate": 1.9842140788127264e-05,
+      "loss": 0.4992,
+      "step": 857
+    },
+    {
+      "epoch": 0.1913896944010707,
+      "grad_norm": 0.18683859705924988,
+      "learning_rate": 1.9841724000838064e-05,
+      "loss": 0.5342,
+      "step": 858
+    },
+    {
+      "epoch": 0.19161275931296007,
+      "grad_norm": 0.18670214712619781,
+      "learning_rate": 1.9841306668452275e-05,
+      "loss": 0.521,
+      "step": 859
+    },
+    {
+      "epoch": 0.19183582422484943,
+      "grad_norm": 0.18776099383831024,
+      "learning_rate": 1.9840888790993023e-05,
+      "loss": 0.5164,
+      "step": 860
+    },
+    {
+      "epoch": 0.1920588891367388,
+      "grad_norm": 0.20680485665798187,
+      "learning_rate": 1.9840470368483448e-05,
+      "loss": 0.5433,
+      "step": 861
+    },
+    {
+      "epoch": 0.19228195404862816,
+      "grad_norm": 0.18562763929367065,
+      "learning_rate": 1.9840051400946724e-05,
+      "loss": 0.5225,
+      "step": 862
+    },
+    {
+      "epoch": 0.1925050189605175,
+      "grad_norm": 0.1958203911781311,
+      "learning_rate": 1.9839631888406055e-05,
+      "loss": 0.5333,
+      "step": 863
+    },
+    {
+      "epoch": 0.19272808387240686,
+      "grad_norm": 0.1797972470521927,
+      "learning_rate": 1.9839211830884682e-05,
+      "loss": 0.5259,
+      "step": 864
+    },
+    {
+      "epoch": 0.19295114878429623,
+      "grad_norm": 0.18006913363933563,
+      "learning_rate": 1.9838791228405866e-05,
+      "loss": 0.5355,
+      "step": 865
+    },
+    {
+      "epoch": 0.1931742136961856,
+      "grad_norm": 0.19286584854125977,
+      "learning_rate": 1.9838370080992902e-05,
+      "loss": 0.5548,
+      "step": 866
+    },
+    {
+      "epoch": 0.19339727860807496,
+      "grad_norm": 0.17611972987651825,
+      "learning_rate": 1.9837948388669118e-05,
+      "loss": 0.4975,
+      "step": 867
+    },
+    {
+      "epoch": 0.19362034351996432,
+      "grad_norm": 0.17204061150550842,
+      "learning_rate": 1.983752615145787e-05,
+      "loss": 0.5353,
+      "step": 868
+    },
+    {
+      "epoch": 0.19384340843185366,
+      "grad_norm": 0.17773252725601196,
+      "learning_rate": 1.9837103369382542e-05,
+      "loss": 0.5621,
+      "step": 869
+    },
+    {
+      "epoch": 0.19406647334374302,
+      "grad_norm": 0.17241743206977844,
+      "learning_rate": 1.983668004246655e-05,
+      "loss": 0.5514,
+      "step": 870
+    },
+    {
+      "epoch": 0.1942895382556324,
+      "grad_norm": 0.1765352487564087,
+      "learning_rate": 1.9836256170733343e-05,
+      "loss": 0.5262,
+      "step": 871
+    },
+    {
+      "epoch": 0.19451260316752175,
+      "grad_norm": 0.18487049639225006,
+      "learning_rate": 1.98358317542064e-05,
+      "loss": 0.4941,
+      "step": 872
+    },
+    {
+      "epoch": 0.19473566807941112,
+      "grad_norm": 0.17895649373531342,
+      "learning_rate": 1.983540679290922e-05,
+      "loss": 0.5501,
+      "step": 873
+    },
+    {
+      "epoch": 0.19495873299130048,
+      "grad_norm": 0.1671830266714096,
+      "learning_rate": 1.9834981286865343e-05,
+      "loss": 0.5207,
+      "step": 874
+    },
+    {
+      "epoch": 0.19518179790318982,
+      "grad_norm": 0.18288585543632507,
+      "learning_rate": 1.9834555236098344e-05,
+      "loss": 0.5405,
+      "step": 875
+    },
+    {
+      "epoch": 0.19540486281507918,
+      "grad_norm": 0.2788625657558441,
+      "learning_rate": 1.983412864063181e-05,
+      "loss": 0.5029,
+      "step": 876
+    },
+    {
+      "epoch": 0.19562792772696855,
+      "grad_norm": 0.17516101896762848,
+      "learning_rate": 1.983370150048938e-05,
+      "loss": 0.5181,
+      "step": 877
+    },
+    {
+      "epoch": 0.1958509926388579,
+      "grad_norm": 0.18724878132343292,
+      "learning_rate": 1.9833273815694695e-05,
+      "loss": 0.5399,
+      "step": 878
+    },
+    {
+      "epoch": 0.19607405755074728,
+      "grad_norm": 0.1845855563879013,
+      "learning_rate": 1.9832845586271456e-05,
+      "loss": 0.5493,
+      "step": 879
+    },
+    {
+      "epoch": 0.19629712246263661,
+      "grad_norm": 0.1939295083284378,
+      "learning_rate": 1.9832416812243377e-05,
+      "loss": 0.5453,
+      "step": 880
+    },
+    {
+      "epoch": 0.19652018737452598,
+      "grad_norm": 0.17018291354179382,
+      "learning_rate": 1.9831987493634207e-05,
+      "loss": 0.5096,
+      "step": 881
+    },
+    {
+      "epoch": 0.19674325228641534,
+      "grad_norm": 0.18090112507343292,
+      "learning_rate": 1.9831557630467725e-05,
+      "loss": 0.5519,
+      "step": 882
+    },
+    {
+      "epoch": 0.1969663171983047,
+      "grad_norm": 0.19264590740203857,
+      "learning_rate": 1.983112722276774e-05,
+      "loss": 0.5469,
+      "step": 883
+    },
+    {
+      "epoch": 0.19718938211019407,
+      "grad_norm": 0.185777947306633,
+      "learning_rate": 1.9830696270558084e-05,
+      "loss": 0.5484,
+      "step": 884
+    },
+    {
+      "epoch": 0.19741244702208344,
+      "grad_norm": 0.15953023731708527,
+      "learning_rate": 1.9830264773862633e-05,
+      "loss": 0.5139,
+      "step": 885
+    },
+    {
+      "epoch": 0.19763551193397277,
+      "grad_norm": 0.17809630930423737,
+      "learning_rate": 1.9829832732705284e-05,
+      "loss": 0.5178,
+      "step": 886
+    },
+    {
+      "epoch": 0.19785857684586214,
+      "grad_norm": 0.20061112940311432,
+      "learning_rate": 1.982940014710997e-05,
+      "loss": 0.5393,
+      "step": 887
+    },
+    {
+      "epoch": 0.1980816417577515,
+      "grad_norm": 0.1748671978712082,
+      "learning_rate": 1.9828967017100642e-05,
+      "loss": 0.5332,
+      "step": 888
+    },
+    {
+      "epoch": 0.19830470666964087,
+      "grad_norm": 0.1713806539773941,
+      "learning_rate": 1.9828533342701296e-05,
+      "loss": 0.5165,
+      "step": 889
+    },
+    {
+      "epoch": 0.19852777158153023,
+      "grad_norm": 0.16470858454704285,
+      "learning_rate": 1.9828099123935948e-05,
+      "loss": 0.5133,
+      "step": 890
+    },
+    {
+      "epoch": 0.1987508364934196,
+      "grad_norm": 0.17374449968338013,
+      "learning_rate": 1.9827664360828647e-05,
+      "loss": 0.5475,
+      "step": 891
+    },
+    {
+      "epoch": 0.19897390140530893,
+      "grad_norm": 0.1877843141555786,
+      "learning_rate": 1.982722905340348e-05,
+      "loss": 0.5253,
+      "step": 892
+    },
+    {
+      "epoch": 0.1991969663171983,
+      "grad_norm": 0.17328353226184845,
+      "learning_rate": 1.982679320168455e-05,
+      "loss": 0.5219,
+      "step": 893
+    },
+    {
+      "epoch": 0.19942003122908766,
+      "grad_norm": 0.17716079950332642,
+      "learning_rate": 1.9826356805696e-05,
+      "loss": 0.535,
+      "step": 894
+    },
+    {
+      "epoch": 0.19964309614097703,
+      "grad_norm": 0.17511911690235138,
+      "learning_rate": 1.9825919865462004e-05,
+      "loss": 0.548,
+      "step": 895
+    },
+    {
+      "epoch": 0.1998661610528664,
+      "grad_norm": 0.1711902767419815,
+      "learning_rate": 1.9825482381006752e-05,
+      "loss": 0.5402,
+      "step": 896
+    },
+    {
+      "epoch": 0.20008922596475576,
+      "grad_norm": 0.17732638120651245,
+      "learning_rate": 1.9825044352354482e-05,
+      "loss": 0.5672,
+      "step": 897
+    },
+    {
+      "epoch": 0.2003122908766451,
+      "grad_norm": 0.17266635596752167,
+      "learning_rate": 1.9824605779529456e-05,
+      "loss": 0.5312,
+      "step": 898
+    },
+    {
+      "epoch": 0.20053535578853446,
+      "grad_norm": 0.16797249019145966,
+      "learning_rate": 1.982416666255596e-05,
+      "loss": 0.5294,
+      "step": 899
+    },
+    {
+      "epoch": 0.20075842070042382,
+      "grad_norm": 0.17062917351722717,
+      "learning_rate": 1.9823727001458318e-05,
+      "loss": 0.51,
+      "step": 900
+    },
+    {
+      "epoch": 0.2009814856123132,
+      "grad_norm": 0.17725898325443268,
+      "learning_rate": 1.9823286796260887e-05,
+      "loss": 0.5284,
+      "step": 901
+    },
+    {
+      "epoch": 0.20120455052420255,
+      "grad_norm": 0.17662313580513,
+      "learning_rate": 1.9822846046988037e-05,
+      "loss": 0.515,
+      "step": 902
+    },
+    {
+      "epoch": 0.2014276154360919,
+      "grad_norm": 0.17648808658123016,
+      "learning_rate": 1.9822404753664183e-05,
+      "loss": 0.5437,
+      "step": 903
+    },
+    {
+      "epoch": 0.20165068034798125,
+      "grad_norm": 0.17179900407791138,
+      "learning_rate": 1.982196291631377e-05,
+      "loss": 0.5332,
+      "step": 904
+    },
+    {
+      "epoch": 0.20187374525987062,
+      "grad_norm": 0.19034990668296814,
+      "learning_rate": 1.982152053496127e-05,
+      "loss": 0.5279,
+      "step": 905
+    },
+    {
+      "epoch": 0.20209681017175998,
+      "grad_norm": 0.17066267132759094,
+      "learning_rate": 1.9821077609631184e-05,
+      "loss": 0.5473,
+      "step": 906
+    },
+    {
+      "epoch": 0.20231987508364935,
+      "grad_norm": 0.1806066334247589,
+      "learning_rate": 1.982063414034804e-05,
+      "loss": 0.5495,
+      "step": 907
+    },
+    {
+      "epoch": 0.2025429399955387,
+      "grad_norm": 0.17627598345279694,
+      "learning_rate": 1.9820190127136403e-05,
+      "loss": 0.5469,
+      "step": 908
+    },
+    {
+      "epoch": 0.20276600490742805,
+      "grad_norm": 0.18282422423362732,
+      "learning_rate": 1.9819745570020867e-05,
+      "loss": 0.5228,
+      "step": 909
+    },
+    {
+      "epoch": 0.2029890698193174,
+      "grad_norm": 0.17409084737300873,
+      "learning_rate": 1.981930046902605e-05,
+      "loss": 0.5183,
+      "step": 910
+    },
+    {
+      "epoch": 0.20321213473120678,
+      "grad_norm": 0.1846904307603836,
+      "learning_rate": 1.9818854824176612e-05,
+      "loss": 0.5198,
+      "step": 911
+    },
+    {
+      "epoch": 0.20343519964309614,
+      "grad_norm": 0.18149109184741974,
+      "learning_rate": 1.9818408635497224e-05,
+      "loss": 0.5078,
+      "step": 912
+    },
+    {
+      "epoch": 0.2036582645549855,
+      "grad_norm": 0.18114425241947174,
+      "learning_rate": 1.981796190301261e-05,
+      "loss": 0.531,
+      "step": 913
+    },
+    {
+      "epoch": 0.20388132946687487,
+      "grad_norm": 0.1718929558992386,
+      "learning_rate": 1.981751462674751e-05,
+      "loss": 0.4939,
+      "step": 914
+    },
+    {
+      "epoch": 0.2041043943787642,
+      "grad_norm": 0.1930830329656601,
+      "learning_rate": 1.9817066806726695e-05,
+      "loss": 0.5055,
+      "step": 915
+    },
+    {
+      "epoch": 0.20432745929065357,
+      "grad_norm": 0.17051468789577484,
+      "learning_rate": 1.9816618442974964e-05,
+      "loss": 0.5058,
+      "step": 916
+    },
+    {
+      "epoch": 0.20455052420254294,
+      "grad_norm": 0.1774812787771225,
+      "learning_rate": 1.9816169535517157e-05,
+      "loss": 0.5341,
+      "step": 917
+    },
+    {
+      "epoch": 0.2047735891144323,
+      "grad_norm": 0.17879648506641388,
+      "learning_rate": 1.9815720084378134e-05,
+      "loss": 0.534,
+      "step": 918
+    },
+    {
+      "epoch": 0.20499665402632167,
+      "grad_norm": 0.1766035556793213,
+      "learning_rate": 1.9815270089582795e-05,
+      "loss": 0.5407,
+      "step": 919
+    },
+    {
+      "epoch": 0.20521971893821103,
+      "grad_norm": 0.17892096936702728,
+      "learning_rate": 1.981481955115605e-05,
+      "loss": 0.4676,
+      "step": 920
+    },
+    {
+      "epoch": 0.20544278385010037,
+      "grad_norm": 0.17289894819259644,
+      "learning_rate": 1.9814368469122866e-05,
+      "loss": 0.5416,
+      "step": 921
+    },
+    {
+      "epoch": 0.20566584876198973,
+      "grad_norm": 0.1802949607372284,
+      "learning_rate": 1.981391684350822e-05,
+      "loss": 0.5425,
+      "step": 922
+    },
+    {
+      "epoch": 0.2058889136738791,
+      "grad_norm": 0.17635071277618408,
+      "learning_rate": 1.9813464674337126e-05,
+      "loss": 0.5294,
+      "step": 923
+    },
+    {
+      "epoch": 0.20611197858576846,
+      "grad_norm": 0.1723651885986328,
+      "learning_rate": 1.981301196163463e-05,
+      "loss": 0.5045,
+      "step": 924
+    },
+    {
+      "epoch": 0.20633504349765783,
+      "grad_norm": 0.18130990862846375,
+      "learning_rate": 1.9812558705425805e-05,
+      "loss": 0.5264,
+      "step": 925
+    },
+    {
+      "epoch": 0.2065581084095472,
+      "grad_norm": 0.1879456490278244,
+      "learning_rate": 1.9812104905735756e-05,
+      "loss": 0.5215,
+      "step": 926
+    },
+    {
+      "epoch": 0.20678117332143653,
+      "grad_norm": 0.1742536425590515,
+      "learning_rate": 1.9811650562589616e-05,
+      "loss": 0.5093,
+      "step": 927
+    },
+    {
+      "epoch": 0.2070042382333259,
+      "grad_norm": 0.18259446322917938,
+      "learning_rate": 1.981119567601255e-05,
+      "loss": 0.5528,
+      "step": 928
+    },
+    {
+      "epoch": 0.20722730314521526,
+      "grad_norm": 0.19506299495697021,
+      "learning_rate": 1.9810740246029755e-05,
+      "loss": 0.5338,
+      "step": 929
+    },
+    {
+      "epoch": 0.20745036805710462,
+      "grad_norm": 0.166092187166214,
+      "learning_rate": 1.981028427266645e-05,
+      "loss": 0.5308,
+      "step": 930
+    },
+    {
+      "epoch": 0.207673432968994,
+      "grad_norm": 0.18397627770900726,
+      "learning_rate": 1.980982775594789e-05,
+      "loss": 0.5593,
+      "step": 931
+    },
+    {
+      "epoch": 0.20789649788088332,
+      "grad_norm": 0.18404512107372284,
+      "learning_rate": 1.980937069589937e-05,
+      "loss": 0.5004,
+      "step": 932
+    },
+    {
+      "epoch": 0.2081195627927727,
+      "grad_norm": 0.1727455109357834,
+      "learning_rate": 1.9808913092546195e-05,
+      "loss": 0.5245,
+      "step": 933
+    },
+    {
+      "epoch": 0.20834262770466205,
+      "grad_norm": 0.18410643935203552,
+      "learning_rate": 1.980845494591371e-05,
+      "loss": 0.532,
+      "step": 934
+    },
+    {
+      "epoch": 0.20856569261655142,
+      "grad_norm": 0.19003835320472717,
+      "learning_rate": 1.9807996256027296e-05,
+      "loss": 0.5129,
+      "step": 935
+    },
+    {
+      "epoch": 0.20878875752844078,
+      "grad_norm": 0.23560728132724762,
+      "learning_rate": 1.980753702291235e-05,
+      "loss": 0.5484,
+      "step": 936
+    },
+    {
+      "epoch": 0.20901182244033015,
+      "grad_norm": 0.17955288290977478,
+      "learning_rate": 1.9807077246594316e-05,
+      "loss": 0.5198,
+      "step": 937
+    },
+    {
+      "epoch": 0.20923488735221948,
+      "grad_norm": 0.19580209255218506,
+      "learning_rate": 1.9806616927098653e-05,
+      "loss": 0.5023,
+      "step": 938
+    },
+    {
+      "epoch": 0.20945795226410885,
+      "grad_norm": 0.17852741479873657,
+      "learning_rate": 1.9806156064450855e-05,
+      "loss": 0.5101,
+      "step": 939
+    },
+    {
+      "epoch": 0.2096810171759982,
+      "grad_norm": 0.17126716673374176,
+      "learning_rate": 1.9805694658676458e-05,
+      "loss": 0.5415,
+      "step": 940
+    },
+    {
+      "epoch": 0.20990408208788758,
+      "grad_norm": 0.16866669058799744,
+      "learning_rate": 1.9805232709801008e-05,
+      "loss": 0.5066,
+      "step": 941
+    },
+    {
+      "epoch": 0.21012714699977694,
+      "grad_norm": 0.17179332673549652,
+      "learning_rate": 1.9804770217850093e-05,
+      "loss": 0.5059,
+      "step": 942
+    },
+    {
+      "epoch": 0.2103502119116663,
+      "grad_norm": 0.1836715042591095,
+      "learning_rate": 1.9804307182849326e-05,
+      "loss": 0.5291,
+      "step": 943
+    },
+    {
+      "epoch": 0.21057327682355564,
+      "grad_norm": 0.2429589480161667,
+      "learning_rate": 1.980384360482436e-05,
+      "loss": 0.5282,
+      "step": 944
+    },
+    {
+      "epoch": 0.210796341735445,
+      "grad_norm": 0.17777171730995178,
+      "learning_rate": 1.9803379483800866e-05,
+      "loss": 0.5338,
+      "step": 945
+    },
+    {
+      "epoch": 0.21101940664733437,
+      "grad_norm": 0.20419363677501678,
+      "learning_rate": 1.9802914819804546e-05,
+      "loss": 0.5345,
+      "step": 946
+    },
+    {
+      "epoch": 0.21124247155922374,
+      "grad_norm": 0.23102112114429474,
+      "learning_rate": 1.9802449612861144e-05,
+      "loss": 0.5272,
+      "step": 947
+    },
+    {
+      "epoch": 0.2114655364711131,
+      "grad_norm": 0.17385543882846832,
+      "learning_rate": 1.9801983862996423e-05,
+      "loss": 0.5206,
+      "step": 948
+    },
+    {
+      "epoch": 0.21168860138300247,
+      "grad_norm": 0.20227134227752686,
+      "learning_rate": 1.980151757023618e-05,
+      "loss": 0.5362,
+      "step": 949
+    },
+    {
+      "epoch": 0.2119116662948918,
+      "grad_norm": 0.3727608025074005,
+      "learning_rate": 1.9801050734606236e-05,
+      "loss": 0.511,
+      "step": 950
+    },
+    {
+      "epoch": 0.21213473120678117,
+      "grad_norm": 0.17483538389205933,
+      "learning_rate": 1.9800583356132453e-05,
+      "loss": 0.5251,
+      "step": 951
+    },
+    {
+      "epoch": 0.21235779611867053,
+      "grad_norm": 0.17400579154491425,
+      "learning_rate": 1.9800115434840716e-05,
+      "loss": 0.541,
+      "step": 952
+    },
+    {
+      "epoch": 0.2125808610305599,
+      "grad_norm": 0.16707202792167664,
+      "learning_rate": 1.979964697075694e-05,
+      "loss": 0.5391,
+      "step": 953
+    },
+    {
+      "epoch": 0.21280392594244926,
+      "grad_norm": 0.16636815667152405,
+      "learning_rate": 1.9799177963907074e-05,
+      "loss": 0.5434,
+      "step": 954
+    },
+    {
+      "epoch": 0.2130269908543386,
+      "grad_norm": 0.17174044251441956,
+      "learning_rate": 1.9798708414317095e-05,
+      "loss": 0.5389,
+      "step": 955
+    },
+    {
+      "epoch": 0.21325005576622796,
+      "grad_norm": 0.16985206305980682,
+      "learning_rate": 1.9798238322013002e-05,
+      "loss": 0.5327,
+      "step": 956
+    },
+    {
+      "epoch": 0.21347312067811733,
+      "grad_norm": 0.17486560344696045,
+      "learning_rate": 1.9797767687020843e-05,
+      "loss": 0.5428,
+      "step": 957
+    },
+    {
+      "epoch": 0.2136961855900067,
+      "grad_norm": 0.17041227221488953,
+      "learning_rate": 1.9797296509366678e-05,
+      "loss": 0.4995,
+      "step": 958
+    },
+    {
+      "epoch": 0.21391925050189606,
+      "grad_norm": 0.1798069328069687,
+      "learning_rate": 1.97968247890766e-05,
+      "loss": 0.5364,
+      "step": 959
+    },
+    {
+      "epoch": 0.21414231541378542,
+      "grad_norm": 0.17648300528526306,
+      "learning_rate": 1.9796352526176746e-05,
+      "loss": 0.5317,
+      "step": 960
+    },
+    {
+      "epoch": 0.21436538032567476,
+      "grad_norm": 0.16647587716579437,
+      "learning_rate": 1.9795879720693264e-05,
+      "loss": 0.4989,
+      "step": 961
+    },
+    {
+      "epoch": 0.21458844523756412,
+      "grad_norm": 0.17619404196739197,
+      "learning_rate": 1.9795406372652345e-05,
+      "loss": 0.5123,
+      "step": 962
+    },
+    {
+      "epoch": 0.2148115101494535,
+      "grad_norm": 0.1789608895778656,
+      "learning_rate": 1.979493248208021e-05,
+      "loss": 0.5421,
+      "step": 963
+    },
+    {
+      "epoch": 0.21503457506134285,
+      "grad_norm": 0.16893987357616425,
+      "learning_rate": 1.97944580490031e-05,
+      "loss": 0.5226,
+      "step": 964
+    },
+    {
+      "epoch": 0.21525763997323222,
+      "grad_norm": 0.1653861552476883,
+      "learning_rate": 1.9793983073447288e-05,
+      "loss": 0.5221,
+      "step": 965
+    },
+    {
+      "epoch": 0.21548070488512158,
+      "grad_norm": 0.17093954980373383,
+      "learning_rate": 1.9793507555439092e-05,
+      "loss": 0.535,
+      "step": 966
+    },
+    {
+      "epoch": 0.21570376979701092,
+      "grad_norm": 0.18254542350769043,
+      "learning_rate": 1.9793031495004845e-05,
+      "loss": 0.5585,
+      "step": 967
+    },
+    {
+      "epoch": 0.21592683470890028,
+      "grad_norm": 0.19674773514270782,
+      "learning_rate": 1.9792554892170908e-05,
+      "loss": 0.5159,
+      "step": 968
+    },
+    {
+      "epoch": 0.21614989962078965,
+      "grad_norm": 0.17958855628967285,
+      "learning_rate": 1.9792077746963686e-05,
+      "loss": 0.5185,
+      "step": 969
+    },
+    {
+      "epoch": 0.216372964532679,
+      "grad_norm": 0.1741204559803009,
+      "learning_rate": 1.9791600059409606e-05,
+      "loss": 0.5325,
+      "step": 970
+    },
+    {
+      "epoch": 0.21659602944456838,
+      "grad_norm": 0.1755409985780716,
+      "learning_rate": 1.9791121829535122e-05,
+      "loss": 0.5005,
+      "step": 971
+    },
+    {
+      "epoch": 0.21681909435645774,
+      "grad_norm": 0.17128707468509674,
+      "learning_rate": 1.979064305736672e-05,
+      "loss": 0.5345,
+      "step": 972
+    },
+    {
+      "epoch": 0.21704215926834708,
+      "grad_norm": 0.17565475404262543,
+      "learning_rate": 1.9790163742930922e-05,
+      "loss": 0.4964,
+      "step": 973
+    },
+    {
+      "epoch": 0.21726522418023644,
+      "grad_norm": 0.1766713559627533,
+      "learning_rate": 1.978968388625427e-05,
+      "loss": 0.5151,
+      "step": 974
+    },
+    {
+      "epoch": 0.2174882890921258,
+      "grad_norm": 0.1699419617652893,
+      "learning_rate": 1.9789203487363352e-05,
+      "loss": 0.5365,
+      "step": 975
+    },
+    {
+      "epoch": 0.21771135400401517,
+      "grad_norm": 0.20117329061031342,
+      "learning_rate": 1.978872254628476e-05,
+      "loss": 0.4942,
+      "step": 976
+    },
+    {
+      "epoch": 0.21793441891590454,
+      "grad_norm": 0.17031805217266083,
+      "learning_rate": 1.9788241063045147e-05,
+      "loss": 0.5262,
+      "step": 977
+    },
+    {
+      "epoch": 0.21815748382779387,
+      "grad_norm": 0.17150211334228516,
+      "learning_rate": 1.9787759037671172e-05,
+      "loss": 0.5169,
+      "step": 978
+    },
+    {
+      "epoch": 0.21838054873968324,
+      "grad_norm": 0.17386960983276367,
+      "learning_rate": 1.978727647018953e-05,
+      "loss": 0.5464,
+      "step": 979
+    },
+    {
+      "epoch": 0.2186036136515726,
+      "grad_norm": 0.17337769269943237,
+      "learning_rate": 1.9786793360626956e-05,
+      "loss": 0.5217,
+      "step": 980
+    },
+    {
+      "epoch": 0.21882667856346197,
+      "grad_norm": 0.17415766417980194,
+      "learning_rate": 1.9786309709010204e-05,
+      "loss": 0.5222,
+      "step": 981
+    },
+    {
+      "epoch": 0.21904974347535133,
+      "grad_norm": 0.17464500665664673,
+      "learning_rate": 1.978582551536606e-05,
+      "loss": 0.504,
+      "step": 982
+    },
+    {
+      "epoch": 0.2192728083872407,
+      "grad_norm": 0.1703118085861206,
+      "learning_rate": 1.9785340779721348e-05,
+      "loss": 0.5419,
+      "step": 983
+    },
+    {
+      "epoch": 0.21949587329913003,
+      "grad_norm": 0.1777229905128479,
+      "learning_rate": 1.9784855502102908e-05,
+      "loss": 0.5396,
+      "step": 984
+    },
+    {
+      "epoch": 0.2197189382110194,
+      "grad_norm": 0.1821225881576538,
+      "learning_rate": 1.978436968253762e-05,
+      "loss": 0.5214,
+      "step": 985
+    },
+    {
+      "epoch": 0.21994200312290876,
+      "grad_norm": 0.17865097522735596,
+      "learning_rate": 1.9783883321052394e-05,
+      "loss": 0.5354,
+      "step": 986
+    },
+    {
+      "epoch": 0.22016506803479813,
+      "grad_norm": 0.17980031669139862,
+      "learning_rate": 1.978339641767417e-05,
+      "loss": 0.5461,
+      "step": 987
+    },
+    {
+      "epoch": 0.2203881329466875,
+      "grad_norm": 0.18906496465206146,
+      "learning_rate": 1.9782908972429906e-05,
+      "loss": 0.5466,
+      "step": 988
+    },
+    {
+      "epoch": 0.22061119785857686,
+      "grad_norm": 0.20680966973304749,
+      "learning_rate": 1.978242098534661e-05,
+      "loss": 0.4944,
+      "step": 989
+    },
+    {
+      "epoch": 0.2208342627704662,
+      "grad_norm": 0.18064817786216736,
+      "learning_rate": 1.978193245645131e-05,
+      "loss": 0.5418,
+      "step": 990
+    },
+    {
+      "epoch": 0.22105732768235556,
+      "grad_norm": 0.18148301541805267,
+      "learning_rate": 1.978144338577105e-05,
+      "loss": 0.5348,
+      "step": 991
+    },
+    {
+      "epoch": 0.22128039259424492,
+      "grad_norm": 0.18182723224163055,
+      "learning_rate": 1.9780953773332933e-05,
+      "loss": 0.5159,
+      "step": 992
+    },
+    {
+      "epoch": 0.2215034575061343,
+      "grad_norm": 0.17886720597743988,
+      "learning_rate": 1.9780463619164073e-05,
+      "loss": 0.5211,
+      "step": 993
+    },
+    {
+      "epoch": 0.22172652241802365,
+      "grad_norm": 0.1731138974428177,
+      "learning_rate": 1.9779972923291615e-05,
+      "loss": 0.5386,
+      "step": 994
+    },
+    {
+      "epoch": 0.22194958732991302,
+      "grad_norm": 0.17276941239833832,
+      "learning_rate": 1.977948168574274e-05,
+      "loss": 0.553,
+      "step": 995
+    },
+    {
+      "epoch": 0.22217265224180235,
+      "grad_norm": 0.1833122968673706,
+      "learning_rate": 1.977898990654465e-05,
+      "loss": 0.5082,
+      "step": 996
+    },
+    {
+      "epoch": 0.22239571715369172,
+      "grad_norm": 0.17474150657653809,
+      "learning_rate": 1.9778497585724586e-05,
+      "loss": 0.5167,
+      "step": 997
+    },
+    {
+      "epoch": 0.22261878206558108,
+      "grad_norm": 0.18032675981521606,
+      "learning_rate": 1.977800472330982e-05,
+      "loss": 0.5351,
+      "step": 998
+    },
+    {
+      "epoch": 0.22284184697747045,
+      "grad_norm": 0.19136174023151398,
+      "learning_rate": 1.9777511319327645e-05,
+      "loss": 0.5387,
+      "step": 999
+    },
+    {
+      "epoch": 0.2230649118893598,
+      "grad_norm": 0.17949000000953674,
+      "learning_rate": 1.977701737380539e-05,
+      "loss": 0.5132,
+      "step": 1000
+    },
+    {
+      "epoch": 0.22328797680124918,
+      "grad_norm": 0.16305360198020935,
+      "learning_rate": 1.9776522886770413e-05,
+      "loss": 0.4798,
+      "step": 1001
+    },
+    {
+      "epoch": 0.22351104171313851,
+      "grad_norm": 0.18127582967281342,
+      "learning_rate": 1.9776027858250102e-05,
+      "loss": 0.5314,
+      "step": 1002
+    },
+    {
+      "epoch": 0.22373410662502788,
+      "grad_norm": 0.17198516428470612,
+      "learning_rate": 1.9775532288271876e-05,
+      "loss": 0.5356,
+      "step": 1003
+    },
+    {
+      "epoch": 0.22395717153691724,
+      "grad_norm": 0.17598840594291687,
+      "learning_rate": 1.9775036176863178e-05,
+      "loss": 0.5243,
+      "step": 1004
+    },
+    {
+      "epoch": 0.2241802364488066,
+      "grad_norm": 0.17470763623714447,
+      "learning_rate": 1.977453952405149e-05,
+      "loss": 0.5138,
+      "step": 1005
+    },
+    {
+      "epoch": 0.22440330136069597,
+      "grad_norm": 0.17710572481155396,
+      "learning_rate": 1.977404232986432e-05,
+      "loss": 0.5103,
+      "step": 1006
+    },
+    {
+      "epoch": 0.2246263662725853,
+      "grad_norm": 0.16758602857589722,
+      "learning_rate": 1.9773544594329202e-05,
+      "loss": 0.5685,
+      "step": 1007
+    },
+    {
+      "epoch": 0.22484943118447467,
+      "grad_norm": 0.17261147499084473,
+      "learning_rate": 1.977304631747371e-05,
+      "loss": 0.5149,
+      "step": 1008
+    },
+    {
+      "epoch": 0.22507249609636404,
+      "grad_norm": 0.1722407191991806,
+      "learning_rate": 1.9772547499325437e-05,
+      "loss": 0.5277,
+      "step": 1009
+    },
+    {
+      "epoch": 0.2252955610082534,
+      "grad_norm": 0.17808939516544342,
+      "learning_rate": 1.9772048139912012e-05,
+      "loss": 0.5392,
+      "step": 1010
+    },
+    {
+      "epoch": 0.22551862592014277,
+      "grad_norm": 0.17333589494228363,
+      "learning_rate": 1.9771548239261088e-05,
+      "loss": 0.5347,
+      "step": 1011
+    },
+    {
+      "epoch": 0.22574169083203213,
+      "grad_norm": 0.1775379180908203,
+      "learning_rate": 1.9771047797400363e-05,
+      "loss": 0.5243,
+      "step": 1012
+    },
+    {
+      "epoch": 0.22596475574392147,
+      "grad_norm": 0.19107598066329956,
+      "learning_rate": 1.9770546814357546e-05,
+      "loss": 0.5464,
+      "step": 1013
+    },
+    {
+      "epoch": 0.22618782065581083,
+      "grad_norm": 0.1707477569580078,
+      "learning_rate": 1.9770045290160388e-05,
+      "loss": 0.5337,
+      "step": 1014
+    },
+    {
+      "epoch": 0.2264108855677002,
+      "grad_norm": 0.19580954313278198,
+      "learning_rate": 1.9769543224836668e-05,
+      "loss": 0.5138,
+      "step": 1015
+    },
+    {
+      "epoch": 0.22663395047958956,
+      "grad_norm": 0.17244374752044678,
+      "learning_rate": 1.9769040618414187e-05,
+      "loss": 0.5232,
+      "step": 1016
+    },
+    {
+      "epoch": 0.22685701539147893,
+      "grad_norm": 0.16292127966880798,
+      "learning_rate": 1.9768537470920788e-05,
+      "loss": 0.4989,
+      "step": 1017
+    },
+    {
+      "epoch": 0.2270800803033683,
+      "grad_norm": 0.18515698611736298,
+      "learning_rate": 1.9768033782384338e-05,
+      "loss": 0.5316,
+      "step": 1018
+    },
+    {
+      "epoch": 0.22730314521525763,
+      "grad_norm": 0.16262215375900269,
+      "learning_rate": 1.9767529552832732e-05,
+      "loss": 0.4855,
+      "step": 1019
+    },
+    {
+      "epoch": 0.227526210127147,
+      "grad_norm": 0.17309491336345673,
+      "learning_rate": 1.9767024782293902e-05,
+      "loss": 0.5041,
+      "step": 1020
+    },
+    {
+      "epoch": 0.22774927503903636,
+      "grad_norm": 0.1719563603401184,
+      "learning_rate": 1.9766519470795803e-05,
+      "loss": 0.5412,
+      "step": 1021
+    },
+    {
+      "epoch": 0.22797233995092572,
+      "grad_norm": 0.2674558460712433,
+      "learning_rate": 1.9766013618366417e-05,
+      "loss": 0.5274,
+      "step": 1022
+    },
+    {
+      "epoch": 0.2281954048628151,
+      "grad_norm": 0.16945339739322662,
+      "learning_rate": 1.9765507225033772e-05,
+      "loss": 0.5034,
+      "step": 1023
+    },
+    {
+      "epoch": 0.22841846977470445,
+      "grad_norm": 0.19735541939735413,
+      "learning_rate": 1.9765000290825908e-05,
+      "loss": 0.5059,
+      "step": 1024
+    },
+    {
+      "epoch": 0.2286415346865938,
+      "grad_norm": 0.17197129130363464,
+      "learning_rate": 1.97644928157709e-05,
+      "loss": 0.5133,
+      "step": 1025
+    },
+    {
+      "epoch": 0.22886459959848315,
+      "grad_norm": 0.18425902724266052,
+      "learning_rate": 1.976398479989686e-05,
+      "loss": 0.5062,
+      "step": 1026
+    },
+    {
+      "epoch": 0.22908766451037252,
+      "grad_norm": 0.4941766858100891,
+      "learning_rate": 1.9763476243231924e-05,
+      "loss": 0.535,
+      "step": 1027
+    },
+    {
+      "epoch": 0.22931072942226188,
+      "grad_norm": 0.16707998514175415,
+      "learning_rate": 1.976296714580426e-05,
+      "loss": 0.5177,
+      "step": 1028
+    },
+    {
+      "epoch": 0.22953379433415125,
+      "grad_norm": 0.173675537109375,
+      "learning_rate": 1.9762457507642066e-05,
+      "loss": 0.5234,
+      "step": 1029
+    },
+    {
+      "epoch": 0.22975685924604058,
+      "grad_norm": 0.18508677184581757,
+      "learning_rate": 1.9761947328773565e-05,
+      "loss": 0.5073,
+      "step": 1030
+    },
+    {
+      "epoch": 0.22997992415792995,
+      "grad_norm": 0.2566058039665222,
+      "learning_rate": 1.9761436609227016e-05,
+      "loss": 0.5176,
+      "step": 1031
+    },
+    {
+      "epoch": 0.2302029890698193,
+      "grad_norm": 0.1702575534582138,
+      "learning_rate": 1.9760925349030704e-05,
+      "loss": 0.5106,
+      "step": 1032
+    },
+    {
+      "epoch": 0.23042605398170868,
+      "grad_norm": 0.18858478963375092,
+      "learning_rate": 1.976041354821295e-05,
+      "loss": 0.5052,
+      "step": 1033
+    },
+    {
+      "epoch": 0.23064911889359804,
+      "grad_norm": 0.191171795129776,
+      "learning_rate": 1.9759901206802098e-05,
+      "loss": 0.5643,
+      "step": 1034
+    },
+    {
+      "epoch": 0.2308721838054874,
+      "grad_norm": 0.17536719143390656,
+      "learning_rate": 1.9759388324826523e-05,
+      "loss": 0.5344,
+      "step": 1035
+    },
+    {
+      "epoch": 0.23109524871737674,
+      "grad_norm": 0.21609671413898468,
+      "learning_rate": 1.9758874902314634e-05,
+      "loss": 0.5197,
+      "step": 1036
+    },
+    {
+      "epoch": 0.2313183136292661,
+      "grad_norm": 0.16850803792476654,
+      "learning_rate": 1.9758360939294867e-05,
+      "loss": 0.4967,
+      "step": 1037
+    },
+    {
+      "epoch": 0.23154137854115547,
+      "grad_norm": 0.19448642432689667,
+      "learning_rate": 1.9757846435795688e-05,
+      "loss": 0.5171,
+      "step": 1038
+    },
+    {
+      "epoch": 0.23176444345304484,
+      "grad_norm": 0.19304804503917694,
+      "learning_rate": 1.9757331391845596e-05,
+      "loss": 0.5491,
+      "step": 1039
+    },
+    {
+      "epoch": 0.2319875083649342,
+      "grad_norm": 0.17683899402618408,
+      "learning_rate": 1.975681580747312e-05,
+      "loss": 0.535,
+      "step": 1040
+    },
+    {
+      "epoch": 0.23221057327682357,
+      "grad_norm": 0.17681746184825897,
+      "learning_rate": 1.9756299682706804e-05,
+      "loss": 0.5053,
+      "step": 1041
+    },
+    {
+      "epoch": 0.2324336381887129,
+      "grad_norm": 0.17382597923278809,
+      "learning_rate": 1.9755783017575244e-05,
+      "loss": 0.5063,
+      "step": 1042
+    },
+    {
+      "epoch": 0.23265670310060227,
+      "grad_norm": 0.17841431498527527,
+      "learning_rate": 1.9755265812107053e-05,
+      "loss": 0.4845,
+      "step": 1043
+    },
+    {
+      "epoch": 0.23287976801249163,
+      "grad_norm": 0.18297810852527618,
+      "learning_rate": 1.9754748066330883e-05,
+      "loss": 0.5418,
+      "step": 1044
+    },
+    {
+      "epoch": 0.233102832924381,
+      "grad_norm": 0.19104409217834473,
+      "learning_rate": 1.97542297802754e-05,
+      "loss": 0.5417,
+      "step": 1045
+    },
+    {
+      "epoch": 0.23332589783627036,
+      "grad_norm": 0.1772020310163498,
+      "learning_rate": 1.975371095396932e-05,
+      "loss": 0.5296,
+      "step": 1046
+    },
+    {
+      "epoch": 0.23354896274815973,
+      "grad_norm": 0.18861500918865204,
+      "learning_rate": 1.9753191587441372e-05,
+      "loss": 0.5667,
+      "step": 1047
+    },
+    {
+      "epoch": 0.23377202766004906,
+      "grad_norm": 0.17586645483970642,
+      "learning_rate": 1.9752671680720324e-05,
+      "loss": 0.5001,
+      "step": 1048
+    },
+    {
+      "epoch": 0.23399509257193843,
+      "grad_norm": 0.7259854674339294,
+      "learning_rate": 1.975215123383497e-05,
+      "loss": 0.5456,
+      "step": 1049
+    },
+    {
+      "epoch": 0.2342181574838278,
+      "grad_norm": 0.18736739456653595,
+      "learning_rate": 1.9751630246814136e-05,
+      "loss": 0.5231,
+      "step": 1050
+    },
+    {
+      "epoch": 0.23444122239571716,
+      "grad_norm": 0.2538264989852905,
+      "learning_rate": 1.9751108719686683e-05,
+      "loss": 0.5387,
+      "step": 1051
+    },
+    {
+      "epoch": 0.23466428730760652,
+      "grad_norm": 0.17714229226112366,
+      "learning_rate": 1.9750586652481492e-05,
+      "loss": 0.5076,
+      "step": 1052
+    },
+    {
+      "epoch": 0.23488735221949586,
+      "grad_norm": 0.3667065501213074,
+      "learning_rate": 1.9750064045227474e-05,
+      "loss": 0.5432,
+      "step": 1053
+    },
+    {
+      "epoch": 0.23511041713138522,
+      "grad_norm": 0.18610940873622894,
+      "learning_rate": 1.9749540897953584e-05,
+      "loss": 0.5307,
+      "step": 1054
+    },
+    {
+      "epoch": 0.2353334820432746,
+      "grad_norm": 0.18711121380329132,
+      "learning_rate": 1.974901721068879e-05,
+      "loss": 0.5256,
+      "step": 1055
+    },
+    {
+      "epoch": 0.23555654695516395,
+      "grad_norm": 0.17803776264190674,
+      "learning_rate": 1.97484929834621e-05,
+      "loss": 0.5076,
+      "step": 1056
+    },
+    {
+      "epoch": 0.23577961186705332,
+      "grad_norm": 0.19948653876781464,
+      "learning_rate": 1.9747968216302545e-05,
+      "loss": 0.5185,
+      "step": 1057
+    },
+    {
+      "epoch": 0.23600267677894268,
+      "grad_norm": 0.3181793987751007,
+      "learning_rate": 1.9747442909239198e-05,
+      "loss": 0.4874,
+      "step": 1058
+    },
+    {
+      "epoch": 0.23622574169083202,
+      "grad_norm": 0.1865629106760025,
+      "learning_rate": 1.9746917062301146e-05,
+      "loss": 0.52,
+      "step": 1059
+    },
+    {
+      "epoch": 0.23644880660272138,
+      "grad_norm": 0.23274224996566772,
+      "learning_rate": 1.9746390675517514e-05,
+      "loss": 0.496,
+      "step": 1060
+    },
+    {
+      "epoch": 0.23667187151461075,
+      "grad_norm": 0.2044648379087448,
+      "learning_rate": 1.974586374891746e-05,
+      "loss": 0.5173,
+      "step": 1061
+    },
+    {
+      "epoch": 0.2368949364265001,
+      "grad_norm": 0.18317201733589172,
+      "learning_rate": 1.974533628253017e-05,
+      "loss": 0.5282,
+      "step": 1062
+    },
+    {
+      "epoch": 0.23711800133838948,
+      "grad_norm": 0.18429698050022125,
+      "learning_rate": 1.9744808276384858e-05,
+      "loss": 0.5395,
+      "step": 1063
+    },
+    {
+      "epoch": 0.23734106625027884,
+      "grad_norm": 0.20926502346992493,
+      "learning_rate": 1.9744279730510764e-05,
+      "loss": 0.5111,
+      "step": 1064
+    },
+    {
+      "epoch": 0.23756413116216818,
+      "grad_norm": 0.17823931574821472,
+      "learning_rate": 1.974375064493716e-05,
+      "loss": 0.5258,
+      "step": 1065
+    },
+    {
+      "epoch": 0.23778719607405754,
+      "grad_norm": 0.19109368324279785,
+      "learning_rate": 1.9743221019693362e-05,
+      "loss": 0.5512,
+      "step": 1066
+    },
+    {
+      "epoch": 0.2380102609859469,
+      "grad_norm": 0.21680758893489838,
+      "learning_rate": 1.9742690854808692e-05,
+      "loss": 0.4951,
+      "step": 1067
+    },
+    {
+      "epoch": 0.23823332589783627,
+      "grad_norm": 0.1768484115600586,
+      "learning_rate": 1.974216015031252e-05,
+      "loss": 0.534,
+      "step": 1068
+    },
+    {
+      "epoch": 0.23845639080972564,
+      "grad_norm": 0.18384598195552826,
+      "learning_rate": 1.974162890623424e-05,
+      "loss": 0.502,
+      "step": 1069
+    },
+    {
+      "epoch": 0.238679455721615,
+      "grad_norm": 0.17585726082324982,
+      "learning_rate": 1.974109712260327e-05,
+      "loss": 0.5177,
+      "step": 1070
+    },
+    {
+      "epoch": 0.23890252063350434,
+      "grad_norm": 0.17469032108783722,
+      "learning_rate": 1.9740564799449073e-05,
+      "loss": 0.553,
+      "step": 1071
+    },
+    {
+      "epoch": 0.2391255855453937,
+      "grad_norm": 0.21859519183635712,
+      "learning_rate": 1.9740031936801122e-05,
+      "loss": 0.5204,
+      "step": 1072
+    },
+    {
+      "epoch": 0.23934865045728307,
+      "grad_norm": 0.36654403805732727,
+      "learning_rate": 1.9739498534688936e-05,
+      "loss": 0.5375,
+      "step": 1073
+    },
+    {
+      "epoch": 0.23957171536917243,
+      "grad_norm": 0.17866800725460052,
+      "learning_rate": 1.973896459314206e-05,
+      "loss": 0.5477,
+      "step": 1074
+    },
+    {
+      "epoch": 0.2397947802810618,
+      "grad_norm": 0.1822461485862732,
+      "learning_rate": 1.973843011219006e-05,
+      "loss": 0.5197,
+      "step": 1075
+    },
+    {
+      "epoch": 0.24001784519295116,
+      "grad_norm": 0.17180079221725464,
+      "learning_rate": 1.9737895091862545e-05,
+      "loss": 0.5269,
+      "step": 1076
+    },
+    {
+      "epoch": 0.2402409101048405,
+      "grad_norm": 0.17935633659362793,
+      "learning_rate": 1.9737359532189147e-05,
+      "loss": 0.5279,
+      "step": 1077
+    },
+    {
+      "epoch": 0.24046397501672986,
+      "grad_norm": 0.22954648733139038,
+      "learning_rate": 1.9736823433199524e-05,
+      "loss": 0.5199,
+      "step": 1078
+    },
+    {
+      "epoch": 0.24068703992861923,
+      "grad_norm": 0.17607101798057556,
+      "learning_rate": 1.973628679492338e-05,
+      "loss": 0.5121,
+      "step": 1079
+    },
+    {
+      "epoch": 0.2409101048405086,
+      "grad_norm": 0.17458245158195496,
+      "learning_rate": 1.9735749617390422e-05,
+      "loss": 0.5363,
+      "step": 1080
+    },
+    {
+      "epoch": 0.24113316975239796,
+      "grad_norm": 0.17682591080665588,
+      "learning_rate": 1.9735211900630414e-05,
+      "loss": 0.5254,
+      "step": 1081
+    },
+    {
+      "epoch": 0.2413562346642873,
+      "grad_norm": 0.17350362241268158,
+      "learning_rate": 1.9734673644673133e-05,
+      "loss": 0.5381,
+      "step": 1082
+    },
+    {
+      "epoch": 0.24157929957617666,
+      "grad_norm": 0.17231930792331696,
+      "learning_rate": 1.973413484954839e-05,
+      "loss": 0.5118,
+      "step": 1083
+    },
+    {
+      "epoch": 0.24180236448806602,
+      "grad_norm": 0.21497975289821625,
+      "learning_rate": 1.9733595515286032e-05,
+      "loss": 0.5353,
+      "step": 1084
+    },
+    {
+      "epoch": 0.2420254293999554,
+      "grad_norm": 0.1669948548078537,
+      "learning_rate": 1.9733055641915926e-05,
+      "loss": 0.5216,
+      "step": 1085
+    },
+    {
+      "epoch": 0.24224849431184475,
+      "grad_norm": 0.18229244649410248,
+      "learning_rate": 1.9732515229467973e-05,
+      "loss": 0.5379,
+      "step": 1086
+    },
+    {
+      "epoch": 0.24247155922373412,
+      "grad_norm": 0.17432808876037598,
+      "learning_rate": 1.973197427797211e-05,
+      "loss": 0.5461,
+      "step": 1087
+    },
+    {
+      "epoch": 0.24269462413562345,
+      "grad_norm": 0.18571536242961884,
+      "learning_rate": 1.9731432787458294e-05,
+      "loss": 0.5469,
+      "step": 1088
+    },
+    {
+      "epoch": 0.24291768904751282,
+      "grad_norm": 0.18052545189857483,
+      "learning_rate": 1.9730890757956517e-05,
+      "loss": 0.5416,
+      "step": 1089
+    },
+    {
+      "epoch": 0.24314075395940218,
+      "grad_norm": 0.23118536174297333,
+      "learning_rate": 1.97303481894968e-05,
+      "loss": 0.5063,
+      "step": 1090
+    },
+    {
+      "epoch": 0.24336381887129155,
+      "grad_norm": 0.16191843152046204,
+      "learning_rate": 1.9729805082109194e-05,
+      "loss": 0.5178,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2435868837831809,
+      "grad_norm": 0.17042016983032227,
+      "learning_rate": 1.9729261435823782e-05,
+      "loss": 0.5024,
+      "step": 1092
+    },
+    {
+      "epoch": 0.24380994869507028,
+      "grad_norm": 0.21003969013690948,
+      "learning_rate": 1.972871725067067e-05,
+      "loss": 0.5106,
+      "step": 1093
+    },
+    {
+      "epoch": 0.24403301360695961,
+      "grad_norm": 0.22672854363918304,
+      "learning_rate": 1.972817252668e-05,
+      "loss": 0.5309,
+      "step": 1094
+    },
+    {
+      "epoch": 0.24425607851884898,
+      "grad_norm": 0.16994351148605347,
+      "learning_rate": 1.9727627263881942e-05,
+      "loss": 0.5023,
+      "step": 1095
+    },
+    {
+      "epoch": 0.24447914343073834,
+      "grad_norm": 0.17019368708133698,
+      "learning_rate": 1.9727081462306697e-05,
+      "loss": 0.5316,
+      "step": 1096
+    },
+    {
+      "epoch": 0.2447022083426277,
+      "grad_norm": 0.17145198583602905,
+      "learning_rate": 1.97265351219845e-05,
+      "loss": 0.5148,
+      "step": 1097
+    },
+    {
+      "epoch": 0.24492527325451707,
+      "grad_norm": 0.20565351843833923,
+      "learning_rate": 1.9725988242945598e-05,
+      "loss": 0.5445,
+      "step": 1098
+    },
+    {
+      "epoch": 0.24514833816640644,
+      "grad_norm": 0.1741577833890915,
+      "learning_rate": 1.9725440825220296e-05,
+      "loss": 0.4958,
+      "step": 1099
+    },
+    {
+      "epoch": 0.24537140307829577,
+      "grad_norm": 0.18705230951309204,
+      "learning_rate": 1.9724892868838902e-05,
+      "loss": 0.5105,
+      "step": 1100
+    },
+    {
+      "epoch": 0.24559446799018514,
+      "grad_norm": 0.5166819095611572,
+      "learning_rate": 1.9724344373831768e-05,
+      "loss": 0.5414,
+      "step": 1101
+    },
+    {
+      "epoch": 0.2458175329020745,
+      "grad_norm": 0.17351533472537994,
+      "learning_rate": 1.9723795340229274e-05,
+      "loss": 0.5024,
+      "step": 1102
+    },
+    {
+      "epoch": 0.24604059781396387,
+      "grad_norm": 0.1658518761396408,
+      "learning_rate": 1.972324576806183e-05,
+      "loss": 0.4977,
+      "step": 1103
+    },
+    {
+      "epoch": 0.24626366272585323,
+      "grad_norm": 0.16890370845794678,
+      "learning_rate": 1.972269565735987e-05,
+      "loss": 0.5359,
+      "step": 1104
+    },
+    {
+      "epoch": 0.24648672763774257,
+      "grad_norm": 0.18016333878040314,
+      "learning_rate": 1.9722145008153873e-05,
+      "loss": 0.5394,
+      "step": 1105
+    },
+    {
+      "epoch": 0.24670979254963193,
+      "grad_norm": 0.16652169823646545,
+      "learning_rate": 1.9721593820474326e-05,
+      "loss": 0.5089,
+      "step": 1106
+    },
+    {
+      "epoch": 0.2469328574615213,
+      "grad_norm": 0.1999325454235077,
+      "learning_rate": 1.9721042094351764e-05,
+      "loss": 0.5541,
+      "step": 1107
+    },
+    {
+      "epoch": 0.24715592237341066,
+      "grad_norm": 0.16372385621070862,
+      "learning_rate": 1.972048982981674e-05,
+      "loss": 0.4942,
+      "step": 1108
+    },
+    {
+      "epoch": 0.24737898728530003,
+      "grad_norm": 0.20245228707790375,
+      "learning_rate": 1.971993702689985e-05,
+      "loss": 0.5034,
+      "step": 1109
+    },
+    {
+      "epoch": 0.2476020521971894,
+      "grad_norm": 0.17487917840480804,
+      "learning_rate": 1.97193836856317e-05,
+      "loss": 0.5231,
+      "step": 1110
+    },
+    {
+      "epoch": 0.24782511710907873,
+      "grad_norm": 0.1696334332227707,
+      "learning_rate": 1.971882980604295e-05,
+      "loss": 0.5141,
+      "step": 1111
+    },
+    {
+      "epoch": 0.2480481820209681,
+      "grad_norm": 0.17021594941616058,
+      "learning_rate": 1.971827538816427e-05,
+      "loss": 0.5081,
+      "step": 1112
+    },
+    {
+      "epoch": 0.24827124693285746,
+      "grad_norm": 0.18088696897029877,
+      "learning_rate": 1.9717720432026367e-05,
+      "loss": 0.5743,
+      "step": 1113
+    },
+    {
+      "epoch": 0.24849431184474682,
+      "grad_norm": 0.17647258937358856,
+      "learning_rate": 1.9717164937659984e-05,
+      "loss": 0.5289,
+      "step": 1114
+    },
+    {
+      "epoch": 0.2487173767566362,
+      "grad_norm": 0.17265263199806213,
+      "learning_rate": 1.971660890509588e-05,
+      "loss": 0.5296,
+      "step": 1115
+    },
+    {
+      "epoch": 0.24894044166852555,
+      "grad_norm": 0.16816379129886627,
+      "learning_rate": 1.971605233436485e-05,
+      "loss": 0.5263,
+      "step": 1116
+    },
+    {
+      "epoch": 0.2491635065804149,
+      "grad_norm": 0.17548470199108124,
+      "learning_rate": 1.9715495225497736e-05,
+      "loss": 0.5315,
+      "step": 1117
+    },
+    {
+      "epoch": 0.24938657149230425,
+      "grad_norm": 0.1775428056716919,
+      "learning_rate": 1.9714937578525374e-05,
+      "loss": 0.5227,
+      "step": 1118
+    },
+    {
+      "epoch": 0.24960963640419362,
+      "grad_norm": 0.17629674077033997,
+      "learning_rate": 1.971437939347866e-05,
+      "loss": 0.5475,
+      "step": 1119
+    },
+    {
+      "epoch": 0.24983270131608298,
+      "grad_norm": 0.20265498757362366,
+      "learning_rate": 1.9713820670388518e-05,
+      "loss": 0.5415,
+      "step": 1120
+    },
+    {
+      "epoch": 0.2500557662279723,
+      "grad_norm": 0.17958010733127594,
+      "learning_rate": 1.9713261409285876e-05,
+      "loss": 0.5491,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2502788311398617,
+      "grad_norm": 0.1667289137840271,
+      "learning_rate": 1.9712701610201723e-05,
+      "loss": 0.5319,
+      "step": 1122
+    },
+    {
+      "epoch": 0.25050189605175105,
+      "grad_norm": 0.17537854611873627,
+      "learning_rate": 1.9712141273167058e-05,
+      "loss": 0.5033,
+      "step": 1123
+    },
+    {
+      "epoch": 0.25072496096364044,
+      "grad_norm": 0.1665065735578537,
+      "learning_rate": 1.9711580398212918e-05,
+      "loss": 0.5043,
+      "step": 1124
+    },
+    {
+      "epoch": 0.2509480258755298,
+      "grad_norm": 0.17893299460411072,
+      "learning_rate": 1.9711018985370366e-05,
+      "loss": 0.5424,
+      "step": 1125
+    },
+    {
+      "epoch": 0.2511710907874191,
+      "grad_norm": 0.17293697595596313,
+      "learning_rate": 1.97104570346705e-05,
+      "loss": 0.546,
+      "step": 1126
+    },
+    {
+      "epoch": 0.2513941556993085,
+      "grad_norm": 0.16615547239780426,
+      "learning_rate": 1.970989454614444e-05,
+      "loss": 0.5163,
+      "step": 1127
+    },
+    {
+      "epoch": 0.25161722061119784,
+      "grad_norm": 0.16944670677185059,
+      "learning_rate": 1.9709331519823343e-05,
+      "loss": 0.5471,
+      "step": 1128
+    },
+    {
+      "epoch": 0.25184028552308724,
+      "grad_norm": 0.179295152425766,
+      "learning_rate": 1.9708767955738394e-05,
+      "loss": 0.5381,
+      "step": 1129
+    },
+    {
+      "epoch": 0.2520633504349766,
+      "grad_norm": 0.1665600687265396,
+      "learning_rate": 1.9708203853920803e-05,
+      "loss": 0.4863,
+      "step": 1130
+    },
+    {
+      "epoch": 0.2522864153468659,
+      "grad_norm": 0.1918114423751831,
+      "learning_rate": 1.970763921440182e-05,
+      "loss": 0.4875,
+      "step": 1131
+    },
+    {
+      "epoch": 0.2525094802587553,
+      "grad_norm": 0.16761860251426697,
+      "learning_rate": 1.9707074037212707e-05,
+      "loss": 0.5314,
+      "step": 1132
+    },
+    {
+      "epoch": 0.25273254517064464,
+      "grad_norm": 0.19442524015903473,
+      "learning_rate": 1.970650832238478e-05,
+      "loss": 0.5186,
+      "step": 1133
+    },
+    {
+      "epoch": 0.25295561008253403,
+      "grad_norm": 0.17174071073532104,
+      "learning_rate": 1.9705942069949362e-05,
+      "loss": 0.5327,
+      "step": 1134
+    },
+    {
+      "epoch": 0.25317867499442337,
+      "grad_norm": 0.18218085169792175,
+      "learning_rate": 1.970537527993782e-05,
+      "loss": 0.5665,
+      "step": 1135
+    },
+    {
+      "epoch": 0.25340173990631276,
+      "grad_norm": 0.17595386505126953,
+      "learning_rate": 1.9704807952381542e-05,
+      "loss": 0.5581,
+      "step": 1136
+    },
+    {
+      "epoch": 0.2536248048182021,
+      "grad_norm": 0.20507752895355225,
+      "learning_rate": 1.9704240087311963e-05,
+      "loss": 0.541,
+      "step": 1137
+    },
+    {
+      "epoch": 0.25384786973009144,
+      "grad_norm": 0.1604541391134262,
+      "learning_rate": 1.970367168476052e-05,
+      "loss": 0.5223,
+      "step": 1138
+    },
+    {
+      "epoch": 0.25407093464198083,
+      "grad_norm": 0.16794823110103607,
+      "learning_rate": 1.9703102744758703e-05,
+      "loss": 0.5444,
+      "step": 1139
+    },
+    {
+      "epoch": 0.25429399955387016,
+      "grad_norm": 0.16887861490249634,
+      "learning_rate": 1.9702533267338015e-05,
+      "loss": 0.5237,
+      "step": 1140
+    },
+    {
+      "epoch": 0.25451706446575956,
+      "grad_norm": 0.16295503079891205,
+      "learning_rate": 1.970196325253001e-05,
+      "loss": 0.5293,
+      "step": 1141
+    },
+    {
+      "epoch": 0.2547401293776489,
+      "grad_norm": 0.16665813326835632,
+      "learning_rate": 1.9701392700366247e-05,
+      "loss": 0.5104,
+      "step": 1142
+    },
+    {
+      "epoch": 0.25496319428953823,
+      "grad_norm": 0.16648006439208984,
+      "learning_rate": 1.970082161087834e-05,
+      "loss": 0.5001,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2551862592014276,
+      "grad_norm": 0.16400828957557678,
+      "learning_rate": 1.9700249984097907e-05,
+      "loss": 0.5106,
+      "step": 1144
+    },
+    {
+      "epoch": 0.25540932411331696,
+      "grad_norm": 0.16578614711761475,
+      "learning_rate": 1.969967782005661e-05,
+      "loss": 0.5134,
+      "step": 1145
+    },
+    {
+      "epoch": 0.25563238902520635,
+      "grad_norm": 0.16847828030586243,
+      "learning_rate": 1.9699105118786145e-05,
+      "loss": 0.4994,
+      "step": 1146
+    },
+    {
+      "epoch": 0.2558554539370957,
+      "grad_norm": 0.18443261086940765,
+      "learning_rate": 1.9698531880318228e-05,
+      "loss": 0.5136,
+      "step": 1147
+    },
+    {
+      "epoch": 0.2560785188489851,
+      "grad_norm": 0.17291708290576935,
+      "learning_rate": 1.969795810468461e-05,
+      "loss": 0.5243,
+      "step": 1148
+    },
+    {
+      "epoch": 0.2563015837608744,
+      "grad_norm": 0.18102054297924042,
+      "learning_rate": 1.9697383791917068e-05,
+      "loss": 0.4966,
+      "step": 1149
+    },
+    {
+      "epoch": 0.25652464867276376,
+      "grad_norm": 0.17706115543842316,
+      "learning_rate": 1.9696808942047414e-05,
+      "loss": 0.5332,
+      "step": 1150
+    },
+    {
+      "epoch": 0.25674771358465315,
+      "grad_norm": 0.1936006247997284,
+      "learning_rate": 1.9696233555107484e-05,
+      "loss": 0.5325,
+      "step": 1151
+    },
+    {
+      "epoch": 0.2569707784965425,
+      "grad_norm": 0.19508236646652222,
+      "learning_rate": 1.969565763112915e-05,
+      "loss": 0.509,
+      "step": 1152
+    },
+    {
+      "epoch": 0.2571938434084319,
+      "grad_norm": 0.16369056701660156,
+      "learning_rate": 1.9695081170144306e-05,
+      "loss": 0.5188,
+      "step": 1153
+    },
+    {
+      "epoch": 0.2574169083203212,
+      "grad_norm": 0.16955684125423431,
+      "learning_rate": 1.9694504172184885e-05,
+      "loss": 0.5316,
+      "step": 1154
+    },
+    {
+      "epoch": 0.25763997323221055,
+      "grad_norm": 0.16252164542675018,
+      "learning_rate": 1.969392663728284e-05,
+      "loss": 0.5291,
+      "step": 1155
+    },
+    {
+      "epoch": 0.25786303814409994,
+      "grad_norm": 0.17399340867996216,
+      "learning_rate": 1.969334856547016e-05,
+      "loss": 0.5153,
+      "step": 1156
+    },
+    {
+      "epoch": 0.2580861030559893,
+      "grad_norm": 0.17638690769672394,
+      "learning_rate": 1.9692769956778867e-05,
+      "loss": 0.5128,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2583091679678787,
+      "grad_norm": 0.1711435467004776,
+      "learning_rate": 1.9692190811241e-05,
+      "loss": 0.5205,
+      "step": 1158
+    },
+    {
+      "epoch": 0.258532232879768,
+      "grad_norm": 0.18419131636619568,
+      "learning_rate": 1.9691611128888643e-05,
+      "loss": 0.5312,
+      "step": 1159
+    },
+    {
+      "epoch": 0.25875529779165735,
+      "grad_norm": 0.1925041675567627,
+      "learning_rate": 1.9691030909753894e-05,
+      "loss": 0.5479,
+      "step": 1160
+    },
+    {
+      "epoch": 0.25897836270354674,
+      "grad_norm": 0.1986900418996811,
+      "learning_rate": 1.9690450153868895e-05,
+      "loss": 0.5095,
+      "step": 1161
+    },
+    {
+      "epoch": 0.2592014276154361,
+      "grad_norm": 0.17978021502494812,
+      "learning_rate": 1.9689868861265816e-05,
+      "loss": 0.5015,
+      "step": 1162
+    },
+    {
+      "epoch": 0.25942449252732547,
+      "grad_norm": 0.16919931769371033,
+      "learning_rate": 1.9689287031976845e-05,
+      "loss": 0.5227,
+      "step": 1163
+    },
+    {
+      "epoch": 0.2596475574392148,
+      "grad_norm": 0.17485311627388,
+      "learning_rate": 1.9688704666034208e-05,
+      "loss": 0.5386,
+      "step": 1164
+    },
+    {
+      "epoch": 0.2598706223511042,
+      "grad_norm": 0.17683085799217224,
+      "learning_rate": 1.9688121763470165e-05,
+      "loss": 0.5201,
+      "step": 1165
+    },
+    {
+      "epoch": 0.26009368726299353,
+      "grad_norm": 0.18902894854545593,
+      "learning_rate": 1.9687538324316997e-05,
+      "loss": 0.51,
+      "step": 1166
+    },
+    {
+      "epoch": 0.26031675217488287,
+      "grad_norm": 0.17308862507343292,
+      "learning_rate": 1.968695434860702e-05,
+      "loss": 0.5112,
+      "step": 1167
+    },
+    {
+      "epoch": 0.26053981708677226,
+      "grad_norm": 0.17456986010074615,
+      "learning_rate": 1.9686369836372577e-05,
+      "loss": 0.5182,
+      "step": 1168
+    },
+    {
+      "epoch": 0.2607628819986616,
+      "grad_norm": 0.17276804149150848,
+      "learning_rate": 1.9685784787646044e-05,
+      "loss": 0.5389,
+      "step": 1169
+    },
+    {
+      "epoch": 0.260985946910551,
+      "grad_norm": 0.1710672229528427,
+      "learning_rate": 1.9685199202459824e-05,
+      "loss": 0.4948,
+      "step": 1170
+    },
+    {
+      "epoch": 0.26120901182244033,
+      "grad_norm": 0.18158847093582153,
+      "learning_rate": 1.9684613080846347e-05,
+      "loss": 0.5332,
+      "step": 1171
+    },
+    {
+      "epoch": 0.26143207673432967,
+      "grad_norm": 0.17782646417617798,
+      "learning_rate": 1.968402642283808e-05,
+      "loss": 0.5519,
+      "step": 1172
+    },
+    {
+      "epoch": 0.26165514164621906,
+      "grad_norm": 0.1716913878917694,
+      "learning_rate": 1.9683439228467515e-05,
+      "loss": 0.5074,
+      "step": 1173
+    },
+    {
+      "epoch": 0.2618782065581084,
+      "grad_norm": 0.1818932741880417,
+      "learning_rate": 1.9682851497767175e-05,
+      "loss": 0.5401,
+      "step": 1174
+    },
+    {
+      "epoch": 0.2621012714699978,
+      "grad_norm": 0.1722068190574646,
+      "learning_rate": 1.9682263230769612e-05,
+      "loss": 0.52,
+      "step": 1175
+    },
+    {
+      "epoch": 0.2623243363818871,
+      "grad_norm": 0.17454420030117035,
+      "learning_rate": 1.9681674427507405e-05,
+      "loss": 0.5063,
+      "step": 1176
+    },
+    {
+      "epoch": 0.2625474012937765,
+      "grad_norm": 0.1726730465888977,
+      "learning_rate": 1.9681085088013174e-05,
+      "loss": 0.5343,
+      "step": 1177
+    },
+    {
+      "epoch": 0.26277046620566585,
+      "grad_norm": 0.1802458018064499,
+      "learning_rate": 1.9680495212319547e-05,
+      "loss": 0.5218,
+      "step": 1178
+    },
+    {
+      "epoch": 0.2629935311175552,
+      "grad_norm": 0.1708204448223114,
+      "learning_rate": 1.9679904800459205e-05,
+      "loss": 0.4943,
+      "step": 1179
+    },
+    {
+      "epoch": 0.2632165960294446,
+      "grad_norm": 0.1644791066646576,
+      "learning_rate": 1.9679313852464846e-05,
+      "loss": 0.534,
+      "step": 1180
+    },
+    {
+      "epoch": 0.2634396609413339,
+      "grad_norm": 0.16891272366046906,
+      "learning_rate": 1.9678722368369203e-05,
+      "loss": 0.519,
+      "step": 1181
+    },
+    {
+      "epoch": 0.2636627258532233,
+      "grad_norm": 0.1720377802848816,
+      "learning_rate": 1.9678130348205032e-05,
+      "loss": 0.5362,
+      "step": 1182
+    },
+    {
+      "epoch": 0.26388579076511265,
+      "grad_norm": 0.17827892303466797,
+      "learning_rate": 1.9677537792005124e-05,
+      "loss": 0.5387,
+      "step": 1183
+    },
+    {
+      "epoch": 0.264108855677002,
+      "grad_norm": 0.17692813277244568,
+      "learning_rate": 1.96769446998023e-05,
+      "loss": 0.5051,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2643319205888914,
+      "grad_norm": 0.17577511072158813,
+      "learning_rate": 1.9676351071629405e-05,
+      "loss": 0.5162,
+      "step": 1185
+    },
+    {
+      "epoch": 0.2645549855007807,
+      "grad_norm": 0.1805581897497177,
+      "learning_rate": 1.9675756907519325e-05,
+      "loss": 0.5528,
+      "step": 1186
+    },
+    {
+      "epoch": 0.2647780504126701,
+      "grad_norm": 0.17259716987609863,
+      "learning_rate": 1.967516220750496e-05,
+      "loss": 0.5381,
+      "step": 1187
+    },
+    {
+      "epoch": 0.26500111532455944,
+      "grad_norm": 0.16992929577827454,
+      "learning_rate": 1.9674566971619256e-05,
+      "loss": 0.51,
+      "step": 1188
+    },
+    {
+      "epoch": 0.2652241802364488,
+      "grad_norm": 0.16945737600326538,
+      "learning_rate": 1.9673971199895177e-05,
+      "loss": 0.4987,
+      "step": 1189
+    },
+    {
+      "epoch": 0.2654472451483382,
+      "grad_norm": 0.16880926489830017,
+      "learning_rate": 1.967337489236572e-05,
+      "loss": 0.5188,
+      "step": 1190
+    },
+    {
+      "epoch": 0.2656703100602275,
+      "grad_norm": 0.17628756165504456,
+      "learning_rate": 1.9672778049063915e-05,
+      "loss": 0.5346,
+      "step": 1191
+    },
+    {
+      "epoch": 0.2658933749721169,
+      "grad_norm": 0.17406663298606873,
+      "learning_rate": 1.967218067002282e-05,
+      "loss": 0.5493,
+      "step": 1192
+    },
+    {
+      "epoch": 0.26611643988400624,
+      "grad_norm": 0.17142999172210693,
+      "learning_rate": 1.9671582755275515e-05,
+      "loss": 0.5256,
+      "step": 1193
+    },
+    {
+      "epoch": 0.26633950479589563,
+      "grad_norm": 0.17752200365066528,
+      "learning_rate": 1.9670984304855125e-05,
+      "loss": 0.5237,
+      "step": 1194
+    },
+    {
+      "epoch": 0.26656256970778497,
+      "grad_norm": 0.17670097947120667,
+      "learning_rate": 1.9670385318794785e-05,
+      "loss": 0.5226,
+      "step": 1195
+    },
+    {
+      "epoch": 0.2667856346196743,
+      "grad_norm": 0.1656324565410614,
+      "learning_rate": 1.966978579712768e-05,
+      "loss": 0.494,
+      "step": 1196
+    },
+    {
+      "epoch": 0.2670086995315637,
+      "grad_norm": 0.17369796335697174,
+      "learning_rate": 1.966918573988701e-05,
+      "loss": 0.506,
+      "step": 1197
+    },
+    {
+      "epoch": 0.26723176444345303,
+      "grad_norm": 0.17040349543094635,
+      "learning_rate": 1.9668585147106017e-05,
+      "loss": 0.5112,
+      "step": 1198
+    },
+    {
+      "epoch": 0.2674548293553424,
+      "grad_norm": 0.1750401258468628,
+      "learning_rate": 1.9667984018817957e-05,
+      "loss": 0.5465,
+      "step": 1199
+    },
+    {
+      "epoch": 0.26767789426723176,
+      "grad_norm": 0.16083772480487823,
+      "learning_rate": 1.9667382355056128e-05,
+      "loss": 0.5105,
+      "step": 1200
+    },
+    {
+      "epoch": 0.2679009591791211,
+      "grad_norm": 0.16108831763267517,
+      "learning_rate": 1.9666780155853854e-05,
+      "loss": 0.512,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2681240240910105,
+      "grad_norm": 0.17286653816699982,
+      "learning_rate": 1.966617742124449e-05,
+      "loss": 0.5686,
+      "step": 1202
+    },
+    {
+      "epoch": 0.26834708900289983,
+      "grad_norm": 0.17905832827091217,
+      "learning_rate": 1.9665574151261418e-05,
+      "loss": 0.5423,
+      "step": 1203
+    },
+    {
+      "epoch": 0.2685701539147892,
+      "grad_norm": 0.16681291162967682,
+      "learning_rate": 1.966497034593805e-05,
+      "loss": 0.5041,
+      "step": 1204
+    },
+    {
+      "epoch": 0.26879321882667856,
+      "grad_norm": 0.17184089124202728,
+      "learning_rate": 1.9664366005307828e-05,
+      "loss": 0.5501,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2690162837385679,
+      "grad_norm": 0.19392690062522888,
+      "learning_rate": 1.9663761129404228e-05,
+      "loss": 0.5398,
+      "step": 1206
+    },
+    {
+      "epoch": 0.2692393486504573,
+      "grad_norm": 0.16529332101345062,
+      "learning_rate": 1.9663155718260746e-05,
+      "loss": 0.5086,
+      "step": 1207
+    },
+    {
+      "epoch": 0.2694624135623466,
+      "grad_norm": 0.1649094521999359,
+      "learning_rate": 1.966254977191092e-05,
+      "loss": 0.515,
+      "step": 1208
+    },
+    {
+      "epoch": 0.269685478474236,
+      "grad_norm": 0.1709623634815216,
+      "learning_rate": 1.9661943290388302e-05,
+      "loss": 0.5315,
+      "step": 1209
+    },
+    {
+      "epoch": 0.26990854338612535,
+      "grad_norm": 0.164725661277771,
+      "learning_rate": 1.9661336273726496e-05,
+      "loss": 0.4817,
+      "step": 1210
+    },
+    {
+      "epoch": 0.27013160829801475,
+      "grad_norm": 0.1715349704027176,
+      "learning_rate": 1.966072872195911e-05,
+      "loss": 0.4706,
+      "step": 1211
+    },
+    {
+      "epoch": 0.2703546732099041,
+      "grad_norm": 0.1747862696647644,
+      "learning_rate": 1.9660120635119798e-05,
+      "loss": 0.5055,
+      "step": 1212
+    },
+    {
+      "epoch": 0.2705777381217934,
+      "grad_norm": 0.17992718517780304,
+      "learning_rate": 1.9659512013242245e-05,
+      "loss": 0.4946,
+      "step": 1213
+    },
+    {
+      "epoch": 0.2708008030336828,
+      "grad_norm": 0.1895056813955307,
+      "learning_rate": 1.9658902856360153e-05,
+      "loss": 0.5258,
+      "step": 1214
+    },
+    {
+      "epoch": 0.27102386794557215,
+      "grad_norm": 0.19161482155323029,
+      "learning_rate": 1.9658293164507265e-05,
+      "loss": 0.5206,
+      "step": 1215
+    },
+    {
+      "epoch": 0.27124693285746154,
+      "grad_norm": 0.1860935389995575,
+      "learning_rate": 1.965768293771735e-05,
+      "loss": 0.5484,
+      "step": 1216
+    },
+    {
+      "epoch": 0.2714699977693509,
+      "grad_norm": 0.17763131856918335,
+      "learning_rate": 1.9657072176024202e-05,
+      "loss": 0.5029,
+      "step": 1217
+    },
+    {
+      "epoch": 0.2716930626812402,
+      "grad_norm": 0.18652020394802094,
+      "learning_rate": 1.9656460879461652e-05,
+      "loss": 0.5161,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2719161275931296,
+      "grad_norm": 0.2644646465778351,
+      "learning_rate": 1.965584904806356e-05,
+      "loss": 0.5371,
+      "step": 1219
+    },
+    {
+      "epoch": 0.27213919250501895,
+      "grad_norm": 0.1769624799489975,
+      "learning_rate": 1.9655236681863806e-05,
+      "loss": 0.4892,
+      "step": 1220
+    },
+    {
+      "epoch": 0.27236225741690834,
+      "grad_norm": 0.1740611046552658,
+      "learning_rate": 1.9654623780896313e-05,
+      "loss": 0.5162,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2725853223287977,
+      "grad_norm": 0.16314013302326202,
+      "learning_rate": 1.9654010345195026e-05,
+      "loss": 0.5304,
+      "step": 1222
+    },
+    {
+      "epoch": 0.27280838724068707,
+      "grad_norm": 0.18545354902744293,
+      "learning_rate": 1.9653396374793915e-05,
+      "loss": 0.5181,
+      "step": 1223
+    },
+    {
+      "epoch": 0.2730314521525764,
+      "grad_norm": 0.17202220857143402,
+      "learning_rate": 1.9652781869726993e-05,
+      "loss": 0.5486,
+      "step": 1224
+    },
+    {
+      "epoch": 0.27325451706446574,
+      "grad_norm": 0.18883365392684937,
+      "learning_rate": 1.9652166830028295e-05,
+      "loss": 0.5565,
+      "step": 1225
+    },
+    {
+      "epoch": 0.27347758197635513,
+      "grad_norm": 0.17693057656288147,
+      "learning_rate": 1.9651551255731884e-05,
+      "loss": 0.5275,
+      "step": 1226
+    },
+    {
+      "epoch": 0.27370064688824447,
+      "grad_norm": 0.17243415117263794,
+      "learning_rate": 1.9650935146871848e-05,
+      "loss": 0.5037,
+      "step": 1227
+    },
+    {
+      "epoch": 0.27392371180013386,
+      "grad_norm": 0.18462933599948883,
+      "learning_rate": 1.9650318503482323e-05,
+      "loss": 0.5573,
+      "step": 1228
+    },
+    {
+      "epoch": 0.2741467767120232,
+      "grad_norm": 0.3310684263706207,
+      "learning_rate": 1.964970132559745e-05,
+      "loss": 0.5186,
+      "step": 1229
+    },
+    {
+      "epoch": 0.27436984162391254,
+      "grad_norm": 0.16663436591625214,
+      "learning_rate": 1.964908361325142e-05,
+      "loss": 0.5067,
+      "step": 1230
+    },
+    {
+      "epoch": 0.27459290653580193,
+      "grad_norm": 0.16989050805568695,
+      "learning_rate": 1.964846536647845e-05,
+      "loss": 0.5084,
+      "step": 1231
+    },
+    {
+      "epoch": 0.27481597144769127,
+      "grad_norm": 0.16925394535064697,
+      "learning_rate": 1.9647846585312775e-05,
+      "loss": 0.5348,
+      "step": 1232
+    },
+    {
+      "epoch": 0.27503903635958066,
+      "grad_norm": 0.176070898771286,
+      "learning_rate": 1.9647227269788665e-05,
+      "loss": 0.5249,
+      "step": 1233
+    },
+    {
+      "epoch": 0.27526210127147,
+      "grad_norm": 0.16656488180160522,
+      "learning_rate": 1.9646607419940428e-05,
+      "loss": 0.5408,
+      "step": 1234
+    },
+    {
+      "epoch": 0.27548516618335933,
+      "grad_norm": 0.16989517211914062,
+      "learning_rate": 1.964598703580239e-05,
+      "loss": 0.5033,
+      "step": 1235
+    },
+    {
+      "epoch": 0.2757082310952487,
+      "grad_norm": 0.18006569147109985,
+      "learning_rate": 1.9645366117408918e-05,
+      "loss": 0.5354,
+      "step": 1236
+    },
+    {
+      "epoch": 0.27593129600713806,
+      "grad_norm": 0.17241814732551575,
+      "learning_rate": 1.9644744664794394e-05,
+      "loss": 0.508,
+      "step": 1237
+    },
+    {
+      "epoch": 0.27615436091902745,
+      "grad_norm": 0.20181423425674438,
+      "learning_rate": 1.9644122677993246e-05,
+      "loss": 0.4536,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2763774258309168,
+      "grad_norm": 0.17486101388931274,
+      "learning_rate": 1.964350015703992e-05,
+      "loss": 0.5597,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2766004907428062,
+      "grad_norm": 0.1774255484342575,
+      "learning_rate": 1.9642877101968894e-05,
+      "loss": 0.5511,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2768235556546955,
+      "grad_norm": 0.3290056586265564,
+      "learning_rate": 1.964225351281468e-05,
+      "loss": 0.5278,
+      "step": 1241
+    },
+    {
+      "epoch": 0.27704662056658486,
+      "grad_norm": 0.20791974663734436,
+      "learning_rate": 1.9641629389611813e-05,
+      "loss": 0.5261,
+      "step": 1242
+    },
+    {
+      "epoch": 0.27726968547847425,
+      "grad_norm": 0.17645849287509918,
+      "learning_rate": 1.9641004732394862e-05,
+      "loss": 0.526,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2774927503903636,
+      "grad_norm": 0.16956675052642822,
+      "learning_rate": 1.9640379541198425e-05,
+      "loss": 0.5489,
+      "step": 1244
+    },
+    {
+      "epoch": 0.277715815302253,
+      "grad_norm": 0.18473365902900696,
+      "learning_rate": 1.9639753816057128e-05,
+      "loss": 0.5422,
+      "step": 1245
+    },
+    {
+      "epoch": 0.2779388802141423,
+      "grad_norm": 0.262015700340271,
+      "learning_rate": 1.9639127557005627e-05,
+      "loss": 0.5031,
+      "step": 1246
+    },
+    {
+      "epoch": 0.27816194512603165,
+      "grad_norm": 0.19820185005664825,
+      "learning_rate": 1.963850076407861e-05,
+      "loss": 0.5132,
+      "step": 1247
+    },
+    {
+      "epoch": 0.27838501003792104,
+      "grad_norm": 0.1682923436164856,
+      "learning_rate": 1.9637873437310795e-05,
+      "loss": 0.5214,
+      "step": 1248
+    },
+    {
+      "epoch": 0.2786080749498104,
+      "grad_norm": 0.17070676386356354,
+      "learning_rate": 1.9637245576736923e-05,
+      "loss": 0.5368,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2788311398616998,
+      "grad_norm": 0.17165376245975494,
+      "learning_rate": 1.9636617182391768e-05,
+      "loss": 0.5282,
+      "step": 1250
+    },
+    {
+      "epoch": 0.2790542047735891,
+      "grad_norm": 0.17272816598415375,
+      "learning_rate": 1.963598825431014e-05,
+      "loss": 0.5657,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2792772696854785,
+      "grad_norm": 0.1659235805273056,
+      "learning_rate": 1.9635358792526865e-05,
+      "loss": 0.5181,
+      "step": 1252
+    },
+    {
+      "epoch": 0.27950033459736784,
+      "grad_norm": 0.1700238287448883,
+      "learning_rate": 1.9634728797076818e-05,
+      "loss": 0.5194,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2797233995092572,
+      "grad_norm": 0.1710444688796997,
+      "learning_rate": 1.9634098267994882e-05,
+      "loss": 0.5405,
+      "step": 1254
+    },
+    {
+      "epoch": 0.27994646442114657,
+      "grad_norm": 0.1748325079679489,
+      "learning_rate": 1.9633467205315983e-05,
+      "loss": 0.5295,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2801695293330359,
+      "grad_norm": 0.1600925624370575,
+      "learning_rate": 1.9632835609075072e-05,
+      "loss": 0.5448,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2803925942449253,
+      "grad_norm": 0.1937401294708252,
+      "learning_rate": 1.9632203479307132e-05,
+      "loss": 0.5145,
+      "step": 1257
+    },
+    {
+      "epoch": 0.28061565915681463,
+      "grad_norm": 0.1888452023267746,
+      "learning_rate": 1.9631570816047176e-05,
+      "loss": 0.5376,
+      "step": 1258
+    },
+    {
+      "epoch": 0.28083872406870397,
+      "grad_norm": 0.16786231100559235,
+      "learning_rate": 1.963093761933024e-05,
+      "loss": 0.5336,
+      "step": 1259
+    },
+    {
+      "epoch": 0.28106178898059336,
+      "grad_norm": 0.16962790489196777,
+      "learning_rate": 1.9630303889191406e-05,
+      "loss": 0.5161,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2812848538924827,
+      "grad_norm": 0.17393000423908234,
+      "learning_rate": 1.9629669625665757e-05,
+      "loss": 0.5098,
+      "step": 1261
+    },
+    {
+      "epoch": 0.2815079188043721,
+      "grad_norm": 0.18124370276927948,
+      "learning_rate": 1.9629034828788435e-05,
+      "loss": 0.5155,
+      "step": 1262
+    },
+    {
+      "epoch": 0.28173098371626143,
+      "grad_norm": 0.16687680780887604,
+      "learning_rate": 1.962839949859459e-05,
+      "loss": 0.5189,
+      "step": 1263
+    },
+    {
+      "epoch": 0.28195404862815077,
+      "grad_norm": 0.1622602790594101,
+      "learning_rate": 1.9627763635119423e-05,
+      "loss": 0.4974,
+      "step": 1264
+    },
+    {
+      "epoch": 0.28217711354004016,
+      "grad_norm": 0.16674260795116425,
+      "learning_rate": 1.9627127238398142e-05,
+      "loss": 0.4923,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2824001784519295,
+      "grad_norm": 0.1815263032913208,
+      "learning_rate": 1.9626490308465996e-05,
+      "loss": 0.5048,
+      "step": 1266
+    },
+    {
+      "epoch": 0.2826232433638189,
+      "grad_norm": 0.17753565311431885,
+      "learning_rate": 1.9625852845358265e-05,
+      "loss": 0.5326,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2828463082757082,
+      "grad_norm": 0.1820516288280487,
+      "learning_rate": 1.9625214849110253e-05,
+      "loss": 0.5289,
+      "step": 1268
+    },
+    {
+      "epoch": 0.2830693731875976,
+      "grad_norm": 0.16515739262104034,
+      "learning_rate": 1.9624576319757302e-05,
+      "loss": 0.5159,
+      "step": 1269
+    },
+    {
+      "epoch": 0.28329243809948695,
+      "grad_norm": 0.17564083635807037,
+      "learning_rate": 1.9623937257334767e-05,
+      "loss": 0.5052,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2835155030113763,
+      "grad_norm": 0.18046805262565613,
+      "learning_rate": 1.9623297661878054e-05,
+      "loss": 0.5349,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2837385679232657,
+      "grad_norm": 0.16942784190177917,
+      "learning_rate": 1.9622657533422583e-05,
+      "loss": 0.4924,
+      "step": 1272
+    },
+    {
+      "epoch": 0.283961632835155,
+      "grad_norm": 0.17093469202518463,
+      "learning_rate": 1.9622016872003807e-05,
+      "loss": 0.5261,
+      "step": 1273
+    },
+    {
+      "epoch": 0.2841846977470444,
+      "grad_norm": 0.17606058716773987,
+      "learning_rate": 1.9621375677657217e-05,
+      "loss": 0.51,
+      "step": 1274
+    },
+    {
+      "epoch": 0.28440776265893375,
+      "grad_norm": 0.16897493600845337,
+      "learning_rate": 1.9620733950418316e-05,
+      "loss": 0.5058,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2846308275708231,
+      "grad_norm": 0.18536695837974548,
+      "learning_rate": 1.9620091690322654e-05,
+      "loss": 0.512,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2848538924827125,
+      "grad_norm": 0.17734606564044952,
+      "learning_rate": 1.96194488974058e-05,
+      "loss": 0.5311,
+      "step": 1277
+    },
+    {
+      "epoch": 0.2850769573946018,
+      "grad_norm": 0.18116462230682373,
+      "learning_rate": 1.9618805571703356e-05,
+      "loss": 0.5153,
+      "step": 1278
+    },
+    {
+      "epoch": 0.2853000223064912,
+      "grad_norm": 0.17743848264217377,
+      "learning_rate": 1.961816171325096e-05,
+      "loss": 0.5238,
+      "step": 1279
+    },
+    {
+      "epoch": 0.28552308721838054,
+      "grad_norm": 0.16778843104839325,
+      "learning_rate": 1.961751732208426e-05,
+      "loss": 0.505,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2857461521302699,
+      "grad_norm": 0.17177841067314148,
+      "learning_rate": 1.961687239823896e-05,
+      "loss": 0.5075,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2859692170421593,
+      "grad_norm": 0.18460653722286224,
+      "learning_rate": 1.9616226941750775e-05,
+      "loss": 0.5109,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2861922819540486,
+      "grad_norm": 0.16801568865776062,
+      "learning_rate": 1.961558095265545e-05,
+      "loss": 0.552,
+      "step": 1283
+    },
+    {
+      "epoch": 0.286415346865938,
+      "grad_norm": 0.19550864398479462,
+      "learning_rate": 1.961493443098877e-05,
+      "loss": 0.5138,
+      "step": 1284
+    },
+    {
+      "epoch": 0.28663841177782734,
+      "grad_norm": 0.17915892601013184,
+      "learning_rate": 1.9614287376786537e-05,
+      "loss": 0.5267,
+      "step": 1285
+    },
+    {
+      "epoch": 0.28686147668971673,
+      "grad_norm": 0.17141470313072205,
+      "learning_rate": 1.9613639790084596e-05,
+      "loss": 0.5398,
+      "step": 1286
+    },
+    {
+      "epoch": 0.28708454160160607,
+      "grad_norm": 0.16843633353710175,
+      "learning_rate": 1.9612991670918808e-05,
+      "loss": 0.5224,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2873076065134954,
+      "grad_norm": 0.17116384208202362,
+      "learning_rate": 1.9612343019325077e-05,
+      "loss": 0.506,
+      "step": 1288
+    },
+    {
+      "epoch": 0.2875306714253848,
+      "grad_norm": 0.19337520003318787,
+      "learning_rate": 1.9611693835339323e-05,
+      "loss": 0.5417,
+      "step": 1289
+    },
+    {
+      "epoch": 0.28775373633727414,
+      "grad_norm": 0.1831638365983963,
+      "learning_rate": 1.9611044118997507e-05,
+      "loss": 0.5487,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2879768012491635,
+      "grad_norm": 0.1741098165512085,
+      "learning_rate": 1.961039387033561e-05,
+      "loss": 0.5098,
+      "step": 1291
+    },
+    {
+      "epoch": 0.28819986616105286,
+      "grad_norm": 0.16796158254146576,
+      "learning_rate": 1.960974308938965e-05,
+      "loss": 0.5105,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2884229310729422,
+      "grad_norm": 0.16848038136959076,
+      "learning_rate": 1.9609091776195667e-05,
+      "loss": 0.5106,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2886459959848316,
+      "grad_norm": 0.1563376486301422,
+      "learning_rate": 1.960843993078974e-05,
+      "loss": 0.4927,
+      "step": 1294
+    },
+    {
+      "epoch": 0.28886906089672093,
+      "grad_norm": 0.16741539537906647,
+      "learning_rate": 1.9607787553207972e-05,
+      "loss": 0.5241,
+      "step": 1295
+    },
+    {
+      "epoch": 0.2890921258086103,
+      "grad_norm": 0.16913025081157684,
+      "learning_rate": 1.9607134643486492e-05,
+      "loss": 0.5183,
+      "step": 1296
+    },
+    {
+      "epoch": 0.28931519072049966,
+      "grad_norm": 0.17226742208003998,
+      "learning_rate": 1.9606481201661466e-05,
+      "loss": 0.5105,
+      "step": 1297
+    },
+    {
+      "epoch": 0.28953825563238905,
+      "grad_norm": 0.17816194891929626,
+      "learning_rate": 1.960582722776908e-05,
+      "loss": 0.5352,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2897613205442784,
+      "grad_norm": 0.16423411667346954,
+      "learning_rate": 1.9605172721845564e-05,
+      "loss": 0.5303,
+      "step": 1299
+    },
+    {
+      "epoch": 0.2899843854561677,
+      "grad_norm": 0.22330845892429352,
+      "learning_rate": 1.9604517683927156e-05,
+      "loss": 0.5115,
+      "step": 1300
+    },
+    {
+      "epoch": 0.2902074503680571,
+      "grad_norm": 0.16385138034820557,
+      "learning_rate": 1.960386211405015e-05,
+      "loss": 0.5375,
+      "step": 1301
+    },
+    {
+      "epoch": 0.29043051527994646,
+      "grad_norm": 0.17269232869148254,
+      "learning_rate": 1.960320601225085e-05,
+      "loss": 0.5291,
+      "step": 1302
+    },
+    {
+      "epoch": 0.29065358019183585,
+      "grad_norm": 0.1625886857509613,
+      "learning_rate": 1.9602549378565592e-05,
+      "loss": 0.4982,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2908766451037252,
+      "grad_norm": 0.17353162169456482,
+      "learning_rate": 1.9601892213030746e-05,
+      "loss": 0.5109,
+      "step": 1304
+    },
+    {
+      "epoch": 0.2910997100156145,
+      "grad_norm": 0.16863249242305756,
+      "learning_rate": 1.9601234515682712e-05,
+      "loss": 0.5409,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2913227749275039,
+      "grad_norm": 0.17428503930568695,
+      "learning_rate": 1.960057628655792e-05,
+      "loss": 0.5155,
+      "step": 1306
+    },
+    {
+      "epoch": 0.29154583983939325,
+      "grad_norm": 0.1825830489397049,
+      "learning_rate": 1.9599917525692816e-05,
+      "loss": 0.5355,
+      "step": 1307
+    },
+    {
+      "epoch": 0.29176890475128264,
+      "grad_norm": 0.8213825821876526,
+      "learning_rate": 1.95992582331239e-05,
+      "loss": 0.5366,
+      "step": 1308
+    },
+    {
+      "epoch": 0.291991969663172,
+      "grad_norm": 0.16370530426502228,
+      "learning_rate": 1.959859840888768e-05,
+      "loss": 0.5071,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2922150345750613,
+      "grad_norm": 0.17610913515090942,
+      "learning_rate": 1.95979380530207e-05,
+      "loss": 0.5346,
+      "step": 1310
+    },
+    {
+      "epoch": 0.2924380994869507,
+      "grad_norm": 0.16303309798240662,
+      "learning_rate": 1.959727716555954e-05,
+      "loss": 0.535,
+      "step": 1311
+    },
+    {
+      "epoch": 0.29266116439884005,
+      "grad_norm": 0.17250092327594757,
+      "learning_rate": 1.9596615746540798e-05,
+      "loss": 0.5322,
+      "step": 1312
+    },
+    {
+      "epoch": 0.29288422931072944,
+      "grad_norm": 0.18915972113609314,
+      "learning_rate": 1.959595379600111e-05,
+      "loss": 0.5293,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2931072942226188,
+      "grad_norm": 0.17842963337898254,
+      "learning_rate": 1.9595291313977144e-05,
+      "loss": 0.5197,
+      "step": 1314
+    },
+    {
+      "epoch": 0.29333035913450817,
+      "grad_norm": 0.1693105846643448,
+      "learning_rate": 1.959462830050559e-05,
+      "loss": 0.5044,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2935534240463975,
+      "grad_norm": 0.23653189837932587,
+      "learning_rate": 1.959396475562316e-05,
+      "loss": 0.5286,
+      "step": 1316
+    },
+    {
+      "epoch": 0.29377648895828684,
+      "grad_norm": 0.1704518347978592,
+      "learning_rate": 1.9593300679366622e-05,
+      "loss": 0.5288,
+      "step": 1317
+    },
+    {
+      "epoch": 0.29399955387017623,
+      "grad_norm": 0.169080451130867,
+      "learning_rate": 1.9592636071772745e-05,
+      "loss": 0.5336,
+      "step": 1318
+    },
+    {
+      "epoch": 0.29422261878206557,
+      "grad_norm": 0.2256423681974411,
+      "learning_rate": 1.959197093287834e-05,
+      "loss": 0.5253,
+      "step": 1319
+    },
+    {
+      "epoch": 0.29444568369395496,
+      "grad_norm": 0.1699950098991394,
+      "learning_rate": 1.9591305262720252e-05,
+      "loss": 0.5308,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2946687486058443,
+      "grad_norm": 0.17314709722995758,
+      "learning_rate": 1.9590639061335345e-05,
+      "loss": 0.5477,
+      "step": 1321
+    },
+    {
+      "epoch": 0.29489181351773364,
+      "grad_norm": 0.16141551733016968,
+      "learning_rate": 1.958997232876052e-05,
+      "loss": 0.5207,
+      "step": 1322
+    },
+    {
+      "epoch": 0.29511487842962303,
+      "grad_norm": 0.15994872152805328,
+      "learning_rate": 1.9589305065032705e-05,
+      "loss": 0.5142,
+      "step": 1323
+    },
+    {
+      "epoch": 0.29533794334151237,
+      "grad_norm": 0.16060283780097961,
+      "learning_rate": 1.9588637270188852e-05,
+      "loss": 0.4967,
+      "step": 1324
+    },
+    {
+      "epoch": 0.29556100825340176,
+      "grad_norm": 0.17949143052101135,
+      "learning_rate": 1.9587968944265955e-05,
+      "loss": 0.5484,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2957840731652911,
+      "grad_norm": 0.17476752400398254,
+      "learning_rate": 1.958730008730103e-05,
+      "loss": 0.5466,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2960071380771805,
+      "grad_norm": 0.205107182264328,
+      "learning_rate": 1.9586630699331115e-05,
+      "loss": 0.5147,
+      "step": 1327
+    },
+    {
+      "epoch": 0.2962302029890698,
+      "grad_norm": 0.17714425921440125,
+      "learning_rate": 1.9585960780393293e-05,
+      "loss": 0.5142,
+      "step": 1328
+    },
+    {
+      "epoch": 0.29645326790095916,
+      "grad_norm": 0.17877231538295746,
+      "learning_rate": 1.9585290330524663e-05,
+      "loss": 0.5073,
+      "step": 1329
+    },
+    {
+      "epoch": 0.29667633281284855,
+      "grad_norm": 0.16752929985523224,
+      "learning_rate": 1.958461934976236e-05,
+      "loss": 0.5381,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2968993977247379,
+      "grad_norm": 0.17550905048847198,
+      "learning_rate": 1.9583947838143553e-05,
+      "loss": 0.5185,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2971224626366273,
+      "grad_norm": 0.17885231971740723,
+      "learning_rate": 1.958327579570542e-05,
+      "loss": 0.5279,
+      "step": 1332
+    },
+    {
+      "epoch": 0.2973455275485166,
+      "grad_norm": 0.16726696491241455,
+      "learning_rate": 1.95826032224852e-05,
+      "loss": 0.5253,
+      "step": 1333
+    },
+    {
+      "epoch": 0.29756859246040596,
+      "grad_norm": 0.16935260593891144,
+      "learning_rate": 1.9581930118520135e-05,
+      "loss": 0.5311,
+      "step": 1334
+    },
+    {
+      "epoch": 0.29779165737229535,
+      "grad_norm": 0.22187648713588715,
+      "learning_rate": 1.9581256483847505e-05,
+      "loss": 0.5124,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2980147222841847,
+      "grad_norm": 0.17254653573036194,
+      "learning_rate": 1.9580582318504623e-05,
+      "loss": 0.5113,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2982377871960741,
+      "grad_norm": 0.1686115860939026,
+      "learning_rate": 1.9579907622528827e-05,
+      "loss": 0.516,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2984608521079634,
+      "grad_norm": 0.17233097553253174,
+      "learning_rate": 1.9579232395957492e-05,
+      "loss": 0.5135,
+      "step": 1338
+    },
+    {
+      "epoch": 0.29868391701985275,
+      "grad_norm": 0.1767144799232483,
+      "learning_rate": 1.957855663882801e-05,
+      "loss": 0.5113,
+      "step": 1339
+    },
+    {
+      "epoch": 0.29890698193174214,
+      "grad_norm": 0.16990788280963898,
+      "learning_rate": 1.9577880351177803e-05,
+      "loss": 0.5213,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2991300468436315,
+      "grad_norm": 0.16583338379859924,
+      "learning_rate": 1.957720353304434e-05,
+      "loss": 0.5059,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2993531117555209,
+      "grad_norm": 0.17070691287517548,
+      "learning_rate": 1.95765261844651e-05,
+      "loss": 0.5154,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2995761766674102,
+      "grad_norm": 0.18274636566638947,
+      "learning_rate": 1.9575848305477606e-05,
+      "loss": 0.5367,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2997992415792996,
+      "grad_norm": 0.16276118159294128,
+      "learning_rate": 1.957516989611939e-05,
+      "loss": 0.5156,
+      "step": 1344
+    },
+    {
+      "epoch": 0.30002230649118894,
+      "grad_norm": 0.16200782358646393,
+      "learning_rate": 1.9574490956428045e-05,
+      "loss": 0.5201,
+      "step": 1345
+    },
+    {
+      "epoch": 0.3002453714030783,
+      "grad_norm": 0.18693317472934723,
+      "learning_rate": 1.9573811486441158e-05,
+      "loss": 0.5181,
+      "step": 1346
+    },
+    {
+      "epoch": 0.30046843631496767,
+      "grad_norm": 0.17908775806427002,
+      "learning_rate": 1.9573131486196372e-05,
+      "loss": 0.5012,
+      "step": 1347
+    },
+    {
+      "epoch": 0.300691501226857,
+      "grad_norm": 0.16340862214565277,
+      "learning_rate": 1.9572450955731346e-05,
+      "loss": 0.4951,
+      "step": 1348
+    },
+    {
+      "epoch": 0.3009145661387464,
+      "grad_norm": 0.16997461020946503,
+      "learning_rate": 1.957176989508377e-05,
+      "loss": 0.5248,
+      "step": 1349
+    },
+    {
+      "epoch": 0.30113763105063573,
+      "grad_norm": 0.1896587759256363,
+      "learning_rate": 1.9571088304291376e-05,
+      "loss": 0.5129,
+      "step": 1350
+    },
+    {
+      "epoch": 0.30136069596252507,
+      "grad_norm": 0.1624515950679779,
+      "learning_rate": 1.95704061833919e-05,
+      "loss": 0.4872,
+      "step": 1351
+    },
+    {
+      "epoch": 0.30158376087441446,
+      "grad_norm": 0.16828244924545288,
+      "learning_rate": 1.956972353242313e-05,
+      "loss": 0.5291,
+      "step": 1352
+    },
+    {
+      "epoch": 0.3018068257863038,
+      "grad_norm": 0.17992524802684784,
+      "learning_rate": 1.9569040351422882e-05,
+      "loss": 0.5396,
+      "step": 1353
+    },
+    {
+      "epoch": 0.3020298906981932,
+      "grad_norm": 0.17174522578716278,
+      "learning_rate": 1.956835664042898e-05,
+      "loss": 0.523,
+      "step": 1354
+    },
+    {
+      "epoch": 0.30225295561008253,
+      "grad_norm": 0.18021540343761444,
+      "learning_rate": 1.9567672399479304e-05,
+      "loss": 0.5091,
+      "step": 1355
+    },
+    {
+      "epoch": 0.30247602052197187,
+      "grad_norm": 0.1889013797044754,
+      "learning_rate": 1.9566987628611748e-05,
+      "loss": 0.521,
+      "step": 1356
+    },
+    {
+      "epoch": 0.30269908543386126,
+      "grad_norm": 0.1710296869277954,
+      "learning_rate": 1.9566302327864233e-05,
+      "loss": 0.51,
+      "step": 1357
+    },
+    {
+      "epoch": 0.3029221503457506,
+      "grad_norm": 0.16925543546676636,
+      "learning_rate": 1.9565616497274725e-05,
+      "loss": 0.5194,
+      "step": 1358
+    },
+    {
+      "epoch": 0.30314521525764,
+      "grad_norm": 0.16558070480823517,
+      "learning_rate": 1.95649301368812e-05,
+      "loss": 0.5284,
+      "step": 1359
+    },
+    {
+      "epoch": 0.3033682801695293,
+      "grad_norm": 0.1646072119474411,
+      "learning_rate": 1.9564243246721686e-05,
+      "loss": 0.53,
+      "step": 1360
+    },
+    {
+      "epoch": 0.3035913450814187,
+      "grad_norm": 0.17670351266860962,
+      "learning_rate": 1.9563555826834214e-05,
+      "loss": 0.527,
+      "step": 1361
+    },
+    {
+      "epoch": 0.30381440999330805,
+      "grad_norm": 0.17414048314094543,
+      "learning_rate": 1.9562867877256867e-05,
+      "loss": 0.5318,
+      "step": 1362
+    },
+    {
+      "epoch": 0.3040374749051974,
+      "grad_norm": 0.18004649877548218,
+      "learning_rate": 1.956217939802774e-05,
+      "loss": 0.5149,
+      "step": 1363
+    },
+    {
+      "epoch": 0.3042605398170868,
+      "grad_norm": 0.16560295224189758,
+      "learning_rate": 1.9561490389184973e-05,
+      "loss": 0.5093,
+      "step": 1364
+    },
+    {
+      "epoch": 0.3044836047289761,
+      "grad_norm": 0.1967393010854721,
+      "learning_rate": 1.956080085076672e-05,
+      "loss": 0.499,
+      "step": 1365
+    },
+    {
+      "epoch": 0.3047066696408655,
+      "grad_norm": 0.17638468742370605,
+      "learning_rate": 1.956011078281118e-05,
+      "loss": 0.5287,
+      "step": 1366
+    },
+    {
+      "epoch": 0.30492973455275485,
+      "grad_norm": 0.18389096856117249,
+      "learning_rate": 1.955942018535657e-05,
+      "loss": 0.5263,
+      "step": 1367
+    },
+    {
+      "epoch": 0.3051527994646442,
+      "grad_norm": 0.16795802116394043,
+      "learning_rate": 1.9558729058441135e-05,
+      "loss": 0.4909,
+      "step": 1368
+    },
+    {
+      "epoch": 0.3053758643765336,
+      "grad_norm": 0.1886921525001526,
+      "learning_rate": 1.955803740210316e-05,
+      "loss": 0.4929,
+      "step": 1369
+    },
+    {
+      "epoch": 0.3055989292884229,
+      "grad_norm": 0.15743811428546906,
+      "learning_rate": 1.9557345216380953e-05,
+      "loss": 0.5252,
+      "step": 1370
+    },
+    {
+      "epoch": 0.3058219942003123,
+      "grad_norm": 0.18187056481838226,
+      "learning_rate": 1.955665250131285e-05,
+      "loss": 0.5307,
+      "step": 1371
+    },
+    {
+      "epoch": 0.30604505911220165,
+      "grad_norm": 0.17258834838867188,
+      "learning_rate": 1.9555959256937214e-05,
+      "loss": 0.5007,
+      "step": 1372
+    },
+    {
+      "epoch": 0.30626812402409104,
+      "grad_norm": 0.16746965050697327,
+      "learning_rate": 1.9555265483292446e-05,
+      "loss": 0.5174,
+      "step": 1373
+    },
+    {
+      "epoch": 0.3064911889359804,
+      "grad_norm": 0.191069096326828,
+      "learning_rate": 1.955457118041697e-05,
+      "loss": 0.5462,
+      "step": 1374
+    },
+    {
+      "epoch": 0.3067142538478697,
+      "grad_norm": 0.17369060218334198,
+      "learning_rate": 1.9553876348349242e-05,
+      "loss": 0.5175,
+      "step": 1375
+    },
+    {
+      "epoch": 0.3069373187597591,
+      "grad_norm": 0.16620078682899475,
+      "learning_rate": 1.9553180987127748e-05,
+      "loss": 0.5298,
+      "step": 1376
+    },
+    {
+      "epoch": 0.30716038367164844,
+      "grad_norm": 0.22137194871902466,
+      "learning_rate": 1.9552485096790996e-05,
+      "loss": 0.5135,
+      "step": 1377
+    },
+    {
+      "epoch": 0.30738344858353783,
+      "grad_norm": 0.20380550622940063,
+      "learning_rate": 1.9551788677377535e-05,
+      "loss": 0.5304,
+      "step": 1378
+    },
+    {
+      "epoch": 0.30760651349542717,
+      "grad_norm": 0.16844195127487183,
+      "learning_rate": 1.955109172892593e-05,
+      "loss": 0.55,
+      "step": 1379
+    },
+    {
+      "epoch": 0.3078295784073165,
+      "grad_norm": 0.1768990010023117,
+      "learning_rate": 1.955039425147479e-05,
+      "loss": 0.5254,
+      "step": 1380
+    },
+    {
+      "epoch": 0.3080526433192059,
+      "grad_norm": 0.16573172807693481,
+      "learning_rate": 1.954969624506274e-05,
+      "loss": 0.5195,
+      "step": 1381
+    },
+    {
+      "epoch": 0.30827570823109524,
+      "grad_norm": 0.16543026268482208,
+      "learning_rate": 1.9548997709728443e-05,
+      "loss": 0.5243,
+      "step": 1382
+    },
+    {
+      "epoch": 0.30849877314298463,
+      "grad_norm": 0.17528237402439117,
+      "learning_rate": 1.9548298645510587e-05,
+      "loss": 0.5133,
+      "step": 1383
+    },
+    {
+      "epoch": 0.30872183805487396,
+      "grad_norm": 0.16765229403972626,
+      "learning_rate": 1.954759905244789e-05,
+      "loss": 0.5185,
+      "step": 1384
+    },
+    {
+      "epoch": 0.3089449029667633,
+      "grad_norm": 0.17179544270038605,
+      "learning_rate": 1.9546898930579102e-05,
+      "loss": 0.5156,
+      "step": 1385
+    },
+    {
+      "epoch": 0.3091679678786527,
+      "grad_norm": 0.17194928228855133,
+      "learning_rate": 1.9546198279942997e-05,
+      "loss": 0.5134,
+      "step": 1386
+    },
+    {
+      "epoch": 0.30939103279054203,
+      "grad_norm": 0.16621114313602448,
+      "learning_rate": 1.9545497100578382e-05,
+      "loss": 0.5101,
+      "step": 1387
+    },
+    {
+      "epoch": 0.3096140977024314,
+      "grad_norm": 0.16408509016036987,
+      "learning_rate": 1.9544795392524096e-05,
+      "loss": 0.515,
+      "step": 1388
+    },
+    {
+      "epoch": 0.30983716261432076,
+      "grad_norm": 0.16583041846752167,
+      "learning_rate": 1.9544093155819004e-05,
+      "loss": 0.4974,
+      "step": 1389
+    },
+    {
+      "epoch": 0.31006022752621015,
+      "grad_norm": 0.16977885365486145,
+      "learning_rate": 1.9543390390502e-05,
+      "loss": 0.5229,
+      "step": 1390
+    },
+    {
+      "epoch": 0.3102832924380995,
+      "grad_norm": 0.1626800149679184,
+      "learning_rate": 1.9542687096611998e-05,
+      "loss": 0.5166,
+      "step": 1391
+    },
+    {
+      "epoch": 0.3105063573499888,
+      "grad_norm": 0.17305392026901245,
+      "learning_rate": 1.9541983274187964e-05,
+      "loss": 0.516,
+      "step": 1392
+    },
+    {
+      "epoch": 0.3107294222618782,
+      "grad_norm": 0.18632498383522034,
+      "learning_rate": 1.9541278923268872e-05,
+      "loss": 0.4985,
+      "step": 1393
+    },
+    {
+      "epoch": 0.31095248717376756,
+      "grad_norm": 0.18472731113433838,
+      "learning_rate": 1.9540574043893738e-05,
+      "loss": 0.5304,
+      "step": 1394
+    },
+    {
+      "epoch": 0.31117555208565695,
+      "grad_norm": 0.1734933704137802,
+      "learning_rate": 1.9539868636101602e-05,
+      "loss": 0.5523,
+      "step": 1395
+    },
+    {
+      "epoch": 0.3113986169975463,
+      "grad_norm": 0.17965291440486908,
+      "learning_rate": 1.9539162699931534e-05,
+      "loss": 0.532,
+      "step": 1396
+    },
+    {
+      "epoch": 0.3116216819094356,
+      "grad_norm": 0.1630621999502182,
+      "learning_rate": 1.9538456235422625e-05,
+      "loss": 0.5351,
+      "step": 1397
+    },
+    {
+      "epoch": 0.311844746821325,
+      "grad_norm": 0.17468804121017456,
+      "learning_rate": 1.9537749242614016e-05,
+      "loss": 0.5144,
+      "step": 1398
+    },
+    {
+      "epoch": 0.31206781173321435,
+      "grad_norm": 0.17356379330158234,
+      "learning_rate": 1.9537041721544862e-05,
+      "loss": 0.5124,
+      "step": 1399
+    },
+    {
+      "epoch": 0.31229087664510374,
+      "grad_norm": 0.1623128354549408,
+      "learning_rate": 1.953633367225434e-05,
+      "loss": 0.5256,
+      "step": 1400
+    },
+    {
+      "epoch": 0.3125139415569931,
+      "grad_norm": 0.17314434051513672,
+      "learning_rate": 1.9535625094781677e-05,
+      "loss": 0.5231,
+      "step": 1401
+    },
+    {
+      "epoch": 0.3127370064688825,
+      "grad_norm": 0.17369012534618378,
+      "learning_rate": 1.9534915989166115e-05,
+      "loss": 0.5266,
+      "step": 1402
+    },
+    {
+      "epoch": 0.3129600713807718,
+      "grad_norm": 0.16464915871620178,
+      "learning_rate": 1.9534206355446927e-05,
+      "loss": 0.4962,
+      "step": 1403
+    },
+    {
+      "epoch": 0.31318313629266115,
+      "grad_norm": 0.16123609244823456,
+      "learning_rate": 1.953349619366342e-05,
+      "loss": 0.4932,
+      "step": 1404
+    },
+    {
+      "epoch": 0.31340620120455054,
+      "grad_norm": 0.16396626830101013,
+      "learning_rate": 1.9532785503854926e-05,
+      "loss": 0.5103,
+      "step": 1405
+    },
+    {
+      "epoch": 0.3136292661164399,
+      "grad_norm": 0.1750822216272354,
+      "learning_rate": 1.9532074286060805e-05,
+      "loss": 0.5108,
+      "step": 1406
+    },
+    {
+      "epoch": 0.31385233102832927,
+      "grad_norm": 0.1627611666917801,
+      "learning_rate": 1.953136254032045e-05,
+      "loss": 0.4933,
+      "step": 1407
+    },
+    {
+      "epoch": 0.3140753959402186,
+      "grad_norm": 0.17482301592826843,
+      "learning_rate": 1.9530650266673286e-05,
+      "loss": 0.5179,
+      "step": 1408
+    },
+    {
+      "epoch": 0.31429846085210794,
+      "grad_norm": 0.17025405168533325,
+      "learning_rate": 1.952993746515876e-05,
+      "loss": 0.4968,
+      "step": 1409
+    },
+    {
+      "epoch": 0.31452152576399733,
+      "grad_norm": 0.17483878135681152,
+      "learning_rate": 1.9529224135816348e-05,
+      "loss": 0.5124,
+      "step": 1410
+    },
+    {
+      "epoch": 0.31474459067588667,
+      "grad_norm": 0.1743018627166748,
+      "learning_rate": 1.9528510278685568e-05,
+      "loss": 0.5207,
+      "step": 1411
+    },
+    {
+      "epoch": 0.31496765558777606,
+      "grad_norm": 0.16575628519058228,
+      "learning_rate": 1.9527795893805947e-05,
+      "loss": 0.4846,
+      "step": 1412
+    },
+    {
+      "epoch": 0.3151907204996654,
+      "grad_norm": 0.16664400696754456,
+      "learning_rate": 1.952708098121706e-05,
+      "loss": 0.5097,
+      "step": 1413
+    },
+    {
+      "epoch": 0.31541378541155474,
+      "grad_norm": 0.17265918850898743,
+      "learning_rate": 1.9526365540958497e-05,
+      "loss": 0.5002,
+      "step": 1414
+    },
+    {
+      "epoch": 0.31563685032344413,
+      "grad_norm": 0.16979828476905823,
+      "learning_rate": 1.952564957306989e-05,
+      "loss": 0.5172,
+      "step": 1415
+    },
+    {
+      "epoch": 0.31585991523533347,
+      "grad_norm": 0.17120909690856934,
+      "learning_rate": 1.952493307759089e-05,
+      "loss": 0.5204,
+      "step": 1416
+    },
+    {
+      "epoch": 0.31608298014722286,
+      "grad_norm": 0.15589144825935364,
+      "learning_rate": 1.9524216054561186e-05,
+      "loss": 0.5001,
+      "step": 1417
+    },
+    {
+      "epoch": 0.3163060450591122,
+      "grad_norm": 0.1638038605451584,
+      "learning_rate": 1.9523498504020486e-05,
+      "loss": 0.495,
+      "step": 1418
+    },
+    {
+      "epoch": 0.3165291099710016,
+      "grad_norm": 0.2510809600353241,
+      "learning_rate": 1.952278042600853e-05,
+      "loss": 0.5177,
+      "step": 1419
+    },
+    {
+      "epoch": 0.3167521748828909,
+      "grad_norm": 0.1806640475988388,
+      "learning_rate": 1.9522061820565093e-05,
+      "loss": 0.5187,
+      "step": 1420
+    },
+    {
+      "epoch": 0.31697523979478026,
+      "grad_norm": 0.16829638183116913,
+      "learning_rate": 1.9521342687729977e-05,
+      "loss": 0.5275,
+      "step": 1421
+    },
+    {
+      "epoch": 0.31719830470666965,
+      "grad_norm": 0.18520011007785797,
+      "learning_rate": 1.9520623027543015e-05,
+      "loss": 0.5654,
+      "step": 1422
+    },
+    {
+      "epoch": 0.317421369618559,
+      "grad_norm": 0.17533186078071594,
+      "learning_rate": 1.951990284004406e-05,
+      "loss": 0.5102,
+      "step": 1423
+    },
+    {
+      "epoch": 0.3176444345304484,
+      "grad_norm": 0.17360709607601166,
+      "learning_rate": 1.9519182125273e-05,
+      "loss": 0.5196,
+      "step": 1424
+    },
+    {
+      "epoch": 0.3178674994423377,
+      "grad_norm": 0.17495203018188477,
+      "learning_rate": 1.951846088326976e-05,
+      "loss": 0.5416,
+      "step": 1425
+    },
+    {
+      "epoch": 0.31809056435422706,
+      "grad_norm": 0.17361585795879364,
+      "learning_rate": 1.9517739114074282e-05,
+      "loss": 0.53,
+      "step": 1426
+    },
+    {
+      "epoch": 0.31831362926611645,
+      "grad_norm": 0.17399337887763977,
+      "learning_rate": 1.9517016817726542e-05,
+      "loss": 0.5207,
+      "step": 1427
+    },
+    {
+      "epoch": 0.3185366941780058,
+      "grad_norm": 0.16763678193092346,
+      "learning_rate": 1.9516293994266548e-05,
+      "loss": 0.5059,
+      "step": 1428
+    },
+    {
+      "epoch": 0.3187597590898952,
+      "grad_norm": 0.16563831269741058,
+      "learning_rate": 1.951557064373433e-05,
+      "loss": 0.5079,
+      "step": 1429
+    },
+    {
+      "epoch": 0.3189828240017845,
+      "grad_norm": 0.1833125650882721,
+      "learning_rate": 1.951484676616996e-05,
+      "loss": 0.5181,
+      "step": 1430
+    },
+    {
+      "epoch": 0.31920588891367385,
+      "grad_norm": 0.15749512612819672,
+      "learning_rate": 1.951412236161352e-05,
+      "loss": 0.4967,
+      "step": 1431
+    },
+    {
+      "epoch": 0.31942895382556324,
+      "grad_norm": 0.17580753564834595,
+      "learning_rate": 1.9513397430105137e-05,
+      "loss": 0.5238,
+      "step": 1432
+    },
+    {
+      "epoch": 0.3196520187374526,
+      "grad_norm": 0.16896769404411316,
+      "learning_rate": 1.9512671971684963e-05,
+      "loss": 0.5218,
+      "step": 1433
+    },
+    {
+      "epoch": 0.319875083649342,
+      "grad_norm": 0.1761835813522339,
+      "learning_rate": 1.951194598639318e-05,
+      "loss": 0.5183,
+      "step": 1434
+    },
+    {
+      "epoch": 0.3200981485612313,
+      "grad_norm": 0.1777181178331375,
+      "learning_rate": 1.9511219474269992e-05,
+      "loss": 0.5283,
+      "step": 1435
+    },
+    {
+      "epoch": 0.3203212134731207,
+      "grad_norm": 0.17135201394557953,
+      "learning_rate": 1.9510492435355647e-05,
+      "loss": 0.5183,
+      "step": 1436
+    },
+    {
+      "epoch": 0.32054427838501004,
+      "grad_norm": 0.18420204520225525,
+      "learning_rate": 1.9509764869690407e-05,
+      "loss": 0.5239,
+      "step": 1437
+    },
+    {
+      "epoch": 0.3207673432968994,
+      "grad_norm": 0.17093034088611603,
+      "learning_rate": 1.9509036777314568e-05,
+      "loss": 0.508,
+      "step": 1438
+    },
+    {
+      "epoch": 0.32099040820878877,
+      "grad_norm": 0.18275004625320435,
+      "learning_rate": 1.9508308158268458e-05,
+      "loss": 0.5257,
+      "step": 1439
+    },
+    {
+      "epoch": 0.3212134731206781,
+      "grad_norm": 0.19184084236621857,
+      "learning_rate": 1.950757901259243e-05,
+      "loss": 0.5138,
+      "step": 1440
+    },
+    {
+      "epoch": 0.3214365380325675,
+      "grad_norm": 0.19820280373096466,
+      "learning_rate": 1.9506849340326876e-05,
+      "loss": 0.5406,
+      "step": 1441
+    },
+    {
+      "epoch": 0.32165960294445683,
+      "grad_norm": 0.1850280463695526,
+      "learning_rate": 1.9506119141512204e-05,
+      "loss": 0.5009,
+      "step": 1442
+    },
+    {
+      "epoch": 0.32188266785634617,
+      "grad_norm": 0.17693570256233215,
+      "learning_rate": 1.9505388416188854e-05,
+      "loss": 0.5502,
+      "step": 1443
+    },
+    {
+      "epoch": 0.32210573276823556,
+      "grad_norm": 0.16360151767730713,
+      "learning_rate": 1.9504657164397307e-05,
+      "loss": 0.5215,
+      "step": 1444
+    },
+    {
+      "epoch": 0.3223287976801249,
+      "grad_norm": 0.1807732880115509,
+      "learning_rate": 1.950392538617806e-05,
+      "loss": 0.5024,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3225518625920143,
+      "grad_norm": 0.17271289229393005,
+      "learning_rate": 1.950319308157164e-05,
+      "loss": 0.4997,
+      "step": 1446
+    },
+    {
+      "epoch": 0.32277492750390363,
+      "grad_norm": 0.16827355325222015,
+      "learning_rate": 1.950246025061861e-05,
+      "loss": 0.5178,
+      "step": 1447
+    },
+    {
+      "epoch": 0.322997992415793,
+      "grad_norm": 0.17736375331878662,
+      "learning_rate": 1.950172689335956e-05,
+      "loss": 0.494,
+      "step": 1448
+    },
+    {
+      "epoch": 0.32322105732768236,
+      "grad_norm": 0.16899362206459045,
+      "learning_rate": 1.9500993009835106e-05,
+      "loss": 0.4701,
+      "step": 1449
+    },
+    {
+      "epoch": 0.3234441222395717,
+      "grad_norm": 0.1732366383075714,
+      "learning_rate": 1.9500258600085894e-05,
+      "loss": 0.4774,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3236671871514611,
+      "grad_norm": 0.16916102170944214,
+      "learning_rate": 1.9499523664152603e-05,
+      "loss": 0.5222,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3238902520633504,
+      "grad_norm": 0.18570592999458313,
+      "learning_rate": 1.9498788202075936e-05,
+      "loss": 0.5611,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3241133169752398,
+      "grad_norm": 0.1690537929534912,
+      "learning_rate": 1.9498052213896627e-05,
+      "loss": 0.5055,
+      "step": 1453
+    },
+    {
+      "epoch": 0.32433638188712915,
+      "grad_norm": 0.17589417099952698,
+      "learning_rate": 1.9497315699655447e-05,
+      "loss": 0.539,
+      "step": 1454
+    },
+    {
+      "epoch": 0.3245594467990185,
+      "grad_norm": 0.1848466843366623,
+      "learning_rate": 1.949657865939318e-05,
+      "loss": 0.5337,
+      "step": 1455
+    },
+    {
+      "epoch": 0.3247825117109079,
+      "grad_norm": 0.17046886682510376,
+      "learning_rate": 1.949584109315065e-05,
+      "loss": 0.507,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3250055766227972,
+      "grad_norm": 0.17708532512187958,
+      "learning_rate": 1.9495103000968708e-05,
+      "loss": 0.5377,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3252286415346866,
+      "grad_norm": 0.18529073894023895,
+      "learning_rate": 1.9494364382888236e-05,
+      "loss": 0.5136,
+      "step": 1458
+    },
+    {
+      "epoch": 0.32545170644657595,
+      "grad_norm": 0.16627921164035797,
+      "learning_rate": 1.9493625238950143e-05,
+      "loss": 0.5149,
+      "step": 1459
+    },
+    {
+      "epoch": 0.3256747713584653,
+      "grad_norm": 0.1823386400938034,
+      "learning_rate": 1.949288556919537e-05,
+      "loss": 0.5422,
+      "step": 1460
+    },
+    {
+      "epoch": 0.3258978362703547,
+      "grad_norm": 0.17173448204994202,
+      "learning_rate": 1.949214537366488e-05,
+      "loss": 0.5315,
+      "step": 1461
+    },
+    {
+      "epoch": 0.326120901182244,
+      "grad_norm": 0.20931746065616608,
+      "learning_rate": 1.949140465239967e-05,
+      "loss": 0.525,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3263439660941334,
+      "grad_norm": 0.3969743549823761,
+      "learning_rate": 1.9490663405440765e-05,
+      "loss": 0.5193,
+      "step": 1463
+    },
+    {
+      "epoch": 0.32656703100602275,
+      "grad_norm": 0.17721976339817047,
+      "learning_rate": 1.9489921632829227e-05,
+      "loss": 0.5115,
+      "step": 1464
+    },
+    {
+      "epoch": 0.32679009591791214,
+      "grad_norm": 0.16409623622894287,
+      "learning_rate": 1.948917933460613e-05,
+      "loss": 0.5123,
+      "step": 1465
+    },
+    {
+      "epoch": 0.3270131608298015,
+      "grad_norm": 0.16853661835193634,
+      "learning_rate": 1.9488436510812594e-05,
+      "loss": 0.5269,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3272362257416908,
+      "grad_norm": 0.1875690519809723,
+      "learning_rate": 1.948769316148976e-05,
+      "loss": 0.4727,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3274592906535802,
+      "grad_norm": 0.1777825504541397,
+      "learning_rate": 1.9486949286678798e-05,
+      "loss": 0.5095,
+      "step": 1468
+    },
+    {
+      "epoch": 0.32768235556546954,
+      "grad_norm": 0.17113567888736725,
+      "learning_rate": 1.948620488642091e-05,
+      "loss": 0.5244,
+      "step": 1469
+    },
+    {
+      "epoch": 0.32790542047735893,
+      "grad_norm": 0.16337715089321136,
+      "learning_rate": 1.9485459960757325e-05,
+      "loss": 0.5194,
+      "step": 1470
+    },
+    {
+      "epoch": 0.32812848538924827,
+      "grad_norm": 0.1793193519115448,
+      "learning_rate": 1.9484714509729305e-05,
+      "loss": 0.527,
+      "step": 1471
+    },
+    {
+      "epoch": 0.3283515503011376,
+      "grad_norm": 0.16597363352775574,
+      "learning_rate": 1.948396853337813e-05,
+      "loss": 0.493,
+      "step": 1472
+    },
+    {
+      "epoch": 0.328574615213027,
+      "grad_norm": 0.16836529970169067,
+      "learning_rate": 1.9483222031745118e-05,
+      "loss": 0.4932,
+      "step": 1473
+    },
+    {
+      "epoch": 0.32879768012491634,
+      "grad_norm": 0.17120935022830963,
+      "learning_rate": 1.9482475004871622e-05,
+      "loss": 0.5156,
+      "step": 1474
+    },
+    {
+      "epoch": 0.32902074503680573,
+      "grad_norm": 0.1692550629377365,
+      "learning_rate": 1.9481727452799013e-05,
+      "loss": 0.5089,
+      "step": 1475
+    },
+    {
+      "epoch": 0.32924380994869507,
+      "grad_norm": 0.16897635161876678,
+      "learning_rate": 1.9480979375568694e-05,
+      "loss": 0.5042,
+      "step": 1476
+    },
+    {
+      "epoch": 0.32946687486058446,
+      "grad_norm": 0.16120664775371552,
+      "learning_rate": 1.9480230773222102e-05,
+      "loss": 0.5105,
+      "step": 1477
+    },
+    {
+      "epoch": 0.3296899397724738,
+      "grad_norm": 0.17156098783016205,
+      "learning_rate": 1.9479481645800694e-05,
+      "loss": 0.5217,
+      "step": 1478
+    },
+    {
+      "epoch": 0.32991300468436313,
+      "grad_norm": 0.17012807726860046,
+      "learning_rate": 1.9478731993345965e-05,
+      "loss": 0.4947,
+      "step": 1479
+    },
+    {
+      "epoch": 0.3301360695962525,
+      "grad_norm": 0.16123747825622559,
+      "learning_rate": 1.9477981815899435e-05,
+      "loss": 0.5181,
+      "step": 1480
+    },
+    {
+      "epoch": 0.33035913450814186,
+      "grad_norm": 0.16029803454875946,
+      "learning_rate": 1.947723111350265e-05,
+      "loss": 0.5089,
+      "step": 1481
+    },
+    {
+      "epoch": 0.33058219942003125,
+      "grad_norm": 0.17824719846248627,
+      "learning_rate": 1.9476479886197198e-05,
+      "loss": 0.5245,
+      "step": 1482
+    },
+    {
+      "epoch": 0.3308052643319206,
+      "grad_norm": 0.1661938726902008,
+      "learning_rate": 1.9475728134024675e-05,
+      "loss": 0.4971,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3310283292438099,
+      "grad_norm": 0.16781049966812134,
+      "learning_rate": 1.9474975857026727e-05,
+      "loss": 0.5187,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3312513941556993,
+      "grad_norm": 0.20398850739002228,
+      "learning_rate": 1.9474223055245014e-05,
+      "loss": 0.5087,
+      "step": 1485
+    },
+    {
+      "epoch": 0.33147445906758866,
+      "grad_norm": 0.19590437412261963,
+      "learning_rate": 1.9473469728721233e-05,
+      "loss": 0.5193,
+      "step": 1486
+    },
+    {
+      "epoch": 0.33169752397947805,
+      "grad_norm": 0.16896989941596985,
+      "learning_rate": 1.947271587749711e-05,
+      "loss": 0.5272,
+      "step": 1487
+    },
+    {
+      "epoch": 0.3319205888913674,
+      "grad_norm": 0.16621822118759155,
+      "learning_rate": 1.9471961501614395e-05,
+      "loss": 0.5356,
+      "step": 1488
+    },
+    {
+      "epoch": 0.3321436538032567,
+      "grad_norm": 0.16570885479450226,
+      "learning_rate": 1.947120660111487e-05,
+      "loss": 0.5105,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3323667187151461,
+      "grad_norm": 0.1804710477590561,
+      "learning_rate": 1.9470451176040343e-05,
+      "loss": 0.5037,
+      "step": 1490
+    },
+    {
+      "epoch": 0.33258978362703545,
+      "grad_norm": 0.16383785009384155,
+      "learning_rate": 1.9469695226432667e-05,
+      "loss": 0.4991,
+      "step": 1491
+    },
+    {
+      "epoch": 0.33281284853892484,
+      "grad_norm": 0.16517230868339539,
+      "learning_rate": 1.9468938752333698e-05,
+      "loss": 0.4858,
+      "step": 1492
+    },
+    {
+      "epoch": 0.3330359134508142,
+      "grad_norm": 0.1879924088716507,
+      "learning_rate": 1.946818175378534e-05,
+      "loss": 0.5003,
+      "step": 1493
+    },
+    {
+      "epoch": 0.3332589783627036,
+      "grad_norm": 0.16963818669319153,
+      "learning_rate": 1.9467424230829514e-05,
+      "loss": 0.5081,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3334820432745929,
+      "grad_norm": 0.1686849743127823,
+      "learning_rate": 1.946666618350819e-05,
+      "loss": 0.5384,
+      "step": 1495
+    },
+    {
+      "epoch": 0.33370510818648225,
+      "grad_norm": 0.18511471152305603,
+      "learning_rate": 1.946590761186334e-05,
+      "loss": 0.5387,
+      "step": 1496
+    },
+    {
+      "epoch": 0.33392817309837164,
+      "grad_norm": 0.15684032440185547,
+      "learning_rate": 1.9465148515936986e-05,
+      "loss": 0.5051,
+      "step": 1497
+    },
+    {
+      "epoch": 0.334151238010261,
+      "grad_norm": 0.18369467556476593,
+      "learning_rate": 1.9464388895771165e-05,
+      "loss": 0.4973,
+      "step": 1498
+    },
+    {
+      "epoch": 0.33437430292215037,
+      "grad_norm": 0.17524507641792297,
+      "learning_rate": 1.9463628751407957e-05,
+      "loss": 0.5204,
+      "step": 1499
+    },
+    {
+      "epoch": 0.3345973678340397,
+      "grad_norm": 0.15974605083465576,
+      "learning_rate": 1.946286808288946e-05,
+      "loss": 0.5128,
+      "step": 1500
+    },
+    {
+      "epoch": 0.33482043274592904,
+      "grad_norm": 0.1720893234014511,
+      "learning_rate": 1.9462106890257805e-05,
+      "loss": 0.531,
+      "step": 1501
+    },
+    {
+      "epoch": 0.33504349765781843,
+      "grad_norm": 0.1723415106534958,
+      "learning_rate": 1.946134517355515e-05,
+      "loss": 0.5344,
+      "step": 1502
+    },
+    {
+      "epoch": 0.33526656256970777,
+      "grad_norm": 0.1691267043352127,
+      "learning_rate": 1.9460582932823685e-05,
+      "loss": 0.4693,
+      "step": 1503
+    },
+    {
+      "epoch": 0.33548962748159716,
+      "grad_norm": 0.18036657571792603,
+      "learning_rate": 1.945982016810563e-05,
+      "loss": 0.5175,
+      "step": 1504
+    },
+    {
+      "epoch": 0.3357126923934865,
+      "grad_norm": 0.16378861665725708,
+      "learning_rate": 1.9459056879443227e-05,
+      "loss": 0.4967,
+      "step": 1505
+    },
+    {
+      "epoch": 0.33593575730537584,
+      "grad_norm": 0.16954581439495087,
+      "learning_rate": 1.9458293066878754e-05,
+      "loss": 0.5024,
+      "step": 1506
+    },
+    {
+      "epoch": 0.33615882221726523,
+      "grad_norm": 0.16542184352874756,
+      "learning_rate": 1.9457528730454516e-05,
+      "loss": 0.525,
+      "step": 1507
+    },
+    {
+      "epoch": 0.33638188712915457,
+      "grad_norm": 0.16997484862804413,
+      "learning_rate": 1.9456763870212853e-05,
+      "loss": 0.5177,
+      "step": 1508
+    },
+    {
+      "epoch": 0.33660495204104396,
+      "grad_norm": 0.1820513904094696,
+      "learning_rate": 1.945599848619611e-05,
+      "loss": 0.5173,
+      "step": 1509
+    },
+    {
+      "epoch": 0.3368280169529333,
+      "grad_norm": 0.1801905781030655,
+      "learning_rate": 1.94552325784467e-05,
+      "loss": 0.4951,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3370510818648227,
+      "grad_norm": 0.16981808841228485,
+      "learning_rate": 1.9454466147007032e-05,
+      "loss": 0.5359,
+      "step": 1511
+    },
+    {
+      "epoch": 0.337274146776712,
+      "grad_norm": 0.16826239228248596,
+      "learning_rate": 1.9453699191919557e-05,
+      "loss": 0.4852,
+      "step": 1512
+    },
+    {
+      "epoch": 0.33749721168860136,
+      "grad_norm": 0.16905133426189423,
+      "learning_rate": 1.9452931713226752e-05,
+      "loss": 0.502,
+      "step": 1513
+    },
+    {
+      "epoch": 0.33772027660049075,
+      "grad_norm": 0.3108493685722351,
+      "learning_rate": 1.945216371097113e-05,
+      "loss": 0.5025,
+      "step": 1514
+    },
+    {
+      "epoch": 0.3379433415123801,
+      "grad_norm": 0.16754211485385895,
+      "learning_rate": 1.9451395185195224e-05,
+      "loss": 0.4913,
+      "step": 1515
+    },
+    {
+      "epoch": 0.3381664064242695,
+      "grad_norm": 0.16909672319889069,
+      "learning_rate": 1.9450626135941603e-05,
+      "loss": 0.5489,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3383894713361588,
+      "grad_norm": 0.17022110521793365,
+      "learning_rate": 1.944985656325286e-05,
+      "loss": 0.5266,
+      "step": 1517
+    },
+    {
+      "epoch": 0.33861253624804816,
+      "grad_norm": 0.16641078889369965,
+      "learning_rate": 1.9449086467171615e-05,
+      "loss": 0.4913,
+      "step": 1518
+    },
+    {
+      "epoch": 0.33883560115993755,
+      "grad_norm": 0.16541585326194763,
+      "learning_rate": 1.9448315847740527e-05,
+      "loss": 0.5116,
+      "step": 1519
+    },
+    {
+      "epoch": 0.3390586660718269,
+      "grad_norm": 0.175320103764534,
+      "learning_rate": 1.9447544705002273e-05,
+      "loss": 0.506,
+      "step": 1520
+    },
+    {
+      "epoch": 0.3392817309837163,
+      "grad_norm": 0.18363285064697266,
+      "learning_rate": 1.9446773038999566e-05,
+      "loss": 0.5119,
+      "step": 1521
+    },
+    {
+      "epoch": 0.3395047958956056,
+      "grad_norm": 0.17000706493854523,
+      "learning_rate": 1.944600084977515e-05,
+      "loss": 0.5286,
+      "step": 1522
+    },
+    {
+      "epoch": 0.339727860807495,
+      "grad_norm": 0.20967601239681244,
+      "learning_rate": 1.9445228137371784e-05,
+      "loss": 0.5366,
+      "step": 1523
+    },
+    {
+      "epoch": 0.33995092571938434,
+      "grad_norm": 0.1751982420682907,
+      "learning_rate": 1.9444454901832273e-05,
+      "loss": 0.5298,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3401739906312737,
+      "grad_norm": 0.16722087562084198,
+      "learning_rate": 1.944368114319944e-05,
+      "loss": 0.5248,
+      "step": 1525
+    },
+    {
+      "epoch": 0.3403970555431631,
+      "grad_norm": 0.16639317572116852,
+      "learning_rate": 1.9442906861516143e-05,
+      "loss": 0.4947,
+      "step": 1526
+    },
+    {
+      "epoch": 0.3406201204550524,
+      "grad_norm": 0.18173882365226746,
+      "learning_rate": 1.9442132056825268e-05,
+      "loss": 0.5198,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3408431853669418,
+      "grad_norm": 0.16811014711856842,
+      "learning_rate": 1.9441356729169725e-05,
+      "loss": 0.5193,
+      "step": 1528
+    },
+    {
+      "epoch": 0.34106625027883114,
+      "grad_norm": 0.18385176360607147,
+      "learning_rate": 1.944058087859246e-05,
+      "loss": 0.5419,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3412893151907205,
+      "grad_norm": 0.18335555493831635,
+      "learning_rate": 1.9439804505136437e-05,
+      "loss": 0.5442,
+      "step": 1530
+    },
+    {
+      "epoch": 0.34151238010260987,
+      "grad_norm": 0.16866059601306915,
+      "learning_rate": 1.9439027608844665e-05,
+      "loss": 0.4893,
+      "step": 1531
+    },
+    {
+      "epoch": 0.3417354450144992,
+      "grad_norm": 0.16770686209201813,
+      "learning_rate": 1.9438250189760168e-05,
+      "loss": 0.5276,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3419585099263886,
+      "grad_norm": 0.17907945811748505,
+      "learning_rate": 1.943747224792601e-05,
+      "loss": 0.5221,
+      "step": 1533
+    },
+    {
+      "epoch": 0.34218157483827794,
+      "grad_norm": 0.16425661742687225,
+      "learning_rate": 1.9436693783385273e-05,
+      "loss": 0.527,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3424046397501673,
+      "grad_norm": 0.20484143495559692,
+      "learning_rate": 1.9435914796181077e-05,
+      "loss": 0.5294,
+      "step": 1535
+    },
+    {
+      "epoch": 0.34262770466205666,
+      "grad_norm": 0.171040877699852,
+      "learning_rate": 1.9435135286356563e-05,
+      "loss": 0.5317,
+      "step": 1536
+    },
+    {
+      "epoch": 0.342850769573946,
+      "grad_norm": 0.1566307246685028,
+      "learning_rate": 1.943435525395491e-05,
+      "loss": 0.4908,
+      "step": 1537
+    },
+    {
+      "epoch": 0.3430738344858354,
+      "grad_norm": 0.15999655425548553,
+      "learning_rate": 1.9433574699019315e-05,
+      "loss": 0.5211,
+      "step": 1538
+    },
+    {
+      "epoch": 0.34329689939772473,
+      "grad_norm": 0.16879580914974213,
+      "learning_rate": 1.9432793621593013e-05,
+      "loss": 0.5473,
+      "step": 1539
+    },
+    {
+      "epoch": 0.3435199643096141,
+      "grad_norm": 0.16014404594898224,
+      "learning_rate": 1.943201202171927e-05,
+      "loss": 0.5369,
+      "step": 1540
+    },
+    {
+      "epoch": 0.34374302922150346,
+      "grad_norm": 0.1585661619901657,
+      "learning_rate": 1.943122989944137e-05,
+      "loss": 0.5203,
+      "step": 1541
+    },
+    {
+      "epoch": 0.3439660941333928,
+      "grad_norm": 0.16618762910366058,
+      "learning_rate": 1.943044725480263e-05,
+      "loss": 0.5163,
+      "step": 1542
+    },
+    {
+      "epoch": 0.3441891590452822,
+      "grad_norm": 0.1790398806333542,
+      "learning_rate": 1.9429664087846407e-05,
+      "loss": 0.492,
+      "step": 1543
+    },
+    {
+      "epoch": 0.3444122239571715,
+      "grad_norm": 0.17689375579357147,
+      "learning_rate": 1.9428880398616065e-05,
+      "loss": 0.5009,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3446352888690609,
+      "grad_norm": 0.16719898581504822,
+      "learning_rate": 1.942809618715502e-05,
+      "loss": 0.5064,
+      "step": 1545
+    },
+    {
+      "epoch": 0.34485835378095026,
+      "grad_norm": 0.1700020283460617,
+      "learning_rate": 1.9427311453506705e-05,
+      "loss": 0.5229,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3450814186928396,
+      "grad_norm": 0.17998361587524414,
+      "learning_rate": 1.9426526197714582e-05,
+      "loss": 0.5188,
+      "step": 1547
+    },
+    {
+      "epoch": 0.345304483604729,
+      "grad_norm": 0.17831408977508545,
+      "learning_rate": 1.9425740419822138e-05,
+      "loss": 0.4967,
+      "step": 1548
+    },
+    {
+      "epoch": 0.3455275485166183,
+      "grad_norm": 0.16397793591022491,
+      "learning_rate": 1.9424954119872904e-05,
+      "loss": 0.5319,
+      "step": 1549
+    },
+    {
+      "epoch": 0.3457506134285077,
+      "grad_norm": 0.17712676525115967,
+      "learning_rate": 1.9424167297910425e-05,
+      "loss": 0.5104,
+      "step": 1550
+    },
+    {
+      "epoch": 0.34597367834039705,
+      "grad_norm": 0.19361375272274017,
+      "learning_rate": 1.9423379953978277e-05,
+      "loss": 0.5113,
+      "step": 1551
+    },
+    {
+      "epoch": 0.34619674325228644,
+      "grad_norm": 0.1744556725025177,
+      "learning_rate": 1.9422592088120074e-05,
+      "loss": 0.5331,
+      "step": 1552
+    },
+    {
+      "epoch": 0.3464198081641758,
+      "grad_norm": 0.1747075766324997,
+      "learning_rate": 1.9421803700379454e-05,
+      "loss": 0.5257,
+      "step": 1553
+    },
+    {
+      "epoch": 0.3466428730760651,
+      "grad_norm": 0.17446283996105194,
+      "learning_rate": 1.9421014790800074e-05,
+      "loss": 0.5133,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3468659379879545,
+      "grad_norm": 0.1626337766647339,
+      "learning_rate": 1.9420225359425637e-05,
+      "loss": 0.4812,
+      "step": 1555
+    },
+    {
+      "epoch": 0.34708900289984385,
+      "grad_norm": 0.16959097981452942,
+      "learning_rate": 1.9419435406299863e-05,
+      "loss": 0.4954,
+      "step": 1556
+    },
+    {
+      "epoch": 0.34731206781173324,
+      "grad_norm": 0.17018267512321472,
+      "learning_rate": 1.9418644931466507e-05,
+      "loss": 0.4785,
+      "step": 1557
+    },
+    {
+      "epoch": 0.3475351327236226,
+      "grad_norm": 0.18296095728874207,
+      "learning_rate": 1.9417853934969347e-05,
+      "loss": 0.5244,
+      "step": 1558
+    },
+    {
+      "epoch": 0.3477581976355119,
+      "grad_norm": 0.1640445590019226,
+      "learning_rate": 1.9417062416852198e-05,
+      "loss": 0.5273,
+      "step": 1559
+    },
+    {
+      "epoch": 0.3479812625474013,
+      "grad_norm": 0.1624521166086197,
+      "learning_rate": 1.9416270377158896e-05,
+      "loss": 0.4939,
+      "step": 1560
+    },
+    {
+      "epoch": 0.34820432745929064,
+      "grad_norm": 0.16685132682323456,
+      "learning_rate": 1.941547781593331e-05,
+      "loss": 0.5118,
+      "step": 1561
+    },
+    {
+      "epoch": 0.34842739237118003,
+      "grad_norm": 0.1717056930065155,
+      "learning_rate": 1.9414684733219334e-05,
+      "loss": 0.5156,
+      "step": 1562
+    },
+    {
+      "epoch": 0.34865045728306937,
+      "grad_norm": 0.18991652131080627,
+      "learning_rate": 1.94138911290609e-05,
+      "loss": 0.5346,
+      "step": 1563
+    },
+    {
+      "epoch": 0.3488735221949587,
+      "grad_norm": 0.16688503324985504,
+      "learning_rate": 1.941309700350196e-05,
+      "loss": 0.5333,
+      "step": 1564
+    },
+    {
+      "epoch": 0.3490965871068481,
+      "grad_norm": 0.17251865565776825,
+      "learning_rate": 1.9412302356586494e-05,
+      "loss": 0.5113,
+      "step": 1565
+    },
+    {
+      "epoch": 0.34931965201873744,
+      "grad_norm": 0.15765444934368134,
+      "learning_rate": 1.941150718835852e-05,
+      "loss": 0.5305,
+      "step": 1566
+    },
+    {
+      "epoch": 0.34954271693062683,
+      "grad_norm": 0.18435880541801453,
+      "learning_rate": 1.9410711498862077e-05,
+      "loss": 0.5087,
+      "step": 1567
+    },
+    {
+      "epoch": 0.34976578184251617,
+      "grad_norm": 0.16191725432872772,
+      "learning_rate": 1.9409915288141235e-05,
+      "loss": 0.5001,
+      "step": 1568
+    },
+    {
+      "epoch": 0.34998884675440556,
+      "grad_norm": 0.19742515683174133,
+      "learning_rate": 1.9409118556240095e-05,
+      "loss": 0.5232,
+      "step": 1569
+    },
+    {
+      "epoch": 0.3502119116662949,
+      "grad_norm": 0.16742932796478271,
+      "learning_rate": 1.940832130320278e-05,
+      "loss": 0.5227,
+      "step": 1570
+    },
+    {
+      "epoch": 0.35043497657818423,
+      "grad_norm": 0.17970915138721466,
+      "learning_rate": 1.9407523529073455e-05,
+      "loss": 0.5228,
+      "step": 1571
+    },
+    {
+      "epoch": 0.3506580414900736,
+      "grad_norm": 0.16648298501968384,
+      "learning_rate": 1.9406725233896297e-05,
+      "loss": 0.4987,
+      "step": 1572
+    },
+    {
+      "epoch": 0.35088110640196296,
+      "grad_norm": 0.1718263030052185,
+      "learning_rate": 1.940592641771553e-05,
+      "loss": 0.5256,
+      "step": 1573
+    },
+    {
+      "epoch": 0.35110417131385235,
+      "grad_norm": 0.17099805176258087,
+      "learning_rate": 1.9405127080575387e-05,
+      "loss": 0.5371,
+      "step": 1574
+    },
+    {
+      "epoch": 0.3513272362257417,
+      "grad_norm": 0.16639132797718048,
+      "learning_rate": 1.9404327222520147e-05,
+      "loss": 0.5055,
+      "step": 1575
+    },
+    {
+      "epoch": 0.351550301137631,
+      "grad_norm": 0.15915724635124207,
+      "learning_rate": 1.9403526843594115e-05,
+      "loss": 0.5123,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3517733660495204,
+      "grad_norm": 0.190689817070961,
+      "learning_rate": 1.9402725943841608e-05,
+      "loss": 0.5149,
+      "step": 1577
+    },
+    {
+      "epoch": 0.35199643096140976,
+      "grad_norm": 0.1728559285402298,
+      "learning_rate": 1.9401924523306998e-05,
+      "loss": 0.518,
+      "step": 1578
+    },
+    {
+      "epoch": 0.35221949587329915,
+      "grad_norm": 0.15996624529361725,
+      "learning_rate": 1.9401122582034664e-05,
+      "loss": 0.4949,
+      "step": 1579
+    },
+    {
+      "epoch": 0.3524425607851885,
+      "grad_norm": 0.17822036147117615,
+      "learning_rate": 1.940032012006903e-05,
+      "loss": 0.5012,
+      "step": 1580
+    },
+    {
+      "epoch": 0.3526656256970778,
+      "grad_norm": 0.18252968788146973,
+      "learning_rate": 1.9399517137454534e-05,
+      "loss": 0.54,
+      "step": 1581
+    },
+    {
+      "epoch": 0.3528886906089672,
+      "grad_norm": 0.1667163223028183,
+      "learning_rate": 1.939871363423566e-05,
+      "loss": 0.5176,
+      "step": 1582
+    },
+    {
+      "epoch": 0.35311175552085655,
+      "grad_norm": 0.17513629794120789,
+      "learning_rate": 1.9397909610456897e-05,
+      "loss": 0.5182,
+      "step": 1583
+    },
+    {
+      "epoch": 0.35333482043274594,
+      "grad_norm": 0.16966688632965088,
+      "learning_rate": 1.939710506616279e-05,
+      "loss": 0.5154,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3535578853446353,
+      "grad_norm": 0.17325864732265472,
+      "learning_rate": 1.9396300001397888e-05,
+      "loss": 0.5321,
+      "step": 1585
+    },
+    {
+      "epoch": 0.3537809502565247,
+      "grad_norm": 0.17322127521038055,
+      "learning_rate": 1.939549441620679e-05,
+      "loss": 0.5206,
+      "step": 1586
+    },
+    {
+      "epoch": 0.354004015168414,
+      "grad_norm": 0.16276715695858002,
+      "learning_rate": 1.9394688310634114e-05,
+      "loss": 0.4874,
+      "step": 1587
+    },
+    {
+      "epoch": 0.35422708008030335,
+      "grad_norm": 0.1607387810945511,
+      "learning_rate": 1.93938816847245e-05,
+      "loss": 0.4813,
+      "step": 1588
+    },
+    {
+      "epoch": 0.35445014499219274,
+      "grad_norm": 0.16621682047843933,
+      "learning_rate": 1.939307453852263e-05,
+      "loss": 0.533,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3546732099040821,
+      "grad_norm": 0.165096253156662,
+      "learning_rate": 1.9392266872073207e-05,
+      "loss": 0.5115,
+      "step": 1590
+    },
+    {
+      "epoch": 0.35489627481597147,
+      "grad_norm": 0.17987790703773499,
+      "learning_rate": 1.9391458685420966e-05,
+      "loss": 0.5439,
+      "step": 1591
+    },
+    {
+      "epoch": 0.3551193397278608,
+      "grad_norm": 0.1641106754541397,
+      "learning_rate": 1.939064997861067e-05,
+      "loss": 0.462,
+      "step": 1592
+    },
+    {
+      "epoch": 0.35534240463975014,
+      "grad_norm": 0.23700258135795593,
+      "learning_rate": 1.9389840751687105e-05,
+      "loss": 0.4945,
+      "step": 1593
+    },
+    {
+      "epoch": 0.35556546955163953,
+      "grad_norm": 0.1738569736480713,
+      "learning_rate": 1.9389031004695095e-05,
+      "loss": 0.5247,
+      "step": 1594
+    },
+    {
+      "epoch": 0.35578853446352887,
+      "grad_norm": 0.16523152589797974,
+      "learning_rate": 1.9388220737679493e-05,
+      "loss": 0.5231,
+      "step": 1595
+    },
+    {
+      "epoch": 0.35601159937541826,
+      "grad_norm": 0.1748376339673996,
+      "learning_rate": 1.9387409950685167e-05,
+      "loss": 0.5048,
+      "step": 1596
+    },
+    {
+      "epoch": 0.3562346642873076,
+      "grad_norm": 0.20013664662837982,
+      "learning_rate": 1.938659864375703e-05,
+      "loss": 0.5284,
+      "step": 1597
+    },
+    {
+      "epoch": 0.356457729199197,
+      "grad_norm": 0.18472377955913544,
+      "learning_rate": 1.938578681694002e-05,
+      "loss": 0.5063,
+      "step": 1598
+    },
+    {
+      "epoch": 0.35668079411108633,
+      "grad_norm": 0.17718833684921265,
+      "learning_rate": 1.9384974470279093e-05,
+      "loss": 0.5156,
+      "step": 1599
+    },
+    {
+      "epoch": 0.35690385902297567,
+      "grad_norm": 0.16550204157829285,
+      "learning_rate": 1.938416160381925e-05,
+      "loss": 0.487,
+      "step": 1600
+    },
+    {
+      "epoch": 0.35712692393486506,
+      "grad_norm": 0.17269675433635712,
+      "learning_rate": 1.93833482176055e-05,
+      "loss": 0.4989,
+      "step": 1601
+    },
+    {
+      "epoch": 0.3573499888467544,
+      "grad_norm": 0.17201650142669678,
+      "learning_rate": 1.938253431168291e-05,
+      "loss": 0.5075,
+      "step": 1602
+    },
+    {
+      "epoch": 0.3575730537586438,
+      "grad_norm": 0.17147138714790344,
+      "learning_rate": 1.938171988609655e-05,
+      "loss": 0.4967,
+      "step": 1603
+    },
+    {
+      "epoch": 0.3577961186705331,
+      "grad_norm": 0.20010094344615936,
+      "learning_rate": 1.938090494089153e-05,
+      "loss": 0.5335,
+      "step": 1604
+    },
+    {
+      "epoch": 0.35801918358242246,
+      "grad_norm": 0.19643662869930267,
+      "learning_rate": 1.9380089476112985e-05,
+      "loss": 0.586,
+      "step": 1605
+    },
+    {
+      "epoch": 0.35824224849431185,
+      "grad_norm": 0.16737660765647888,
+      "learning_rate": 1.937927349180608e-05,
+      "loss": 0.5203,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3584653134062012,
+      "grad_norm": 0.1626511663198471,
+      "learning_rate": 1.9378456988016015e-05,
+      "loss": 0.5057,
+      "step": 1607
+    },
+    {
+      "epoch": 0.3586883783180906,
+      "grad_norm": 0.18112361431121826,
+      "learning_rate": 1.9377639964788005e-05,
+      "loss": 0.5709,
+      "step": 1608
+    },
+    {
+      "epoch": 0.3589114432299799,
+      "grad_norm": 0.1701660305261612,
+      "learning_rate": 1.937682242216731e-05,
+      "loss": 0.5427,
+      "step": 1609
+    },
+    {
+      "epoch": 0.35913450814186926,
+      "grad_norm": 0.16982769966125488,
+      "learning_rate": 1.9376004360199202e-05,
+      "loss": 0.5414,
+      "step": 1610
+    },
+    {
+      "epoch": 0.35935757305375865,
+      "grad_norm": 0.1677013635635376,
+      "learning_rate": 1.9375185778928997e-05,
+      "loss": 0.5189,
+      "step": 1611
+    },
+    {
+      "epoch": 0.359580637965648,
+      "grad_norm": 0.16802898049354553,
+      "learning_rate": 1.9374366678402032e-05,
+      "loss": 0.5359,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3598037028775374,
+      "grad_norm": 0.1629040390253067,
+      "learning_rate": 1.9373547058663674e-05,
+      "loss": 0.5239,
+      "step": 1613
+    },
+    {
+      "epoch": 0.3600267677894267,
+      "grad_norm": 0.15942612290382385,
+      "learning_rate": 1.9372726919759318e-05,
+      "loss": 0.4927,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3602498327013161,
+      "grad_norm": 0.16228771209716797,
+      "learning_rate": 1.9371906261734387e-05,
+      "loss": 0.4795,
+      "step": 1615
+    },
+    {
+      "epoch": 0.36047289761320545,
+      "grad_norm": 0.18663759529590607,
+      "learning_rate": 1.9371085084634337e-05,
+      "loss": 0.4523,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3606959625250948,
+      "grad_norm": 0.1616244614124298,
+      "learning_rate": 1.9370263388504647e-05,
+      "loss": 0.5221,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3609190274369842,
+      "grad_norm": 0.16261187195777893,
+      "learning_rate": 1.936944117339083e-05,
+      "loss": 0.507,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3611420923488735,
+      "grad_norm": 0.16541439294815063,
+      "learning_rate": 1.9368618439338424e-05,
+      "loss": 0.5318,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3613651572607629,
+      "grad_norm": 0.17135834693908691,
+      "learning_rate": 1.9367795186392996e-05,
+      "loss": 0.5071,
+      "step": 1620
+    },
+    {
+      "epoch": 0.36158822217265224,
+      "grad_norm": 0.1599961221218109,
+      "learning_rate": 1.936697141460015e-05,
+      "loss": 0.5031,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3618112870845416,
+      "grad_norm": 0.17239384353160858,
+      "learning_rate": 1.9366147124005504e-05,
+      "loss": 0.5138,
+      "step": 1622
+    },
+    {
+      "epoch": 0.36203435199643097,
+      "grad_norm": 0.1636582612991333,
+      "learning_rate": 1.9365322314654714e-05,
+      "loss": 0.5197,
+      "step": 1623
+    },
+    {
+      "epoch": 0.3622574169083203,
+      "grad_norm": 0.17068350315093994,
+      "learning_rate": 1.9364496986593463e-05,
+      "loss": 0.4886,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3624804818202097,
+      "grad_norm": 0.17033697664737701,
+      "learning_rate": 1.9363671139867467e-05,
+      "loss": 0.5232,
+      "step": 1625
+    },
+    {
+      "epoch": 0.36270354673209904,
+      "grad_norm": 0.16417935490608215,
+      "learning_rate": 1.936284477452246e-05,
+      "loss": 0.5384,
+      "step": 1626
+    },
+    {
+      "epoch": 0.36292661164398843,
+      "grad_norm": 0.1705051213502884,
+      "learning_rate": 1.9362017890604215e-05,
+      "loss": 0.517,
+      "step": 1627
+    },
+    {
+      "epoch": 0.36314967655587777,
+      "grad_norm": 0.17601191997528076,
+      "learning_rate": 1.9361190488158535e-05,
+      "loss": 0.5141,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3633727414677671,
+      "grad_norm": 0.173202782869339,
+      "learning_rate": 1.936036256723124e-05,
+      "loss": 0.5133,
+      "step": 1629
+    },
+    {
+      "epoch": 0.3635958063796565,
+      "grad_norm": 0.1655133217573166,
+      "learning_rate": 1.935953412786818e-05,
+      "loss": 0.5278,
+      "step": 1630
+    },
+    {
+      "epoch": 0.36381887129154583,
+      "grad_norm": 0.1649170219898224,
+      "learning_rate": 1.9358705170115253e-05,
+      "loss": 0.5101,
+      "step": 1631
+    },
+    {
+      "epoch": 0.3640419362034352,
+      "grad_norm": 0.4769546389579773,
+      "learning_rate": 1.9357875694018364e-05,
+      "loss": 0.5325,
+      "step": 1632
+    },
+    {
+      "epoch": 0.36426500111532456,
+      "grad_norm": 0.16675199568271637,
+      "learning_rate": 1.9357045699623452e-05,
+      "loss": 0.5373,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3644880660272139,
+      "grad_norm": 0.18537083268165588,
+      "learning_rate": 1.9356215186976496e-05,
+      "loss": 0.5077,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3647111309391033,
+      "grad_norm": 0.1646578013896942,
+      "learning_rate": 1.935538415612349e-05,
+      "loss": 0.5163,
+      "step": 1635
+    },
+    {
+      "epoch": 0.3649341958509926,
+      "grad_norm": 0.17558851838111877,
+      "learning_rate": 1.935455260711046e-05,
+      "loss": 0.5269,
+      "step": 1636
+    },
+    {
+      "epoch": 0.365157260762882,
+      "grad_norm": 0.1750616431236267,
+      "learning_rate": 1.9353720539983462e-05,
+      "loss": 0.5185,
+      "step": 1637
+    },
+    {
+      "epoch": 0.36538032567477136,
+      "grad_norm": 0.1686229407787323,
+      "learning_rate": 1.9352887954788583e-05,
+      "loss": 0.5033,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3656033905866607,
+      "grad_norm": 0.17237605154514313,
+      "learning_rate": 1.935205485157194e-05,
+      "loss": 0.5071,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3658264554985501,
+      "grad_norm": 0.17209577560424805,
+      "learning_rate": 1.9351221230379673e-05,
+      "loss": 0.5049,
+      "step": 1640
+    },
+    {
+      "epoch": 0.3660495204104394,
+      "grad_norm": 0.1662733256816864,
+      "learning_rate": 1.9350387091257952e-05,
+      "loss": 0.5399,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3662725853223288,
+      "grad_norm": 0.17135699093341827,
+      "learning_rate": 1.9349552434252976e-05,
+      "loss": 0.5175,
+      "step": 1642
+    },
+    {
+      "epoch": 0.36649565023421815,
+      "grad_norm": 0.17738810181617737,
+      "learning_rate": 1.9348717259410975e-05,
+      "loss": 0.5464,
+      "step": 1643
+    },
+    {
+      "epoch": 0.36671871514610754,
+      "grad_norm": 0.17101503908634186,
+      "learning_rate": 1.9347881566778208e-05,
+      "loss": 0.5162,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3669417800579969,
+      "grad_norm": 0.16841377317905426,
+      "learning_rate": 1.934704535640096e-05,
+      "loss": 0.5244,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3671648449698862,
+      "grad_norm": 0.17288081347942352,
+      "learning_rate": 1.9346208628325543e-05,
+      "loss": 0.5487,
+      "step": 1646
+    },
+    {
+      "epoch": 0.3673879098817756,
+      "grad_norm": 0.1731618493795395,
+      "learning_rate": 1.93453713825983e-05,
+      "loss": 0.4924,
+      "step": 1647
+    },
+    {
+      "epoch": 0.36761097479366495,
+      "grad_norm": 0.18122443556785583,
+      "learning_rate": 1.934453361926561e-05,
+      "loss": 0.5105,
+      "step": 1648
+    },
+    {
+      "epoch": 0.36783403970555434,
+      "grad_norm": 0.17434357106685638,
+      "learning_rate": 1.9343695338373866e-05,
+      "loss": 0.5185,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3680571046174437,
+      "grad_norm": 0.19195139408111572,
+      "learning_rate": 1.93428565399695e-05,
+      "loss": 0.5074,
+      "step": 1650
+    },
+    {
+      "epoch": 0.368280169529333,
+      "grad_norm": 0.17529140412807465,
+      "learning_rate": 1.9342017224098974e-05,
+      "loss": 0.5355,
+      "step": 1651
+    },
+    {
+      "epoch": 0.3685032344412224,
+      "grad_norm": 0.16739797592163086,
+      "learning_rate": 1.9341177390808768e-05,
+      "loss": 0.5128,
+      "step": 1652
+    },
+    {
+      "epoch": 0.36872629935311174,
+      "grad_norm": 0.17157401144504547,
+      "learning_rate": 1.9340337040145397e-05,
+      "loss": 0.5117,
+      "step": 1653
+    },
+    {
+      "epoch": 0.36894936426500113,
+      "grad_norm": 0.166087806224823,
+      "learning_rate": 1.933949617215541e-05,
+      "loss": 0.5182,
+      "step": 1654
+    },
+    {
+      "epoch": 0.36917242917689047,
+      "grad_norm": 0.17426824569702148,
+      "learning_rate": 1.9338654786885377e-05,
+      "loss": 0.5141,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3693954940887798,
+      "grad_norm": 0.18022432923316956,
+      "learning_rate": 1.93378128843819e-05,
+      "loss": 0.5097,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3696185590006692,
+      "grad_norm": 0.1715417206287384,
+      "learning_rate": 1.933697046469161e-05,
+      "loss": 0.525,
+      "step": 1657
+    },
+    {
+      "epoch": 0.36984162391255854,
+      "grad_norm": 0.17939765751361847,
+      "learning_rate": 1.9336127527861158e-05,
+      "loss": 0.5166,
+      "step": 1658
+    },
+    {
+      "epoch": 0.37006468882444793,
+      "grad_norm": 0.1856374740600586,
+      "learning_rate": 1.9335284073937242e-05,
+      "loss": 0.5078,
+      "step": 1659
+    },
+    {
+      "epoch": 0.37028775373633727,
+      "grad_norm": 0.16677772998809814,
+      "learning_rate": 1.9334440102966567e-05,
+      "loss": 0.5427,
+      "step": 1660
+    },
+    {
+      "epoch": 0.37051081864822666,
+      "grad_norm": 0.17052651941776276,
+      "learning_rate": 1.933359561499589e-05,
+      "loss": 0.4976,
+      "step": 1661
+    },
+    {
+      "epoch": 0.370733883560116,
+      "grad_norm": 0.180454283952713,
+      "learning_rate": 1.9332750610071972e-05,
+      "loss": 0.4807,
+      "step": 1662
+    },
+    {
+      "epoch": 0.37095694847200533,
+      "grad_norm": 0.16270607709884644,
+      "learning_rate": 1.9331905088241623e-05,
+      "loss": 0.5079,
+      "step": 1663
+    },
+    {
+      "epoch": 0.3711800133838947,
+      "grad_norm": 0.1687788963317871,
+      "learning_rate": 1.9331059049551668e-05,
+      "loss": 0.5208,
+      "step": 1664
+    },
+    {
+      "epoch": 0.37140307829578406,
+      "grad_norm": 0.16753818094730377,
+      "learning_rate": 1.933021249404897e-05,
+      "loss": 0.4991,
+      "step": 1665
+    },
+    {
+      "epoch": 0.37162614320767345,
+      "grad_norm": 0.1763564944267273,
+      "learning_rate": 1.9329365421780414e-05,
+      "loss": 0.5321,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3718492081195628,
+      "grad_norm": 0.1733461618423462,
+      "learning_rate": 1.932851783279292e-05,
+      "loss": 0.5331,
+      "step": 1667
+    },
+    {
+      "epoch": 0.3720722730314521,
+      "grad_norm": 0.1711835116147995,
+      "learning_rate": 1.9327669727133424e-05,
+      "loss": 0.4764,
+      "step": 1668
+    },
+    {
+      "epoch": 0.3722953379433415,
+      "grad_norm": 0.190887913107872,
+      "learning_rate": 1.932682110484891e-05,
+      "loss": 0.5254,
+      "step": 1669
+    },
+    {
+      "epoch": 0.37251840285523086,
+      "grad_norm": 0.16770336031913757,
+      "learning_rate": 1.9325971965986373e-05,
+      "loss": 0.4989,
+      "step": 1670
+    },
+    {
+      "epoch": 0.37274146776712025,
+      "grad_norm": 0.2075570970773697,
+      "learning_rate": 1.9325122310592846e-05,
+      "loss": 0.515,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3729645326790096,
+      "grad_norm": 0.17692680656909943,
+      "learning_rate": 1.9324272138715388e-05,
+      "loss": 0.5442,
+      "step": 1672
+    },
+    {
+      "epoch": 0.373187597590899,
+      "grad_norm": 0.17124991118907928,
+      "learning_rate": 1.932342145040109e-05,
+      "loss": 0.5066,
+      "step": 1673
+    },
+    {
+      "epoch": 0.3734106625027883,
+      "grad_norm": 0.16621904075145721,
+      "learning_rate": 1.932257024569706e-05,
+      "loss": 0.5066,
+      "step": 1674
+    },
+    {
+      "epoch": 0.37363372741467765,
+      "grad_norm": 0.17721469700336456,
+      "learning_rate": 1.932171852465045e-05,
+      "loss": 0.4789,
+      "step": 1675
+    },
+    {
+      "epoch": 0.37385679232656704,
+      "grad_norm": 0.16972602903842926,
+      "learning_rate": 1.9320866287308433e-05,
+      "loss": 0.4922,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3740798572384564,
+      "grad_norm": 0.17205440998077393,
+      "learning_rate": 1.9320013533718208e-05,
+      "loss": 0.4909,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3743029221503458,
+      "grad_norm": 0.16761687397956848,
+      "learning_rate": 1.9319160263927013e-05,
+      "loss": 0.513,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3745259870622351,
+      "grad_norm": 0.16278418898582458,
+      "learning_rate": 1.93183064779821e-05,
+      "loss": 0.508,
+      "step": 1679
+    },
+    {
+      "epoch": 0.37474905197412445,
+      "grad_norm": 0.17359983921051025,
+      "learning_rate": 1.931745217593076e-05,
+      "loss": 0.5472,
+      "step": 1680
+    },
+    {
+      "epoch": 0.37497211688601384,
+      "grad_norm": 0.16624867916107178,
+      "learning_rate": 1.931659735782031e-05,
+      "loss": 0.53,
+      "step": 1681
+    },
+    {
+      "epoch": 0.3751951817979032,
+      "grad_norm": 0.16128243505954742,
+      "learning_rate": 1.9315742023698095e-05,
+      "loss": 0.474,
+      "step": 1682
+    },
+    {
+      "epoch": 0.37541824670979257,
+      "grad_norm": 0.19452917575836182,
+      "learning_rate": 1.9314886173611487e-05,
+      "loss": 0.5527,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3756413116216819,
+      "grad_norm": 0.18000726401805878,
+      "learning_rate": 1.931402980760789e-05,
+      "loss": 0.5185,
+      "step": 1684
+    },
+    {
+      "epoch": 0.37586437653357124,
+      "grad_norm": 0.1593390852212906,
+      "learning_rate": 1.9313172925734736e-05,
+      "loss": 0.503,
+      "step": 1685
+    },
+    {
+      "epoch": 0.37608744144546064,
+      "grad_norm": 0.1637982428073883,
+      "learning_rate": 1.931231552803948e-05,
+      "loss": 0.5011,
+      "step": 1686
+    },
+    {
+      "epoch": 0.37631050635734997,
+      "grad_norm": 0.18159234523773193,
+      "learning_rate": 1.931145761456962e-05,
+      "loss": 0.501,
+      "step": 1687
+    },
+    {
+      "epoch": 0.37653357126923936,
+      "grad_norm": 0.17507293820381165,
+      "learning_rate": 1.9310599185372657e-05,
+      "loss": 0.5219,
+      "step": 1688
+    },
+    {
+      "epoch": 0.3767566361811287,
+      "grad_norm": 0.16377007961273193,
+      "learning_rate": 1.9309740240496152e-05,
+      "loss": 0.5096,
+      "step": 1689
+    },
+    {
+      "epoch": 0.3769797010930181,
+      "grad_norm": 0.171824112534523,
+      "learning_rate": 1.930888077998767e-05,
+      "loss": 0.5366,
+      "step": 1690
+    },
+    {
+      "epoch": 0.37720276600490743,
+      "grad_norm": 0.18775586783885956,
+      "learning_rate": 1.9308020803894813e-05,
+      "loss": 0.5018,
+      "step": 1691
+    },
+    {
+      "epoch": 0.37742583091679677,
+      "grad_norm": 0.1640700101852417,
+      "learning_rate": 1.9307160312265216e-05,
+      "loss": 0.5159,
+      "step": 1692
+    },
+    {
+      "epoch": 0.37764889582868616,
+      "grad_norm": 0.1758948117494583,
+      "learning_rate": 1.9306299305146535e-05,
+      "loss": 0.5171,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3778719607405755,
+      "grad_norm": 0.1625639647245407,
+      "learning_rate": 1.9305437782586463e-05,
+      "loss": 0.4989,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3780950256524649,
+      "grad_norm": 0.15426993370056152,
+      "learning_rate": 1.9304575744632708e-05,
+      "loss": 0.4781,
+      "step": 1695
+    },
+    {
+      "epoch": 0.3783180905643542,
+      "grad_norm": 0.18176917731761932,
+      "learning_rate": 1.9303713191333025e-05,
+      "loss": 0.5307,
+      "step": 1696
+    },
+    {
+      "epoch": 0.37854115547624356,
+      "grad_norm": 0.16359803080558777,
+      "learning_rate": 1.930285012273518e-05,
+      "loss": 0.5067,
+      "step": 1697
+    },
+    {
+      "epoch": 0.37876422038813296,
+      "grad_norm": 0.16793343424797058,
+      "learning_rate": 1.930198653888698e-05,
+      "loss": 0.5234,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3789872853000223,
+      "grad_norm": 0.17032268643379211,
+      "learning_rate": 1.930112243983625e-05,
+      "loss": 0.5521,
+      "step": 1699
+    },
+    {
+      "epoch": 0.3792103502119117,
+      "grad_norm": 0.18838275969028473,
+      "learning_rate": 1.930025782563086e-05,
+      "loss": 0.5203,
+      "step": 1700
+    },
+    {
+      "epoch": 0.379433415123801,
+      "grad_norm": 0.1813000738620758,
+      "learning_rate": 1.9299392696318683e-05,
+      "loss": 0.5299,
+      "step": 1701
+    },
+    {
+      "epoch": 0.3796564800356904,
+      "grad_norm": 0.15768998861312866,
+      "learning_rate": 1.9298527051947645e-05,
+      "loss": 0.4802,
+      "step": 1702
+    },
+    {
+      "epoch": 0.37987954494757975,
+      "grad_norm": 0.15893371403217316,
+      "learning_rate": 1.9297660892565692e-05,
+      "loss": 0.4939,
+      "step": 1703
+    },
+    {
+      "epoch": 0.3801026098594691,
+      "grad_norm": 0.16448427736759186,
+      "learning_rate": 1.929679421822079e-05,
+      "loss": 0.503,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3803256747713585,
+      "grad_norm": 0.1799025982618332,
+      "learning_rate": 1.9295927028960947e-05,
+      "loss": 0.5146,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3805487396832478,
+      "grad_norm": 0.17270612716674805,
+      "learning_rate": 1.9295059324834193e-05,
+      "loss": 0.5472,
+      "step": 1706
+    },
+    {
+      "epoch": 0.3807718045951372,
+      "grad_norm": 0.1640714704990387,
+      "learning_rate": 1.9294191105888586e-05,
+      "loss": 0.4847,
+      "step": 1707
+    },
+    {
+      "epoch": 0.38099486950702655,
+      "grad_norm": 0.16522546112537384,
+      "learning_rate": 1.9293322372172207e-05,
+      "loss": 0.4752,
+      "step": 1708
+    },
+    {
+      "epoch": 0.3812179344189159,
+      "grad_norm": 0.17741802334785461,
+      "learning_rate": 1.9292453123733184e-05,
+      "loss": 0.5246,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3814409993308053,
+      "grad_norm": 0.3004036545753479,
+      "learning_rate": 1.9291583360619653e-05,
+      "loss": 0.5032,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3816640642426946,
+      "grad_norm": 0.17262016236782074,
+      "learning_rate": 1.9290713082879786e-05,
+      "loss": 0.5208,
+      "step": 1711
+    },
+    {
+      "epoch": 0.381887129154584,
+      "grad_norm": 0.1745826005935669,
+      "learning_rate": 1.928984229056179e-05,
+      "loss": 0.5245,
+      "step": 1712
+    },
+    {
+      "epoch": 0.38211019406647334,
+      "grad_norm": 0.19315177202224731,
+      "learning_rate": 1.9288970983713893e-05,
+      "loss": 0.5029,
+      "step": 1713
+    },
+    {
+      "epoch": 0.3823332589783627,
+      "grad_norm": 0.17079593241214752,
+      "learning_rate": 1.9288099162384354e-05,
+      "loss": 0.5204,
+      "step": 1714
+    },
+    {
+      "epoch": 0.38255632389025207,
+      "grad_norm": 0.17363616824150085,
+      "learning_rate": 1.9287226826621457e-05,
+      "loss": 0.5195,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3827793888021414,
+      "grad_norm": 0.1668912172317505,
+      "learning_rate": 1.928635397647352e-05,
+      "loss": 0.5259,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3830024537140308,
+      "grad_norm": 0.162687286734581,
+      "learning_rate": 1.9285480611988886e-05,
+      "loss": 0.5135,
+      "step": 1717
+    },
+    {
+      "epoch": 0.38322551862592014,
+      "grad_norm": 0.1736724078655243,
+      "learning_rate": 1.9284606733215925e-05,
+      "loss": 0.4855,
+      "step": 1718
+    },
+    {
+      "epoch": 0.38344858353780953,
+      "grad_norm": 0.16885711252689362,
+      "learning_rate": 1.9283732340203045e-05,
+      "loss": 0.4934,
+      "step": 1719
+    },
+    {
+      "epoch": 0.38367164844969887,
+      "grad_norm": 0.17074720561504364,
+      "learning_rate": 1.928285743299867e-05,
+      "loss": 0.5167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3838947133615882,
+      "grad_norm": 0.17049984633922577,
+      "learning_rate": 1.9281982011651257e-05,
+      "loss": 0.498,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3841177782734776,
+      "grad_norm": 0.21770413219928741,
+      "learning_rate": 1.9281106076209296e-05,
+      "loss": 0.4869,
+      "step": 1722
+    },
+    {
+      "epoch": 0.38434084318536693,
+      "grad_norm": 0.17190971970558167,
+      "learning_rate": 1.9280229626721302e-05,
+      "loss": 0.5476,
+      "step": 1723
+    },
+    {
+      "epoch": 0.3845639080972563,
+      "grad_norm": 0.20358806848526,
+      "learning_rate": 1.9279352663235813e-05,
+      "loss": 0.5074,
+      "step": 1724
+    },
+    {
+      "epoch": 0.38478697300914566,
+      "grad_norm": 0.19548457860946655,
+      "learning_rate": 1.9278475185801404e-05,
+      "loss": 0.5246,
+      "step": 1725
+    },
+    {
+      "epoch": 0.385010037921035,
+      "grad_norm": 0.18699929118156433,
+      "learning_rate": 1.9277597194466674e-05,
+      "loss": 0.5267,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3852331028329244,
+      "grad_norm": 0.16484257578849792,
+      "learning_rate": 1.9276718689280258e-05,
+      "loss": 0.5108,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3854561677448137,
+      "grad_norm": 0.16136226058006287,
+      "learning_rate": 1.9275839670290804e-05,
+      "loss": 0.4973,
+      "step": 1728
+    },
+    {
+      "epoch": 0.3856792326567031,
+      "grad_norm": 0.16811032593250275,
+      "learning_rate": 1.9274960137547002e-05,
+      "loss": 0.4971,
+      "step": 1729
+    },
+    {
+      "epoch": 0.38590229756859246,
+      "grad_norm": 0.17258323729038239,
+      "learning_rate": 1.9274080091097568e-05,
+      "loss": 0.5513,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3861253624804818,
+      "grad_norm": 0.16782884299755096,
+      "learning_rate": 1.927319953099124e-05,
+      "loss": 0.5032,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3863484273923712,
+      "grad_norm": 0.1599043756723404,
+      "learning_rate": 1.9272318457276792e-05,
+      "loss": 0.5117,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3865714923042605,
+      "grad_norm": 0.18118049204349518,
+      "learning_rate": 1.9271436870003022e-05,
+      "loss": 0.5476,
+      "step": 1733
+    },
+    {
+      "epoch": 0.3867945572161499,
+      "grad_norm": 0.16125431656837463,
+      "learning_rate": 1.927055476921876e-05,
+      "loss": 0.496,
+      "step": 1734
+    },
+    {
+      "epoch": 0.38701762212803925,
+      "grad_norm": 0.17001016438007355,
+      "learning_rate": 1.9269672154972863e-05,
+      "loss": 0.5274,
+      "step": 1735
+    },
+    {
+      "epoch": 0.38724068703992864,
+      "grad_norm": 0.18982788920402527,
+      "learning_rate": 1.9268789027314208e-05,
+      "loss": 0.5423,
+      "step": 1736
+    },
+    {
+      "epoch": 0.387463751951818,
+      "grad_norm": 0.20157238841056824,
+      "learning_rate": 1.9267905386291716e-05,
+      "loss": 0.5204,
+      "step": 1737
+    },
+    {
+      "epoch": 0.3876868168637073,
+      "grad_norm": 0.16580817103385925,
+      "learning_rate": 1.926702123195433e-05,
+      "loss": 0.523,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3879098817755967,
+      "grad_norm": 0.1722583770751953,
+      "learning_rate": 1.926613656435101e-05,
+      "loss": 0.4689,
+      "step": 1739
+    },
+    {
+      "epoch": 0.38813294668748605,
+      "grad_norm": 0.16248932480812073,
+      "learning_rate": 1.9265251383530765e-05,
+      "loss": 0.4838,
+      "step": 1740
+    },
+    {
+      "epoch": 0.38835601159937544,
+      "grad_norm": 0.16789276897907257,
+      "learning_rate": 1.9264365689542616e-05,
+      "loss": 0.5191,
+      "step": 1741
+    },
+    {
+      "epoch": 0.3885790765112648,
+      "grad_norm": 0.16488412022590637,
+      "learning_rate": 1.926347948243562e-05,
+      "loss": 0.5124,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3888021414231541,
+      "grad_norm": 0.17527227103710175,
+      "learning_rate": 1.926259276225886e-05,
+      "loss": 0.5042,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3890252063350435,
+      "grad_norm": 0.16100718080997467,
+      "learning_rate": 1.926170552906145e-05,
+      "loss": 0.5103,
+      "step": 1744
+    },
+    {
+      "epoch": 0.38924827124693284,
+      "grad_norm": 0.1704183667898178,
+      "learning_rate": 1.926081778289253e-05,
+      "loss": 0.5252,
+      "step": 1745
+    },
+    {
+      "epoch": 0.38947133615882223,
+      "grad_norm": 0.15948736667633057,
+      "learning_rate": 1.9259929523801266e-05,
+      "loss": 0.4972,
+      "step": 1746
+    },
+    {
+      "epoch": 0.38969440107071157,
+      "grad_norm": 0.15783725678920746,
+      "learning_rate": 1.9259040751836858e-05,
+      "loss": 0.4862,
+      "step": 1747
+    },
+    {
+      "epoch": 0.38991746598260096,
+      "grad_norm": 0.22207237780094147,
+      "learning_rate": 1.9258151467048533e-05,
+      "loss": 0.5058,
+      "step": 1748
+    },
+    {
+      "epoch": 0.3901405308944903,
+      "grad_norm": 0.18270133435726166,
+      "learning_rate": 1.9257261669485544e-05,
+      "loss": 0.4872,
+      "step": 1749
+    },
+    {
+      "epoch": 0.39036359580637964,
+      "grad_norm": 0.15779659152030945,
+      "learning_rate": 1.925637135919717e-05,
+      "loss": 0.4972,
+      "step": 1750
+    },
+    {
+      "epoch": 0.39058666071826903,
+      "grad_norm": 0.16197752952575684,
+      "learning_rate": 1.9255480536232728e-05,
+      "loss": 0.5001,
+      "step": 1751
+    },
+    {
+      "epoch": 0.39080972563015837,
+      "grad_norm": 0.1598835587501526,
+      "learning_rate": 1.9254589200641556e-05,
+      "loss": 0.4991,
+      "step": 1752
+    },
+    {
+      "epoch": 0.39103279054204776,
+      "grad_norm": 0.1754835844039917,
+      "learning_rate": 1.925369735247302e-05,
+      "loss": 0.4858,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3912558554539371,
+      "grad_norm": 0.17721553146839142,
+      "learning_rate": 1.9252804991776513e-05,
+      "loss": 0.4954,
+      "step": 1754
+    },
+    {
+      "epoch": 0.39147892036582643,
+      "grad_norm": 0.16600117087364197,
+      "learning_rate": 1.9251912118601466e-05,
+      "loss": 0.5006,
+      "step": 1755
+    },
+    {
+      "epoch": 0.3917019852777158,
+      "grad_norm": 0.18240424990653992,
+      "learning_rate": 1.925101873299733e-05,
+      "loss": 0.5563,
+      "step": 1756
+    },
+    {
+      "epoch": 0.39192505018960516,
+      "grad_norm": 0.2154020071029663,
+      "learning_rate": 1.9250124835013583e-05,
+      "loss": 0.5213,
+      "step": 1757
+    },
+    {
+      "epoch": 0.39214811510149455,
+      "grad_norm": 0.17148853838443756,
+      "learning_rate": 1.9249230424699735e-05,
+      "loss": 0.5378,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3923711800133839,
+      "grad_norm": 0.17164389789104462,
+      "learning_rate": 1.9248335502105328e-05,
+      "loss": 0.5272,
+      "step": 1759
+    },
+    {
+      "epoch": 0.39259424492527323,
+      "grad_norm": 0.16211992502212524,
+      "learning_rate": 1.924744006727993e-05,
+      "loss": 0.5543,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3928173098371626,
+      "grad_norm": 0.17015716433525085,
+      "learning_rate": 1.924654412027313e-05,
+      "loss": 0.5281,
+      "step": 1761
+    },
+    {
+      "epoch": 0.39304037474905196,
+      "grad_norm": 0.15979216992855072,
+      "learning_rate": 1.924564766113455e-05,
+      "loss": 0.4938,
+      "step": 1762
+    },
+    {
+      "epoch": 0.39326343966094135,
+      "grad_norm": 0.16782401502132416,
+      "learning_rate": 1.924475068991385e-05,
+      "loss": 0.4876,
+      "step": 1763
+    },
+    {
+      "epoch": 0.3934865045728307,
+      "grad_norm": 0.1688111275434494,
+      "learning_rate": 1.9243853206660703e-05,
+      "loss": 0.503,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3937095694847201,
+      "grad_norm": 0.16509543359279633,
+      "learning_rate": 1.924295521142482e-05,
+      "loss": 0.4955,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3939326343966094,
+      "grad_norm": 0.17914824187755585,
+      "learning_rate": 1.9242056704255935e-05,
+      "loss": 0.5334,
+      "step": 1766
+    },
+    {
+      "epoch": 0.39415569930849875,
+      "grad_norm": 0.18346786499023438,
+      "learning_rate": 1.9241157685203817e-05,
+      "loss": 0.5135,
+      "step": 1767
+    },
+    {
+      "epoch": 0.39437876422038814,
+      "grad_norm": 0.18024152517318726,
+      "learning_rate": 1.9240258154318257e-05,
+      "loss": 0.5284,
+      "step": 1768
+    },
+    {
+      "epoch": 0.3946018291322775,
+      "grad_norm": 0.17404595017433167,
+      "learning_rate": 1.923935811164908e-05,
+      "loss": 0.5069,
+      "step": 1769
+    },
+    {
+      "epoch": 0.3948248940441669,
+      "grad_norm": 0.15942560136318207,
+      "learning_rate": 1.9238457557246128e-05,
+      "loss": 0.5034,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3950479589560562,
+      "grad_norm": 0.15909984707832336,
+      "learning_rate": 1.9237556491159285e-05,
+      "loss": 0.4762,
+      "step": 1771
+    },
+    {
+      "epoch": 0.39527102386794555,
+      "grad_norm": 0.16344551742076874,
+      "learning_rate": 1.9236654913438456e-05,
+      "loss": 0.5243,
+      "step": 1772
+    },
+    {
+      "epoch": 0.39549408877983494,
+      "grad_norm": 0.1672784835100174,
+      "learning_rate": 1.923575282413358e-05,
+      "loss": 0.5057,
+      "step": 1773
+    },
+    {
+      "epoch": 0.3957171536917243,
+      "grad_norm": 0.1585850566625595,
+      "learning_rate": 1.9234850223294613e-05,
+      "loss": 0.5074,
+      "step": 1774
+    },
+    {
+      "epoch": 0.39594021860361367,
+      "grad_norm": 0.17053529620170593,
+      "learning_rate": 1.9233947110971556e-05,
+      "loss": 0.5556,
+      "step": 1775
+    },
+    {
+      "epoch": 0.396163283515503,
+      "grad_norm": 0.16313889622688293,
+      "learning_rate": 1.9233043487214423e-05,
+      "loss": 0.491,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3963863484273924,
+      "grad_norm": 0.15941911935806274,
+      "learning_rate": 1.9232139352073265e-05,
+      "loss": 0.4862,
+      "step": 1777
+    },
+    {
+      "epoch": 0.39660941333928174,
+      "grad_norm": 0.1699916571378708,
+      "learning_rate": 1.9231234705598153e-05,
+      "loss": 0.542,
+      "step": 1778
+    },
+    {
+      "epoch": 0.3968324782511711,
+      "grad_norm": 0.17430098354816437,
+      "learning_rate": 1.9230329547839196e-05,
+      "loss": 0.5006,
+      "step": 1779
+    },
+    {
+      "epoch": 0.39705554316306046,
+      "grad_norm": 0.16042235493659973,
+      "learning_rate": 1.9229423878846535e-05,
+      "loss": 0.5087,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3972786080749498,
+      "grad_norm": 0.1637001633644104,
+      "learning_rate": 1.9228517698670316e-05,
+      "loss": 0.4966,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3975016729868392,
+      "grad_norm": 0.1627058982849121,
+      "learning_rate": 1.922761100736074e-05,
+      "loss": 0.4882,
+      "step": 1782
+    },
+    {
+      "epoch": 0.39772473789872853,
+      "grad_norm": 0.1667163074016571,
+      "learning_rate": 1.9226703804968022e-05,
+      "loss": 0.5295,
+      "step": 1783
+    },
+    {
+      "epoch": 0.39794780281061787,
+      "grad_norm": 0.16532334685325623,
+      "learning_rate": 1.9225796091542412e-05,
+      "loss": 0.5062,
+      "step": 1784
+    },
+    {
+      "epoch": 0.39817086772250726,
+      "grad_norm": 0.16788703203201294,
+      "learning_rate": 1.9224887867134178e-05,
+      "loss": 0.5276,
+      "step": 1785
+    },
+    {
+      "epoch": 0.3983939326343966,
+      "grad_norm": 0.17480318248271942,
+      "learning_rate": 1.9223979131793627e-05,
+      "loss": 0.52,
+      "step": 1786
+    },
+    {
+      "epoch": 0.398616997546286,
+      "grad_norm": 0.1601472944021225,
+      "learning_rate": 1.9223069885571094e-05,
+      "loss": 0.5263,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3988400624581753,
+      "grad_norm": 0.16812476515769958,
+      "learning_rate": 1.9222160128516932e-05,
+      "loss": 0.4831,
+      "step": 1788
+    },
+    {
+      "epoch": 0.39906312737006466,
+      "grad_norm": 0.1809302121400833,
+      "learning_rate": 1.9221249860681537e-05,
+      "loss": 0.4944,
+      "step": 1789
+    },
+    {
+      "epoch": 0.39928619228195406,
+      "grad_norm": 0.16555048525333405,
+      "learning_rate": 1.9220339082115317e-05,
+      "loss": 0.4851,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3995092571938434,
+      "grad_norm": 0.16044898331165314,
+      "learning_rate": 1.9219427792868722e-05,
+      "loss": 0.5377,
+      "step": 1791
+    },
+    {
+      "epoch": 0.3997323221057328,
+      "grad_norm": 0.16331616044044495,
+      "learning_rate": 1.921851599299222e-05,
+      "loss": 0.4978,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3999553870176221,
+      "grad_norm": 0.16315311193466187,
+      "learning_rate": 1.9217603682536315e-05,
+      "loss": 0.51,
+      "step": 1793
+    },
+    {
+      "epoch": 0.4001784519295115,
+      "grad_norm": 0.16615985333919525,
+      "learning_rate": 1.9216690861551544e-05,
+      "loss": 0.5343,
+      "step": 1794
+    },
+    {
+      "epoch": 0.40040151684140085,
+      "grad_norm": 0.1581142395734787,
+      "learning_rate": 1.9215777530088452e-05,
+      "loss": 0.5276,
+      "step": 1795
+    },
+    {
+      "epoch": 0.4006245817532902,
+      "grad_norm": 0.25778672099113464,
+      "learning_rate": 1.9214863688197634e-05,
+      "loss": 0.5265,
+      "step": 1796
+    },
+    {
+      "epoch": 0.4008476466651796,
+      "grad_norm": 0.21218818426132202,
+      "learning_rate": 1.92139493359297e-05,
+      "loss": 0.4929,
+      "step": 1797
+    },
+    {
+      "epoch": 0.4010707115770689,
+      "grad_norm": 0.1664579063653946,
+      "learning_rate": 1.9213034473335293e-05,
+      "loss": 0.5298,
+      "step": 1798
+    },
+    {
+      "epoch": 0.4012937764889583,
+      "grad_norm": 0.158608078956604,
+      "learning_rate": 1.9212119100465084e-05,
+      "loss": 0.513,
+      "step": 1799
+    },
+    {
+      "epoch": 0.40151684140084765,
+      "grad_norm": 0.15767832100391388,
+      "learning_rate": 1.9211203217369774e-05,
+      "loss": 0.5037,
+      "step": 1800
+    },
+    {
+      "epoch": 0.401739906312737,
+      "grad_norm": 0.1640487164258957,
+      "learning_rate": 1.921028682410009e-05,
+      "loss": 0.5102,
+      "step": 1801
+    },
+    {
+      "epoch": 0.4019629712246264,
+      "grad_norm": 0.1980050951242447,
+      "learning_rate": 1.9209369920706783e-05,
+      "loss": 0.4926,
+      "step": 1802
+    },
+    {
+      "epoch": 0.4021860361365157,
+      "grad_norm": 0.16443881392478943,
+      "learning_rate": 1.9208452507240642e-05,
+      "loss": 0.5165,
+      "step": 1803
+    },
+    {
+      "epoch": 0.4024091010484051,
+      "grad_norm": 0.17739985883235931,
+      "learning_rate": 1.920753458375248e-05,
+      "loss": 0.5141,
+      "step": 1804
+    },
+    {
+      "epoch": 0.40263216596029444,
+      "grad_norm": 0.1755398064851761,
+      "learning_rate": 1.9206616150293132e-05,
+      "loss": 0.5279,
+      "step": 1805
+    },
+    {
+      "epoch": 0.4028552308721838,
+      "grad_norm": 0.16583852469921112,
+      "learning_rate": 1.9205697206913473e-05,
+      "loss": 0.4838,
+      "step": 1806
+    },
+    {
+      "epoch": 0.40307829578407317,
+      "grad_norm": 0.1888515055179596,
+      "learning_rate": 1.9204777753664397e-05,
+      "loss": 0.5381,
+      "step": 1807
+    },
+    {
+      "epoch": 0.4033013606959625,
+      "grad_norm": 0.17295043170452118,
+      "learning_rate": 1.9203857790596826e-05,
+      "loss": 0.5062,
+      "step": 1808
+    },
+    {
+      "epoch": 0.4035244256078519,
+      "grad_norm": 0.16970674693584442,
+      "learning_rate": 1.9202937317761713e-05,
+      "loss": 0.5138,
+      "step": 1809
+    },
+    {
+      "epoch": 0.40374749051974124,
+      "grad_norm": 0.16952745616436005,
+      "learning_rate": 1.9202016335210047e-05,
+      "loss": 0.4829,
+      "step": 1810
+    },
+    {
+      "epoch": 0.40397055543163063,
+      "grad_norm": 0.16743586957454681,
+      "learning_rate": 1.9201094842992832e-05,
+      "loss": 0.5085,
+      "step": 1811
+    },
+    {
+      "epoch": 0.40419362034351997,
+      "grad_norm": 0.16140130162239075,
+      "learning_rate": 1.9200172841161108e-05,
+      "loss": 0.4983,
+      "step": 1812
+    },
+    {
+      "epoch": 0.4044166852554093,
+      "grad_norm": 0.1676797717809677,
+      "learning_rate": 1.9199250329765943e-05,
+      "loss": 0.5055,
+      "step": 1813
+    },
+    {
+      "epoch": 0.4046397501672987,
+      "grad_norm": 0.1718011051416397,
+      "learning_rate": 1.9198327308858427e-05,
+      "loss": 0.5021,
+      "step": 1814
+    },
+    {
+      "epoch": 0.40486281507918803,
+      "grad_norm": 0.1699821949005127,
+      "learning_rate": 1.9197403778489684e-05,
+      "loss": 0.5312,
+      "step": 1815
+    },
+    {
+      "epoch": 0.4050858799910774,
+      "grad_norm": 0.16701120138168335,
+      "learning_rate": 1.9196479738710865e-05,
+      "loss": 0.5028,
+      "step": 1816
+    },
+    {
+      "epoch": 0.40530894490296676,
+      "grad_norm": 0.17370650172233582,
+      "learning_rate": 1.9195555189573153e-05,
+      "loss": 0.5075,
+      "step": 1817
+    },
+    {
+      "epoch": 0.4055320098148561,
+      "grad_norm": 0.16744117438793182,
+      "learning_rate": 1.919463013112775e-05,
+      "loss": 0.4972,
+      "step": 1818
+    },
+    {
+      "epoch": 0.4057550747267455,
+      "grad_norm": 0.1703552007675171,
+      "learning_rate": 1.9193704563425896e-05,
+      "loss": 0.5215,
+      "step": 1819
+    },
+    {
+      "epoch": 0.4059781396386348,
+      "grad_norm": 0.1808227002620697,
+      "learning_rate": 1.919277848651885e-05,
+      "loss": 0.5048,
+      "step": 1820
+    },
+    {
+      "epoch": 0.4062012045505242,
+      "grad_norm": 0.17191959917545319,
+      "learning_rate": 1.9191851900457905e-05,
+      "loss": 0.5293,
+      "step": 1821
+    },
+    {
+      "epoch": 0.40642426946241356,
+      "grad_norm": 0.16464729607105255,
+      "learning_rate": 1.9190924805294388e-05,
+      "loss": 0.531,
+      "step": 1822
+    },
+    {
+      "epoch": 0.40664733437430295,
+      "grad_norm": 0.17217296361923218,
+      "learning_rate": 1.9189997201079638e-05,
+      "loss": 0.5221,
+      "step": 1823
+    },
+    {
+      "epoch": 0.4068703992861923,
+      "grad_norm": 0.15952859818935394,
+      "learning_rate": 1.918906908786504e-05,
+      "loss": 0.4898,
+      "step": 1824
+    },
+    {
+      "epoch": 0.4070934641980816,
+      "grad_norm": 0.15873339772224426,
+      "learning_rate": 1.9188140465701987e-05,
+      "loss": 0.4771,
+      "step": 1825
+    },
+    {
+      "epoch": 0.407316529109971,
+      "grad_norm": 0.16631244122982025,
+      "learning_rate": 1.9187211334641923e-05,
+      "loss": 0.5216,
+      "step": 1826
+    },
+    {
+      "epoch": 0.40753959402186035,
+      "grad_norm": 0.19430842995643616,
+      "learning_rate": 1.918628169473631e-05,
+      "loss": 0.532,
+      "step": 1827
+    },
+    {
+      "epoch": 0.40776265893374974,
+      "grad_norm": 0.1647733449935913,
+      "learning_rate": 1.9185351546036625e-05,
+      "loss": 0.4991,
+      "step": 1828
+    },
+    {
+      "epoch": 0.4079857238456391,
+      "grad_norm": 0.17242863774299622,
+      "learning_rate": 1.9184420888594398e-05,
+      "loss": 0.5095,
+      "step": 1829
+    },
+    {
+      "epoch": 0.4082087887575284,
+      "grad_norm": 0.15754511952400208,
+      "learning_rate": 1.9183489722461167e-05,
+      "loss": 0.5185,
+      "step": 1830
+    },
+    {
+      "epoch": 0.4084318536694178,
+      "grad_norm": 0.16371309757232666,
+      "learning_rate": 1.918255804768851e-05,
+      "loss": 0.5146,
+      "step": 1831
+    },
+    {
+      "epoch": 0.40865491858130715,
+      "grad_norm": 0.17599989473819733,
+      "learning_rate": 1.918162586432803e-05,
+      "loss": 0.5509,
+      "step": 1832
+    },
+    {
+      "epoch": 0.40887798349319654,
+      "grad_norm": 0.17950280010700226,
+      "learning_rate": 1.9180693172431353e-05,
+      "loss": 0.5236,
+      "step": 1833
+    },
+    {
+      "epoch": 0.4091010484050859,
+      "grad_norm": 0.1845085769891739,
+      "learning_rate": 1.917975997205014e-05,
+      "loss": 0.5155,
+      "step": 1834
+    },
+    {
+      "epoch": 0.4093241133169752,
+      "grad_norm": 0.16801810264587402,
+      "learning_rate": 1.9178826263236076e-05,
+      "loss": 0.5265,
+      "step": 1835
+    },
+    {
+      "epoch": 0.4095471782288646,
+      "grad_norm": 0.16314861178398132,
+      "learning_rate": 1.9177892046040875e-05,
+      "loss": 0.4922,
+      "step": 1836
+    },
+    {
+      "epoch": 0.40977024314075394,
+      "grad_norm": 0.16442064940929413,
+      "learning_rate": 1.9176957320516287e-05,
+      "loss": 0.5004,
+      "step": 1837
+    },
+    {
+      "epoch": 0.40999330805264333,
+      "grad_norm": 0.16156339645385742,
+      "learning_rate": 1.917602208671407e-05,
+      "loss": 0.5172,
+      "step": 1838
+    },
+    {
+      "epoch": 0.41021637296453267,
+      "grad_norm": 0.17251324653625488,
+      "learning_rate": 1.9175086344686035e-05,
+      "loss": 0.5432,
+      "step": 1839
+    },
+    {
+      "epoch": 0.41043943787642206,
+      "grad_norm": 0.1891004890203476,
+      "learning_rate": 1.9174150094484e-05,
+      "loss": 0.5276,
+      "step": 1840
+    },
+    {
+      "epoch": 0.4106625027883114,
+      "grad_norm": 0.1750495731830597,
+      "learning_rate": 1.917321333615983e-05,
+      "loss": 0.548,
+      "step": 1841
+    },
+    {
+      "epoch": 0.41088556770020074,
+      "grad_norm": 0.1698615849018097,
+      "learning_rate": 1.91722760697654e-05,
+      "loss": 0.5276,
+      "step": 1842
+    },
+    {
+      "epoch": 0.41110863261209013,
+      "grad_norm": 0.16396887600421906,
+      "learning_rate": 1.917133829535263e-05,
+      "loss": 0.5211,
+      "step": 1843
+    },
+    {
+      "epoch": 0.41133169752397947,
+      "grad_norm": 0.18432539701461792,
+      "learning_rate": 1.917040001297345e-05,
+      "loss": 0.5276,
+      "step": 1844
+    },
+    {
+      "epoch": 0.41155476243586886,
+      "grad_norm": 0.17017246782779694,
+      "learning_rate": 1.9169461222679836e-05,
+      "loss": 0.5312,
+      "step": 1845
+    },
+    {
+      "epoch": 0.4117778273477582,
+      "grad_norm": 0.16993820667266846,
+      "learning_rate": 1.9168521924523782e-05,
+      "loss": 0.4981,
+      "step": 1846
+    },
+    {
+      "epoch": 0.41200089225964753,
+      "grad_norm": 0.1594426929950714,
+      "learning_rate": 1.916758211855731e-05,
+      "loss": 0.5162,
+      "step": 1847
+    },
+    {
+      "epoch": 0.4122239571715369,
+      "grad_norm": 0.16228000819683075,
+      "learning_rate": 1.9166641804832474e-05,
+      "loss": 0.5086,
+      "step": 1848
+    },
+    {
+      "epoch": 0.41244702208342626,
+      "grad_norm": 0.2065616101026535,
+      "learning_rate": 1.9165700983401354e-05,
+      "loss": 0.4916,
+      "step": 1849
+    },
+    {
+      "epoch": 0.41267008699531565,
+      "grad_norm": 0.16846513748168945,
+      "learning_rate": 1.916475965431606e-05,
+      "loss": 0.5246,
+      "step": 1850
+    },
+    {
+      "epoch": 0.412893151907205,
+      "grad_norm": 0.19557945430278778,
+      "learning_rate": 1.9163817817628728e-05,
+      "loss": 0.5208,
+      "step": 1851
+    },
+    {
+      "epoch": 0.4131162168190944,
+      "grad_norm": 0.17081156373023987,
+      "learning_rate": 1.916287547339152e-05,
+      "loss": 0.5534,
+      "step": 1852
+    },
+    {
+      "epoch": 0.4133392817309837,
+      "grad_norm": 0.16580162942409515,
+      "learning_rate": 1.9161932621656634e-05,
+      "loss": 0.5152,
+      "step": 1853
+    },
+    {
+      "epoch": 0.41356234664287306,
+      "grad_norm": 0.18416714668273926,
+      "learning_rate": 1.9160989262476288e-05,
+      "loss": 0.5064,
+      "step": 1854
+    },
+    {
+      "epoch": 0.41378541155476245,
+      "grad_norm": 0.16161711513996124,
+      "learning_rate": 1.916004539590273e-05,
+      "loss": 0.5008,
+      "step": 1855
+    },
+    {
+      "epoch": 0.4140084764666518,
+      "grad_norm": 0.17786547541618347,
+      "learning_rate": 1.9159101021988244e-05,
+      "loss": 0.5168,
+      "step": 1856
+    },
+    {
+      "epoch": 0.4142315413785412,
+      "grad_norm": 0.15692083537578583,
+      "learning_rate": 1.9158156140785125e-05,
+      "loss": 0.5026,
+      "step": 1857
+    },
+    {
+      "epoch": 0.4144546062904305,
+      "grad_norm": 0.22050227224826813,
+      "learning_rate": 1.9157210752345713e-05,
+      "loss": 0.4936,
+      "step": 1858
+    },
+    {
+      "epoch": 0.41467767120231985,
+      "grad_norm": 0.20180164277553558,
+      "learning_rate": 1.915626485672237e-05,
+      "loss": 0.5347,
+      "step": 1859
+    },
+    {
+      "epoch": 0.41490073611420925,
+      "grad_norm": 0.156357541680336,
+      "learning_rate": 1.9155318453967483e-05,
+      "loss": 0.4816,
+      "step": 1860
+    },
+    {
+      "epoch": 0.4151238010260986,
+      "grad_norm": 0.16347983479499817,
+      "learning_rate": 1.9154371544133472e-05,
+      "loss": 0.5032,
+      "step": 1861
+    },
+    {
+      "epoch": 0.415346865937988,
+      "grad_norm": 0.16571176052093506,
+      "learning_rate": 1.9153424127272783e-05,
+      "loss": 0.4875,
+      "step": 1862
+    },
+    {
+      "epoch": 0.4155699308498773,
+      "grad_norm": 0.16103731095790863,
+      "learning_rate": 1.9152476203437884e-05,
+      "loss": 0.5266,
+      "step": 1863
+    },
+    {
+      "epoch": 0.41579299576176665,
+      "grad_norm": 0.17397384345531464,
+      "learning_rate": 1.915152777268128e-05,
+      "loss": 0.5114,
+      "step": 1864
+    },
+    {
+      "epoch": 0.41601606067365604,
+      "grad_norm": 0.16579222679138184,
+      "learning_rate": 1.9150578835055507e-05,
+      "loss": 0.5219,
+      "step": 1865
+    },
+    {
+      "epoch": 0.4162391255855454,
+      "grad_norm": 0.1724756360054016,
+      "learning_rate": 1.914962939061312e-05,
+      "loss": 0.4942,
+      "step": 1866
+    },
+    {
+      "epoch": 0.41646219049743477,
+      "grad_norm": 0.16709351539611816,
+      "learning_rate": 1.9148679439406704e-05,
+      "loss": 0.514,
+      "step": 1867
+    },
+    {
+      "epoch": 0.4166852554093241,
+      "grad_norm": 0.15628811717033386,
+      "learning_rate": 1.914772898148887e-05,
+      "loss": 0.5153,
+      "step": 1868
+    },
+    {
+      "epoch": 0.4169083203212135,
+      "grad_norm": 0.16881687939167023,
+      "learning_rate": 1.914677801691226e-05,
+      "loss": 0.4924,
+      "step": 1869
+    },
+    {
+      "epoch": 0.41713138523310284,
+      "grad_norm": 0.16054266691207886,
+      "learning_rate": 1.9145826545729555e-05,
+      "loss": 0.5238,
+      "step": 1870
+    },
+    {
+      "epoch": 0.4173544501449922,
+      "grad_norm": 0.1636224240064621,
+      "learning_rate": 1.9144874567993446e-05,
+      "loss": 0.4813,
+      "step": 1871
+    },
+    {
+      "epoch": 0.41757751505688157,
+      "grad_norm": 0.16189956665039062,
+      "learning_rate": 1.9143922083756656e-05,
+      "loss": 0.5211,
+      "step": 1872
+    },
+    {
+      "epoch": 0.4178005799687709,
+      "grad_norm": 0.16065020859241486,
+      "learning_rate": 1.9142969093071944e-05,
+      "loss": 0.5116,
+      "step": 1873
+    },
+    {
+      "epoch": 0.4180236448806603,
+      "grad_norm": 0.17001986503601074,
+      "learning_rate": 1.9142015595992096e-05,
+      "loss": 0.5202,
+      "step": 1874
+    },
+    {
+      "epoch": 0.41824670979254963,
+      "grad_norm": 0.16956806182861328,
+      "learning_rate": 1.9141061592569913e-05,
+      "loss": 0.4941,
+      "step": 1875
+    },
+    {
+      "epoch": 0.41846977470443897,
+      "grad_norm": 0.16168902814388275,
+      "learning_rate": 1.9140107082858243e-05,
+      "loss": 0.5134,
+      "step": 1876
+    },
+    {
+      "epoch": 0.41869283961632836,
+      "grad_norm": 0.16588400304317474,
+      "learning_rate": 1.9139152066909948e-05,
+      "loss": 0.4899,
+      "step": 1877
+    },
+    {
+      "epoch": 0.4189159045282177,
+      "grad_norm": 0.16289415955543518,
+      "learning_rate": 1.9138196544777925e-05,
+      "loss": 0.482,
+      "step": 1878
+    },
+    {
+      "epoch": 0.4191389694401071,
+      "grad_norm": 0.17892742156982422,
+      "learning_rate": 1.9137240516515094e-05,
+      "loss": 0.5178,
+      "step": 1879
+    },
+    {
+      "epoch": 0.4193620343519964,
+      "grad_norm": 0.17155860364437103,
+      "learning_rate": 1.913628398217441e-05,
+      "loss": 0.5099,
+      "step": 1880
+    },
+    {
+      "epoch": 0.41958509926388576,
+      "grad_norm": 0.1801706701517105,
+      "learning_rate": 1.913532694180885e-05,
+      "loss": 0.4966,
+      "step": 1881
+    },
+    {
+      "epoch": 0.41980816417577516,
+      "grad_norm": 0.16115552186965942,
+      "learning_rate": 1.9134369395471416e-05,
+      "loss": 0.4839,
+      "step": 1882
+    },
+    {
+      "epoch": 0.4200312290876645,
+      "grad_norm": 0.16582123935222626,
+      "learning_rate": 1.913341134321515e-05,
+      "loss": 0.4719,
+      "step": 1883
+    },
+    {
+      "epoch": 0.4202542939995539,
+      "grad_norm": 0.16009938716888428,
+      "learning_rate": 1.9132452785093113e-05,
+      "loss": 0.5096,
+      "step": 1884
+    },
+    {
+      "epoch": 0.4204773589114432,
+      "grad_norm": 0.17301099002361298,
+      "learning_rate": 1.9131493721158395e-05,
+      "loss": 0.5295,
+      "step": 1885
+    },
+    {
+      "epoch": 0.4207004238233326,
+      "grad_norm": 0.16522350907325745,
+      "learning_rate": 1.9130534151464116e-05,
+      "loss": 0.5386,
+      "step": 1886
+    },
+    {
+      "epoch": 0.42092348873522195,
+      "grad_norm": 0.1624733954668045,
+      "learning_rate": 1.9129574076063423e-05,
+      "loss": 0.5226,
+      "step": 1887
+    },
+    {
+      "epoch": 0.4211465536471113,
+      "grad_norm": 0.16734281182289124,
+      "learning_rate": 1.9128613495009487e-05,
+      "loss": 0.5367,
+      "step": 1888
+    },
+    {
+      "epoch": 0.4213696185590007,
+      "grad_norm": 0.15870822966098785,
+      "learning_rate": 1.912765240835552e-05,
+      "loss": 0.4597,
+      "step": 1889
+    },
+    {
+      "epoch": 0.42159268347089,
+      "grad_norm": 0.16024811565876007,
+      "learning_rate": 1.912669081615474e-05,
+      "loss": 0.4794,
+      "step": 1890
+    },
+    {
+      "epoch": 0.4218157483827794,
+      "grad_norm": 0.16861672699451447,
+      "learning_rate": 1.912572871846042e-05,
+      "loss": 0.4918,
+      "step": 1891
+    },
+    {
+      "epoch": 0.42203881329466875,
+      "grad_norm": 0.16772149503231049,
+      "learning_rate": 1.9124766115325837e-05,
+      "loss": 0.5262,
+      "step": 1892
+    },
+    {
+      "epoch": 0.4222618782065581,
+      "grad_norm": 0.16915294528007507,
+      "learning_rate": 1.912380300680431e-05,
+      "loss": 0.4926,
+      "step": 1893
+    },
+    {
+      "epoch": 0.4224849431184475,
+      "grad_norm": 0.16086746752262115,
+      "learning_rate": 1.912283939294918e-05,
+      "loss": 0.5091,
+      "step": 1894
+    },
+    {
+      "epoch": 0.4227080080303368,
+      "grad_norm": 0.18025851249694824,
+      "learning_rate": 1.912187527381382e-05,
+      "loss": 0.5147,
+      "step": 1895
+    },
+    {
+      "epoch": 0.4229310729422262,
+      "grad_norm": 0.15781170129776,
+      "learning_rate": 1.9120910649451632e-05,
+      "loss": 0.501,
+      "step": 1896
+    },
+    {
+      "epoch": 0.42315413785411554,
+      "grad_norm": 0.15790554881095886,
+      "learning_rate": 1.9119945519916036e-05,
+      "loss": 0.4963,
+      "step": 1897
+    },
+    {
+      "epoch": 0.42337720276600493,
+      "grad_norm": 0.16987471282482147,
+      "learning_rate": 1.9118979885260493e-05,
+      "loss": 0.5148,
+      "step": 1898
+    },
+    {
+      "epoch": 0.42360026767789427,
+      "grad_norm": 0.16438226401805878,
+      "learning_rate": 1.9118013745538483e-05,
+      "loss": 0.5348,
+      "step": 1899
+    },
+    {
+      "epoch": 0.4238233325897836,
+      "grad_norm": 0.18299153447151184,
+      "learning_rate": 1.9117047100803513e-05,
+      "loss": 0.5055,
+      "step": 1900
+    },
+    {
+      "epoch": 0.424046397501673,
+      "grad_norm": 0.18117006123065948,
+      "learning_rate": 1.911607995110913e-05,
+      "loss": 0.517,
+      "step": 1901
+    },
+    {
+      "epoch": 0.42426946241356234,
+      "grad_norm": 0.19496211409568787,
+      "learning_rate": 1.9115112296508896e-05,
+      "loss": 0.5216,
+      "step": 1902
+    },
+    {
+      "epoch": 0.42449252732545173,
+      "grad_norm": 0.17908519506454468,
+      "learning_rate": 1.9114144137056406e-05,
+      "loss": 0.5386,
+      "step": 1903
+    },
+    {
+      "epoch": 0.42471559223734107,
+      "grad_norm": 0.18684661388397217,
+      "learning_rate": 1.9113175472805284e-05,
+      "loss": 0.5341,
+      "step": 1904
+    },
+    {
+      "epoch": 0.4249386571492304,
+      "grad_norm": 0.1662161499261856,
+      "learning_rate": 1.9112206303809183e-05,
+      "loss": 0.4824,
+      "step": 1905
+    },
+    {
+      "epoch": 0.4251617220611198,
+      "grad_norm": 0.1687028408050537,
+      "learning_rate": 1.9111236630121775e-05,
+      "loss": 0.5026,
+      "step": 1906
+    },
+    {
+      "epoch": 0.42538478697300913,
+      "grad_norm": 0.20177949965000153,
+      "learning_rate": 1.9110266451796772e-05,
+      "loss": 0.5112,
+      "step": 1907
+    },
+    {
+      "epoch": 0.4256078518848985,
+      "grad_norm": 0.16266430914402008,
+      "learning_rate": 1.9109295768887907e-05,
+      "loss": 0.5011,
+      "step": 1908
+    },
+    {
+      "epoch": 0.42583091679678786,
+      "grad_norm": 0.1639285683631897,
+      "learning_rate": 1.910832458144894e-05,
+      "loss": 0.5147,
+      "step": 1909
+    },
+    {
+      "epoch": 0.4260539817086772,
+      "grad_norm": 0.16959701478481293,
+      "learning_rate": 1.9107352889533667e-05,
+      "loss": 0.513,
+      "step": 1910
+    },
+    {
+      "epoch": 0.4262770466205666,
+      "grad_norm": 0.16716282069683075,
+      "learning_rate": 1.9106380693195903e-05,
+      "loss": 0.5023,
+      "step": 1911
+    },
+    {
+      "epoch": 0.42650011153245593,
+      "grad_norm": 0.16967961192131042,
+      "learning_rate": 1.9105407992489495e-05,
+      "loss": 0.5185,
+      "step": 1912
+    },
+    {
+      "epoch": 0.4267231764443453,
+      "grad_norm": 0.16723939776420593,
+      "learning_rate": 1.9104434787468316e-05,
+      "loss": 0.5013,
+      "step": 1913
+    },
+    {
+      "epoch": 0.42694624135623466,
+      "grad_norm": 0.16064821183681488,
+      "learning_rate": 1.9103461078186268e-05,
+      "loss": 0.5109,
+      "step": 1914
+    },
+    {
+      "epoch": 0.42716930626812405,
+      "grad_norm": 0.17482173442840576,
+      "learning_rate": 1.9102486864697285e-05,
+      "loss": 0.5148,
+      "step": 1915
+    },
+    {
+      "epoch": 0.4273923711800134,
+      "grad_norm": 0.1638958603143692,
+      "learning_rate": 1.910151214705532e-05,
+      "loss": 0.527,
+      "step": 1916
+    },
+    {
+      "epoch": 0.4276154360919027,
+      "grad_norm": 0.18999075889587402,
+      "learning_rate": 1.9100536925314363e-05,
+      "loss": 0.5,
+      "step": 1917
+    },
+    {
+      "epoch": 0.4278385010037921,
+      "grad_norm": 0.18148307502269745,
+      "learning_rate": 1.9099561199528425e-05,
+      "loss": 0.4862,
+      "step": 1918
+    },
+    {
+      "epoch": 0.42806156591568145,
+      "grad_norm": 0.17788200080394745,
+      "learning_rate": 1.909858496975155e-05,
+      "loss": 0.4991,
+      "step": 1919
+    },
+    {
+      "epoch": 0.42828463082757084,
+      "grad_norm": 0.1629716157913208,
+      "learning_rate": 1.9097608236037813e-05,
+      "loss": 0.5292,
+      "step": 1920
+    },
+    {
+      "epoch": 0.4285076957394602,
+      "grad_norm": 0.1686851680278778,
+      "learning_rate": 1.9096630998441298e-05,
+      "loss": 0.5156,
+      "step": 1921
+    },
+    {
+      "epoch": 0.4287307606513495,
+      "grad_norm": 0.1803978681564331,
+      "learning_rate": 1.909565325701614e-05,
+      "loss": 0.5193,
+      "step": 1922
+    },
+    {
+      "epoch": 0.4289538255632389,
+      "grad_norm": 0.17601296305656433,
+      "learning_rate": 1.9094675011816496e-05,
+      "loss": 0.5193,
+      "step": 1923
+    },
+    {
+      "epoch": 0.42917689047512825,
+      "grad_norm": 0.17354577779769897,
+      "learning_rate": 1.9093696262896535e-05,
+      "loss": 0.5055,
+      "step": 1924
+    },
+    {
+      "epoch": 0.42939995538701764,
+      "grad_norm": 0.1659156084060669,
+      "learning_rate": 1.9092717010310476e-05,
+      "loss": 0.5232,
+      "step": 1925
+    },
+    {
+      "epoch": 0.429623020298907,
+      "grad_norm": 0.1605677753686905,
+      "learning_rate": 1.909173725411255e-05,
+      "loss": 0.4916,
+      "step": 1926
+    },
+    {
+      "epoch": 0.42984608521079637,
+      "grad_norm": 0.17374806106090546,
+      "learning_rate": 1.9090756994357035e-05,
+      "loss": 0.5259,
+      "step": 1927
+    },
+    {
+      "epoch": 0.4300691501226857,
+      "grad_norm": 0.16589613258838654,
+      "learning_rate": 1.9089776231098204e-05,
+      "loss": 0.5192,
+      "step": 1928
+    },
+    {
+      "epoch": 0.43029221503457504,
+      "grad_norm": 0.16348575055599213,
+      "learning_rate": 1.9088794964390395e-05,
+      "loss": 0.4963,
+      "step": 1929
+    },
+    {
+      "epoch": 0.43051527994646444,
+      "grad_norm": 0.16494914889335632,
+      "learning_rate": 1.9087813194287948e-05,
+      "loss": 0.5359,
+      "step": 1930
+    },
+    {
+      "epoch": 0.43073834485835377,
+      "grad_norm": 0.16134381294250488,
+      "learning_rate": 1.9086830920845242e-05,
+      "loss": 0.5087,
+      "step": 1931
+    },
+    {
+      "epoch": 0.43096140977024316,
+      "grad_norm": 0.16159029304981232,
+      "learning_rate": 1.908584814411668e-05,
+      "loss": 0.5242,
+      "step": 1932
+    },
+    {
+      "epoch": 0.4311844746821325,
+      "grad_norm": 0.1648871749639511,
+      "learning_rate": 1.9084864864156696e-05,
+      "loss": 0.5069,
+      "step": 1933
+    },
+    {
+      "epoch": 0.43140753959402184,
+      "grad_norm": 0.17623218894004822,
+      "learning_rate": 1.9083881081019752e-05,
+      "loss": 0.5049,
+      "step": 1934
+    },
+    {
+      "epoch": 0.43163060450591123,
+      "grad_norm": 0.166509211063385,
+      "learning_rate": 1.9082896794760327e-05,
+      "loss": 0.4966,
+      "step": 1935
+    },
+    {
+      "epoch": 0.43185366941780057,
+      "grad_norm": 0.15902282297611237,
+      "learning_rate": 1.908191200543295e-05,
+      "loss": 0.5343,
+      "step": 1936
+    },
+    {
+      "epoch": 0.43207673432968996,
+      "grad_norm": 0.16570866107940674,
+      "learning_rate": 1.908092671309216e-05,
+      "loss": 0.4797,
+      "step": 1937
+    },
+    {
+      "epoch": 0.4322997992415793,
+      "grad_norm": 0.15843339264392853,
+      "learning_rate": 1.9079940917792524e-05,
+      "loss": 0.5359,
+      "step": 1938
+    },
+    {
+      "epoch": 0.43252286415346863,
+      "grad_norm": 0.16115322709083557,
+      "learning_rate": 1.9078954619588645e-05,
+      "loss": 0.5032,
+      "step": 1939
+    },
+    {
+      "epoch": 0.432745929065358,
+      "grad_norm": 0.16374099254608154,
+      "learning_rate": 1.9077967818535153e-05,
+      "loss": 0.4998,
+      "step": 1940
+    },
+    {
+      "epoch": 0.43296899397724736,
+      "grad_norm": 0.17378082871437073,
+      "learning_rate": 1.9076980514686695e-05,
+      "loss": 0.5403,
+      "step": 1941
+    },
+    {
+      "epoch": 0.43319205888913676,
+      "grad_norm": 0.16688857972621918,
+      "learning_rate": 1.9075992708097965e-05,
+      "loss": 0.5103,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4334151238010261,
+      "grad_norm": 0.16233094036579132,
+      "learning_rate": 1.9075004398823665e-05,
+      "loss": 0.5045,
+      "step": 1943
+    },
+    {
+      "epoch": 0.4336381887129155,
+      "grad_norm": 0.16693229973316193,
+      "learning_rate": 1.907401558691854e-05,
+      "loss": 0.5167,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4338612536248048,
+      "grad_norm": 0.16341574490070343,
+      "learning_rate": 1.9073026272437353e-05,
+      "loss": 0.5001,
+      "step": 1945
+    },
+    {
+      "epoch": 0.43408431853669416,
+      "grad_norm": 0.16043853759765625,
+      "learning_rate": 1.90720364554349e-05,
+      "loss": 0.5109,
+      "step": 1946
+    },
+    {
+      "epoch": 0.43430738344858355,
+      "grad_norm": 0.15996608138084412,
+      "learning_rate": 1.9071046135966e-05,
+      "loss": 0.4826,
+      "step": 1947
+    },
+    {
+      "epoch": 0.4345304483604729,
+      "grad_norm": 0.15902268886566162,
+      "learning_rate": 1.9070055314085508e-05,
+      "loss": 0.4989,
+      "step": 1948
+    },
+    {
+      "epoch": 0.4347535132723623,
+      "grad_norm": 0.16526652872562408,
+      "learning_rate": 1.9069063989848298e-05,
+      "loss": 0.5209,
+      "step": 1949
+    },
+    {
+      "epoch": 0.4349765781842516,
+      "grad_norm": 0.16759642958641052,
+      "learning_rate": 1.9068072163309282e-05,
+      "loss": 0.5246,
+      "step": 1950
+    },
+    {
+      "epoch": 0.43519964309614095,
+      "grad_norm": 0.17135794460773468,
+      "learning_rate": 1.9067079834523387e-05,
+      "loss": 0.5198,
+      "step": 1951
+    },
+    {
+      "epoch": 0.43542270800803035,
+      "grad_norm": 0.16336077451705933,
+      "learning_rate": 1.9066087003545576e-05,
+      "loss": 0.5098,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4356457729199197,
+      "grad_norm": 0.1512858122587204,
+      "learning_rate": 1.9065093670430836e-05,
+      "loss": 0.4843,
+      "step": 1953
+    },
+    {
+      "epoch": 0.4358688378318091,
+      "grad_norm": 0.17053045332431793,
+      "learning_rate": 1.9064099835234188e-05,
+      "loss": 0.527,
+      "step": 1954
+    },
+    {
+      "epoch": 0.4360919027436984,
+      "grad_norm": 0.16126702725887299,
+      "learning_rate": 1.9063105498010678e-05,
+      "loss": 0.5222,
+      "step": 1955
+    },
+    {
+      "epoch": 0.43631496765558775,
+      "grad_norm": 0.1633397489786148,
+      "learning_rate": 1.9062110658815375e-05,
+      "loss": 0.5124,
+      "step": 1956
+    },
+    {
+      "epoch": 0.43653803256747714,
+      "grad_norm": 0.18700194358825684,
+      "learning_rate": 1.9061115317703384e-05,
+      "loss": 0.5123,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4367610974793665,
+      "grad_norm": 0.1907283067703247,
+      "learning_rate": 1.9060119474729826e-05,
+      "loss": 0.5223,
+      "step": 1958
+    },
+    {
+      "epoch": 0.43698416239125587,
+      "grad_norm": 0.1634737253189087,
+      "learning_rate": 1.9059123129949865e-05,
+      "loss": 0.4879,
+      "step": 1959
+    },
+    {
+      "epoch": 0.4372072273031452,
+      "grad_norm": 0.18314914405345917,
+      "learning_rate": 1.9058126283418675e-05,
+      "loss": 0.5018,
+      "step": 1960
+    },
+    {
+      "epoch": 0.4374302922150346,
+      "grad_norm": 0.1680321991443634,
+      "learning_rate": 1.9057128935191477e-05,
+      "loss": 0.4975,
+      "step": 1961
+    },
+    {
+      "epoch": 0.43765335712692394,
+      "grad_norm": 0.1635400801897049,
+      "learning_rate": 1.9056131085323506e-05,
+      "loss": 0.4895,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4378764220388133,
+      "grad_norm": 0.15011341869831085,
+      "learning_rate": 1.905513273387003e-05,
+      "loss": 0.4765,
+      "step": 1963
+    },
+    {
+      "epoch": 0.43809948695070267,
+      "grad_norm": 0.16991405189037323,
+      "learning_rate": 1.9054133880886348e-05,
+      "loss": 0.518,
+      "step": 1964
+    },
+    {
+      "epoch": 0.438322551862592,
+      "grad_norm": 0.17523062229156494,
+      "learning_rate": 1.9053134526427777e-05,
+      "loss": 0.4764,
+      "step": 1965
+    },
+    {
+      "epoch": 0.4385456167744814,
+      "grad_norm": 0.1695917248725891,
+      "learning_rate": 1.905213467054967e-05,
+      "loss": 0.5111,
+      "step": 1966
+    },
+    {
+      "epoch": 0.43876868168637073,
+      "grad_norm": 0.16954153776168823,
+      "learning_rate": 1.90511343133074e-05,
+      "loss": 0.505,
+      "step": 1967
+    },
+    {
+      "epoch": 0.43899174659826007,
+      "grad_norm": 0.3032020628452301,
+      "learning_rate": 1.905013345475638e-05,
+      "loss": 0.5037,
+      "step": 1968
+    },
+    {
+      "epoch": 0.43921481151014946,
+      "grad_norm": 0.16867168247699738,
+      "learning_rate": 1.9049132094952046e-05,
+      "loss": 0.5596,
+      "step": 1969
+    },
+    {
+      "epoch": 0.4394378764220388,
+      "grad_norm": 0.16780449450016022,
+      "learning_rate": 1.904813023394985e-05,
+      "loss": 0.4937,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4396609413339282,
+      "grad_norm": 0.19413304328918457,
+      "learning_rate": 1.904712787180529e-05,
+      "loss": 0.4823,
+      "step": 1971
+    },
+    {
+      "epoch": 0.4398840062458175,
+      "grad_norm": 0.16119445860385895,
+      "learning_rate": 1.9046125008573876e-05,
+      "loss": 0.5451,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4401070711577069,
+      "grad_norm": 0.1607835441827774,
+      "learning_rate": 1.904512164431116e-05,
+      "loss": 0.5066,
+      "step": 1973
+    },
+    {
+      "epoch": 0.44033013606959626,
+      "grad_norm": 0.1697167158126831,
+      "learning_rate": 1.9044117779072708e-05,
+      "loss": 0.5166,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4405532009814856,
+      "grad_norm": 0.16189232468605042,
+      "learning_rate": 1.9043113412914128e-05,
+      "loss": 0.4968,
+      "step": 1975
+    },
+    {
+      "epoch": 0.440776265893375,
+      "grad_norm": 0.1717143952846527,
+      "learning_rate": 1.904210854589104e-05,
+      "loss": 0.481,
+      "step": 1976
+    },
+    {
+      "epoch": 0.4409993308052643,
+      "grad_norm": 0.17401723563671112,
+      "learning_rate": 1.9041103178059107e-05,
+      "loss": 0.526,
+      "step": 1977
+    },
+    {
+      "epoch": 0.4412223957171537,
+      "grad_norm": 0.16598494350910187,
+      "learning_rate": 1.9040097309474007e-05,
+      "loss": 0.5195,
+      "step": 1978
+    },
+    {
+      "epoch": 0.44144546062904305,
+      "grad_norm": 0.21129997074604034,
+      "learning_rate": 1.9039090940191455e-05,
+      "loss": 0.5096,
+      "step": 1979
+    },
+    {
+      "epoch": 0.4416685255409324,
+      "grad_norm": 0.17546051740646362,
+      "learning_rate": 1.9038084070267186e-05,
+      "loss": 0.5313,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4418915904528218,
+      "grad_norm": 0.15967412292957306,
+      "learning_rate": 1.9037076699756973e-05,
+      "loss": 0.5183,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4421146553647111,
+      "grad_norm": 0.17214882373809814,
+      "learning_rate": 1.9036068828716603e-05,
+      "loss": 0.5091,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4423377202766005,
+      "grad_norm": 0.184258371591568,
+      "learning_rate": 1.9035060457201904e-05,
+      "loss": 0.4785,
+      "step": 1983
+    },
+    {
+      "epoch": 0.44256078518848985,
+      "grad_norm": 0.1570109874010086,
+      "learning_rate": 1.9034051585268725e-05,
+      "loss": 0.515,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4427838501003792,
+      "grad_norm": 0.17051473259925842,
+      "learning_rate": 1.903304221297294e-05,
+      "loss": 0.5031,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4430069150122686,
+      "grad_norm": 0.19998317956924438,
+      "learning_rate": 1.903203234037046e-05,
+      "loss": 0.499,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4432299799241579,
+      "grad_norm": 0.16017475724220276,
+      "learning_rate": 1.9031021967517213e-05,
+      "loss": 0.5237,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4434530448360473,
+      "grad_norm": 0.18142235279083252,
+      "learning_rate": 1.9030011094469164e-05,
+      "loss": 0.5373,
+      "step": 1988
+    },
+    {
+      "epoch": 0.44367610974793664,
+      "grad_norm": 0.16957718133926392,
+      "learning_rate": 1.90289997212823e-05,
+      "loss": 0.5038,
+      "step": 1989
+    },
+    {
+      "epoch": 0.44389917465982603,
+      "grad_norm": 0.16631929576396942,
+      "learning_rate": 1.9027987848012635e-05,
+      "loss": 0.4873,
+      "step": 1990
+    },
+    {
+      "epoch": 0.44412223957171537,
+      "grad_norm": 0.16362012922763824,
+      "learning_rate": 1.9026975474716215e-05,
+      "loss": 0.5204,
+      "step": 1991
+    },
+    {
+      "epoch": 0.4443453044836047,
+      "grad_norm": 0.39765313267707825,
+      "learning_rate": 1.902596260144911e-05,
+      "loss": 0.4736,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4445683693954941,
+      "grad_norm": 0.17083793878555298,
+      "learning_rate": 1.9024949228267423e-05,
+      "loss": 0.5152,
+      "step": 1993
+    },
+    {
+      "epoch": 0.44479143430738344,
+      "grad_norm": 0.1833147555589676,
+      "learning_rate": 1.902393535522728e-05,
+      "loss": 0.536,
+      "step": 1994
+    },
+    {
+      "epoch": 0.44501449921927283,
+      "grad_norm": 0.17427003383636475,
+      "learning_rate": 1.902292098238483e-05,
+      "loss": 0.514,
+      "step": 1995
+    },
+    {
+      "epoch": 0.44523756413116217,
+      "grad_norm": 0.15868180990219116,
+      "learning_rate": 1.902190610979626e-05,
+      "loss": 0.4869,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4454606290430515,
+      "grad_norm": 0.17407076060771942,
+      "learning_rate": 1.9020890737517783e-05,
+      "loss": 0.5134,
+      "step": 1997
+    },
+    {
+      "epoch": 0.4456836939549409,
+      "grad_norm": 0.16776099801063538,
+      "learning_rate": 1.901987486560563e-05,
+      "loss": 0.5079,
+      "step": 1998
+    },
+    {
+      "epoch": 0.44590675886683023,
+      "grad_norm": 0.2472338080406189,
+      "learning_rate": 1.9018858494116074e-05,
+      "loss": 0.5128,
+      "step": 1999
+    },
+    {
+      "epoch": 0.4461298237787196,
+      "grad_norm": 0.16303667426109314,
+      "learning_rate": 1.90178416231054e-05,
+      "loss": 0.5225,
+      "step": 2000
+    },
+    {
+      "epoch": 0.44635288869060896,
+      "grad_norm": 0.17127907276153564,
+      "learning_rate": 1.901682425262993e-05,
+      "loss": 0.5092,
+      "step": 2001
+    },
+    {
+      "epoch": 0.44657595360249835,
+      "grad_norm": 0.17813526093959808,
+      "learning_rate": 1.9015806382746018e-05,
+      "loss": 0.5,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4467990185143877,
+      "grad_norm": 0.1653776615858078,
+      "learning_rate": 1.901478801351004e-05,
+      "loss": 0.5051,
+      "step": 2003
+    },
+    {
+      "epoch": 0.44702208342627703,
+      "grad_norm": 0.16295598447322845,
+      "learning_rate": 1.9013769144978392e-05,
+      "loss": 0.4955,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4472451483381664,
+      "grad_norm": 0.18259268999099731,
+      "learning_rate": 1.901274977720751e-05,
+      "loss": 0.5504,
+      "step": 2005
+    },
+    {
+      "epoch": 0.44746821325005576,
+      "grad_norm": 0.19138063490390778,
+      "learning_rate": 1.9011729910253856e-05,
+      "loss": 0.5129,
+      "step": 2006
+    },
+    {
+      "epoch": 0.44769127816194515,
+      "grad_norm": 0.1588236689567566,
+      "learning_rate": 1.9010709544173913e-05,
+      "loss": 0.5078,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4479143430738345,
+      "grad_norm": 0.162478968501091,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.5121,
+      "step": 2008
+    },
+    {
+      "epoch": 0.4481374079857238,
+      "grad_norm": 0.17752927541732788,
+      "learning_rate": 1.900866731486124e-05,
+      "loss": 0.4983,
+      "step": 2009
+    },
+    {
+      "epoch": 0.4483604728976132,
+      "grad_norm": 0.17127980291843414,
+      "learning_rate": 1.900764545174163e-05,
+      "loss": 0.5347,
+      "step": 2010
+    },
+    {
+      "epoch": 0.44858353780950255,
+      "grad_norm": 0.1686343103647232,
+      "learning_rate": 1.900662308972195e-05,
+      "loss": 0.5107,
+      "step": 2011
+    },
+    {
+      "epoch": 0.44880660272139195,
+      "grad_norm": 0.16008520126342773,
+      "learning_rate": 1.9005600228858832e-05,
+      "loss": 0.5036,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4490296676332813,
+      "grad_norm": 0.16494229435920715,
+      "learning_rate": 1.9004576869208922e-05,
+      "loss": 0.5135,
+      "step": 2013
+    },
+    {
+      "epoch": 0.4492527325451706,
+      "grad_norm": 0.17318038642406464,
+      "learning_rate": 1.9003553010828906e-05,
+      "loss": 0.5279,
+      "step": 2014
+    },
+    {
+      "epoch": 0.44947579745706,
+      "grad_norm": 0.252083420753479,
+      "learning_rate": 1.9002528653775492e-05,
+      "loss": 0.5176,
+      "step": 2015
+    },
+    {
+      "epoch": 0.44969886236894935,
+      "grad_norm": 0.1737111508846283,
+      "learning_rate": 1.900150379810541e-05,
+      "loss": 0.5051,
+      "step": 2016
+    },
+    {
+      "epoch": 0.44992192728083874,
+      "grad_norm": 0.172331303358078,
+      "learning_rate": 1.9000478443875427e-05,
+      "loss": 0.5161,
+      "step": 2017
+    },
+    {
+      "epoch": 0.4501449921927281,
+      "grad_norm": 0.17037436366081238,
+      "learning_rate": 1.899945259114233e-05,
+      "loss": 0.5282,
+      "step": 2018
+    },
+    {
+      "epoch": 0.45036805710461747,
+      "grad_norm": 0.16774620115756989,
+      "learning_rate": 1.8998426239962945e-05,
+      "loss": 0.5056,
+      "step": 2019
+    },
+    {
+      "epoch": 0.4505911220165068,
+      "grad_norm": 0.17856718599796295,
+      "learning_rate": 1.899739939039411e-05,
+      "loss": 0.514,
+      "step": 2020
+    },
+    {
+      "epoch": 0.45081418692839614,
+      "grad_norm": 0.18140466511249542,
+      "learning_rate": 1.89963720424927e-05,
+      "loss": 0.5372,
+      "step": 2021
+    },
+    {
+      "epoch": 0.45103725184028554,
+      "grad_norm": 0.1641564816236496,
+      "learning_rate": 1.8995344196315618e-05,
+      "loss": 0.5205,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4512603167521749,
+      "grad_norm": 0.16866321861743927,
+      "learning_rate": 1.899431585191979e-05,
+      "loss": 0.4906,
+      "step": 2023
+    },
+    {
+      "epoch": 0.45148338166406426,
+      "grad_norm": 0.16418704390525818,
+      "learning_rate": 1.8993287009362175e-05,
+      "loss": 0.5275,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4517064465759536,
+      "grad_norm": 0.15996740758419037,
+      "learning_rate": 1.8992257668699756e-05,
+      "loss": 0.5124,
+      "step": 2025
+    },
+    {
+      "epoch": 0.45192951148784294,
+      "grad_norm": 0.16520540416240692,
+      "learning_rate": 1.899122782998954e-05,
+      "loss": 0.5045,
+      "step": 2026
+    },
+    {
+      "epoch": 0.45215257639973233,
+      "grad_norm": 0.16536462306976318,
+      "learning_rate": 1.8990197493288575e-05,
+      "loss": 0.5003,
+      "step": 2027
+    },
+    {
+      "epoch": 0.45237564131162167,
+      "grad_norm": 0.16777758300304413,
+      "learning_rate": 1.8989166658653916e-05,
+      "loss": 0.5443,
+      "step": 2028
+    },
+    {
+      "epoch": 0.45259870622351106,
+      "grad_norm": 0.16215604543685913,
+      "learning_rate": 1.8988135326142668e-05,
+      "loss": 0.4965,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4528217711354004,
+      "grad_norm": 0.1622023582458496,
+      "learning_rate": 1.8987103495811947e-05,
+      "loss": 0.4606,
+      "step": 2030
+    },
+    {
+      "epoch": 0.45304483604728973,
+      "grad_norm": 0.16461379826068878,
+      "learning_rate": 1.8986071167718902e-05,
+      "loss": 0.5273,
+      "step": 2031
+    },
+    {
+      "epoch": 0.4532679009591791,
+      "grad_norm": 0.16976390779018402,
+      "learning_rate": 1.8985038341920715e-05,
+      "loss": 0.4889,
+      "step": 2032
+    },
+    {
+      "epoch": 0.45349096587106846,
+      "grad_norm": 0.1569637656211853,
+      "learning_rate": 1.898400501847458e-05,
+      "loss": 0.4828,
+      "step": 2033
+    },
+    {
+      "epoch": 0.45371403078295786,
+      "grad_norm": 0.16640910506248474,
+      "learning_rate": 1.898297119743774e-05,
+      "loss": 0.5351,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4539370956948472,
+      "grad_norm": 0.1644599735736847,
+      "learning_rate": 1.898193687886745e-05,
+      "loss": 0.4869,
+      "step": 2035
+    },
+    {
+      "epoch": 0.4541601606067366,
+      "grad_norm": 0.16732025146484375,
+      "learning_rate": 1.8980902062820997e-05,
+      "loss": 0.4893,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4543832255186259,
+      "grad_norm": 0.16019105911254883,
+      "learning_rate": 1.8979866749355694e-05,
+      "loss": 0.5038,
+      "step": 2037
+    },
+    {
+      "epoch": 0.45460629043051526,
+      "grad_norm": 0.17239058017730713,
+      "learning_rate": 1.8978830938528884e-05,
+      "loss": 0.5106,
+      "step": 2038
+    },
+    {
+      "epoch": 0.45482935534240465,
+      "grad_norm": 0.15983308851718903,
+      "learning_rate": 1.8977794630397942e-05,
+      "loss": 0.5163,
+      "step": 2039
+    },
+    {
+      "epoch": 0.455052420254294,
+      "grad_norm": 0.16848912835121155,
+      "learning_rate": 1.8976757825020255e-05,
+      "loss": 0.5462,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4552754851661834,
+      "grad_norm": 0.15303179621696472,
+      "learning_rate": 1.897572052245326e-05,
+      "loss": 0.4594,
+      "step": 2041
+    },
+    {
+      "epoch": 0.4554985500780727,
+      "grad_norm": 0.17675776779651642,
+      "learning_rate": 1.8974682722754397e-05,
+      "loss": 0.5385,
+      "step": 2042
+    },
+    {
+      "epoch": 0.45572161498996205,
+      "grad_norm": 0.17349328100681305,
+      "learning_rate": 1.8973644425981154e-05,
+      "loss": 0.4953,
+      "step": 2043
+    },
+    {
+      "epoch": 0.45594467990185145,
+      "grad_norm": 0.15753738582134247,
+      "learning_rate": 1.897260563219104e-05,
+      "loss": 0.4851,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4561677448137408,
+      "grad_norm": 0.16278420388698578,
+      "learning_rate": 1.897156634144158e-05,
+      "loss": 0.4994,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4563908097256302,
+      "grad_norm": 0.1592012643814087,
+      "learning_rate": 1.8970526553790346e-05,
+      "loss": 0.513,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4566138746375195,
+      "grad_norm": 0.18133483827114105,
+      "learning_rate": 1.8969486269294922e-05,
+      "loss": 0.4734,
+      "step": 2047
+    },
+    {
+      "epoch": 0.4568369395494089,
+      "grad_norm": 0.170905202627182,
+      "learning_rate": 1.8968445488012933e-05,
+      "loss": 0.5401,
+      "step": 2048
+    },
+    {
+      "epoch": 0.45706000446129824,
+      "grad_norm": 0.16522808372974396,
+      "learning_rate": 1.8967404210002014e-05,
+      "loss": 0.5008,
+      "step": 2049
+    },
+    {
+      "epoch": 0.4572830693731876,
+      "grad_norm": 0.18945086002349854,
+      "learning_rate": 1.8966362435319845e-05,
+      "loss": 0.5064,
+      "step": 2050
+    },
+    {
+      "epoch": 0.45750613428507697,
+      "grad_norm": 0.17545610666275024,
+      "learning_rate": 1.8965320164024123e-05,
+      "loss": 0.5105,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4577291991969663,
+      "grad_norm": 0.1666799783706665,
+      "learning_rate": 1.8964277396172577e-05,
+      "loss": 0.5233,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4579522641088557,
+      "grad_norm": 0.16869474947452545,
+      "learning_rate": 1.896323413182296e-05,
+      "loss": 0.4921,
+      "step": 2053
+    },
+    {
+      "epoch": 0.45817532902074504,
+      "grad_norm": 0.17108272016048431,
+      "learning_rate": 1.8962190371033057e-05,
+      "loss": 0.5358,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4583983939326344,
+      "grad_norm": 0.16527149081230164,
+      "learning_rate": 1.8961146113860676e-05,
+      "loss": 0.5033,
+      "step": 2055
+    },
+    {
+      "epoch": 0.45862145884452377,
+      "grad_norm": 0.17335473001003265,
+      "learning_rate": 1.8960101360363656e-05,
+      "loss": 0.5156,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4588445237564131,
+      "grad_norm": 0.17064844071865082,
+      "learning_rate": 1.895905611059986e-05,
+      "loss": 0.5118,
+      "step": 2057
+    },
+    {
+      "epoch": 0.4590675886683025,
+      "grad_norm": 0.16565892100334167,
+      "learning_rate": 1.8958010364627183e-05,
+      "loss": 0.5152,
+      "step": 2058
+    },
+    {
+      "epoch": 0.45929065358019183,
+      "grad_norm": 0.1769896149635315,
+      "learning_rate": 1.8956964122503546e-05,
+      "loss": 0.5483,
+      "step": 2059
+    },
+    {
+      "epoch": 0.45951371849208117,
+      "grad_norm": 0.1574466973543167,
+      "learning_rate": 1.895591738428689e-05,
+      "loss": 0.5205,
+      "step": 2060
+    },
+    {
+      "epoch": 0.45973678340397056,
+      "grad_norm": 0.16593651473522186,
+      "learning_rate": 1.8954870150035195e-05,
+      "loss": 0.4863,
+      "step": 2061
+    },
+    {
+      "epoch": 0.4599598483158599,
+      "grad_norm": 0.16003793478012085,
+      "learning_rate": 1.8953822419806468e-05,
+      "loss": 0.5051,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4601829132277493,
+      "grad_norm": 0.17091605067253113,
+      "learning_rate": 1.895277419365873e-05,
+      "loss": 0.5037,
+      "step": 2063
+    },
+    {
+      "epoch": 0.4604059781396386,
+      "grad_norm": 0.17469583451747894,
+      "learning_rate": 1.895172547165004e-05,
+      "loss": 0.4878,
+      "step": 2064
+    },
+    {
+      "epoch": 0.460629043051528,
+      "grad_norm": 0.16121211647987366,
+      "learning_rate": 1.895067625383849e-05,
+      "loss": 0.4768,
+      "step": 2065
+    },
+    {
+      "epoch": 0.46085210796341736,
+      "grad_norm": 0.1591392159461975,
+      "learning_rate": 1.894962654028218e-05,
+      "loss": 0.5138,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4610751728753067,
+      "grad_norm": 1.2021454572677612,
+      "learning_rate": 1.8948576331039264e-05,
+      "loss": 0.5438,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4612982377871961,
+      "grad_norm": 0.1606753021478653,
+      "learning_rate": 1.8947525626167896e-05,
+      "loss": 0.5376,
+      "step": 2068
+    },
+    {
+      "epoch": 0.4615213026990854,
+      "grad_norm": 0.17104879021644592,
+      "learning_rate": 1.894647442572628e-05,
+      "loss": 0.5164,
+      "step": 2069
+    },
+    {
+      "epoch": 0.4617443676109748,
+      "grad_norm": 0.16374671459197998,
+      "learning_rate": 1.8945422729772633e-05,
+      "loss": 0.516,
+      "step": 2070
+    },
+    {
+      "epoch": 0.46196743252286415,
+      "grad_norm": 0.16128143668174744,
+      "learning_rate": 1.8944370538365206e-05,
+      "loss": 0.5157,
+      "step": 2071
+    },
+    {
+      "epoch": 0.4621904974347535,
+      "grad_norm": 0.16720320284366608,
+      "learning_rate": 1.8943317851562278e-05,
+      "loss": 0.5291,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4624135623466429,
+      "grad_norm": 0.1879906952381134,
+      "learning_rate": 1.8942264669422154e-05,
+      "loss": 0.5145,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4626366272585322,
+      "grad_norm": 0.17528733611106873,
+      "learning_rate": 1.894121099200316e-05,
+      "loss": 0.5235,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4628596921704216,
+      "grad_norm": 0.16449770331382751,
+      "learning_rate": 1.894015681936366e-05,
+      "loss": 0.5165,
+      "step": 2075
+    },
+    {
+      "epoch": 0.46308275708231095,
+      "grad_norm": 0.16646511852741241,
+      "learning_rate": 1.8939102151562036e-05,
+      "loss": 0.5093,
+      "step": 2076
+    },
+    {
+      "epoch": 0.46330582199420034,
+      "grad_norm": 0.17490622401237488,
+      "learning_rate": 1.893804698865671e-05,
+      "loss": 0.5385,
+      "step": 2077
+    },
+    {
+      "epoch": 0.4635288869060897,
+      "grad_norm": 0.2023608237504959,
+      "learning_rate": 1.893699133070612e-05,
+      "loss": 0.4972,
+      "step": 2078
+    },
+    {
+      "epoch": 0.463751951817979,
+      "grad_norm": 0.16244065761566162,
+      "learning_rate": 1.893593517776873e-05,
+      "loss": 0.5269,
+      "step": 2079
+    },
+    {
+      "epoch": 0.4639750167298684,
+      "grad_norm": 0.16335059702396393,
+      "learning_rate": 1.8934878529903043e-05,
+      "loss": 0.5107,
+      "step": 2080
+    },
+    {
+      "epoch": 0.46419808164175774,
+      "grad_norm": 0.16115811467170715,
+      "learning_rate": 1.8933821387167582e-05,
+      "loss": 0.5214,
+      "step": 2081
+    },
+    {
+      "epoch": 0.46442114655364714,
+      "grad_norm": 0.1833171248435974,
+      "learning_rate": 1.8932763749620894e-05,
+      "loss": 0.4664,
+      "step": 2082
+    },
+    {
+      "epoch": 0.46464421146553647,
+      "grad_norm": 0.1721174716949463,
+      "learning_rate": 1.893170561732156e-05,
+      "loss": 0.5224,
+      "step": 2083
+    },
+    {
+      "epoch": 0.4648672763774258,
+      "grad_norm": 0.16398632526397705,
+      "learning_rate": 1.8930646990328188e-05,
+      "loss": 0.4968,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4650903412893152,
+      "grad_norm": 0.17112642526626587,
+      "learning_rate": 1.892958786869941e-05,
+      "loss": 0.5241,
+      "step": 2085
+    },
+    {
+      "epoch": 0.46531340620120454,
+      "grad_norm": 0.18161650002002716,
+      "learning_rate": 1.8928528252493884e-05,
+      "loss": 0.5047,
+      "step": 2086
+    },
+    {
+      "epoch": 0.46553647111309393,
+      "grad_norm": 0.1646048128604889,
+      "learning_rate": 1.8927468141770304e-05,
+      "loss": 0.5271,
+      "step": 2087
+    },
+    {
+      "epoch": 0.46575953602498327,
+      "grad_norm": 0.16542619466781616,
+      "learning_rate": 1.8926407536587378e-05,
+      "loss": 0.5338,
+      "step": 2088
+    },
+    {
+      "epoch": 0.4659826009368726,
+      "grad_norm": 0.17052631080150604,
+      "learning_rate": 1.8925346437003856e-05,
+      "loss": 0.5276,
+      "step": 2089
+    },
+    {
+      "epoch": 0.466205665848762,
+      "grad_norm": 0.17085763812065125,
+      "learning_rate": 1.8924284843078503e-05,
+      "loss": 0.5277,
+      "step": 2090
+    },
+    {
+      "epoch": 0.46642873076065133,
+      "grad_norm": 0.1689939796924591,
+      "learning_rate": 1.8923222754870124e-05,
+      "loss": 0.4889,
+      "step": 2091
+    },
+    {
+      "epoch": 0.4666517956725407,
+      "grad_norm": 0.16189821064472198,
+      "learning_rate": 1.8922160172437535e-05,
+      "loss": 0.5009,
+      "step": 2092
+    },
+    {
+      "epoch": 0.46687486058443006,
+      "grad_norm": 0.16525843739509583,
+      "learning_rate": 1.89210970958396e-05,
+      "loss": 0.5073,
+      "step": 2093
+    },
+    {
+      "epoch": 0.46709792549631945,
+      "grad_norm": 0.159906804561615,
+      "learning_rate": 1.8920033525135184e-05,
+      "loss": 0.5107,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4673209904082088,
+      "grad_norm": 0.15631826221942902,
+      "learning_rate": 1.8918969460383205e-05,
+      "loss": 0.4839,
+      "step": 2095
+    },
+    {
+      "epoch": 0.46754405532009813,
+      "grad_norm": 0.1734917014837265,
+      "learning_rate": 1.8917904901642593e-05,
+      "loss": 0.5121,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4677671202319875,
+      "grad_norm": 0.17475415766239166,
+      "learning_rate": 1.8916839848972315e-05,
+      "loss": 0.5198,
+      "step": 2097
+    },
+    {
+      "epoch": 0.46799018514387686,
+      "grad_norm": 0.15618795156478882,
+      "learning_rate": 1.8915774302431357e-05,
+      "loss": 0.4733,
+      "step": 2098
+    },
+    {
+      "epoch": 0.46821325005576625,
+      "grad_norm": 0.17231950163841248,
+      "learning_rate": 1.8914708262078735e-05,
+      "loss": 0.4975,
+      "step": 2099
+    },
+    {
+      "epoch": 0.4684363149676556,
+      "grad_norm": 0.1630786657333374,
+      "learning_rate": 1.891364172797349e-05,
+      "loss": 0.4906,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4686593798795449,
+      "grad_norm": 0.20263253152370453,
+      "learning_rate": 1.89125747001747e-05,
+      "loss": 0.4961,
+      "step": 2101
+    },
+    {
+      "epoch": 0.4688824447914343,
+      "grad_norm": 0.1711629182100296,
+      "learning_rate": 1.891150717874146e-05,
+      "loss": 0.5197,
+      "step": 2102
+    },
+    {
+      "epoch": 0.46910550970332365,
+      "grad_norm": 0.15791350603103638,
+      "learning_rate": 1.89104391637329e-05,
+      "loss": 0.5151,
+      "step": 2103
+    },
+    {
+      "epoch": 0.46932857461521305,
+      "grad_norm": 0.1648074984550476,
+      "learning_rate": 1.890937065520817e-05,
+      "loss": 0.5206,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4695516395271024,
+      "grad_norm": 0.16917914152145386,
+      "learning_rate": 1.8908301653226448e-05,
+      "loss": 0.5308,
+      "step": 2105
+    },
+    {
+      "epoch": 0.4697747044389917,
+      "grad_norm": 0.16751410067081451,
+      "learning_rate": 1.8907232157846946e-05,
+      "loss": 0.512,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4699977693508811,
+      "grad_norm": 0.16895310580730438,
+      "learning_rate": 1.89061621691289e-05,
+      "loss": 0.5134,
+      "step": 2107
+    },
+    {
+      "epoch": 0.47022083426277045,
+      "grad_norm": 0.1659642457962036,
+      "learning_rate": 1.8905091687131567e-05,
+      "loss": 0.5125,
+      "step": 2108
+    },
+    {
+      "epoch": 0.47044389917465984,
+      "grad_norm": 0.18306776881217957,
+      "learning_rate": 1.8904020711914243e-05,
+      "loss": 0.5404,
+      "step": 2109
+    },
+    {
+      "epoch": 0.4706669640865492,
+      "grad_norm": 0.15722155570983887,
+      "learning_rate": 1.8902949243536245e-05,
+      "loss": 0.4929,
+      "step": 2110
+    },
+    {
+      "epoch": 0.47089002899843857,
+      "grad_norm": 0.17024749517440796,
+      "learning_rate": 1.8901877282056916e-05,
+      "loss": 0.5129,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4711130939103279,
+      "grad_norm": 0.1668287068605423,
+      "learning_rate": 1.8900804827535626e-05,
+      "loss": 0.5386,
+      "step": 2112
+    },
+    {
+      "epoch": 0.47133615882221724,
+      "grad_norm": 0.17657139897346497,
+      "learning_rate": 1.8899731880031778e-05,
+      "loss": 0.4682,
+      "step": 2113
+    },
+    {
+      "epoch": 0.47155922373410664,
+      "grad_norm": 0.17689798772335052,
+      "learning_rate": 1.8898658439604798e-05,
+      "loss": 0.4814,
+      "step": 2114
+    },
+    {
+      "epoch": 0.471782288645996,
+      "grad_norm": 0.16755974292755127,
+      "learning_rate": 1.8897584506314137e-05,
+      "loss": 0.4956,
+      "step": 2115
+    },
+    {
+      "epoch": 0.47200535355788537,
+      "grad_norm": 0.16352280974388123,
+      "learning_rate": 1.8896510080219277e-05,
+      "loss": 0.5034,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4722284184697747,
+      "grad_norm": 0.16078925132751465,
+      "learning_rate": 1.889543516137973e-05,
+      "loss": 0.5017,
+      "step": 2117
+    },
+    {
+      "epoch": 0.47245148338166404,
+      "grad_norm": 0.17038793861865997,
+      "learning_rate": 1.8894359749855027e-05,
+      "loss": 0.5128,
+      "step": 2118
+    },
+    {
+      "epoch": 0.47267454829355343,
+      "grad_norm": 0.16477574408054352,
+      "learning_rate": 1.8893283845704733e-05,
+      "loss": 0.4765,
+      "step": 2119
+    },
+    {
+      "epoch": 0.47289761320544277,
+      "grad_norm": 0.18243131041526794,
+      "learning_rate": 1.889220744898844e-05,
+      "loss": 0.5266,
+      "step": 2120
+    },
+    {
+      "epoch": 0.47312067811733216,
+      "grad_norm": 0.15715399384498596,
+      "learning_rate": 1.8891130559765763e-05,
+      "loss": 0.5092,
+      "step": 2121
+    },
+    {
+      "epoch": 0.4733437430292215,
+      "grad_norm": 0.16996978223323822,
+      "learning_rate": 1.8890053178096353e-05,
+      "loss": 0.5166,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4735668079411109,
+      "grad_norm": 0.16575470566749573,
+      "learning_rate": 1.888897530403987e-05,
+      "loss": 0.5212,
+      "step": 2123
+    },
+    {
+      "epoch": 0.4737898728530002,
+      "grad_norm": 0.1607031673192978,
+      "learning_rate": 1.8887896937656028e-05,
+      "loss": 0.5067,
+      "step": 2124
+    },
+    {
+      "epoch": 0.47401293776488956,
+      "grad_norm": 0.157185897231102,
+      "learning_rate": 1.8886818079004545e-05,
+      "loss": 0.4801,
+      "step": 2125
+    },
+    {
+      "epoch": 0.47423600267677896,
+      "grad_norm": 0.1661345213651657,
+      "learning_rate": 1.8885738728145173e-05,
+      "loss": 0.5212,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4744590675886683,
+      "grad_norm": 0.17212405800819397,
+      "learning_rate": 1.8884658885137698e-05,
+      "loss": 0.4872,
+      "step": 2127
+    },
+    {
+      "epoch": 0.4746821325005577,
+      "grad_norm": 0.1681511402130127,
+      "learning_rate": 1.8883578550041925e-05,
+      "loss": 0.5266,
+      "step": 2128
+    },
+    {
+      "epoch": 0.474905197412447,
+      "grad_norm": 0.15793408453464508,
+      "learning_rate": 1.8882497722917697e-05,
+      "loss": 0.505,
+      "step": 2129
+    },
+    {
+      "epoch": 0.47512826232433636,
+      "grad_norm": 0.1688978672027588,
+      "learning_rate": 1.8881416403824867e-05,
+      "loss": 0.5005,
+      "step": 2130
+    },
+    {
+      "epoch": 0.47535132723622575,
+      "grad_norm": 0.17982251942157745,
+      "learning_rate": 1.8880334592823333e-05,
+      "loss": 0.5177,
+      "step": 2131
+    },
+    {
+      "epoch": 0.4755743921481151,
+      "grad_norm": 0.15668730437755585,
+      "learning_rate": 1.8879252289973008e-05,
+      "loss": 0.5076,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4757974570600045,
+      "grad_norm": 0.15427713096141815,
+      "learning_rate": 1.8878169495333843e-05,
+      "loss": 0.4916,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4760205219718938,
+      "grad_norm": 0.1573878973722458,
+      "learning_rate": 1.88770862089658e-05,
+      "loss": 0.4851,
+      "step": 2134
+    },
+    {
+      "epoch": 0.47624358688378315,
+      "grad_norm": 0.16141542792320251,
+      "learning_rate": 1.887600243092889e-05,
+      "loss": 0.5065,
+      "step": 2135
+    },
+    {
+      "epoch": 0.47646665179567255,
+      "grad_norm": 0.1874970644712448,
+      "learning_rate": 1.8874918161283127e-05,
+      "loss": 0.4987,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4766897167075619,
+      "grad_norm": 0.1556268036365509,
+      "learning_rate": 1.887383340008857e-05,
+      "loss": 0.5001,
+      "step": 2137
+    },
+    {
+      "epoch": 0.4769127816194513,
+      "grad_norm": 0.16177254915237427,
+      "learning_rate": 1.8872748147405303e-05,
+      "loss": 0.5145,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4771358465313406,
+      "grad_norm": 0.16272678971290588,
+      "learning_rate": 1.8871662403293434e-05,
+      "loss": 0.5111,
+      "step": 2139
+    },
+    {
+      "epoch": 0.47735891144323,
+      "grad_norm": 0.157213494181633,
+      "learning_rate": 1.8870576167813096e-05,
+      "loss": 0.4975,
+      "step": 2140
+    },
+    {
+      "epoch": 0.47758197635511934,
+      "grad_norm": 0.16583600640296936,
+      "learning_rate": 1.886948944102445e-05,
+      "loss": 0.4835,
+      "step": 2141
+    },
+    {
+      "epoch": 0.4778050412670087,
+      "grad_norm": 0.15528440475463867,
+      "learning_rate": 1.8868402222987687e-05,
+      "loss": 0.4821,
+      "step": 2142
+    },
+    {
+      "epoch": 0.47802810617889807,
+      "grad_norm": 0.1678980439901352,
+      "learning_rate": 1.8867314513763023e-05,
+      "loss": 0.5126,
+      "step": 2143
+    },
+    {
+      "epoch": 0.4782511710907874,
+      "grad_norm": 0.1719445437192917,
+      "learning_rate": 1.886622631341071e-05,
+      "loss": 0.5278,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4784742360026768,
+      "grad_norm": 0.16177958250045776,
+      "learning_rate": 1.886513762199101e-05,
+      "loss": 0.4716,
+      "step": 2145
+    },
+    {
+      "epoch": 0.47869730091456614,
+      "grad_norm": 0.17055903375148773,
+      "learning_rate": 1.886404843956422e-05,
+      "loss": 0.5155,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4789203658264555,
+      "grad_norm": 0.16013142466545105,
+      "learning_rate": 1.8862958766190673e-05,
+      "loss": 0.4882,
+      "step": 2147
+    },
+    {
+      "epoch": 0.47914343073834487,
+      "grad_norm": 0.15844453871250153,
+      "learning_rate": 1.886186860193072e-05,
+      "loss": 0.5125,
+      "step": 2148
+    },
+    {
+      "epoch": 0.4793664956502342,
+      "grad_norm": 0.16520345211029053,
+      "learning_rate": 1.886077794684474e-05,
+      "loss": 0.5355,
+      "step": 2149
+    },
+    {
+      "epoch": 0.4795895605621236,
+      "grad_norm": 0.16560572385787964,
+      "learning_rate": 1.885968680099314e-05,
+      "loss": 0.5011,
+      "step": 2150
+    },
+    {
+      "epoch": 0.47981262547401293,
+      "grad_norm": 0.16460593044757843,
+      "learning_rate": 1.885859516443636e-05,
+      "loss": 0.5186,
+      "step": 2151
+    },
+    {
+      "epoch": 0.4800356903859023,
+      "grad_norm": 0.16416500508785248,
+      "learning_rate": 1.885750303723485e-05,
+      "loss": 0.5121,
+      "step": 2152
+    },
+    {
+      "epoch": 0.48025875529779166,
+      "grad_norm": 0.1615086942911148,
+      "learning_rate": 1.8856410419449108e-05,
+      "loss": 0.5291,
+      "step": 2153
+    },
+    {
+      "epoch": 0.480481820209681,
+      "grad_norm": 0.16628290712833405,
+      "learning_rate": 1.885531731113965e-05,
+      "loss": 0.5166,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4807048851215704,
+      "grad_norm": 0.212614044547081,
+      "learning_rate": 1.8854223712367017e-05,
+      "loss": 0.4687,
+      "step": 2155
+    },
+    {
+      "epoch": 0.48092795003345973,
+      "grad_norm": 0.16417112946510315,
+      "learning_rate": 1.8853129623191775e-05,
+      "loss": 0.4839,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4811510149453491,
+      "grad_norm": 0.16616222262382507,
+      "learning_rate": 1.8852035043674534e-05,
+      "loss": 0.5013,
+      "step": 2157
+    },
+    {
+      "epoch": 0.48137407985723846,
+      "grad_norm": 0.16397857666015625,
+      "learning_rate": 1.8850939973875907e-05,
+      "loss": 0.5163,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4815971447691278,
+      "grad_norm": 0.1688777208328247,
+      "learning_rate": 1.8849844413856548e-05,
+      "loss": 0.5154,
+      "step": 2159
+    },
+    {
+      "epoch": 0.4818202096810172,
+      "grad_norm": 0.16248802840709686,
+      "learning_rate": 1.884874836367714e-05,
+      "loss": 0.49,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4820432745929065,
+      "grad_norm": 0.17419035732746124,
+      "learning_rate": 1.8847651823398385e-05,
+      "loss": 0.5105,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4822663395047959,
+      "grad_norm": 0.16489480435848236,
+      "learning_rate": 1.884655479308102e-05,
+      "loss": 0.4967,
+      "step": 2162
+    },
+    {
+      "epoch": 0.48248940441668525,
+      "grad_norm": 0.16695787012577057,
+      "learning_rate": 1.8845457272785802e-05,
+      "loss": 0.5316,
+      "step": 2163
+    },
+    {
+      "epoch": 0.4827124693285746,
+      "grad_norm": 0.1630323827266693,
+      "learning_rate": 1.884435926257352e-05,
+      "loss": 0.5144,
+      "step": 2164
+    },
+    {
+      "epoch": 0.482935534240464,
+      "grad_norm": 0.16861893236637115,
+      "learning_rate": 1.8843260762504985e-05,
+      "loss": 0.5402,
+      "step": 2165
+    },
+    {
+      "epoch": 0.4831585991523533,
+      "grad_norm": 0.16794002056121826,
+      "learning_rate": 1.884216177264105e-05,
+      "loss": 0.5581,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4833816640642427,
+      "grad_norm": 0.16783180832862854,
+      "learning_rate": 1.8841062293042572e-05,
+      "loss": 0.4884,
+      "step": 2167
+    },
+    {
+      "epoch": 0.48360472897613205,
+      "grad_norm": 0.16696153581142426,
+      "learning_rate": 1.8839962323770455e-05,
+      "loss": 0.4893,
+      "step": 2168
+    },
+    {
+      "epoch": 0.48382779388802144,
+      "grad_norm": 0.17243477702140808,
+      "learning_rate": 1.8838861864885617e-05,
+      "loss": 0.5134,
+      "step": 2169
+    },
+    {
+      "epoch": 0.4840508587999108,
+      "grad_norm": 0.16844888031482697,
+      "learning_rate": 1.883776091644901e-05,
+      "loss": 0.5203,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4842739237118001,
+      "grad_norm": 0.16011419892311096,
+      "learning_rate": 1.8836659478521614e-05,
+      "loss": 0.4837,
+      "step": 2171
+    },
+    {
+      "epoch": 0.4844969886236895,
+      "grad_norm": 0.16117213666439056,
+      "learning_rate": 1.883555755116443e-05,
+      "loss": 0.5124,
+      "step": 2172
+    },
+    {
+      "epoch": 0.48472005353557884,
+      "grad_norm": 0.15577594935894012,
+      "learning_rate": 1.883445513443849e-05,
+      "loss": 0.4939,
+      "step": 2173
+    },
+    {
+      "epoch": 0.48494311844746824,
+      "grad_norm": 0.17128697037696838,
+      "learning_rate": 1.883335222840485e-05,
+      "loss": 0.5331,
+      "step": 2174
+    },
+    {
+      "epoch": 0.4851661833593576,
+      "grad_norm": 0.16078589856624603,
+      "learning_rate": 1.8832248833124606e-05,
+      "loss": 0.5348,
+      "step": 2175
+    },
+    {
+      "epoch": 0.4853892482712469,
+      "grad_norm": 0.17158381640911102,
+      "learning_rate": 1.8831144948658863e-05,
+      "loss": 0.5016,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4856123131831363,
+      "grad_norm": 0.17395475506782532,
+      "learning_rate": 1.883004057506876e-05,
+      "loss": 0.5145,
+      "step": 2177
+    },
+    {
+      "epoch": 0.48583537809502564,
+      "grad_norm": 0.15187759697437286,
+      "learning_rate": 1.882893571241547e-05,
+      "loss": 0.4747,
+      "step": 2178
+    },
+    {
+      "epoch": 0.48605844300691503,
+      "grad_norm": 0.1744384616613388,
+      "learning_rate": 1.8827830360760184e-05,
+      "loss": 0.488,
+      "step": 2179
+    },
+    {
+      "epoch": 0.48628150791880437,
+      "grad_norm": 0.16714079678058624,
+      "learning_rate": 1.8826724520164118e-05,
+      "loss": 0.5117,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4865045728306937,
+      "grad_norm": 0.1541585475206375,
+      "learning_rate": 1.8825618190688534e-05,
+      "loss": 0.4808,
+      "step": 2181
+    },
+    {
+      "epoch": 0.4867276377425831,
+      "grad_norm": 0.16642670333385468,
+      "learning_rate": 1.8824511372394694e-05,
+      "loss": 0.5223,
+      "step": 2182
+    },
+    {
+      "epoch": 0.48695070265447243,
+      "grad_norm": 0.16418053209781647,
+      "learning_rate": 1.8823404065343904e-05,
+      "loss": 0.5022,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4871737675663618,
+      "grad_norm": 0.1558452993631363,
+      "learning_rate": 1.88222962695975e-05,
+      "loss": 0.5009,
+      "step": 2184
+    },
+    {
+      "epoch": 0.48739683247825116,
+      "grad_norm": 0.16112567484378815,
+      "learning_rate": 1.8821187985216835e-05,
+      "loss": 0.4851,
+      "step": 2185
+    },
+    {
+      "epoch": 0.48761989739014056,
+      "grad_norm": 0.15951795876026154,
+      "learning_rate": 1.8820079212263287e-05,
+      "loss": 0.5364,
+      "step": 2186
+    },
+    {
+      "epoch": 0.4878429623020299,
+      "grad_norm": 0.15660803020000458,
+      "learning_rate": 1.8818969950798274e-05,
+      "loss": 0.4687,
+      "step": 2187
+    },
+    {
+      "epoch": 0.48806602721391923,
+      "grad_norm": 0.16717955470085144,
+      "learning_rate": 1.881786020088323e-05,
+      "loss": 0.5184,
+      "step": 2188
+    },
+    {
+      "epoch": 0.4882890921258086,
+      "grad_norm": 0.4413386583328247,
+      "learning_rate": 1.8816749962579625e-05,
+      "loss": 0.49,
+      "step": 2189
+    },
+    {
+      "epoch": 0.48851215703769796,
+      "grad_norm": 0.18018092215061188,
+      "learning_rate": 1.8815639235948945e-05,
+      "loss": 0.4948,
+      "step": 2190
+    },
+    {
+      "epoch": 0.48873522194958735,
+      "grad_norm": 0.17363835871219635,
+      "learning_rate": 1.881452802105271e-05,
+      "loss": 0.5255,
+      "step": 2191
+    },
+    {
+      "epoch": 0.4889582868614767,
+      "grad_norm": 0.16417300701141357,
+      "learning_rate": 1.8813416317952474e-05,
+      "loss": 0.4984,
+      "step": 2192
+    },
+    {
+      "epoch": 0.489181351773366,
+      "grad_norm": 0.1698407083749771,
+      "learning_rate": 1.8812304126709797e-05,
+      "loss": 0.5155,
+      "step": 2193
+    },
+    {
+      "epoch": 0.4894044166852554,
+      "grad_norm": 0.16547353565692902,
+      "learning_rate": 1.881119144738629e-05,
+      "loss": 0.504,
+      "step": 2194
+    },
+    {
+      "epoch": 0.48962748159714475,
+      "grad_norm": 0.15916848182678223,
+      "learning_rate": 1.8810078280043574e-05,
+      "loss": 0.5158,
+      "step": 2195
+    },
+    {
+      "epoch": 0.48985054650903415,
+      "grad_norm": 0.16655400395393372,
+      "learning_rate": 1.8808964624743303e-05,
+      "loss": 0.521,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4900736114209235,
+      "grad_norm": 0.16791512072086334,
+      "learning_rate": 1.8807850481547165e-05,
+      "loss": 0.5327,
+      "step": 2197
+    },
+    {
+      "epoch": 0.4902966763328129,
+      "grad_norm": 0.1612214297056198,
+      "learning_rate": 1.880673585051686e-05,
+      "loss": 0.5074,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4905197412447022,
+      "grad_norm": 0.33009010553359985,
+      "learning_rate": 1.880562073171413e-05,
+      "loss": 0.4991,
+      "step": 2199
+    },
+    {
+      "epoch": 0.49074280615659155,
+      "grad_norm": 0.17632457613945007,
+      "learning_rate": 1.8804505125200732e-05,
+      "loss": 0.5094,
+      "step": 2200
+    },
+    {
+      "epoch": 0.49096587106848094,
+      "grad_norm": 0.17380337417125702,
+      "learning_rate": 1.8803389031038462e-05,
+      "loss": 0.5132,
+      "step": 2201
+    },
+    {
+      "epoch": 0.4911889359803703,
+      "grad_norm": 0.16013850271701813,
+      "learning_rate": 1.880227244928913e-05,
+      "loss": 0.4987,
+      "step": 2202
+    },
+    {
+      "epoch": 0.49141200089225967,
+      "grad_norm": 0.18471701443195343,
+      "learning_rate": 1.8801155380014578e-05,
+      "loss": 0.5203,
+      "step": 2203
+    },
+    {
+      "epoch": 0.491635065804149,
+      "grad_norm": 0.16928328573703766,
+      "learning_rate": 1.8800037823276683e-05,
+      "loss": 0.4868,
+      "step": 2204
+    },
+    {
+      "epoch": 0.49185813071603834,
+      "grad_norm": 0.1749187409877777,
+      "learning_rate": 1.8798919779137337e-05,
+      "loss": 0.5049,
+      "step": 2205
+    },
+    {
+      "epoch": 0.49208119562792774,
+      "grad_norm": 0.173324853181839,
+      "learning_rate": 1.8797801247658465e-05,
+      "loss": 0.4963,
+      "step": 2206
+    },
+    {
+      "epoch": 0.4923042605398171,
+      "grad_norm": 0.16650435328483582,
+      "learning_rate": 1.8796682228902024e-05,
+      "loss": 0.5232,
+      "step": 2207
+    },
+    {
+      "epoch": 0.49252732545170647,
+      "grad_norm": 0.16675227880477905,
+      "learning_rate": 1.8795562722929986e-05,
+      "loss": 0.4887,
+      "step": 2208
+    },
+    {
+      "epoch": 0.4927503903635958,
+      "grad_norm": 0.20501257479190826,
+      "learning_rate": 1.8794442729804356e-05,
+      "loss": 0.5045,
+      "step": 2209
+    },
+    {
+      "epoch": 0.49297345527548514,
+      "grad_norm": 0.15837359428405762,
+      "learning_rate": 1.879332224958717e-05,
+      "loss": 0.4704,
+      "step": 2210
+    },
+    {
+      "epoch": 0.49319652018737453,
+      "grad_norm": 0.15580782294273376,
+      "learning_rate": 1.8792201282340485e-05,
+      "loss": 0.4753,
+      "step": 2211
+    },
+    {
+      "epoch": 0.49341958509926387,
+      "grad_norm": 0.16155748069286346,
+      "learning_rate": 1.879107982812639e-05,
+      "loss": 0.5178,
+      "step": 2212
+    },
+    {
+      "epoch": 0.49364265001115326,
+      "grad_norm": 0.17078126966953278,
+      "learning_rate": 1.8789957887006994e-05,
+      "loss": 0.5208,
+      "step": 2213
+    },
+    {
+      "epoch": 0.4938657149230426,
+      "grad_norm": 0.16327013075351715,
+      "learning_rate": 1.8788835459044438e-05,
+      "loss": 0.5446,
+      "step": 2214
+    },
+    {
+      "epoch": 0.494088779834932,
+      "grad_norm": 0.1604888141155243,
+      "learning_rate": 1.878771254430089e-05,
+      "loss": 0.5142,
+      "step": 2215
+    },
+    {
+      "epoch": 0.4943118447468213,
+      "grad_norm": 0.16108398139476776,
+      "learning_rate": 1.8786589142838548e-05,
+      "loss": 0.517,
+      "step": 2216
+    },
+    {
+      "epoch": 0.49453490965871066,
+      "grad_norm": 0.15508361160755157,
+      "learning_rate": 1.8785465254719625e-05,
+      "loss": 0.4782,
+      "step": 2217
+    },
+    {
+      "epoch": 0.49475797457060006,
+      "grad_norm": 0.1612912118434906,
+      "learning_rate": 1.878434088000638e-05,
+      "loss": 0.5104,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4949810394824894,
+      "grad_norm": 0.1641056090593338,
+      "learning_rate": 1.8783216018761075e-05,
+      "loss": 0.5227,
+      "step": 2219
+    },
+    {
+      "epoch": 0.4952041043943788,
+      "grad_norm": 0.16892564296722412,
+      "learning_rate": 1.878209067104602e-05,
+      "loss": 0.5142,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4954271693062681,
+      "grad_norm": 0.18777111172676086,
+      "learning_rate": 1.8780964836923545e-05,
+      "loss": 0.5091,
+      "step": 2221
+    },
+    {
+      "epoch": 0.49565023421815746,
+      "grad_norm": 0.15524406731128693,
+      "learning_rate": 1.8779838516455998e-05,
+      "loss": 0.5155,
+      "step": 2222
+    },
+    {
+      "epoch": 0.49587329913004685,
+      "grad_norm": 0.1688852459192276,
+      "learning_rate": 1.877871170970577e-05,
+      "loss": 0.5266,
+      "step": 2223
+    },
+    {
+      "epoch": 0.4960963640419362,
+      "grad_norm": 0.2029954195022583,
+      "learning_rate": 1.8777584416735268e-05,
+      "loss": 0.4751,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4963194289538256,
+      "grad_norm": 0.168426051735878,
+      "learning_rate": 1.8776456637606926e-05,
+      "loss": 0.5263,
+      "step": 2225
+    },
+    {
+      "epoch": 0.4965424938657149,
+      "grad_norm": 0.18985538184642792,
+      "learning_rate": 1.877532837238321e-05,
+      "loss": 0.528,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4967655587776043,
+      "grad_norm": 0.15908297896385193,
+      "learning_rate": 1.8774199621126605e-05,
+      "loss": 0.5226,
+      "step": 2227
+    },
+    {
+      "epoch": 0.49698862368949365,
+      "grad_norm": 0.15700525045394897,
+      "learning_rate": 1.8773070383899638e-05,
+      "loss": 0.4857,
+      "step": 2228
+    },
+    {
+      "epoch": 0.497211688601383,
+      "grad_norm": 0.17015953361988068,
+      "learning_rate": 1.877194066076485e-05,
+      "loss": 0.488,
+      "step": 2229
+    },
+    {
+      "epoch": 0.4974347535132724,
+      "grad_norm": 0.21846015751361847,
+      "learning_rate": 1.8770810451784806e-05,
+      "loss": 0.5305,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4976578184251617,
+      "grad_norm": 0.16297823190689087,
+      "learning_rate": 1.8769679757022114e-05,
+      "loss": 0.494,
+      "step": 2231
+    },
+    {
+      "epoch": 0.4978808833370511,
+      "grad_norm": 0.1758735030889511,
+      "learning_rate": 1.876854857653939e-05,
+      "loss": 0.5394,
+      "step": 2232
+    },
+    {
+      "epoch": 0.49810394824894044,
+      "grad_norm": 0.1673547923564911,
+      "learning_rate": 1.876741691039929e-05,
+      "loss": 0.5035,
+      "step": 2233
+    },
+    {
+      "epoch": 0.4983270131608298,
+      "grad_norm": 0.16478639841079712,
+      "learning_rate": 1.8766284758664487e-05,
+      "loss": 0.5,
+      "step": 2234
+    },
+    {
+      "epoch": 0.49855007807271917,
+      "grad_norm": 0.16910696029663086,
+      "learning_rate": 1.8765152121397697e-05,
+      "loss": 0.5087,
+      "step": 2235
+    },
+    {
+      "epoch": 0.4987731429846085,
+      "grad_norm": 0.16947272419929504,
+      "learning_rate": 1.876401899866165e-05,
+      "loss": 0.5064,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4989962078964979,
+      "grad_norm": 0.16172346472740173,
+      "learning_rate": 1.87628853905191e-05,
+      "loss": 0.4901,
+      "step": 2237
+    },
+    {
+      "epoch": 0.49921927280838724,
+      "grad_norm": 0.16061736643314362,
+      "learning_rate": 1.8761751297032838e-05,
+      "loss": 0.5038,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4994423377202766,
+      "grad_norm": 0.15990641713142395,
+      "learning_rate": 1.8760616718265676e-05,
+      "loss": 0.4854,
+      "step": 2239
+    },
+    {
+      "epoch": 0.49966540263216597,
+      "grad_norm": 0.16735920310020447,
+      "learning_rate": 1.875948165428045e-05,
+      "loss": 0.5027,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4998884675440553,
+      "grad_norm": 0.16122539341449738,
+      "learning_rate": 1.8758346105140033e-05,
+      "loss": 0.4922,
+      "step": 2241
+    },
+    {
+      "epoch": 0.5001115324559446,
+      "grad_norm": 0.1718963235616684,
+      "learning_rate": 1.8757210070907315e-05,
+      "loss": 0.5138,
+      "step": 2242
+    },
+    {
+      "epoch": 0.500334597367834,
+      "grad_norm": 0.17143313586711884,
+      "learning_rate": 1.875607355164522e-05,
+      "loss": 0.5308,
+      "step": 2243
+    },
+    {
+      "epoch": 0.5005576622797234,
+      "grad_norm": 0.15493686497211456,
+      "learning_rate": 1.875493654741669e-05,
+      "loss": 0.5243,
+      "step": 2244
+    },
+    {
+      "epoch": 0.5007807271916127,
+      "grad_norm": 0.162288635969162,
+      "learning_rate": 1.8753799058284707e-05,
+      "loss": 0.5052,
+      "step": 2245
+    },
+    {
+      "epoch": 0.5010037921035021,
+      "grad_norm": 0.17301948368549347,
+      "learning_rate": 1.8752661084312268e-05,
+      "loss": 0.5119,
+      "step": 2246
+    },
+    {
+      "epoch": 0.5012268570153915,
+      "grad_norm": 0.16729313135147095,
+      "learning_rate": 1.8751522625562405e-05,
+      "loss": 0.5372,
+      "step": 2247
+    },
+    {
+      "epoch": 0.5014499219272809,
+      "grad_norm": 0.16277892887592316,
+      "learning_rate": 1.8750383682098166e-05,
+      "loss": 0.4847,
+      "step": 2248
+    },
+    {
+      "epoch": 0.5016729868391702,
+      "grad_norm": 0.16257484257221222,
+      "learning_rate": 1.8749244253982633e-05,
+      "loss": 0.5105,
+      "step": 2249
+    },
+    {
+      "epoch": 0.5018960517510596,
+      "grad_norm": 0.16304615139961243,
+      "learning_rate": 1.8748104341278924e-05,
+      "loss": 0.5073,
+      "step": 2250
+    },
+    {
+      "epoch": 0.502119116662949,
+      "grad_norm": 0.1586243063211441,
+      "learning_rate": 1.874696394405017e-05,
+      "loss": 0.5096,
+      "step": 2251
+    },
+    {
+      "epoch": 0.5023421815748382,
+      "grad_norm": 0.16425400972366333,
+      "learning_rate": 1.874582306235953e-05,
+      "loss": 0.4817,
+      "step": 2252
+    },
+    {
+      "epoch": 0.5025652464867276,
+      "grad_norm": 0.16808679699897766,
+      "learning_rate": 1.874468169627019e-05,
+      "loss": 0.4866,
+      "step": 2253
+    },
+    {
+      "epoch": 0.502788311398617,
+      "grad_norm": 0.1656448394060135,
+      "learning_rate": 1.8743539845845378e-05,
+      "loss": 0.4903,
+      "step": 2254
+    },
+    {
+      "epoch": 0.5030113763105064,
+      "grad_norm": 0.16811887919902802,
+      "learning_rate": 1.8742397511148328e-05,
+      "loss": 0.4954,
+      "step": 2255
+    },
+    {
+      "epoch": 0.5032344412223957,
+      "grad_norm": 0.17424499988555908,
+      "learning_rate": 1.8741254692242315e-05,
+      "loss": 0.5301,
+      "step": 2256
+    },
+    {
+      "epoch": 0.5034575061342851,
+      "grad_norm": 0.17697711288928986,
+      "learning_rate": 1.874011138919063e-05,
+      "loss": 0.5103,
+      "step": 2257
+    },
+    {
+      "epoch": 0.5036805710461745,
+      "grad_norm": 0.16454057395458221,
+      "learning_rate": 1.8738967602056597e-05,
+      "loss": 0.499,
+      "step": 2258
+    },
+    {
+      "epoch": 0.5039036359580638,
+      "grad_norm": 0.16936853528022766,
+      "learning_rate": 1.873782333090357e-05,
+      "loss": 0.5029,
+      "step": 2259
+    },
+    {
+      "epoch": 0.5041267008699531,
+      "grad_norm": 0.16321606934070587,
+      "learning_rate": 1.873667857579492e-05,
+      "loss": 0.5043,
+      "step": 2260
+    },
+    {
+      "epoch": 0.5043497657818425,
+      "grad_norm": 0.17789475619792938,
+      "learning_rate": 1.873553333679406e-05,
+      "loss": 0.5206,
+      "step": 2261
+    },
+    {
+      "epoch": 0.5045728306937318,
+      "grad_norm": 0.17053046822547913,
+      "learning_rate": 1.8734387613964414e-05,
+      "loss": 0.5241,
+      "step": 2262
+    },
+    {
+      "epoch": 0.5047958956056212,
+      "grad_norm": 0.16408482193946838,
+      "learning_rate": 1.8733241407369438e-05,
+      "loss": 0.4999,
+      "step": 2263
+    },
+    {
+      "epoch": 0.5050189605175106,
+      "grad_norm": 0.1780211627483368,
+      "learning_rate": 1.873209471707262e-05,
+      "loss": 0.5014,
+      "step": 2264
+    },
+    {
+      "epoch": 0.5052420254294,
+      "grad_norm": 0.17059922218322754,
+      "learning_rate": 1.873094754313747e-05,
+      "loss": 0.5315,
+      "step": 2265
+    },
+    {
+      "epoch": 0.5054650903412893,
+      "grad_norm": 0.17030328512191772,
+      "learning_rate": 1.8729799885627528e-05,
+      "loss": 0.5295,
+      "step": 2266
+    },
+    {
+      "epoch": 0.5056881552531787,
+      "grad_norm": 0.1972958743572235,
+      "learning_rate": 1.872865174460635e-05,
+      "loss": 0.5094,
+      "step": 2267
+    },
+    {
+      "epoch": 0.5059112201650681,
+      "grad_norm": 0.1663295030593872,
+      "learning_rate": 1.8727503120137537e-05,
+      "loss": 0.5023,
+      "step": 2268
+    },
+    {
+      "epoch": 0.5061342850769573,
+      "grad_norm": 0.1621520221233368,
+      "learning_rate": 1.87263540122847e-05,
+      "loss": 0.5141,
+      "step": 2269
+    },
+    {
+      "epoch": 0.5063573499888467,
+      "grad_norm": 0.35805073380470276,
+      "learning_rate": 1.872520442111149e-05,
+      "loss": 0.539,
+      "step": 2270
+    },
+    {
+      "epoch": 0.5065804149007361,
+      "grad_norm": 0.16250286996364594,
+      "learning_rate": 1.8724054346681573e-05,
+      "loss": 0.5173,
+      "step": 2271
+    },
+    {
+      "epoch": 0.5068034798126255,
+      "grad_norm": 0.16895325481891632,
+      "learning_rate": 1.872290378905865e-05,
+      "loss": 0.5041,
+      "step": 2272
+    },
+    {
+      "epoch": 0.5070265447245148,
+      "grad_norm": 0.1690344661474228,
+      "learning_rate": 1.872175274830645e-05,
+      "loss": 0.5337,
+      "step": 2273
+    },
+    {
+      "epoch": 0.5072496096364042,
+      "grad_norm": 0.15611621737480164,
+      "learning_rate": 1.8720601224488716e-05,
+      "loss": 0.4819,
+      "step": 2274
+    },
+    {
+      "epoch": 0.5074726745482936,
+      "grad_norm": 0.18844662606716156,
+      "learning_rate": 1.871944921766923e-05,
+      "loss": 0.5082,
+      "step": 2275
+    },
+    {
+      "epoch": 0.5076957394601829,
+      "grad_norm": 0.16826370358467102,
+      "learning_rate": 1.8718296727911803e-05,
+      "loss": 0.4914,
+      "step": 2276
+    },
+    {
+      "epoch": 0.5079188043720723,
+      "grad_norm": 0.15561175346374512,
+      "learning_rate": 1.871714375528026e-05,
+      "loss": 0.5147,
+      "step": 2277
+    },
+    {
+      "epoch": 0.5081418692839617,
+      "grad_norm": 0.1746017336845398,
+      "learning_rate": 1.8715990299838463e-05,
+      "loss": 0.5224,
+      "step": 2278
+    },
+    {
+      "epoch": 0.508364934195851,
+      "grad_norm": 0.1721627563238144,
+      "learning_rate": 1.8714836361650303e-05,
+      "loss": 0.509,
+      "step": 2279
+    },
+    {
+      "epoch": 0.5085879991077403,
+      "grad_norm": 0.15273146331310272,
+      "learning_rate": 1.871368194077968e-05,
+      "loss": 0.4933,
+      "step": 2280
+    },
+    {
+      "epoch": 0.5088110640196297,
+      "grad_norm": 0.17017610371112823,
+      "learning_rate": 1.8712527037290546e-05,
+      "loss": 0.518,
+      "step": 2281
+    },
+    {
+      "epoch": 0.5090341289315191,
+      "grad_norm": 0.16407892107963562,
+      "learning_rate": 1.8711371651246854e-05,
+      "loss": 0.5152,
+      "step": 2282
+    },
+    {
+      "epoch": 0.5092571938434084,
+      "grad_norm": 0.18274806439876556,
+      "learning_rate": 1.8710215782712606e-05,
+      "loss": 0.4959,
+      "step": 2283
+    },
+    {
+      "epoch": 0.5094802587552978,
+      "grad_norm": 0.17501875758171082,
+      "learning_rate": 1.870905943175182e-05,
+      "loss": 0.5372,
+      "step": 2284
+    },
+    {
+      "epoch": 0.5097033236671872,
+      "grad_norm": 0.1640838235616684,
+      "learning_rate": 1.870790259842854e-05,
+      "loss": 0.5102,
+      "step": 2285
+    },
+    {
+      "epoch": 0.5099263885790765,
+      "grad_norm": 0.15836408734321594,
+      "learning_rate": 1.870674528280684e-05,
+      "loss": 0.5194,
+      "step": 2286
+    },
+    {
+      "epoch": 0.5101494534909659,
+      "grad_norm": 0.16688112914562225,
+      "learning_rate": 1.8705587484950815e-05,
+      "loss": 0.5077,
+      "step": 2287
+    },
+    {
+      "epoch": 0.5103725184028552,
+      "grad_norm": 0.16494499146938324,
+      "learning_rate": 1.8704429204924598e-05,
+      "loss": 0.4954,
+      "step": 2288
+    },
+    {
+      "epoch": 0.5105955833147446,
+      "grad_norm": 0.1580001264810562,
+      "learning_rate": 1.8703270442792337e-05,
+      "loss": 0.507,
+      "step": 2289
+    },
+    {
+      "epoch": 0.5108186482266339,
+      "grad_norm": 0.17867796123027802,
+      "learning_rate": 1.8702111198618213e-05,
+      "loss": 0.5081,
+      "step": 2290
+    },
+    {
+      "epoch": 0.5110417131385233,
+      "grad_norm": 0.28131747245788574,
+      "learning_rate": 1.8700951472466435e-05,
+      "loss": 0.5188,
+      "step": 2291
+    },
+    {
+      "epoch": 0.5112647780504127,
+      "grad_norm": 0.1570468544960022,
+      "learning_rate": 1.869979126440123e-05,
+      "loss": 0.4992,
+      "step": 2292
+    },
+    {
+      "epoch": 0.511487842962302,
+      "grad_norm": 0.1686072200536728,
+      "learning_rate": 1.8698630574486862e-05,
+      "loss": 0.4951,
+      "step": 2293
+    },
+    {
+      "epoch": 0.5117109078741914,
+      "grad_norm": 0.1702122986316681,
+      "learning_rate": 1.869746940278762e-05,
+      "loss": 0.4965,
+      "step": 2294
+    },
+    {
+      "epoch": 0.5119339727860808,
+      "grad_norm": 0.1632847636938095,
+      "learning_rate": 1.8696307749367807e-05,
+      "loss": 0.48,
+      "step": 2295
+    },
+    {
+      "epoch": 0.5121570376979702,
+      "grad_norm": 0.20230402052402496,
+      "learning_rate": 1.8695145614291773e-05,
+      "loss": 0.5262,
+      "step": 2296
+    },
+    {
+      "epoch": 0.5123801026098594,
+      "grad_norm": 0.17471125721931458,
+      "learning_rate": 1.8693982997623877e-05,
+      "loss": 0.5207,
+      "step": 2297
+    },
+    {
+      "epoch": 0.5126031675217488,
+      "grad_norm": 0.17141500115394592,
+      "learning_rate": 1.869281989942852e-05,
+      "loss": 0.4995,
+      "step": 2298
+    },
+    {
+      "epoch": 0.5128262324336382,
+      "grad_norm": 0.1641875058412552,
+      "learning_rate": 1.8691656319770112e-05,
+      "loss": 0.4773,
+      "step": 2299
+    },
+    {
+      "epoch": 0.5130492973455275,
+      "grad_norm": 0.1655300408601761,
+      "learning_rate": 1.8690492258713107e-05,
+      "loss": 0.5171,
+      "step": 2300
+    },
+    {
+      "epoch": 0.5132723622574169,
+      "grad_norm": 0.15635208785533905,
+      "learning_rate": 1.8689327716321975e-05,
+      "loss": 0.5005,
+      "step": 2301
+    },
+    {
+      "epoch": 0.5134954271693063,
+      "grad_norm": 0.1540510356426239,
+      "learning_rate": 1.8688162692661214e-05,
+      "loss": 0.4925,
+      "step": 2302
+    },
+    {
+      "epoch": 0.5137184920811956,
+      "grad_norm": 0.16273614764213562,
+      "learning_rate": 1.8686997187795354e-05,
+      "loss": 0.5153,
+      "step": 2303
+    },
+    {
+      "epoch": 0.513941556993085,
+      "grad_norm": 0.16257858276367188,
+      "learning_rate": 1.8685831201788945e-05,
+      "loss": 0.495,
+      "step": 2304
+    },
+    {
+      "epoch": 0.5141646219049744,
+      "grad_norm": 0.1771703064441681,
+      "learning_rate": 1.8684664734706572e-05,
+      "loss": 0.5499,
+      "step": 2305
+    },
+    {
+      "epoch": 0.5143876868168638,
+      "grad_norm": 0.17417144775390625,
+      "learning_rate": 1.8683497786612834e-05,
+      "loss": 0.4777,
+      "step": 2306
+    },
+    {
+      "epoch": 0.514610751728753,
+      "grad_norm": 0.2703874707221985,
+      "learning_rate": 1.8682330357572368e-05,
+      "loss": 0.5032,
+      "step": 2307
+    },
+    {
+      "epoch": 0.5148338166406424,
+      "grad_norm": 0.16738004982471466,
+      "learning_rate": 1.8681162447649834e-05,
+      "loss": 0.514,
+      "step": 2308
+    },
+    {
+      "epoch": 0.5150568815525318,
+      "grad_norm": 0.17614148557186127,
+      "learning_rate": 1.8679994056909915e-05,
+      "loss": 0.5193,
+      "step": 2309
+    },
+    {
+      "epoch": 0.5152799464644211,
+      "grad_norm": 0.1655927300453186,
+      "learning_rate": 1.8678825185417328e-05,
+      "loss": 0.5253,
+      "step": 2310
+    },
+    {
+      "epoch": 0.5155030113763105,
+      "grad_norm": 0.16509908437728882,
+      "learning_rate": 1.867765583323681e-05,
+      "loss": 0.5108,
+      "step": 2311
+    },
+    {
+      "epoch": 0.5157260762881999,
+      "grad_norm": 0.17742401361465454,
+      "learning_rate": 1.8676486000433123e-05,
+      "loss": 0.5024,
+      "step": 2312
+    },
+    {
+      "epoch": 0.5159491412000893,
+      "grad_norm": 0.1574956476688385,
+      "learning_rate": 1.8675315687071068e-05,
+      "loss": 0.4944,
+      "step": 2313
+    },
+    {
+      "epoch": 0.5161722061119786,
+      "grad_norm": 0.16360540688037872,
+      "learning_rate": 1.867414489321546e-05,
+      "loss": 0.5054,
+      "step": 2314
+    },
+    {
+      "epoch": 0.516395271023868,
+      "grad_norm": 0.1697058230638504,
+      "learning_rate": 1.8672973618931144e-05,
+      "loss": 0.5198,
+      "step": 2315
+    },
+    {
+      "epoch": 0.5166183359357573,
+      "grad_norm": 0.16391722857952118,
+      "learning_rate": 1.8671801864282996e-05,
+      "loss": 0.5138,
+      "step": 2316
+    },
+    {
+      "epoch": 0.5168414008476466,
+      "grad_norm": 0.1696222871541977,
+      "learning_rate": 1.867062962933591e-05,
+      "loss": 0.4929,
+      "step": 2317
+    },
+    {
+      "epoch": 0.517064465759536,
+      "grad_norm": 0.16633786261081696,
+      "learning_rate": 1.8669456914154817e-05,
+      "loss": 0.4723,
+      "step": 2318
+    },
+    {
+      "epoch": 0.5172875306714254,
+      "grad_norm": 0.16019189357757568,
+      "learning_rate": 1.8668283718804664e-05,
+      "loss": 0.5093,
+      "step": 2319
+    },
+    {
+      "epoch": 0.5175105955833147,
+      "grad_norm": 0.1600850522518158,
+      "learning_rate": 1.8667110043350435e-05,
+      "loss": 0.4998,
+      "step": 2320
+    },
+    {
+      "epoch": 0.5177336604952041,
+      "grad_norm": 0.1658647060394287,
+      "learning_rate": 1.8665935887857136e-05,
+      "loss": 0.5201,
+      "step": 2321
+    },
+    {
+      "epoch": 0.5179567254070935,
+      "grad_norm": 0.16224202513694763,
+      "learning_rate": 1.8664761252389795e-05,
+      "loss": 0.4979,
+      "step": 2322
+    },
+    {
+      "epoch": 0.5181797903189829,
+      "grad_norm": 0.16525104641914368,
+      "learning_rate": 1.866358613701347e-05,
+      "loss": 0.5168,
+      "step": 2323
+    },
+    {
+      "epoch": 0.5184028552308722,
+      "grad_norm": 0.16819415986537933,
+      "learning_rate": 1.866241054179325e-05,
+      "loss": 0.5189,
+      "step": 2324
+    },
+    {
+      "epoch": 0.5186259201427615,
+      "grad_norm": 0.1534615010023117,
+      "learning_rate": 1.8661234466794246e-05,
+      "loss": 0.4935,
+      "step": 2325
+    },
+    {
+      "epoch": 0.5188489850546509,
+      "grad_norm": 0.15495775640010834,
+      "learning_rate": 1.8660057912081598e-05,
+      "loss": 0.5039,
+      "step": 2326
+    },
+    {
+      "epoch": 0.5190720499665402,
+      "grad_norm": 0.16145430505275726,
+      "learning_rate": 1.8658880877720467e-05,
+      "loss": 0.4758,
+      "step": 2327
+    },
+    {
+      "epoch": 0.5192951148784296,
+      "grad_norm": 0.16488440334796906,
+      "learning_rate": 1.8657703363776044e-05,
+      "loss": 0.5183,
+      "step": 2328
+    },
+    {
+      "epoch": 0.519518179790319,
+      "grad_norm": 0.16661033034324646,
+      "learning_rate": 1.8656525370313553e-05,
+      "loss": 0.519,
+      "step": 2329
+    },
+    {
+      "epoch": 0.5197412447022084,
+      "grad_norm": 0.15961341559886932,
+      "learning_rate": 1.8655346897398234e-05,
+      "loss": 0.478,
+      "step": 2330
+    },
+    {
+      "epoch": 0.5199643096140977,
+      "grad_norm": 0.15573342144489288,
+      "learning_rate": 1.865416794509536e-05,
+      "loss": 0.4843,
+      "step": 2331
+    },
+    {
+      "epoch": 0.5201873745259871,
+      "grad_norm": 0.16535112261772156,
+      "learning_rate": 1.8652988513470227e-05,
+      "loss": 0.4943,
+      "step": 2332
+    },
+    {
+      "epoch": 0.5204104394378765,
+      "grad_norm": 0.16046664118766785,
+      "learning_rate": 1.865180860258816e-05,
+      "loss": 0.5445,
+      "step": 2333
+    },
+    {
+      "epoch": 0.5206335043497657,
+      "grad_norm": 0.16381436586380005,
+      "learning_rate": 1.8650628212514516e-05,
+      "loss": 0.4956,
+      "step": 2334
+    },
+    {
+      "epoch": 0.5208565692616551,
+      "grad_norm": 0.17565996944904327,
+      "learning_rate": 1.864944734331466e-05,
+      "loss": 0.5186,
+      "step": 2335
+    },
+    {
+      "epoch": 0.5210796341735445,
+      "grad_norm": 0.15597450733184814,
+      "learning_rate": 1.8648265995054005e-05,
+      "loss": 0.4993,
+      "step": 2336
+    },
+    {
+      "epoch": 0.5213026990854338,
+      "grad_norm": 0.22045518457889557,
+      "learning_rate": 1.8647084167797982e-05,
+      "loss": 0.4938,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5215257639973232,
+      "grad_norm": 0.14948628842830658,
+      "learning_rate": 1.8645901861612044e-05,
+      "loss": 0.4637,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5217488289092126,
+      "grad_norm": 0.16708412766456604,
+      "learning_rate": 1.8644719076561675e-05,
+      "loss": 0.5093,
+      "step": 2339
+    },
+    {
+      "epoch": 0.521971893821102,
+      "grad_norm": 0.1535186767578125,
+      "learning_rate": 1.8643535812712386e-05,
+      "loss": 0.4894,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5221949587329913,
+      "grad_norm": 0.19612246751785278,
+      "learning_rate": 1.8642352070129715e-05,
+      "loss": 0.4719,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5224180236448807,
+      "grad_norm": 0.1691182553768158,
+      "learning_rate": 1.8641167848879225e-05,
+      "loss": 0.5119,
+      "step": 2342
+    },
+    {
+      "epoch": 0.52264108855677,
+      "grad_norm": 0.16633757948875427,
+      "learning_rate": 1.86399831490265e-05,
+      "loss": 0.4941,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5228641534686593,
+      "grad_norm": 0.15868477523326874,
+      "learning_rate": 1.8638797970637162e-05,
+      "loss": 0.5114,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5230872183805487,
+      "grad_norm": 0.1649598479270935,
+      "learning_rate": 1.8637612313776856e-05,
+      "loss": 0.5446,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5233102832924381,
+      "grad_norm": 0.1706569939851761,
+      "learning_rate": 1.8636426178511246e-05,
+      "loss": 0.509,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5235333482043275,
+      "grad_norm": 0.16140028834342957,
+      "learning_rate": 1.8635239564906026e-05,
+      "loss": 0.4963,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5237564131162168,
+      "grad_norm": 0.16061624884605408,
+      "learning_rate": 1.8634052473026925e-05,
+      "loss": 0.5149,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5239794780281062,
+      "grad_norm": 0.1699933260679245,
+      "learning_rate": 1.8632864902939684e-05,
+      "loss": 0.5436,
+      "step": 2349
+    },
+    {
+      "epoch": 0.5242025429399956,
+      "grad_norm": 0.16812016069889069,
+      "learning_rate": 1.8631676854710082e-05,
+      "loss": 0.5179,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5244256078518849,
+      "grad_norm": 0.1655721366405487,
+      "learning_rate": 1.8630488328403924e-05,
+      "loss": 0.5069,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5246486727637742,
+      "grad_norm": 0.28273990750312805,
+      "learning_rate": 1.8629299324087032e-05,
+      "loss": 0.5207,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5248717376756636,
+      "grad_norm": 0.167250394821167,
+      "learning_rate": 1.8628109841825263e-05,
+      "loss": 0.4995,
+      "step": 2353
+    },
+    {
+      "epoch": 0.525094802587553,
+      "grad_norm": 0.15988002717494965,
+      "learning_rate": 1.8626919881684497e-05,
+      "loss": 0.5331,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5253178674994423,
+      "grad_norm": 0.18735967576503754,
+      "learning_rate": 1.8625729443730643e-05,
+      "loss": 0.508,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5255409324113317,
+      "grad_norm": 0.16121627390384674,
+      "learning_rate": 1.8624538528029638e-05,
+      "loss": 0.4926,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5257639973232211,
+      "grad_norm": 0.16314919292926788,
+      "learning_rate": 1.8623347134647437e-05,
+      "loss": 0.4936,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5259870622351104,
+      "grad_norm": 0.16220073401927948,
+      "learning_rate": 1.862215526365003e-05,
+      "loss": 0.4866,
+      "step": 2358
+    },
+    {
+      "epoch": 0.5262101271469998,
+      "grad_norm": 0.18198350071907043,
+      "learning_rate": 1.8620962915103425e-05,
+      "loss": 0.4969,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5264331920588892,
+      "grad_norm": 0.16916455328464508,
+      "learning_rate": 1.8619770089073665e-05,
+      "loss": 0.4872,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5266562569707784,
+      "grad_norm": 0.19399681687355042,
+      "learning_rate": 1.861857678562682e-05,
+      "loss": 0.5037,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5268793218826678,
+      "grad_norm": 0.16876184940338135,
+      "learning_rate": 1.8617383004828978e-05,
+      "loss": 0.5265,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5271023867945572,
+      "grad_norm": 0.17060886323451996,
+      "learning_rate": 1.8616188746746262e-05,
+      "loss": 0.5247,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5273254517064466,
+      "grad_norm": 0.17065522074699402,
+      "learning_rate": 1.8614994011444812e-05,
+      "loss": 0.5258,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5275485166183359,
+      "grad_norm": 0.41077789664268494,
+      "learning_rate": 1.8613798798990806e-05,
+      "loss": 0.4971,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5277715815302253,
+      "grad_norm": 0.1597769409418106,
+      "learning_rate": 1.8612603109450437e-05,
+      "loss": 0.4971,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5279946464421147,
+      "grad_norm": 0.15872445702552795,
+      "learning_rate": 1.8611406942889934e-05,
+      "loss": 0.4987,
+      "step": 2367
+    },
+    {
+      "epoch": 0.528217711354004,
+      "grad_norm": 0.1601647585630417,
+      "learning_rate": 1.861021029937555e-05,
+      "loss": 0.4791,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5284407762658934,
+      "grad_norm": 0.1533002108335495,
+      "learning_rate": 1.8609013178973555e-05,
+      "loss": 0.4842,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5286638411777828,
+      "grad_norm": 0.16213175654411316,
+      "learning_rate": 1.8607815581750257e-05,
+      "loss": 0.5063,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5288869060896721,
+      "grad_norm": 0.1895373910665512,
+      "learning_rate": 1.860661750777199e-05,
+      "loss": 0.4885,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5291099710015614,
+      "grad_norm": 0.1578717827796936,
+      "learning_rate": 1.8605418957105105e-05,
+      "loss": 0.4541,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5293330359134508,
+      "grad_norm": 0.17269465327262878,
+      "learning_rate": 1.8604219929815987e-05,
+      "loss": 0.5234,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5295561008253402,
+      "grad_norm": 0.16912835836410522,
+      "learning_rate": 1.860302042597105e-05,
+      "loss": 0.5295,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5297791657372295,
+      "grad_norm": 0.16457562148571014,
+      "learning_rate": 1.8601820445636722e-05,
+      "loss": 0.5096,
+      "step": 2375
+    },
+    {
+      "epoch": 0.5300022306491189,
+      "grad_norm": 0.17353565990924835,
+      "learning_rate": 1.860061998887947e-05,
+      "loss": 0.5086,
+      "step": 2376
+    },
+    {
+      "epoch": 0.5302252955610083,
+      "grad_norm": 0.1719389110803604,
+      "learning_rate": 1.859941905576579e-05,
+      "loss": 0.5045,
+      "step": 2377
+    },
+    {
+      "epoch": 0.5304483604728976,
+      "grad_norm": 0.15863987803459167,
+      "learning_rate": 1.8598217646362183e-05,
+      "loss": 0.5007,
+      "step": 2378
+    },
+    {
+      "epoch": 0.530671425384787,
+      "grad_norm": 0.16614408791065216,
+      "learning_rate": 1.85970157607352e-05,
+      "loss": 0.5123,
+      "step": 2379
+    },
+    {
+      "epoch": 0.5308944902966763,
+      "grad_norm": 0.1657867580652237,
+      "learning_rate": 1.859581339895141e-05,
+      "loss": 0.5237,
+      "step": 2380
+    },
+    {
+      "epoch": 0.5311175552085657,
+      "grad_norm": 0.1731244921684265,
+      "learning_rate": 1.85946105610774e-05,
+      "loss": 0.5407,
+      "step": 2381
+    },
+    {
+      "epoch": 0.531340620120455,
+      "grad_norm": 0.15216992795467377,
+      "learning_rate": 1.85934072471798e-05,
+      "loss": 0.4709,
+      "step": 2382
+    },
+    {
+      "epoch": 0.5315636850323444,
+      "grad_norm": 0.1583755612373352,
+      "learning_rate": 1.8592203457325248e-05,
+      "loss": 0.491,
+      "step": 2383
+    },
+    {
+      "epoch": 0.5317867499442338,
+      "grad_norm": 0.16302180290222168,
+      "learning_rate": 1.859099919158042e-05,
+      "loss": 0.4984,
+      "step": 2384
+    },
+    {
+      "epoch": 0.5320098148561231,
+      "grad_norm": 0.3964548110961914,
+      "learning_rate": 1.858979445001202e-05,
+      "loss": 0.5208,
+      "step": 2385
+    },
+    {
+      "epoch": 0.5322328797680125,
+      "grad_norm": 0.1690683513879776,
+      "learning_rate": 1.8588589232686768e-05,
+      "loss": 0.4932,
+      "step": 2386
+    },
+    {
+      "epoch": 0.5324559446799019,
+      "grad_norm": 0.18451258540153503,
+      "learning_rate": 1.8587383539671424e-05,
+      "loss": 0.5261,
+      "step": 2387
+    },
+    {
+      "epoch": 0.5326790095917913,
+      "grad_norm": 0.18741649389266968,
+      "learning_rate": 1.858617737103276e-05,
+      "loss": 0.499,
+      "step": 2388
+    },
+    {
+      "epoch": 0.5329020745036805,
+      "grad_norm": 0.16601239144802094,
+      "learning_rate": 1.8584970726837587e-05,
+      "loss": 0.5176,
+      "step": 2389
+    },
+    {
+      "epoch": 0.5331251394155699,
+      "grad_norm": 0.16434918344020844,
+      "learning_rate": 1.858376360715273e-05,
+      "loss": 0.497,
+      "step": 2390
+    },
+    {
+      "epoch": 0.5333482043274593,
+      "grad_norm": 0.16978149116039276,
+      "learning_rate": 1.8582556012045053e-05,
+      "loss": 0.4946,
+      "step": 2391
+    },
+    {
+      "epoch": 0.5335712692393486,
+      "grad_norm": 0.17810384929180145,
+      "learning_rate": 1.8581347941581438e-05,
+      "loss": 0.5223,
+      "step": 2392
+    },
+    {
+      "epoch": 0.533794334151238,
+      "grad_norm": 0.1722012311220169,
+      "learning_rate": 1.8580139395828795e-05,
+      "loss": 0.5082,
+      "step": 2393
+    },
+    {
+      "epoch": 0.5340173990631274,
+      "grad_norm": 0.16442382335662842,
+      "learning_rate": 1.857893037485406e-05,
+      "loss": 0.4838,
+      "step": 2394
+    },
+    {
+      "epoch": 0.5342404639750167,
+      "grad_norm": 0.17343981564044952,
+      "learning_rate": 1.8577720878724195e-05,
+      "loss": 0.5024,
+      "step": 2395
+    },
+    {
+      "epoch": 0.5344635288869061,
+      "grad_norm": 0.1579180806875229,
+      "learning_rate": 1.8576510907506192e-05,
+      "loss": 0.5168,
+      "step": 2396
+    },
+    {
+      "epoch": 0.5346865937987955,
+      "grad_norm": 0.1650589406490326,
+      "learning_rate": 1.8575300461267073e-05,
+      "loss": 0.5063,
+      "step": 2397
+    },
+    {
+      "epoch": 0.5349096587106849,
+      "grad_norm": 0.17452824115753174,
+      "learning_rate": 1.8574089540073868e-05,
+      "loss": 0.5272,
+      "step": 2398
+    },
+    {
+      "epoch": 0.5351327236225741,
+      "grad_norm": 0.16344332695007324,
+      "learning_rate": 1.8572878143993652e-05,
+      "loss": 0.4876,
+      "step": 2399
+    },
+    {
+      "epoch": 0.5353557885344635,
+      "grad_norm": 0.16399069130420685,
+      "learning_rate": 1.857166627309352e-05,
+      "loss": 0.4929,
+      "step": 2400
+    },
+    {
+      "epoch": 0.5355788534463529,
+      "grad_norm": 0.1590677946805954,
+      "learning_rate": 1.857045392744059e-05,
+      "loss": 0.4672,
+      "step": 2401
+    },
+    {
+      "epoch": 0.5358019183582422,
+      "grad_norm": 0.15974955260753632,
+      "learning_rate": 1.8569241107102014e-05,
+      "loss": 0.504,
+      "step": 2402
+    },
+    {
+      "epoch": 0.5360249832701316,
+      "grad_norm": 0.15898270905017853,
+      "learning_rate": 1.856802781214496e-05,
+      "loss": 0.5031,
+      "step": 2403
+    },
+    {
+      "epoch": 0.536248048182021,
+      "grad_norm": 0.15949919819831848,
+      "learning_rate": 1.856681404263663e-05,
+      "loss": 0.4957,
+      "step": 2404
+    },
+    {
+      "epoch": 0.5364711130939104,
+      "grad_norm": 0.1708550751209259,
+      "learning_rate": 1.8565599798644253e-05,
+      "loss": 0.5325,
+      "step": 2405
+    },
+    {
+      "epoch": 0.5366941780057997,
+      "grad_norm": 0.1631123423576355,
+      "learning_rate": 1.856438508023508e-05,
+      "loss": 0.5044,
+      "step": 2406
+    },
+    {
+      "epoch": 0.536917242917689,
+      "grad_norm": 0.17329680919647217,
+      "learning_rate": 1.8563169887476386e-05,
+      "loss": 0.4905,
+      "step": 2407
+    },
+    {
+      "epoch": 0.5371403078295784,
+      "grad_norm": 0.1555010825395584,
+      "learning_rate": 1.8561954220435483e-05,
+      "loss": 0.4896,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5373633727414677,
+      "grad_norm": 0.1617012768983841,
+      "learning_rate": 1.85607380791797e-05,
+      "loss": 0.5199,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5375864376533571,
+      "grad_norm": 0.16019612550735474,
+      "learning_rate": 1.8559521463776388e-05,
+      "loss": 0.5331,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5378095025652465,
+      "grad_norm": 0.15812602639198303,
+      "learning_rate": 1.855830437429294e-05,
+      "loss": 0.4946,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5380325674771358,
+      "grad_norm": 0.16577614843845367,
+      "learning_rate": 1.8557086810796756e-05,
+      "loss": 0.5049,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5382556323890252,
+      "grad_norm": 3.4729576110839844,
+      "learning_rate": 1.8555868773355283e-05,
+      "loss": 0.5572,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5384786973009146,
+      "grad_norm": 0.177010640501976,
+      "learning_rate": 1.8554650262035975e-05,
+      "loss": 0.4871,
+      "step": 2414
+    },
+    {
+      "epoch": 0.538701762212804,
+      "grad_norm": 0.16966521739959717,
+      "learning_rate": 1.8553431276906328e-05,
+      "loss": 0.5043,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5389248271246933,
+      "grad_norm": 0.16035056114196777,
+      "learning_rate": 1.855221181803385e-05,
+      "loss": 0.4951,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5391478920365826,
+      "grad_norm": 0.1640123724937439,
+      "learning_rate": 1.8550991885486093e-05,
+      "loss": 0.5175,
+      "step": 2417
+    },
+    {
+      "epoch": 0.539370956948472,
+      "grad_norm": 0.16451038420200348,
+      "learning_rate": 1.8549771479330612e-05,
+      "loss": 0.5192,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5395940218603613,
+      "grad_norm": 0.16037864983081818,
+      "learning_rate": 1.8548550599635007e-05,
+      "loss": 0.5174,
+      "step": 2419
+    },
+    {
+      "epoch": 0.5398170867722507,
+      "grad_norm": 0.1533394753932953,
+      "learning_rate": 1.85473292464669e-05,
+      "loss": 0.502,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5400401516841401,
+      "grad_norm": 0.16923432052135468,
+      "learning_rate": 1.854610741989393e-05,
+      "loss": 0.4917,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5402632165960295,
+      "grad_norm": 0.16241200268268585,
+      "learning_rate": 1.8544885119983774e-05,
+      "loss": 0.4922,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5404862815079188,
+      "grad_norm": 0.1572006493806839,
+      "learning_rate": 1.8543662346804138e-05,
+      "loss": 0.4771,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5407093464198082,
+      "grad_norm": 0.18193307518959045,
+      "learning_rate": 1.8542439100422733e-05,
+      "loss": 0.512,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5409324113316976,
+      "grad_norm": 0.1868167519569397,
+      "learning_rate": 1.8541215380907317e-05,
+      "loss": 0.5112,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5411554762435868,
+      "grad_norm": 0.1707518994808197,
+      "learning_rate": 1.8539991188325664e-05,
+      "loss": 0.4923,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5413785411554762,
+      "grad_norm": 0.16489239037036896,
+      "learning_rate": 1.8538766522745587e-05,
+      "loss": 0.4787,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5416016060673656,
+      "grad_norm": 0.16489636898040771,
+      "learning_rate": 1.8537541384234906e-05,
+      "loss": 0.4966,
+      "step": 2428
+    },
+    {
+      "epoch": 0.541824670979255,
+      "grad_norm": 0.17048819363117218,
+      "learning_rate": 1.853631577286148e-05,
+      "loss": 0.5079,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5420477358911443,
+      "grad_norm": 0.16793492436408997,
+      "learning_rate": 1.853508968869319e-05,
+      "loss": 0.5122,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5422708008030337,
+      "grad_norm": 0.16388513147830963,
+      "learning_rate": 1.8533863131797948e-05,
+      "loss": 0.4653,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5424938657149231,
+      "grad_norm": 0.16700832545757294,
+      "learning_rate": 1.853263610224368e-05,
+      "loss": 0.5208,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5427169306268124,
+      "grad_norm": 0.17072373628616333,
+      "learning_rate": 1.8531408600098356e-05,
+      "loss": 0.5032,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5429399955387018,
+      "grad_norm": 0.16046811640262604,
+      "learning_rate": 1.8530180625429958e-05,
+      "loss": 0.515,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5431630604505912,
+      "grad_norm": 0.16855685412883759,
+      "learning_rate": 1.8528952178306504e-05,
+      "loss": 0.4915,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5433861253624804,
+      "grad_norm": 0.17628896236419678,
+      "learning_rate": 1.8527723258796025e-05,
+      "loss": 0.5221,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5436091902743698,
+      "grad_norm": 0.1593201607465744,
+      "learning_rate": 1.852649386696659e-05,
+      "loss": 0.5027,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5438322551862592,
+      "grad_norm": 0.18933804333209991,
+      "learning_rate": 1.852526400288629e-05,
+      "loss": 0.5187,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5440553200981486,
+      "grad_norm": 0.16613252460956573,
+      "learning_rate": 1.852403366662325e-05,
+      "loss": 0.4939,
+      "step": 2439
+    },
+    {
+      "epoch": 0.5442783850100379,
+      "grad_norm": 0.16360655426979065,
+      "learning_rate": 1.85228028582456e-05,
+      "loss": 0.4835,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5445014499219273,
+      "grad_norm": 0.1624102145433426,
+      "learning_rate": 1.8521571577821522e-05,
+      "loss": 0.5308,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5447245148338167,
+      "grad_norm": 0.16573922336101532,
+      "learning_rate": 1.8520339825419204e-05,
+      "loss": 0.5148,
+      "step": 2442
+    },
+    {
+      "epoch": 0.544947579745706,
+      "grad_norm": 0.17061501741409302,
+      "learning_rate": 1.8519107601106875e-05,
+      "loss": 0.5025,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5451706446575953,
+      "grad_norm": 0.16382652521133423,
+      "learning_rate": 1.851787490495278e-05,
+      "loss": 0.5214,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5453937095694847,
+      "grad_norm": 0.16731955111026764,
+      "learning_rate": 1.8516641737025187e-05,
+      "loss": 0.4915,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5456167744813741,
+      "grad_norm": 0.1665150374174118,
+      "learning_rate": 1.8515408097392408e-05,
+      "loss": 0.4585,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5458398393932634,
+      "grad_norm": 0.16348014771938324,
+      "learning_rate": 1.851417398612276e-05,
+      "loss": 0.5042,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5460629043051528,
+      "grad_norm": 0.1641550064086914,
+      "learning_rate": 1.85129394032846e-05,
+      "loss": 0.5103,
+      "step": 2448
+    },
+    {
+      "epoch": 0.5462859692170422,
+      "grad_norm": 0.16449302434921265,
+      "learning_rate": 1.8511704348946314e-05,
+      "loss": 0.5043,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5465090341289315,
+      "grad_norm": 0.1623837649822235,
+      "learning_rate": 1.85104688231763e-05,
+      "loss": 0.4831,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5467320990408209,
+      "grad_norm": 0.16712717711925507,
+      "learning_rate": 1.8509232826042983e-05,
+      "loss": 0.5146,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5469551639527103,
+      "grad_norm": 0.14933447539806366,
+      "learning_rate": 1.850799635761483e-05,
+      "loss": 0.5011,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5471782288645995,
+      "grad_norm": 0.17392055690288544,
+      "learning_rate": 1.8506759417960322e-05,
+      "loss": 0.482,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5474012937764889,
+      "grad_norm": 0.1616058647632599,
+      "learning_rate": 1.850552200714797e-05,
+      "loss": 0.5152,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5476243586883783,
+      "grad_norm": 0.16296276450157166,
+      "learning_rate": 1.8504284125246304e-05,
+      "loss": 0.5073,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5478474236002677,
+      "grad_norm": 0.16247029602527618,
+      "learning_rate": 1.850304577232389e-05,
+      "loss": 0.5194,
+      "step": 2456
+    },
+    {
+      "epoch": 0.548070488512157,
+      "grad_norm": 0.16784507036209106,
+      "learning_rate": 1.8501806948449316e-05,
+      "loss": 0.5212,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5482935534240464,
+      "grad_norm": 0.1538800299167633,
+      "learning_rate": 1.8500567653691192e-05,
+      "loss": 0.4907,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5485166183359358,
+      "grad_norm": 0.1586543768644333,
+      "learning_rate": 1.8499327888118163e-05,
+      "loss": 0.4873,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5487396832478251,
+      "grad_norm": 0.15387628972530365,
+      "learning_rate": 1.8498087651798893e-05,
+      "loss": 0.5102,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5489627481597145,
+      "grad_norm": 0.15858915448188782,
+      "learning_rate": 1.8496846944802072e-05,
+      "loss": 0.4983,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5491858130716039,
+      "grad_norm": 0.15983784198760986,
+      "learning_rate": 1.849560576719642e-05,
+      "loss": 0.4728,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5494088779834932,
+      "grad_norm": 0.1605166792869568,
+      "learning_rate": 1.849436411905068e-05,
+      "loss": 0.5083,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5496319428953825,
+      "grad_norm": 0.15921124815940857,
+      "learning_rate": 1.8493122000433628e-05,
+      "loss": 0.5054,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5498550078072719,
+      "grad_norm": 0.15854284167289734,
+      "learning_rate": 1.849187941141405e-05,
+      "loss": 0.5111,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5500780727191613,
+      "grad_norm": 0.16846506297588348,
+      "learning_rate": 1.8490636352060778e-05,
+      "loss": 0.526,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5503011376310506,
+      "grad_norm": 0.1619185209274292,
+      "learning_rate": 1.8489392822442657e-05,
+      "loss": 0.4965,
+      "step": 2467
+    },
+    {
+      "epoch": 0.55052420254294,
+      "grad_norm": 0.17305007576942444,
+      "learning_rate": 1.8488148822628557e-05,
+      "loss": 0.5171,
+      "step": 2468
+    },
+    {
+      "epoch": 0.5507472674548294,
+      "grad_norm": 0.16538777947425842,
+      "learning_rate": 1.8486904352687384e-05,
+      "loss": 0.5187,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5509703323667187,
+      "grad_norm": 0.1914980560541153,
+      "learning_rate": 1.8485659412688065e-05,
+      "loss": 0.5105,
+      "step": 2470
+    },
+    {
+      "epoch": 0.551193397278608,
+      "grad_norm": 0.16565662622451782,
+      "learning_rate": 1.8484414002699552e-05,
+      "loss": 0.4949,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5514164621904974,
+      "grad_norm": 0.17171315848827362,
+      "learning_rate": 1.848316812279082e-05,
+      "loss": 0.507,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5516395271023868,
+      "grad_norm": 0.1598774790763855,
+      "learning_rate": 1.8481921773030878e-05,
+      "loss": 0.5101,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5518625920142761,
+      "grad_norm": 0.16791561245918274,
+      "learning_rate": 1.8480674953488752e-05,
+      "loss": 0.4902,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5520856569261655,
+      "grad_norm": 0.17411862313747406,
+      "learning_rate": 1.8479427664233505e-05,
+      "loss": 0.5017,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5523087218380549,
+      "grad_norm": 0.16108182072639465,
+      "learning_rate": 1.8478179905334213e-05,
+      "loss": 0.4886,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5525317867499442,
+      "grad_norm": 0.1858782172203064,
+      "learning_rate": 1.847693167685999e-05,
+      "loss": 0.5047,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5527548516618336,
+      "grad_norm": 0.17761607468128204,
+      "learning_rate": 1.847568297887997e-05,
+      "loss": 0.5233,
+      "step": 2478
+    },
+    {
+      "epoch": 0.552977916573723,
+      "grad_norm": 0.1658019870519638,
+      "learning_rate": 1.8474433811463307e-05,
+      "loss": 0.5263,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5532009814856124,
+      "grad_norm": 1.4999321699142456,
+      "learning_rate": 1.84731841746792e-05,
+      "loss": 0.5359,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5534240463975016,
+      "grad_norm": 0.17992182075977325,
+      "learning_rate": 1.847193406859685e-05,
+      "loss": 0.5033,
+      "step": 2481
+    },
+    {
+      "epoch": 0.553647111309391,
+      "grad_norm": 0.17310257256031036,
+      "learning_rate": 1.8470683493285503e-05,
+      "loss": 0.4905,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5538701762212804,
+      "grad_norm": 0.16551616787910461,
+      "learning_rate": 1.846943244881442e-05,
+      "loss": 0.4925,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5540932411331697,
+      "grad_norm": 0.17117410898208618,
+      "learning_rate": 1.846818093525289e-05,
+      "loss": 0.5128,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5543163060450591,
+      "grad_norm": 0.15640629827976227,
+      "learning_rate": 1.8466928952670242e-05,
+      "loss": 0.5054,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5545393709569485,
+      "grad_norm": 0.16013699769973755,
+      "learning_rate": 1.8465676501135804e-05,
+      "loss": 0.4966,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5547624358688378,
+      "grad_norm": 0.16586218774318695,
+      "learning_rate": 1.846442358071895e-05,
+      "loss": 0.506,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5549855007807272,
+      "grad_norm": 0.1749541014432907,
+      "learning_rate": 1.8463170191489075e-05,
+      "loss": 0.4733,
+      "step": 2488
+    },
+    {
+      "epoch": 0.5552085656926166,
+      "grad_norm": 0.16143982112407684,
+      "learning_rate": 1.84619163335156e-05,
+      "loss": 0.5014,
+      "step": 2489
+    },
+    {
+      "epoch": 0.555431630604506,
+      "grad_norm": 0.1761699765920639,
+      "learning_rate": 1.846066200686797e-05,
+      "loss": 0.5305,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5556546955163952,
+      "grad_norm": 0.25962182879447937,
+      "learning_rate": 1.8459407211615658e-05,
+      "loss": 0.5047,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5558777604282846,
+      "grad_norm": 0.15808863937854767,
+      "learning_rate": 1.8458151947828165e-05,
+      "loss": 0.5079,
+      "step": 2492
+    },
+    {
+      "epoch": 0.556100825340174,
+      "grad_norm": 0.1679726392030716,
+      "learning_rate": 1.8456896215575013e-05,
+      "loss": 0.5097,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5563238902520633,
+      "grad_norm": 0.15282638370990753,
+      "learning_rate": 1.845564001492575e-05,
+      "loss": 0.5196,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5565469551639527,
+      "grad_norm": 0.1774146854877472,
+      "learning_rate": 1.8454383345949954e-05,
+      "loss": 0.5227,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5567700200758421,
+      "grad_norm": 0.1605686992406845,
+      "learning_rate": 1.8453126208717235e-05,
+      "loss": 0.4751,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5569930849877315,
+      "grad_norm": 0.16172072291374207,
+      "learning_rate": 1.845186860329721e-05,
+      "loss": 0.5158,
+      "step": 2497
+    },
+    {
+      "epoch": 0.5572161498996208,
+      "grad_norm": 0.16925646364688873,
+      "learning_rate": 1.8450610529759535e-05,
+      "loss": 0.5148,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5574392148115102,
+      "grad_norm": 0.16925497353076935,
+      "learning_rate": 1.8449351988173894e-05,
+      "loss": 0.4947,
+      "step": 2499
+    },
+    {
+      "epoch": 0.5576622797233995,
+      "grad_norm": 1.354637861251831,
+      "learning_rate": 1.8448092978609993e-05,
+      "loss": 0.5136,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5578853446352888,
+      "grad_norm": 0.16736604273319244,
+      "learning_rate": 1.844683350113756e-05,
+      "loss": 0.5389,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5581084095471782,
+      "grad_norm": 0.16498203575611115,
+      "learning_rate": 1.8445573555826355e-05,
+      "loss": 0.5018,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5583314744590676,
+      "grad_norm": 0.17059291899204254,
+      "learning_rate": 1.8444313142746164e-05,
+      "loss": 0.5282,
+      "step": 2503
+    },
+    {
+      "epoch": 0.558554539370957,
+      "grad_norm": 0.21622738242149353,
+      "learning_rate": 1.844305226196679e-05,
+      "loss": 0.4923,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5587776042828463,
+      "grad_norm": 0.16687920689582825,
+      "learning_rate": 1.844179091355808e-05,
+      "loss": 0.4998,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5590006691947357,
+      "grad_norm": 0.1704476922750473,
+      "learning_rate": 1.8440529097589885e-05,
+      "loss": 0.5267,
+      "step": 2506
+    },
+    {
+      "epoch": 0.5592237341066251,
+      "grad_norm": 0.16384254395961761,
+      "learning_rate": 1.8439266814132092e-05,
+      "loss": 0.5053,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5594467990185144,
+      "grad_norm": 0.16077467799186707,
+      "learning_rate": 1.843800406325462e-05,
+      "loss": 0.5151,
+      "step": 2508
+    },
+    {
+      "epoch": 0.5596698639304037,
+      "grad_norm": 0.16944189369678497,
+      "learning_rate": 1.843674084502741e-05,
+      "loss": 0.5054,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5598929288422931,
+      "grad_norm": 0.16107740998268127,
+      "learning_rate": 1.8435477159520418e-05,
+      "loss": 0.4861,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5601159937541824,
+      "grad_norm": 0.17457795143127441,
+      "learning_rate": 1.843421300680364e-05,
+      "loss": 0.486,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5603390586660718,
+      "grad_norm": 0.16926448047161102,
+      "learning_rate": 1.8432948386947092e-05,
+      "loss": 0.5238,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5605621235779612,
+      "grad_norm": 0.16914579272270203,
+      "learning_rate": 1.8431683300020817e-05,
+      "loss": 0.5021,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5607851884898506,
+      "grad_norm": 0.16727939248085022,
+      "learning_rate": 1.8430417746094886e-05,
+      "loss": 0.53,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5610082534017399,
+      "grad_norm": 0.1788937747478485,
+      "learning_rate": 1.842915172523939e-05,
+      "loss": 0.5226,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5612313183136293,
+      "grad_norm": 0.17947039008140564,
+      "learning_rate": 1.8427885237524446e-05,
+      "loss": 0.4914,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5614543832255187,
+      "grad_norm": 0.17127928137779236,
+      "learning_rate": 1.842661828302021e-05,
+      "loss": 0.4802,
+      "step": 2517
+    },
+    {
+      "epoch": 0.5616774481374079,
+      "grad_norm": 0.20038791000843048,
+      "learning_rate": 1.8425350861796845e-05,
+      "loss": 0.4911,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5619005130492973,
+      "grad_norm": 0.16238313913345337,
+      "learning_rate": 1.842408297392455e-05,
+      "loss": 0.492,
+      "step": 2519
+    },
+    {
+      "epoch": 0.5621235779611867,
+      "grad_norm": 0.17273001372814178,
+      "learning_rate": 1.8422814619473556e-05,
+      "loss": 0.555,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5623466428730761,
+      "grad_norm": 0.19861721992492676,
+      "learning_rate": 1.84215457985141e-05,
+      "loss": 0.5245,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5625697077849654,
+      "grad_norm": 0.15922409296035767,
+      "learning_rate": 1.8420276511116467e-05,
+      "loss": 0.4968,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5627927726968548,
+      "grad_norm": 0.15657752752304077,
+      "learning_rate": 1.8419006757350956e-05,
+      "loss": 0.4923,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5630158376087442,
+      "grad_norm": 0.1599670946598053,
+      "learning_rate": 1.8417736537287893e-05,
+      "loss": 0.5381,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5632389025206335,
+      "grad_norm": 0.16177695989608765,
+      "learning_rate": 1.841646585099763e-05,
+      "loss": 0.5025,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5634619674325229,
+      "grad_norm": 0.17016616463661194,
+      "learning_rate": 1.8415194698550548e-05,
+      "loss": 0.5024,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5636850323444123,
+      "grad_norm": 0.15452872216701508,
+      "learning_rate": 1.8413923080017047e-05,
+      "loss": 0.4838,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5639080972563015,
+      "grad_norm": 0.17372727394104004,
+      "learning_rate": 1.8412650995467564e-05,
+      "loss": 0.5272,
+      "step": 2528
+    },
+    {
+      "epoch": 0.5641311621681909,
+      "grad_norm": 0.20483483374118805,
+      "learning_rate": 1.8411378444972548e-05,
+      "loss": 0.4984,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5643542270800803,
+      "grad_norm": 0.16677002608776093,
+      "learning_rate": 1.8410105428602485e-05,
+      "loss": 0.5217,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5645772919919697,
+      "grad_norm": 0.16841678321361542,
+      "learning_rate": 1.840883194642788e-05,
+      "loss": 0.5331,
+      "step": 2531
+    },
+    {
+      "epoch": 0.564800356903859,
+      "grad_norm": 0.16345760226249695,
+      "learning_rate": 1.8407557998519273e-05,
+      "loss": 0.5372,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5650234218157484,
+      "grad_norm": 0.16982464492321014,
+      "learning_rate": 1.840628358494721e-05,
+      "loss": 0.4795,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5652464867276378,
+      "grad_norm": 0.16752153635025024,
+      "learning_rate": 1.840500870578229e-05,
+      "loss": 0.5088,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5654695516395271,
+      "grad_norm": 0.16022369265556335,
+      "learning_rate": 1.840373336109512e-05,
+      "loss": 0.4676,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5656926165514164,
+      "grad_norm": 0.15749748051166534,
+      "learning_rate": 1.8402457550956336e-05,
+      "loss": 0.5325,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5659156814633058,
+      "grad_norm": 0.16572564840316772,
+      "learning_rate": 1.8401181275436596e-05,
+      "loss": 0.4972,
+      "step": 2537
+    },
+    {
+      "epoch": 0.5661387463751952,
+      "grad_norm": 0.15949216485023499,
+      "learning_rate": 1.839990453460659e-05,
+      "loss": 0.5088,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5663618112870845,
+      "grad_norm": 0.15577349066734314,
+      "learning_rate": 1.8398627328537037e-05,
+      "loss": 0.4871,
+      "step": 2539
+    },
+    {
+      "epoch": 0.5665848761989739,
+      "grad_norm": 0.22663244605064392,
+      "learning_rate": 1.839734965729867e-05,
+      "loss": 0.51,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5668079411108633,
+      "grad_norm": 0.16689926385879517,
+      "learning_rate": 1.8396071520962256e-05,
+      "loss": 0.4998,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5670310060227526,
+      "grad_norm": 0.16619545221328735,
+      "learning_rate": 1.8394792919598592e-05,
+      "loss": 0.5208,
+      "step": 2542
+    },
+    {
+      "epoch": 0.567254070934642,
+      "grad_norm": 0.15518589317798615,
+      "learning_rate": 1.8393513853278492e-05,
+      "loss": 0.4967,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5674771358465314,
+      "grad_norm": 0.19833387434482574,
+      "learning_rate": 1.8392234322072792e-05,
+      "loss": 0.5029,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5677002007584206,
+      "grad_norm": 0.15093770623207092,
+      "learning_rate": 1.839095432605237e-05,
+      "loss": 0.5017,
+      "step": 2545
+    },
+    {
+      "epoch": 0.56792326567031,
+      "grad_norm": 0.19351163506507874,
+      "learning_rate": 1.8389673865288114e-05,
+      "loss": 0.494,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5681463305821994,
+      "grad_norm": 0.16114504635334015,
+      "learning_rate": 1.8388392939850946e-05,
+      "loss": 0.5118,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5683693954940888,
+      "grad_norm": 0.16146999597549438,
+      "learning_rate": 1.8387111549811812e-05,
+      "loss": 0.4732,
+      "step": 2548
+    },
+    {
+      "epoch": 0.5685924604059781,
+      "grad_norm": 0.1651589721441269,
+      "learning_rate": 1.8385829695241687e-05,
+      "loss": 0.5086,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5688155253178675,
+      "grad_norm": 0.15596236288547516,
+      "learning_rate": 1.838454737621156e-05,
+      "loss": 0.5017,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5690385902297569,
+      "grad_norm": 0.1615191251039505,
+      "learning_rate": 1.838326459279246e-05,
+      "loss": 0.5212,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5692616551416462,
+      "grad_norm": 0.16675913333892822,
+      "learning_rate": 1.8381981345055435e-05,
+      "loss": 0.5229,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5694847200535356,
+      "grad_norm": 0.16236376762390137,
+      "learning_rate": 1.8380697633071558e-05,
+      "loss": 0.4955,
+      "step": 2553
+    },
+    {
+      "epoch": 0.569707784965425,
+      "grad_norm": 0.17616188526153564,
+      "learning_rate": 1.837941345691193e-05,
+      "loss": 0.5067,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5699308498773143,
+      "grad_norm": 0.16186662018299103,
+      "learning_rate": 1.8378128816647676e-05,
+      "loss": 0.5054,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5701539147892036,
+      "grad_norm": 0.1674317866563797,
+      "learning_rate": 1.8376843712349946e-05,
+      "loss": 0.5009,
+      "step": 2556
+    },
+    {
+      "epoch": 0.570376979701093,
+      "grad_norm": 0.17086781561374664,
+      "learning_rate": 1.837555814408992e-05,
+      "loss": 0.4998,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5706000446129824,
+      "grad_norm": 0.1624806523323059,
+      "learning_rate": 1.8374272111938797e-05,
+      "loss": 0.519,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5708231095248717,
+      "grad_norm": 0.1701088398694992,
+      "learning_rate": 1.837298561596781e-05,
+      "loss": 0.5173,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5710461744367611,
+      "grad_norm": 0.15940968692302704,
+      "learning_rate": 1.8371698656248212e-05,
+      "loss": 0.515,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5712692393486505,
+      "grad_norm": 0.17150092124938965,
+      "learning_rate": 1.837041123285128e-05,
+      "loss": 0.4826,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5714923042605398,
+      "grad_norm": 0.16845233738422394,
+      "learning_rate": 1.836912334584833e-05,
+      "loss": 0.5018,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5717153691724292,
+      "grad_norm": 0.18033871054649353,
+      "learning_rate": 1.8367834995310676e-05,
+      "loss": 0.5177,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5719384340843185,
+      "grad_norm": 0.17002245783805847,
+      "learning_rate": 1.8366546181309686e-05,
+      "loss": 0.5287,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5721614989962079,
+      "grad_norm": 0.16821357607841492,
+      "learning_rate": 1.836525690391674e-05,
+      "loss": 0.522,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5723845639080972,
+      "grad_norm": 0.1578892022371292,
+      "learning_rate": 1.836396716320325e-05,
+      "loss": 0.4929,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5726076288199866,
+      "grad_norm": 0.15691736340522766,
+      "learning_rate": 1.836267695924065e-05,
+      "loss": 0.4775,
+      "step": 2567
+    },
+    {
+      "epoch": 0.572830693731876,
+      "grad_norm": 0.16604529321193695,
+      "learning_rate": 1.8361386292100394e-05,
+      "loss": 0.4854,
+      "step": 2568
+    },
+    {
+      "epoch": 0.5730537586437653,
+      "grad_norm": 0.16472390294075012,
+      "learning_rate": 1.8360095161853966e-05,
+      "loss": 0.486,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5732768235556547,
+      "grad_norm": 0.1708393096923828,
+      "learning_rate": 1.8358803568572885e-05,
+      "loss": 0.5086,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5734998884675441,
+      "grad_norm": 0.1598411649465561,
+      "learning_rate": 1.8357511512328683e-05,
+      "loss": 0.5071,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5737229533794335,
+      "grad_norm": 0.19416102766990662,
+      "learning_rate": 1.8356218993192922e-05,
+      "loss": 0.5092,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5739460182913227,
+      "grad_norm": 0.1696127951145172,
+      "learning_rate": 1.835492601123719e-05,
+      "loss": 0.5157,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5741690832032121,
+      "grad_norm": 0.16586486995220184,
+      "learning_rate": 1.8353632566533102e-05,
+      "loss": 0.4948,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5743921481151015,
+      "grad_norm": 0.17883430421352386,
+      "learning_rate": 1.8352338659152296e-05,
+      "loss": 0.4989,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5746152130269908,
+      "grad_norm": 0.15909375250339508,
+      "learning_rate": 1.8351044289166435e-05,
+      "loss": 0.4892,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5748382779388802,
+      "grad_norm": 0.17203538119792938,
+      "learning_rate": 1.834974945664721e-05,
+      "loss": 0.5039,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5750613428507696,
+      "grad_norm": 0.17230457067489624,
+      "learning_rate": 1.834845416166634e-05,
+      "loss": 0.5102,
+      "step": 2578
+    },
+    {
+      "epoch": 0.575284407762659,
+      "grad_norm": 0.16863304376602173,
+      "learning_rate": 1.8347158404295566e-05,
+      "loss": 0.5365,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5755074726745483,
+      "grad_norm": 0.15985195338726044,
+      "learning_rate": 1.8345862184606653e-05,
+      "loss": 0.4991,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5757305375864377,
+      "grad_norm": 0.16000472009181976,
+      "learning_rate": 1.8344565502671396e-05,
+      "loss": 0.4898,
+      "step": 2581
+    },
+    {
+      "epoch": 0.575953602498327,
+      "grad_norm": 0.16451792418956757,
+      "learning_rate": 1.8343268358561607e-05,
+      "loss": 0.5382,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5761766674102163,
+      "grad_norm": 0.16820542514324188,
+      "learning_rate": 1.834197075234914e-05,
+      "loss": 0.4993,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5763997323221057,
+      "grad_norm": 0.16512157022953033,
+      "learning_rate": 1.834067268410586e-05,
+      "loss": 0.5139,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5766227972339951,
+      "grad_norm": 0.16488297283649445,
+      "learning_rate": 1.833937415390366e-05,
+      "loss": 0.5371,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5768458621458844,
+      "grad_norm": 0.1621064841747284,
+      "learning_rate": 1.8338075161814462e-05,
+      "loss": 0.5218,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5770689270577738,
+      "grad_norm": 0.1790938526391983,
+      "learning_rate": 1.8336775707910214e-05,
+      "loss": 0.5357,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5772919919696632,
+      "grad_norm": 0.16097491979599,
+      "learning_rate": 1.8335475792262888e-05,
+      "loss": 0.4874,
+      "step": 2588
+    },
+    {
+      "epoch": 0.5775150568815526,
+      "grad_norm": 0.17577522993087769,
+      "learning_rate": 1.8334175414944476e-05,
+      "loss": 0.5097,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5777381217934419,
+      "grad_norm": 0.17199388146400452,
+      "learning_rate": 1.833287457602701e-05,
+      "loss": 0.4969,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5779611867053313,
+      "grad_norm": 0.16499276459217072,
+      "learning_rate": 1.833157327558253e-05,
+      "loss": 0.5207,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5781842516172206,
+      "grad_norm": 0.16708426177501678,
+      "learning_rate": 1.8330271513683118e-05,
+      "loss": 0.5077,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5784073165291099,
+      "grad_norm": 0.1643514633178711,
+      "learning_rate": 1.8328969290400867e-05,
+      "loss": 0.4884,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5786303814409993,
+      "grad_norm": 0.1548931747674942,
+      "learning_rate": 1.832766660580791e-05,
+      "loss": 0.5047,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5788534463528887,
+      "grad_norm": 0.16089920699596405,
+      "learning_rate": 1.832636345997639e-05,
+      "loss": 0.4916,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5790765112647781,
+      "grad_norm": 0.15845981240272522,
+      "learning_rate": 1.8325059852978485e-05,
+      "loss": 0.4832,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5792995761766674,
+      "grad_norm": 0.1613757461309433,
+      "learning_rate": 1.83237557848864e-05,
+      "loss": 0.5192,
+      "step": 2597
+    },
+    {
+      "epoch": 0.5795226410885568,
+      "grad_norm": 0.15937237441539764,
+      "learning_rate": 1.8322451255772365e-05,
+      "loss": 0.5028,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5797457060004462,
+      "grad_norm": 0.16132612526416779,
+      "learning_rate": 1.8321146265708627e-05,
+      "loss": 0.4948,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5799687709123355,
+      "grad_norm": 0.18276172876358032,
+      "learning_rate": 1.8319840814767463e-05,
+      "loss": 0.4845,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5801918358242248,
+      "grad_norm": 0.1661817729473114,
+      "learning_rate": 1.8318534903021182e-05,
+      "loss": 0.5344,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5804149007361142,
+      "grad_norm": 0.15862146019935608,
+      "learning_rate": 1.8317228530542117e-05,
+      "loss": 0.5043,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5806379656480035,
+      "grad_norm": 0.15994656085968018,
+      "learning_rate": 1.8315921697402618e-05,
+      "loss": 0.4765,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5808610305598929,
+      "grad_norm": 0.16988401114940643,
+      "learning_rate": 1.8314614403675063e-05,
+      "loss": 0.5194,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5810840954717823,
+      "grad_norm": 0.16828812658786774,
+      "learning_rate": 1.831330664943186e-05,
+      "loss": 0.5357,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5813071603836717,
+      "grad_norm": 0.19693851470947266,
+      "learning_rate": 1.8311998434745445e-05,
+      "loss": 0.4976,
+      "step": 2606
+    },
+    {
+      "epoch": 0.581530225295561,
+      "grad_norm": 0.1689445823431015,
+      "learning_rate": 1.831068975968827e-05,
+      "loss": 0.4992,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5817532902074504,
+      "grad_norm": 0.15751276910305023,
+      "learning_rate": 1.830938062433282e-05,
+      "loss": 0.4711,
+      "step": 2608
+    },
+    {
+      "epoch": 0.5819763551193398,
+      "grad_norm": 0.16275669634342194,
+      "learning_rate": 1.8308071028751608e-05,
+      "loss": 0.5184,
+      "step": 2609
+    },
+    {
+      "epoch": 0.582199420031229,
+      "grad_norm": 0.16536571085453033,
+      "learning_rate": 1.8306760973017158e-05,
+      "loss": 0.5172,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5824224849431184,
+      "grad_norm": 0.18836647272109985,
+      "learning_rate": 1.830545045720203e-05,
+      "loss": 0.5019,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5826455498550078,
+      "grad_norm": 0.16014382243156433,
+      "learning_rate": 1.830413948137882e-05,
+      "loss": 0.5182,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5828686147668972,
+      "grad_norm": 0.1588478833436966,
+      "learning_rate": 1.8302828045620128e-05,
+      "loss": 0.5072,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5830916796787865,
+      "grad_norm": 0.17403388023376465,
+      "learning_rate": 1.830151614999859e-05,
+      "loss": 0.5093,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5833147445906759,
+      "grad_norm": 0.17456629872322083,
+      "learning_rate": 1.830020379458687e-05,
+      "loss": 0.4922,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5835378095025653,
+      "grad_norm": 0.15623332560062408,
+      "learning_rate": 1.829889097945765e-05,
+      "loss": 0.5133,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5837608744144546,
+      "grad_norm": 0.18376080691814423,
+      "learning_rate": 1.8297577704683653e-05,
+      "loss": 0.5154,
+      "step": 2617
+    },
+    {
+      "epoch": 0.583983939326344,
+      "grad_norm": 0.17060859501361847,
+      "learning_rate": 1.8296263970337602e-05,
+      "loss": 0.5058,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5842070042382334,
+      "grad_norm": 0.1654299646615982,
+      "learning_rate": 1.829494977649227e-05,
+      "loss": 0.5072,
+      "step": 2619
+    },
+    {
+      "epoch": 0.5844300691501226,
+      "grad_norm": 0.1547902524471283,
+      "learning_rate": 1.829363512322044e-05,
+      "loss": 0.4967,
+      "step": 2620
+    },
+    {
+      "epoch": 0.584653134062012,
+      "grad_norm": 0.1600826382637024,
+      "learning_rate": 1.829232001059493e-05,
+      "loss": 0.5396,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5848761989739014,
+      "grad_norm": 0.16009269654750824,
+      "learning_rate": 1.8291004438688578e-05,
+      "loss": 0.525,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5850992638857908,
+      "grad_norm": 0.2055627703666687,
+      "learning_rate": 1.8289688407574246e-05,
+      "loss": 0.4993,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5853223287976801,
+      "grad_norm": 0.1680591106414795,
+      "learning_rate": 1.8288371917324827e-05,
+      "loss": 0.5184,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5855453937095695,
+      "grad_norm": 0.1637965738773346,
+      "learning_rate": 1.828705496801323e-05,
+      "loss": 0.5205,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5857684586214589,
+      "grad_norm": 0.172703817486763,
+      "learning_rate": 1.828573755971241e-05,
+      "loss": 0.5201,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5859915235333482,
+      "grad_norm": 0.18580362200737,
+      "learning_rate": 1.8284419692495316e-05,
+      "loss": 0.5246,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5862145884452375,
+      "grad_norm": 0.16274447739124298,
+      "learning_rate": 1.8283101366434954e-05,
+      "loss": 0.5172,
+      "step": 2628
+    },
+    {
+      "epoch": 0.5864376533571269,
+      "grad_norm": 0.16170893609523773,
+      "learning_rate": 1.8281782581604334e-05,
+      "loss": 0.506,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5866607182690163,
+      "grad_norm": 0.15339437127113342,
+      "learning_rate": 1.82804633380765e-05,
+      "loss": 0.4706,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5868837831809056,
+      "grad_norm": 0.17161330580711365,
+      "learning_rate": 1.827914363592452e-05,
+      "loss": 0.5217,
+      "step": 2631
+    },
+    {
+      "epoch": 0.587106848092795,
+      "grad_norm": 0.16379369795322418,
+      "learning_rate": 1.8277823475221485e-05,
+      "loss": 0.4954,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5873299130046844,
+      "grad_norm": 0.16465184092521667,
+      "learning_rate": 1.827650285604052e-05,
+      "loss": 0.4995,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5875529779165737,
+      "grad_norm": 0.1681264042854309,
+      "learning_rate": 1.8275181778454767e-05,
+      "loss": 0.5391,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5877760428284631,
+      "grad_norm": 0.17729119956493378,
+      "learning_rate": 1.827386024253739e-05,
+      "loss": 0.5313,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5879991077403525,
+      "grad_norm": 0.1703905314207077,
+      "learning_rate": 1.8272538248361592e-05,
+      "loss": 0.5407,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5882221726522417,
+      "grad_norm": 0.1546456664800644,
+      "learning_rate": 1.8271215796000588e-05,
+      "loss": 0.4965,
+      "step": 2637
+    },
+    {
+      "epoch": 0.5884452375641311,
+      "grad_norm": 0.15894514322280884,
+      "learning_rate": 1.8269892885527624e-05,
+      "loss": 0.498,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5886683024760205,
+      "grad_norm": 0.15367206931114197,
+      "learning_rate": 1.826856951701597e-05,
+      "loss": 0.482,
+      "step": 2639
+    },
+    {
+      "epoch": 0.5888913673879099,
+      "grad_norm": 0.16719990968704224,
+      "learning_rate": 1.826724569053893e-05,
+      "loss": 0.5287,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5891144322997992,
+      "grad_norm": 0.1742510050535202,
+      "learning_rate": 1.8265921406169816e-05,
+      "loss": 0.4773,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5893374972116886,
+      "grad_norm": 0.16305537521839142,
+      "learning_rate": 1.8264596663981985e-05,
+      "loss": 0.47,
+      "step": 2642
+    },
+    {
+      "epoch": 0.589560562123578,
+      "grad_norm": 0.1746242344379425,
+      "learning_rate": 1.82632714640488e-05,
+      "loss": 0.4978,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5897836270354673,
+      "grad_norm": 0.16970300674438477,
+      "learning_rate": 1.8261945806443666e-05,
+      "loss": 0.5225,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5900066919473567,
+      "grad_norm": 0.16862879693508148,
+      "learning_rate": 1.826061969124e-05,
+      "loss": 0.5299,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5902297568592461,
+      "grad_norm": 0.1616392582654953,
+      "learning_rate": 1.825929311851126e-05,
+      "loss": 0.5237,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5904528217711354,
+      "grad_norm": 0.15673169493675232,
+      "learning_rate": 1.8257966088330907e-05,
+      "loss": 0.5005,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5906758866830247,
+      "grad_norm": 0.16055789589881897,
+      "learning_rate": 1.825663860077245e-05,
+      "loss": 0.5179,
+      "step": 2648
+    },
+    {
+      "epoch": 0.5908989515949141,
+      "grad_norm": 0.1535293161869049,
+      "learning_rate": 1.8255310655909414e-05,
+      "loss": 0.4926,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5911220165068035,
+      "grad_norm": 0.16450564563274384,
+      "learning_rate": 1.8253982253815343e-05,
+      "loss": 0.5016,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5913450814186928,
+      "grad_norm": 0.1581798493862152,
+      "learning_rate": 1.8252653394563814e-05,
+      "loss": 0.5031,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5915681463305822,
+      "grad_norm": 0.16955383121967316,
+      "learning_rate": 1.825132407822843e-05,
+      "loss": 0.5408,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5917912112424716,
+      "grad_norm": 0.16490530967712402,
+      "learning_rate": 1.8249994304882818e-05,
+      "loss": 0.5352,
+      "step": 2653
+    },
+    {
+      "epoch": 0.592014276154361,
+      "grad_norm": 0.1716136932373047,
+      "learning_rate": 1.8248664074600626e-05,
+      "loss": 0.4936,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5922373410662503,
+      "grad_norm": 0.16612227261066437,
+      "learning_rate": 1.8247333387455534e-05,
+      "loss": 0.4886,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5924604059781396,
+      "grad_norm": 0.16571937501430511,
+      "learning_rate": 1.8246002243521234e-05,
+      "loss": 0.5171,
+      "step": 2656
+    },
+    {
+      "epoch": 0.592683470890029,
+      "grad_norm": 0.15452441573143005,
+      "learning_rate": 1.8244670642871464e-05,
+      "loss": 0.4825,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5929065358019183,
+      "grad_norm": 0.16785788536071777,
+      "learning_rate": 1.8243338585579974e-05,
+      "loss": 0.4993,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5931296007138077,
+      "grad_norm": 0.16040126979351044,
+      "learning_rate": 1.824200607172054e-05,
+      "loss": 0.498,
+      "step": 2659
+    },
+    {
+      "epoch": 0.5933526656256971,
+      "grad_norm": 0.16780081391334534,
+      "learning_rate": 1.8240673101366963e-05,
+      "loss": 0.509,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5935757305375864,
+      "grad_norm": 0.16159792244434357,
+      "learning_rate": 1.823933967459308e-05,
+      "loss": 0.5054,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5937987954494758,
+      "grad_norm": 0.1585383266210556,
+      "learning_rate": 1.823800579147273e-05,
+      "loss": 0.4861,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5940218603613652,
+      "grad_norm": 0.1630701869726181,
+      "learning_rate": 1.8236671452079805e-05,
+      "loss": 0.5226,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5942449252732546,
+      "grad_norm": 0.16565611958503723,
+      "learning_rate": 1.8235336656488203e-05,
+      "loss": 0.4905,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5944679901851438,
+      "grad_norm": 0.15455248951911926,
+      "learning_rate": 1.8234001404771856e-05,
+      "loss": 0.4844,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5946910550970332,
+      "grad_norm": 0.1679016500711441,
+      "learning_rate": 1.8232665697004713e-05,
+      "loss": 0.5312,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5949141200089226,
+      "grad_norm": 0.16260258853435516,
+      "learning_rate": 1.823132953326076e-05,
+      "loss": 0.5106,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5951371849208119,
+      "grad_norm": 0.15242820978164673,
+      "learning_rate": 1.8229992913614004e-05,
+      "loss": 0.4641,
+      "step": 2668
+    },
+    {
+      "epoch": 0.5953602498327013,
+      "grad_norm": 0.17028586566448212,
+      "learning_rate": 1.822865583813847e-05,
+      "loss": 0.5183,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5955833147445907,
+      "grad_norm": 0.14964601397514343,
+      "learning_rate": 1.8227318306908216e-05,
+      "loss": 0.4843,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5958063796564801,
+      "grad_norm": 0.1675204187631607,
+      "learning_rate": 1.822598031999732e-05,
+      "loss": 0.5098,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5960294445683694,
+      "grad_norm": 0.15445132553577423,
+      "learning_rate": 1.822464187747989e-05,
+      "loss": 0.4759,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5962525094802588,
+      "grad_norm": 0.1613750010728836,
+      "learning_rate": 1.822330297943006e-05,
+      "loss": 0.5154,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5964755743921482,
+      "grad_norm": 0.16264407336711884,
+      "learning_rate": 1.8221963625921984e-05,
+      "loss": 0.4758,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5966986393040374,
+      "grad_norm": 0.16100256145000458,
+      "learning_rate": 1.8220623817029843e-05,
+      "loss": 0.4946,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5969217042159268,
+      "grad_norm": 0.1642381250858307,
+      "learning_rate": 1.8219283552827847e-05,
+      "loss": 0.5029,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5971447691278162,
+      "grad_norm": 0.16491542756557465,
+      "learning_rate": 1.8217942833390227e-05,
+      "loss": 0.5077,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5973678340397055,
+      "grad_norm": 0.1611475944519043,
+      "learning_rate": 1.821660165879124e-05,
+      "loss": 0.5346,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5975908989515949,
+      "grad_norm": 0.16811859607696533,
+      "learning_rate": 1.8215260029105166e-05,
+      "loss": 0.5173,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5978139638634843,
+      "grad_norm": 0.17261561751365662,
+      "learning_rate": 1.8213917944406315e-05,
+      "loss": 0.5142,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5980370287753737,
+      "grad_norm": 0.1671370267868042,
+      "learning_rate": 1.8212575404769023e-05,
+      "loss": 0.5289,
+      "step": 2681
+    },
+    {
+      "epoch": 0.598260093687263,
+      "grad_norm": 0.16574646532535553,
+      "learning_rate": 1.8211232410267645e-05,
+      "loss": 0.5179,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5984831585991524,
+      "grad_norm": 0.1583874672651291,
+      "learning_rate": 1.8209888960976565e-05,
+      "loss": 0.503,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5987062235110417,
+      "grad_norm": 0.15771937370300293,
+      "learning_rate": 1.8208545056970193e-05,
+      "loss": 0.4811,
+      "step": 2684
+    },
+    {
+      "epoch": 0.598929288422931,
+      "grad_norm": 0.16496342420578003,
+      "learning_rate": 1.820720069832296e-05,
+      "loss": 0.5178,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5991523533348204,
+      "grad_norm": 0.1662297397851944,
+      "learning_rate": 1.820585588510933e-05,
+      "loss": 0.5286,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5993754182467098,
+      "grad_norm": 0.15326498448848724,
+      "learning_rate": 1.8204510617403785e-05,
+      "loss": 0.4983,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5995984831585992,
+      "grad_norm": 0.1545604169368744,
+      "learning_rate": 1.820316489528083e-05,
+      "loss": 0.4746,
+      "step": 2688
+    },
+    {
+      "epoch": 0.5998215480704885,
+      "grad_norm": 0.16388210654258728,
+      "learning_rate": 1.8201818718815004e-05,
+      "loss": 0.4766,
+      "step": 2689
+    },
+    {
+      "epoch": 0.6000446129823779,
+      "grad_norm": 0.16617290675640106,
+      "learning_rate": 1.820047208808087e-05,
+      "loss": 0.5301,
+      "step": 2690
+    },
+    {
+      "epoch": 0.6002676778942673,
+      "grad_norm": 0.16141341626644135,
+      "learning_rate": 1.8199125003153e-05,
+      "loss": 0.5095,
+      "step": 2691
+    },
+    {
+      "epoch": 0.6004907428061566,
+      "grad_norm": 0.1575475037097931,
+      "learning_rate": 1.8197777464106022e-05,
+      "loss": 0.5164,
+      "step": 2692
+    },
+    {
+      "epoch": 0.600713807718046,
+      "grad_norm": 0.1886652559041977,
+      "learning_rate": 1.8196429471014558e-05,
+      "loss": 0.4922,
+      "step": 2693
+    },
+    {
+      "epoch": 0.6009368726299353,
+      "grad_norm": 0.16036200523376465,
+      "learning_rate": 1.8195081023953268e-05,
+      "loss": 0.5116,
+      "step": 2694
+    },
+    {
+      "epoch": 0.6011599375418246,
+      "grad_norm": 0.16269893944263458,
+      "learning_rate": 1.8193732122996847e-05,
+      "loss": 0.5135,
+      "step": 2695
+    },
+    {
+      "epoch": 0.601383002453714,
+      "grad_norm": 0.1868923157453537,
+      "learning_rate": 1.819238276822e-05,
+      "loss": 0.5052,
+      "step": 2696
+    },
+    {
+      "epoch": 0.6016060673656034,
+      "grad_norm": 0.16269947588443756,
+      "learning_rate": 1.8191032959697464e-05,
+      "loss": 0.4829,
+      "step": 2697
+    },
+    {
+      "epoch": 0.6018291322774928,
+      "grad_norm": 0.1679493635892868,
+      "learning_rate": 1.8189682697504e-05,
+      "loss": 0.4967,
+      "step": 2698
+    },
+    {
+      "epoch": 0.6020521971893821,
+      "grad_norm": 0.16471365094184875,
+      "learning_rate": 1.8188331981714386e-05,
+      "loss": 0.5189,
+      "step": 2699
+    },
+    {
+      "epoch": 0.6022752621012715,
+      "grad_norm": 0.16562478244304657,
+      "learning_rate": 1.8186980812403448e-05,
+      "loss": 0.5217,
+      "step": 2700
+    },
+    {
+      "epoch": 0.6024983270131609,
+      "grad_norm": 0.16145823895931244,
+      "learning_rate": 1.818562918964601e-05,
+      "loss": 0.5008,
+      "step": 2701
+    },
+    {
+      "epoch": 0.6027213919250501,
+      "grad_norm": 0.17072793841362,
+      "learning_rate": 1.8184277113516938e-05,
+      "loss": 0.5302,
+      "step": 2702
+    },
+    {
+      "epoch": 0.6029444568369395,
+      "grad_norm": 0.16824592649936676,
+      "learning_rate": 1.8182924584091122e-05,
+      "loss": 0.5358,
+      "step": 2703
+    },
+    {
+      "epoch": 0.6031675217488289,
+      "grad_norm": 0.1661488264799118,
+      "learning_rate": 1.8181571601443465e-05,
+      "loss": 0.5391,
+      "step": 2704
+    },
+    {
+      "epoch": 0.6033905866607183,
+      "grad_norm": 0.15754079818725586,
+      "learning_rate": 1.8180218165648913e-05,
+      "loss": 0.5013,
+      "step": 2705
+    },
+    {
+      "epoch": 0.6036136515726076,
+      "grad_norm": 0.1813340038061142,
+      "learning_rate": 1.817886427678242e-05,
+      "loss": 0.5301,
+      "step": 2706
+    },
+    {
+      "epoch": 0.603836716484497,
+      "grad_norm": 0.16588057577610016,
+      "learning_rate": 1.817750993491898e-05,
+      "loss": 0.4924,
+      "step": 2707
+    },
+    {
+      "epoch": 0.6040597813963864,
+      "grad_norm": 0.17572402954101562,
+      "learning_rate": 1.8176155140133596e-05,
+      "loss": 0.5075,
+      "step": 2708
+    },
+    {
+      "epoch": 0.6042828463082757,
+      "grad_norm": 0.157828688621521,
+      "learning_rate": 1.8174799892501315e-05,
+      "loss": 0.5133,
+      "step": 2709
+    },
+    {
+      "epoch": 0.6045059112201651,
+      "grad_norm": 0.1617002636194229,
+      "learning_rate": 1.817344419209719e-05,
+      "loss": 0.5092,
+      "step": 2710
+    },
+    {
+      "epoch": 0.6047289761320545,
+      "grad_norm": 0.1876181662082672,
+      "learning_rate": 1.817208803899632e-05,
+      "loss": 0.5326,
+      "step": 2711
+    },
+    {
+      "epoch": 0.6049520410439437,
+      "grad_norm": 0.15636108815670013,
+      "learning_rate": 1.8170731433273802e-05,
+      "loss": 0.4922,
+      "step": 2712
+    },
+    {
+      "epoch": 0.6051751059558331,
+      "grad_norm": 0.15785710513591766,
+      "learning_rate": 1.8169374375004784e-05,
+      "loss": 0.4956,
+      "step": 2713
+    },
+    {
+      "epoch": 0.6053981708677225,
+      "grad_norm": 0.17369434237480164,
+      "learning_rate": 1.8168016864264426e-05,
+      "loss": 0.493,
+      "step": 2714
+    },
+    {
+      "epoch": 0.6056212357796119,
+      "grad_norm": 0.16383057832717896,
+      "learning_rate": 1.8166658901127915e-05,
+      "loss": 0.4638,
+      "step": 2715
+    },
+    {
+      "epoch": 0.6058443006915012,
+      "grad_norm": 0.17268022894859314,
+      "learning_rate": 1.8165300485670464e-05,
+      "loss": 0.5056,
+      "step": 2716
+    },
+    {
+      "epoch": 0.6060673656033906,
+      "grad_norm": 0.1691250205039978,
+      "learning_rate": 1.8163941617967313e-05,
+      "loss": 0.5161,
+      "step": 2717
+    },
+    {
+      "epoch": 0.60629043051528,
+      "grad_norm": 0.1628870815038681,
+      "learning_rate": 1.8162582298093715e-05,
+      "loss": 0.5015,
+      "step": 2718
+    },
+    {
+      "epoch": 0.6065134954271693,
+      "grad_norm": 0.17182950675487518,
+      "learning_rate": 1.816122252612497e-05,
+      "loss": 0.5237,
+      "step": 2719
+    },
+    {
+      "epoch": 0.6067365603390587,
+      "grad_norm": 0.17092610895633698,
+      "learning_rate": 1.8159862302136386e-05,
+      "loss": 0.4941,
+      "step": 2720
+    },
+    {
+      "epoch": 0.606959625250948,
+      "grad_norm": 0.16498810052871704,
+      "learning_rate": 1.8158501626203298e-05,
+      "loss": 0.5197,
+      "step": 2721
+    },
+    {
+      "epoch": 0.6071826901628374,
+      "grad_norm": 0.16107720136642456,
+      "learning_rate": 1.815714049840107e-05,
+      "loss": 0.5186,
+      "step": 2722
+    },
+    {
+      "epoch": 0.6074057550747267,
+      "grad_norm": 0.1599961370229721,
+      "learning_rate": 1.8155778918805095e-05,
+      "loss": 0.5035,
+      "step": 2723
+    },
+    {
+      "epoch": 0.6076288199866161,
+      "grad_norm": 0.16911983489990234,
+      "learning_rate": 1.815441688749078e-05,
+      "loss": 0.5009,
+      "step": 2724
+    },
+    {
+      "epoch": 0.6078518848985055,
+      "grad_norm": 0.16482600569725037,
+      "learning_rate": 1.8153054404533562e-05,
+      "loss": 0.4991,
+      "step": 2725
+    },
+    {
+      "epoch": 0.6080749498103948,
+      "grad_norm": 0.16007374227046967,
+      "learning_rate": 1.8151691470008906e-05,
+      "loss": 0.4837,
+      "step": 2726
+    },
+    {
+      "epoch": 0.6082980147222842,
+      "grad_norm": 0.16472531855106354,
+      "learning_rate": 1.81503280839923e-05,
+      "loss": 0.4997,
+      "step": 2727
+    },
+    {
+      "epoch": 0.6085210796341736,
+      "grad_norm": 0.1605815589427948,
+      "learning_rate": 1.814896424655926e-05,
+      "loss": 0.5011,
+      "step": 2728
+    },
+    {
+      "epoch": 0.608744144546063,
+      "grad_norm": 0.16523852944374084,
+      "learning_rate": 1.814759995778532e-05,
+      "loss": 0.4947,
+      "step": 2729
+    },
+    {
+      "epoch": 0.6089672094579522,
+      "grad_norm": 0.1722632497549057,
+      "learning_rate": 1.8146235217746043e-05,
+      "loss": 0.495,
+      "step": 2730
+    },
+    {
+      "epoch": 0.6091902743698416,
+      "grad_norm": 0.16873040795326233,
+      "learning_rate": 1.8144870026517018e-05,
+      "loss": 0.5273,
+      "step": 2731
+    },
+    {
+      "epoch": 0.609413339281731,
+      "grad_norm": 0.16078028082847595,
+      "learning_rate": 1.8143504384173858e-05,
+      "loss": 0.5051,
+      "step": 2732
+    },
+    {
+      "epoch": 0.6096364041936203,
+      "grad_norm": 0.18267033994197845,
+      "learning_rate": 1.8142138290792202e-05,
+      "loss": 0.4946,
+      "step": 2733
+    },
+    {
+      "epoch": 0.6098594691055097,
+      "grad_norm": 0.17000767588615417,
+      "learning_rate": 1.814077174644771e-05,
+      "loss": 0.5079,
+      "step": 2734
+    },
+    {
+      "epoch": 0.6100825340173991,
+      "grad_norm": 0.1608540564775467,
+      "learning_rate": 1.813940475121607e-05,
+      "loss": 0.4912,
+      "step": 2735
+    },
+    {
+      "epoch": 0.6103055989292884,
+      "grad_norm": 0.160196915268898,
+      "learning_rate": 1.8138037305172997e-05,
+      "loss": 0.5315,
+      "step": 2736
+    },
+    {
+      "epoch": 0.6105286638411778,
+      "grad_norm": 0.16966816782951355,
+      "learning_rate": 1.813666940839423e-05,
+      "loss": 0.532,
+      "step": 2737
+    },
+    {
+      "epoch": 0.6107517287530672,
+      "grad_norm": 0.19911563396453857,
+      "learning_rate": 1.8135301060955525e-05,
+      "loss": 0.4847,
+      "step": 2738
+    },
+    {
+      "epoch": 0.6109747936649566,
+      "grad_norm": 0.16148284077644348,
+      "learning_rate": 1.8133932262932678e-05,
+      "loss": 0.4756,
+      "step": 2739
+    },
+    {
+      "epoch": 0.6111978585768458,
+      "grad_norm": 0.16302482783794403,
+      "learning_rate": 1.8132563014401497e-05,
+      "loss": 0.5021,
+      "step": 2740
+    },
+    {
+      "epoch": 0.6114209234887352,
+      "grad_norm": 0.1716173142194748,
+      "learning_rate": 1.813119331543782e-05,
+      "loss": 0.5303,
+      "step": 2741
+    },
+    {
+      "epoch": 0.6116439884006246,
+      "grad_norm": 0.16563881933689117,
+      "learning_rate": 1.812982316611751e-05,
+      "loss": 0.4991,
+      "step": 2742
+    },
+    {
+      "epoch": 0.6118670533125139,
+      "grad_norm": 0.1720925122499466,
+      "learning_rate": 1.812845256651645e-05,
+      "loss": 0.5026,
+      "step": 2743
+    },
+    {
+      "epoch": 0.6120901182244033,
+      "grad_norm": 0.16051127016544342,
+      "learning_rate": 1.8127081516710565e-05,
+      "loss": 0.4837,
+      "step": 2744
+    },
+    {
+      "epoch": 0.6123131831362927,
+      "grad_norm": 0.16201861202716827,
+      "learning_rate": 1.8125710016775778e-05,
+      "loss": 0.5161,
+      "step": 2745
+    },
+    {
+      "epoch": 0.6125362480481821,
+      "grad_norm": 0.17014898359775543,
+      "learning_rate": 1.812433806678806e-05,
+      "loss": 0.5346,
+      "step": 2746
+    },
+    {
+      "epoch": 0.6127593129600714,
+      "grad_norm": 0.1703520119190216,
+      "learning_rate": 1.8122965666823398e-05,
+      "loss": 0.5182,
+      "step": 2747
+    },
+    {
+      "epoch": 0.6129823778719607,
+      "grad_norm": 0.19158673286437988,
+      "learning_rate": 1.8121592816957797e-05,
+      "loss": 0.5043,
+      "step": 2748
+    },
+    {
+      "epoch": 0.6132054427838501,
+      "grad_norm": 0.14746567606925964,
+      "learning_rate": 1.8120219517267302e-05,
+      "loss": 0.4841,
+      "step": 2749
+    },
+    {
+      "epoch": 0.6134285076957394,
+      "grad_norm": 0.16160626709461212,
+      "learning_rate": 1.811884576782797e-05,
+      "loss": 0.5158,
+      "step": 2750
+    },
+    {
+      "epoch": 0.6136515726076288,
+      "grad_norm": 0.17975735664367676,
+      "learning_rate": 1.8117471568715893e-05,
+      "loss": 0.5235,
+      "step": 2751
+    },
+    {
+      "epoch": 0.6138746375195182,
+      "grad_norm": 0.17119070887565613,
+      "learning_rate": 1.8116096920007177e-05,
+      "loss": 0.4951,
+      "step": 2752
+    },
+    {
+      "epoch": 0.6140977024314075,
+      "grad_norm": 0.16311194002628326,
+      "learning_rate": 1.8114721821777964e-05,
+      "loss": 0.499,
+      "step": 2753
+    },
+    {
+      "epoch": 0.6143207673432969,
+      "grad_norm": 0.16315460205078125,
+      "learning_rate": 1.811334627410441e-05,
+      "loss": 0.51,
+      "step": 2754
+    },
+    {
+      "epoch": 0.6145438322551863,
+      "grad_norm": 0.44980496168136597,
+      "learning_rate": 1.81119702770627e-05,
+      "loss": 0.5005,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6147668971670757,
+      "grad_norm": 0.17961294949054718,
+      "learning_rate": 1.8110593830729057e-05,
+      "loss": 0.5204,
+      "step": 2756
+    },
+    {
+      "epoch": 0.614989962078965,
+      "grad_norm": 0.16227351129055023,
+      "learning_rate": 1.8109216935179712e-05,
+      "loss": 0.5081,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6152130269908543,
+      "grad_norm": 0.1761593520641327,
+      "learning_rate": 1.810783959049092e-05,
+      "loss": 0.5163,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6154360919027437,
+      "grad_norm": 0.17095595598220825,
+      "learning_rate": 1.810646179673897e-05,
+      "loss": 0.4948,
+      "step": 2759
+    },
+    {
+      "epoch": 0.615659156814633,
+      "grad_norm": 0.16665005683898926,
+      "learning_rate": 1.8105083554000175e-05,
+      "loss": 0.4937,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6158822217265224,
+      "grad_norm": 0.16761796176433563,
+      "learning_rate": 1.810370486235087e-05,
+      "loss": 0.5081,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6161052866384118,
+      "grad_norm": 0.17292195558547974,
+      "learning_rate": 1.8102325721867417e-05,
+      "loss": 0.5027,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6163283515503012,
+      "grad_norm": 0.1606137752532959,
+      "learning_rate": 1.8100946132626197e-05,
+      "loss": 0.4867,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6165514164621905,
+      "grad_norm": 0.1619354635477066,
+      "learning_rate": 1.8099566094703626e-05,
+      "loss": 0.474,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6167744813740799,
+      "grad_norm": 0.16491201519966125,
+      "learning_rate": 1.8098185608176132e-05,
+      "loss": 0.4953,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6169975462859693,
+      "grad_norm": 0.16283756494522095,
+      "learning_rate": 1.8096804673120183e-05,
+      "loss": 0.4934,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6172206111978585,
+      "grad_norm": 0.15283644199371338,
+      "learning_rate": 1.809542328961226e-05,
+      "loss": 0.4691,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6174436761097479,
+      "grad_norm": 0.16907426714897156,
+      "learning_rate": 1.809404145772887e-05,
+      "loss": 0.4901,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6176667410216373,
+      "grad_norm": 0.16500405967235565,
+      "learning_rate": 1.8092659177546554e-05,
+      "loss": 0.5042,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6178898059335266,
+      "grad_norm": 0.16616129875183105,
+      "learning_rate": 1.8091276449141868e-05,
+      "loss": 0.528,
+      "step": 2770
+    },
+    {
+      "epoch": 0.618112870845416,
+      "grad_norm": 0.15396897494792938,
+      "learning_rate": 1.8089893272591393e-05,
+      "loss": 0.4797,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6183359357573054,
+      "grad_norm": 0.1576898992061615,
+      "learning_rate": 1.8088509647971744e-05,
+      "loss": 0.5007,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6185590006691948,
+      "grad_norm": 0.1703861653804779,
+      "learning_rate": 1.808712557535955e-05,
+      "loss": 0.5114,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6187820655810841,
+      "grad_norm": 0.15141689777374268,
+      "learning_rate": 1.8085741054831472e-05,
+      "loss": 0.457,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6190051304929735,
+      "grad_norm": 0.19050109386444092,
+      "learning_rate": 1.8084356086464197e-05,
+      "loss": 0.474,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6192281954048628,
+      "grad_norm": 0.17825506627559662,
+      "learning_rate": 1.8082970670334425e-05,
+      "loss": 0.5164,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6194512603167521,
+      "grad_norm": 0.16141526401042938,
+      "learning_rate": 1.8081584806518897e-05,
+      "loss": 0.5025,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6196743252286415,
+      "grad_norm": 0.23704519867897034,
+      "learning_rate": 1.8080198495094364e-05,
+      "loss": 0.5315,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6198973901405309,
+      "grad_norm": 0.15994718670845032,
+      "learning_rate": 1.8078811736137612e-05,
+      "loss": 0.4853,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6201204550524203,
+      "grad_norm": 0.1649598479270935,
+      "learning_rate": 1.807742452972545e-05,
+      "loss": 0.5262,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6203435199643096,
+      "grad_norm": 0.14894191920757294,
+      "learning_rate": 1.8076036875934707e-05,
+      "loss": 0.4816,
+      "step": 2781
+    },
+    {
+      "epoch": 0.620566584876199,
+      "grad_norm": 0.17157211899757385,
+      "learning_rate": 1.807464877484224e-05,
+      "loss": 0.4847,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6207896497880884,
+      "grad_norm": 0.1663542538881302,
+      "learning_rate": 1.8073260226524937e-05,
+      "loss": 0.4931,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6210127146999777,
+      "grad_norm": 0.1610024869441986,
+      "learning_rate": 1.8071871231059695e-05,
+      "loss": 0.4924,
+      "step": 2784
+    },
+    {
+      "epoch": 0.621235779611867,
+      "grad_norm": 0.16384142637252808,
+      "learning_rate": 1.807048178852345e-05,
+      "loss": 0.5143,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6214588445237564,
+      "grad_norm": 0.16362221539020538,
+      "learning_rate": 1.8069091898993162e-05,
+      "loss": 0.5172,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6216819094356457,
+      "grad_norm": 0.19173026084899902,
+      "learning_rate": 1.8067701562545808e-05,
+      "loss": 0.512,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6219049743475351,
+      "grad_norm": 0.15911833941936493,
+      "learning_rate": 1.8066310779258393e-05,
+      "loss": 0.4874,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6221280392594245,
+      "grad_norm": 0.15391753613948822,
+      "learning_rate": 1.8064919549207946e-05,
+      "loss": 0.4966,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6223511041713139,
+      "grad_norm": 0.21308940649032593,
+      "learning_rate": 1.8063527872471523e-05,
+      "loss": 0.5172,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6225741690832032,
+      "grad_norm": 0.15762938559055328,
+      "learning_rate": 1.8062135749126208e-05,
+      "loss": 0.4904,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6227972339950926,
+      "grad_norm": 0.18538956344127655,
+      "learning_rate": 1.80607431792491e-05,
+      "loss": 0.4907,
+      "step": 2792
+    },
+    {
+      "epoch": 0.623020298906982,
+      "grad_norm": 0.1694374829530716,
+      "learning_rate": 1.8059350162917333e-05,
+      "loss": 0.4892,
+      "step": 2793
+    },
+    {
+      "epoch": 0.6232433638188712,
+      "grad_norm": 0.1632799506187439,
+      "learning_rate": 1.8057956700208055e-05,
+      "loss": 0.531,
+      "step": 2794
+    },
+    {
+      "epoch": 0.6234664287307606,
+      "grad_norm": 0.19030988216400146,
+      "learning_rate": 1.805656279119845e-05,
+      "loss": 0.4801,
+      "step": 2795
+    },
+    {
+      "epoch": 0.62368949364265,
+      "grad_norm": 0.17578838765621185,
+      "learning_rate": 1.8055168435965722e-05,
+      "loss": 0.5129,
+      "step": 2796
+    },
+    {
+      "epoch": 0.6239125585545394,
+      "grad_norm": 0.17337745428085327,
+      "learning_rate": 1.8053773634587095e-05,
+      "loss": 0.5252,
+      "step": 2797
+    },
+    {
+      "epoch": 0.6241356234664287,
+      "grad_norm": 0.16400404274463654,
+      "learning_rate": 1.8052378387139827e-05,
+      "loss": 0.4977,
+      "step": 2798
+    },
+    {
+      "epoch": 0.6243586883783181,
+      "grad_norm": 0.16884362697601318,
+      "learning_rate": 1.8050982693701188e-05,
+      "loss": 0.5209,
+      "step": 2799
+    },
+    {
+      "epoch": 0.6245817532902075,
+      "grad_norm": 0.16655333340168,
+      "learning_rate": 1.8049586554348487e-05,
+      "loss": 0.5236,
+      "step": 2800
+    },
+    {
+      "epoch": 0.6248048182020968,
+      "grad_norm": 0.1815658062696457,
+      "learning_rate": 1.804818996915905e-05,
+      "loss": 0.5155,
+      "step": 2801
+    },
+    {
+      "epoch": 0.6250278831139862,
+      "grad_norm": 0.1637655794620514,
+      "learning_rate": 1.8046792938210226e-05,
+      "loss": 0.5108,
+      "step": 2802
+    },
+    {
+      "epoch": 0.6252509480258756,
+      "grad_norm": 0.16214872896671295,
+      "learning_rate": 1.804539546157939e-05,
+      "loss": 0.5015,
+      "step": 2803
+    },
+    {
+      "epoch": 0.625474012937765,
+      "grad_norm": 0.1624768078327179,
+      "learning_rate": 1.804399753934395e-05,
+      "loss": 0.4852,
+      "step": 2804
+    },
+    {
+      "epoch": 0.6256970778496542,
+      "grad_norm": 0.1509588062763214,
+      "learning_rate": 1.8042599171581322e-05,
+      "loss": 0.4895,
+      "step": 2805
+    },
+    {
+      "epoch": 0.6259201427615436,
+      "grad_norm": 0.15983150899410248,
+      "learning_rate": 1.804120035836897e-05,
+      "loss": 0.5066,
+      "step": 2806
+    },
+    {
+      "epoch": 0.626143207673433,
+      "grad_norm": 0.1639653593301773,
+      "learning_rate": 1.8039801099784356e-05,
+      "loss": 0.5081,
+      "step": 2807
+    },
+    {
+      "epoch": 0.6263662725853223,
+      "grad_norm": 0.16054677963256836,
+      "learning_rate": 1.8038401395904984e-05,
+      "loss": 0.5019,
+      "step": 2808
+    },
+    {
+      "epoch": 0.6265893374972117,
+      "grad_norm": 0.16557928919792175,
+      "learning_rate": 1.8037001246808382e-05,
+      "loss": 0.5085,
+      "step": 2809
+    },
+    {
+      "epoch": 0.6268124024091011,
+      "grad_norm": 0.17172253131866455,
+      "learning_rate": 1.8035600652572093e-05,
+      "loss": 0.5069,
+      "step": 2810
+    },
+    {
+      "epoch": 0.6270354673209904,
+      "grad_norm": 0.16289275884628296,
+      "learning_rate": 1.80341996132737e-05,
+      "loss": 0.5054,
+      "step": 2811
+    },
+    {
+      "epoch": 0.6272585322328798,
+      "grad_norm": 0.159224271774292,
+      "learning_rate": 1.8032798128990788e-05,
+      "loss": 0.4884,
+      "step": 2812
+    },
+    {
+      "epoch": 0.6274815971447691,
+      "grad_norm": 0.1516553908586502,
+      "learning_rate": 1.803139619980099e-05,
+      "loss": 0.4823,
+      "step": 2813
+    },
+    {
+      "epoch": 0.6277046620566585,
+      "grad_norm": 0.164667546749115,
+      "learning_rate": 1.802999382578195e-05,
+      "loss": 0.5184,
+      "step": 2814
+    },
+    {
+      "epoch": 0.6279277269685478,
+      "grad_norm": 0.17175255715847015,
+      "learning_rate": 1.8028591007011343e-05,
+      "loss": 0.5166,
+      "step": 2815
+    },
+    {
+      "epoch": 0.6281507918804372,
+      "grad_norm": 0.18747776746749878,
+      "learning_rate": 1.8027187743566867e-05,
+      "loss": 0.499,
+      "step": 2816
+    },
+    {
+      "epoch": 0.6283738567923266,
+      "grad_norm": 0.16748382151126862,
+      "learning_rate": 1.8025784035526235e-05,
+      "loss": 0.5099,
+      "step": 2817
+    },
+    {
+      "epoch": 0.6285969217042159,
+      "grad_norm": 0.19813144207000732,
+      "learning_rate": 1.80243798829672e-05,
+      "loss": 0.5045,
+      "step": 2818
+    },
+    {
+      "epoch": 0.6288199866161053,
+      "grad_norm": 0.16501356661319733,
+      "learning_rate": 1.8022975285967534e-05,
+      "loss": 0.514,
+      "step": 2819
+    },
+    {
+      "epoch": 0.6290430515279947,
+      "grad_norm": 0.17181000113487244,
+      "learning_rate": 1.8021570244605028e-05,
+      "loss": 0.4963,
+      "step": 2820
+    },
+    {
+      "epoch": 0.6292661164398841,
+      "grad_norm": 0.16481012105941772,
+      "learning_rate": 1.8020164758957505e-05,
+      "loss": 0.4831,
+      "step": 2821
+    },
+    {
+      "epoch": 0.6294891813517733,
+      "grad_norm": 0.1659911572933197,
+      "learning_rate": 1.8018758829102808e-05,
+      "loss": 0.5026,
+      "step": 2822
+    },
+    {
+      "epoch": 0.6297122462636627,
+      "grad_norm": 0.18394367396831512,
+      "learning_rate": 1.8017352455118812e-05,
+      "loss": 0.5072,
+      "step": 2823
+    },
+    {
+      "epoch": 0.6299353111755521,
+      "grad_norm": 0.1666680872440338,
+      "learning_rate": 1.80159456370834e-05,
+      "loss": 0.4986,
+      "step": 2824
+    },
+    {
+      "epoch": 0.6301583760874414,
+      "grad_norm": 0.184769406914711,
+      "learning_rate": 1.80145383750745e-05,
+      "loss": 0.4863,
+      "step": 2825
+    },
+    {
+      "epoch": 0.6303814409993308,
+      "grad_norm": 0.15709228813648224,
+      "learning_rate": 1.801313066917005e-05,
+      "loss": 0.4945,
+      "step": 2826
+    },
+    {
+      "epoch": 0.6306045059112202,
+      "grad_norm": 0.16040046513080597,
+      "learning_rate": 1.801172251944802e-05,
+      "loss": 0.4704,
+      "step": 2827
+    },
+    {
+      "epoch": 0.6308275708231095,
+      "grad_norm": 0.174714133143425,
+      "learning_rate": 1.8010313925986398e-05,
+      "loss": 0.5224,
+      "step": 2828
+    },
+    {
+      "epoch": 0.6310506357349989,
+      "grad_norm": 0.17879654467105865,
+      "learning_rate": 1.8008904888863206e-05,
+      "loss": 0.5168,
+      "step": 2829
+    },
+    {
+      "epoch": 0.6312737006468883,
+      "grad_norm": 0.16159847378730774,
+      "learning_rate": 1.8007495408156483e-05,
+      "loss": 0.4905,
+      "step": 2830
+    },
+    {
+      "epoch": 0.6314967655587777,
+      "grad_norm": 0.16279365122318268,
+      "learning_rate": 1.8006085483944295e-05,
+      "loss": 0.5004,
+      "step": 2831
+    },
+    {
+      "epoch": 0.6317198304706669,
+      "grad_norm": 0.16124996542930603,
+      "learning_rate": 1.800467511630473e-05,
+      "loss": 0.498,
+      "step": 2832
+    },
+    {
+      "epoch": 0.6319428953825563,
+      "grad_norm": 0.16358453035354614,
+      "learning_rate": 1.800326430531591e-05,
+      "loss": 0.5172,
+      "step": 2833
+    },
+    {
+      "epoch": 0.6321659602944457,
+      "grad_norm": 0.16300733387470245,
+      "learning_rate": 1.8001853051055967e-05,
+      "loss": 0.5024,
+      "step": 2834
+    },
+    {
+      "epoch": 0.632389025206335,
+      "grad_norm": 0.1758897304534912,
+      "learning_rate": 1.8000441353603072e-05,
+      "loss": 0.4946,
+      "step": 2835
+    },
+    {
+      "epoch": 0.6326120901182244,
+      "grad_norm": 0.1550527960062027,
+      "learning_rate": 1.7999029213035408e-05,
+      "loss": 0.5027,
+      "step": 2836
+    },
+    {
+      "epoch": 0.6328351550301138,
+      "grad_norm": 0.15276572108268738,
+      "learning_rate": 1.799761662943119e-05,
+      "loss": 0.4968,
+      "step": 2837
+    },
+    {
+      "epoch": 0.6330582199420032,
+      "grad_norm": 0.16600914299488068,
+      "learning_rate": 1.7996203602868657e-05,
+      "loss": 0.5278,
+      "step": 2838
+    },
+    {
+      "epoch": 0.6332812848538925,
+      "grad_norm": 0.16788353025913239,
+      "learning_rate": 1.799479013342607e-05,
+      "loss": 0.5071,
+      "step": 2839
+    },
+    {
+      "epoch": 0.6335043497657818,
+      "grad_norm": 0.16952337324619293,
+      "learning_rate": 1.7993376221181716e-05,
+      "loss": 0.4852,
+      "step": 2840
+    },
+    {
+      "epoch": 0.6337274146776712,
+      "grad_norm": 0.28554677963256836,
+      "learning_rate": 1.7991961866213907e-05,
+      "loss": 0.474,
+      "step": 2841
+    },
+    {
+      "epoch": 0.6339504795895605,
+      "grad_norm": 0.17595677077770233,
+      "learning_rate": 1.7990547068600977e-05,
+      "loss": 0.5045,
+      "step": 2842
+    },
+    {
+      "epoch": 0.6341735445014499,
+      "grad_norm": 0.1631106734275818,
+      "learning_rate": 1.798913182842129e-05,
+      "loss": 0.4943,
+      "step": 2843
+    },
+    {
+      "epoch": 0.6343966094133393,
+      "grad_norm": 0.1630406528711319,
+      "learning_rate": 1.7987716145753226e-05,
+      "loss": 0.5095,
+      "step": 2844
+    },
+    {
+      "epoch": 0.6346196743252286,
+      "grad_norm": 0.17511357367038727,
+      "learning_rate": 1.7986300020675198e-05,
+      "loss": 0.5203,
+      "step": 2845
+    },
+    {
+      "epoch": 0.634842739237118,
+      "grad_norm": 0.1971133053302765,
+      "learning_rate": 1.798488345326564e-05,
+      "loss": 0.5061,
+      "step": 2846
+    },
+    {
+      "epoch": 0.6350658041490074,
+      "grad_norm": 0.15673868358135223,
+      "learning_rate": 1.7983466443603008e-05,
+      "loss": 0.4957,
+      "step": 2847
+    },
+    {
+      "epoch": 0.6352888690608968,
+      "grad_norm": 0.1608821600675583,
+      "learning_rate": 1.798204899176579e-05,
+      "loss": 0.4848,
+      "step": 2848
+    },
+    {
+      "epoch": 0.635511933972786,
+      "grad_norm": 0.15686407685279846,
+      "learning_rate": 1.7980631097832485e-05,
+      "loss": 0.4986,
+      "step": 2849
+    },
+    {
+      "epoch": 0.6357349988846754,
+      "grad_norm": 0.16254571080207825,
+      "learning_rate": 1.797921276188163e-05,
+      "loss": 0.5255,
+      "step": 2850
+    },
+    {
+      "epoch": 0.6359580637965648,
+      "grad_norm": 0.17136751115322113,
+      "learning_rate": 1.7977793983991785e-05,
+      "loss": 0.5029,
+      "step": 2851
+    },
+    {
+      "epoch": 0.6361811287084541,
+      "grad_norm": 0.2001720815896988,
+      "learning_rate": 1.7976374764241523e-05,
+      "loss": 0.5208,
+      "step": 2852
+    },
+    {
+      "epoch": 0.6364041936203435,
+      "grad_norm": 0.15888839960098267,
+      "learning_rate": 1.7974955102709457e-05,
+      "loss": 0.5125,
+      "step": 2853
+    },
+    {
+      "epoch": 0.6366272585322329,
+      "grad_norm": 0.1562851518392563,
+      "learning_rate": 1.797353499947421e-05,
+      "loss": 0.4647,
+      "step": 2854
+    },
+    {
+      "epoch": 0.6368503234441223,
+      "grad_norm": 0.157870814204216,
+      "learning_rate": 1.7972114454614436e-05,
+      "loss": 0.5072,
+      "step": 2855
+    },
+    {
+      "epoch": 0.6370733883560116,
+      "grad_norm": 0.16168496012687683,
+      "learning_rate": 1.7970693468208823e-05,
+      "loss": 0.4946,
+      "step": 2856
+    },
+    {
+      "epoch": 0.637296453267901,
+      "grad_norm": 0.23463092744350433,
+      "learning_rate": 1.796927204033607e-05,
+      "loss": 0.5073,
+      "step": 2857
+    },
+    {
+      "epoch": 0.6375195181797904,
+      "grad_norm": 0.18455378711223602,
+      "learning_rate": 1.7967850171074896e-05,
+      "loss": 0.5278,
+      "step": 2858
+    },
+    {
+      "epoch": 0.6377425830916796,
+      "grad_norm": 0.16380822658538818,
+      "learning_rate": 1.796642786050406e-05,
+      "loss": 0.5048,
+      "step": 2859
+    },
+    {
+      "epoch": 0.637965648003569,
+      "grad_norm": 0.15988902747631073,
+      "learning_rate": 1.7965005108702342e-05,
+      "loss": 0.4787,
+      "step": 2860
+    },
+    {
+      "epoch": 0.6381887129154584,
+      "grad_norm": 0.16252587735652924,
+      "learning_rate": 1.796358191574854e-05,
+      "loss": 0.5075,
+      "step": 2861
+    },
+    {
+      "epoch": 0.6384117778273477,
+      "grad_norm": 0.1922154724597931,
+      "learning_rate": 1.7962158281721475e-05,
+      "loss": 0.4616,
+      "step": 2862
+    },
+    {
+      "epoch": 0.6386348427392371,
+      "grad_norm": 0.17007192969322205,
+      "learning_rate": 1.7960734206700002e-05,
+      "loss": 0.4741,
+      "step": 2863
+    },
+    {
+      "epoch": 0.6388579076511265,
+      "grad_norm": 0.16017115116119385,
+      "learning_rate": 1.7959309690762992e-05,
+      "loss": 0.4757,
+      "step": 2864
+    },
+    {
+      "epoch": 0.6390809725630159,
+      "grad_norm": 0.16466043889522552,
+      "learning_rate": 1.795788473398935e-05,
+      "loss": 0.4856,
+      "step": 2865
+    },
+    {
+      "epoch": 0.6393040374749052,
+      "grad_norm": 0.16355857253074646,
+      "learning_rate": 1.795645933645799e-05,
+      "loss": 0.5055,
+      "step": 2866
+    },
+    {
+      "epoch": 0.6395271023867946,
+      "grad_norm": 0.1593533307313919,
+      "learning_rate": 1.7955033498247863e-05,
+      "loss": 0.5103,
+      "step": 2867
+    },
+    {
+      "epoch": 0.639750167298684,
+      "grad_norm": 0.16536985337734222,
+      "learning_rate": 1.7953607219437942e-05,
+      "loss": 0.5029,
+      "step": 2868
+    },
+    {
+      "epoch": 0.6399732322105732,
+      "grad_norm": 0.1590854823589325,
+      "learning_rate": 1.7952180500107225e-05,
+      "loss": 0.4687,
+      "step": 2869
+    },
+    {
+      "epoch": 0.6401962971224626,
+      "grad_norm": 0.16659674048423767,
+      "learning_rate": 1.7950753340334734e-05,
+      "loss": 0.4999,
+      "step": 2870
+    },
+    {
+      "epoch": 0.640419362034352,
+      "grad_norm": 0.17274489998817444,
+      "learning_rate": 1.7949325740199507e-05,
+      "loss": 0.4995,
+      "step": 2871
+    },
+    {
+      "epoch": 0.6406424269462414,
+      "grad_norm": 0.16115543246269226,
+      "learning_rate": 1.7947897699780616e-05,
+      "loss": 0.484,
+      "step": 2872
+    },
+    {
+      "epoch": 0.6408654918581307,
+      "grad_norm": 0.16735753417015076,
+      "learning_rate": 1.7946469219157158e-05,
+      "loss": 0.4917,
+      "step": 2873
+    },
+    {
+      "epoch": 0.6410885567700201,
+      "grad_norm": 0.1667783111333847,
+      "learning_rate": 1.7945040298408248e-05,
+      "loss": 0.5106,
+      "step": 2874
+    },
+    {
+      "epoch": 0.6413116216819095,
+      "grad_norm": 0.16274797916412354,
+      "learning_rate": 1.794361093761303e-05,
+      "loss": 0.5006,
+      "step": 2875
+    },
+    {
+      "epoch": 0.6415346865937988,
+      "grad_norm": 0.15851566195487976,
+      "learning_rate": 1.7942181136850672e-05,
+      "loss": 0.5007,
+      "step": 2876
+    },
+    {
+      "epoch": 0.6417577515056881,
+      "grad_norm": 0.16172711551189423,
+      "learning_rate": 1.7940750896200363e-05,
+      "loss": 0.5043,
+      "step": 2877
+    },
+    {
+      "epoch": 0.6419808164175775,
+      "grad_norm": 0.18239612877368927,
+      "learning_rate": 1.7939320215741322e-05,
+      "loss": 0.4888,
+      "step": 2878
+    },
+    {
+      "epoch": 0.6422038813294669,
+      "grad_norm": 0.17973902821540833,
+      "learning_rate": 1.7937889095552787e-05,
+      "loss": 0.4813,
+      "step": 2879
+    },
+    {
+      "epoch": 0.6424269462413562,
+      "grad_norm": 0.1610574722290039,
+      "learning_rate": 1.7936457535714023e-05,
+      "loss": 0.5138,
+      "step": 2880
+    },
+    {
+      "epoch": 0.6426500111532456,
+      "grad_norm": 0.16427603363990784,
+      "learning_rate": 1.7935025536304317e-05,
+      "loss": 0.483,
+      "step": 2881
+    },
+    {
+      "epoch": 0.642873076065135,
+      "grad_norm": 0.18190504610538483,
+      "learning_rate": 1.7933593097402983e-05,
+      "loss": 0.4968,
+      "step": 2882
+    },
+    {
+      "epoch": 0.6430961409770243,
+      "grad_norm": 0.16263769567012787,
+      "learning_rate": 1.793216021908936e-05,
+      "loss": 0.5245,
+      "step": 2883
+    },
+    {
+      "epoch": 0.6433192058889137,
+      "grad_norm": 0.16960637271404266,
+      "learning_rate": 1.793072690144281e-05,
+      "loss": 0.5353,
+      "step": 2884
+    },
+    {
+      "epoch": 0.6435422708008031,
+      "grad_norm": 0.1611735224723816,
+      "learning_rate": 1.7929293144542715e-05,
+      "loss": 0.4701,
+      "step": 2885
+    },
+    {
+      "epoch": 0.6437653357126923,
+      "grad_norm": 0.16312386095523834,
+      "learning_rate": 1.792785894846849e-05,
+      "loss": 0.5167,
+      "step": 2886
+    },
+    {
+      "epoch": 0.6439884006245817,
+      "grad_norm": 0.17161010205745697,
+      "learning_rate": 1.7926424313299568e-05,
+      "loss": 0.5092,
+      "step": 2887
+    },
+    {
+      "epoch": 0.6442114655364711,
+      "grad_norm": 0.1609477698802948,
+      "learning_rate": 1.7924989239115407e-05,
+      "loss": 0.5295,
+      "step": 2888
+    },
+    {
+      "epoch": 0.6444345304483605,
+      "grad_norm": 0.1607387214899063,
+      "learning_rate": 1.7923553725995494e-05,
+      "loss": 0.5205,
+      "step": 2889
+    },
+    {
+      "epoch": 0.6446575953602498,
+      "grad_norm": 0.16367650032043457,
+      "learning_rate": 1.7922117774019333e-05,
+      "loss": 0.4875,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6448806602721392,
+      "grad_norm": 0.17014965415000916,
+      "learning_rate": 1.7920681383266458e-05,
+      "loss": 0.5218,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6451037251840286,
+      "grad_norm": 0.1545102298259735,
+      "learning_rate": 1.7919244553816426e-05,
+      "loss": 0.4954,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6453267900959179,
+      "grad_norm": 0.2512567937374115,
+      "learning_rate": 1.7917807285748817e-05,
+      "loss": 0.5173,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6455498550078073,
+      "grad_norm": 0.17359793186187744,
+      "learning_rate": 1.7916369579143235e-05,
+      "loss": 0.5253,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6457729199196967,
+      "grad_norm": 0.17602702975273132,
+      "learning_rate": 1.7914931434079305e-05,
+      "loss": 0.492,
+      "step": 2895
+    },
+    {
+      "epoch": 0.645995984831586,
+      "grad_norm": 0.1598963886499405,
+      "learning_rate": 1.791349285063669e-05,
+      "loss": 0.5071,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6462190497434753,
+      "grad_norm": 0.16161389648914337,
+      "learning_rate": 1.7912053828895064e-05,
+      "loss": 0.4896,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6464421146553647,
+      "grad_norm": 0.16500675678253174,
+      "learning_rate": 1.7910614368934127e-05,
+      "loss": 0.5142,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6466651795672541,
+      "grad_norm": 0.15490639209747314,
+      "learning_rate": 1.7909174470833604e-05,
+      "loss": 0.4708,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6468882444791434,
+      "grad_norm": 0.1680765300989151,
+      "learning_rate": 1.7907734134673252e-05,
+      "loss": 0.4978,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6471113093910328,
+      "grad_norm": 0.16419318318367004,
+      "learning_rate": 1.790629336053284e-05,
+      "loss": 0.5106,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6473343743029222,
+      "grad_norm": 0.1587122082710266,
+      "learning_rate": 1.790485214849217e-05,
+      "loss": 0.456,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6475574392148115,
+      "grad_norm": 0.17150086164474487,
+      "learning_rate": 1.7903410498631063e-05,
+      "loss": 0.5066,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6477805041267009,
+      "grad_norm": 0.16548244655132294,
+      "learning_rate": 1.790196841102937e-05,
+      "loss": 0.5297,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6480035690385902,
+      "grad_norm": 0.1526126116514206,
+      "learning_rate": 1.790052588576696e-05,
+      "loss": 0.4693,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6482266339504796,
+      "grad_norm": 0.16352921724319458,
+      "learning_rate": 1.7899082922923732e-05,
+      "loss": 0.5291,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6484496988623689,
+      "grad_norm": 0.16485454142093658,
+      "learning_rate": 1.78976395225796e-05,
+      "loss": 0.4752,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6486727637742583,
+      "grad_norm": 0.1610778123140335,
+      "learning_rate": 1.7896195684814516e-05,
+      "loss": 0.479,
+      "step": 2908
+    },
+    {
+      "epoch": 0.6488958286861477,
+      "grad_norm": 0.16191574931144714,
+      "learning_rate": 1.7894751409708447e-05,
+      "loss": 0.5059,
+      "step": 2909
+    },
+    {
+      "epoch": 0.649118893598037,
+      "grad_norm": 0.1598249077796936,
+      "learning_rate": 1.7893306697341385e-05,
+      "loss": 0.487,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6493419585099264,
+      "grad_norm": 0.1673557460308075,
+      "learning_rate": 1.7891861547793345e-05,
+      "loss": 0.4926,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6495650234218158,
+      "grad_norm": 0.16955086588859558,
+      "learning_rate": 1.789041596114437e-05,
+      "loss": 0.4912,
+      "step": 2912
+    },
+    {
+      "epoch": 0.6497880883337052,
+      "grad_norm": 0.15351925790309906,
+      "learning_rate": 1.788896993747453e-05,
+      "loss": 0.4758,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6500111532455944,
+      "grad_norm": 0.16994906961917877,
+      "learning_rate": 1.7887523476863907e-05,
+      "loss": 0.5038,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6502342181574838,
+      "grad_norm": 0.16765102744102478,
+      "learning_rate": 1.7886076579392622e-05,
+      "loss": 0.4787,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6504572830693732,
+      "grad_norm": 0.16875889897346497,
+      "learning_rate": 1.7884629245140812e-05,
+      "loss": 0.5267,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6506803479812625,
+      "grad_norm": 0.16783274710178375,
+      "learning_rate": 1.7883181474188637e-05,
+      "loss": 0.4913,
+      "step": 2917
+    },
+    {
+      "epoch": 0.6509034128931519,
+      "grad_norm": 0.16633883118629456,
+      "learning_rate": 1.7881733266616284e-05,
+      "loss": 0.5019,
+      "step": 2918
+    },
+    {
+      "epoch": 0.6511264778050413,
+      "grad_norm": 0.1654990315437317,
+      "learning_rate": 1.7880284622503966e-05,
+      "loss": 0.5033,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6513495427169306,
+      "grad_norm": 0.16043652594089508,
+      "learning_rate": 1.7878835541931915e-05,
+      "loss": 0.4957,
+      "step": 2920
+    },
+    {
+      "epoch": 0.65157260762882,
+      "grad_norm": 0.16924645006656647,
+      "learning_rate": 1.7877386024980392e-05,
+      "loss": 0.5223,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6517956725407094,
+      "grad_norm": 0.1635693907737732,
+      "learning_rate": 1.7875936071729682e-05,
+      "loss": 0.4908,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6520187374525988,
+      "grad_norm": 0.1630011945962906,
+      "learning_rate": 1.7874485682260087e-05,
+      "loss": 0.5138,
+      "step": 2923
+    },
+    {
+      "epoch": 0.652241802364488,
+      "grad_norm": 0.16332125663757324,
+      "learning_rate": 1.7873034856651944e-05,
+      "loss": 0.5119,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6524648672763774,
+      "grad_norm": 0.1802121102809906,
+      "learning_rate": 1.787158359498561e-05,
+      "loss": 0.4766,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6526879321882668,
+      "grad_norm": 0.16062189638614655,
+      "learning_rate": 1.7870131897341458e-05,
+      "loss": 0.4747,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6529109971001561,
+      "grad_norm": 0.16367454826831818,
+      "learning_rate": 1.7868679763799898e-05,
+      "loss": 0.4698,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6531340620120455,
+      "grad_norm": 0.16663846373558044,
+      "learning_rate": 1.786722719444136e-05,
+      "loss": 0.4903,
+      "step": 2928
+    },
+    {
+      "epoch": 0.6533571269239349,
+      "grad_norm": 0.16041219234466553,
+      "learning_rate": 1.786577418934629e-05,
+      "loss": 0.496,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6535801918358243,
+      "grad_norm": 0.1660057008266449,
+      "learning_rate": 1.7864320748595168e-05,
+      "loss": 0.497,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6538032567477136,
+      "grad_norm": 0.17890667915344238,
+      "learning_rate": 1.7862866872268493e-05,
+      "loss": 0.5183,
+      "step": 2931
+    },
+    {
+      "epoch": 0.654026321659603,
+      "grad_norm": 0.15913154184818268,
+      "learning_rate": 1.7861412560446794e-05,
+      "loss": 0.4827,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6542493865714923,
+      "grad_norm": 0.16374540328979492,
+      "learning_rate": 1.7859957813210614e-05,
+      "loss": 0.5119,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6544724514833816,
+      "grad_norm": 0.1748073399066925,
+      "learning_rate": 1.7858502630640533e-05,
+      "loss": 0.493,
+      "step": 2934
+    },
+    {
+      "epoch": 0.654695516395271,
+      "grad_norm": 0.1679995357990265,
+      "learning_rate": 1.7857047012817144e-05,
+      "loss": 0.4878,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6549185813071604,
+      "grad_norm": 0.16782502830028534,
+      "learning_rate": 1.7855590959821068e-05,
+      "loss": 0.4942,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6551416462190497,
+      "grad_norm": 0.1666112244129181,
+      "learning_rate": 1.785413447173295e-05,
+      "loss": 0.5109,
+      "step": 2937
+    },
+    {
+      "epoch": 0.6553647111309391,
+      "grad_norm": 0.16338278353214264,
+      "learning_rate": 1.785267754863346e-05,
+      "loss": 0.509,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6555877760428285,
+      "grad_norm": 0.16831496357917786,
+      "learning_rate": 1.7851220190603295e-05,
+      "loss": 0.4933,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6558108409547179,
+      "grad_norm": 0.16522802412509918,
+      "learning_rate": 1.7849762397723168e-05,
+      "loss": 0.5204,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6560339058666071,
+      "grad_norm": 0.1579723209142685,
+      "learning_rate": 1.7848304170073822e-05,
+      "loss": 0.4747,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6562569707784965,
+      "grad_norm": 0.16369852423667908,
+      "learning_rate": 1.784684550773602e-05,
+      "loss": 0.4763,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6564800356903859,
+      "grad_norm": 0.16621820628643036,
+      "learning_rate": 1.7845386410790558e-05,
+      "loss": 0.5113,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6567031006022752,
+      "grad_norm": 0.161569282412529,
+      "learning_rate": 1.784392687931825e-05,
+      "loss": 0.4899,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6569261655141646,
+      "grad_norm": 0.1616351306438446,
+      "learning_rate": 1.7842466913399928e-05,
+      "loss": 0.4879,
+      "step": 2945
+    },
+    {
+      "epoch": 0.657149230426054,
+      "grad_norm": 0.16426199674606323,
+      "learning_rate": 1.7841006513116456e-05,
+      "loss": 0.5265,
+      "step": 2946
+    },
+    {
+      "epoch": 0.6573722953379434,
+      "grad_norm": 0.1592702567577362,
+      "learning_rate": 1.7839545678548727e-05,
+      "loss": 0.5036,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6575953602498327,
+      "grad_norm": 0.15806902945041656,
+      "learning_rate": 1.7838084409777637e-05,
+      "loss": 0.4954,
+      "step": 2948
+    },
+    {
+      "epoch": 0.6578184251617221,
+      "grad_norm": 0.17478401958942413,
+      "learning_rate": 1.7836622706884138e-05,
+      "loss": 0.5046,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6580414900736115,
+      "grad_norm": 0.16117317974567413,
+      "learning_rate": 1.7835160569949174e-05,
+      "loss": 0.4925,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6582645549855007,
+      "grad_norm": 0.1660463809967041,
+      "learning_rate": 1.783369799905373e-05,
+      "loss": 0.5003,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6584876198973901,
+      "grad_norm": 0.1700252741575241,
+      "learning_rate": 1.7832234994278822e-05,
+      "loss": 0.5169,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6587106848092795,
+      "grad_norm": 0.17790904641151428,
+      "learning_rate": 1.7830771555705468e-05,
+      "loss": 0.5088,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6589337497211689,
+      "grad_norm": 0.16493864357471466,
+      "learning_rate": 1.782930768341473e-05,
+      "loss": 0.5082,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6591568146330582,
+      "grad_norm": 0.16437609493732452,
+      "learning_rate": 1.7827843377487683e-05,
+      "loss": 0.4984,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6593798795449476,
+      "grad_norm": 0.16439390182495117,
+      "learning_rate": 1.7826378638005432e-05,
+      "loss": 0.5005,
+      "step": 2956
+    },
+    {
+      "epoch": 0.659602944456837,
+      "grad_norm": 0.15785926580429077,
+      "learning_rate": 1.78249134650491e-05,
+      "loss": 0.5005,
+      "step": 2957
+    },
+    {
+      "epoch": 0.6598260093687263,
+      "grad_norm": 0.20435695350170135,
+      "learning_rate": 1.782344785869984e-05,
+      "loss": 0.4711,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6600490742806157,
+      "grad_norm": 0.15359720587730408,
+      "learning_rate": 1.7821981819038828e-05,
+      "loss": 0.4738,
+      "step": 2959
+    },
+    {
+      "epoch": 0.660272139192505,
+      "grad_norm": 0.17044633626937866,
+      "learning_rate": 1.7820515346147262e-05,
+      "loss": 0.4951,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6604952041043943,
+      "grad_norm": 0.17282849550247192,
+      "learning_rate": 1.781904844010636e-05,
+      "loss": 0.496,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6607182690162837,
+      "grad_norm": 0.1486678570508957,
+      "learning_rate": 1.7817581100997374e-05,
+      "loss": 0.4671,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6609413339281731,
+      "grad_norm": 0.16091682016849518,
+      "learning_rate": 1.781611332890157e-05,
+      "loss": 0.4824,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6611643988400625,
+      "grad_norm": 0.15881048142910004,
+      "learning_rate": 1.7814645123900246e-05,
+      "loss": 0.4864,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6613874637519518,
+      "grad_norm": 0.17700685560703278,
+      "learning_rate": 1.781317648607472e-05,
+      "loss": 0.4962,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6616105286638412,
+      "grad_norm": 0.15473395586013794,
+      "learning_rate": 1.781170741550633e-05,
+      "loss": 0.4737,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6618335935757306,
+      "grad_norm": 0.1670142114162445,
+      "learning_rate": 1.781023791227645e-05,
+      "loss": 0.5105,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6620566584876199,
+      "grad_norm": 0.18345218896865845,
+      "learning_rate": 1.780876797646646e-05,
+      "loss": 0.5086,
+      "step": 2968
+    },
+    {
+      "epoch": 0.6622797233995092,
+      "grad_norm": 0.16229334473609924,
+      "learning_rate": 1.7807297608157784e-05,
+      "loss": 0.5007,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6625027883113986,
+      "grad_norm": 0.16335418820381165,
+      "learning_rate": 1.7805826807431856e-05,
+      "loss": 0.4976,
+      "step": 2970
+    },
+    {
+      "epoch": 0.662725853223288,
+      "grad_norm": 0.16784057021141052,
+      "learning_rate": 1.780435557437014e-05,
+      "loss": 0.5106,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6629489181351773,
+      "grad_norm": 0.16552112996578217,
+      "learning_rate": 1.7802883909054118e-05,
+      "loss": 0.5087,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6631719830470667,
+      "grad_norm": 0.206945538520813,
+      "learning_rate": 1.7801411811565308e-05,
+      "loss": 0.4753,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6633950479589561,
+      "grad_norm": 0.1575002670288086,
+      "learning_rate": 1.7799939281985236e-05,
+      "loss": 0.4875,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6636181128708454,
+      "grad_norm": 0.1623086929321289,
+      "learning_rate": 1.7798466320395463e-05,
+      "loss": 0.4933,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6638411777827348,
+      "grad_norm": 0.1668468862771988,
+      "learning_rate": 1.779699292687757e-05,
+      "loss": 0.462,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6640642426946242,
+      "grad_norm": 0.43826133012771606,
+      "learning_rate": 1.7795519101513166e-05,
+      "loss": 0.5073,
+      "step": 2977
+    },
+    {
+      "epoch": 0.6642873076065134,
+      "grad_norm": 0.18710243701934814,
+      "learning_rate": 1.779404484438388e-05,
+      "loss": 0.5103,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6645103725184028,
+      "grad_norm": 0.17884770035743713,
+      "learning_rate": 1.7792570155571358e-05,
+      "loss": 0.5219,
+      "step": 2979
+    },
+    {
+      "epoch": 0.6647334374302922,
+      "grad_norm": 0.1652306616306305,
+      "learning_rate": 1.7791095035157288e-05,
+      "loss": 0.513,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6649565023421816,
+      "grad_norm": 0.175328329205513,
+      "learning_rate": 1.7789619483223367e-05,
+      "loss": 0.5064,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6651795672540709,
+      "grad_norm": 0.15750913321971893,
+      "learning_rate": 1.7788143499851318e-05,
+      "loss": 0.4608,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6654026321659603,
+      "grad_norm": 0.15582314133644104,
+      "learning_rate": 1.7786667085122895e-05,
+      "loss": 0.4793,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6656256970778497,
+      "grad_norm": 0.18893486261367798,
+      "learning_rate": 1.7785190239119864e-05,
+      "loss": 0.5277,
+      "step": 2984
+    },
+    {
+      "epoch": 0.665848761989739,
+      "grad_norm": 0.16801688075065613,
+      "learning_rate": 1.7783712961924032e-05,
+      "loss": 0.5281,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6660718269016284,
+      "grad_norm": 0.16510803997516632,
+      "learning_rate": 1.778223525361721e-05,
+      "loss": 0.4675,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6662948918135178,
+      "grad_norm": 0.16860422492027283,
+      "learning_rate": 1.778075711428125e-05,
+      "loss": 0.4712,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6665179567254071,
+      "grad_norm": 0.17091190814971924,
+      "learning_rate": 1.777927854399802e-05,
+      "loss": 0.5208,
+      "step": 2988
+    },
+    {
+      "epoch": 0.6667410216372964,
+      "grad_norm": 0.19528479874134064,
+      "learning_rate": 1.7777799542849408e-05,
+      "loss": 0.482,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6669640865491858,
+      "grad_norm": 0.16672208905220032,
+      "learning_rate": 1.7776320110917334e-05,
+      "loss": 0.5002,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6671871514610752,
+      "grad_norm": 0.1595972180366516,
+      "learning_rate": 1.777484024828374e-05,
+      "loss": 0.4868,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6674102163729645,
+      "grad_norm": 0.16304026544094086,
+      "learning_rate": 1.7773359955030583e-05,
+      "loss": 0.4839,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6676332812848539,
+      "grad_norm": 0.17252400517463684,
+      "learning_rate": 1.7771879231239857e-05,
+      "loss": 0.4737,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6678563461967433,
+      "grad_norm": 0.17409226298332214,
+      "learning_rate": 1.777039807699357e-05,
+      "loss": 0.5152,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6680794111086326,
+      "grad_norm": 0.15897680819034576,
+      "learning_rate": 1.7768916492373763e-05,
+      "loss": 0.5081,
+      "step": 2995
+    },
+    {
+      "epoch": 0.668302476020522,
+      "grad_norm": 0.15727804601192474,
+      "learning_rate": 1.7767434477462493e-05,
+      "loss": 0.4902,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6685255409324113,
+      "grad_norm": 0.16594363749027252,
+      "learning_rate": 1.776595203234184e-05,
+      "loss": 0.4982,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6687486058443007,
+      "grad_norm": 0.16226732730865479,
+      "learning_rate": 1.7764469157093916e-05,
+      "loss": 0.5304,
+      "step": 2998
+    },
+    {
+      "epoch": 0.66897167075619,
+      "grad_norm": 0.15772786736488342,
+      "learning_rate": 1.7762985851800846e-05,
+      "loss": 0.4707,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6691947356680794,
+      "grad_norm": 0.19541241228580475,
+      "learning_rate": 1.776150211654479e-05,
+      "loss": 0.4826,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6694178005799688,
+      "grad_norm": 0.17381907999515533,
+      "learning_rate": 1.7760017951407924e-05,
+      "loss": 0.5304,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6696408654918581,
+      "grad_norm": 0.16170580685138702,
+      "learning_rate": 1.7758533356472454e-05,
+      "loss": 0.4722,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6698639304037475,
+      "grad_norm": 0.16046100854873657,
+      "learning_rate": 1.7757048331820604e-05,
+      "loss": 0.4852,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6700869953156369,
+      "grad_norm": 0.17672637104988098,
+      "learning_rate": 1.775556287753462e-05,
+      "loss": 0.5,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6703100602275263,
+      "grad_norm": 0.15894585847854614,
+      "learning_rate": 1.7754076993696784e-05,
+      "loss": 0.4819,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6705331251394155,
+      "grad_norm": 0.16397017240524292,
+      "learning_rate": 1.7752590680389382e-05,
+      "loss": 0.5348,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6707561900513049,
+      "grad_norm": 0.17470906674861908,
+      "learning_rate": 1.7751103937694748e-05,
+      "loss": 0.5056,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6709792549631943,
+      "grad_norm": 0.16147972643375397,
+      "learning_rate": 1.774961676569522e-05,
+      "loss": 0.4845,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6712023198750836,
+      "grad_norm": 0.15075084567070007,
+      "learning_rate": 1.774812916447317e-05,
+      "loss": 0.4873,
+      "step": 3009
+    },
+    {
+      "epoch": 0.671425384786973,
+      "grad_norm": 0.1605006605386734,
+      "learning_rate": 1.774664113411099e-05,
+      "loss": 0.4885,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6716484496988624,
+      "grad_norm": 0.17120316624641418,
+      "learning_rate": 1.7745152674691093e-05,
+      "loss": 0.4952,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6718715146107517,
+      "grad_norm": 0.17164325714111328,
+      "learning_rate": 1.774366378629592e-05,
+      "loss": 0.5225,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6720945795226411,
+      "grad_norm": 0.16912591457366943,
+      "learning_rate": 1.774217446900794e-05,
+      "loss": 0.5165,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6723176444345305,
+      "grad_norm": 0.1592075079679489,
+      "learning_rate": 1.7740684722909638e-05,
+      "loss": 0.481,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6725407093464199,
+      "grad_norm": 0.15714260935783386,
+      "learning_rate": 1.7739194548083526e-05,
+      "loss": 0.5122,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6727637742583091,
+      "grad_norm": 0.15991204977035522,
+      "learning_rate": 1.7737703944612135e-05,
+      "loss": 0.5006,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6729868391701985,
+      "grad_norm": 0.15693789720535278,
+      "learning_rate": 1.7736212912578028e-05,
+      "loss": 0.4867,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6732099040820879,
+      "grad_norm": 0.15840266644954681,
+      "learning_rate": 1.773472145206379e-05,
+      "loss": 0.4937,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6734329689939772,
+      "grad_norm": 0.16329781711101532,
+      "learning_rate": 1.7733229563152024e-05,
+      "loss": 0.489,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6736560339058666,
+      "grad_norm": 0.17248262465000153,
+      "learning_rate": 1.7731737245925357e-05,
+      "loss": 0.5112,
+      "step": 3020
+    },
+    {
+      "epoch": 0.673879098817756,
+      "grad_norm": 0.1815134584903717,
+      "learning_rate": 1.7730244500466454e-05,
+      "loss": 0.5027,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6741021637296454,
+      "grad_norm": 0.17082563042640686,
+      "learning_rate": 1.772875132685798e-05,
+      "loss": 0.5018,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6743252286415347,
+      "grad_norm": 0.17142242193222046,
+      "learning_rate": 1.772725772518264e-05,
+      "loss": 0.5106,
+      "step": 3023
+    },
+    {
+      "epoch": 0.674548293553424,
+      "grad_norm": 0.17516618967056274,
+      "learning_rate": 1.7725763695523166e-05,
+      "loss": 0.4743,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6747713584653134,
+      "grad_norm": 0.16038702428340912,
+      "learning_rate": 1.77242692379623e-05,
+      "loss": 0.5071,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6749944233772027,
+      "grad_norm": 0.16445225477218628,
+      "learning_rate": 1.7722774352582816e-05,
+      "loss": 0.4992,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6752174882890921,
+      "grad_norm": 0.16889688372612,
+      "learning_rate": 1.772127903946751e-05,
+      "loss": 0.4909,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6754405532009815,
+      "grad_norm": 0.17117194831371307,
+      "learning_rate": 1.77197832986992e-05,
+      "loss": 0.4927,
+      "step": 3028
+    },
+    {
+      "epoch": 0.6756636181128709,
+      "grad_norm": 0.16765987873077393,
+      "learning_rate": 1.7718287130360733e-05,
+      "loss": 0.5065,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6758866830247602,
+      "grad_norm": 0.1618220955133438,
+      "learning_rate": 1.7716790534534977e-05,
+      "loss": 0.4931,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6761097479366496,
+      "grad_norm": 0.16639363765716553,
+      "learning_rate": 1.7715293511304815e-05,
+      "loss": 0.5044,
+      "step": 3031
+    },
+    {
+      "epoch": 0.676332812848539,
+      "grad_norm": 0.167351633310318,
+      "learning_rate": 1.7713796060753173e-05,
+      "loss": 0.5188,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6765558777604282,
+      "grad_norm": 0.1656593531370163,
+      "learning_rate": 1.771229818296298e-05,
+      "loss": 0.5271,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6767789426723176,
+      "grad_norm": 0.168874591588974,
+      "learning_rate": 1.7710799878017203e-05,
+      "loss": 0.4947,
+      "step": 3034
+    },
+    {
+      "epoch": 0.677002007584207,
+      "grad_norm": 0.16307704150676727,
+      "learning_rate": 1.7709301145998827e-05,
+      "loss": 0.5071,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6772250724960963,
+      "grad_norm": 0.1524697244167328,
+      "learning_rate": 1.7707801986990857e-05,
+      "loss": 0.497,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6774481374079857,
+      "grad_norm": 0.16599062085151672,
+      "learning_rate": 1.7706302401076327e-05,
+      "loss": 0.4726,
+      "step": 3037
+    },
+    {
+      "epoch": 0.6776712023198751,
+      "grad_norm": 0.16413775086402893,
+      "learning_rate": 1.77048023883383e-05,
+      "loss": 0.527,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6778942672317645,
+      "grad_norm": 0.1584494262933731,
+      "learning_rate": 1.770330194885985e-05,
+      "loss": 0.5095,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6781173321436538,
+      "grad_norm": 0.15526416897773743,
+      "learning_rate": 1.7701801082724084e-05,
+      "loss": 0.4914,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6783403970555432,
+      "grad_norm": 0.16402754187583923,
+      "learning_rate": 1.7700299790014126e-05,
+      "loss": 0.5184,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6785634619674326,
+      "grad_norm": 0.16508348286151886,
+      "learning_rate": 1.769879807081313e-05,
+      "loss": 0.4783,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6787865268793218,
+      "grad_norm": 0.15416789054870605,
+      "learning_rate": 1.769729592520427e-05,
+      "loss": 0.4818,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6790095917912112,
+      "grad_norm": 0.27602240443229675,
+      "learning_rate": 1.769579335327074e-05,
+      "loss": 0.49,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6792326567031006,
+      "grad_norm": 0.15309108793735504,
+      "learning_rate": 1.7694290355095768e-05,
+      "loss": 0.4778,
+      "step": 3045
+    },
+    {
+      "epoch": 0.67945572161499,
+      "grad_norm": 0.15830327570438385,
+      "learning_rate": 1.76927869307626e-05,
+      "loss": 0.4909,
+      "step": 3046
+    },
+    {
+      "epoch": 0.6796787865268793,
+      "grad_norm": 0.16737417876720428,
+      "learning_rate": 1.76912830803545e-05,
+      "loss": 0.4642,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6799018514387687,
+      "grad_norm": 0.16713353991508484,
+      "learning_rate": 1.7689778803954764e-05,
+      "loss": 0.5076,
+      "step": 3048
+    },
+    {
+      "epoch": 0.6801249163506581,
+      "grad_norm": 0.15644113719463348,
+      "learning_rate": 1.7688274101646702e-05,
+      "loss": 0.4746,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6803479812625474,
+      "grad_norm": 0.1714516133069992,
+      "learning_rate": 1.7686768973513663e-05,
+      "loss": 0.4924,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6805710461744368,
+      "grad_norm": 0.1625642329454422,
+      "learning_rate": 1.7685263419639008e-05,
+      "loss": 0.5103,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6807941110863261,
+      "grad_norm": 0.15343210101127625,
+      "learning_rate": 1.768375744010612e-05,
+      "loss": 0.4871,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6810171759982154,
+      "grad_norm": 0.1622699499130249,
+      "learning_rate": 1.7682251034998413e-05,
+      "loss": 0.5093,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6812402409101048,
+      "grad_norm": 0.15701012313365936,
+      "learning_rate": 1.768074420439932e-05,
+      "loss": 0.5052,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6814633058219942,
+      "grad_norm": 0.1647741198539734,
+      "learning_rate": 1.76792369483923e-05,
+      "loss": 0.528,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6816863707338836,
+      "grad_norm": 0.16413739323616028,
+      "learning_rate": 1.7677729267060836e-05,
+      "loss": 0.5029,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6819094356457729,
+      "grad_norm": 0.1663079708814621,
+      "learning_rate": 1.7676221160488426e-05,
+      "loss": 0.4961,
+      "step": 3057
+    },
+    {
+      "epoch": 0.6821325005576623,
+      "grad_norm": 0.1598016768693924,
+      "learning_rate": 1.7674712628758603e-05,
+      "loss": 0.5201,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6823555654695517,
+      "grad_norm": 0.18910476565361023,
+      "learning_rate": 1.767320367195492e-05,
+      "loss": 0.5203,
+      "step": 3059
+    },
+    {
+      "epoch": 0.682578630381441,
+      "grad_norm": 0.1525489091873169,
+      "learning_rate": 1.767169429016095e-05,
+      "loss": 0.4609,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6828016952933303,
+      "grad_norm": 0.22233475744724274,
+      "learning_rate": 1.7670184483460296e-05,
+      "loss": 0.4936,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6830247602052197,
+      "grad_norm": 0.16714778542518616,
+      "learning_rate": 1.766867425193658e-05,
+      "loss": 0.5248,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6832478251171091,
+      "grad_norm": 0.1673208624124527,
+      "learning_rate": 1.766716359567344e-05,
+      "loss": 0.5056,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6834708900289984,
+      "grad_norm": 0.1797683835029602,
+      "learning_rate": 1.7665652514754554e-05,
+      "loss": 0.5039,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6836939549408878,
+      "grad_norm": 0.1645476371049881,
+      "learning_rate": 1.7664141009263614e-05,
+      "loss": 0.4907,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6839170198527772,
+      "grad_norm": 0.15605735778808594,
+      "learning_rate": 1.7662629079284336e-05,
+      "loss": 0.4821,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6841400847646665,
+      "grad_norm": 0.16608816385269165,
+      "learning_rate": 1.7661116724900456e-05,
+      "loss": 0.5028,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6843631496765559,
+      "grad_norm": 0.16926267743110657,
+      "learning_rate": 1.7659603946195746e-05,
+      "loss": 0.4933,
+      "step": 3068
+    },
+    {
+      "epoch": 0.6845862145884453,
+      "grad_norm": 0.15846210718154907,
+      "learning_rate": 1.7658090743253985e-05,
+      "loss": 0.4991,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6848092795003345,
+      "grad_norm": 0.16848038136959076,
+      "learning_rate": 1.7656577116158988e-05,
+      "loss": 0.5002,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6850323444122239,
+      "grad_norm": 0.16113083064556122,
+      "learning_rate": 1.765506306499459e-05,
+      "loss": 0.4999,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6852554093241133,
+      "grad_norm": 0.1538555920124054,
+      "learning_rate": 1.7653548589844648e-05,
+      "loss": 0.4812,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6854784742360027,
+      "grad_norm": 0.16473166644573212,
+      "learning_rate": 1.765203369079304e-05,
+      "loss": 0.4885,
+      "step": 3073
+    },
+    {
+      "epoch": 0.685701539147892,
+      "grad_norm": 0.15544745326042175,
+      "learning_rate": 1.765051836792367e-05,
+      "loss": 0.4901,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6859246040597814,
+      "grad_norm": 0.1728777289390564,
+      "learning_rate": 1.764900262132048e-05,
+      "loss": 0.4578,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6861476689716708,
+      "grad_norm": 0.16069582104682922,
+      "learning_rate": 1.76474864510674e-05,
+      "loss": 0.4904,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6863707338835601,
+      "grad_norm": 0.15537656843662262,
+      "learning_rate": 1.764596985724842e-05,
+      "loss": 0.4899,
+      "step": 3077
+    },
+    {
+      "epoch": 0.6865937987954495,
+      "grad_norm": 0.16955626010894775,
+      "learning_rate": 1.7644452839947536e-05,
+      "loss": 0.5011,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6868168637073389,
+      "grad_norm": 0.16193504631519318,
+      "learning_rate": 1.7642935399248765e-05,
+      "loss": 0.5128,
+      "step": 3079
+    },
+    {
+      "epoch": 0.6870399286192282,
+      "grad_norm": 0.1631263643503189,
+      "learning_rate": 1.7641417535236155e-05,
+      "loss": 0.5026,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6872629935311175,
+      "grad_norm": 0.18346811830997467,
+      "learning_rate": 1.7639899247993775e-05,
+      "loss": 0.4857,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6874860584430069,
+      "grad_norm": 0.16757658123970032,
+      "learning_rate": 1.7638380537605722e-05,
+      "loss": 0.4935,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6877091233548963,
+      "grad_norm": 0.18027780950069427,
+      "learning_rate": 1.7636861404156106e-05,
+      "loss": 0.5448,
+      "step": 3083
+    },
+    {
+      "epoch": 0.6879321882667856,
+      "grad_norm": 0.17672498524188995,
+      "learning_rate": 1.763534184772907e-05,
+      "loss": 0.5054,
+      "step": 3084
+    },
+    {
+      "epoch": 0.688155253178675,
+      "grad_norm": 0.15495164692401886,
+      "learning_rate": 1.763382186840877e-05,
+      "loss": 0.4778,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6883783180905644,
+      "grad_norm": 0.16750416159629822,
+      "learning_rate": 1.76323014662794e-05,
+      "loss": 0.5147,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6886013830024537,
+      "grad_norm": 0.16243582963943481,
+      "learning_rate": 1.763078064142516e-05,
+      "loss": 0.5029,
+      "step": 3087
+    },
+    {
+      "epoch": 0.688824447914343,
+      "grad_norm": 0.1765555739402771,
+      "learning_rate": 1.7629259393930292e-05,
+      "loss": 0.5402,
+      "step": 3088
+    },
+    {
+      "epoch": 0.6890475128262324,
+      "grad_norm": 0.17327693104743958,
+      "learning_rate": 1.7627737723879048e-05,
+      "loss": 0.5223,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6892705777381218,
+      "grad_norm": 0.16204944252967834,
+      "learning_rate": 1.762621563135571e-05,
+      "loss": 0.4926,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6894936426500111,
+      "grad_norm": 0.16001787781715393,
+      "learning_rate": 1.762469311644458e-05,
+      "loss": 0.5038,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6897167075619005,
+      "grad_norm": 0.15672995150089264,
+      "learning_rate": 1.7623170179229982e-05,
+      "loss": 0.4752,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6899397724737899,
+      "grad_norm": 0.16134774684906006,
+      "learning_rate": 1.7621646819796264e-05,
+      "loss": 0.4911,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6901628373856792,
+      "grad_norm": 0.18240399658679962,
+      "learning_rate": 1.762012303822781e-05,
+      "loss": 0.4956,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6903859022975686,
+      "grad_norm": 0.16389527916908264,
+      "learning_rate": 1.761859883460901e-05,
+      "loss": 0.4713,
+      "step": 3095
+    },
+    {
+      "epoch": 0.690608967209458,
+      "grad_norm": 0.16408023238182068,
+      "learning_rate": 1.761707420902428e-05,
+      "loss": 0.4925,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6908320321213474,
+      "grad_norm": 0.16707156598567963,
+      "learning_rate": 1.761554916155807e-05,
+      "loss": 0.5094,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6910550970332366,
+      "grad_norm": 0.1580830216407776,
+      "learning_rate": 1.7614023692294838e-05,
+      "loss": 0.4896,
+      "step": 3098
+    },
+    {
+      "epoch": 0.691278161945126,
+      "grad_norm": 0.1949886530637741,
+      "learning_rate": 1.7612497801319084e-05,
+      "loss": 0.4919,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6915012268570154,
+      "grad_norm": 0.1662502884864807,
+      "learning_rate": 1.7610971488715315e-05,
+      "loss": 0.5104,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6917242917689047,
+      "grad_norm": 0.15772640705108643,
+      "learning_rate": 1.760944475456807e-05,
+      "loss": 0.4699,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6919473566807941,
+      "grad_norm": 0.17815501987934113,
+      "learning_rate": 1.760791759896191e-05,
+      "loss": 0.5133,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6921704215926835,
+      "grad_norm": 0.16980375349521637,
+      "learning_rate": 1.760639002198142e-05,
+      "loss": 0.4874,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6923934865045729,
+      "grad_norm": 0.16098888218402863,
+      "learning_rate": 1.7604862023711204e-05,
+      "loss": 0.508,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6926165514164622,
+      "grad_norm": 0.16175471246242523,
+      "learning_rate": 1.760333360423589e-05,
+      "loss": 0.4963,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6928396163283516,
+      "grad_norm": 0.1580919772386551,
+      "learning_rate": 1.7601804763640137e-05,
+      "loss": 0.4674,
+      "step": 3106
+    },
+    {
+      "epoch": 0.693062681240241,
+      "grad_norm": 0.15538008511066437,
+      "learning_rate": 1.7600275502008618e-05,
+      "loss": 0.4907,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6932857461521302,
+      "grad_norm": 0.15990300476551056,
+      "learning_rate": 1.7598745819426034e-05,
+      "loss": 0.5035,
+      "step": 3108
+    },
+    {
+      "epoch": 0.6935088110640196,
+      "grad_norm": 0.1725892871618271,
+      "learning_rate": 1.759721571597711e-05,
+      "loss": 0.5196,
+      "step": 3109
+    },
+    {
+      "epoch": 0.693731875975909,
+      "grad_norm": 0.16250944137573242,
+      "learning_rate": 1.7595685191746586e-05,
+      "loss": 0.4907,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6939549408877983,
+      "grad_norm": 0.16884207725524902,
+      "learning_rate": 1.759415424681924e-05,
+      "loss": 0.535,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6941780057996877,
+      "grad_norm": 0.1701684594154358,
+      "learning_rate": 1.7592622881279867e-05,
+      "loss": 0.4878,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6944010707115771,
+      "grad_norm": 0.17077569663524628,
+      "learning_rate": 1.7591091095213277e-05,
+      "loss": 0.4851,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6946241356234665,
+      "grad_norm": 0.33712977170944214,
+      "learning_rate": 1.758955888870431e-05,
+      "loss": 0.476,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6948472005353558,
+      "grad_norm": 0.16534623503684998,
+      "learning_rate": 1.7588026261837833e-05,
+      "loss": 0.4962,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6950702654472452,
+      "grad_norm": 0.17375873029232025,
+      "learning_rate": 1.758649321469873e-05,
+      "loss": 0.4883,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6952933303591345,
+      "grad_norm": 0.16607031226158142,
+      "learning_rate": 1.758495974737191e-05,
+      "loss": 0.4727,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6955163952710238,
+      "grad_norm": 0.16579467058181763,
+      "learning_rate": 1.7583425859942312e-05,
+      "loss": 0.5156,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6957394601829132,
+      "grad_norm": 0.16067442297935486,
+      "learning_rate": 1.7581891552494886e-05,
+      "loss": 0.4797,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6959625250948026,
+      "grad_norm": 0.16372907161712646,
+      "learning_rate": 1.7580356825114616e-05,
+      "loss": 0.4918,
+      "step": 3120
+    },
+    {
+      "epoch": 0.696185590006692,
+      "grad_norm": 0.1595798134803772,
+      "learning_rate": 1.75788216778865e-05,
+      "loss": 0.4963,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6964086549185813,
+      "grad_norm": 0.16394098103046417,
+      "learning_rate": 1.757728611089557e-05,
+      "loss": 0.5189,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6966317198304707,
+      "grad_norm": 0.16349278390407562,
+      "learning_rate": 1.757575012422687e-05,
+      "loss": 0.5392,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6968547847423601,
+      "grad_norm": 0.16474361717700958,
+      "learning_rate": 1.7574213717965473e-05,
+      "loss": 0.498,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6970778496542493,
+      "grad_norm": 0.16759975254535675,
+      "learning_rate": 1.757267689219648e-05,
+      "loss": 0.4909,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6973009145661387,
+      "grad_norm": 0.18048445880413055,
+      "learning_rate": 1.7571139647005004e-05,
+      "loss": 0.5313,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6975239794780281,
+      "grad_norm": 0.21055570244789124,
+      "learning_rate": 1.7569601982476194e-05,
+      "loss": 0.4773,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6977470443899174,
+      "grad_norm": 0.16457606852054596,
+      "learning_rate": 1.7568063898695205e-05,
+      "loss": 0.4817,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6979701093018068,
+      "grad_norm": 0.15786704421043396,
+      "learning_rate": 1.7566525395747237e-05,
+      "loss": 0.4793,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6981931742136962,
+      "grad_norm": 0.17408685386180878,
+      "learning_rate": 1.7564986473717498e-05,
+      "loss": 0.5238,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6984162391255856,
+      "grad_norm": 0.1875070482492447,
+      "learning_rate": 1.7563447132691222e-05,
+      "loss": 0.5133,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6986393040374749,
+      "grad_norm": 0.15555784106254578,
+      "learning_rate": 1.7561907372753665e-05,
+      "loss": 0.4851,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6988623689493643,
+      "grad_norm": 0.15845829248428345,
+      "learning_rate": 1.756036719399011e-05,
+      "loss": 0.4989,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6990854338612537,
+      "grad_norm": 0.20709070563316345,
+      "learning_rate": 1.7558826596485866e-05,
+      "loss": 0.4939,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6993084987731429,
+      "grad_norm": 0.17586678266525269,
+      "learning_rate": 1.755728558032626e-05,
+      "loss": 0.5028,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6995315636850323,
+      "grad_norm": 0.16700804233551025,
+      "learning_rate": 1.7555744145596638e-05,
+      "loss": 0.4955,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6997546285969217,
+      "grad_norm": 0.17724819481372833,
+      "learning_rate": 1.755420229238238e-05,
+      "loss": 0.4571,
+      "step": 3137
+    },
+    {
+      "epoch": 0.6999776935088111,
+      "grad_norm": 0.17517662048339844,
+      "learning_rate": 1.755266002076888e-05,
+      "loss": 0.5144,
+      "step": 3138
+    },
+    {
+      "epoch": 0.7002007584207004,
+      "grad_norm": 0.16020509600639343,
+      "learning_rate": 1.755111733084156e-05,
+      "loss": 0.4523,
+      "step": 3139
+    },
+    {
+      "epoch": 0.7004238233325898,
+      "grad_norm": 0.15964041650295258,
+      "learning_rate": 1.7549574222685864e-05,
+      "loss": 0.4695,
+      "step": 3140
+    },
+    {
+      "epoch": 0.7006468882444792,
+      "grad_norm": 0.17395976185798645,
+      "learning_rate": 1.754803069638726e-05,
+      "loss": 0.515,
+      "step": 3141
+    },
+    {
+      "epoch": 0.7008699531563685,
+      "grad_norm": 0.15984676778316498,
+      "learning_rate": 1.7546486752031237e-05,
+      "loss": 0.4909,
+      "step": 3142
+    },
+    {
+      "epoch": 0.7010930180682579,
+      "grad_norm": 0.16784314811229706,
+      "learning_rate": 1.7544942389703305e-05,
+      "loss": 0.4973,
+      "step": 3143
+    },
+    {
+      "epoch": 0.7013160829801472,
+      "grad_norm": 0.16454213857650757,
+      "learning_rate": 1.754339760948901e-05,
+      "loss": 0.503,
+      "step": 3144
+    },
+    {
+      "epoch": 0.7015391478920365,
+      "grad_norm": 0.1578417867422104,
+      "learning_rate": 1.7541852411473902e-05,
+      "loss": 0.4632,
+      "step": 3145
+    },
+    {
+      "epoch": 0.7017622128039259,
+      "grad_norm": 0.16670000553131104,
+      "learning_rate": 1.7540306795743566e-05,
+      "loss": 0.4804,
+      "step": 3146
+    },
+    {
+      "epoch": 0.7019852777158153,
+      "grad_norm": 0.15188740193843842,
+      "learning_rate": 1.753876076238361e-05,
+      "loss": 0.4884,
+      "step": 3147
+    },
+    {
+      "epoch": 0.7022083426277047,
+      "grad_norm": 0.16512979567050934,
+      "learning_rate": 1.7537214311479663e-05,
+      "loss": 0.5002,
+      "step": 3148
+    },
+    {
+      "epoch": 0.702431407539594,
+      "grad_norm": 0.15965288877487183,
+      "learning_rate": 1.7535667443117377e-05,
+      "loss": 0.4885,
+      "step": 3149
+    },
+    {
+      "epoch": 0.7026544724514834,
+      "grad_norm": 0.1552652269601822,
+      "learning_rate": 1.7534120157382425e-05,
+      "loss": 0.4955,
+      "step": 3150
+    },
+    {
+      "epoch": 0.7028775373633728,
+      "grad_norm": 0.16739974915981293,
+      "learning_rate": 1.7532572454360506e-05,
+      "loss": 0.4902,
+      "step": 3151
+    },
+    {
+      "epoch": 0.703100602275262,
+      "grad_norm": 0.18262702226638794,
+      "learning_rate": 1.7531024334137348e-05,
+      "loss": 0.5283,
+      "step": 3152
+    },
+    {
+      "epoch": 0.7033236671871514,
+      "grad_norm": 0.15477107465267181,
+      "learning_rate": 1.7529475796798686e-05,
+      "loss": 0.4969,
+      "step": 3153
+    },
+    {
+      "epoch": 0.7035467320990408,
+      "grad_norm": 0.16841940581798553,
+      "learning_rate": 1.7527926842430295e-05,
+      "loss": 0.5351,
+      "step": 3154
+    },
+    {
+      "epoch": 0.7037697970109302,
+      "grad_norm": 0.17480523884296417,
+      "learning_rate": 1.7526377471117963e-05,
+      "loss": 0.5279,
+      "step": 3155
+    },
+    {
+      "epoch": 0.7039928619228195,
+      "grad_norm": 0.1635589897632599,
+      "learning_rate": 1.75248276829475e-05,
+      "loss": 0.5076,
+      "step": 3156
+    },
+    {
+      "epoch": 0.7042159268347089,
+      "grad_norm": 0.18227113783359528,
+      "learning_rate": 1.7523277478004747e-05,
+      "loss": 0.4927,
+      "step": 3157
+    },
+    {
+      "epoch": 0.7044389917465983,
+      "grad_norm": 0.17305073142051697,
+      "learning_rate": 1.7521726856375568e-05,
+      "loss": 0.4967,
+      "step": 3158
+    },
+    {
+      "epoch": 0.7046620566584876,
+      "grad_norm": 0.16051623225212097,
+      "learning_rate": 1.7520175818145838e-05,
+      "loss": 0.5031,
+      "step": 3159
+    },
+    {
+      "epoch": 0.704885121570377,
+      "grad_norm": 0.16289980709552765,
+      "learning_rate": 1.751862436340147e-05,
+      "loss": 0.5101,
+      "step": 3160
+    },
+    {
+      "epoch": 0.7051081864822664,
+      "grad_norm": 0.16630062460899353,
+      "learning_rate": 1.751707249222839e-05,
+      "loss": 0.4928,
+      "step": 3161
+    },
+    {
+      "epoch": 0.7053312513941556,
+      "grad_norm": 0.154288649559021,
+      "learning_rate": 1.7515520204712552e-05,
+      "loss": 0.4921,
+      "step": 3162
+    },
+    {
+      "epoch": 0.705554316306045,
+      "grad_norm": 0.1542695015668869,
+      "learning_rate": 1.751396750093993e-05,
+      "loss": 0.4929,
+      "step": 3163
+    },
+    {
+      "epoch": 0.7057773812179344,
+      "grad_norm": 0.1700357347726822,
+      "learning_rate": 1.7512414380996524e-05,
+      "loss": 0.5014,
+      "step": 3164
+    },
+    {
+      "epoch": 0.7060004461298238,
+      "grad_norm": 0.1531635820865631,
+      "learning_rate": 1.7510860844968355e-05,
+      "loss": 0.4764,
+      "step": 3165
+    },
+    {
+      "epoch": 0.7062235110417131,
+      "grad_norm": 0.1919977366924286,
+      "learning_rate": 1.7509306892941464e-05,
+      "loss": 0.5453,
+      "step": 3166
+    },
+    {
+      "epoch": 0.7064465759536025,
+      "grad_norm": 0.16512922942638397,
+      "learning_rate": 1.7507752525001924e-05,
+      "loss": 0.5015,
+      "step": 3167
+    },
+    {
+      "epoch": 0.7066696408654919,
+      "grad_norm": 0.1677495837211609,
+      "learning_rate": 1.7506197741235822e-05,
+      "loss": 0.5164,
+      "step": 3168
+    },
+    {
+      "epoch": 0.7068927057773812,
+      "grad_norm": 0.15319664776325226,
+      "learning_rate": 1.7504642541729273e-05,
+      "loss": 0.4815,
+      "step": 3169
+    },
+    {
+      "epoch": 0.7071157706892706,
+      "grad_norm": 0.16099566221237183,
+      "learning_rate": 1.7503086926568416e-05,
+      "loss": 0.4966,
+      "step": 3170
+    },
+    {
+      "epoch": 0.70733883560116,
+      "grad_norm": 0.15989680588245392,
+      "learning_rate": 1.750153089583941e-05,
+      "loss": 0.4949,
+      "step": 3171
+    },
+    {
+      "epoch": 0.7075619005130493,
+      "grad_norm": 0.16961508989334106,
+      "learning_rate": 1.7499974449628433e-05,
+      "loss": 0.4916,
+      "step": 3172
+    },
+    {
+      "epoch": 0.7077849654249386,
+      "grad_norm": 0.1632029265165329,
+      "learning_rate": 1.74984175880217e-05,
+      "loss": 0.4921,
+      "step": 3173
+    },
+    {
+      "epoch": 0.708008030336828,
+      "grad_norm": 0.1647975593805313,
+      "learning_rate": 1.7496860311105426e-05,
+      "loss": 0.4635,
+      "step": 3174
+    },
+    {
+      "epoch": 0.7082310952487174,
+      "grad_norm": 0.15906676650047302,
+      "learning_rate": 1.7495302618965874e-05,
+      "loss": 0.4686,
+      "step": 3175
+    },
+    {
+      "epoch": 0.7084541601606067,
+      "grad_norm": 0.16585782170295715,
+      "learning_rate": 1.7493744511689316e-05,
+      "loss": 0.5159,
+      "step": 3176
+    },
+    {
+      "epoch": 0.7086772250724961,
+      "grad_norm": 0.16121666133403778,
+      "learning_rate": 1.7492185989362052e-05,
+      "loss": 0.5,
+      "step": 3177
+    },
+    {
+      "epoch": 0.7089002899843855,
+      "grad_norm": 0.1639820635318756,
+      "learning_rate": 1.7490627052070394e-05,
+      "loss": 0.4959,
+      "step": 3178
+    },
+    {
+      "epoch": 0.7091233548962749,
+      "grad_norm": 0.17450456321239471,
+      "learning_rate": 1.74890676999007e-05,
+      "loss": 0.4932,
+      "step": 3179
+    },
+    {
+      "epoch": 0.7093464198081642,
+      "grad_norm": 0.1635378748178482,
+      "learning_rate": 1.7487507932939324e-05,
+      "loss": 0.5107,
+      "step": 3180
+    },
+    {
+      "epoch": 0.7095694847200535,
+      "grad_norm": 0.16000831127166748,
+      "learning_rate": 1.7485947751272657e-05,
+      "loss": 0.462,
+      "step": 3181
+    },
+    {
+      "epoch": 0.7097925496319429,
+      "grad_norm": 0.16015185415744781,
+      "learning_rate": 1.748438715498712e-05,
+      "loss": 0.4934,
+      "step": 3182
+    },
+    {
+      "epoch": 0.7100156145438322,
+      "grad_norm": 0.1680067628622055,
+      "learning_rate": 1.7482826144169144e-05,
+      "loss": 0.5074,
+      "step": 3183
+    },
+    {
+      "epoch": 0.7102386794557216,
+      "grad_norm": 0.16120266914367676,
+      "learning_rate": 1.7481264718905187e-05,
+      "loss": 0.4853,
+      "step": 3184
+    },
+    {
+      "epoch": 0.710461744367611,
+      "grad_norm": 0.1535530686378479,
+      "learning_rate": 1.747970287928173e-05,
+      "loss": 0.4722,
+      "step": 3185
+    },
+    {
+      "epoch": 0.7106848092795003,
+      "grad_norm": 0.1595955789089203,
+      "learning_rate": 1.747814062538528e-05,
+      "loss": 0.5031,
+      "step": 3186
+    },
+    {
+      "epoch": 0.7109078741913897,
+      "grad_norm": 0.16170883178710938,
+      "learning_rate": 1.7476577957302358e-05,
+      "loss": 0.4947,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7111309391032791,
+      "grad_norm": 0.16378958523273468,
+      "learning_rate": 1.747501487511952e-05,
+      "loss": 0.4722,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7113540040151685,
+      "grad_norm": 0.17210274934768677,
+      "learning_rate": 1.7473451378923344e-05,
+      "loss": 0.5183,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7115770689270577,
+      "grad_norm": 0.15549586713314056,
+      "learning_rate": 1.7471887468800416e-05,
+      "loss": 0.48,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7118001338389471,
+      "grad_norm": 0.17286068201065063,
+      "learning_rate": 1.747032314483736e-05,
+      "loss": 0.4768,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7120231987508365,
+      "grad_norm": 0.1790837049484253,
+      "learning_rate": 1.746875840712082e-05,
+      "loss": 0.5181,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7122462636627258,
+      "grad_norm": 0.1648358702659607,
+      "learning_rate": 1.746719325573746e-05,
+      "loss": 0.508,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7124693285746152,
+      "grad_norm": 0.16890966892242432,
+      "learning_rate": 1.7465627690773964e-05,
+      "loss": 0.4928,
+      "step": 3194
+    },
+    {
+      "epoch": 0.7126923934865046,
+      "grad_norm": 0.1561020165681839,
+      "learning_rate": 1.7464061712317047e-05,
+      "loss": 0.4783,
+      "step": 3195
+    },
+    {
+      "epoch": 0.712915458398394,
+      "grad_norm": 0.17456288635730743,
+      "learning_rate": 1.7462495320453442e-05,
+      "loss": 0.5203,
+      "step": 3196
+    },
+    {
+      "epoch": 0.7131385233102833,
+      "grad_norm": 0.16500715911388397,
+      "learning_rate": 1.7460928515269902e-05,
+      "loss": 0.5259,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7133615882221727,
+      "grad_norm": 0.1663879156112671,
+      "learning_rate": 1.7459361296853217e-05,
+      "loss": 0.5204,
+      "step": 3198
+    },
+    {
+      "epoch": 0.713584653134062,
+      "grad_norm": 0.1611357033252716,
+      "learning_rate": 1.745779366529018e-05,
+      "loss": 0.498,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7138077180459513,
+      "grad_norm": 0.1556338518857956,
+      "learning_rate": 1.7456225620667613e-05,
+      "loss": 0.4753,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7140307829578407,
+      "grad_norm": 0.16986200213432312,
+      "learning_rate": 1.7454657163072372e-05,
+      "loss": 0.5075,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7142538478697301,
+      "grad_norm": 0.16040728986263275,
+      "learning_rate": 1.7453088292591327e-05,
+      "loss": 0.5016,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7144769127816194,
+      "grad_norm": 0.16356582939624786,
+      "learning_rate": 1.7451519009311368e-05,
+      "loss": 0.5118,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7146999776935088,
+      "grad_norm": 0.16113781929016113,
+      "learning_rate": 1.744994931331942e-05,
+      "loss": 0.4869,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7149230426053982,
+      "grad_norm": 0.17623494565486908,
+      "learning_rate": 1.744837920470241e-05,
+      "loss": 0.5247,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7151461075172876,
+      "grad_norm": 0.16005289554595947,
+      "learning_rate": 1.744680868354731e-05,
+      "loss": 0.4722,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7153691724291769,
+      "grad_norm": 0.18001532554626465,
+      "learning_rate": 1.7445237749941106e-05,
+      "loss": 0.4879,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7155922373410663,
+      "grad_norm": 0.1617603451013565,
+      "learning_rate": 1.74436664039708e-05,
+      "loss": 0.4962,
+      "step": 3208
+    },
+    {
+      "epoch": 0.7158153022529556,
+      "grad_norm": 0.16146957874298096,
+      "learning_rate": 1.7442094645723425e-05,
+      "loss": 0.4849,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7160383671648449,
+      "grad_norm": 0.16159506142139435,
+      "learning_rate": 1.744052247528604e-05,
+      "loss": 0.4814,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7162614320767343,
+      "grad_norm": 0.16017811000347137,
+      "learning_rate": 1.7438949892745717e-05,
+      "loss": 0.4779,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7164844969886237,
+      "grad_norm": 0.16643457114696503,
+      "learning_rate": 1.7437376898189554e-05,
+      "loss": 0.5058,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7167075619005131,
+      "grad_norm": 0.174330934882164,
+      "learning_rate": 1.7435803491704674e-05,
+      "loss": 0.5037,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7169306268124024,
+      "grad_norm": 0.16516099870204926,
+      "learning_rate": 1.7434229673378226e-05,
+      "loss": 0.4929,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7171536917242918,
+      "grad_norm": 0.16369006037712097,
+      "learning_rate": 1.7432655443297377e-05,
+      "loss": 0.4788,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7173767566361812,
+      "grad_norm": 0.19648444652557373,
+      "learning_rate": 1.7431080801549313e-05,
+      "loss": 0.4808,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7175998215480704,
+      "grad_norm": 0.16290031373500824,
+      "learning_rate": 1.742950574822125e-05,
+      "loss": 0.4811,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7178228864599598,
+      "grad_norm": 0.15613336861133575,
+      "learning_rate": 1.7427930283400428e-05,
+      "loss": 0.4948,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7180459513718492,
+      "grad_norm": 0.17377032339572906,
+      "learning_rate": 1.7426354407174102e-05,
+      "loss": 0.5127,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7182690162837385,
+      "grad_norm": 0.16334526240825653,
+      "learning_rate": 1.7424778119629556e-05,
+      "loss": 0.4865,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7184920811956279,
+      "grad_norm": 0.15451814234256744,
+      "learning_rate": 1.7423201420854092e-05,
+      "loss": 0.4711,
+      "step": 3221
+    },
+    {
+      "epoch": 0.7187151461075173,
+      "grad_norm": 0.17903545498847961,
+      "learning_rate": 1.7421624310935043e-05,
+      "loss": 0.5265,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7189382110194067,
+      "grad_norm": 0.15851649641990662,
+      "learning_rate": 1.7420046789959754e-05,
+      "loss": 0.5125,
+      "step": 3223
+    },
+    {
+      "epoch": 0.719161275931296,
+      "grad_norm": 0.15990275144577026,
+      "learning_rate": 1.74184688580156e-05,
+      "loss": 0.4983,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7193843408431854,
+      "grad_norm": 0.16905120015144348,
+      "learning_rate": 1.7416890515189977e-05,
+      "loss": 0.514,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7196074057550748,
+      "grad_norm": 0.16614043712615967,
+      "learning_rate": 1.74153117615703e-05,
+      "loss": 0.5071,
+      "step": 3226
+    },
+    {
+      "epoch": 0.719830470666964,
+      "grad_norm": 0.1659470647573471,
+      "learning_rate": 1.741373259724402e-05,
+      "loss": 0.5101,
+      "step": 3227
+    },
+    {
+      "epoch": 0.7200535355788534,
+      "grad_norm": 0.17081058025360107,
+      "learning_rate": 1.7412153022298587e-05,
+      "loss": 0.5345,
+      "step": 3228
+    },
+    {
+      "epoch": 0.7202766004907428,
+      "grad_norm": 0.17581294476985931,
+      "learning_rate": 1.74105730368215e-05,
+      "loss": 0.5119,
+      "step": 3229
+    },
+    {
+      "epoch": 0.7204996654026322,
+      "grad_norm": 0.15835914015769958,
+      "learning_rate": 1.7408992640900263e-05,
+      "loss": 0.5,
+      "step": 3230
+    },
+    {
+      "epoch": 0.7207227303145215,
+      "grad_norm": 0.16518062353134155,
+      "learning_rate": 1.740741183462241e-05,
+      "loss": 0.4791,
+      "step": 3231
+    },
+    {
+      "epoch": 0.7209457952264109,
+      "grad_norm": 0.19645792245864868,
+      "learning_rate": 1.7405830618075494e-05,
+      "loss": 0.4851,
+      "step": 3232
+    },
+    {
+      "epoch": 0.7211688601383003,
+      "grad_norm": 0.1701032668352127,
+      "learning_rate": 1.7404248991347093e-05,
+      "loss": 0.5165,
+      "step": 3233
+    },
+    {
+      "epoch": 0.7213919250501896,
+      "grad_norm": 0.16348305344581604,
+      "learning_rate": 1.740266695452481e-05,
+      "loss": 0.5236,
+      "step": 3234
+    },
+    {
+      "epoch": 0.721614989962079,
+      "grad_norm": 0.1572495847940445,
+      "learning_rate": 1.7401084507696263e-05,
+      "loss": 0.5141,
+      "step": 3235
+    },
+    {
+      "epoch": 0.7218380548739683,
+      "grad_norm": 0.16454187035560608,
+      "learning_rate": 1.7399501650949107e-05,
+      "loss": 0.4714,
+      "step": 3236
+    },
+    {
+      "epoch": 0.7220611197858576,
+      "grad_norm": 0.17999160289764404,
+      "learning_rate": 1.7397918384371003e-05,
+      "loss": 0.5035,
+      "step": 3237
+    },
+    {
+      "epoch": 0.722284184697747,
+      "grad_norm": 0.16409459710121155,
+      "learning_rate": 1.739633470804964e-05,
+      "loss": 0.4903,
+      "step": 3238
+    },
+    {
+      "epoch": 0.7225072496096364,
+      "grad_norm": 0.15939798951148987,
+      "learning_rate": 1.739475062207274e-05,
+      "loss": 0.4968,
+      "step": 3239
+    },
+    {
+      "epoch": 0.7227303145215258,
+      "grad_norm": 0.1598650962114334,
+      "learning_rate": 1.7393166126528035e-05,
+      "loss": 0.4829,
+      "step": 3240
+    },
+    {
+      "epoch": 0.7229533794334151,
+      "grad_norm": 0.1741870641708374,
+      "learning_rate": 1.7391581221503286e-05,
+      "loss": 0.5268,
+      "step": 3241
+    },
+    {
+      "epoch": 0.7231764443453045,
+      "grad_norm": 0.16262130439281464,
+      "learning_rate": 1.7389995907086273e-05,
+      "loss": 0.4814,
+      "step": 3242
+    },
+    {
+      "epoch": 0.7233995092571939,
+      "grad_norm": 0.16130922734737396,
+      "learning_rate": 1.73884101833648e-05,
+      "loss": 0.48,
+      "step": 3243
+    },
+    {
+      "epoch": 0.7236225741690832,
+      "grad_norm": 0.167790949344635,
+      "learning_rate": 1.7386824050426697e-05,
+      "loss": 0.4774,
+      "step": 3244
+    },
+    {
+      "epoch": 0.7238456390809725,
+      "grad_norm": 0.16130533814430237,
+      "learning_rate": 1.7385237508359812e-05,
+      "loss": 0.4765,
+      "step": 3245
+    },
+    {
+      "epoch": 0.7240687039928619,
+      "grad_norm": 0.16216439008712769,
+      "learning_rate": 1.7383650557252023e-05,
+      "loss": 0.5197,
+      "step": 3246
+    },
+    {
+      "epoch": 0.7242917689047513,
+      "grad_norm": 0.17525294423103333,
+      "learning_rate": 1.7382063197191218e-05,
+      "loss": 0.5063,
+      "step": 3247
+    },
+    {
+      "epoch": 0.7245148338166406,
+      "grad_norm": 0.16659271717071533,
+      "learning_rate": 1.738047542826532e-05,
+      "loss": 0.5077,
+      "step": 3248
+    },
+    {
+      "epoch": 0.72473789872853,
+      "grad_norm": 0.15952268242835999,
+      "learning_rate": 1.7378887250562268e-05,
+      "loss": 0.4991,
+      "step": 3249
+    },
+    {
+      "epoch": 0.7249609636404194,
+      "grad_norm": 0.16580569744110107,
+      "learning_rate": 1.737729866417002e-05,
+      "loss": 0.5026,
+      "step": 3250
+    },
+    {
+      "epoch": 0.7251840285523087,
+      "grad_norm": 0.16428150236606598,
+      "learning_rate": 1.7375709669176572e-05,
+      "loss": 0.4993,
+      "step": 3251
+    },
+    {
+      "epoch": 0.7254070934641981,
+      "grad_norm": 0.17543014883995056,
+      "learning_rate": 1.7374120265669927e-05,
+      "loss": 0.483,
+      "step": 3252
+    },
+    {
+      "epoch": 0.7256301583760875,
+      "grad_norm": 0.16254562139511108,
+      "learning_rate": 1.7372530453738113e-05,
+      "loss": 0.5051,
+      "step": 3253
+    },
+    {
+      "epoch": 0.7258532232879769,
+      "grad_norm": 0.16587981581687927,
+      "learning_rate": 1.737094023346919e-05,
+      "loss": 0.4717,
+      "step": 3254
+    },
+    {
+      "epoch": 0.7260762881998661,
+      "grad_norm": 0.15644344687461853,
+      "learning_rate": 1.7369349604951233e-05,
+      "loss": 0.4918,
+      "step": 3255
+    },
+    {
+      "epoch": 0.7262993531117555,
+      "grad_norm": 0.16663451492786407,
+      "learning_rate": 1.736775856827234e-05,
+      "loss": 0.4973,
+      "step": 3256
+    },
+    {
+      "epoch": 0.7265224180236449,
+      "grad_norm": 0.16686449944972992,
+      "learning_rate": 1.736616712352063e-05,
+      "loss": 0.5115,
+      "step": 3257
+    },
+    {
+      "epoch": 0.7267454829355342,
+      "grad_norm": 0.15312035381793976,
+      "learning_rate": 1.736457527078425e-05,
+      "loss": 0.4498,
+      "step": 3258
+    },
+    {
+      "epoch": 0.7269685478474236,
+      "grad_norm": 0.17026790976524353,
+      "learning_rate": 1.7362983010151368e-05,
+      "loss": 0.4958,
+      "step": 3259
+    },
+    {
+      "epoch": 0.727191612759313,
+      "grad_norm": 0.16290387511253357,
+      "learning_rate": 1.7361390341710173e-05,
+      "loss": 0.4785,
+      "step": 3260
+    },
+    {
+      "epoch": 0.7274146776712023,
+      "grad_norm": 0.15380245447158813,
+      "learning_rate": 1.7359797265548876e-05,
+      "loss": 0.466,
+      "step": 3261
+    },
+    {
+      "epoch": 0.7276377425830917,
+      "grad_norm": 0.1699349284172058,
+      "learning_rate": 1.7358203781755707e-05,
+      "loss": 0.4858,
+      "step": 3262
+    },
+    {
+      "epoch": 0.727860807494981,
+      "grad_norm": 0.16431613266468048,
+      "learning_rate": 1.735660989041893e-05,
+      "loss": 0.5011,
+      "step": 3263
+    },
+    {
+      "epoch": 0.7280838724068704,
+      "grad_norm": 0.16305620968341827,
+      "learning_rate": 1.735501559162682e-05,
+      "loss": 0.5167,
+      "step": 3264
+    },
+    {
+      "epoch": 0.7283069373187597,
+      "grad_norm": 0.15818987786769867,
+      "learning_rate": 1.7353420885467688e-05,
+      "loss": 0.4953,
+      "step": 3265
+    },
+    {
+      "epoch": 0.7285300022306491,
+      "grad_norm": 0.164206400513649,
+      "learning_rate": 1.7351825772029847e-05,
+      "loss": 0.5128,
+      "step": 3266
+    },
+    {
+      "epoch": 0.7287530671425385,
+      "grad_norm": 0.15800419449806213,
+      "learning_rate": 1.7350230251401653e-05,
+      "loss": 0.4654,
+      "step": 3267
+    },
+    {
+      "epoch": 0.7289761320544278,
+      "grad_norm": 0.16068902611732483,
+      "learning_rate": 1.734863432367147e-05,
+      "loss": 0.4997,
+      "step": 3268
+    },
+    {
+      "epoch": 0.7291991969663172,
+      "grad_norm": 0.16787229478359222,
+      "learning_rate": 1.7347037988927696e-05,
+      "loss": 0.5244,
+      "step": 3269
+    },
+    {
+      "epoch": 0.7294222618782066,
+      "grad_norm": 0.1621689349412918,
+      "learning_rate": 1.7345441247258743e-05,
+      "loss": 0.4855,
+      "step": 3270
+    },
+    {
+      "epoch": 0.729645326790096,
+      "grad_norm": 0.16722342371940613,
+      "learning_rate": 1.734384409875305e-05,
+      "loss": 0.4992,
+      "step": 3271
+    },
+    {
+      "epoch": 0.7298683917019853,
+      "grad_norm": 0.15855719149112701,
+      "learning_rate": 1.7342246543499074e-05,
+      "loss": 0.5074,
+      "step": 3272
+    },
+    {
+      "epoch": 0.7300914566138746,
+      "grad_norm": 0.17619627714157104,
+      "learning_rate": 1.7340648581585296e-05,
+      "loss": 0.4834,
+      "step": 3273
+    },
+    {
+      "epoch": 0.730314521525764,
+      "grad_norm": 0.17650535702705383,
+      "learning_rate": 1.7339050213100233e-05,
+      "loss": 0.5393,
+      "step": 3274
+    },
+    {
+      "epoch": 0.7305375864376533,
+      "grad_norm": 0.17073291540145874,
+      "learning_rate": 1.73374514381324e-05,
+      "loss": 0.4873,
+      "step": 3275
+    },
+    {
+      "epoch": 0.7307606513495427,
+      "grad_norm": 0.1601572334766388,
+      "learning_rate": 1.733585225677035e-05,
+      "loss": 0.485,
+      "step": 3276
+    },
+    {
+      "epoch": 0.7309837162614321,
+      "grad_norm": 0.1590786874294281,
+      "learning_rate": 1.7334252669102665e-05,
+      "loss": 0.4668,
+      "step": 3277
+    },
+    {
+      "epoch": 0.7312067811733214,
+      "grad_norm": 0.15466643869876862,
+      "learning_rate": 1.7332652675217928e-05,
+      "loss": 0.4476,
+      "step": 3278
+    },
+    {
+      "epoch": 0.7314298460852108,
+      "grad_norm": 0.15521705150604248,
+      "learning_rate": 1.733105227520476e-05,
+      "loss": 0.4878,
+      "step": 3279
+    },
+    {
+      "epoch": 0.7316529109971002,
+      "grad_norm": 0.21078747510910034,
+      "learning_rate": 1.7329451469151807e-05,
+      "loss": 0.5203,
+      "step": 3280
+    },
+    {
+      "epoch": 0.7318759759089896,
+      "grad_norm": 0.1662643849849701,
+      "learning_rate": 1.7327850257147724e-05,
+      "loss": 0.5099,
+      "step": 3281
+    },
+    {
+      "epoch": 0.7320990408208788,
+      "grad_norm": 0.16159231960773468,
+      "learning_rate": 1.73262486392812e-05,
+      "loss": 0.4944,
+      "step": 3282
+    },
+    {
+      "epoch": 0.7323221057327682,
+      "grad_norm": 0.16965213418006897,
+      "learning_rate": 1.7324646615640947e-05,
+      "loss": 0.4966,
+      "step": 3283
+    },
+    {
+      "epoch": 0.7325451706446576,
+      "grad_norm": 0.1666761040687561,
+      "learning_rate": 1.732304418631569e-05,
+      "loss": 0.4846,
+      "step": 3284
+    },
+    {
+      "epoch": 0.7327682355565469,
+      "grad_norm": 0.16207394003868103,
+      "learning_rate": 1.7321441351394178e-05,
+      "loss": 0.5124,
+      "step": 3285
+    },
+    {
+      "epoch": 0.7329913004684363,
+      "grad_norm": 0.15828841924667358,
+      "learning_rate": 1.7319838110965192e-05,
+      "loss": 0.4887,
+      "step": 3286
+    },
+    {
+      "epoch": 0.7332143653803257,
+      "grad_norm": 0.16760720312595367,
+      "learning_rate": 1.731823446511753e-05,
+      "loss": 0.4715,
+      "step": 3287
+    },
+    {
+      "epoch": 0.7334374302922151,
+      "grad_norm": 0.1632707566022873,
+      "learning_rate": 1.7316630413940005e-05,
+      "loss": 0.4803,
+      "step": 3288
+    },
+    {
+      "epoch": 0.7336604952041044,
+      "grad_norm": 0.1794753223657608,
+      "learning_rate": 1.7315025957521468e-05,
+      "loss": 0.5025,
+      "step": 3289
+    },
+    {
+      "epoch": 0.7338835601159938,
+      "grad_norm": 0.1697871834039688,
+      "learning_rate": 1.7313421095950778e-05,
+      "loss": 0.4725,
+      "step": 3290
+    },
+    {
+      "epoch": 0.7341066250278832,
+      "grad_norm": 0.16839033365249634,
+      "learning_rate": 1.7311815829316826e-05,
+      "loss": 0.5406,
+      "step": 3291
+    },
+    {
+      "epoch": 0.7343296899397724,
+      "grad_norm": 0.16735753417015076,
+      "learning_rate": 1.731021015770852e-05,
+      "loss": 0.5238,
+      "step": 3292
+    },
+    {
+      "epoch": 0.7345527548516618,
+      "grad_norm": 0.16878098249435425,
+      "learning_rate": 1.7308604081214793e-05,
+      "loss": 0.4978,
+      "step": 3293
+    },
+    {
+      "epoch": 0.7347758197635512,
+      "grad_norm": 0.16095402836799622,
+      "learning_rate": 1.7306997599924597e-05,
+      "loss": 0.4696,
+      "step": 3294
+    },
+    {
+      "epoch": 0.7349988846754405,
+      "grad_norm": 0.16328908503055573,
+      "learning_rate": 1.730539071392691e-05,
+      "loss": 0.4965,
+      "step": 3295
+    },
+    {
+      "epoch": 0.7352219495873299,
+      "grad_norm": 0.1627175509929657,
+      "learning_rate": 1.7303783423310735e-05,
+      "loss": 0.4822,
+      "step": 3296
+    },
+    {
+      "epoch": 0.7354450144992193,
+      "grad_norm": 0.1619403213262558,
+      "learning_rate": 1.730217572816509e-05,
+      "loss": 0.5027,
+      "step": 3297
+    },
+    {
+      "epoch": 0.7356680794111087,
+      "grad_norm": 0.16253367066383362,
+      "learning_rate": 1.7300567628579025e-05,
+      "loss": 0.5028,
+      "step": 3298
+    },
+    {
+      "epoch": 0.735891144322998,
+      "grad_norm": 0.16807517409324646,
+      "learning_rate": 1.72989591246416e-05,
+      "loss": 0.4756,
+      "step": 3299
+    },
+    {
+      "epoch": 0.7361142092348874,
+      "grad_norm": 0.18981721997261047,
+      "learning_rate": 1.7297350216441903e-05,
+      "loss": 0.5209,
+      "step": 3300
+    },
+    {
+      "epoch": 0.7363372741467767,
+      "grad_norm": 0.1598469614982605,
+      "learning_rate": 1.7295740904069053e-05,
+      "loss": 0.4872,
+      "step": 3301
+    },
+    {
+      "epoch": 0.736560339058666,
+      "grad_norm": 0.20748665928840637,
+      "learning_rate": 1.7294131187612176e-05,
+      "loss": 0.4603,
+      "step": 3302
+    },
+    {
+      "epoch": 0.7367834039705554,
+      "grad_norm": 0.16444922983646393,
+      "learning_rate": 1.7292521067160434e-05,
+      "loss": 0.4772,
+      "step": 3303
+    },
+    {
+      "epoch": 0.7370064688824448,
+      "grad_norm": 0.16779246926307678,
+      "learning_rate": 1.7290910542803004e-05,
+      "loss": 0.4948,
+      "step": 3304
+    },
+    {
+      "epoch": 0.7372295337943342,
+      "grad_norm": 0.16438250243663788,
+      "learning_rate": 1.7289299614629083e-05,
+      "loss": 0.5103,
+      "step": 3305
+    },
+    {
+      "epoch": 0.7374525987062235,
+      "grad_norm": 0.17737194895744324,
+      "learning_rate": 1.7287688282727903e-05,
+      "loss": 0.4994,
+      "step": 3306
+    },
+    {
+      "epoch": 0.7376756636181129,
+      "grad_norm": 0.15918989479541779,
+      "learning_rate": 1.7286076547188703e-05,
+      "loss": 0.4978,
+      "step": 3307
+    },
+    {
+      "epoch": 0.7378987285300023,
+      "grad_norm": 0.3185868561267853,
+      "learning_rate": 1.728446440810075e-05,
+      "loss": 0.489,
+      "step": 3308
+    },
+    {
+      "epoch": 0.7381217934418915,
+      "grad_norm": 0.1764698624610901,
+      "learning_rate": 1.728285186555334e-05,
+      "loss": 0.506,
+      "step": 3309
+    },
+    {
+      "epoch": 0.7383448583537809,
+      "grad_norm": 0.16116134822368622,
+      "learning_rate": 1.7281238919635784e-05,
+      "loss": 0.519,
+      "step": 3310
+    },
+    {
+      "epoch": 0.7385679232656703,
+      "grad_norm": 0.1670210361480713,
+      "learning_rate": 1.7279625570437413e-05,
+      "loss": 0.5228,
+      "step": 3311
+    },
+    {
+      "epoch": 0.7387909881775596,
+      "grad_norm": 0.16115038096904755,
+      "learning_rate": 1.7278011818047588e-05,
+      "loss": 0.4882,
+      "step": 3312
+    },
+    {
+      "epoch": 0.739014053089449,
+      "grad_norm": 0.16242891550064087,
+      "learning_rate": 1.7276397662555685e-05,
+      "loss": 0.5139,
+      "step": 3313
+    },
+    {
+      "epoch": 0.7392371180013384,
+      "grad_norm": 0.16696135699748993,
+      "learning_rate": 1.7274783104051112e-05,
+      "loss": 0.521,
+      "step": 3314
+    },
+    {
+      "epoch": 0.7394601829132278,
+      "grad_norm": 0.1544819325208664,
+      "learning_rate": 1.727316814262329e-05,
+      "loss": 0.4765,
+      "step": 3315
+    },
+    {
+      "epoch": 0.7396832478251171,
+      "grad_norm": 0.16446231305599213,
+      "learning_rate": 1.727155277836167e-05,
+      "loss": 0.4683,
+      "step": 3316
+    },
+    {
+      "epoch": 0.7399063127370065,
+      "grad_norm": 0.1666785031557083,
+      "learning_rate": 1.7269937011355713e-05,
+      "loss": 0.503,
+      "step": 3317
+    },
+    {
+      "epoch": 0.7401293776488959,
+      "grad_norm": 0.16166886687278748,
+      "learning_rate": 1.7268320841694915e-05,
+      "loss": 0.5014,
+      "step": 3318
+    },
+    {
+      "epoch": 0.7403524425607851,
+      "grad_norm": 0.27264273166656494,
+      "learning_rate": 1.7266704269468786e-05,
+      "loss": 0.4834,
+      "step": 3319
+    },
+    {
+      "epoch": 0.7405755074726745,
+      "grad_norm": 0.16175110638141632,
+      "learning_rate": 1.7265087294766872e-05,
+      "loss": 0.4926,
+      "step": 3320
+    },
+    {
+      "epoch": 0.7407985723845639,
+      "grad_norm": 0.1768079400062561,
+      "learning_rate": 1.726346991767872e-05,
+      "loss": 0.4933,
+      "step": 3321
+    },
+    {
+      "epoch": 0.7410216372964533,
+      "grad_norm": 0.15689828991889954,
+      "learning_rate": 1.7261852138293918e-05,
+      "loss": 0.4928,
+      "step": 3322
+    },
+    {
+      "epoch": 0.7412447022083426,
+      "grad_norm": 0.16128939390182495,
+      "learning_rate": 1.7260233956702062e-05,
+      "loss": 0.4739,
+      "step": 3323
+    },
+    {
+      "epoch": 0.741467767120232,
+      "grad_norm": 0.16142840683460236,
+      "learning_rate": 1.7258615372992783e-05,
+      "loss": 0.5123,
+      "step": 3324
+    },
+    {
+      "epoch": 0.7416908320321214,
+      "grad_norm": 0.15940620005130768,
+      "learning_rate": 1.7256996387255725e-05,
+      "loss": 0.5013,
+      "step": 3325
+    },
+    {
+      "epoch": 0.7419138969440107,
+      "grad_norm": 0.16134943068027496,
+      "learning_rate": 1.7255376999580557e-05,
+      "loss": 0.4651,
+      "step": 3326
+    },
+    {
+      "epoch": 0.7421369618559001,
+      "grad_norm": 0.15544810891151428,
+      "learning_rate": 1.7253757210056978e-05,
+      "loss": 0.4657,
+      "step": 3327
+    },
+    {
+      "epoch": 0.7423600267677894,
+      "grad_norm": 0.15278422832489014,
+      "learning_rate": 1.7252137018774694e-05,
+      "loss": 0.4604,
+      "step": 3328
+    },
+    {
+      "epoch": 0.7425830916796788,
+      "grad_norm": 0.15711228549480438,
+      "learning_rate": 1.7250516425823443e-05,
+      "loss": 0.4903,
+      "step": 3329
+    },
+    {
+      "epoch": 0.7428061565915681,
+      "grad_norm": 0.15838393568992615,
+      "learning_rate": 1.7248895431292988e-05,
+      "loss": 0.484,
+      "step": 3330
+    },
+    {
+      "epoch": 0.7430292215034575,
+      "grad_norm": 0.15732166171073914,
+      "learning_rate": 1.72472740352731e-05,
+      "loss": 0.4637,
+      "step": 3331
+    },
+    {
+      "epoch": 0.7432522864153469,
+      "grad_norm": 0.16680875420570374,
+      "learning_rate": 1.7245652237853593e-05,
+      "loss": 0.4999,
+      "step": 3332
+    },
+    {
+      "epoch": 0.7434753513272362,
+      "grad_norm": 0.16169287264347076,
+      "learning_rate": 1.7244030039124287e-05,
+      "loss": 0.4829,
+      "step": 3333
+    },
+    {
+      "epoch": 0.7436984162391256,
+      "grad_norm": 0.15638568997383118,
+      "learning_rate": 1.7242407439175035e-05,
+      "loss": 0.4731,
+      "step": 3334
+    },
+    {
+      "epoch": 0.743921481151015,
+      "grad_norm": 0.1748877912759781,
+      "learning_rate": 1.72407844380957e-05,
+      "loss": 0.4979,
+      "step": 3335
+    },
+    {
+      "epoch": 0.7441445460629043,
+      "grad_norm": 0.15920549631118774,
+      "learning_rate": 1.7239161035976175e-05,
+      "loss": 0.5024,
+      "step": 3336
+    },
+    {
+      "epoch": 0.7443676109747936,
+      "grad_norm": 0.16807498037815094,
+      "learning_rate": 1.7237537232906376e-05,
+      "loss": 0.5414,
+      "step": 3337
+    },
+    {
+      "epoch": 0.744590675886683,
+      "grad_norm": 0.16192768514156342,
+      "learning_rate": 1.723591302897624e-05,
+      "loss": 0.4883,
+      "step": 3338
+    },
+    {
+      "epoch": 0.7448137407985724,
+      "grad_norm": 0.15893089771270752,
+      "learning_rate": 1.7234288424275726e-05,
+      "loss": 0.5002,
+      "step": 3339
+    },
+    {
+      "epoch": 0.7450368057104617,
+      "grad_norm": 0.1622776985168457,
+      "learning_rate": 1.7232663418894812e-05,
+      "loss": 0.4859,
+      "step": 3340
+    },
+    {
+      "epoch": 0.7452598706223511,
+      "grad_norm": 0.1694253832101822,
+      "learning_rate": 1.72310380129235e-05,
+      "loss": 0.4903,
+      "step": 3341
+    },
+    {
+      "epoch": 0.7454829355342405,
+      "grad_norm": 0.1649332493543625,
+      "learning_rate": 1.722941220645182e-05,
+      "loss": 0.5145,
+      "step": 3342
+    },
+    {
+      "epoch": 0.7457060004461298,
+      "grad_norm": 0.19938278198242188,
+      "learning_rate": 1.722778599956982e-05,
+      "loss": 0.4949,
+      "step": 3343
+    },
+    {
+      "epoch": 0.7459290653580192,
+      "grad_norm": 0.15563727915287018,
+      "learning_rate": 1.7226159392367564e-05,
+      "loss": 0.4911,
+      "step": 3344
+    },
+    {
+      "epoch": 0.7461521302699086,
+      "grad_norm": 0.15972435474395752,
+      "learning_rate": 1.7224532384935148e-05,
+      "loss": 0.4748,
+      "step": 3345
+    },
+    {
+      "epoch": 0.746375195181798,
+      "grad_norm": 0.16170111298561096,
+      "learning_rate": 1.722290497736268e-05,
+      "loss": 0.4701,
+      "step": 3346
+    },
+    {
+      "epoch": 0.7465982600936872,
+      "grad_norm": 0.1537921279668808,
+      "learning_rate": 1.7221277169740305e-05,
+      "loss": 0.478,
+      "step": 3347
+    },
+    {
+      "epoch": 0.7468213250055766,
+      "grad_norm": 0.15656088292598724,
+      "learning_rate": 1.7219648962158174e-05,
+      "loss": 0.4851,
+      "step": 3348
+    },
+    {
+      "epoch": 0.747044389917466,
+      "grad_norm": 0.16441093385219574,
+      "learning_rate": 1.7218020354706473e-05,
+      "loss": 0.4846,
+      "step": 3349
+    },
+    {
+      "epoch": 0.7472674548293553,
+      "grad_norm": 0.17135939002037048,
+      "learning_rate": 1.72163913474754e-05,
+      "loss": 0.5044,
+      "step": 3350
+    },
+    {
+      "epoch": 0.7474905197412447,
+      "grad_norm": 0.17070244252681732,
+      "learning_rate": 1.721476194055518e-05,
+      "loss": 0.537,
+      "step": 3351
+    },
+    {
+      "epoch": 0.7477135846531341,
+      "grad_norm": 0.1651678830385208,
+      "learning_rate": 1.7213132134036063e-05,
+      "loss": 0.5161,
+      "step": 3352
+    },
+    {
+      "epoch": 0.7479366495650234,
+      "grad_norm": 0.16997648775577545,
+      "learning_rate": 1.7211501928008317e-05,
+      "loss": 0.5026,
+      "step": 3353
+    },
+    {
+      "epoch": 0.7481597144769128,
+      "grad_norm": 0.17357714474201202,
+      "learning_rate": 1.7209871322562232e-05,
+      "loss": 0.4899,
+      "step": 3354
+    },
+    {
+      "epoch": 0.7483827793888022,
+      "grad_norm": 0.16555210947990417,
+      "learning_rate": 1.7208240317788115e-05,
+      "loss": 0.5176,
+      "step": 3355
+    },
+    {
+      "epoch": 0.7486058443006915,
+      "grad_norm": 0.15799009799957275,
+      "learning_rate": 1.7206608913776315e-05,
+      "loss": 0.4923,
+      "step": 3356
+    },
+    {
+      "epoch": 0.7488289092125808,
+      "grad_norm": 0.15853382647037506,
+      "learning_rate": 1.720497711061718e-05,
+      "loss": 0.4881,
+      "step": 3357
+    },
+    {
+      "epoch": 0.7490519741244702,
+      "grad_norm": 0.15760891139507294,
+      "learning_rate": 1.720334490840109e-05,
+      "loss": 0.5008,
+      "step": 3358
+    },
+    {
+      "epoch": 0.7492750390363596,
+      "grad_norm": 0.16732452809810638,
+      "learning_rate": 1.720171230721845e-05,
+      "loss": 0.4923,
+      "step": 3359
+    },
+    {
+      "epoch": 0.7494981039482489,
+      "grad_norm": 0.17124591767787933,
+      "learning_rate": 1.7200079307159677e-05,
+      "loss": 0.5272,
+      "step": 3360
+    },
+    {
+      "epoch": 0.7497211688601383,
+      "grad_norm": 0.16751034557819366,
+      "learning_rate": 1.7198445908315226e-05,
+      "loss": 0.4801,
+      "step": 3361
+    },
+    {
+      "epoch": 0.7499442337720277,
+      "grad_norm": 0.16137580573558807,
+      "learning_rate": 1.719681211077556e-05,
+      "loss": 0.5371,
+      "step": 3362
+    },
+    {
+      "epoch": 0.7501672986839171,
+      "grad_norm": 0.16350311040878296,
+      "learning_rate": 1.7195177914631172e-05,
+      "loss": 0.5152,
+      "step": 3363
+    },
+    {
+      "epoch": 0.7503903635958064,
+      "grad_norm": 0.1560794562101364,
+      "learning_rate": 1.719354331997257e-05,
+      "loss": 0.4788,
+      "step": 3364
+    },
+    {
+      "epoch": 0.7506134285076957,
+      "grad_norm": 0.16271278262138367,
+      "learning_rate": 1.7191908326890288e-05,
+      "loss": 0.4975,
+      "step": 3365
+    },
+    {
+      "epoch": 0.7508364934195851,
+      "grad_norm": 0.21892762184143066,
+      "learning_rate": 1.7190272935474883e-05,
+      "loss": 0.511,
+      "step": 3366
+    },
+    {
+      "epoch": 0.7510595583314744,
+      "grad_norm": 0.1879514753818512,
+      "learning_rate": 1.7188637145816937e-05,
+      "loss": 0.5277,
+      "step": 3367
+    },
+    {
+      "epoch": 0.7512826232433638,
+      "grad_norm": 0.17167915403842926,
+      "learning_rate": 1.718700095800705e-05,
+      "loss": 0.5128,
+      "step": 3368
+    },
+    {
+      "epoch": 0.7515056881552532,
+      "grad_norm": 0.16950075328350067,
+      "learning_rate": 1.718536437213584e-05,
+      "loss": 0.5175,
+      "step": 3369
+    },
+    {
+      "epoch": 0.7517287530671425,
+      "grad_norm": 0.1571718156337738,
+      "learning_rate": 1.718372738829395e-05,
+      "loss": 0.489,
+      "step": 3370
+    },
+    {
+      "epoch": 0.7519518179790319,
+      "grad_norm": 0.161586731672287,
+      "learning_rate": 1.718209000657205e-05,
+      "loss": 0.5112,
+      "step": 3371
+    },
+    {
+      "epoch": 0.7521748828909213,
+      "grad_norm": 0.16906411945819855,
+      "learning_rate": 1.718045222706083e-05,
+      "loss": 0.5171,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7523979478028107,
+      "grad_norm": 0.17176932096481323,
+      "learning_rate": 1.7178814049851e-05,
+      "loss": 0.5352,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7526210127146999,
+      "grad_norm": 0.15188109874725342,
+      "learning_rate": 1.717717547503329e-05,
+      "loss": 0.4857,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7528440776265893,
+      "grad_norm": 0.1867418736219406,
+      "learning_rate": 1.7175536502698456e-05,
+      "loss": 0.5199,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7530671425384787,
+      "grad_norm": 0.1905105859041214,
+      "learning_rate": 1.7173897132937274e-05,
+      "loss": 0.5077,
+      "step": 3376
+    },
+    {
+      "epoch": 0.753290207450368,
+      "grad_norm": 0.16345377266407013,
+      "learning_rate": 1.7172257365840544e-05,
+      "loss": 0.5362,
+      "step": 3377
+    },
+    {
+      "epoch": 0.7535132723622574,
+      "grad_norm": 0.17045968770980835,
+      "learning_rate": 1.7170617201499083e-05,
+      "loss": 0.5057,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7537363372741468,
+      "grad_norm": 0.16535595059394836,
+      "learning_rate": 1.716897664000374e-05,
+      "loss": 0.5141,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7539594021860362,
+      "grad_norm": 0.16590653359889984,
+      "learning_rate": 1.716733568144538e-05,
+      "loss": 0.5146,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7541824670979255,
+      "grad_norm": 0.156137153506279,
+      "learning_rate": 1.716569432591488e-05,
+      "loss": 0.4984,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7544055320098149,
+      "grad_norm": 0.17076677083969116,
+      "learning_rate": 1.7164052573503155e-05,
+      "loss": 0.4968,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7546285969217043,
+      "grad_norm": 0.15686576068401337,
+      "learning_rate": 1.7162410424301132e-05,
+      "loss": 0.4868,
+      "step": 3383
+    },
+    {
+      "epoch": 0.7548516618335935,
+      "grad_norm": 0.19442743062973022,
+      "learning_rate": 1.716076787839977e-05,
+      "loss": 0.5058,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7550747267454829,
+      "grad_norm": 0.16424842178821564,
+      "learning_rate": 1.715912493589004e-05,
+      "loss": 0.4857,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7552977916573723,
+      "grad_norm": 0.6959996819496155,
+      "learning_rate": 1.7157481596862936e-05,
+      "loss": 0.5091,
+      "step": 3386
+    },
+    {
+      "epoch": 0.7555208565692616,
+      "grad_norm": 0.17369645833969116,
+      "learning_rate": 1.7155837861409482e-05,
+      "loss": 0.5116,
+      "step": 3387
+    },
+    {
+      "epoch": 0.755743921481151,
+      "grad_norm": 0.2050577700138092,
+      "learning_rate": 1.7154193729620713e-05,
+      "loss": 0.4982,
+      "step": 3388
+    },
+    {
+      "epoch": 0.7559669863930404,
+      "grad_norm": 0.15558819472789764,
+      "learning_rate": 1.7152549201587695e-05,
+      "loss": 0.4873,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7561900513049298,
+      "grad_norm": 0.16767403483390808,
+      "learning_rate": 1.715090427740151e-05,
+      "loss": 0.4806,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7564131162168191,
+      "grad_norm": 0.1653711050748825,
+      "learning_rate": 1.714925895715326e-05,
+      "loss": 0.4858,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7566361811287085,
+      "grad_norm": 0.1628393530845642,
+      "learning_rate": 1.7147613240934087e-05,
+      "loss": 0.5036,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7568592460405978,
+      "grad_norm": 0.1690330058336258,
+      "learning_rate": 1.714596712883513e-05,
+      "loss": 0.4788,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7570823109524871,
+      "grad_norm": 0.18115870654582977,
+      "learning_rate": 1.714432062094756e-05,
+      "loss": 0.5009,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7573053758643765,
+      "grad_norm": 0.17084982991218567,
+      "learning_rate": 1.7142673717362578e-05,
+      "loss": 0.5281,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7575284407762659,
+      "grad_norm": 0.16751186549663544,
+      "learning_rate": 1.7141026418171396e-05,
+      "loss": 0.5377,
+      "step": 3396
+    },
+    {
+      "epoch": 0.7577515056881553,
+      "grad_norm": 0.16510246694087982,
+      "learning_rate": 1.713937872346525e-05,
+      "loss": 0.4995,
+      "step": 3397
+    },
+    {
+      "epoch": 0.7579745706000446,
+      "grad_norm": 0.1689106822013855,
+      "learning_rate": 1.7137730633335404e-05,
+      "loss": 0.4891,
+      "step": 3398
+    },
+    {
+      "epoch": 0.758197635511934,
+      "grad_norm": 0.16398201882839203,
+      "learning_rate": 1.7136082147873136e-05,
+      "loss": 0.5013,
+      "step": 3399
+    },
+    {
+      "epoch": 0.7584207004238234,
+      "grad_norm": 0.16415871679782867,
+      "learning_rate": 1.713443326716975e-05,
+      "loss": 0.5343,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7586437653357126,
+      "grad_norm": 0.164589986205101,
+      "learning_rate": 1.7132783991316577e-05,
+      "loss": 0.4897,
+      "step": 3401
+    },
+    {
+      "epoch": 0.758866830247602,
+      "grad_norm": 0.17511430382728577,
+      "learning_rate": 1.7131134320404953e-05,
+      "loss": 0.5088,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7590898951594914,
+      "grad_norm": 0.1833367496728897,
+      "learning_rate": 1.7129484254526257e-05,
+      "loss": 0.4811,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7593129600713808,
+      "grad_norm": 0.16160368919372559,
+      "learning_rate": 1.7127833793771874e-05,
+      "loss": 0.508,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7595360249832701,
+      "grad_norm": 0.1855868101119995,
+      "learning_rate": 1.7126182938233228e-05,
+      "loss": 0.5091,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7597590898951595,
+      "grad_norm": 0.16717424988746643,
+      "learning_rate": 1.7124531688001735e-05,
+      "loss": 0.482,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7599821548070489,
+      "grad_norm": 0.1635364592075348,
+      "learning_rate": 1.7122880043168872e-05,
+      "loss": 0.4706,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7602052197189382,
+      "grad_norm": 0.15802796185016632,
+      "learning_rate": 1.71212280038261e-05,
+      "loss": 0.4835,
+      "step": 3408
+    },
+    {
+      "epoch": 0.7604282846308276,
+      "grad_norm": 0.16960306465625763,
+      "learning_rate": 1.7119575570064926e-05,
+      "loss": 0.474,
+      "step": 3409
+    },
+    {
+      "epoch": 0.760651349542717,
+      "grad_norm": 0.16367004811763763,
+      "learning_rate": 1.7117922741976878e-05,
+      "loss": 0.5132,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7608744144546062,
+      "grad_norm": 0.1700059324502945,
+      "learning_rate": 1.7116269519653493e-05,
+      "loss": 0.5124,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7610974793664956,
+      "grad_norm": 0.19023141264915466,
+      "learning_rate": 1.711461590318634e-05,
+      "loss": 0.487,
+      "step": 3412
+    },
+    {
+      "epoch": 0.761320544278385,
+      "grad_norm": 0.16813622415065765,
+      "learning_rate": 1.7112961892667003e-05,
+      "loss": 0.4749,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7615436091902744,
+      "grad_norm": 0.16135910153388977,
+      "learning_rate": 1.7111307488187096e-05,
+      "loss": 0.5041,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7617666741021637,
+      "grad_norm": 0.16579851508140564,
+      "learning_rate": 1.710965268983825e-05,
+      "loss": 0.5062,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7619897390140531,
+      "grad_norm": 0.15735271573066711,
+      "learning_rate": 1.7107997497712113e-05,
+      "loss": 0.5071,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7622128039259425,
+      "grad_norm": 0.1686030477285385,
+      "learning_rate": 1.7106341911900365e-05,
+      "loss": 0.5271,
+      "step": 3417
+    },
+    {
+      "epoch": 0.7624358688378318,
+      "grad_norm": 0.17071577906608582,
+      "learning_rate": 1.7104685932494704e-05,
+      "loss": 0.5179,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7626589337497212,
+      "grad_norm": 0.15663312375545502,
+      "learning_rate": 1.7103029559586843e-05,
+      "loss": 0.4928,
+      "step": 3419
+    },
+    {
+      "epoch": 0.7628819986616105,
+      "grad_norm": 0.17715679109096527,
+      "learning_rate": 1.7101372793268526e-05,
+      "loss": 0.5371,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7631050635734999,
+      "grad_norm": 0.1651579737663269,
+      "learning_rate": 1.709971563363151e-05,
+      "loss": 0.5072,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7633281284853892,
+      "grad_norm": 0.15921546518802643,
+      "learning_rate": 1.7098058080767587e-05,
+      "loss": 0.5008,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7635511933972786,
+      "grad_norm": 0.17586684226989746,
+      "learning_rate": 1.709640013476856e-05,
+      "loss": 0.4962,
+      "step": 3423
+    },
+    {
+      "epoch": 0.763774258309168,
+      "grad_norm": 0.17664553225040436,
+      "learning_rate": 1.7094741795726254e-05,
+      "loss": 0.5107,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7639973232210573,
+      "grad_norm": 0.1621769219636917,
+      "learning_rate": 1.7093083063732518e-05,
+      "loss": 0.5017,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7642203881329467,
+      "grad_norm": 0.1619473695755005,
+      "learning_rate": 1.7091423938879227e-05,
+      "loss": 0.4858,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7644434530448361,
+      "grad_norm": 0.1741994023323059,
+      "learning_rate": 1.7089764421258272e-05,
+      "loss": 0.5226,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7646665179567254,
+      "grad_norm": 0.16964323818683624,
+      "learning_rate": 1.7088104510961564e-05,
+      "loss": 0.4987,
+      "step": 3428
+    },
+    {
+      "epoch": 0.7648895828686147,
+      "grad_norm": 0.1797408014535904,
+      "learning_rate": 1.7086444208081047e-05,
+      "loss": 0.4833,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7651126477805041,
+      "grad_norm": 0.19941550493240356,
+      "learning_rate": 1.708478351270867e-05,
+      "loss": 0.5138,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7653357126923935,
+      "grad_norm": 0.17348411679267883,
+      "learning_rate": 1.708312242493642e-05,
+      "loss": 0.4866,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7655587776042828,
+      "grad_norm": 0.1664038747549057,
+      "learning_rate": 1.7081460944856294e-05,
+      "loss": 0.5124,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7657818425161722,
+      "grad_norm": 0.15917468070983887,
+      "learning_rate": 1.7079799072560318e-05,
+      "loss": 0.5066,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7660049074280616,
+      "grad_norm": 0.16369706392288208,
+      "learning_rate": 1.7078136808140532e-05,
+      "loss": 0.4903,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7662279723399509,
+      "grad_norm": 0.16175441443920135,
+      "learning_rate": 1.707647415168901e-05,
+      "loss": 0.525,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7664510372518403,
+      "grad_norm": 0.19323702156543732,
+      "learning_rate": 1.707481110329783e-05,
+      "loss": 0.4865,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7666741021637297,
+      "grad_norm": 0.15811483561992645,
+      "learning_rate": 1.707314766305912e-05,
+      "loss": 0.4709,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7668971670756191,
+      "grad_norm": 0.1613859087228775,
+      "learning_rate": 1.707148383106499e-05,
+      "loss": 0.4969,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7671202319875083,
+      "grad_norm": 0.16382627189159393,
+      "learning_rate": 1.706981960740761e-05,
+      "loss": 0.4893,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7673432968993977,
+      "grad_norm": 0.17395590245723724,
+      "learning_rate": 1.706815499217915e-05,
+      "loss": 0.5073,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7675663618112871,
+      "grad_norm": 0.16161273419857025,
+      "learning_rate": 1.7066489985471802e-05,
+      "loss": 0.4766,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7677894267231764,
+      "grad_norm": 0.16916361451148987,
+      "learning_rate": 1.706482458737779e-05,
+      "loss": 0.5102,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7680124916350658,
+      "grad_norm": 0.16515444219112396,
+      "learning_rate": 1.7063158797989355e-05,
+      "loss": 0.5006,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7682355565469552,
+      "grad_norm": 0.16712456941604614,
+      "learning_rate": 1.7061492617398755e-05,
+      "loss": 0.4757,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7684586214588445,
+      "grad_norm": 0.1649584323167801,
+      "learning_rate": 1.7059826045698275e-05,
+      "loss": 0.4979,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7686816863707339,
+      "grad_norm": 0.16432853043079376,
+      "learning_rate": 1.7058159082980223e-05,
+      "loss": 0.5155,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7689047512826233,
+      "grad_norm": 0.15966151654720306,
+      "learning_rate": 1.7056491729336917e-05,
+      "loss": 0.4908,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7691278161945126,
+      "grad_norm": 0.1597587913274765,
+      "learning_rate": 1.7054823984860716e-05,
+      "loss": 0.4912,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7693508811064019,
+      "grad_norm": 0.16364127397537231,
+      "learning_rate": 1.705315584964399e-05,
+      "loss": 0.5045,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7695739460182913,
+      "grad_norm": 0.15808305144309998,
+      "learning_rate": 1.7051487323779122e-05,
+      "loss": 0.5025,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7697970109301807,
+      "grad_norm": 0.173845574259758,
+      "learning_rate": 1.704981840735853e-05,
+      "loss": 0.5166,
+      "step": 3451
+    },
+    {
+      "epoch": 0.77002007584207,
+      "grad_norm": 0.1586831957101822,
+      "learning_rate": 1.7048149100474653e-05,
+      "loss": 0.4758,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7702431407539594,
+      "grad_norm": 0.16501812636852264,
+      "learning_rate": 1.704647940321994e-05,
+      "loss": 0.5051,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7704662056658488,
+      "grad_norm": 0.17648212611675262,
+      "learning_rate": 1.704480931568688e-05,
+      "loss": 0.5014,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7706892705777382,
+      "grad_norm": 0.15847481787204742,
+      "learning_rate": 1.704313883796796e-05,
+      "loss": 0.5246,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7709123354896275,
+      "grad_norm": 0.17660297453403473,
+      "learning_rate": 1.704146797015571e-05,
+      "loss": 0.5028,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7711354004015168,
+      "grad_norm": 0.1632460057735443,
+      "learning_rate": 1.7039796712342672e-05,
+      "loss": 0.4647,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7713584653134062,
+      "grad_norm": 0.15266932547092438,
+      "learning_rate": 1.7038125064621408e-05,
+      "loss": 0.4829,
+      "step": 3458
+    },
+    {
+      "epoch": 0.7715815302252955,
+      "grad_norm": 0.16995325684547424,
+      "learning_rate": 1.703645302708451e-05,
+      "loss": 0.5057,
+      "step": 3459
+    },
+    {
+      "epoch": 0.7718045951371849,
+      "grad_norm": 0.15925444662570953,
+      "learning_rate": 1.703478059982458e-05,
+      "loss": 0.5128,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7720276600490743,
+      "grad_norm": 0.17199598252773285,
+      "learning_rate": 1.703310778293425e-05,
+      "loss": 0.5088,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7722507249609636,
+      "grad_norm": 0.17313992977142334,
+      "learning_rate": 1.7031434576506173e-05,
+      "loss": 0.4796,
+      "step": 3462
+    },
+    {
+      "epoch": 0.772473789872853,
+      "grad_norm": 0.1665591597557068,
+      "learning_rate": 1.7029760980633016e-05,
+      "loss": 0.4978,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7726968547847424,
+      "grad_norm": 0.16291576623916626,
+      "learning_rate": 1.7028086995407477e-05,
+      "loss": 0.4892,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7729199196966318,
+      "grad_norm": 0.15629801154136658,
+      "learning_rate": 1.7026412620922276e-05,
+      "loss": 0.483,
+      "step": 3465
+    },
+    {
+      "epoch": 0.773142984608521,
+      "grad_norm": 0.16080623865127563,
+      "learning_rate": 1.702473785727014e-05,
+      "loss": 0.4968,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7733660495204104,
+      "grad_norm": 0.16567450761795044,
+      "learning_rate": 1.702306270454384e-05,
+      "loss": 0.5023,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7735891144322998,
+      "grad_norm": 0.16630107164382935,
+      "learning_rate": 1.702138716283615e-05,
+      "loss": 0.4554,
+      "step": 3468
+    },
+    {
+      "epoch": 0.7738121793441891,
+      "grad_norm": 0.1635141223669052,
+      "learning_rate": 1.7019711232239872e-05,
+      "loss": 0.4947,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7740352442560785,
+      "grad_norm": 0.16662491858005524,
+      "learning_rate": 1.7018034912847826e-05,
+      "loss": 0.4947,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7742583091679679,
+      "grad_norm": 0.16234828531742096,
+      "learning_rate": 1.7016358204752865e-05,
+      "loss": 0.5424,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7744813740798573,
+      "grad_norm": 0.16206470131874084,
+      "learning_rate": 1.701468110804785e-05,
+      "loss": 0.4761,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7747044389917466,
+      "grad_norm": 0.1729600876569748,
+      "learning_rate": 1.7013003622825674e-05,
+      "loss": 0.4867,
+      "step": 3473
+    },
+    {
+      "epoch": 0.774927503903636,
+      "grad_norm": 0.15577882528305054,
+      "learning_rate": 1.7011325749179245e-05,
+      "loss": 0.4935,
+      "step": 3474
+    },
+    {
+      "epoch": 0.7751505688155254,
+      "grad_norm": 0.16553224623203278,
+      "learning_rate": 1.7009647487201492e-05,
+      "loss": 0.5052,
+      "step": 3475
+    },
+    {
+      "epoch": 0.7753736337274146,
+      "grad_norm": 0.16988618671894073,
+      "learning_rate": 1.700796883698536e-05,
+      "loss": 0.4999,
+      "step": 3476
+    },
+    {
+      "epoch": 0.775596698639304,
+      "grad_norm": 0.15535961091518402,
+      "learning_rate": 1.7006289798623842e-05,
+      "loss": 0.494,
+      "step": 3477
+    },
+    {
+      "epoch": 0.7758197635511934,
+      "grad_norm": 0.16982175409793854,
+      "learning_rate": 1.700461037220992e-05,
+      "loss": 0.5057,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7760428284630828,
+      "grad_norm": 0.17564809322357178,
+      "learning_rate": 1.7002930557836615e-05,
+      "loss": 0.5088,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7762658933749721,
+      "grad_norm": 0.16570429503917694,
+      "learning_rate": 1.7001250355596967e-05,
+      "loss": 0.5223,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7764889582868615,
+      "grad_norm": 0.16374748945236206,
+      "learning_rate": 1.6999569765584035e-05,
+      "loss": 0.4533,
+      "step": 3481
+    },
+    {
+      "epoch": 0.7767120231987509,
+      "grad_norm": 0.15578052401542664,
+      "learning_rate": 1.69978887878909e-05,
+      "loss": 0.4797,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7769350881106402,
+      "grad_norm": 0.15627166628837585,
+      "learning_rate": 1.6996207422610664e-05,
+      "loss": 0.4523,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7771581530225296,
+      "grad_norm": 0.1620151549577713,
+      "learning_rate": 1.6994525669836453e-05,
+      "loss": 0.4999,
+      "step": 3484
+    },
+    {
+      "epoch": 0.7773812179344189,
+      "grad_norm": 0.15806102752685547,
+      "learning_rate": 1.6992843529661413e-05,
+      "loss": 0.5013,
+      "step": 3485
+    },
+    {
+      "epoch": 0.7776042828463082,
+      "grad_norm": 0.25901806354522705,
+      "learning_rate": 1.6991161002178712e-05,
+      "loss": 0.5003,
+      "step": 3486
+    },
+    {
+      "epoch": 0.7778273477581976,
+      "grad_norm": 0.15611301362514496,
+      "learning_rate": 1.698947808748154e-05,
+      "loss": 0.5079,
+      "step": 3487
+    },
+    {
+      "epoch": 0.778050412670087,
+      "grad_norm": 0.1656789630651474,
+      "learning_rate": 1.6987794785663107e-05,
+      "loss": 0.4983,
+      "step": 3488
+    },
+    {
+      "epoch": 0.7782734775819764,
+      "grad_norm": 0.16347582638263702,
+      "learning_rate": 1.698611109681664e-05,
+      "loss": 0.4887,
+      "step": 3489
+    },
+    {
+      "epoch": 0.7784965424938657,
+      "grad_norm": 0.15885215997695923,
+      "learning_rate": 1.69844270210354e-05,
+      "loss": 0.4837,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7787196074057551,
+      "grad_norm": 0.17082704603672028,
+      "learning_rate": 1.698274255841265e-05,
+      "loss": 0.5165,
+      "step": 3491
+    },
+    {
+      "epoch": 0.7789426723176445,
+      "grad_norm": 0.16731634736061096,
+      "learning_rate": 1.6981057709041703e-05,
+      "loss": 0.485,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7791657372295338,
+      "grad_norm": 0.16583065688610077,
+      "learning_rate": 1.697937247301586e-05,
+      "loss": 0.5117,
+      "step": 3493
+    },
+    {
+      "epoch": 0.7793888021414231,
+      "grad_norm": 0.16518862545490265,
+      "learning_rate": 1.6977686850428475e-05,
+      "loss": 0.5141,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7796118670533125,
+      "grad_norm": 0.15372084081172943,
+      "learning_rate": 1.69760008413729e-05,
+      "loss": 0.4787,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7798349319652019,
+      "grad_norm": 0.16989074647426605,
+      "learning_rate": 1.6974314445942514e-05,
+      "loss": 0.4906,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7800579968770912,
+      "grad_norm": 0.16484175622463226,
+      "learning_rate": 1.697262766423072e-05,
+      "loss": 0.5,
+      "step": 3497
+    },
+    {
+      "epoch": 0.7802810617889806,
+      "grad_norm": 0.16868092119693756,
+      "learning_rate": 1.6970940496330953e-05,
+      "loss": 0.5106,
+      "step": 3498
+    },
+    {
+      "epoch": 0.78050412670087,
+      "grad_norm": 0.16630426049232483,
+      "learning_rate": 1.6969252942336648e-05,
+      "loss": 0.505,
+      "step": 3499
+    },
+    {
+      "epoch": 0.7807271916127593,
+      "grad_norm": 0.15219101309776306,
+      "learning_rate": 1.696756500234128e-05,
+      "loss": 0.4698,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7809502565246487,
+      "grad_norm": 0.16838675737380981,
+      "learning_rate": 1.6965876676438334e-05,
+      "loss": 0.4897,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7811733214365381,
+      "grad_norm": 0.15935415029525757,
+      "learning_rate": 1.696418796472132e-05,
+      "loss": 0.508,
+      "step": 3502
+    },
+    {
+      "epoch": 0.7813963863484273,
+      "grad_norm": 0.16184543073177338,
+      "learning_rate": 1.696249886728377e-05,
+      "loss": 0.5083,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7816194512603167,
+      "grad_norm": 0.15746904909610748,
+      "learning_rate": 1.6960809384219237e-05,
+      "loss": 0.4662,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7818425161722061,
+      "grad_norm": 0.1662943959236145,
+      "learning_rate": 1.6959119515621295e-05,
+      "loss": 0.4853,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7820655810840955,
+      "grad_norm": 0.16866953670978546,
+      "learning_rate": 1.695742926158354e-05,
+      "loss": 0.4919,
+      "step": 3506
+    },
+    {
+      "epoch": 0.7822886459959848,
+      "grad_norm": 0.1712365448474884,
+      "learning_rate": 1.695573862219959e-05,
+      "loss": 0.4882,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7825117109078742,
+      "grad_norm": 0.17081387341022491,
+      "learning_rate": 1.6954047597563078e-05,
+      "loss": 0.5068,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7827347758197636,
+      "grad_norm": 0.16168633103370667,
+      "learning_rate": 1.695235618776767e-05,
+      "loss": 0.5119,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7829578407316529,
+      "grad_norm": 0.2024863213300705,
+      "learning_rate": 1.6950664392907042e-05,
+      "loss": 0.5039,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7831809056435423,
+      "grad_norm": 0.1648394763469696,
+      "learning_rate": 1.6948972213074902e-05,
+      "loss": 0.5304,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7834039705554317,
+      "grad_norm": 0.15931276977062225,
+      "learning_rate": 1.6947279648364966e-05,
+      "loss": 0.5197,
+      "step": 3512
+    },
+    {
+      "epoch": 0.783627035467321,
+      "grad_norm": 0.15827037394046783,
+      "learning_rate": 1.6945586698870985e-05,
+      "loss": 0.513,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7838501003792103,
+      "grad_norm": 0.15963312983512878,
+      "learning_rate": 1.694389336468672e-05,
+      "loss": 0.5096,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7840731652910997,
+      "grad_norm": 0.1624821275472641,
+      "learning_rate": 1.694219964590597e-05,
+      "loss": 0.5086,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7842962302029891,
+      "grad_norm": 0.15603917837142944,
+      "learning_rate": 1.694050554262253e-05,
+      "loss": 0.499,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7845192951148784,
+      "grad_norm": 0.16541822254657745,
+      "learning_rate": 1.6938811054930237e-05,
+      "loss": 0.4843,
+      "step": 3517
+    },
+    {
+      "epoch": 0.7847423600267678,
+      "grad_norm": 0.16056667268276215,
+      "learning_rate": 1.693711618292294e-05,
+      "loss": 0.4992,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7849654249386572,
+      "grad_norm": 0.15387259423732758,
+      "learning_rate": 1.693542092669451e-05,
+      "loss": 0.5056,
+      "step": 3519
+    },
+    {
+      "epoch": 0.7851884898505465,
+      "grad_norm": 0.15904875099658966,
+      "learning_rate": 1.6933725286338846e-05,
+      "loss": 0.4718,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7854115547624358,
+      "grad_norm": 0.16445663571357727,
+      "learning_rate": 1.693202926194986e-05,
+      "loss": 0.4752,
+      "step": 3521
+    },
+    {
+      "epoch": 0.7856346196743252,
+      "grad_norm": 0.16833317279815674,
+      "learning_rate": 1.693033285362149e-05,
+      "loss": 0.4626,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7858576845862146,
+      "grad_norm": 0.17628587782382965,
+      "learning_rate": 1.692863606144769e-05,
+      "loss": 0.5001,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7860807494981039,
+      "grad_norm": 0.16575254499912262,
+      "learning_rate": 1.692693888552245e-05,
+      "loss": 0.4987,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7863038144099933,
+      "grad_norm": 0.1609877198934555,
+      "learning_rate": 1.6925241325939756e-05,
+      "loss": 0.4897,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7865268793218827,
+      "grad_norm": 0.1630864441394806,
+      "learning_rate": 1.6923543382793636e-05,
+      "loss": 0.473,
+      "step": 3526
+    },
+    {
+      "epoch": 0.786749944233772,
+      "grad_norm": 0.16511790454387665,
+      "learning_rate": 1.6921845056178133e-05,
+      "loss": 0.511,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7869730091456614,
+      "grad_norm": 0.15955112874507904,
+      "learning_rate": 1.6920146346187312e-05,
+      "loss": 0.4702,
+      "step": 3528
+    },
+    {
+      "epoch": 0.7871960740575508,
+      "grad_norm": 0.16080255806446075,
+      "learning_rate": 1.691844725291526e-05,
+      "loss": 0.4897,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7874191389694402,
+      "grad_norm": 0.15729734301567078,
+      "learning_rate": 1.6916747776456074e-05,
+      "loss": 0.477,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7876422038813294,
+      "grad_norm": 0.18572328984737396,
+      "learning_rate": 1.691504791690389e-05,
+      "loss": 0.4953,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7878652687932188,
+      "grad_norm": 0.17920143902301788,
+      "learning_rate": 1.6913347674352855e-05,
+      "loss": 0.4932,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7880883337051082,
+      "grad_norm": 0.14651532471179962,
+      "learning_rate": 1.691164704889714e-05,
+      "loss": 0.4531,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7883113986169975,
+      "grad_norm": 0.16487324237823486,
+      "learning_rate": 1.6909946040630935e-05,
+      "loss": 0.487,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7885344635288869,
+      "grad_norm": 0.16055698692798615,
+      "learning_rate": 1.6908244649648455e-05,
+      "loss": 0.4754,
+      "step": 3535
+    },
+    {
+      "epoch": 0.7887575284407763,
+      "grad_norm": 0.1739615648984909,
+      "learning_rate": 1.690654287604393e-05,
+      "loss": 0.5076,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7889805933526656,
+      "grad_norm": 0.17668306827545166,
+      "learning_rate": 1.690484071991162e-05,
+      "loss": 0.5177,
+      "step": 3537
+    },
+    {
+      "epoch": 0.789203658264555,
+      "grad_norm": 0.17033684253692627,
+      "learning_rate": 1.690313818134579e-05,
+      "loss": 0.4884,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7894267231764444,
+      "grad_norm": 0.19986379146575928,
+      "learning_rate": 1.690143526044075e-05,
+      "loss": 0.4885,
+      "step": 3539
+    },
+    {
+      "epoch": 0.7896497880883337,
+      "grad_norm": 0.17267881333827972,
+      "learning_rate": 1.6899731957290814e-05,
+      "loss": 0.498,
+      "step": 3540
+    },
+    {
+      "epoch": 0.789872853000223,
+      "grad_norm": 0.15866464376449585,
+      "learning_rate": 1.689802827199032e-05,
+      "loss": 0.4856,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7900959179121124,
+      "grad_norm": 0.17956510186195374,
+      "learning_rate": 1.689632420463363e-05,
+      "loss": 0.5046,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7903189828240018,
+      "grad_norm": 0.17033924162387848,
+      "learning_rate": 1.6894619755315127e-05,
+      "loss": 0.5272,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7905420477358911,
+      "grad_norm": 0.15526439249515533,
+      "learning_rate": 1.6892914924129212e-05,
+      "loss": 0.4858,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7907651126477805,
+      "grad_norm": 0.16875073313713074,
+      "learning_rate": 1.689120971117031e-05,
+      "loss": 0.4957,
+      "step": 3545
+    },
+    {
+      "epoch": 0.7909881775596699,
+      "grad_norm": 0.16168349981307983,
+      "learning_rate": 1.6889504116532868e-05,
+      "loss": 0.5041,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7912112424715593,
+      "grad_norm": 0.1654532253742218,
+      "learning_rate": 1.688779814031135e-05,
+      "loss": 0.5032,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7914343073834486,
+      "grad_norm": 0.16922220587730408,
+      "learning_rate": 1.6886091782600248e-05,
+      "loss": 0.5098,
+      "step": 3548
+    },
+    {
+      "epoch": 0.791657372295338,
+      "grad_norm": 0.17434757947921753,
+      "learning_rate": 1.6884385043494064e-05,
+      "loss": 0.5232,
+      "step": 3549
+    },
+    {
+      "epoch": 0.7918804372072273,
+      "grad_norm": 0.1557532101869583,
+      "learning_rate": 1.688267792308733e-05,
+      "loss": 0.4865,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7921035021191166,
+      "grad_norm": 0.1586635261774063,
+      "learning_rate": 1.6880970421474604e-05,
+      "loss": 0.4967,
+      "step": 3551
+    },
+    {
+      "epoch": 0.792326567031006,
+      "grad_norm": 0.16100507974624634,
+      "learning_rate": 1.6879262538750453e-05,
+      "loss": 0.4804,
+      "step": 3552
+    },
+    {
+      "epoch": 0.7925496319428954,
+      "grad_norm": 0.15817134082317352,
+      "learning_rate": 1.6877554275009467e-05,
+      "loss": 0.5066,
+      "step": 3553
+    },
+    {
+      "epoch": 0.7927726968547848,
+      "grad_norm": 0.15588125586509705,
+      "learning_rate": 1.6875845630346265e-05,
+      "loss": 0.5074,
+      "step": 3554
+    },
+    {
+      "epoch": 0.7929957617666741,
+      "grad_norm": 0.29957878589630127,
+      "learning_rate": 1.687413660485548e-05,
+      "loss": 0.4962,
+      "step": 3555
+    },
+    {
+      "epoch": 0.7932188266785635,
+      "grad_norm": 0.15556854009628296,
+      "learning_rate": 1.6872427198631772e-05,
+      "loss": 0.4746,
+      "step": 3556
+    },
+    {
+      "epoch": 0.7934418915904529,
+      "grad_norm": 0.1643342226743698,
+      "learning_rate": 1.6870717411769818e-05,
+      "loss": 0.5085,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7936649565023421,
+      "grad_norm": 0.16817674040794373,
+      "learning_rate": 1.686900724436431e-05,
+      "loss": 0.5078,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7938880214142315,
+      "grad_norm": 0.17203262448310852,
+      "learning_rate": 1.6867296696509978e-05,
+      "loss": 0.4775,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7941110863261209,
+      "grad_norm": 0.1592828631401062,
+      "learning_rate": 1.6865585768301556e-05,
+      "loss": 0.487,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7943341512380102,
+      "grad_norm": 0.15609320998191833,
+      "learning_rate": 1.6863874459833806e-05,
+      "loss": 0.4696,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7945572161498996,
+      "grad_norm": 0.17543616890907288,
+      "learning_rate": 1.6862162771201515e-05,
+      "loss": 0.5098,
+      "step": 3562
+    },
+    {
+      "epoch": 0.794780281061789,
+      "grad_norm": 0.17788106203079224,
+      "learning_rate": 1.6860450702499486e-05,
+      "loss": 0.4826,
+      "step": 3563
+    },
+    {
+      "epoch": 0.7950033459736784,
+      "grad_norm": 0.16301631927490234,
+      "learning_rate": 1.685873825382254e-05,
+      "loss": 0.5087,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7952264108855677,
+      "grad_norm": 0.16601480543613434,
+      "learning_rate": 1.685702542526553e-05,
+      "loss": 0.4624,
+      "step": 3565
+    },
+    {
+      "epoch": 0.7954494757974571,
+      "grad_norm": 0.7922313809394836,
+      "learning_rate": 1.6855312216923316e-05,
+      "loss": 0.5045,
+      "step": 3566
+    },
+    {
+      "epoch": 0.7956725407093465,
+      "grad_norm": 0.2716256380081177,
+      "learning_rate": 1.685359862889079e-05,
+      "loss": 0.5,
+      "step": 3567
+    },
+    {
+      "epoch": 0.7958956056212357,
+      "grad_norm": 0.1633540540933609,
+      "learning_rate": 1.6851884661262864e-05,
+      "loss": 0.4944,
+      "step": 3568
+    },
+    {
+      "epoch": 0.7961186705331251,
+      "grad_norm": 0.17535462975502014,
+      "learning_rate": 1.6850170314134465e-05,
+      "loss": 0.5251,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7963417354450145,
+      "grad_norm": 0.18369325995445251,
+      "learning_rate": 1.6848455587600542e-05,
+      "loss": 0.5043,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7965648003569039,
+      "grad_norm": 0.15560777485370636,
+      "learning_rate": 1.6846740481756072e-05,
+      "loss": 0.4686,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7967878652687932,
+      "grad_norm": 0.18801093101501465,
+      "learning_rate": 1.6845024996696047e-05,
+      "loss": 0.4946,
+      "step": 3572
+    },
+    {
+      "epoch": 0.7970109301806826,
+      "grad_norm": 0.17000144720077515,
+      "learning_rate": 1.684330913251548e-05,
+      "loss": 0.5,
+      "step": 3573
+    },
+    {
+      "epoch": 0.797233995092572,
+      "grad_norm": 0.16179999709129333,
+      "learning_rate": 1.6841592889309405e-05,
+      "loss": 0.4769,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7974570600044613,
+      "grad_norm": 0.16142041981220245,
+      "learning_rate": 1.6839876267172883e-05,
+      "loss": 0.4644,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7976801249163507,
+      "grad_norm": 0.16446426510810852,
+      "learning_rate": 1.683815926620099e-05,
+      "loss": 0.4835,
+      "step": 3576
+    },
+    {
+      "epoch": 0.79790318982824,
+      "grad_norm": 0.16954201459884644,
+      "learning_rate": 1.6836441886488822e-05,
+      "loss": 0.5151,
+      "step": 3577
+    },
+    {
+      "epoch": 0.7981262547401293,
+      "grad_norm": 0.16908320784568787,
+      "learning_rate": 1.6834724128131496e-05,
+      "loss": 0.4714,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7983493196520187,
+      "grad_norm": 0.1670169085264206,
+      "learning_rate": 1.683300599122416e-05,
+      "loss": 0.5004,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7985723845639081,
+      "grad_norm": 0.17246994376182556,
+      "learning_rate": 1.683128747586197e-05,
+      "loss": 0.5066,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7987954494757975,
+      "grad_norm": 0.16627082228660583,
+      "learning_rate": 1.6829568582140108e-05,
+      "loss": 0.4813,
+      "step": 3581
+    },
+    {
+      "epoch": 0.7990185143876868,
+      "grad_norm": 0.16429170966148376,
+      "learning_rate": 1.6827849310153778e-05,
+      "loss": 0.4916,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7992415792995762,
+      "grad_norm": 0.1685648411512375,
+      "learning_rate": 1.6826129659998204e-05,
+      "loss": 0.4821,
+      "step": 3583
+    },
+    {
+      "epoch": 0.7994646442114656,
+      "grad_norm": 0.1623144894838333,
+      "learning_rate": 1.6824409631768633e-05,
+      "loss": 0.5092,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7996877091233549,
+      "grad_norm": 0.16186010837554932,
+      "learning_rate": 1.6822689225560328e-05,
+      "loss": 0.4647,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7999107740352442,
+      "grad_norm": 0.15835130214691162,
+      "learning_rate": 1.682096844146858e-05,
+      "loss": 0.5139,
+      "step": 3586
+    },
+    {
+      "epoch": 0.8001338389471336,
+      "grad_norm": 0.1640879362821579,
+      "learning_rate": 1.681924727958869e-05,
+      "loss": 0.4815,
+      "step": 3587
+    },
+    {
+      "epoch": 0.800356903859023,
+      "grad_norm": 0.18087539076805115,
+      "learning_rate": 1.681752574001599e-05,
+      "loss": 0.5057,
+      "step": 3588
+    },
+    {
+      "epoch": 0.8005799687709123,
+      "grad_norm": 0.16318152844905853,
+      "learning_rate": 1.6815803822845834e-05,
+      "loss": 0.4945,
+      "step": 3589
+    },
+    {
+      "epoch": 0.8008030336828017,
+      "grad_norm": 0.16909107565879822,
+      "learning_rate": 1.681408152817359e-05,
+      "loss": 0.5123,
+      "step": 3590
+    },
+    {
+      "epoch": 0.8010260985946911,
+      "grad_norm": 0.16913791000843048,
+      "learning_rate": 1.6812358856094652e-05,
+      "loss": 0.4933,
+      "step": 3591
+    },
+    {
+      "epoch": 0.8012491635065804,
+      "grad_norm": 0.16670222580432892,
+      "learning_rate": 1.681063580670442e-05,
+      "loss": 0.5109,
+      "step": 3592
+    },
+    {
+      "epoch": 0.8014722284184698,
+      "grad_norm": 0.16838552057743073,
+      "learning_rate": 1.680891238009834e-05,
+      "loss": 0.4926,
+      "step": 3593
+    },
+    {
+      "epoch": 0.8016952933303592,
+      "grad_norm": 0.15705366432666779,
+      "learning_rate": 1.6807188576371864e-05,
+      "loss": 0.4762,
+      "step": 3594
+    },
+    {
+      "epoch": 0.8019183582422484,
+      "grad_norm": 0.15505351126194,
+      "learning_rate": 1.6805464395620465e-05,
+      "loss": 0.4769,
+      "step": 3595
+    },
+    {
+      "epoch": 0.8021414231541378,
+      "grad_norm": 0.17056378722190857,
+      "learning_rate": 1.6803739837939642e-05,
+      "loss": 0.5038,
+      "step": 3596
+    },
+    {
+      "epoch": 0.8023644880660272,
+      "grad_norm": 0.15908926725387573,
+      "learning_rate": 1.6802014903424905e-05,
+      "loss": 0.5079,
+      "step": 3597
+    },
+    {
+      "epoch": 0.8025875529779166,
+      "grad_norm": 0.16009296476840973,
+      "learning_rate": 1.68002895921718e-05,
+      "loss": 0.497,
+      "step": 3598
+    },
+    {
+      "epoch": 0.8028106178898059,
+      "grad_norm": 0.17202340066432953,
+      "learning_rate": 1.6798563904275882e-05,
+      "loss": 0.5097,
+      "step": 3599
+    },
+    {
+      "epoch": 0.8030336828016953,
+      "grad_norm": 0.16179756820201874,
+      "learning_rate": 1.679683783983273e-05,
+      "loss": 0.4992,
+      "step": 3600
+    },
+    {
+      "epoch": 0.8032567477135847,
+      "grad_norm": 0.16165940463542938,
+      "learning_rate": 1.6795111398937944e-05,
+      "loss": 0.4959,
+      "step": 3601
+    },
+    {
+      "epoch": 0.803479812625474,
+      "grad_norm": 0.17612482607364655,
+      "learning_rate": 1.679338458168714e-05,
+      "loss": 0.4937,
+      "step": 3602
+    },
+    {
+      "epoch": 0.8037028775373634,
+      "grad_norm": 0.16228753328323364,
+      "learning_rate": 1.679165738817597e-05,
+      "loss": 0.5287,
+      "step": 3603
+    },
+    {
+      "epoch": 0.8039259424492528,
+      "grad_norm": 0.17843377590179443,
+      "learning_rate": 1.6789929818500096e-05,
+      "loss": 0.4851,
+      "step": 3604
+    },
+    {
+      "epoch": 0.8041490073611421,
+      "grad_norm": 0.1637219339609146,
+      "learning_rate": 1.6788201872755196e-05,
+      "loss": 0.4633,
+      "step": 3605
+    },
+    {
+      "epoch": 0.8043720722730314,
+      "grad_norm": 0.16961325705051422,
+      "learning_rate": 1.6786473551036978e-05,
+      "loss": 0.5084,
+      "step": 3606
+    },
+    {
+      "epoch": 0.8045951371849208,
+      "grad_norm": 0.15785901248455048,
+      "learning_rate": 1.6784744853441167e-05,
+      "loss": 0.4855,
+      "step": 3607
+    },
+    {
+      "epoch": 0.8048182020968102,
+      "grad_norm": 0.1613706648349762,
+      "learning_rate": 1.6783015780063503e-05,
+      "loss": 0.4674,
+      "step": 3608
+    },
+    {
+      "epoch": 0.8050412670086995,
+      "grad_norm": 0.15770243108272552,
+      "learning_rate": 1.678128633099976e-05,
+      "loss": 0.4802,
+      "step": 3609
+    },
+    {
+      "epoch": 0.8052643319205889,
+      "grad_norm": 0.16045016050338745,
+      "learning_rate": 1.677955650634573e-05,
+      "loss": 0.4965,
+      "step": 3610
+    },
+    {
+      "epoch": 0.8054873968324783,
+      "grad_norm": 0.15433759987354279,
+      "learning_rate": 1.6777826306197208e-05,
+      "loss": 0.4752,
+      "step": 3611
+    },
+    {
+      "epoch": 0.8057104617443676,
+      "grad_norm": 0.16770930588245392,
+      "learning_rate": 1.6776095730650034e-05,
+      "loss": 0.5025,
+      "step": 3612
+    },
+    {
+      "epoch": 0.805933526656257,
+      "grad_norm": 0.15930598974227905,
+      "learning_rate": 1.6774364779800057e-05,
+      "loss": 0.4995,
+      "step": 3613
+    },
+    {
+      "epoch": 0.8061565915681463,
+      "grad_norm": 0.16089414060115814,
+      "learning_rate": 1.6772633453743142e-05,
+      "loss": 0.4973,
+      "step": 3614
+    },
+    {
+      "epoch": 0.8063796564800357,
+      "grad_norm": 0.1723753660917282,
+      "learning_rate": 1.6770901752575186e-05,
+      "loss": 0.5267,
+      "step": 3615
+    },
+    {
+      "epoch": 0.806602721391925,
+      "grad_norm": 0.15985791385173798,
+      "learning_rate": 1.6769169676392103e-05,
+      "loss": 0.4873,
+      "step": 3616
+    },
+    {
+      "epoch": 0.8068257863038144,
+      "grad_norm": 0.15705527365207672,
+      "learning_rate": 1.676743722528982e-05,
+      "loss": 0.48,
+      "step": 3617
+    },
+    {
+      "epoch": 0.8070488512157038,
+      "grad_norm": 0.17080652713775635,
+      "learning_rate": 1.6765704399364297e-05,
+      "loss": 0.507,
+      "step": 3618
+    },
+    {
+      "epoch": 0.8072719161275931,
+      "grad_norm": 0.16177797317504883,
+      "learning_rate": 1.6763971198711505e-05,
+      "loss": 0.5168,
+      "step": 3619
+    },
+    {
+      "epoch": 0.8074949810394825,
+      "grad_norm": 0.15763692557811737,
+      "learning_rate": 1.6762237623427445e-05,
+      "loss": 0.4736,
+      "step": 3620
+    },
+    {
+      "epoch": 0.8077180459513719,
+      "grad_norm": 0.15657658874988556,
+      "learning_rate": 1.6760503673608123e-05,
+      "loss": 0.479,
+      "step": 3621
+    },
+    {
+      "epoch": 0.8079411108632613,
+      "grad_norm": 0.15933813154697418,
+      "learning_rate": 1.6758769349349586e-05,
+      "loss": 0.519,
+      "step": 3622
+    },
+    {
+      "epoch": 0.8081641757751505,
+      "grad_norm": 0.173330619931221,
+      "learning_rate": 1.675703465074789e-05,
+      "loss": 0.5056,
+      "step": 3623
+    },
+    {
+      "epoch": 0.8083872406870399,
+      "grad_norm": 0.16727612912654877,
+      "learning_rate": 1.6755299577899107e-05,
+      "loss": 0.4949,
+      "step": 3624
+    },
+    {
+      "epoch": 0.8086103055989293,
+      "grad_norm": 0.23466536402702332,
+      "learning_rate": 1.6753564130899343e-05,
+      "loss": 0.5203,
+      "step": 3625
+    },
+    {
+      "epoch": 0.8088333705108186,
+      "grad_norm": 0.16051769256591797,
+      "learning_rate": 1.6751828309844714e-05,
+      "loss": 0.4907,
+      "step": 3626
+    },
+    {
+      "epoch": 0.809056435422708,
+      "grad_norm": 0.16254782676696777,
+      "learning_rate": 1.6750092114831368e-05,
+      "loss": 0.4976,
+      "step": 3627
+    },
+    {
+      "epoch": 0.8092795003345974,
+      "grad_norm": 0.1722305417060852,
+      "learning_rate": 1.6748355545955456e-05,
+      "loss": 0.4932,
+      "step": 3628
+    },
+    {
+      "epoch": 0.8095025652464868,
+      "grad_norm": 0.16842518746852875,
+      "learning_rate": 1.6746618603313165e-05,
+      "loss": 0.4934,
+      "step": 3629
+    },
+    {
+      "epoch": 0.8097256301583761,
+      "grad_norm": 0.16397298872470856,
+      "learning_rate": 1.67448812870007e-05,
+      "loss": 0.4824,
+      "step": 3630
+    },
+    {
+      "epoch": 0.8099486950702655,
+      "grad_norm": 0.15751919150352478,
+      "learning_rate": 1.674314359711428e-05,
+      "loss": 0.5057,
+      "step": 3631
+    },
+    {
+      "epoch": 0.8101717599821548,
+      "grad_norm": 0.15799839794635773,
+      "learning_rate": 1.6741405533750154e-05,
+      "loss": 0.4941,
+      "step": 3632
+    },
+    {
+      "epoch": 0.8103948248940441,
+      "grad_norm": 0.16038256883621216,
+      "learning_rate": 1.6739667097004583e-05,
+      "loss": 0.4765,
+      "step": 3633
+    },
+    {
+      "epoch": 0.8106178898059335,
+      "grad_norm": 0.18798737227916718,
+      "learning_rate": 1.6737928286973852e-05,
+      "loss": 0.5119,
+      "step": 3634
+    },
+    {
+      "epoch": 0.8108409547178229,
+      "grad_norm": 0.1568623036146164,
+      "learning_rate": 1.673618910375427e-05,
+      "loss": 0.4881,
+      "step": 3635
+    },
+    {
+      "epoch": 0.8110640196297122,
+      "grad_norm": 0.16038067638874054,
+      "learning_rate": 1.6734449547442165e-05,
+      "loss": 0.491,
+      "step": 3636
+    },
+    {
+      "epoch": 0.8112870845416016,
+      "grad_norm": 0.1699703484773636,
+      "learning_rate": 1.6732709618133882e-05,
+      "loss": 0.4772,
+      "step": 3637
+    },
+    {
+      "epoch": 0.811510149453491,
+      "grad_norm": 0.16090309619903564,
+      "learning_rate": 1.673096931592579e-05,
+      "loss": 0.4703,
+      "step": 3638
+    },
+    {
+      "epoch": 0.8117332143653804,
+      "grad_norm": 0.16035638749599457,
+      "learning_rate": 1.672922864091428e-05,
+      "loss": 0.5076,
+      "step": 3639
+    },
+    {
+      "epoch": 0.8119562792772697,
+      "grad_norm": 0.16750063002109528,
+      "learning_rate": 1.6727487593195757e-05,
+      "loss": 0.4961,
+      "step": 3640
+    },
+    {
+      "epoch": 0.812179344189159,
+      "grad_norm": 0.15547265112400055,
+      "learning_rate": 1.6725746172866652e-05,
+      "loss": 0.4759,
+      "step": 3641
+    },
+    {
+      "epoch": 0.8124024091010484,
+      "grad_norm": 0.15655042231082916,
+      "learning_rate": 1.672400438002342e-05,
+      "loss": 0.511,
+      "step": 3642
+    },
+    {
+      "epoch": 0.8126254740129377,
+      "grad_norm": 0.1612492948770523,
+      "learning_rate": 1.6722262214762527e-05,
+      "loss": 0.4901,
+      "step": 3643
+    },
+    {
+      "epoch": 0.8128485389248271,
+      "grad_norm": 0.15349607169628143,
+      "learning_rate": 1.6720519677180472e-05,
+      "loss": 0.5094,
+      "step": 3644
+    },
+    {
+      "epoch": 0.8130716038367165,
+      "grad_norm": 0.16423405706882477,
+      "learning_rate": 1.671877676737376e-05,
+      "loss": 0.513,
+      "step": 3645
+    },
+    {
+      "epoch": 0.8132946687486059,
+      "grad_norm": 0.1583424061536789,
+      "learning_rate": 1.671703348543893e-05,
+      "loss": 0.4795,
+      "step": 3646
+    },
+    {
+      "epoch": 0.8135177336604952,
+      "grad_norm": 0.16294583678245544,
+      "learning_rate": 1.671528983147253e-05,
+      "loss": 0.5019,
+      "step": 3647
+    },
+    {
+      "epoch": 0.8137407985723846,
+      "grad_norm": 0.1648416668176651,
+      "learning_rate": 1.671354580557114e-05,
+      "loss": 0.4879,
+      "step": 3648
+    },
+    {
+      "epoch": 0.813963863484274,
+      "grad_norm": 0.16462133824825287,
+      "learning_rate": 1.6711801407831356e-05,
+      "loss": 0.4911,
+      "step": 3649
+    },
+    {
+      "epoch": 0.8141869283961632,
+      "grad_norm": 0.26390138268470764,
+      "learning_rate": 1.671005663834979e-05,
+      "loss": 0.4822,
+      "step": 3650
+    },
+    {
+      "epoch": 0.8144099933080526,
+      "grad_norm": 0.17741946876049042,
+      "learning_rate": 1.670831149722308e-05,
+      "loss": 0.5237,
+      "step": 3651
+    },
+    {
+      "epoch": 0.814633058219942,
+      "grad_norm": 0.18031084537506104,
+      "learning_rate": 1.670656598454788e-05,
+      "loss": 0.4956,
+      "step": 3652
+    },
+    {
+      "epoch": 0.8148561231318313,
+      "grad_norm": 0.1602051854133606,
+      "learning_rate": 1.670482010042087e-05,
+      "loss": 0.4695,
+      "step": 3653
+    },
+    {
+      "epoch": 0.8150791880437207,
+      "grad_norm": 0.15887601673603058,
+      "learning_rate": 1.670307384493875e-05,
+      "loss": 0.4907,
+      "step": 3654
+    },
+    {
+      "epoch": 0.8153022529556101,
+      "grad_norm": 0.16547198593616486,
+      "learning_rate": 1.6701327218198234e-05,
+      "loss": 0.4944,
+      "step": 3655
+    },
+    {
+      "epoch": 0.8155253178674995,
+      "grad_norm": 0.16335895657539368,
+      "learning_rate": 1.6699580220296065e-05,
+      "loss": 0.484,
+      "step": 3656
+    },
+    {
+      "epoch": 0.8157483827793888,
+      "grad_norm": 0.1715383529663086,
+      "learning_rate": 1.6697832851329002e-05,
+      "loss": 0.5086,
+      "step": 3657
+    },
+    {
+      "epoch": 0.8159714476912782,
+      "grad_norm": 0.16271516680717468,
+      "learning_rate": 1.6696085111393825e-05,
+      "loss": 0.4771,
+      "step": 3658
+    },
+    {
+      "epoch": 0.8161945126031676,
+      "grad_norm": 0.16480712592601776,
+      "learning_rate": 1.6694337000587334e-05,
+      "loss": 0.5099,
+      "step": 3659
+    },
+    {
+      "epoch": 0.8164175775150568,
+      "grad_norm": 0.15898552536964417,
+      "learning_rate": 1.669258851900635e-05,
+      "loss": 0.4987,
+      "step": 3660
+    },
+    {
+      "epoch": 0.8166406424269462,
+      "grad_norm": 0.16049543023109436,
+      "learning_rate": 1.6690839666747717e-05,
+      "loss": 0.4762,
+      "step": 3661
+    },
+    {
+      "epoch": 0.8168637073388356,
+      "grad_norm": 0.15958286821842194,
+      "learning_rate": 1.6689090443908296e-05,
+      "loss": 0.4827,
+      "step": 3662
+    },
+    {
+      "epoch": 0.817086772250725,
+      "grad_norm": 0.16123978793621063,
+      "learning_rate": 1.668734085058497e-05,
+      "loss": 0.4803,
+      "step": 3663
+    },
+    {
+      "epoch": 0.8173098371626143,
+      "grad_norm": 0.1554545909166336,
+      "learning_rate": 1.668559088687464e-05,
+      "loss": 0.4813,
+      "step": 3664
+    },
+    {
+      "epoch": 0.8175329020745037,
+      "grad_norm": 0.15960079431533813,
+      "learning_rate": 1.6683840552874235e-05,
+      "loss": 0.4947,
+      "step": 3665
+    },
+    {
+      "epoch": 0.8177559669863931,
+      "grad_norm": 0.15964864194393158,
+      "learning_rate": 1.6682089848680698e-05,
+      "loss": 0.4815,
+      "step": 3666
+    },
+    {
+      "epoch": 0.8179790318982824,
+      "grad_norm": 0.16264641284942627,
+      "learning_rate": 1.6680338774390993e-05,
+      "loss": 0.509,
+      "step": 3667
+    },
+    {
+      "epoch": 0.8182020968101718,
+      "grad_norm": 0.164497509598732,
+      "learning_rate": 1.6678587330102103e-05,
+      "loss": 0.5247,
+      "step": 3668
+    },
+    {
+      "epoch": 0.8184251617220611,
+      "grad_norm": 0.16061222553253174,
+      "learning_rate": 1.667683551591104e-05,
+      "loss": 0.5161,
+      "step": 3669
+    },
+    {
+      "epoch": 0.8186482266339504,
+      "grad_norm": 0.19114501774311066,
+      "learning_rate": 1.6675083331914823e-05,
+      "loss": 0.5119,
+      "step": 3670
+    },
+    {
+      "epoch": 0.8188712915458398,
+      "grad_norm": 0.18399201333522797,
+      "learning_rate": 1.6673330778210508e-05,
+      "loss": 0.4907,
+      "step": 3671
+    },
+    {
+      "epoch": 0.8190943564577292,
+      "grad_norm": 0.15932810306549072,
+      "learning_rate": 1.6671577854895153e-05,
+      "loss": 0.5102,
+      "step": 3672
+    },
+    {
+      "epoch": 0.8193174213696186,
+      "grad_norm": 0.1565459966659546,
+      "learning_rate": 1.6669824562065856e-05,
+      "loss": 0.4752,
+      "step": 3673
+    },
+    {
+      "epoch": 0.8195404862815079,
+      "grad_norm": 0.1646818071603775,
+      "learning_rate": 1.6668070899819714e-05,
+      "loss": 0.5099,
+      "step": 3674
+    },
+    {
+      "epoch": 0.8197635511933973,
+      "grad_norm": 0.1591091752052307,
+      "learning_rate": 1.6666316868253867e-05,
+      "loss": 0.4969,
+      "step": 3675
+    },
+    {
+      "epoch": 0.8199866161052867,
+      "grad_norm": 0.16319642961025238,
+      "learning_rate": 1.6664562467465455e-05,
+      "loss": 0.5129,
+      "step": 3676
+    },
+    {
+      "epoch": 0.820209681017176,
+      "grad_norm": 0.16143767535686493,
+      "learning_rate": 1.6662807697551654e-05,
+      "loss": 0.4952,
+      "step": 3677
+    },
+    {
+      "epoch": 0.8204327459290653,
+      "grad_norm": 0.15746799111366272,
+      "learning_rate": 1.666105255860965e-05,
+      "loss": 0.4947,
+      "step": 3678
+    },
+    {
+      "epoch": 0.8206558108409547,
+      "grad_norm": 0.16585499048233032,
+      "learning_rate": 1.6659297050736657e-05,
+      "loss": 0.5134,
+      "step": 3679
+    },
+    {
+      "epoch": 0.8208788757528441,
+      "grad_norm": 0.16004452109336853,
+      "learning_rate": 1.6657541174029902e-05,
+      "loss": 0.4995,
+      "step": 3680
+    },
+    {
+      "epoch": 0.8211019406647334,
+      "grad_norm": 0.1682739406824112,
+      "learning_rate": 1.665578492858664e-05,
+      "loss": 0.4796,
+      "step": 3681
+    },
+    {
+      "epoch": 0.8213250055766228,
+      "grad_norm": 0.1618606299161911,
+      "learning_rate": 1.6654028314504147e-05,
+      "loss": 0.4849,
+      "step": 3682
+    },
+    {
+      "epoch": 0.8215480704885122,
+      "grad_norm": 0.1632058471441269,
+      "learning_rate": 1.6652271331879706e-05,
+      "loss": 0.499,
+      "step": 3683
+    },
+    {
+      "epoch": 0.8217711354004015,
+      "grad_norm": 0.16575373709201813,
+      "learning_rate": 1.665051398081064e-05,
+      "loss": 0.5193,
+      "step": 3684
+    },
+    {
+      "epoch": 0.8219942003122909,
+      "grad_norm": 0.16228437423706055,
+      "learning_rate": 1.664875626139427e-05,
+      "loss": 0.5065,
+      "step": 3685
+    },
+    {
+      "epoch": 0.8222172652241803,
+      "grad_norm": 0.1525840312242508,
+      "learning_rate": 1.6646998173727955e-05,
+      "loss": 0.4904,
+      "step": 3686
+    },
+    {
+      "epoch": 0.8224403301360695,
+      "grad_norm": 0.16036051511764526,
+      "learning_rate": 1.6645239717909074e-05,
+      "loss": 0.4798,
+      "step": 3687
+    },
+    {
+      "epoch": 0.8226633950479589,
+      "grad_norm": 0.16049040853977203,
+      "learning_rate": 1.6643480894035015e-05,
+      "loss": 0.4985,
+      "step": 3688
+    },
+    {
+      "epoch": 0.8228864599598483,
+      "grad_norm": 0.16121259331703186,
+      "learning_rate": 1.6641721702203196e-05,
+      "loss": 0.4586,
+      "step": 3689
+    },
+    {
+      "epoch": 0.8231095248717377,
+      "grad_norm": 0.1610342562198639,
+      "learning_rate": 1.663996214251105e-05,
+      "loss": 0.4781,
+      "step": 3690
+    },
+    {
+      "epoch": 0.823332589783627,
+      "grad_norm": 0.17599108815193176,
+      "learning_rate": 1.6638202215056036e-05,
+      "loss": 0.509,
+      "step": 3691
+    },
+    {
+      "epoch": 0.8235556546955164,
+      "grad_norm": 0.15026211738586426,
+      "learning_rate": 1.6636441919935627e-05,
+      "loss": 0.4718,
+      "step": 3692
+    },
+    {
+      "epoch": 0.8237787196074058,
+      "grad_norm": 0.1567082703113556,
+      "learning_rate": 1.6634681257247314e-05,
+      "loss": 0.5084,
+      "step": 3693
+    },
+    {
+      "epoch": 0.8240017845192951,
+      "grad_norm": 0.1553567349910736,
+      "learning_rate": 1.6632920227088628e-05,
+      "loss": 0.481,
+      "step": 3694
+    },
+    {
+      "epoch": 0.8242248494311845,
+      "grad_norm": 0.16551898419857025,
+      "learning_rate": 1.663115882955709e-05,
+      "loss": 0.5145,
+      "step": 3695
+    },
+    {
+      "epoch": 0.8244479143430739,
+      "grad_norm": 0.1742102950811386,
+      "learning_rate": 1.6629397064750267e-05,
+      "loss": 0.5073,
+      "step": 3696
+    },
+    {
+      "epoch": 0.8246709792549632,
+      "grad_norm": 0.16208507120609283,
+      "learning_rate": 1.6627634932765735e-05,
+      "loss": 0.4637,
+      "step": 3697
+    },
+    {
+      "epoch": 0.8248940441668525,
+      "grad_norm": 0.1664377748966217,
+      "learning_rate": 1.662587243370109e-05,
+      "loss": 0.4927,
+      "step": 3698
+    },
+    {
+      "epoch": 0.8251171090787419,
+      "grad_norm": 0.16273631155490875,
+      "learning_rate": 1.662410956765395e-05,
+      "loss": 0.4861,
+      "step": 3699
+    },
+    {
+      "epoch": 0.8253401739906313,
+      "grad_norm": 0.16939525306224823,
+      "learning_rate": 1.6622346334721956e-05,
+      "loss": 0.4973,
+      "step": 3700
+    },
+    {
+      "epoch": 0.8255632389025206,
+      "grad_norm": 0.16064801812171936,
+      "learning_rate": 1.6620582735002762e-05,
+      "loss": 0.4815,
+      "step": 3701
+    },
+    {
+      "epoch": 0.82578630381441,
+      "grad_norm": 0.17662890255451202,
+      "learning_rate": 1.6618818768594058e-05,
+      "loss": 0.4851,
+      "step": 3702
+    },
+    {
+      "epoch": 0.8260093687262994,
+      "grad_norm": 0.16551104187965393,
+      "learning_rate": 1.6617054435593535e-05,
+      "loss": 0.4872,
+      "step": 3703
+    },
+    {
+      "epoch": 0.8262324336381888,
+      "grad_norm": 0.16171106696128845,
+      "learning_rate": 1.6615289736098912e-05,
+      "loss": 0.4801,
+      "step": 3704
+    },
+    {
+      "epoch": 0.826455498550078,
+      "grad_norm": 0.15718747675418854,
+      "learning_rate": 1.6613524670207933e-05,
+      "loss": 0.4687,
+      "step": 3705
+    },
+    {
+      "epoch": 0.8266785634619674,
+      "grad_norm": 0.16042965650558472,
+      "learning_rate": 1.6611759238018356e-05,
+      "loss": 0.5026,
+      "step": 3706
+    },
+    {
+      "epoch": 0.8269016283738568,
+      "grad_norm": 0.1648973822593689,
+      "learning_rate": 1.660999343962796e-05,
+      "loss": 0.5002,
+      "step": 3707
+    },
+    {
+      "epoch": 0.8271246932857461,
+      "grad_norm": 0.16398411989212036,
+      "learning_rate": 1.6608227275134555e-05,
+      "loss": 0.4601,
+      "step": 3708
+    },
+    {
+      "epoch": 0.8273477581976355,
+      "grad_norm": 0.16151902079582214,
+      "learning_rate": 1.6606460744635952e-05,
+      "loss": 0.5171,
+      "step": 3709
+    },
+    {
+      "epoch": 0.8275708231095249,
+      "grad_norm": 0.15724126994609833,
+      "learning_rate": 1.660469384823e-05,
+      "loss": 0.4794,
+      "step": 3710
+    },
+    {
+      "epoch": 0.8277938880214142,
+      "grad_norm": 0.16321682929992676,
+      "learning_rate": 1.6602926586014555e-05,
+      "loss": 0.4987,
+      "step": 3711
+    },
+    {
+      "epoch": 0.8280169529333036,
+      "grad_norm": 0.17783483862876892,
+      "learning_rate": 1.66011589580875e-05,
+      "loss": 0.4795,
+      "step": 3712
+    },
+    {
+      "epoch": 0.828240017845193,
+      "grad_norm": 0.16518917679786682,
+      "learning_rate": 1.659939096454674e-05,
+      "loss": 0.5131,
+      "step": 3713
+    },
+    {
+      "epoch": 0.8284630827570824,
+      "grad_norm": 0.1636318564414978,
+      "learning_rate": 1.6597622605490198e-05,
+      "loss": 0.4772,
+      "step": 3714
+    },
+    {
+      "epoch": 0.8286861476689716,
+      "grad_norm": 0.16061849892139435,
+      "learning_rate": 1.6595853881015814e-05,
+      "loss": 0.5001,
+      "step": 3715
+    },
+    {
+      "epoch": 0.828909212580861,
+      "grad_norm": 0.15872059762477875,
+      "learning_rate": 1.6594084791221554e-05,
+      "loss": 0.4987,
+      "step": 3716
+    },
+    {
+      "epoch": 0.8291322774927504,
+      "grad_norm": 0.15593315660953522,
+      "learning_rate": 1.65923153362054e-05,
+      "loss": 0.4752,
+      "step": 3717
+    },
+    {
+      "epoch": 0.8293553424046397,
+      "grad_norm": 0.15845991671085358,
+      "learning_rate": 1.6590545516065353e-05,
+      "loss": 0.5206,
+      "step": 3718
+    },
+    {
+      "epoch": 0.8295784073165291,
+      "grad_norm": 0.15479490160942078,
+      "learning_rate": 1.658877533089944e-05,
+      "loss": 0.4706,
+      "step": 3719
+    },
+    {
+      "epoch": 0.8298014722284185,
+      "grad_norm": 0.16462849080562592,
+      "learning_rate": 1.6587004780805704e-05,
+      "loss": 0.4816,
+      "step": 3720
+    },
+    {
+      "epoch": 0.8300245371403079,
+      "grad_norm": 0.17479833960533142,
+      "learning_rate": 1.658523386588221e-05,
+      "loss": 0.5083,
+      "step": 3721
+    },
+    {
+      "epoch": 0.8302476020521972,
+      "grad_norm": 0.16387352347373962,
+      "learning_rate": 1.658346258622704e-05,
+      "loss": 0.4897,
+      "step": 3722
+    },
+    {
+      "epoch": 0.8304706669640866,
+      "grad_norm": 0.16821224987506866,
+      "learning_rate": 1.6581690941938307e-05,
+      "loss": 0.4928,
+      "step": 3723
+    },
+    {
+      "epoch": 0.830693731875976,
+      "grad_norm": 0.15833254158496857,
+      "learning_rate": 1.657991893311412e-05,
+      "loss": 0.4523,
+      "step": 3724
+    },
+    {
+      "epoch": 0.8309167967878652,
+      "grad_norm": 0.15844541788101196,
+      "learning_rate": 1.657814655985264e-05,
+      "loss": 0.4989,
+      "step": 3725
+    },
+    {
+      "epoch": 0.8311398616997546,
+      "grad_norm": 0.16405586898326874,
+      "learning_rate": 1.657637382225202e-05,
+      "loss": 0.512,
+      "step": 3726
+    },
+    {
+      "epoch": 0.831362926611644,
+      "grad_norm": 0.16231343150138855,
+      "learning_rate": 1.6574600720410455e-05,
+      "loss": 0.4856,
+      "step": 3727
+    },
+    {
+      "epoch": 0.8315859915235333,
+      "grad_norm": 0.16474612057209015,
+      "learning_rate": 1.6572827254426145e-05,
+      "loss": 0.495,
+      "step": 3728
+    },
+    {
+      "epoch": 0.8318090564354227,
+      "grad_norm": 0.16462446749210358,
+      "learning_rate": 1.6571053424397316e-05,
+      "loss": 0.4931,
+      "step": 3729
+    },
+    {
+      "epoch": 0.8320321213473121,
+      "grad_norm": 0.1530153751373291,
+      "learning_rate": 1.6569279230422215e-05,
+      "loss": 0.4752,
+      "step": 3730
+    },
+    {
+      "epoch": 0.8322551862592015,
+      "grad_norm": 0.15689024329185486,
+      "learning_rate": 1.656750467259911e-05,
+      "loss": 0.4899,
+      "step": 3731
+    },
+    {
+      "epoch": 0.8324782511710908,
+      "grad_norm": 0.17177921533584595,
+      "learning_rate": 1.656572975102628e-05,
+      "loss": 0.4941,
+      "step": 3732
+    },
+    {
+      "epoch": 0.8327013160829801,
+      "grad_norm": 0.15630176663398743,
+      "learning_rate": 1.6563954465802042e-05,
+      "loss": 0.4768,
+      "step": 3733
+    },
+    {
+      "epoch": 0.8329243809948695,
+      "grad_norm": 0.17884239554405212,
+      "learning_rate": 1.6562178817024713e-05,
+      "loss": 0.5076,
+      "step": 3734
+    },
+    {
+      "epoch": 0.8331474459067588,
+      "grad_norm": 0.15849219262599945,
+      "learning_rate": 1.6560402804792644e-05,
+      "loss": 0.4948,
+      "step": 3735
+    },
+    {
+      "epoch": 0.8333705108186482,
+      "grad_norm": 0.15310537815093994,
+      "learning_rate": 1.65586264292042e-05,
+      "loss": 0.467,
+      "step": 3736
+    },
+    {
+      "epoch": 0.8335935757305376,
+      "grad_norm": 0.1629456728696823,
+      "learning_rate": 1.6556849690357776e-05,
+      "loss": 0.4794,
+      "step": 3737
+    },
+    {
+      "epoch": 0.833816640642427,
+      "grad_norm": 0.16876748204231262,
+      "learning_rate": 1.6555072588351765e-05,
+      "loss": 0.497,
+      "step": 3738
+    },
+    {
+      "epoch": 0.8340397055543163,
+      "grad_norm": 0.17010155320167542,
+      "learning_rate": 1.6553295123284605e-05,
+      "loss": 0.5199,
+      "step": 3739
+    },
+    {
+      "epoch": 0.8342627704662057,
+      "grad_norm": 0.16115908324718475,
+      "learning_rate": 1.6551517295254732e-05,
+      "loss": 0.5094,
+      "step": 3740
+    },
+    {
+      "epoch": 0.8344858353780951,
+      "grad_norm": 0.18651996552944183,
+      "learning_rate": 1.6549739104360627e-05,
+      "loss": 0.4963,
+      "step": 3741
+    },
+    {
+      "epoch": 0.8347089002899843,
+      "grad_norm": 0.16856126487255096,
+      "learning_rate": 1.6547960550700766e-05,
+      "loss": 0.4729,
+      "step": 3742
+    },
+    {
+      "epoch": 0.8349319652018737,
+      "grad_norm": 0.16244658827781677,
+      "learning_rate": 1.6546181634373666e-05,
+      "loss": 0.4964,
+      "step": 3743
+    },
+    {
+      "epoch": 0.8351550301137631,
+      "grad_norm": 0.15980856120586395,
+      "learning_rate": 1.654440235547785e-05,
+      "loss": 0.4741,
+      "step": 3744
+    },
+    {
+      "epoch": 0.8353780950256524,
+      "grad_norm": 0.15810216963291168,
+      "learning_rate": 1.6542622714111865e-05,
+      "loss": 0.4905,
+      "step": 3745
+    },
+    {
+      "epoch": 0.8356011599375418,
+      "grad_norm": 0.15649911761283875,
+      "learning_rate": 1.654084271037428e-05,
+      "loss": 0.4883,
+      "step": 3746
+    },
+    {
+      "epoch": 0.8358242248494312,
+      "grad_norm": 0.16376326978206635,
+      "learning_rate": 1.653906234436368e-05,
+      "loss": 0.4788,
+      "step": 3747
+    },
+    {
+      "epoch": 0.8360472897613206,
+      "grad_norm": 0.17131556570529938,
+      "learning_rate": 1.6537281616178674e-05,
+      "loss": 0.5121,
+      "step": 3748
+    },
+    {
+      "epoch": 0.8362703546732099,
+      "grad_norm": 0.1694851517677307,
+      "learning_rate": 1.6535500525917893e-05,
+      "loss": 0.5165,
+      "step": 3749
+    },
+    {
+      "epoch": 0.8364934195850993,
+      "grad_norm": 0.16306069493293762,
+      "learning_rate": 1.653371907367998e-05,
+      "loss": 0.493,
+      "step": 3750
+    },
+    {
+      "epoch": 0.8367164844969887,
+      "grad_norm": 0.1630067676305771,
+      "learning_rate": 1.6531937259563612e-05,
+      "loss": 0.4992,
+      "step": 3751
+    },
+    {
+      "epoch": 0.8369395494088779,
+      "grad_norm": 0.15753702819347382,
+      "learning_rate": 1.6530155083667468e-05,
+      "loss": 0.4767,
+      "step": 3752
+    },
+    {
+      "epoch": 0.8371626143207673,
+      "grad_norm": 0.1651187390089035,
+      "learning_rate": 1.6528372546090258e-05,
+      "loss": 0.4987,
+      "step": 3753
+    },
+    {
+      "epoch": 0.8373856792326567,
+      "grad_norm": 0.1615830808877945,
+      "learning_rate": 1.6526589646930712e-05,
+      "loss": 0.4932,
+      "step": 3754
+    },
+    {
+      "epoch": 0.8376087441445461,
+      "grad_norm": 0.1610192209482193,
+      "learning_rate": 1.6524806386287578e-05,
+      "loss": 0.5062,
+      "step": 3755
+    },
+    {
+      "epoch": 0.8378318090564354,
+      "grad_norm": 0.16058041155338287,
+      "learning_rate": 1.652302276425962e-05,
+      "loss": 0.4937,
+      "step": 3756
+    },
+    {
+      "epoch": 0.8380548739683248,
+      "grad_norm": 0.15800607204437256,
+      "learning_rate": 1.6521238780945635e-05,
+      "loss": 0.4814,
+      "step": 3757
+    },
+    {
+      "epoch": 0.8382779388802142,
+      "grad_norm": 0.16751538217067719,
+      "learning_rate": 1.6519454436444423e-05,
+      "loss": 0.4904,
+      "step": 3758
+    },
+    {
+      "epoch": 0.8385010037921035,
+      "grad_norm": 0.16399426758289337,
+      "learning_rate": 1.651766973085482e-05,
+      "loss": 0.5086,
+      "step": 3759
+    },
+    {
+      "epoch": 0.8387240687039929,
+      "grad_norm": 0.16619375348091125,
+      "learning_rate": 1.6515884664275663e-05,
+      "loss": 0.5192,
+      "step": 3760
+    },
+    {
+      "epoch": 0.8389471336158822,
+      "grad_norm": 0.1645645648241043,
+      "learning_rate": 1.651409923680583e-05,
+      "loss": 0.4894,
+      "step": 3761
+    },
+    {
+      "epoch": 0.8391701985277715,
+      "grad_norm": 0.16611702740192413,
+      "learning_rate": 1.6512313448544207e-05,
+      "loss": 0.4858,
+      "step": 3762
+    },
+    {
+      "epoch": 0.8393932634396609,
+      "grad_norm": 0.15403629839420319,
+      "learning_rate": 1.6510527299589696e-05,
+      "loss": 0.4844,
+      "step": 3763
+    },
+    {
+      "epoch": 0.8396163283515503,
+      "grad_norm": 0.15380257368087769,
+      "learning_rate": 1.6508740790041236e-05,
+      "loss": 0.491,
+      "step": 3764
+    },
+    {
+      "epoch": 0.8398393932634397,
+      "grad_norm": 0.1628074049949646,
+      "learning_rate": 1.650695391999777e-05,
+      "loss": 0.4666,
+      "step": 3765
+    },
+    {
+      "epoch": 0.840062458175329,
+      "grad_norm": 0.16377408802509308,
+      "learning_rate": 1.650516668955826e-05,
+      "loss": 0.4695,
+      "step": 3766
+    },
+    {
+      "epoch": 0.8402855230872184,
+      "grad_norm": 0.16199228167533875,
+      "learning_rate": 1.6503379098821705e-05,
+      "loss": 0.48,
+      "step": 3767
+    },
+    {
+      "epoch": 0.8405085879991078,
+      "grad_norm": 0.17420238256454468,
+      "learning_rate": 1.6501591147887108e-05,
+      "loss": 0.4983,
+      "step": 3768
+    },
+    {
+      "epoch": 0.840731652910997,
+      "grad_norm": 0.1635866016149521,
+      "learning_rate": 1.649980283685349e-05,
+      "loss": 0.4877,
+      "step": 3769
+    },
+    {
+      "epoch": 0.8409547178228864,
+      "grad_norm": 0.16857163608074188,
+      "learning_rate": 1.6498014165819908e-05,
+      "loss": 0.4956,
+      "step": 3770
+    },
+    {
+      "epoch": 0.8411777827347758,
+      "grad_norm": 0.17237527668476105,
+      "learning_rate": 1.649622513488543e-05,
+      "loss": 0.515,
+      "step": 3771
+    },
+    {
+      "epoch": 0.8414008476466652,
+      "grad_norm": 0.1601458191871643,
+      "learning_rate": 1.6494435744149142e-05,
+      "loss": 0.4839,
+      "step": 3772
+    },
+    {
+      "epoch": 0.8416239125585545,
+      "grad_norm": 0.1580028235912323,
+      "learning_rate": 1.6492645993710148e-05,
+      "loss": 0.4529,
+      "step": 3773
+    },
+    {
+      "epoch": 0.8418469774704439,
+      "grad_norm": 0.16191108524799347,
+      "learning_rate": 1.649085588366758e-05,
+      "loss": 0.495,
+      "step": 3774
+    },
+    {
+      "epoch": 0.8420700423823333,
+      "grad_norm": 0.16524077951908112,
+      "learning_rate": 1.6489065414120583e-05,
+      "loss": 0.4832,
+      "step": 3775
+    },
+    {
+      "epoch": 0.8422931072942226,
+      "grad_norm": 0.16424553096294403,
+      "learning_rate": 1.6487274585168327e-05,
+      "loss": 0.4811,
+      "step": 3776
+    },
+    {
+      "epoch": 0.842516172206112,
+      "grad_norm": 0.16102652251720428,
+      "learning_rate": 1.6485483396909997e-05,
+      "loss": 0.486,
+      "step": 3777
+    },
+    {
+      "epoch": 0.8427392371180014,
+      "grad_norm": 0.16395264863967896,
+      "learning_rate": 1.64836918494448e-05,
+      "loss": 0.4956,
+      "step": 3778
+    },
+    {
+      "epoch": 0.8429623020298908,
+      "grad_norm": 0.17286796867847443,
+      "learning_rate": 1.6481899942871967e-05,
+      "loss": 0.4859,
+      "step": 3779
+    },
+    {
+      "epoch": 0.84318536694178,
+      "grad_norm": 0.16825571656227112,
+      "learning_rate": 1.648010767729074e-05,
+      "loss": 0.4873,
+      "step": 3780
+    },
+    {
+      "epoch": 0.8434084318536694,
+      "grad_norm": 0.14970429241657257,
+      "learning_rate": 1.647831505280039e-05,
+      "loss": 0.4725,
+      "step": 3781
+    },
+    {
+      "epoch": 0.8436314967655588,
+      "grad_norm": 0.1689714640378952,
+      "learning_rate": 1.64765220695002e-05,
+      "loss": 0.4865,
+      "step": 3782
+    },
+    {
+      "epoch": 0.8438545616774481,
+      "grad_norm": 0.18679800629615784,
+      "learning_rate": 1.647472872748948e-05,
+      "loss": 0.5144,
+      "step": 3783
+    },
+    {
+      "epoch": 0.8440776265893375,
+      "grad_norm": 0.15957824885845184,
+      "learning_rate": 1.6472935026867555e-05,
+      "loss": 0.4969,
+      "step": 3784
+    },
+    {
+      "epoch": 0.8443006915012269,
+      "grad_norm": 0.1743471622467041,
+      "learning_rate": 1.6471140967733772e-05,
+      "loss": 0.4945,
+      "step": 3785
+    },
+    {
+      "epoch": 0.8445237564131162,
+      "grad_norm": 0.17243602871894836,
+      "learning_rate": 1.64693465501875e-05,
+      "loss": 0.4963,
+      "step": 3786
+    },
+    {
+      "epoch": 0.8447468213250056,
+      "grad_norm": 0.1854105442762375,
+      "learning_rate": 1.646755177432812e-05,
+      "loss": 0.4655,
+      "step": 3787
+    },
+    {
+      "epoch": 0.844969886236895,
+      "grad_norm": 0.1698131114244461,
+      "learning_rate": 1.6465756640255038e-05,
+      "loss": 0.5208,
+      "step": 3788
+    },
+    {
+      "epoch": 0.8451929511487843,
+      "grad_norm": 0.17987042665481567,
+      "learning_rate": 1.6463961148067685e-05,
+      "loss": 0.5031,
+      "step": 3789
+    },
+    {
+      "epoch": 0.8454160160606736,
+      "grad_norm": 0.15817251801490784,
+      "learning_rate": 1.6462165297865503e-05,
+      "loss": 0.4892,
+      "step": 3790
+    },
+    {
+      "epoch": 0.845639080972563,
+      "grad_norm": 0.16425161063671112,
+      "learning_rate": 1.6460369089747956e-05,
+      "loss": 0.4867,
+      "step": 3791
+    },
+    {
+      "epoch": 0.8458621458844524,
+      "grad_norm": 0.164134219288826,
+      "learning_rate": 1.6458572523814535e-05,
+      "loss": 0.5063,
+      "step": 3792
+    },
+    {
+      "epoch": 0.8460852107963417,
+      "grad_norm": 0.1651495099067688,
+      "learning_rate": 1.6456775600164737e-05,
+      "loss": 0.5168,
+      "step": 3793
+    },
+    {
+      "epoch": 0.8463082757082311,
+      "grad_norm": 0.1596038043498993,
+      "learning_rate": 1.6454978318898093e-05,
+      "loss": 0.4948,
+      "step": 3794
+    },
+    {
+      "epoch": 0.8465313406201205,
+      "grad_norm": 0.16954916715621948,
+      "learning_rate": 1.645318068011415e-05,
+      "loss": 0.5364,
+      "step": 3795
+    },
+    {
+      "epoch": 0.8467544055320099,
+      "grad_norm": 0.16513592004776,
+      "learning_rate": 1.6451382683912468e-05,
+      "loss": 0.4905,
+      "step": 3796
+    },
+    {
+      "epoch": 0.8469774704438991,
+      "grad_norm": 0.17096330225467682,
+      "learning_rate": 1.6449584330392627e-05,
+      "loss": 0.4687,
+      "step": 3797
+    },
+    {
+      "epoch": 0.8472005353557885,
+      "grad_norm": 0.16846898198127747,
+      "learning_rate": 1.644778561965424e-05,
+      "loss": 0.4803,
+      "step": 3798
+    },
+    {
+      "epoch": 0.8474236002676779,
+      "grad_norm": 0.16746586561203003,
+      "learning_rate": 1.644598655179693e-05,
+      "loss": 0.5015,
+      "step": 3799
+    },
+    {
+      "epoch": 0.8476466651795672,
+      "grad_norm": 0.16072294116020203,
+      "learning_rate": 1.6444187126920334e-05,
+      "loss": 0.4993,
+      "step": 3800
+    },
+    {
+      "epoch": 0.8478697300914566,
+      "grad_norm": 0.18298223614692688,
+      "learning_rate": 1.644238734512412e-05,
+      "loss": 0.4923,
+      "step": 3801
+    },
+    {
+      "epoch": 0.848092795003346,
+      "grad_norm": 0.1531648337841034,
+      "learning_rate": 1.6440587206507972e-05,
+      "loss": 0.4731,
+      "step": 3802
+    },
+    {
+      "epoch": 0.8483158599152353,
+      "grad_norm": 0.16385571658611298,
+      "learning_rate": 1.6438786711171588e-05,
+      "loss": 0.4718,
+      "step": 3803
+    },
+    {
+      "epoch": 0.8485389248271247,
+      "grad_norm": 0.15524934232234955,
+      "learning_rate": 1.6436985859214698e-05,
+      "loss": 0.4886,
+      "step": 3804
+    },
+    {
+      "epoch": 0.8487619897390141,
+      "grad_norm": 0.16199147701263428,
+      "learning_rate": 1.643518465073704e-05,
+      "loss": 0.5056,
+      "step": 3805
+    },
+    {
+      "epoch": 0.8489850546509035,
+      "grad_norm": 0.17070268094539642,
+      "learning_rate": 1.6433383085838378e-05,
+      "loss": 0.5157,
+      "step": 3806
+    },
+    {
+      "epoch": 0.8492081195627927,
+      "grad_norm": 0.19035401940345764,
+      "learning_rate": 1.643158116461849e-05,
+      "loss": 0.5135,
+      "step": 3807
+    },
+    {
+      "epoch": 0.8494311844746821,
+      "grad_norm": 0.16606462001800537,
+      "learning_rate": 1.6429778887177182e-05,
+      "loss": 0.5058,
+      "step": 3808
+    },
+    {
+      "epoch": 0.8496542493865715,
+      "grad_norm": 0.16338834166526794,
+      "learning_rate": 1.6427976253614275e-05,
+      "loss": 0.4914,
+      "step": 3809
+    },
+    {
+      "epoch": 0.8498773142984608,
+      "grad_norm": 0.16145442426204681,
+      "learning_rate": 1.6426173264029614e-05,
+      "loss": 0.4999,
+      "step": 3810
+    },
+    {
+      "epoch": 0.8501003792103502,
+      "grad_norm": 0.16404959559440613,
+      "learning_rate": 1.642436991852305e-05,
+      "loss": 0.5014,
+      "step": 3811
+    },
+    {
+      "epoch": 0.8503234441222396,
+      "grad_norm": 0.16108281910419464,
+      "learning_rate": 1.642256621719447e-05,
+      "loss": 0.4702,
+      "step": 3812
+    },
+    {
+      "epoch": 0.850546509034129,
+      "grad_norm": 0.15342526137828827,
+      "learning_rate": 1.642076216014377e-05,
+      "loss": 0.4686,
+      "step": 3813
+    },
+    {
+      "epoch": 0.8507695739460183,
+      "grad_norm": 0.16050684452056885,
+      "learning_rate": 1.6418957747470877e-05,
+      "loss": 0.4729,
+      "step": 3814
+    },
+    {
+      "epoch": 0.8509926388579077,
+      "grad_norm": 0.17533640563488007,
+      "learning_rate": 1.641715297927573e-05,
+      "loss": 0.4926,
+      "step": 3815
+    },
+    {
+      "epoch": 0.851215703769797,
+      "grad_norm": 0.16416087746620178,
+      "learning_rate": 1.641534785565828e-05,
+      "loss": 0.513,
+      "step": 3816
+    },
+    {
+      "epoch": 0.8514387686816863,
+      "grad_norm": 0.1623440384864807,
+      "learning_rate": 1.6413542376718513e-05,
+      "loss": 0.49,
+      "step": 3817
+    },
+    {
+      "epoch": 0.8516618335935757,
+      "grad_norm": 0.1671840250492096,
+      "learning_rate": 1.641173654255643e-05,
+      "loss": 0.5044,
+      "step": 3818
+    },
+    {
+      "epoch": 0.8518848985054651,
+      "grad_norm": 0.1557224541902542,
+      "learning_rate": 1.640993035327204e-05,
+      "loss": 0.4984,
+      "step": 3819
+    },
+    {
+      "epoch": 0.8521079634173544,
+      "grad_norm": 0.15791340172290802,
+      "learning_rate": 1.6408123808965392e-05,
+      "loss": 0.4801,
+      "step": 3820
+    },
+    {
+      "epoch": 0.8523310283292438,
+      "grad_norm": 0.1637008786201477,
+      "learning_rate": 1.6406316909736536e-05,
+      "loss": 0.4989,
+      "step": 3821
+    },
+    {
+      "epoch": 0.8525540932411332,
+      "grad_norm": 0.16228072345256805,
+      "learning_rate": 1.6404509655685555e-05,
+      "loss": 0.4568,
+      "step": 3822
+    },
+    {
+      "epoch": 0.8527771581530226,
+      "grad_norm": 0.16301074624061584,
+      "learning_rate": 1.640270204691254e-05,
+      "loss": 0.4741,
+      "step": 3823
+    },
+    {
+      "epoch": 0.8530002230649119,
+      "grad_norm": 0.15945035219192505,
+      "learning_rate": 1.6400894083517612e-05,
+      "loss": 0.4885,
+      "step": 3824
+    },
+    {
+      "epoch": 0.8532232879768012,
+      "grad_norm": 0.16819296777248383,
+      "learning_rate": 1.639908576560091e-05,
+      "loss": 0.5147,
+      "step": 3825
+    },
+    {
+      "epoch": 0.8534463528886906,
+      "grad_norm": 0.16270814836025238,
+      "learning_rate": 1.6397277093262583e-05,
+      "loss": 0.5147,
+      "step": 3826
+    },
+    {
+      "epoch": 0.8536694178005799,
+      "grad_norm": 0.16252903640270233,
+      "learning_rate": 1.6395468066602812e-05,
+      "loss": 0.5019,
+      "step": 3827
+    },
+    {
+      "epoch": 0.8538924827124693,
+      "grad_norm": 0.16933606564998627,
+      "learning_rate": 1.6393658685721787e-05,
+      "loss": 0.5085,
+      "step": 3828
+    },
+    {
+      "epoch": 0.8541155476243587,
+      "grad_norm": 0.17316611111164093,
+      "learning_rate": 1.639184895071973e-05,
+      "loss": 0.5141,
+      "step": 3829
+    },
+    {
+      "epoch": 0.8543386125362481,
+      "grad_norm": 0.17270614206790924,
+      "learning_rate": 1.6390038861696868e-05,
+      "loss": 0.4894,
+      "step": 3830
+    },
+    {
+      "epoch": 0.8545616774481374,
+      "grad_norm": 0.1740892380475998,
+      "learning_rate": 1.638822841875346e-05,
+      "loss": 0.5061,
+      "step": 3831
+    },
+    {
+      "epoch": 0.8547847423600268,
+      "grad_norm": 0.18361349403858185,
+      "learning_rate": 1.638641762198978e-05,
+      "loss": 0.4868,
+      "step": 3832
+    },
+    {
+      "epoch": 0.8550078072719162,
+      "grad_norm": 0.18127956986427307,
+      "learning_rate": 1.638460647150612e-05,
+      "loss": 0.5208,
+      "step": 3833
+    },
+    {
+      "epoch": 0.8552308721838054,
+      "grad_norm": 0.1610213816165924,
+      "learning_rate": 1.6382794967402792e-05,
+      "loss": 0.4749,
+      "step": 3834
+    },
+    {
+      "epoch": 0.8554539370956948,
+      "grad_norm": 0.16226479411125183,
+      "learning_rate": 1.638098310978013e-05,
+      "loss": 0.4872,
+      "step": 3835
+    },
+    {
+      "epoch": 0.8556770020075842,
+      "grad_norm": 0.15979908406734467,
+      "learning_rate": 1.6379170898738483e-05,
+      "loss": 0.5036,
+      "step": 3836
+    },
+    {
+      "epoch": 0.8559000669194735,
+      "grad_norm": 0.1549946367740631,
+      "learning_rate": 1.6377358334378228e-05,
+      "loss": 0.4858,
+      "step": 3837
+    },
+    {
+      "epoch": 0.8561231318313629,
+      "grad_norm": 0.16230742633342743,
+      "learning_rate": 1.6375545416799756e-05,
+      "loss": 0.5084,
+      "step": 3838
+    },
+    {
+      "epoch": 0.8563461967432523,
+      "grad_norm": 0.15997640788555145,
+      "learning_rate": 1.6373732146103466e-05,
+      "loss": 0.5162,
+      "step": 3839
+    },
+    {
+      "epoch": 0.8565692616551417,
+      "grad_norm": 0.1616683304309845,
+      "learning_rate": 1.6371918522389804e-05,
+      "loss": 0.471,
+      "step": 3840
+    },
+    {
+      "epoch": 0.856792326567031,
+      "grad_norm": 0.15914194285869598,
+      "learning_rate": 1.637010454575921e-05,
+      "loss": 0.5039,
+      "step": 3841
+    },
+    {
+      "epoch": 0.8570153914789204,
+      "grad_norm": 0.16176219284534454,
+      "learning_rate": 1.636829021631216e-05,
+      "loss": 0.4903,
+      "step": 3842
+    },
+    {
+      "epoch": 0.8572384563908098,
+      "grad_norm": 0.1591506451368332,
+      "learning_rate": 1.636647553414914e-05,
+      "loss": 0.4833,
+      "step": 3843
+    },
+    {
+      "epoch": 0.857461521302699,
+      "grad_norm": 0.16955386102199554,
+      "learning_rate": 1.6364660499370656e-05,
+      "loss": 0.5088,
+      "step": 3844
+    },
+    {
+      "epoch": 0.8576845862145884,
+      "grad_norm": 0.15826305747032166,
+      "learning_rate": 1.636284511207724e-05,
+      "loss": 0.4985,
+      "step": 3845
+    },
+    {
+      "epoch": 0.8579076511264778,
+      "grad_norm": 0.16054044663906097,
+      "learning_rate": 1.6361029372369433e-05,
+      "loss": 0.4991,
+      "step": 3846
+    },
+    {
+      "epoch": 0.8581307160383672,
+      "grad_norm": 0.16432225704193115,
+      "learning_rate": 1.6359213280347814e-05,
+      "loss": 0.4927,
+      "step": 3847
+    },
+    {
+      "epoch": 0.8583537809502565,
+      "grad_norm": 0.17022813856601715,
+      "learning_rate": 1.6357396836112957e-05,
+      "loss": 0.4863,
+      "step": 3848
+    },
+    {
+      "epoch": 0.8585768458621459,
+      "grad_norm": 0.15264475345611572,
+      "learning_rate": 1.6355580039765478e-05,
+      "loss": 0.4891,
+      "step": 3849
+    },
+    {
+      "epoch": 0.8587999107740353,
+      "grad_norm": 0.1790938675403595,
+      "learning_rate": 1.6353762891405993e-05,
+      "loss": 0.517,
+      "step": 3850
+    },
+    {
+      "epoch": 0.8590229756859246,
+      "grad_norm": 0.15989623963832855,
+      "learning_rate": 1.6351945391135154e-05,
+      "loss": 0.4684,
+      "step": 3851
+    },
+    {
+      "epoch": 0.859246040597814,
+      "grad_norm": 0.16801711916923523,
+      "learning_rate": 1.6350127539053626e-05,
+      "loss": 0.4782,
+      "step": 3852
+    },
+    {
+      "epoch": 0.8594691055097033,
+      "grad_norm": 0.15932013094425201,
+      "learning_rate": 1.634830933526209e-05,
+      "loss": 0.4805,
+      "step": 3853
+    },
+    {
+      "epoch": 0.8596921704215927,
+      "grad_norm": 0.16254423558712006,
+      "learning_rate": 1.6346490779861252e-05,
+      "loss": 0.5063,
+      "step": 3854
+    },
+    {
+      "epoch": 0.859915235333482,
+      "grad_norm": 0.1786786913871765,
+      "learning_rate": 1.634467187295183e-05,
+      "loss": 0.5457,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8601383002453714,
+      "grad_norm": 0.15343046188354492,
+      "learning_rate": 1.6342852614634575e-05,
+      "loss": 0.4681,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8603613651572608,
+      "grad_norm": 0.16010713577270508,
+      "learning_rate": 1.634103300501024e-05,
+      "loss": 0.5131,
+      "step": 3857
+    },
+    {
+      "epoch": 0.8605844300691501,
+      "grad_norm": 0.1694883406162262,
+      "learning_rate": 1.6339213044179612e-05,
+      "loss": 0.5117,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8608074949810395,
+      "grad_norm": 0.1592147946357727,
+      "learning_rate": 1.6337392732243488e-05,
+      "loss": 0.5026,
+      "step": 3859
+    },
+    {
+      "epoch": 0.8610305598929289,
+      "grad_norm": 0.15784434974193573,
+      "learning_rate": 1.6335572069302694e-05,
+      "loss": 0.4705,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8612536248048182,
+      "grad_norm": 0.15071116387844086,
+      "learning_rate": 1.6333751055458065e-05,
+      "loss": 0.4781,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8614766897167075,
+      "grad_norm": 0.16125838458538055,
+      "learning_rate": 1.6331929690810464e-05,
+      "loss": 0.5079,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8616997546285969,
+      "grad_norm": 0.16050268709659576,
+      "learning_rate": 1.6330107975460764e-05,
+      "loss": 0.4983,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8619228195404863,
+      "grad_norm": 0.15100105106830597,
+      "learning_rate": 1.632828590950987e-05,
+      "loss": 0.46,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8621458844523756,
+      "grad_norm": 0.17107254266738892,
+      "learning_rate": 1.632646349305869e-05,
+      "loss": 0.4708,
+      "step": 3865
+    },
+    {
+      "epoch": 0.862368949364265,
+      "grad_norm": 0.15665309131145477,
+      "learning_rate": 1.6324640726208172e-05,
+      "loss": 0.4786,
+      "step": 3866
+    },
+    {
+      "epoch": 0.8625920142761544,
+      "grad_norm": 0.1617649793624878,
+      "learning_rate": 1.6322817609059267e-05,
+      "loss": 0.5078,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8628150791880437,
+      "grad_norm": 0.15603747963905334,
+      "learning_rate": 1.6320994141712948e-05,
+      "loss": 0.4939,
+      "step": 3868
+    },
+    {
+      "epoch": 0.8630381440999331,
+      "grad_norm": 0.14941559731960297,
+      "learning_rate": 1.6319170324270212e-05,
+      "loss": 0.4637,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8632612090118225,
+      "grad_norm": 0.1626354604959488,
+      "learning_rate": 1.631734615683208e-05,
+      "loss": 0.4715,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8634842739237119,
+      "grad_norm": 0.15691153705120087,
+      "learning_rate": 1.6315521639499573e-05,
+      "loss": 0.5017,
+      "step": 3871
+    },
+    {
+      "epoch": 0.8637073388356011,
+      "grad_norm": 0.19187089800834656,
+      "learning_rate": 1.6313696772373754e-05,
+      "loss": 0.5369,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8639304037474905,
+      "grad_norm": 0.16823086142539978,
+      "learning_rate": 1.6311871555555696e-05,
+      "loss": 0.4892,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8641534686593799,
+      "grad_norm": 0.17205961048603058,
+      "learning_rate": 1.6310045989146486e-05,
+      "loss": 0.5472,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8643765335712692,
+      "grad_norm": 0.1610354632139206,
+      "learning_rate": 1.6308220073247237e-05,
+      "loss": 0.4887,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8645995984831586,
+      "grad_norm": 0.21832773089408875,
+      "learning_rate": 1.6306393807959078e-05,
+      "loss": 0.5081,
+      "step": 3876
+    },
+    {
+      "epoch": 0.864822663395048,
+      "grad_norm": 0.17104685306549072,
+      "learning_rate": 1.6304567193383164e-05,
+      "loss": 0.5179,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8650457283069373,
+      "grad_norm": 0.15571469068527222,
+      "learning_rate": 1.6302740229620662e-05,
+      "loss": 0.4773,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8652687932188267,
+      "grad_norm": 0.16199520230293274,
+      "learning_rate": 1.630091291677276e-05,
+      "loss": 0.4972,
+      "step": 3879
+    },
+    {
+      "epoch": 0.865491858130716,
+      "grad_norm": 0.17316444218158722,
+      "learning_rate": 1.6299085254940664e-05,
+      "loss": 0.4961,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8657149230426054,
+      "grad_norm": 0.16597145795822144,
+      "learning_rate": 1.6297257244225602e-05,
+      "loss": 0.4893,
+      "step": 3881
+    },
+    {
+      "epoch": 0.8659379879544947,
+      "grad_norm": 0.1571418046951294,
+      "learning_rate": 1.6295428884728827e-05,
+      "loss": 0.4957,
+      "step": 3882
+    },
+    {
+      "epoch": 0.8661610528663841,
+      "grad_norm": 0.15437713265419006,
+      "learning_rate": 1.62936001765516e-05,
+      "loss": 0.4912,
+      "step": 3883
+    },
+    {
+      "epoch": 0.8663841177782735,
+      "grad_norm": 0.15803179144859314,
+      "learning_rate": 1.6291771119795202e-05,
+      "loss": 0.5012,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8666071826901628,
+      "grad_norm": 0.1526332050561905,
+      "learning_rate": 1.628994171456095e-05,
+      "loss": 0.48,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8668302476020522,
+      "grad_norm": 0.16080142557621002,
+      "learning_rate": 1.628811196095016e-05,
+      "loss": 0.4885,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8670533125139416,
+      "grad_norm": 0.16212552785873413,
+      "learning_rate": 1.628628185906417e-05,
+      "loss": 0.5049,
+      "step": 3887
+    },
+    {
+      "epoch": 0.867276377425831,
+      "grad_norm": 0.1618081033229828,
+      "learning_rate": 1.6284451409004352e-05,
+      "loss": 0.4955,
+      "step": 3888
+    },
+    {
+      "epoch": 0.8674994423377203,
+      "grad_norm": 0.15760111808776855,
+      "learning_rate": 1.628262061087208e-05,
+      "loss": 0.4808,
+      "step": 3889
+    },
+    {
+      "epoch": 0.8677225072496096,
+      "grad_norm": 0.15690281987190247,
+      "learning_rate": 1.6280789464768765e-05,
+      "loss": 0.5198,
+      "step": 3890
+    },
+    {
+      "epoch": 0.867945572161499,
+      "grad_norm": 0.17230992019176483,
+      "learning_rate": 1.6278957970795818e-05,
+      "loss": 0.4762,
+      "step": 3891
+    },
+    {
+      "epoch": 0.8681686370733883,
+      "grad_norm": 0.15698418021202087,
+      "learning_rate": 1.6277126129054687e-05,
+      "loss": 0.4969,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8683917019852777,
+      "grad_norm": 0.15693382918834686,
+      "learning_rate": 1.6275293939646822e-05,
+      "loss": 0.4856,
+      "step": 3893
+    },
+    {
+      "epoch": 0.8686147668971671,
+      "grad_norm": 0.1605808138847351,
+      "learning_rate": 1.6273461402673706e-05,
+      "loss": 0.4739,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8688378318090564,
+      "grad_norm": 0.2007637321949005,
+      "learning_rate": 1.6271628518236836e-05,
+      "loss": 0.5115,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8690608967209458,
+      "grad_norm": 0.17741620540618896,
+      "learning_rate": 1.6269795286437728e-05,
+      "loss": 0.5034,
+      "step": 3896
+    },
+    {
+      "epoch": 0.8692839616328352,
+      "grad_norm": 0.1685972362756729,
+      "learning_rate": 1.6267961707377923e-05,
+      "loss": 0.5316,
+      "step": 3897
+    },
+    {
+      "epoch": 0.8695070265447246,
+      "grad_norm": 0.16251643002033234,
+      "learning_rate": 1.6266127781158965e-05,
+      "loss": 0.5018,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8697300914566138,
+      "grad_norm": 0.15942230820655823,
+      "learning_rate": 1.626429350788244e-05,
+      "loss": 0.4799,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8699531563685032,
+      "grad_norm": 0.16205990314483643,
+      "learning_rate": 1.6262458887649933e-05,
+      "loss": 0.4721,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8701762212803926,
+      "grad_norm": 0.16467221081256866,
+      "learning_rate": 1.6260623920563062e-05,
+      "loss": 0.5137,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8703992861922819,
+      "grad_norm": 0.16359251737594604,
+      "learning_rate": 1.6258788606723457e-05,
+      "loss": 0.5063,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8706223511041713,
+      "grad_norm": 0.16454778611660004,
+      "learning_rate": 1.625695294623277e-05,
+      "loss": 0.4847,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8708454160160607,
+      "grad_norm": 0.1625732183456421,
+      "learning_rate": 1.625511693919267e-05,
+      "loss": 0.5324,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8710684809279501,
+      "grad_norm": 0.16343548893928528,
+      "learning_rate": 1.625328058570485e-05,
+      "loss": 0.5012,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8712915458398394,
+      "grad_norm": 0.15868264436721802,
+      "learning_rate": 1.6251443885871013e-05,
+      "loss": 0.4921,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8715146107517288,
+      "grad_norm": 0.1692667156457901,
+      "learning_rate": 1.6249606839792897e-05,
+      "loss": 0.4861,
+      "step": 3907
+    },
+    {
+      "epoch": 0.8717376756636182,
+      "grad_norm": 0.15825539827346802,
+      "learning_rate": 1.6247769447572235e-05,
+      "loss": 0.4828,
+      "step": 3908
+    },
+    {
+      "epoch": 0.8719607405755074,
+      "grad_norm": 0.1585260033607483,
+      "learning_rate": 1.6245931709310806e-05,
+      "loss": 0.4885,
+      "step": 3909
+    },
+    {
+      "epoch": 0.8721838054873968,
+      "grad_norm": 0.1593310832977295,
+      "learning_rate": 1.624409362511039e-05,
+      "loss": 0.5158,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8724068703992862,
+      "grad_norm": 0.15775549411773682,
+      "learning_rate": 1.624225519507279e-05,
+      "loss": 0.4812,
+      "step": 3911
+    },
+    {
+      "epoch": 0.8726299353111755,
+      "grad_norm": 0.16018901765346527,
+      "learning_rate": 1.624041641929983e-05,
+      "loss": 0.5123,
+      "step": 3912
+    },
+    {
+      "epoch": 0.8728530002230649,
+      "grad_norm": 0.16512146592140198,
+      "learning_rate": 1.6238577297893357e-05,
+      "loss": 0.4886,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8730760651349543,
+      "grad_norm": 0.16052314639091492,
+      "learning_rate": 1.6236737830955233e-05,
+      "loss": 0.5078,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8732991300468437,
+      "grad_norm": 0.16342321038246155,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.5151,
+      "step": 3915
+    },
+    {
+      "epoch": 0.873522194958733,
+      "grad_norm": 0.16125398874282837,
+      "learning_rate": 1.6233057860891566e-05,
+      "loss": 0.4786,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8737452598706223,
+      "grad_norm": 0.17200647294521332,
+      "learning_rate": 1.623121735796985e-05,
+      "loss": 0.5013,
+      "step": 3917
+    },
+    {
+      "epoch": 0.8739683247825117,
+      "grad_norm": 0.17798694968223572,
+      "learning_rate": 1.6229376509924116e-05,
+      "loss": 0.5257,
+      "step": 3918
+    },
+    {
+      "epoch": 0.874191389694401,
+      "grad_norm": 0.15707524120807648,
+      "learning_rate": 1.6227535316856326e-05,
+      "loss": 0.4889,
+      "step": 3919
+    },
+    {
+      "epoch": 0.8744144546062904,
+      "grad_norm": 0.16776537895202637,
+      "learning_rate": 1.622569377886846e-05,
+      "loss": 0.5196,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8746375195181798,
+      "grad_norm": 0.16779324412345886,
+      "learning_rate": 1.622385189606251e-05,
+      "loss": 0.512,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8748605844300692,
+      "grad_norm": 0.16673165559768677,
+      "learning_rate": 1.622200966854049e-05,
+      "loss": 0.4888,
+      "step": 3922
+    },
+    {
+      "epoch": 0.8750836493419585,
+      "grad_norm": 0.18797388672828674,
+      "learning_rate": 1.622016709640444e-05,
+      "loss": 0.4789,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8753067142538479,
+      "grad_norm": 0.21027544140815735,
+      "learning_rate": 1.621832417975641e-05,
+      "loss": 0.4818,
+      "step": 3924
+    },
+    {
+      "epoch": 0.8755297791657373,
+      "grad_norm": 0.16859135031700134,
+      "learning_rate": 1.621648091869847e-05,
+      "loss": 0.4949,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8757528440776265,
+      "grad_norm": 0.1632431447505951,
+      "learning_rate": 1.6214637313332714e-05,
+      "loss": 0.4869,
+      "step": 3926
+    },
+    {
+      "epoch": 0.8759759089895159,
+      "grad_norm": 0.1545240879058838,
+      "learning_rate": 1.6212793363761253e-05,
+      "loss": 0.495,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8761989739014053,
+      "grad_norm": 0.162541463971138,
+      "learning_rate": 1.621094907008621e-05,
+      "loss": 0.507,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8764220388132947,
+      "grad_norm": 0.16767984628677368,
+      "learning_rate": 1.6209104432409745e-05,
+      "loss": 0.5098,
+      "step": 3929
+    },
+    {
+      "epoch": 0.876645103725184,
+      "grad_norm": 0.17964600026607513,
+      "learning_rate": 1.6207259450834022e-05,
+      "loss": 0.5055,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8768681686370734,
+      "grad_norm": 0.16593998670578003,
+      "learning_rate": 1.620541412546122e-05,
+      "loss": 0.4944,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8770912335489628,
+      "grad_norm": 0.15954278409481049,
+      "learning_rate": 1.6203568456393554e-05,
+      "loss": 0.5158,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8773142984608521,
+      "grad_norm": 0.1770762950181961,
+      "learning_rate": 1.620172244373324e-05,
+      "loss": 0.4886,
+      "step": 3933
+    },
+    {
+      "epoch": 0.8775373633727415,
+      "grad_norm": 0.16538506746292114,
+      "learning_rate": 1.619987608758253e-05,
+      "loss": 0.4563,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8777604282846309,
+      "grad_norm": 0.1733943372964859,
+      "learning_rate": 1.6198029388043685e-05,
+      "loss": 0.5131,
+      "step": 3935
+    },
+    {
+      "epoch": 0.8779834931965201,
+      "grad_norm": 0.15919508039951324,
+      "learning_rate": 1.619618234521898e-05,
+      "loss": 0.4904,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8782065581084095,
+      "grad_norm": 0.1546212136745453,
+      "learning_rate": 1.6194334959210726e-05,
+      "loss": 0.4827,
+      "step": 3937
+    },
+    {
+      "epoch": 0.8784296230202989,
+      "grad_norm": 0.16514717042446136,
+      "learning_rate": 1.6192487230121236e-05,
+      "loss": 0.501,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8786526879321883,
+      "grad_norm": 0.1485379934310913,
+      "learning_rate": 1.6190639158052852e-05,
+      "loss": 0.47,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8788757528440776,
+      "grad_norm": 0.1519124060869217,
+      "learning_rate": 1.618879074310793e-05,
+      "loss": 0.484,
+      "step": 3940
+    },
+    {
+      "epoch": 0.879098817755967,
+      "grad_norm": 0.16282908618450165,
+      "learning_rate": 1.618694198538885e-05,
+      "loss": 0.5071,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8793218826678564,
+      "grad_norm": 0.16473978757858276,
+      "learning_rate": 1.6185092884998e-05,
+      "loss": 0.5203,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8795449475797457,
+      "grad_norm": 0.15753906965255737,
+      "learning_rate": 1.6183243442037807e-05,
+      "loss": 0.4989,
+      "step": 3943
+    },
+    {
+      "epoch": 0.879768012491635,
+      "grad_norm": 0.1654537469148636,
+      "learning_rate": 1.6181393656610693e-05,
+      "loss": 0.4749,
+      "step": 3944
+    },
+    {
+      "epoch": 0.8799910774035244,
+      "grad_norm": 0.17788641154766083,
+      "learning_rate": 1.6179543528819116e-05,
+      "loss": 0.4961,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8802141423154138,
+      "grad_norm": 0.16031092405319214,
+      "learning_rate": 1.617769305876555e-05,
+      "loss": 0.5006,
+      "step": 3946
+    },
+    {
+      "epoch": 0.8804372072273031,
+      "grad_norm": 0.16216473281383514,
+      "learning_rate": 1.6175842246552484e-05,
+      "loss": 0.5012,
+      "step": 3947
+    },
+    {
+      "epoch": 0.8806602721391925,
+      "grad_norm": 0.3736167848110199,
+      "learning_rate": 1.6173991092282424e-05,
+      "loss": 0.481,
+      "step": 3948
+    },
+    {
+      "epoch": 0.8808833370510819,
+      "grad_norm": 0.16511203348636627,
+      "learning_rate": 1.6172139596057902e-05,
+      "loss": 0.4837,
+      "step": 3949
+    },
+    {
+      "epoch": 0.8811064019629712,
+      "grad_norm": 0.16273178160190582,
+      "learning_rate": 1.6170287757981468e-05,
+      "loss": 0.494,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8813294668748606,
+      "grad_norm": 0.17184849083423615,
+      "learning_rate": 1.616843557815568e-05,
+      "loss": 0.5257,
+      "step": 3951
+    },
+    {
+      "epoch": 0.88155253178675,
+      "grad_norm": 0.16385000944137573,
+      "learning_rate": 1.6166583056683132e-05,
+      "loss": 0.5086,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8817755966986393,
+      "grad_norm": 0.16628234088420868,
+      "learning_rate": 1.6164730193666423e-05,
+      "loss": 0.5035,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8819986616105286,
+      "grad_norm": 0.16098076105117798,
+      "learning_rate": 1.616287698920818e-05,
+      "loss": 0.4979,
+      "step": 3954
+    },
+    {
+      "epoch": 0.882221726522418,
+      "grad_norm": 0.16197021305561066,
+      "learning_rate": 1.6161023443411044e-05,
+      "loss": 0.4883,
+      "step": 3955
+    },
+    {
+      "epoch": 0.8824447914343074,
+      "grad_norm": 0.16163140535354614,
+      "learning_rate": 1.6159169556377672e-05,
+      "loss": 0.4899,
+      "step": 3956
+    },
+    {
+      "epoch": 0.8826678563461967,
+      "grad_norm": 0.16022367775440216,
+      "learning_rate": 1.615731532821075e-05,
+      "loss": 0.4832,
+      "step": 3957
+    },
+    {
+      "epoch": 0.8828909212580861,
+      "grad_norm": 0.16164539754390717,
+      "learning_rate": 1.615546075901297e-05,
+      "loss": 0.5016,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8831139861699755,
+      "grad_norm": 0.17649579048156738,
+      "learning_rate": 1.615360584888706e-05,
+      "loss": 0.4918,
+      "step": 3959
+    },
+    {
+      "epoch": 0.8833370510818648,
+      "grad_norm": 0.16985541582107544,
+      "learning_rate": 1.6151750597935746e-05,
+      "loss": 0.4947,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8835601159937542,
+      "grad_norm": 0.1681845337152481,
+      "learning_rate": 1.6149895006261788e-05,
+      "loss": 0.5169,
+      "step": 3961
+    },
+    {
+      "epoch": 0.8837831809056436,
+      "grad_norm": 0.1974424421787262,
+      "learning_rate": 1.6148039073967964e-05,
+      "loss": 0.5077,
+      "step": 3962
+    },
+    {
+      "epoch": 0.884006245817533,
+      "grad_norm": 0.1644642949104309,
+      "learning_rate": 1.614618280115706e-05,
+      "loss": 0.4981,
+      "step": 3963
+    },
+    {
+      "epoch": 0.8842293107294222,
+      "grad_norm": 0.16741390526294708,
+      "learning_rate": 1.6144326187931893e-05,
+      "loss": 0.4821,
+      "step": 3964
+    },
+    {
+      "epoch": 0.8844523756413116,
+      "grad_norm": 0.17722152173519135,
+      "learning_rate": 1.614246923439529e-05,
+      "loss": 0.4782,
+      "step": 3965
+    },
+    {
+      "epoch": 0.884675440553201,
+      "grad_norm": 0.17287255823612213,
+      "learning_rate": 1.6140611940650104e-05,
+      "loss": 0.4996,
+      "step": 3966
+    },
+    {
+      "epoch": 0.8848985054650903,
+      "grad_norm": 0.17982237040996552,
+      "learning_rate": 1.6138754306799206e-05,
+      "loss": 0.4567,
+      "step": 3967
+    },
+    {
+      "epoch": 0.8851215703769797,
+      "grad_norm": 0.1712944060564041,
+      "learning_rate": 1.6136896332945474e-05,
+      "loss": 0.4789,
+      "step": 3968
+    },
+    {
+      "epoch": 0.8853446352888691,
+      "grad_norm": 0.16616348922252655,
+      "learning_rate": 1.6135038019191823e-05,
+      "loss": 0.509,
+      "step": 3969
+    },
+    {
+      "epoch": 0.8855677002007584,
+      "grad_norm": 0.16956603527069092,
+      "learning_rate": 1.6133179365641178e-05,
+      "loss": 0.4906,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8857907651126478,
+      "grad_norm": 0.15731406211853027,
+      "learning_rate": 1.613132037239648e-05,
+      "loss": 0.4911,
+      "step": 3971
+    },
+    {
+      "epoch": 0.8860138300245372,
+      "grad_norm": 0.16865961253643036,
+      "learning_rate": 1.6129461039560693e-05,
+      "loss": 0.5072,
+      "step": 3972
+    },
+    {
+      "epoch": 0.8862368949364265,
+      "grad_norm": 0.16574735939502716,
+      "learning_rate": 1.6127601367236793e-05,
+      "loss": 0.4807,
+      "step": 3973
+    },
+    {
+      "epoch": 0.8864599598483158,
+      "grad_norm": 0.16178300976753235,
+      "learning_rate": 1.6125741355527788e-05,
+      "loss": 0.5138,
+      "step": 3974
+    },
+    {
+      "epoch": 0.8866830247602052,
+      "grad_norm": 0.163814976811409,
+      "learning_rate": 1.6123881004536696e-05,
+      "loss": 0.4999,
+      "step": 3975
+    },
+    {
+      "epoch": 0.8869060896720946,
+      "grad_norm": 0.1648682951927185,
+      "learning_rate": 1.612202031436655e-05,
+      "loss": 0.4621,
+      "step": 3976
+    },
+    {
+      "epoch": 0.8871291545839839,
+      "grad_norm": 0.1537158489227295,
+      "learning_rate": 1.6120159285120417e-05,
+      "loss": 0.4678,
+      "step": 3977
+    },
+    {
+      "epoch": 0.8873522194958733,
+      "grad_norm": 0.17030583322048187,
+      "learning_rate": 1.6118297916901357e-05,
+      "loss": 0.5142,
+      "step": 3978
+    },
+    {
+      "epoch": 0.8875752844077627,
+      "grad_norm": 0.15777969360351562,
+      "learning_rate": 1.6116436209812476e-05,
+      "loss": 0.4857,
+      "step": 3979
+    },
+    {
+      "epoch": 0.8877983493196521,
+      "grad_norm": 0.16635335981845856,
+      "learning_rate": 1.6114574163956883e-05,
+      "loss": 0.4945,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8880214142315414,
+      "grad_norm": 0.16205914318561554,
+      "learning_rate": 1.611271177943771e-05,
+      "loss": 0.5013,
+      "step": 3981
+    },
+    {
+      "epoch": 0.8882444791434307,
+      "grad_norm": 0.16518031060695648,
+      "learning_rate": 1.6110849056358112e-05,
+      "loss": 0.492,
+      "step": 3982
+    },
+    {
+      "epoch": 0.8884675440553201,
+      "grad_norm": 0.17391857504844666,
+      "learning_rate": 1.610898599482125e-05,
+      "loss": 0.5147,
+      "step": 3983
+    },
+    {
+      "epoch": 0.8886906089672094,
+      "grad_norm": 0.16679394245147705,
+      "learning_rate": 1.610712259493032e-05,
+      "loss": 0.5079,
+      "step": 3984
+    },
+    {
+      "epoch": 0.8889136738790988,
+      "grad_norm": 0.15235210955142975,
+      "learning_rate": 1.6105258856788525e-05,
+      "loss": 0.4616,
+      "step": 3985
+    },
+    {
+      "epoch": 0.8891367387909882,
+      "grad_norm": 0.15558315813541412,
+      "learning_rate": 1.6103394780499088e-05,
+      "loss": 0.5129,
+      "step": 3986
+    },
+    {
+      "epoch": 0.8893598037028775,
+      "grad_norm": 0.16309772431850433,
+      "learning_rate": 1.610153036616526e-05,
+      "loss": 0.4981,
+      "step": 3987
+    },
+    {
+      "epoch": 0.8895828686147669,
+      "grad_norm": 0.1611076146364212,
+      "learning_rate": 1.60996656138903e-05,
+      "loss": 0.4858,
+      "step": 3988
+    },
+    {
+      "epoch": 0.8898059335266563,
+      "grad_norm": 0.16238349676132202,
+      "learning_rate": 1.6097800523777487e-05,
+      "loss": 0.5327,
+      "step": 3989
+    },
+    {
+      "epoch": 0.8900289984385457,
+      "grad_norm": 0.1624278724193573,
+      "learning_rate": 1.6095935095930125e-05,
+      "loss": 0.4988,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8902520633504349,
+      "grad_norm": 0.16069476306438446,
+      "learning_rate": 1.609406933045153e-05,
+      "loss": 0.5061,
+      "step": 3991
+    },
+    {
+      "epoch": 0.8904751282623243,
+      "grad_norm": 0.17323002219200134,
+      "learning_rate": 1.6092203227445046e-05,
+      "loss": 0.5126,
+      "step": 3992
+    },
+    {
+      "epoch": 0.8906981931742137,
+      "grad_norm": 0.16106821596622467,
+      "learning_rate": 1.6090336787014028e-05,
+      "loss": 0.5075,
+      "step": 3993
+    },
+    {
+      "epoch": 0.890921258086103,
+      "grad_norm": 0.20618608593940735,
+      "learning_rate": 1.6088470009261846e-05,
+      "loss": 0.4934,
+      "step": 3994
+    },
+    {
+      "epoch": 0.8911443229979924,
+      "grad_norm": 0.15415968000888824,
+      "learning_rate": 1.6086602894291895e-05,
+      "loss": 0.4685,
+      "step": 3995
+    },
+    {
+      "epoch": 0.8913673879098818,
+      "grad_norm": 0.16283094882965088,
+      "learning_rate": 1.608473544220759e-05,
+      "loss": 0.4893,
+      "step": 3996
+    },
+    {
+      "epoch": 0.8915904528217712,
+      "grad_norm": 0.15671685338020325,
+      "learning_rate": 1.6082867653112365e-05,
+      "loss": 0.4866,
+      "step": 3997
+    },
+    {
+      "epoch": 0.8918135177336605,
+      "grad_norm": 0.14689774811267853,
+      "learning_rate": 1.6080999527109665e-05,
+      "loss": 0.4596,
+      "step": 3998
+    },
+    {
+      "epoch": 0.8920365826455499,
+      "grad_norm": 0.1579175591468811,
+      "learning_rate": 1.6079131064302958e-05,
+      "loss": 0.5135,
+      "step": 3999
+    },
+    {
+      "epoch": 0.8922596475574393,
+      "grad_norm": 0.19000330567359924,
+      "learning_rate": 1.6077262264795735e-05,
+      "loss": 0.4955,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8924827124693285,
+      "grad_norm": 0.17095667123794556,
+      "learning_rate": 1.6075393128691497e-05,
+      "loss": 0.492,
+      "step": 4001
+    },
+    {
+      "epoch": 0.8927057773812179,
+      "grad_norm": 0.1627168506383896,
+      "learning_rate": 1.6073523656093778e-05,
+      "loss": 0.4954,
+      "step": 4002
+    },
+    {
+      "epoch": 0.8929288422931073,
+      "grad_norm": 0.16945117712020874,
+      "learning_rate": 1.6071653847106113e-05,
+      "loss": 0.5073,
+      "step": 4003
+    },
+    {
+      "epoch": 0.8931519072049967,
+      "grad_norm": 0.16621045768260956,
+      "learning_rate": 1.6069783701832066e-05,
+      "loss": 0.509,
+      "step": 4004
+    },
+    {
+      "epoch": 0.893374972116886,
+      "grad_norm": 0.16034524142742157,
+      "learning_rate": 1.6067913220375216e-05,
+      "loss": 0.4839,
+      "step": 4005
+    },
+    {
+      "epoch": 0.8935980370287754,
+      "grad_norm": 0.15669786930084229,
+      "learning_rate": 1.6066042402839163e-05,
+      "loss": 0.4953,
+      "step": 4006
+    },
+    {
+      "epoch": 0.8938211019406648,
+      "grad_norm": 0.15480557084083557,
+      "learning_rate": 1.606417124932752e-05,
+      "loss": 0.459,
+      "step": 4007
+    },
+    {
+      "epoch": 0.8940441668525541,
+      "grad_norm": 0.15904684364795685,
+      "learning_rate": 1.6062299759943938e-05,
+      "loss": 0.4977,
+      "step": 4008
+    },
+    {
+      "epoch": 0.8942672317644434,
+      "grad_norm": 0.16506803035736084,
+      "learning_rate": 1.6060427934792056e-05,
+      "loss": 0.4792,
+      "step": 4009
+    },
+    {
+      "epoch": 0.8944902966763328,
+      "grad_norm": 0.16208292543888092,
+      "learning_rate": 1.6058555773975552e-05,
+      "loss": 0.5113,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8947133615882221,
+      "grad_norm": 0.15986153483390808,
+      "learning_rate": 1.6056683277598123e-05,
+      "loss": 0.4839,
+      "step": 4011
+    },
+    {
+      "epoch": 0.8949364265001115,
+      "grad_norm": 0.17647914588451385,
+      "learning_rate": 1.6054810445763474e-05,
+      "loss": 0.4845,
+      "step": 4012
+    },
+    {
+      "epoch": 0.8951594914120009,
+      "grad_norm": 0.16568545997142792,
+      "learning_rate": 1.6052937278575338e-05,
+      "loss": 0.5114,
+      "step": 4013
+    },
+    {
+      "epoch": 0.8953825563238903,
+      "grad_norm": 0.16345763206481934,
+      "learning_rate": 1.605106377613746e-05,
+      "loss": 0.5004,
+      "step": 4014
+    },
+    {
+      "epoch": 0.8956056212357796,
+      "grad_norm": 0.15674430131912231,
+      "learning_rate": 1.6049189938553606e-05,
+      "loss": 0.4745,
+      "step": 4015
+    },
+    {
+      "epoch": 0.895828686147669,
+      "grad_norm": 0.25220704078674316,
+      "learning_rate": 1.6047315765927566e-05,
+      "loss": 0.4988,
+      "step": 4016
+    },
+    {
+      "epoch": 0.8960517510595584,
+      "grad_norm": 0.16052506864070892,
+      "learning_rate": 1.6045441258363138e-05,
+      "loss": 0.468,
+      "step": 4017
+    },
+    {
+      "epoch": 0.8962748159714476,
+      "grad_norm": 0.1760246306657791,
+      "learning_rate": 1.6043566415964145e-05,
+      "loss": 0.5213,
+      "step": 4018
+    },
+    {
+      "epoch": 0.896497880883337,
+      "grad_norm": 0.16761453449726105,
+      "learning_rate": 1.6041691238834426e-05,
+      "loss": 0.5223,
+      "step": 4019
+    },
+    {
+      "epoch": 0.8967209457952264,
+      "grad_norm": 0.1628294140100479,
+      "learning_rate": 1.6039815727077845e-05,
+      "loss": 0.4642,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8969440107071158,
+      "grad_norm": 0.16300807893276215,
+      "learning_rate": 1.6037939880798277e-05,
+      "loss": 0.4907,
+      "step": 4021
+    },
+    {
+      "epoch": 0.8971670756190051,
+      "grad_norm": 0.16601742804050446,
+      "learning_rate": 1.603606370009962e-05,
+      "loss": 0.465,
+      "step": 4022
+    },
+    {
+      "epoch": 0.8973901405308945,
+      "grad_norm": 0.16860070824623108,
+      "learning_rate": 1.6034187185085783e-05,
+      "loss": 0.4817,
+      "step": 4023
+    },
+    {
+      "epoch": 0.8976132054427839,
+      "grad_norm": 0.16450795531272888,
+      "learning_rate": 1.6032310335860706e-05,
+      "loss": 0.4991,
+      "step": 4024
+    },
+    {
+      "epoch": 0.8978362703546732,
+      "grad_norm": 0.18083292245864868,
+      "learning_rate": 1.603043315252834e-05,
+      "loss": 0.49,
+      "step": 4025
+    },
+    {
+      "epoch": 0.8980593352665626,
+      "grad_norm": 0.16025912761688232,
+      "learning_rate": 1.6028555635192648e-05,
+      "loss": 0.492,
+      "step": 4026
+    },
+    {
+      "epoch": 0.898282400178452,
+      "grad_norm": 0.17619994282722473,
+      "learning_rate": 1.6026677783957626e-05,
+      "loss": 0.5244,
+      "step": 4027
+    },
+    {
+      "epoch": 0.8985054650903412,
+      "grad_norm": 0.16059327125549316,
+      "learning_rate": 1.602479959892728e-05,
+      "loss": 0.4875,
+      "step": 4028
+    },
+    {
+      "epoch": 0.8987285300022306,
+      "grad_norm": 0.166759192943573,
+      "learning_rate": 1.6022921080205634e-05,
+      "loss": 0.4953,
+      "step": 4029
+    },
+    {
+      "epoch": 0.89895159491412,
+      "grad_norm": 0.17170590162277222,
+      "learning_rate": 1.602104222789673e-05,
+      "loss": 0.5209,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8991746598260094,
+      "grad_norm": 0.17768704891204834,
+      "learning_rate": 1.601916304210464e-05,
+      "loss": 0.4861,
+      "step": 4031
+    },
+    {
+      "epoch": 0.8993977247378987,
+      "grad_norm": 0.1798562854528427,
+      "learning_rate": 1.6017283522933432e-05,
+      "loss": 0.5049,
+      "step": 4032
+    },
+    {
+      "epoch": 0.8996207896497881,
+      "grad_norm": 0.15688154101371765,
+      "learning_rate": 1.6015403670487216e-05,
+      "loss": 0.4527,
+      "step": 4033
+    },
+    {
+      "epoch": 0.8998438545616775,
+      "grad_norm": 0.1658952832221985,
+      "learning_rate": 1.6013523484870107e-05,
+      "loss": 0.4687,
+      "step": 4034
+    },
+    {
+      "epoch": 0.9000669194735668,
+      "grad_norm": 0.20666684210300446,
+      "learning_rate": 1.6011642966186237e-05,
+      "loss": 0.4883,
+      "step": 4035
+    },
+    {
+      "epoch": 0.9002899843854562,
+      "grad_norm": 0.1706281453371048,
+      "learning_rate": 1.600976211453977e-05,
+      "loss": 0.4989,
+      "step": 4036
+    },
+    {
+      "epoch": 0.9005130492973455,
+      "grad_norm": 0.16239850223064423,
+      "learning_rate": 1.600788093003487e-05,
+      "loss": 0.4944,
+      "step": 4037
+    },
+    {
+      "epoch": 0.9007361142092349,
+      "grad_norm": 0.17332103848457336,
+      "learning_rate": 1.6005999412775736e-05,
+      "loss": 0.5225,
+      "step": 4038
+    },
+    {
+      "epoch": 0.9009591791211242,
+      "grad_norm": 0.16437414288520813,
+      "learning_rate": 1.600411756286657e-05,
+      "loss": 0.5197,
+      "step": 4039
+    },
+    {
+      "epoch": 0.9011822440330136,
+      "grad_norm": 0.1621236801147461,
+      "learning_rate": 1.6002235380411614e-05,
+      "loss": 0.4943,
+      "step": 4040
+    },
+    {
+      "epoch": 0.901405308944903,
+      "grad_norm": 0.17484545707702637,
+      "learning_rate": 1.60003528655151e-05,
+      "loss": 0.5032,
+      "step": 4041
+    },
+    {
+      "epoch": 0.9016283738567923,
+      "grad_norm": 0.16205301880836487,
+      "learning_rate": 1.5998470018281303e-05,
+      "loss": 0.4848,
+      "step": 4042
+    },
+    {
+      "epoch": 0.9018514387686817,
+      "grad_norm": 0.15417253971099854,
+      "learning_rate": 1.5996586838814505e-05,
+      "loss": 0.4656,
+      "step": 4043
+    },
+    {
+      "epoch": 0.9020745036805711,
+      "grad_norm": 0.15824946761131287,
+      "learning_rate": 1.5994703327219008e-05,
+      "loss": 0.4926,
+      "step": 4044
+    },
+    {
+      "epoch": 0.9022975685924604,
+      "grad_norm": 0.17363391816616058,
+      "learning_rate": 1.5992819483599132e-05,
+      "loss": 0.5254,
+      "step": 4045
+    },
+    {
+      "epoch": 0.9025206335043497,
+      "grad_norm": 0.16660279035568237,
+      "learning_rate": 1.599093530805922e-05,
+      "loss": 0.5396,
+      "step": 4046
+    },
+    {
+      "epoch": 0.9027436984162391,
+      "grad_norm": 0.16047632694244385,
+      "learning_rate": 1.5989050800703622e-05,
+      "loss": 0.4782,
+      "step": 4047
+    },
+    {
+      "epoch": 0.9029667633281285,
+      "grad_norm": 0.1667635589838028,
+      "learning_rate": 1.5987165961636718e-05,
+      "loss": 0.5084,
+      "step": 4048
+    },
+    {
+      "epoch": 0.9031898282400178,
+      "grad_norm": 0.16391973197460175,
+      "learning_rate": 1.5985280790962903e-05,
+      "loss": 0.4967,
+      "step": 4049
+    },
+    {
+      "epoch": 0.9034128931519072,
+      "grad_norm": 0.15799559652805328,
+      "learning_rate": 1.598339528878659e-05,
+      "loss": 0.4929,
+      "step": 4050
+    },
+    {
+      "epoch": 0.9036359580637966,
+      "grad_norm": 0.15577222406864166,
+      "learning_rate": 1.5981509455212207e-05,
+      "loss": 0.4766,
+      "step": 4051
+    },
+    {
+      "epoch": 0.9038590229756859,
+      "grad_norm": 0.15642912685871124,
+      "learning_rate": 1.5979623290344207e-05,
+      "loss": 0.4714,
+      "step": 4052
+    },
+    {
+      "epoch": 0.9040820878875753,
+      "grad_norm": 0.1563492864370346,
+      "learning_rate": 1.5977736794287057e-05,
+      "loss": 0.4891,
+      "step": 4053
+    },
+    {
+      "epoch": 0.9043051527994647,
+      "grad_norm": 0.16570942103862762,
+      "learning_rate": 1.597584996714524e-05,
+      "loss": 0.4857,
+      "step": 4054
+    },
+    {
+      "epoch": 0.904528217711354,
+      "grad_norm": 0.16158944368362427,
+      "learning_rate": 1.5973962809023258e-05,
+      "loss": 0.487,
+      "step": 4055
+    },
+    {
+      "epoch": 0.9047512826232433,
+      "grad_norm": 0.16446934640407562,
+      "learning_rate": 1.5972075320025643e-05,
+      "loss": 0.5012,
+      "step": 4056
+    },
+    {
+      "epoch": 0.9049743475351327,
+      "grad_norm": 0.20039531588554382,
+      "learning_rate": 1.597018750025693e-05,
+      "loss": 0.5092,
+      "step": 4057
+    },
+    {
+      "epoch": 0.9051974124470221,
+      "grad_norm": 0.16635029017925262,
+      "learning_rate": 1.5968299349821678e-05,
+      "loss": 0.5425,
+      "step": 4058
+    },
+    {
+      "epoch": 0.9054204773589114,
+      "grad_norm": 0.162612184882164,
+      "learning_rate": 1.596641086882447e-05,
+      "loss": 0.4561,
+      "step": 4059
+    },
+    {
+      "epoch": 0.9056435422708008,
+      "grad_norm": 0.15874888002872467,
+      "learning_rate": 1.5964522057369897e-05,
+      "loss": 0.503,
+      "step": 4060
+    },
+    {
+      "epoch": 0.9058666071826902,
+      "grad_norm": 0.16984137892723083,
+      "learning_rate": 1.596263291556257e-05,
+      "loss": 0.4931,
+      "step": 4061
+    },
+    {
+      "epoch": 0.9060896720945795,
+      "grad_norm": 0.15440818667411804,
+      "learning_rate": 1.5960743443507128e-05,
+      "loss": 0.4834,
+      "step": 4062
+    },
+    {
+      "epoch": 0.9063127370064689,
+      "grad_norm": 0.17957209050655365,
+      "learning_rate": 1.595885364130822e-05,
+      "loss": 0.5049,
+      "step": 4063
+    },
+    {
+      "epoch": 0.9065358019183583,
+      "grad_norm": 0.15549324452877045,
+      "learning_rate": 1.5956963509070513e-05,
+      "loss": 0.4733,
+      "step": 4064
+    },
+    {
+      "epoch": 0.9067588668302476,
+      "grad_norm": 0.1653992384672165,
+      "learning_rate": 1.59550730468987e-05,
+      "loss": 0.4925,
+      "step": 4065
+    },
+    {
+      "epoch": 0.9069819317421369,
+      "grad_norm": 0.1693269908428192,
+      "learning_rate": 1.5953182254897478e-05,
+      "loss": 0.4908,
+      "step": 4066
+    },
+    {
+      "epoch": 0.9072049966540263,
+      "grad_norm": 0.16473425924777985,
+      "learning_rate": 1.5951291133171577e-05,
+      "loss": 0.4741,
+      "step": 4067
+    },
+    {
+      "epoch": 0.9074280615659157,
+      "grad_norm": 0.16339989006519318,
+      "learning_rate": 1.5949399681825738e-05,
+      "loss": 0.4574,
+      "step": 4068
+    },
+    {
+      "epoch": 0.907651126477805,
+      "grad_norm": 0.16389763355255127,
+      "learning_rate": 1.5947507900964723e-05,
+      "loss": 0.4951,
+      "step": 4069
+    },
+    {
+      "epoch": 0.9078741913896944,
+      "grad_norm": 0.16346803307533264,
+      "learning_rate": 1.594561579069331e-05,
+      "loss": 0.4794,
+      "step": 4070
+    },
+    {
+      "epoch": 0.9080972563015838,
+      "grad_norm": 0.16627097129821777,
+      "learning_rate": 1.5943723351116293e-05,
+      "loss": 0.4726,
+      "step": 4071
+    },
+    {
+      "epoch": 0.9083203212134732,
+      "grad_norm": 0.17151188850402832,
+      "learning_rate": 1.5941830582338488e-05,
+      "loss": 0.5246,
+      "step": 4072
+    },
+    {
+      "epoch": 0.9085433861253625,
+      "grad_norm": 0.15683230757713318,
+      "learning_rate": 1.593993748446473e-05,
+      "loss": 0.4771,
+      "step": 4073
+    },
+    {
+      "epoch": 0.9087664510372518,
+      "grad_norm": 0.1879899501800537,
+      "learning_rate": 1.5938044057599873e-05,
+      "loss": 0.5102,
+      "step": 4074
+    },
+    {
+      "epoch": 0.9089895159491412,
+      "grad_norm": 0.16234296560287476,
+      "learning_rate": 1.593615030184878e-05,
+      "loss": 0.4906,
+      "step": 4075
+    },
+    {
+      "epoch": 0.9092125808610305,
+      "grad_norm": 0.15749451518058777,
+      "learning_rate": 1.593425621731635e-05,
+      "loss": 0.4867,
+      "step": 4076
+    },
+    {
+      "epoch": 0.9094356457729199,
+      "grad_norm": 0.1519862711429596,
+      "learning_rate": 1.593236180410748e-05,
+      "loss": 0.4644,
+      "step": 4077
+    },
+    {
+      "epoch": 0.9096587106848093,
+      "grad_norm": 0.20612524449825287,
+      "learning_rate": 1.5930467062327096e-05,
+      "loss": 0.5017,
+      "step": 4078
+    },
+    {
+      "epoch": 0.9098817755966987,
+      "grad_norm": 0.16134196519851685,
+      "learning_rate": 1.5928571992080142e-05,
+      "loss": 0.4801,
+      "step": 4079
+    },
+    {
+      "epoch": 0.910104840508588,
+      "grad_norm": 0.15148447453975677,
+      "learning_rate": 1.592667659347158e-05,
+      "loss": 0.4697,
+      "step": 4080
+    },
+    {
+      "epoch": 0.9103279054204774,
+      "grad_norm": 0.1598382443189621,
+      "learning_rate": 1.5924780866606387e-05,
+      "loss": 0.4976,
+      "step": 4081
+    },
+    {
+      "epoch": 0.9105509703323668,
+      "grad_norm": 0.15935340523719788,
+      "learning_rate": 1.592288481158956e-05,
+      "loss": 0.4806,
+      "step": 4082
+    },
+    {
+      "epoch": 0.910774035244256,
+      "grad_norm": 0.16446979343891144,
+      "learning_rate": 1.5920988428526117e-05,
+      "loss": 0.4966,
+      "step": 4083
+    },
+    {
+      "epoch": 0.9109971001561454,
+      "grad_norm": 0.15920083224773407,
+      "learning_rate": 1.591909171752109e-05,
+      "loss": 0.4917,
+      "step": 4084
+    },
+    {
+      "epoch": 0.9112201650680348,
+      "grad_norm": 0.16009603440761566,
+      "learning_rate": 1.5917194678679532e-05,
+      "loss": 0.5169,
+      "step": 4085
+    },
+    {
+      "epoch": 0.9114432299799241,
+      "grad_norm": 0.17751926183700562,
+      "learning_rate": 1.5915297312106513e-05,
+      "loss": 0.4804,
+      "step": 4086
+    },
+    {
+      "epoch": 0.9116662948918135,
+      "grad_norm": 0.18244434893131256,
+      "learning_rate": 1.5913399617907116e-05,
+      "loss": 0.487,
+      "step": 4087
+    },
+    {
+      "epoch": 0.9118893598037029,
+      "grad_norm": 0.15786734223365784,
+      "learning_rate": 1.5911501596186455e-05,
+      "loss": 0.4986,
+      "step": 4088
+    },
+    {
+      "epoch": 0.9121124247155923,
+      "grad_norm": 0.15740512311458588,
+      "learning_rate": 1.5909603247049654e-05,
+      "loss": 0.4895,
+      "step": 4089
+    },
+    {
+      "epoch": 0.9123354896274816,
+      "grad_norm": 0.1518784463405609,
+      "learning_rate": 1.5907704570601845e-05,
+      "loss": 0.4522,
+      "step": 4090
+    },
+    {
+      "epoch": 0.912558554539371,
+      "grad_norm": 0.1830586940050125,
+      "learning_rate": 1.59058055669482e-05,
+      "loss": 0.513,
+      "step": 4091
+    },
+    {
+      "epoch": 0.9127816194512604,
+      "grad_norm": 0.16128897666931152,
+      "learning_rate": 1.5903906236193892e-05,
+      "loss": 0.4706,
+      "step": 4092
+    },
+    {
+      "epoch": 0.9130046843631496,
+      "grad_norm": 0.17136983573436737,
+      "learning_rate": 1.5902006578444123e-05,
+      "loss": 0.4867,
+      "step": 4093
+    },
+    {
+      "epoch": 0.913227749275039,
+      "grad_norm": 0.15806463360786438,
+      "learning_rate": 1.59001065938041e-05,
+      "loss": 0.5066,
+      "step": 4094
+    },
+    {
+      "epoch": 0.9134508141869284,
+      "grad_norm": 0.17267145216464996,
+      "learning_rate": 1.5898206282379063e-05,
+      "loss": 0.4805,
+      "step": 4095
+    },
+    {
+      "epoch": 0.9136738790988178,
+      "grad_norm": 0.15600056946277618,
+      "learning_rate": 1.5896305644274262e-05,
+      "loss": 0.4865,
+      "step": 4096
+    },
+    {
+      "epoch": 0.9138969440107071,
+      "grad_norm": 0.1608215868473053,
+      "learning_rate": 1.5894404679594963e-05,
+      "loss": 0.5102,
+      "step": 4097
+    },
+    {
+      "epoch": 0.9141200089225965,
+      "grad_norm": 0.15523095428943634,
+      "learning_rate": 1.5892503388446456e-05,
+      "loss": 0.4642,
+      "step": 4098
+    },
+    {
+      "epoch": 0.9143430738344859,
+      "grad_norm": 0.16346514225006104,
+      "learning_rate": 1.589060177093405e-05,
+      "loss": 0.4655,
+      "step": 4099
+    },
+    {
+      "epoch": 0.9145661387463752,
+      "grad_norm": 0.1508352905511856,
+      "learning_rate": 1.588869982716306e-05,
+      "loss": 0.4826,
+      "step": 4100
+    },
+    {
+      "epoch": 0.9147892036582645,
+      "grad_norm": 0.1622077375650406,
+      "learning_rate": 1.5886797557238832e-05,
+      "loss": 0.4876,
+      "step": 4101
+    },
+    {
+      "epoch": 0.9150122685701539,
+      "grad_norm": 0.17161275446414948,
+      "learning_rate": 1.588489496126673e-05,
+      "loss": 0.4963,
+      "step": 4102
+    },
+    {
+      "epoch": 0.9152353334820432,
+      "grad_norm": 0.15936551988124847,
+      "learning_rate": 1.5882992039352122e-05,
+      "loss": 0.4768,
+      "step": 4103
+    },
+    {
+      "epoch": 0.9154583983939326,
+      "grad_norm": 0.1664397269487381,
+      "learning_rate": 1.588108879160041e-05,
+      "loss": 0.4927,
+      "step": 4104
+    },
+    {
+      "epoch": 0.915681463305822,
+      "grad_norm": 0.16500705480575562,
+      "learning_rate": 1.5879185218117012e-05,
+      "loss": 0.4909,
+      "step": 4105
+    },
+    {
+      "epoch": 0.9159045282177114,
+      "grad_norm": 0.16480299830436707,
+      "learning_rate": 1.5877281319007352e-05,
+      "loss": 0.468,
+      "step": 4106
+    },
+    {
+      "epoch": 0.9161275931296007,
+      "grad_norm": 0.17092856764793396,
+      "learning_rate": 1.5875377094376883e-05,
+      "loss": 0.5188,
+      "step": 4107
+    },
+    {
+      "epoch": 0.9163506580414901,
+      "grad_norm": 0.16929516196250916,
+      "learning_rate": 1.5873472544331073e-05,
+      "loss": 0.4932,
+      "step": 4108
+    },
+    {
+      "epoch": 0.9165737229533795,
+      "grad_norm": 0.1531393676996231,
+      "learning_rate": 1.5871567668975406e-05,
+      "loss": 0.4512,
+      "step": 4109
+    },
+    {
+      "epoch": 0.9167967878652687,
+      "grad_norm": 0.15918409824371338,
+      "learning_rate": 1.586966246841539e-05,
+      "loss": 0.5019,
+      "step": 4110
+    },
+    {
+      "epoch": 0.9170198527771581,
+      "grad_norm": 0.19668515026569366,
+      "learning_rate": 1.5867756942756548e-05,
+      "loss": 0.5106,
+      "step": 4111
+    },
+    {
+      "epoch": 0.9172429176890475,
+      "grad_norm": 0.15807564556598663,
+      "learning_rate": 1.5865851092104414e-05,
+      "loss": 0.4738,
+      "step": 4112
+    },
+    {
+      "epoch": 0.9174659826009369,
+      "grad_norm": 0.16922055184841156,
+      "learning_rate": 1.586394491656455e-05,
+      "loss": 0.5086,
+      "step": 4113
+    },
+    {
+      "epoch": 0.9176890475128262,
+      "grad_norm": 0.18552662432193756,
+      "learning_rate": 1.586203841624253e-05,
+      "loss": 0.5246,
+      "step": 4114
+    },
+    {
+      "epoch": 0.9179121124247156,
+      "grad_norm": 0.16202445328235626,
+      "learning_rate": 1.5860131591243945e-05,
+      "loss": 0.5113,
+      "step": 4115
+    },
+    {
+      "epoch": 0.918135177336605,
+      "grad_norm": 0.16352488100528717,
+      "learning_rate": 1.5858224441674416e-05,
+      "loss": 0.4739,
+      "step": 4116
+    },
+    {
+      "epoch": 0.9183582422484943,
+      "grad_norm": 0.16634705662727356,
+      "learning_rate": 1.5856316967639566e-05,
+      "loss": 0.5221,
+      "step": 4117
+    },
+    {
+      "epoch": 0.9185813071603837,
+      "grad_norm": 0.16568778455257416,
+      "learning_rate": 1.5854409169245043e-05,
+      "loss": 0.4476,
+      "step": 4118
+    },
+    {
+      "epoch": 0.9188043720722731,
+      "grad_norm": 0.160440593957901,
+      "learning_rate": 1.5852501046596516e-05,
+      "loss": 0.5054,
+      "step": 4119
+    },
+    {
+      "epoch": 0.9190274369841623,
+      "grad_norm": 0.1603170782327652,
+      "learning_rate": 1.5850592599799668e-05,
+      "loss": 0.4947,
+      "step": 4120
+    },
+    {
+      "epoch": 0.9192505018960517,
+      "grad_norm": 0.16584816575050354,
+      "learning_rate": 1.5848683828960195e-05,
+      "loss": 0.5014,
+      "step": 4121
+    },
+    {
+      "epoch": 0.9194735668079411,
+      "grad_norm": 0.1588207632303238,
+      "learning_rate": 1.584677473418383e-05,
+      "loss": 0.4973,
+      "step": 4122
+    },
+    {
+      "epoch": 0.9196966317198305,
+      "grad_norm": 0.1574103981256485,
+      "learning_rate": 1.5844865315576296e-05,
+      "loss": 0.5214,
+      "step": 4123
+    },
+    {
+      "epoch": 0.9199196966317198,
+      "grad_norm": 0.17080260813236237,
+      "learning_rate": 1.584295557324336e-05,
+      "loss": 0.5067,
+      "step": 4124
+    },
+    {
+      "epoch": 0.9201427615436092,
+      "grad_norm": 0.15937574207782745,
+      "learning_rate": 1.584104550729079e-05,
+      "loss": 0.4971,
+      "step": 4125
+    },
+    {
+      "epoch": 0.9203658264554986,
+      "grad_norm": 0.15593832731246948,
+      "learning_rate": 1.5839135117824375e-05,
+      "loss": 0.4844,
+      "step": 4126
+    },
+    {
+      "epoch": 0.9205888913673879,
+      "grad_norm": 0.18794505298137665,
+      "learning_rate": 1.583722440494993e-05,
+      "loss": 0.5135,
+      "step": 4127
+    },
+    {
+      "epoch": 0.9208119562792773,
+      "grad_norm": 0.183371901512146,
+      "learning_rate": 1.5835313368773276e-05,
+      "loss": 0.4861,
+      "step": 4128
+    },
+    {
+      "epoch": 0.9210350211911666,
+      "grad_norm": 0.153816357254982,
+      "learning_rate": 1.583340200940027e-05,
+      "loss": 0.4921,
+      "step": 4129
+    },
+    {
+      "epoch": 0.921258086103056,
+      "grad_norm": 0.16074557602405548,
+      "learning_rate": 1.583149032693676e-05,
+      "loss": 0.4566,
+      "step": 4130
+    },
+    {
+      "epoch": 0.9214811510149453,
+      "grad_norm": 0.15913258492946625,
+      "learning_rate": 1.5829578321488636e-05,
+      "loss": 0.4838,
+      "step": 4131
+    },
+    {
+      "epoch": 0.9217042159268347,
+      "grad_norm": 0.15266257524490356,
+      "learning_rate": 1.58276659931618e-05,
+      "loss": 0.4704,
+      "step": 4132
+    },
+    {
+      "epoch": 0.9219272808387241,
+      "grad_norm": 0.15182676911354065,
+      "learning_rate": 1.5825753342062155e-05,
+      "loss": 0.4825,
+      "step": 4133
+    },
+    {
+      "epoch": 0.9221503457506134,
+      "grad_norm": 0.15525878965854645,
+      "learning_rate": 1.582384036829565e-05,
+      "loss": 0.4636,
+      "step": 4134
+    },
+    {
+      "epoch": 0.9223734106625028,
+      "grad_norm": 0.15554243326187134,
+      "learning_rate": 1.582192707196823e-05,
+      "loss": 0.4945,
+      "step": 4135
+    },
+    {
+      "epoch": 0.9225964755743922,
+      "grad_norm": 0.15928377211093903,
+      "learning_rate": 1.582001345318587e-05,
+      "loss": 0.4772,
+      "step": 4136
+    },
+    {
+      "epoch": 0.9228195404862815,
+      "grad_norm": 0.15914230048656464,
+      "learning_rate": 1.581809951205455e-05,
+      "loss": 0.4723,
+      "step": 4137
+    },
+    {
+      "epoch": 0.9230426053981708,
+      "grad_norm": 0.16770337522029877,
+      "learning_rate": 1.581618524868029e-05,
+      "loss": 0.4579,
+      "step": 4138
+    },
+    {
+      "epoch": 0.9232656703100602,
+      "grad_norm": 0.16558465361595154,
+      "learning_rate": 1.58142706631691e-05,
+      "loss": 0.5186,
+      "step": 4139
+    },
+    {
+      "epoch": 0.9234887352219496,
+      "grad_norm": 0.16763773560523987,
+      "learning_rate": 1.5812355755627028e-05,
+      "loss": 0.4887,
+      "step": 4140
+    },
+    {
+      "epoch": 0.9237118001338389,
+      "grad_norm": 0.1562553197145462,
+      "learning_rate": 1.5810440526160133e-05,
+      "loss": 0.4953,
+      "step": 4141
+    },
+    {
+      "epoch": 0.9239348650457283,
+      "grad_norm": 0.1571418195962906,
+      "learning_rate": 1.5808524974874493e-05,
+      "loss": 0.4753,
+      "step": 4142
+    },
+    {
+      "epoch": 0.9241579299576177,
+      "grad_norm": 0.17362381517887115,
+      "learning_rate": 1.5806609101876203e-05,
+      "loss": 0.4704,
+      "step": 4143
+    },
+    {
+      "epoch": 0.924380994869507,
+      "grad_norm": 0.1619691401720047,
+      "learning_rate": 1.580469290727138e-05,
+      "loss": 0.476,
+      "step": 4144
+    },
+    {
+      "epoch": 0.9246040597813964,
+      "grad_norm": 0.16508683562278748,
+      "learning_rate": 1.5802776391166146e-05,
+      "loss": 0.5118,
+      "step": 4145
+    },
+    {
+      "epoch": 0.9248271246932858,
+      "grad_norm": 0.16421370208263397,
+      "learning_rate": 1.5800859553666655e-05,
+      "loss": 0.4876,
+      "step": 4146
+    },
+    {
+      "epoch": 0.9250501896051752,
+      "grad_norm": 0.16035978496074677,
+      "learning_rate": 1.5798942394879073e-05,
+      "loss": 0.5048,
+      "step": 4147
+    },
+    {
+      "epoch": 0.9252732545170644,
+      "grad_norm": 0.16965007781982422,
+      "learning_rate": 1.5797024914909584e-05,
+      "loss": 0.4932,
+      "step": 4148
+    },
+    {
+      "epoch": 0.9254963194289538,
+      "grad_norm": 0.16695380210876465,
+      "learning_rate": 1.5795107113864393e-05,
+      "loss": 0.5188,
+      "step": 4149
+    },
+    {
+      "epoch": 0.9257193843408432,
+      "grad_norm": 0.1638556569814682,
+      "learning_rate": 1.5793188991849717e-05,
+      "loss": 0.4764,
+      "step": 4150
+    },
+    {
+      "epoch": 0.9259424492527325,
+      "grad_norm": 0.16190040111541748,
+      "learning_rate": 1.579127054897179e-05,
+      "loss": 0.4877,
+      "step": 4151
+    },
+    {
+      "epoch": 0.9261655141646219,
+      "grad_norm": 0.16219079494476318,
+      "learning_rate": 1.5789351785336874e-05,
+      "loss": 0.4823,
+      "step": 4152
+    },
+    {
+      "epoch": 0.9263885790765113,
+      "grad_norm": 0.15557697415351868,
+      "learning_rate": 1.5787432701051242e-05,
+      "loss": 0.4826,
+      "step": 4153
+    },
+    {
+      "epoch": 0.9266116439884007,
+      "grad_norm": 0.15617480874061584,
+      "learning_rate": 1.578551329622118e-05,
+      "loss": 0.5012,
+      "step": 4154
+    },
+    {
+      "epoch": 0.92683470890029,
+      "grad_norm": 0.16417647898197174,
+      "learning_rate": 1.5783593570953e-05,
+      "loss": 0.4692,
+      "step": 4155
+    },
+    {
+      "epoch": 0.9270577738121794,
+      "grad_norm": 0.16066023707389832,
+      "learning_rate": 1.578167352535303e-05,
+      "loss": 0.4932,
+      "step": 4156
+    },
+    {
+      "epoch": 0.9272808387240687,
+      "grad_norm": 0.15020349621772766,
+      "learning_rate": 1.577975315952761e-05,
+      "loss": 0.4593,
+      "step": 4157
+    },
+    {
+      "epoch": 0.927503903635958,
+      "grad_norm": 0.15949909389019012,
+      "learning_rate": 1.57778324735831e-05,
+      "loss": 0.4952,
+      "step": 4158
+    },
+    {
+      "epoch": 0.9277269685478474,
+      "grad_norm": 0.169538676738739,
+      "learning_rate": 1.577591146762589e-05,
+      "loss": 0.4989,
+      "step": 4159
+    },
+    {
+      "epoch": 0.9279500334597368,
+      "grad_norm": 0.1595795601606369,
+      "learning_rate": 1.5773990141762366e-05,
+      "loss": 0.4886,
+      "step": 4160
+    },
+    {
+      "epoch": 0.9281730983716261,
+      "grad_norm": 0.17213213443756104,
+      "learning_rate": 1.577206849609895e-05,
+      "loss": 0.4383,
+      "step": 4161
+    },
+    {
+      "epoch": 0.9283961632835155,
+      "grad_norm": 0.15668156743049622,
+      "learning_rate": 1.5770146530742075e-05,
+      "loss": 0.4672,
+      "step": 4162
+    },
+    {
+      "epoch": 0.9286192281954049,
+      "grad_norm": 0.15910851955413818,
+      "learning_rate": 1.576822424579819e-05,
+      "loss": 0.5136,
+      "step": 4163
+    },
+    {
+      "epoch": 0.9288422931072943,
+      "grad_norm": 0.1524379998445511,
+      "learning_rate": 1.5766301641373755e-05,
+      "loss": 0.4752,
+      "step": 4164
+    },
+    {
+      "epoch": 0.9290653580191836,
+      "grad_norm": 0.1620722860097885,
+      "learning_rate": 1.5764378717575272e-05,
+      "loss": 0.46,
+      "step": 4165
+    },
+    {
+      "epoch": 0.9292884229310729,
+      "grad_norm": 0.16785915195941925,
+      "learning_rate": 1.576245547450923e-05,
+      "loss": 0.5097,
+      "step": 4166
+    },
+    {
+      "epoch": 0.9295114878429623,
+      "grad_norm": 0.15819083154201508,
+      "learning_rate": 1.5760531912282163e-05,
+      "loss": 0.5042,
+      "step": 4167
+    },
+    {
+      "epoch": 0.9297345527548516,
+      "grad_norm": 0.15982572734355927,
+      "learning_rate": 1.57586080310006e-05,
+      "loss": 0.4832,
+      "step": 4168
+    },
+    {
+      "epoch": 0.929957617666741,
+      "grad_norm": 0.16882368922233582,
+      "learning_rate": 1.57566838307711e-05,
+      "loss": 0.495,
+      "step": 4169
+    },
+    {
+      "epoch": 0.9301806825786304,
+      "grad_norm": 0.15450328588485718,
+      "learning_rate": 1.575475931170024e-05,
+      "loss": 0.4831,
+      "step": 4170
+    },
+    {
+      "epoch": 0.9304037474905198,
+      "grad_norm": 0.15884611010551453,
+      "learning_rate": 1.575283447389461e-05,
+      "loss": 0.4447,
+      "step": 4171
+    },
+    {
+      "epoch": 0.9306268124024091,
+      "grad_norm": 0.16536371409893036,
+      "learning_rate": 1.575090931746082e-05,
+      "loss": 0.4708,
+      "step": 4172
+    },
+    {
+      "epoch": 0.9308498773142985,
+      "grad_norm": 0.16749803721904755,
+      "learning_rate": 1.57489838425055e-05,
+      "loss": 0.4927,
+      "step": 4173
+    },
+    {
+      "epoch": 0.9310729422261879,
+      "grad_norm": 0.17306271195411682,
+      "learning_rate": 1.5747058049135286e-05,
+      "loss": 0.5355,
+      "step": 4174
+    },
+    {
+      "epoch": 0.9312960071380771,
+      "grad_norm": 0.1781659871339798,
+      "learning_rate": 1.5745131937456853e-05,
+      "loss": 0.5054,
+      "step": 4175
+    },
+    {
+      "epoch": 0.9315190720499665,
+      "grad_norm": 0.16959606111049652,
+      "learning_rate": 1.5743205507576873e-05,
+      "loss": 0.4799,
+      "step": 4176
+    },
+    {
+      "epoch": 0.9317421369618559,
+      "grad_norm": 0.16488516330718994,
+      "learning_rate": 1.5741278759602045e-05,
+      "loss": 0.4835,
+      "step": 4177
+    },
+    {
+      "epoch": 0.9319652018737452,
+      "grad_norm": 0.1639404445886612,
+      "learning_rate": 1.5739351693639085e-05,
+      "loss": 0.4757,
+      "step": 4178
+    },
+    {
+      "epoch": 0.9321882667856346,
+      "grad_norm": 0.17086337506771088,
+      "learning_rate": 1.573742430979473e-05,
+      "loss": 0.4982,
+      "step": 4179
+    },
+    {
+      "epoch": 0.932411331697524,
+      "grad_norm": 0.162103071808815,
+      "learning_rate": 1.5735496608175722e-05,
+      "loss": 0.4593,
+      "step": 4180
+    },
+    {
+      "epoch": 0.9326343966094134,
+      "grad_norm": 0.1637817919254303,
+      "learning_rate": 1.5733568588888835e-05,
+      "loss": 0.4776,
+      "step": 4181
+    },
+    {
+      "epoch": 0.9328574615213027,
+      "grad_norm": 0.1821950078010559,
+      "learning_rate": 1.5731640252040857e-05,
+      "loss": 0.5159,
+      "step": 4182
+    },
+    {
+      "epoch": 0.9330805264331921,
+      "grad_norm": 0.21807077527046204,
+      "learning_rate": 1.5729711597738587e-05,
+      "loss": 0.4721,
+      "step": 4183
+    },
+    {
+      "epoch": 0.9333035913450815,
+      "grad_norm": 0.1628894805908203,
+      "learning_rate": 1.5727782626088844e-05,
+      "loss": 0.5033,
+      "step": 4184
+    },
+    {
+      "epoch": 0.9335266562569707,
+      "grad_norm": 0.158598393201828,
+      "learning_rate": 1.5725853337198476e-05,
+      "loss": 0.4743,
+      "step": 4185
+    },
+    {
+      "epoch": 0.9337497211688601,
+      "grad_norm": 0.16272245347499847,
+      "learning_rate": 1.5723923731174327e-05,
+      "loss": 0.4952,
+      "step": 4186
+    },
+    {
+      "epoch": 0.9339727860807495,
+      "grad_norm": 0.16067437827587128,
+      "learning_rate": 1.5721993808123283e-05,
+      "loss": 0.4879,
+      "step": 4187
+    },
+    {
+      "epoch": 0.9341958509926389,
+      "grad_norm": 0.16409751772880554,
+      "learning_rate": 1.5720063568152222e-05,
+      "loss": 0.5051,
+      "step": 4188
+    },
+    {
+      "epoch": 0.9344189159045282,
+      "grad_norm": 0.15659625828266144,
+      "learning_rate": 1.5718133011368065e-05,
+      "loss": 0.4497,
+      "step": 4189
+    },
+    {
+      "epoch": 0.9346419808164176,
+      "grad_norm": 0.17328283190727234,
+      "learning_rate": 1.5716202137877732e-05,
+      "loss": 0.4927,
+      "step": 4190
+    },
+    {
+      "epoch": 0.934865045728307,
+      "grad_norm": 0.15970586240291595,
+      "learning_rate": 1.5714270947788168e-05,
+      "loss": 0.4706,
+      "step": 4191
+    },
+    {
+      "epoch": 0.9350881106401963,
+      "grad_norm": 0.16526605188846588,
+      "learning_rate": 1.5712339441206335e-05,
+      "loss": 0.5266,
+      "step": 4192
+    },
+    {
+      "epoch": 0.9353111755520856,
+      "grad_norm": 0.16756314039230347,
+      "learning_rate": 1.5710407618239215e-05,
+      "loss": 0.4818,
+      "step": 4193
+    },
+    {
+      "epoch": 0.935534240463975,
+      "grad_norm": 0.16548283398151398,
+      "learning_rate": 1.57084754789938e-05,
+      "loss": 0.4586,
+      "step": 4194
+    },
+    {
+      "epoch": 0.9357573053758643,
+      "grad_norm": 0.1704016625881195,
+      "learning_rate": 1.57065430235771e-05,
+      "loss": 0.4999,
+      "step": 4195
+    },
+    {
+      "epoch": 0.9359803702877537,
+      "grad_norm": 0.16027678549289703,
+      "learning_rate": 1.5704610252096158e-05,
+      "loss": 0.5155,
+      "step": 4196
+    },
+    {
+      "epoch": 0.9362034351996431,
+      "grad_norm": 0.16171659529209137,
+      "learning_rate": 1.5702677164658013e-05,
+      "loss": 0.482,
+      "step": 4197
+    },
+    {
+      "epoch": 0.9364265001115325,
+      "grad_norm": 0.15903575718402863,
+      "learning_rate": 1.5700743761369735e-05,
+      "loss": 0.4862,
+      "step": 4198
+    },
+    {
+      "epoch": 0.9366495650234218,
+      "grad_norm": 0.16277752816677094,
+      "learning_rate": 1.569881004233841e-05,
+      "loss": 0.5011,
+      "step": 4199
+    },
+    {
+      "epoch": 0.9368726299353112,
+      "grad_norm": 0.18112027645111084,
+      "learning_rate": 1.5696876007671137e-05,
+      "loss": 0.5124,
+      "step": 4200
+    },
+    {
+      "epoch": 0.9370956948472006,
+      "grad_norm": 0.1702015995979309,
+      "learning_rate": 1.5694941657475037e-05,
+      "loss": 0.5233,
+      "step": 4201
+    },
+    {
+      "epoch": 0.9373187597590898,
+      "grad_norm": 0.15752531588077545,
+      "learning_rate": 1.5693006991857248e-05,
+      "loss": 0.486,
+      "step": 4202
+    },
+    {
+      "epoch": 0.9375418246709792,
+      "grad_norm": 0.16015774011611938,
+      "learning_rate": 1.5691072010924915e-05,
+      "loss": 0.4816,
+      "step": 4203
+    },
+    {
+      "epoch": 0.9377648895828686,
+      "grad_norm": 0.17005962133407593,
+      "learning_rate": 1.568913671478522e-05,
+      "loss": 0.4986,
+      "step": 4204
+    },
+    {
+      "epoch": 0.937987954494758,
+      "grad_norm": 0.15905898809432983,
+      "learning_rate": 1.5687201103545343e-05,
+      "loss": 0.5031,
+      "step": 4205
+    },
+    {
+      "epoch": 0.9382110194066473,
+      "grad_norm": 0.16087795794010162,
+      "learning_rate": 1.56852651773125e-05,
+      "loss": 0.4957,
+      "step": 4206
+    },
+    {
+      "epoch": 0.9384340843185367,
+      "grad_norm": 0.16135592758655548,
+      "learning_rate": 1.5683328936193908e-05,
+      "loss": 0.5094,
+      "step": 4207
+    },
+    {
+      "epoch": 0.9386571492304261,
+      "grad_norm": 0.18367910385131836,
+      "learning_rate": 1.568139238029681e-05,
+      "loss": 0.515,
+      "step": 4208
+    },
+    {
+      "epoch": 0.9388802141423154,
+      "grad_norm": 0.16034136712551117,
+      "learning_rate": 1.5679455509728468e-05,
+      "loss": 0.5168,
+      "step": 4209
+    },
+    {
+      "epoch": 0.9391032790542048,
+      "grad_norm": 0.17193713784217834,
+      "learning_rate": 1.567751832459615e-05,
+      "loss": 0.5078,
+      "step": 4210
+    },
+    {
+      "epoch": 0.9393263439660942,
+      "grad_norm": 0.15431612730026245,
+      "learning_rate": 1.5675580825007158e-05,
+      "loss": 0.4718,
+      "step": 4211
+    },
+    {
+      "epoch": 0.9395494088779834,
+      "grad_norm": 0.16114309430122375,
+      "learning_rate": 1.5673643011068796e-05,
+      "loss": 0.4779,
+      "step": 4212
+    },
+    {
+      "epoch": 0.9397724737898728,
+      "grad_norm": 0.15799474716186523,
+      "learning_rate": 1.5671704882888396e-05,
+      "loss": 0.5086,
+      "step": 4213
+    },
+    {
+      "epoch": 0.9399955387017622,
+      "grad_norm": 0.1653827577829361,
+      "learning_rate": 1.5669766440573302e-05,
+      "loss": 0.488,
+      "step": 4214
+    },
+    {
+      "epoch": 0.9402186036136516,
+      "grad_norm": 0.16851937770843506,
+      "learning_rate": 1.566782768423088e-05,
+      "loss": 0.4806,
+      "step": 4215
+    },
+    {
+      "epoch": 0.9404416685255409,
+      "grad_norm": 0.17142115533351898,
+      "learning_rate": 1.566588861396851e-05,
+      "loss": 0.5131,
+      "step": 4216
+    },
+    {
+      "epoch": 0.9406647334374303,
+      "grad_norm": 0.15409086644649506,
+      "learning_rate": 1.5663949229893587e-05,
+      "loss": 0.4624,
+      "step": 4217
+    },
+    {
+      "epoch": 0.9408877983493197,
+      "grad_norm": 0.16042175889015198,
+      "learning_rate": 1.566200953211353e-05,
+      "loss": 0.4816,
+      "step": 4218
+    },
+    {
+      "epoch": 0.941110863261209,
+      "grad_norm": 0.15703439712524414,
+      "learning_rate": 1.5660069520735766e-05,
+      "loss": 0.504,
+      "step": 4219
+    },
+    {
+      "epoch": 0.9413339281730984,
+      "grad_norm": 0.16937057673931122,
+      "learning_rate": 1.565812919586775e-05,
+      "loss": 0.4766,
+      "step": 4220
+    },
+    {
+      "epoch": 0.9415569930849877,
+      "grad_norm": 0.17867791652679443,
+      "learning_rate": 1.565618855761695e-05,
+      "loss": 0.4872,
+      "step": 4221
+    },
+    {
+      "epoch": 0.9417800579968771,
+      "grad_norm": 0.15998175740242004,
+      "learning_rate": 1.5654247606090846e-05,
+      "loss": 0.4844,
+      "step": 4222
+    },
+    {
+      "epoch": 0.9420031229087664,
+      "grad_norm": 0.15649773180484772,
+      "learning_rate": 1.5652306341396943e-05,
+      "loss": 0.4774,
+      "step": 4223
+    },
+    {
+      "epoch": 0.9422261878206558,
+      "grad_norm": 0.15873044729232788,
+      "learning_rate": 1.5650364763642764e-05,
+      "loss": 0.5239,
+      "step": 4224
+    },
+    {
+      "epoch": 0.9424492527325452,
+      "grad_norm": 0.15709929168224335,
+      "learning_rate": 1.564842287293584e-05,
+      "loss": 0.4875,
+      "step": 4225
+    },
+    {
+      "epoch": 0.9426723176444345,
+      "grad_norm": 0.1562758833169937,
+      "learning_rate": 1.5646480669383726e-05,
+      "loss": 0.5132,
+      "step": 4226
+    },
+    {
+      "epoch": 0.9428953825563239,
+      "grad_norm": 0.162166565656662,
+      "learning_rate": 1.5644538153093995e-05,
+      "loss": 0.4901,
+      "step": 4227
+    },
+    {
+      "epoch": 0.9431184474682133,
+      "grad_norm": 0.1618107706308365,
+      "learning_rate": 1.564259532417424e-05,
+      "loss": 0.4722,
+      "step": 4228
+    },
+    {
+      "epoch": 0.9433415123801027,
+      "grad_norm": 0.15450115501880646,
+      "learning_rate": 1.5640652182732057e-05,
+      "loss": 0.4498,
+      "step": 4229
+    },
+    {
+      "epoch": 0.943564577291992,
+      "grad_norm": 0.1606828272342682,
+      "learning_rate": 1.563870872887508e-05,
+      "loss": 0.4798,
+      "step": 4230
+    },
+    {
+      "epoch": 0.9437876422038813,
+      "grad_norm": 0.15939275920391083,
+      "learning_rate": 1.5636764962710936e-05,
+      "loss": 0.4773,
+      "step": 4231
+    },
+    {
+      "epoch": 0.9440107071157707,
+      "grad_norm": 0.16131429374217987,
+      "learning_rate": 1.5634820884347303e-05,
+      "loss": 0.4861,
+      "step": 4232
+    },
+    {
+      "epoch": 0.94423377202766,
+      "grad_norm": 0.15643377602100372,
+      "learning_rate": 1.563287649389184e-05,
+      "loss": 0.5172,
+      "step": 4233
+    },
+    {
+      "epoch": 0.9444568369395494,
+      "grad_norm": 0.1604384183883667,
+      "learning_rate": 1.5630931791452246e-05,
+      "loss": 0.4808,
+      "step": 4234
+    },
+    {
+      "epoch": 0.9446799018514388,
+      "grad_norm": 0.15961483120918274,
+      "learning_rate": 1.5628986777136223e-05,
+      "loss": 0.4648,
+      "step": 4235
+    },
+    {
+      "epoch": 0.9449029667633281,
+      "grad_norm": 0.16450265049934387,
+      "learning_rate": 1.562704145105151e-05,
+      "loss": 0.5006,
+      "step": 4236
+    },
+    {
+      "epoch": 0.9451260316752175,
+      "grad_norm": 0.1608218103647232,
+      "learning_rate": 1.5625095813305847e-05,
+      "loss": 0.4983,
+      "step": 4237
+    },
+    {
+      "epoch": 0.9453490965871069,
+      "grad_norm": 0.1569865345954895,
+      "learning_rate": 1.5623149864006993e-05,
+      "loss": 0.4969,
+      "step": 4238
+    },
+    {
+      "epoch": 0.9455721614989963,
+      "grad_norm": 0.1610192060470581,
+      "learning_rate": 1.5621203603262727e-05,
+      "loss": 0.4842,
+      "step": 4239
+    },
+    {
+      "epoch": 0.9457952264108855,
+      "grad_norm": 0.18263404071331024,
+      "learning_rate": 1.561925703118085e-05,
+      "loss": 0.5053,
+      "step": 4240
+    },
+    {
+      "epoch": 0.9460182913227749,
+      "grad_norm": 0.16298796236515045,
+      "learning_rate": 1.561731014786917e-05,
+      "loss": 0.4906,
+      "step": 4241
+    },
+    {
+      "epoch": 0.9462413562346643,
+      "grad_norm": 0.16036361455917358,
+      "learning_rate": 1.5615362953435517e-05,
+      "loss": 0.4647,
+      "step": 4242
+    },
+    {
+      "epoch": 0.9464644211465536,
+      "grad_norm": 0.16408760845661163,
+      "learning_rate": 1.5613415447987743e-05,
+      "loss": 0.4836,
+      "step": 4243
+    },
+    {
+      "epoch": 0.946687486058443,
+      "grad_norm": 0.1685844510793686,
+      "learning_rate": 1.5611467631633713e-05,
+      "loss": 0.5509,
+      "step": 4244
+    },
+    {
+      "epoch": 0.9469105509703324,
+      "grad_norm": 0.16026711463928223,
+      "learning_rate": 1.5609519504481306e-05,
+      "loss": 0.4926,
+      "step": 4245
+    },
+    {
+      "epoch": 0.9471336158822218,
+      "grad_norm": 0.19063352048397064,
+      "learning_rate": 1.560757106663843e-05,
+      "loss": 0.4975,
+      "step": 4246
+    },
+    {
+      "epoch": 0.9473566807941111,
+      "grad_norm": 0.1648816168308258,
+      "learning_rate": 1.560562231821299e-05,
+      "loss": 0.5142,
+      "step": 4247
+    },
+    {
+      "epoch": 0.9475797457060005,
+      "grad_norm": 0.1545959860086441,
+      "learning_rate": 1.5603673259312927e-05,
+      "loss": 0.4862,
+      "step": 4248
+    },
+    {
+      "epoch": 0.9478028106178898,
+      "grad_norm": 0.15768404304981232,
+      "learning_rate": 1.5601723890046188e-05,
+      "loss": 0.4708,
+      "step": 4249
+    },
+    {
+      "epoch": 0.9480258755297791,
+      "grad_norm": 0.1632533222436905,
+      "learning_rate": 1.5599774210520747e-05,
+      "loss": 0.5113,
+      "step": 4250
+    },
+    {
+      "epoch": 0.9482489404416685,
+      "grad_norm": 0.1614227592945099,
+      "learning_rate": 1.5597824220844583e-05,
+      "loss": 0.5081,
+      "step": 4251
+    },
+    {
+      "epoch": 0.9484720053535579,
+      "grad_norm": 0.15559621155261993,
+      "learning_rate": 1.55958739211257e-05,
+      "loss": 0.5012,
+      "step": 4252
+    },
+    {
+      "epoch": 0.9486950702654472,
+      "grad_norm": 0.16258159279823303,
+      "learning_rate": 1.5593923311472127e-05,
+      "loss": 0.5018,
+      "step": 4253
+    },
+    {
+      "epoch": 0.9489181351773366,
+      "grad_norm": 0.15811897814273834,
+      "learning_rate": 1.559197239199189e-05,
+      "loss": 0.4694,
+      "step": 4254
+    },
+    {
+      "epoch": 0.949141200089226,
+      "grad_norm": 0.15983614325523376,
+      "learning_rate": 1.5590021162793047e-05,
+      "loss": 0.4758,
+      "step": 4255
+    },
+    {
+      "epoch": 0.9493642650011154,
+      "grad_norm": 0.16571156680583954,
+      "learning_rate": 1.558806962398367e-05,
+      "loss": 0.4952,
+      "step": 4256
+    },
+    {
+      "epoch": 0.9495873299130047,
+      "grad_norm": 0.17233605682849884,
+      "learning_rate": 1.5586117775671844e-05,
+      "loss": 0.5252,
+      "step": 4257
+    },
+    {
+      "epoch": 0.949810394824894,
+      "grad_norm": 0.17330588400363922,
+      "learning_rate": 1.558416561796568e-05,
+      "loss": 0.5099,
+      "step": 4258
+    },
+    {
+      "epoch": 0.9500334597367834,
+      "grad_norm": 0.15399448573589325,
+      "learning_rate": 1.5582213150973296e-05,
+      "loss": 0.4697,
+      "step": 4259
+    },
+    {
+      "epoch": 0.9502565246486727,
+      "grad_norm": 0.17017517983913422,
+      "learning_rate": 1.5580260374802837e-05,
+      "loss": 0.4981,
+      "step": 4260
+    },
+    {
+      "epoch": 0.9504795895605621,
+      "grad_norm": 0.1677882969379425,
+      "learning_rate": 1.5578307289562457e-05,
+      "loss": 0.5042,
+      "step": 4261
+    },
+    {
+      "epoch": 0.9507026544724515,
+      "grad_norm": 0.15946677327156067,
+      "learning_rate": 1.557635389536033e-05,
+      "loss": 0.4697,
+      "step": 4262
+    },
+    {
+      "epoch": 0.9509257193843409,
+      "grad_norm": 1.8394982814788818,
+      "learning_rate": 1.557440019230465e-05,
+      "loss": 0.5449,
+      "step": 4263
+    },
+    {
+      "epoch": 0.9511487842962302,
+      "grad_norm": 0.16612844169139862,
+      "learning_rate": 1.5572446180503618e-05,
+      "loss": 0.4826,
+      "step": 4264
+    },
+    {
+      "epoch": 0.9513718492081196,
+      "grad_norm": 0.15885433554649353,
+      "learning_rate": 1.557049186006547e-05,
+      "loss": 0.4837,
+      "step": 4265
+    },
+    {
+      "epoch": 0.951594914120009,
+      "grad_norm": 0.16099439561367035,
+      "learning_rate": 1.5568537231098438e-05,
+      "loss": 0.4863,
+      "step": 4266
+    },
+    {
+      "epoch": 0.9518179790318982,
+      "grad_norm": 0.15327222645282745,
+      "learning_rate": 1.5566582293710787e-05,
+      "loss": 0.4865,
+      "step": 4267
+    },
+    {
+      "epoch": 0.9520410439437876,
+      "grad_norm": 0.1621992439031601,
+      "learning_rate": 1.5564627048010797e-05,
+      "loss": 0.5082,
+      "step": 4268
+    },
+    {
+      "epoch": 0.952264108855677,
+      "grad_norm": 0.16586080193519592,
+      "learning_rate": 1.5562671494106756e-05,
+      "loss": 0.5066,
+      "step": 4269
+    },
+    {
+      "epoch": 0.9524871737675663,
+      "grad_norm": 0.16953197121620178,
+      "learning_rate": 1.5560715632106976e-05,
+      "loss": 0.5127,
+      "step": 4270
+    },
+    {
+      "epoch": 0.9527102386794557,
+      "grad_norm": 0.1806926429271698,
+      "learning_rate": 1.555875946211979e-05,
+      "loss": 0.5091,
+      "step": 4271
+    },
+    {
+      "epoch": 0.9529333035913451,
+      "grad_norm": 0.16707123816013336,
+      "learning_rate": 1.5556802984253534e-05,
+      "loss": 0.5048,
+      "step": 4272
+    },
+    {
+      "epoch": 0.9531563685032345,
+      "grad_norm": 0.29071560502052307,
+      "learning_rate": 1.5554846198616576e-05,
+      "loss": 0.4998,
+      "step": 4273
+    },
+    {
+      "epoch": 0.9533794334151238,
+      "grad_norm": 0.19042329490184784,
+      "learning_rate": 1.5552889105317296e-05,
+      "loss": 0.4883,
+      "step": 4274
+    },
+    {
+      "epoch": 0.9536024983270132,
+      "grad_norm": 0.1610180288553238,
+      "learning_rate": 1.555093170446409e-05,
+      "loss": 0.4816,
+      "step": 4275
+    },
+    {
+      "epoch": 0.9538255632389026,
+      "grad_norm": 0.16079892218112946,
+      "learning_rate": 1.5548973996165365e-05,
+      "loss": 0.5118,
+      "step": 4276
+    },
+    {
+      "epoch": 0.9540486281507918,
+      "grad_norm": 0.1768450289964676,
+      "learning_rate": 1.5547015980529558e-05,
+      "loss": 0.5249,
+      "step": 4277
+    },
+    {
+      "epoch": 0.9542716930626812,
+      "grad_norm": 0.1556175947189331,
+      "learning_rate": 1.5545057657665115e-05,
+      "loss": 0.5081,
+      "step": 4278
+    },
+    {
+      "epoch": 0.9544947579745706,
+      "grad_norm": 0.16020707786083221,
+      "learning_rate": 1.5543099027680496e-05,
+      "loss": 0.4846,
+      "step": 4279
+    },
+    {
+      "epoch": 0.95471782288646,
+      "grad_norm": 0.1518724262714386,
+      "learning_rate": 1.554114009068419e-05,
+      "loss": 0.4866,
+      "step": 4280
+    },
+    {
+      "epoch": 0.9549408877983493,
+      "grad_norm": 0.1620427668094635,
+      "learning_rate": 1.5539180846784686e-05,
+      "loss": 0.5054,
+      "step": 4281
+    },
+    {
+      "epoch": 0.9551639527102387,
+      "grad_norm": 0.1567380130290985,
+      "learning_rate": 1.5537221296090506e-05,
+      "loss": 0.4857,
+      "step": 4282
+    },
+    {
+      "epoch": 0.9553870176221281,
+      "grad_norm": 0.1572207808494568,
+      "learning_rate": 1.553526143871018e-05,
+      "loss": 0.4803,
+      "step": 4283
+    },
+    {
+      "epoch": 0.9556100825340174,
+      "grad_norm": 0.1630953699350357,
+      "learning_rate": 1.553330127475226e-05,
+      "loss": 0.5076,
+      "step": 4284
+    },
+    {
+      "epoch": 0.9558331474459068,
+      "grad_norm": 0.16126613318920135,
+      "learning_rate": 1.5531340804325303e-05,
+      "loss": 0.4566,
+      "step": 4285
+    },
+    {
+      "epoch": 0.9560562123577961,
+      "grad_norm": 0.16429363191127777,
+      "learning_rate": 1.5529380027537904e-05,
+      "loss": 0.4905,
+      "step": 4286
+    },
+    {
+      "epoch": 0.9562792772696854,
+      "grad_norm": 0.16202637553215027,
+      "learning_rate": 1.5527418944498656e-05,
+      "loss": 0.4962,
+      "step": 4287
+    },
+    {
+      "epoch": 0.9565023421815748,
+      "grad_norm": 0.16903194785118103,
+      "learning_rate": 1.5525457555316177e-05,
+      "loss": 0.4917,
+      "step": 4288
+    },
+    {
+      "epoch": 0.9567254070934642,
+      "grad_norm": 0.16751469671726227,
+      "learning_rate": 1.5523495860099102e-05,
+      "loss": 0.4762,
+      "step": 4289
+    },
+    {
+      "epoch": 0.9569484720053536,
+      "grad_norm": 0.16918997466564178,
+      "learning_rate": 1.5521533858956085e-05,
+      "loss": 0.5005,
+      "step": 4290
+    },
+    {
+      "epoch": 0.9571715369172429,
+      "grad_norm": 0.268764466047287,
+      "learning_rate": 1.551957155199579e-05,
+      "loss": 0.4734,
+      "step": 4291
+    },
+    {
+      "epoch": 0.9573946018291323,
+      "grad_norm": 0.1583995223045349,
+      "learning_rate": 1.55176089393269e-05,
+      "loss": 0.4776,
+      "step": 4292
+    },
+    {
+      "epoch": 0.9576176667410217,
+      "grad_norm": 0.16260802745819092,
+      "learning_rate": 1.5515646021058124e-05,
+      "loss": 0.487,
+      "step": 4293
+    },
+    {
+      "epoch": 0.957840731652911,
+      "grad_norm": 0.16519910097122192,
+      "learning_rate": 1.5513682797298172e-05,
+      "loss": 0.4916,
+      "step": 4294
+    },
+    {
+      "epoch": 0.9580637965648003,
+      "grad_norm": 0.15430989861488342,
+      "learning_rate": 1.551171926815579e-05,
+      "loss": 0.4645,
+      "step": 4295
+    },
+    {
+      "epoch": 0.9582868614766897,
+      "grad_norm": 0.15664781630039215,
+      "learning_rate": 1.5509755433739723e-05,
+      "loss": 0.4824,
+      "step": 4296
+    },
+    {
+      "epoch": 0.9585099263885791,
+      "grad_norm": 0.16775622963905334,
+      "learning_rate": 1.550779129415874e-05,
+      "loss": 0.4993,
+      "step": 4297
+    },
+    {
+      "epoch": 0.9587329913004684,
+      "grad_norm": 0.163412943482399,
+      "learning_rate": 1.550582684952163e-05,
+      "loss": 0.5176,
+      "step": 4298
+    },
+    {
+      "epoch": 0.9589560562123578,
+      "grad_norm": 0.1565362513065338,
+      "learning_rate": 1.5503862099937198e-05,
+      "loss": 0.4667,
+      "step": 4299
+    },
+    {
+      "epoch": 0.9591791211242472,
+      "grad_norm": 0.16091248393058777,
+      "learning_rate": 1.550189704551426e-05,
+      "loss": 0.4823,
+      "step": 4300
+    },
+    {
+      "epoch": 0.9594021860361365,
+      "grad_norm": 0.1619873195886612,
+      "learning_rate": 1.5499931686361658e-05,
+      "loss": 0.4753,
+      "step": 4301
+    },
+    {
+      "epoch": 0.9596252509480259,
+      "grad_norm": 0.18374527990818024,
+      "learning_rate": 1.549796602258824e-05,
+      "loss": 0.465,
+      "step": 4302
+    },
+    {
+      "epoch": 0.9598483158599153,
+      "grad_norm": 0.16035234928131104,
+      "learning_rate": 1.549600005430288e-05,
+      "loss": 0.4913,
+      "step": 4303
+    },
+    {
+      "epoch": 0.9600713807718046,
+      "grad_norm": 0.16635017096996307,
+      "learning_rate": 1.549403378161447e-05,
+      "loss": 0.4869,
+      "step": 4304
+    },
+    {
+      "epoch": 0.9602944456836939,
+      "grad_norm": 0.16243912279605865,
+      "learning_rate": 1.5492067204631908e-05,
+      "loss": 0.4787,
+      "step": 4305
+    },
+    {
+      "epoch": 0.9605175105955833,
+      "grad_norm": 0.1565091758966446,
+      "learning_rate": 1.5490100323464118e-05,
+      "loss": 0.4805,
+      "step": 4306
+    },
+    {
+      "epoch": 0.9607405755074727,
+      "grad_norm": 0.1620541661977768,
+      "learning_rate": 1.5488133138220038e-05,
+      "loss": 0.5016,
+      "step": 4307
+    },
+    {
+      "epoch": 0.960963640419362,
+      "grad_norm": 0.16399456560611725,
+      "learning_rate": 1.5486165649008623e-05,
+      "loss": 0.4977,
+      "step": 4308
+    },
+    {
+      "epoch": 0.9611867053312514,
+      "grad_norm": 0.1687907576560974,
+      "learning_rate": 1.5484197855938847e-05,
+      "loss": 0.4966,
+      "step": 4309
+    },
+    {
+      "epoch": 0.9614097702431408,
+      "grad_norm": 0.16494713723659515,
+      "learning_rate": 1.548222975911969e-05,
+      "loss": 0.5044,
+      "step": 4310
+    },
+    {
+      "epoch": 0.9616328351550301,
+      "grad_norm": 0.16072864830493927,
+      "learning_rate": 1.5480261358660172e-05,
+      "loss": 0.4969,
+      "step": 4311
+    },
+    {
+      "epoch": 0.9618559000669195,
+      "grad_norm": 0.1517607718706131,
+      "learning_rate": 1.5478292654669304e-05,
+      "loss": 0.4789,
+      "step": 4312
+    },
+    {
+      "epoch": 0.9620789649788088,
+      "grad_norm": 0.15903200209140778,
+      "learning_rate": 1.547632364725613e-05,
+      "loss": 0.4877,
+      "step": 4313
+    },
+    {
+      "epoch": 0.9623020298906982,
+      "grad_norm": 0.2090669870376587,
+      "learning_rate": 1.5474354336529706e-05,
+      "loss": 0.4699,
+      "step": 4314
+    },
+    {
+      "epoch": 0.9625250948025875,
+      "grad_norm": 0.17890144884586334,
+      "learning_rate": 1.5472384722599102e-05,
+      "loss": 0.4921,
+      "step": 4315
+    },
+    {
+      "epoch": 0.9627481597144769,
+      "grad_norm": 0.1629069298505783,
+      "learning_rate": 1.547041480557341e-05,
+      "loss": 0.5049,
+      "step": 4316
+    },
+    {
+      "epoch": 0.9629712246263663,
+      "grad_norm": 0.1475018411874771,
+      "learning_rate": 1.5468444585561736e-05,
+      "loss": 0.4795,
+      "step": 4317
+    },
+    {
+      "epoch": 0.9631942895382556,
+      "grad_norm": 0.15612593293190002,
+      "learning_rate": 1.54664740626732e-05,
+      "loss": 0.4402,
+      "step": 4318
+    },
+    {
+      "epoch": 0.963417354450145,
+      "grad_norm": 0.15412123501300812,
+      "learning_rate": 1.546450323701695e-05,
+      "loss": 0.4493,
+      "step": 4319
+    },
+    {
+      "epoch": 0.9636404193620344,
+      "grad_norm": 0.16448107361793518,
+      "learning_rate": 1.5462532108702134e-05,
+      "loss": 0.4979,
+      "step": 4320
+    },
+    {
+      "epoch": 0.9638634842739238,
+      "grad_norm": 0.16529002785682678,
+      "learning_rate": 1.546056067783793e-05,
+      "loss": 0.5127,
+      "step": 4321
+    },
+    {
+      "epoch": 0.964086549185813,
+      "grad_norm": 0.15807853639125824,
+      "learning_rate": 1.545858894453353e-05,
+      "loss": 0.4638,
+      "step": 4322
+    },
+    {
+      "epoch": 0.9643096140977024,
+      "grad_norm": 0.16824275255203247,
+      "learning_rate": 1.5456616908898134e-05,
+      "loss": 0.4877,
+      "step": 4323
+    },
+    {
+      "epoch": 0.9645326790095918,
+      "grad_norm": 0.15887348353862762,
+      "learning_rate": 1.5454644571040973e-05,
+      "loss": 0.4688,
+      "step": 4324
+    },
+    {
+      "epoch": 0.9647557439214811,
+      "grad_norm": 0.16340267658233643,
+      "learning_rate": 1.545267193107128e-05,
+      "loss": 0.4779,
+      "step": 4325
+    },
+    {
+      "epoch": 0.9649788088333705,
+      "grad_norm": 0.16982871294021606,
+      "learning_rate": 1.545069898909832e-05,
+      "loss": 0.5058,
+      "step": 4326
+    },
+    {
+      "epoch": 0.9652018737452599,
+      "grad_norm": 0.15784378349781036,
+      "learning_rate": 1.544872574523137e-05,
+      "loss": 0.4563,
+      "step": 4327
+    },
+    {
+      "epoch": 0.9654249386571492,
+      "grad_norm": 0.15926995873451233,
+      "learning_rate": 1.5446752199579703e-05,
+      "loss": 0.4893,
+      "step": 4328
+    },
+    {
+      "epoch": 0.9656480035690386,
+      "grad_norm": 0.16018277406692505,
+      "learning_rate": 1.544477835225265e-05,
+      "loss": 0.4876,
+      "step": 4329
+    },
+    {
+      "epoch": 0.965871068480928,
+      "grad_norm": 0.16839507222175598,
+      "learning_rate": 1.544280420335951e-05,
+      "loss": 0.5184,
+      "step": 4330
+    },
+    {
+      "epoch": 0.9660941333928174,
+      "grad_norm": 0.1582273244857788,
+      "learning_rate": 1.5440829753009646e-05,
+      "loss": 0.4895,
+      "step": 4331
+    },
+    {
+      "epoch": 0.9663171983047066,
+      "grad_norm": 0.15812306106090546,
+      "learning_rate": 1.5438855001312402e-05,
+      "loss": 0.5244,
+      "step": 4332
+    },
+    {
+      "epoch": 0.966540263216596,
+      "grad_norm": 0.15881675481796265,
+      "learning_rate": 1.5436879948377157e-05,
+      "loss": 0.4541,
+      "step": 4333
+    },
+    {
+      "epoch": 0.9667633281284854,
+      "grad_norm": 0.1649298518896103,
+      "learning_rate": 1.5434904594313303e-05,
+      "loss": 0.4898,
+      "step": 4334
+    },
+    {
+      "epoch": 0.9669863930403747,
+      "grad_norm": 0.1670350432395935,
+      "learning_rate": 1.5432928939230243e-05,
+      "loss": 0.4961,
+      "step": 4335
+    },
+    {
+      "epoch": 0.9672094579522641,
+      "grad_norm": 0.17490597069263458,
+      "learning_rate": 1.5430952983237404e-05,
+      "loss": 0.4996,
+      "step": 4336
+    },
+    {
+      "epoch": 0.9674325228641535,
+      "grad_norm": 0.18037396669387817,
+      "learning_rate": 1.542897672644423e-05,
+      "loss": 0.4856,
+      "step": 4337
+    },
+    {
+      "epoch": 0.9676555877760429,
+      "grad_norm": 0.16085660457611084,
+      "learning_rate": 1.5427000168960172e-05,
+      "loss": 0.4741,
+      "step": 4338
+    },
+    {
+      "epoch": 0.9678786526879322,
+      "grad_norm": 0.16227804124355316,
+      "learning_rate": 1.5425023310894707e-05,
+      "loss": 0.4858,
+      "step": 4339
+    },
+    {
+      "epoch": 0.9681017175998216,
+      "grad_norm": 0.21843907237052917,
+      "learning_rate": 1.5423046152357328e-05,
+      "loss": 0.4705,
+      "step": 4340
+    },
+    {
+      "epoch": 0.968324782511711,
+      "grad_norm": 0.16902711987495422,
+      "learning_rate": 1.542106869345754e-05,
+      "loss": 0.4673,
+      "step": 4341
+    },
+    {
+      "epoch": 0.9685478474236002,
+      "grad_norm": 0.15176647901535034,
+      "learning_rate": 1.5419090934304865e-05,
+      "loss": 0.477,
+      "step": 4342
+    },
+    {
+      "epoch": 0.9687709123354896,
+      "grad_norm": 0.15946295857429504,
+      "learning_rate": 1.5417112875008854e-05,
+      "loss": 0.4768,
+      "step": 4343
+    },
+    {
+      "epoch": 0.968993977247379,
+      "grad_norm": 0.1705644428730011,
+      "learning_rate": 1.5415134515679053e-05,
+      "loss": 0.4806,
+      "step": 4344
+    },
+    {
+      "epoch": 0.9692170421592683,
+      "grad_norm": 0.1598884016275406,
+      "learning_rate": 1.541315585642504e-05,
+      "loss": 0.5018,
+      "step": 4345
+    },
+    {
+      "epoch": 0.9694401070711577,
+      "grad_norm": 0.15806621313095093,
+      "learning_rate": 1.54111768973564e-05,
+      "loss": 0.4854,
+      "step": 4346
+    },
+    {
+      "epoch": 0.9696631719830471,
+      "grad_norm": 0.15730160474777222,
+      "learning_rate": 1.5409197638582753e-05,
+      "loss": 0.499,
+      "step": 4347
+    },
+    {
+      "epoch": 0.9698862368949365,
+      "grad_norm": 0.16367729008197784,
+      "learning_rate": 1.540721808021371e-05,
+      "loss": 0.4784,
+      "step": 4348
+    },
+    {
+      "epoch": 0.9701093018068258,
+      "grad_norm": 0.16696509718894958,
+      "learning_rate": 1.5405238222358925e-05,
+      "loss": 0.4887,
+      "step": 4349
+    },
+    {
+      "epoch": 0.9703323667187151,
+      "grad_norm": 0.1621687263250351,
+      "learning_rate": 1.5403258065128042e-05,
+      "loss": 0.5113,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9705554316306045,
+      "grad_norm": 0.1656588315963745,
+      "learning_rate": 1.5401277608630742e-05,
+      "loss": 0.5021,
+      "step": 4351
+    },
+    {
+      "epoch": 0.9707784965424938,
+      "grad_norm": 0.17137381434440613,
+      "learning_rate": 1.539929685297671e-05,
+      "loss": 0.4965,
+      "step": 4352
+    },
+    {
+      "epoch": 0.9710015614543832,
+      "grad_norm": 0.16810309886932373,
+      "learning_rate": 1.5397315798275654e-05,
+      "loss": 0.4961,
+      "step": 4353
+    },
+    {
+      "epoch": 0.9712246263662726,
+      "grad_norm": 0.17059317231178284,
+      "learning_rate": 1.5395334444637306e-05,
+      "loss": 0.4708,
+      "step": 4354
+    },
+    {
+      "epoch": 0.971447691278162,
+      "grad_norm": 0.16224807500839233,
+      "learning_rate": 1.539335279217139e-05,
+      "loss": 0.4593,
+      "step": 4355
+    },
+    {
+      "epoch": 0.9716707561900513,
+      "grad_norm": 0.17209084331989288,
+      "learning_rate": 1.5391370840987674e-05,
+      "loss": 0.4959,
+      "step": 4356
+    },
+    {
+      "epoch": 0.9718938211019407,
+      "grad_norm": 0.16699501872062683,
+      "learning_rate": 1.5389388591195928e-05,
+      "loss": 0.4897,
+      "step": 4357
+    },
+    {
+      "epoch": 0.9721168860138301,
+      "grad_norm": 0.15935075283050537,
+      "learning_rate": 1.538740604290594e-05,
+      "loss": 0.4833,
+      "step": 4358
+    },
+    {
+      "epoch": 0.9723399509257193,
+      "grad_norm": 0.16760936379432678,
+      "learning_rate": 1.538542319622752e-05,
+      "loss": 0.4707,
+      "step": 4359
+    },
+    {
+      "epoch": 0.9725630158376087,
+      "grad_norm": 0.19774286448955536,
+      "learning_rate": 1.5383440051270486e-05,
+      "loss": 0.4799,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9727860807494981,
+      "grad_norm": 0.1727793663740158,
+      "learning_rate": 1.5381456608144677e-05,
+      "loss": 0.4761,
+      "step": 4361
+    },
+    {
+      "epoch": 0.9730091456613874,
+      "grad_norm": 0.1611672341823578,
+      "learning_rate": 1.5379472866959954e-05,
+      "loss": 0.4984,
+      "step": 4362
+    },
+    {
+      "epoch": 0.9732322105732768,
+      "grad_norm": 0.15909035503864288,
+      "learning_rate": 1.537748882782618e-05,
+      "loss": 0.487,
+      "step": 4363
+    },
+    {
+      "epoch": 0.9734552754851662,
+      "grad_norm": 0.17754100263118744,
+      "learning_rate": 1.5375504490853255e-05,
+      "loss": 0.4782,
+      "step": 4364
+    },
+    {
+      "epoch": 0.9736783403970556,
+      "grad_norm": 0.19527114927768707,
+      "learning_rate": 1.5373519856151077e-05,
+      "loss": 0.4783,
+      "step": 4365
+    },
+    {
+      "epoch": 0.9739014053089449,
+      "grad_norm": 0.1601899415254593,
+      "learning_rate": 1.5371534923829562e-05,
+      "loss": 0.5042,
+      "step": 4366
+    },
+    {
+      "epoch": 0.9741244702208343,
+      "grad_norm": 0.16773121058940887,
+      "learning_rate": 1.536954969399866e-05,
+      "loss": 0.5025,
+      "step": 4367
+    },
+    {
+      "epoch": 0.9743475351327237,
+      "grad_norm": 0.1618761271238327,
+      "learning_rate": 1.5367564166768322e-05,
+      "loss": 0.5125,
+      "step": 4368
+    },
+    {
+      "epoch": 0.9745706000446129,
+      "grad_norm": 0.1555018573999405,
+      "learning_rate": 1.5365578342248515e-05,
+      "loss": 0.4754,
+      "step": 4369
+    },
+    {
+      "epoch": 0.9747936649565023,
+      "grad_norm": 0.16692039370536804,
+      "learning_rate": 1.5363592220549227e-05,
+      "loss": 0.5025,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9750167298683917,
+      "grad_norm": 0.16272681951522827,
+      "learning_rate": 1.5361605801780465e-05,
+      "loss": 0.513,
+      "step": 4371
+    },
+    {
+      "epoch": 0.9752397947802811,
+      "grad_norm": 0.17153267562389374,
+      "learning_rate": 1.535961908605225e-05,
+      "loss": 0.4716,
+      "step": 4372
+    },
+    {
+      "epoch": 0.9754628596921704,
+      "grad_norm": 0.162004753947258,
+      "learning_rate": 1.5357632073474614e-05,
+      "loss": 0.4752,
+      "step": 4373
+    },
+    {
+      "epoch": 0.9756859246040598,
+      "grad_norm": 0.17653700709342957,
+      "learning_rate": 1.535564476415761e-05,
+      "loss": 0.4875,
+      "step": 4374
+    },
+    {
+      "epoch": 0.9759089895159492,
+      "grad_norm": 0.16345418989658356,
+      "learning_rate": 1.535365715821132e-05,
+      "loss": 0.5036,
+      "step": 4375
+    },
+    {
+      "epoch": 0.9761320544278385,
+      "grad_norm": 0.16799689829349518,
+      "learning_rate": 1.535166925574581e-05,
+      "loss": 0.5035,
+      "step": 4376
+    },
+    {
+      "epoch": 0.9763551193397279,
+      "grad_norm": 0.204716295003891,
+      "learning_rate": 1.53496810568712e-05,
+      "loss": 0.501,
+      "step": 4377
+    },
+    {
+      "epoch": 0.9765781842516172,
+      "grad_norm": 0.15521782636642456,
+      "learning_rate": 1.53476925616976e-05,
+      "loss": 0.4864,
+      "step": 4378
+    },
+    {
+      "epoch": 0.9768012491635066,
+      "grad_norm": 0.16439463198184967,
+      "learning_rate": 1.5345703770335147e-05,
+      "loss": 0.491,
+      "step": 4379
+    },
+    {
+      "epoch": 0.9770243140753959,
+      "grad_norm": 0.16255317628383636,
+      "learning_rate": 1.5343714682893997e-05,
+      "loss": 0.5108,
+      "step": 4380
+    },
+    {
+      "epoch": 0.9772473789872853,
+      "grad_norm": 0.15555396676063538,
+      "learning_rate": 1.534172529948431e-05,
+      "loss": 0.4506,
+      "step": 4381
+    },
+    {
+      "epoch": 0.9774704438991747,
+      "grad_norm": 0.16245706379413605,
+      "learning_rate": 1.5339735620216275e-05,
+      "loss": 0.4892,
+      "step": 4382
+    },
+    {
+      "epoch": 0.977693508811064,
+      "grad_norm": 0.16190017759799957,
+      "learning_rate": 1.5337745645200097e-05,
+      "loss": 0.4716,
+      "step": 4383
+    },
+    {
+      "epoch": 0.9779165737229534,
+      "grad_norm": 0.1614699810743332,
+      "learning_rate": 1.5335755374545985e-05,
+      "loss": 0.5041,
+      "step": 4384
+    },
+    {
+      "epoch": 0.9781396386348428,
+      "grad_norm": 0.1543981432914734,
+      "learning_rate": 1.533376480836418e-05,
+      "loss": 0.4837,
+      "step": 4385
+    },
+    {
+      "epoch": 0.978362703546732,
+      "grad_norm": 0.15510593354701996,
+      "learning_rate": 1.5331773946764928e-05,
+      "loss": 0.5073,
+      "step": 4386
+    },
+    {
+      "epoch": 0.9785857684586214,
+      "grad_norm": 0.16036750376224518,
+      "learning_rate": 1.5329782789858495e-05,
+      "loss": 0.511,
+      "step": 4387
+    },
+    {
+      "epoch": 0.9788088333705108,
+      "grad_norm": 0.1586884707212448,
+      "learning_rate": 1.532779133775517e-05,
+      "loss": 0.4783,
+      "step": 4388
+    },
+    {
+      "epoch": 0.9790318982824002,
+      "grad_norm": 0.1595693677663803,
+      "learning_rate": 1.5325799590565247e-05,
+      "loss": 0.4847,
+      "step": 4389
+    },
+    {
+      "epoch": 0.9792549631942895,
+      "grad_norm": 0.1676463484764099,
+      "learning_rate": 1.532380754839904e-05,
+      "loss": 0.4857,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9794780281061789,
+      "grad_norm": 0.1569330245256424,
+      "learning_rate": 1.532181521136688e-05,
+      "loss": 0.5038,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9797010930180683,
+      "grad_norm": 0.15425747632980347,
+      "learning_rate": 1.5319822579579125e-05,
+      "loss": 0.4687,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9799241579299576,
+      "grad_norm": 0.1658424586057663,
+      "learning_rate": 1.5317829653146127e-05,
+      "loss": 0.4727,
+      "step": 4393
+    },
+    {
+      "epoch": 0.980147222841847,
+      "grad_norm": 0.16950277984142303,
+      "learning_rate": 1.5315836432178275e-05,
+      "loss": 0.5056,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9803702877537364,
+      "grad_norm": 0.1796022653579712,
+      "learning_rate": 1.5313842916785965e-05,
+      "loss": 0.5062,
+      "step": 4395
+    },
+    {
+      "epoch": 0.9805933526656258,
+      "grad_norm": 0.16499702632427216,
+      "learning_rate": 1.5311849107079603e-05,
+      "loss": 0.5142,
+      "step": 4396
+    },
+    {
+      "epoch": 0.980816417577515,
+      "grad_norm": 0.1709393560886383,
+      "learning_rate": 1.5309855003169632e-05,
+      "loss": 0.5001,
+      "step": 4397
+    },
+    {
+      "epoch": 0.9810394824894044,
+      "grad_norm": 0.16077296435832977,
+      "learning_rate": 1.5307860605166487e-05,
+      "loss": 0.4908,
+      "step": 4398
+    },
+    {
+      "epoch": 0.9812625474012938,
+      "grad_norm": 0.17582206428050995,
+      "learning_rate": 1.5305865913180633e-05,
+      "loss": 0.4844,
+      "step": 4399
+    },
+    {
+      "epoch": 0.9814856123131831,
+      "grad_norm": 0.16168825328350067,
+      "learning_rate": 1.5303870927322552e-05,
+      "loss": 0.4856,
+      "step": 4400
+    },
+    {
+      "epoch": 0.9817086772250725,
+      "grad_norm": 0.15929754078388214,
+      "learning_rate": 1.5301875647702732e-05,
+      "loss": 0.4878,
+      "step": 4401
+    },
+    {
+      "epoch": 0.9819317421369619,
+      "grad_norm": 0.16380812227725983,
+      "learning_rate": 1.5299880074431693e-05,
+      "loss": 0.5009,
+      "step": 4402
+    },
+    {
+      "epoch": 0.9821548070488512,
+      "grad_norm": 0.17140789330005646,
+      "learning_rate": 1.5297884207619957e-05,
+      "loss": 0.4604,
+      "step": 4403
+    },
+    {
+      "epoch": 0.9823778719607406,
+      "grad_norm": 0.17358756065368652,
+      "learning_rate": 1.5295888047378064e-05,
+      "loss": 0.4905,
+      "step": 4404
+    },
+    {
+      "epoch": 0.98260093687263,
+      "grad_norm": 0.16201473772525787,
+      "learning_rate": 1.5293891593816583e-05,
+      "loss": 0.5156,
+      "step": 4405
+    },
+    {
+      "epoch": 0.9828240017845193,
+      "grad_norm": 0.16091175377368927,
+      "learning_rate": 1.529189484704608e-05,
+      "loss": 0.4883,
+      "step": 4406
+    },
+    {
+      "epoch": 0.9830470666964086,
+      "grad_norm": 0.16278314590454102,
+      "learning_rate": 1.528989780717716e-05,
+      "loss": 0.4727,
+      "step": 4407
+    },
+    {
+      "epoch": 0.983270131608298,
+      "grad_norm": 0.17446744441986084,
+      "learning_rate": 1.5287900474320422e-05,
+      "loss": 0.4824,
+      "step": 4408
+    },
+    {
+      "epoch": 0.9834931965201874,
+      "grad_norm": 0.3287087678909302,
+      "learning_rate": 1.5285902848586495e-05,
+      "loss": 0.49,
+      "step": 4409
+    },
+    {
+      "epoch": 0.9837162614320767,
+      "grad_norm": 0.16612185537815094,
+      "learning_rate": 1.5283904930086017e-05,
+      "loss": 0.4911,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9839393263439661,
+      "grad_norm": 0.16220030188560486,
+      "learning_rate": 1.528190671892965e-05,
+      "loss": 0.4537,
+      "step": 4411
+    },
+    {
+      "epoch": 0.9841623912558555,
+      "grad_norm": 0.1631203591823578,
+      "learning_rate": 1.5279908215228058e-05,
+      "loss": 0.4725,
+      "step": 4412
+    },
+    {
+      "epoch": 0.9843854561677449,
+      "grad_norm": 0.15897974371910095,
+      "learning_rate": 1.5277909419091942e-05,
+      "loss": 0.4794,
+      "step": 4413
+    },
+    {
+      "epoch": 0.9846085210796341,
+      "grad_norm": 0.18077795207500458,
+      "learning_rate": 1.5275910330632e-05,
+      "loss": 0.4764,
+      "step": 4414
+    },
+    {
+      "epoch": 0.9848315859915235,
+      "grad_norm": 0.1682848483324051,
+      "learning_rate": 1.5273910949958963e-05,
+      "loss": 0.5024,
+      "step": 4415
+    },
+    {
+      "epoch": 0.9850546509034129,
+      "grad_norm": 0.1584632247686386,
+      "learning_rate": 1.527191127718356e-05,
+      "loss": 0.5007,
+      "step": 4416
+    },
+    {
+      "epoch": 0.9852777158153022,
+      "grad_norm": 0.1735336184501648,
+      "learning_rate": 1.5269911312416547e-05,
+      "loss": 0.4887,
+      "step": 4417
+    },
+    {
+      "epoch": 0.9855007807271916,
+      "grad_norm": 0.16184300184249878,
+      "learning_rate": 1.5267911055768697e-05,
+      "loss": 0.5064,
+      "step": 4418
+    },
+    {
+      "epoch": 0.985723845639081,
+      "grad_norm": 0.15283328294754028,
+      "learning_rate": 1.5265910507350797e-05,
+      "loss": 0.5012,
+      "step": 4419
+    },
+    {
+      "epoch": 0.9859469105509703,
+      "grad_norm": 0.1493072211742401,
+      "learning_rate": 1.526390966727365e-05,
+      "loss": 0.4763,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9861699754628597,
+      "grad_norm": 0.15723635256290436,
+      "learning_rate": 1.526190853564807e-05,
+      "loss": 0.4893,
+      "step": 4421
+    },
+    {
+      "epoch": 0.9863930403747491,
+      "grad_norm": 0.17012642323970795,
+      "learning_rate": 1.52599071125849e-05,
+      "loss": 0.4905,
+      "step": 4422
+    },
+    {
+      "epoch": 0.9866161052866385,
+      "grad_norm": 0.15649664402008057,
+      "learning_rate": 1.5257905398194988e-05,
+      "loss": 0.4841,
+      "step": 4423
+    },
+    {
+      "epoch": 0.9868391701985277,
+      "grad_norm": 0.1696402132511139,
+      "learning_rate": 1.5255903392589204e-05,
+      "loss": 0.4792,
+      "step": 4424
+    },
+    {
+      "epoch": 0.9870622351104171,
+      "grad_norm": 0.17853881418704987,
+      "learning_rate": 1.5253901095878423e-05,
+      "loss": 0.5112,
+      "step": 4425
+    },
+    {
+      "epoch": 0.9872853000223065,
+      "grad_norm": 0.1598142683506012,
+      "learning_rate": 1.5251898508173558e-05,
+      "loss": 0.4907,
+      "step": 4426
+    },
+    {
+      "epoch": 0.9875083649341958,
+      "grad_norm": 0.1565473973751068,
+      "learning_rate": 1.5249895629585511e-05,
+      "loss": 0.4922,
+      "step": 4427
+    },
+    {
+      "epoch": 0.9877314298460852,
+      "grad_norm": 0.16049052774906158,
+      "learning_rate": 1.5247892460225226e-05,
+      "loss": 0.4947,
+      "step": 4428
+    },
+    {
+      "epoch": 0.9879544947579746,
+      "grad_norm": 0.16537557542324066,
+      "learning_rate": 1.5245889000203644e-05,
+      "loss": 0.4547,
+      "step": 4429
+    },
+    {
+      "epoch": 0.988177559669864,
+      "grad_norm": 0.16105584800243378,
+      "learning_rate": 1.5243885249631732e-05,
+      "loss": 0.4697,
+      "step": 4430
+    },
+    {
+      "epoch": 0.9884006245817533,
+      "grad_norm": 0.5197194814682007,
+      "learning_rate": 1.5241881208620468e-05,
+      "loss": 0.4723,
+      "step": 4431
+    },
+    {
+      "epoch": 0.9886236894936427,
+      "grad_norm": 0.15987588465213776,
+      "learning_rate": 1.5239876877280852e-05,
+      "loss": 0.5388,
+      "step": 4432
+    },
+    {
+      "epoch": 0.988846754405532,
+      "grad_norm": 0.16223685443401337,
+      "learning_rate": 1.5237872255723894e-05,
+      "loss": 0.5185,
+      "step": 4433
+    },
+    {
+      "epoch": 0.9890698193174213,
+      "grad_norm": 0.16801123321056366,
+      "learning_rate": 1.5235867344060622e-05,
+      "loss": 0.4825,
+      "step": 4434
+    },
+    {
+      "epoch": 0.9892928842293107,
+      "grad_norm": 0.16588987410068512,
+      "learning_rate": 1.523386214240208e-05,
+      "loss": 0.4933,
+      "step": 4435
+    },
+    {
+      "epoch": 0.9895159491412001,
+      "grad_norm": 0.1642870157957077,
+      "learning_rate": 1.5231856650859334e-05,
+      "loss": 0.4572,
+      "step": 4436
+    },
+    {
+      "epoch": 0.9897390140530894,
+      "grad_norm": 0.1569276601076126,
+      "learning_rate": 1.5229850869543454e-05,
+      "loss": 0.4844,
+      "step": 4437
+    },
+    {
+      "epoch": 0.9899620789649788,
+      "grad_norm": 0.16232363879680634,
+      "learning_rate": 1.5227844798565538e-05,
+      "loss": 0.5041,
+      "step": 4438
+    },
+    {
+      "epoch": 0.9901851438768682,
+      "grad_norm": 0.18178559839725494,
+      "learning_rate": 1.5225838438036693e-05,
+      "loss": 0.5203,
+      "step": 4439
+    },
+    {
+      "epoch": 0.9904082087887576,
+      "grad_norm": 0.1483062505722046,
+      "learning_rate": 1.5223831788068039e-05,
+      "loss": 0.4601,
+      "step": 4440
+    },
+    {
+      "epoch": 0.9906312737006469,
+      "grad_norm": 0.16904570162296295,
+      "learning_rate": 1.5221824848770728e-05,
+      "loss": 0.4692,
+      "step": 4441
+    },
+    {
+      "epoch": 0.9908543386125362,
+      "grad_norm": 0.16170957684516907,
+      "learning_rate": 1.5219817620255906e-05,
+      "loss": 0.476,
+      "step": 4442
+    },
+    {
+      "epoch": 0.9910774035244256,
+      "grad_norm": 0.15581966936588287,
+      "learning_rate": 1.521781010263475e-05,
+      "loss": 0.4685,
+      "step": 4443
+    },
+    {
+      "epoch": 0.9913004684363149,
+      "grad_norm": 0.17382118105888367,
+      "learning_rate": 1.521580229601845e-05,
+      "loss": 0.511,
+      "step": 4444
+    },
+    {
+      "epoch": 0.9915235333482043,
+      "grad_norm": 0.16696421802043915,
+      "learning_rate": 1.521379420051821e-05,
+      "loss": 0.5072,
+      "step": 4445
+    },
+    {
+      "epoch": 0.9917465982600937,
+      "grad_norm": 0.16701167821884155,
+      "learning_rate": 1.521178581624525e-05,
+      "loss": 0.521,
+      "step": 4446
+    },
+    {
+      "epoch": 0.9919696631719831,
+      "grad_norm": 0.16650201380252838,
+      "learning_rate": 1.520977714331081e-05,
+      "loss": 0.4945,
+      "step": 4447
+    },
+    {
+      "epoch": 0.9921927280838724,
+      "grad_norm": 0.1575930416584015,
+      "learning_rate": 1.5207768181826138e-05,
+      "loss": 0.4668,
+      "step": 4448
+    },
+    {
+      "epoch": 0.9924157929957618,
+      "grad_norm": 0.16714414954185486,
+      "learning_rate": 1.5205758931902507e-05,
+      "loss": 0.4739,
+      "step": 4449
+    },
+    {
+      "epoch": 0.9926388579076512,
+      "grad_norm": 0.1591663658618927,
+      "learning_rate": 1.5203749393651204e-05,
+      "loss": 0.5024,
+      "step": 4450
+    },
+    {
+      "epoch": 0.9928619228195404,
+      "grad_norm": 0.16330404579639435,
+      "learning_rate": 1.5201739567183525e-05,
+      "loss": 0.5104,
+      "step": 4451
+    },
+    {
+      "epoch": 0.9930849877314298,
+      "grad_norm": 0.16072112321853638,
+      "learning_rate": 1.519972945261079e-05,
+      "loss": 0.4722,
+      "step": 4452
+    },
+    {
+      "epoch": 0.9933080526433192,
+      "grad_norm": 0.1594439148902893,
+      "learning_rate": 1.5197719050044328e-05,
+      "loss": 0.4951,
+      "step": 4453
+    },
+    {
+      "epoch": 0.9935311175552086,
+      "grad_norm": 0.17731733620166779,
+      "learning_rate": 1.519570835959549e-05,
+      "loss": 0.4799,
+      "step": 4454
+    },
+    {
+      "epoch": 0.9937541824670979,
+      "grad_norm": 0.1658770889043808,
+      "learning_rate": 1.5193697381375641e-05,
+      "loss": 0.471,
+      "step": 4455
+    },
+    {
+      "epoch": 0.9939772473789873,
+      "grad_norm": 0.168108731508255,
+      "learning_rate": 1.5191686115496163e-05,
+      "loss": 0.4856,
+      "step": 4456
+    },
+    {
+      "epoch": 0.9942003122908767,
+      "grad_norm": 0.1631343513727188,
+      "learning_rate": 1.5189674562068448e-05,
+      "loss": 0.4795,
+      "step": 4457
+    },
+    {
+      "epoch": 0.994423377202766,
+      "grad_norm": 0.17591005563735962,
+      "learning_rate": 1.5187662721203916e-05,
+      "loss": 0.4431,
+      "step": 4458
+    },
+    {
+      "epoch": 0.9946464421146554,
+      "grad_norm": 0.16236771643161774,
+      "learning_rate": 1.5185650593013984e-05,
+      "loss": 0.5097,
+      "step": 4459
+    },
+    {
+      "epoch": 0.9948695070265448,
+      "grad_norm": 0.15711717307567596,
+      "learning_rate": 1.5183638177610109e-05,
+      "loss": 0.4896,
+      "step": 4460
+    },
+    {
+      "epoch": 0.995092571938434,
+      "grad_norm": 0.16357070207595825,
+      "learning_rate": 1.5181625475103744e-05,
+      "loss": 0.4878,
+      "step": 4461
+    },
+    {
+      "epoch": 0.9953156368503234,
+      "grad_norm": 0.1608823835849762,
+      "learning_rate": 1.5179612485606366e-05,
+      "loss": 0.4924,
+      "step": 4462
+    },
+    {
+      "epoch": 0.9955387017622128,
+      "grad_norm": 0.1610344648361206,
+      "learning_rate": 1.5177599209229468e-05,
+      "loss": 0.4564,
+      "step": 4463
+    },
+    {
+      "epoch": 0.9957617666741022,
+      "grad_norm": 0.1570533961057663,
+      "learning_rate": 1.5175585646084557e-05,
+      "loss": 0.4665,
+      "step": 4464
+    },
+    {
+      "epoch": 0.9959848315859915,
+      "grad_norm": 0.16878454387187958,
+      "learning_rate": 1.5173571796283155e-05,
+      "loss": 0.4832,
+      "step": 4465
+    },
+    {
+      "epoch": 0.9962078964978809,
+      "grad_norm": 0.15934909880161285,
+      "learning_rate": 1.5171557659936806e-05,
+      "loss": 0.4629,
+      "step": 4466
+    },
+    {
+      "epoch": 0.9964309614097703,
+      "grad_norm": 0.17097942531108856,
+      "learning_rate": 1.5169543237157062e-05,
+      "loss": 0.5009,
+      "step": 4467
+    },
+    {
+      "epoch": 0.9966540263216596,
+      "grad_norm": 0.1563706398010254,
+      "learning_rate": 1.5167528528055498e-05,
+      "loss": 0.4864,
+      "step": 4468
+    },
+    {
+      "epoch": 0.996877091233549,
+      "grad_norm": 0.16761906445026398,
+      "learning_rate": 1.5165513532743696e-05,
+      "loss": 0.4688,
+      "step": 4469
+    },
+    {
+      "epoch": 0.9971001561454383,
+      "grad_norm": 0.16788896918296814,
+      "learning_rate": 1.5163498251333267e-05,
+      "loss": 0.4975,
+      "step": 4470
+    },
+    {
+      "epoch": 0.9973232210573277,
+      "grad_norm": 0.16318921744823456,
+      "learning_rate": 1.516148268393582e-05,
+      "loss": 0.483,
+      "step": 4471
+    },
+    {
+      "epoch": 0.997546285969217,
+      "grad_norm": 0.1533585637807846,
+      "learning_rate": 1.5159466830662997e-05,
+      "loss": 0.4888,
+      "step": 4472
+    },
+    {
+      "epoch": 0.9977693508811064,
+      "grad_norm": 0.16081759333610535,
+      "learning_rate": 1.515745069162645e-05,
+      "loss": 0.4997,
+      "step": 4473
+    },
+    {
+      "epoch": 0.9979924157929958,
+      "grad_norm": 0.17346571385860443,
+      "learning_rate": 1.5155434266937836e-05,
+      "loss": 0.4918,
+      "step": 4474
+    },
+    {
+      "epoch": 0.9982154807048851,
+      "grad_norm": 0.16051939129829407,
+      "learning_rate": 1.515341755670885e-05,
+      "loss": 0.4911,
+      "step": 4475
+    },
+    {
+      "epoch": 0.9984385456167745,
+      "grad_norm": 0.15489095449447632,
+      "learning_rate": 1.5151400561051177e-05,
+      "loss": 0.4573,
+      "step": 4476
+    },
+    {
+      "epoch": 0.9986616105286639,
+      "grad_norm": 0.17051677405834198,
+      "learning_rate": 1.5149383280076544e-05,
+      "loss": 0.47,
+      "step": 4477
+    },
+    {
+      "epoch": 0.9988846754405531,
+      "grad_norm": 0.16898037493228912,
+      "learning_rate": 1.5147365713896669e-05,
+      "loss": 0.5265,
+      "step": 4478
+    },
+    {
+      "epoch": 0.9991077403524425,
+      "grad_norm": 0.1665852814912796,
+      "learning_rate": 1.5145347862623303e-05,
+      "loss": 0.4967,
+      "step": 4479
+    },
+    {
+      "epoch": 0.9993308052643319,
+      "grad_norm": 0.16627469658851624,
+      "learning_rate": 1.5143329726368205e-05,
+      "loss": 0.4755,
+      "step": 4480
+    },
+    {
+      "epoch": 0.9995538701762213,
+      "grad_norm": 0.15166249871253967,
+      "learning_rate": 1.5141311305243158e-05,
+      "loss": 0.4762,
+      "step": 4481
+    },
+    {
+      "epoch": 0.9997769350881106,
+      "grad_norm": 0.15809613466262817,
+      "learning_rate": 1.513929259935995e-05,
+      "loss": 0.4825,
+      "step": 4482
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.24375846982002258,
+      "learning_rate": 1.5137273608830387e-05,
+      "loss": 0.4902,
+      "step": 4483
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 0.3229251801967621,
+      "eval_runtime": 666.2343,
+      "eval_samples_per_second": 94.726,
+      "eval_steps_per_second": 1.481,
+      "step": 4483
+    },
+    {
+      "epoch": 1.0002230649118893,
+      "grad_norm": 0.2046872079372406,
+      "learning_rate": 1.5135254333766302e-05,
+      "loss": 0.5093,
+      "step": 4484
+    },
+    {
+      "epoch": 1.0004461298237788,
+      "grad_norm": 0.16689886152744293,
+      "learning_rate": 1.5133234774279526e-05,
+      "loss": 0.5118,
+      "step": 4485
+    },
+    {
+      "epoch": 1.000669194735668,
+      "grad_norm": 0.16698423027992249,
+      "learning_rate": 1.5131214930481922e-05,
+      "loss": 0.4917,
+      "step": 4486
+    },
+    {
+      "epoch": 1.0008922596475573,
+      "grad_norm": 0.17248550057411194,
+      "learning_rate": 1.5129194802485354e-05,
+      "loss": 0.4693,
+      "step": 4487
+    },
+    {
+      "epoch": 1.0011153245594469,
+      "grad_norm": 0.16011400520801544,
+      "learning_rate": 1.5127174390401717e-05,
+      "loss": 0.4788,
+      "step": 4488
+    },
+    {
+      "epoch": 1.0013383894713361,
+      "grad_norm": 0.16476082801818848,
+      "learning_rate": 1.512515369434291e-05,
+      "loss": 0.4723,
+      "step": 4489
+    },
+    {
+      "epoch": 1.0015614543832254,
+      "grad_norm": 0.1672191321849823,
+      "learning_rate": 1.5123132714420856e-05,
+      "loss": 0.484,
+      "step": 4490
+    },
+    {
+      "epoch": 1.001784519295115,
+      "grad_norm": 0.15264827013015747,
+      "learning_rate": 1.5121111450747483e-05,
+      "loss": 0.4577,
+      "step": 4491
+    },
+    {
+      "epoch": 1.0020075842070042,
+      "grad_norm": 0.15963797271251678,
+      "learning_rate": 1.5119089903434746e-05,
+      "loss": 0.4866,
+      "step": 4492
+    },
+    {
+      "epoch": 1.0022306491188937,
+      "grad_norm": 0.16446684300899506,
+      "learning_rate": 1.511706807259461e-05,
+      "loss": 0.4976,
+      "step": 4493
+    },
+    {
+      "epoch": 1.002453714030783,
+      "grad_norm": 0.1584995537996292,
+      "learning_rate": 1.5115045958339056e-05,
+      "loss": 0.4921,
+      "step": 4494
+    },
+    {
+      "epoch": 1.0026767789426723,
+      "grad_norm": 0.16608034074306488,
+      "learning_rate": 1.5113023560780083e-05,
+      "loss": 0.4946,
+      "step": 4495
+    },
+    {
+      "epoch": 1.0028998438545618,
+      "grad_norm": 0.15869465470314026,
+      "learning_rate": 1.5111000880029703e-05,
+      "loss": 0.5022,
+      "step": 4496
+    },
+    {
+      "epoch": 1.003122908766451,
+      "grad_norm": 0.16816754639148712,
+      "learning_rate": 1.5108977916199941e-05,
+      "loss": 0.4791,
+      "step": 4497
+    },
+    {
+      "epoch": 1.0033459736783403,
+      "grad_norm": 0.17300903797149658,
+      "learning_rate": 1.5106954669402849e-05,
+      "loss": 0.4817,
+      "step": 4498
+    },
+    {
+      "epoch": 1.0035690385902298,
+      "grad_norm": 0.17333939671516418,
+      "learning_rate": 1.510493113975048e-05,
+      "loss": 0.4671,
+      "step": 4499
+    },
+    {
+      "epoch": 1.0037921035021191,
+      "grad_norm": 0.2043878585100174,
+      "learning_rate": 1.5102907327354919e-05,
+      "loss": 0.5001,
+      "step": 4500
+    },
+    {
+      "epoch": 1.0040151684140084,
+      "grad_norm": 0.15702618658542633,
+      "learning_rate": 1.5100883232328247e-05,
+      "loss": 0.5172,
+      "step": 4501
+    },
+    {
+      "epoch": 1.004238233325898,
+      "grad_norm": 0.1576346904039383,
+      "learning_rate": 1.5098858854782576e-05,
+      "loss": 0.4813,
+      "step": 4502
+    },
+    {
+      "epoch": 1.0044612982377872,
+      "grad_norm": 0.171005517244339,
+      "learning_rate": 1.509683419483003e-05,
+      "loss": 0.5054,
+      "step": 4503
+    },
+    {
+      "epoch": 1.0046843631496765,
+      "grad_norm": 0.16706731915473938,
+      "learning_rate": 1.5094809252582744e-05,
+      "loss": 0.4791,
+      "step": 4504
+    },
+    {
+      "epoch": 1.004907428061566,
+      "grad_norm": 0.1601940393447876,
+      "learning_rate": 1.5092784028152878e-05,
+      "loss": 0.4788,
+      "step": 4505
+    },
+    {
+      "epoch": 1.0051304929734552,
+      "grad_norm": 0.1597285270690918,
+      "learning_rate": 1.5090758521652596e-05,
+      "loss": 0.4669,
+      "step": 4506
+    },
+    {
+      "epoch": 1.0053535578853445,
+      "grad_norm": 0.16775481402873993,
+      "learning_rate": 1.5088732733194085e-05,
+      "loss": 0.5091,
+      "step": 4507
+    },
+    {
+      "epoch": 1.005576622797234,
+      "grad_norm": 0.16889511048793793,
+      "learning_rate": 1.5086706662889544e-05,
+      "loss": 0.5208,
+      "step": 4508
+    },
+    {
+      "epoch": 1.0057996877091233,
+      "grad_norm": 0.1600525975227356,
+      "learning_rate": 1.5084680310851192e-05,
+      "loss": 0.4861,
+      "step": 4509
+    },
+    {
+      "epoch": 1.0060227526210128,
+      "grad_norm": 0.18192386627197266,
+      "learning_rate": 1.508265367719126e-05,
+      "loss": 0.486,
+      "step": 4510
+    },
+    {
+      "epoch": 1.006245817532902,
+      "grad_norm": 0.1748298555612564,
+      "learning_rate": 1.5080626762021997e-05,
+      "loss": 0.4623,
+      "step": 4511
+    },
+    {
+      "epoch": 1.0064688824447914,
+      "grad_norm": 0.1619860827922821,
+      "learning_rate": 1.5078599565455666e-05,
+      "loss": 0.4707,
+      "step": 4512
+    },
+    {
+      "epoch": 1.0066919473566809,
+      "grad_norm": 0.16041533648967743,
+      "learning_rate": 1.5076572087604544e-05,
+      "loss": 0.4663,
+      "step": 4513
+    },
+    {
+      "epoch": 1.0069150122685702,
+      "grad_norm": 0.1654667854309082,
+      "learning_rate": 1.5074544328580928e-05,
+      "loss": 0.4746,
+      "step": 4514
+    },
+    {
+      "epoch": 1.0071380771804594,
+      "grad_norm": 0.1631074994802475,
+      "learning_rate": 1.5072516288497127e-05,
+      "loss": 0.508,
+      "step": 4515
+    },
+    {
+      "epoch": 1.007361142092349,
+      "grad_norm": 0.16078722476959229,
+      "learning_rate": 1.5070487967465466e-05,
+      "loss": 0.5053,
+      "step": 4516
+    },
+    {
+      "epoch": 1.0075842070042382,
+      "grad_norm": 0.15384818613529205,
+      "learning_rate": 1.5068459365598286e-05,
+      "loss": 0.4896,
+      "step": 4517
+    },
+    {
+      "epoch": 1.0078072719161275,
+      "grad_norm": 0.15592217445373535,
+      "learning_rate": 1.5066430483007949e-05,
+      "loss": 0.4911,
+      "step": 4518
+    },
+    {
+      "epoch": 1.008030336828017,
+      "grad_norm": 0.1650250107049942,
+      "learning_rate": 1.5064401319806817e-05,
+      "loss": 0.4914,
+      "step": 4519
+    },
+    {
+      "epoch": 1.0082534017399063,
+      "grad_norm": 0.16215115785598755,
+      "learning_rate": 1.5062371876107286e-05,
+      "loss": 0.468,
+      "step": 4520
+    },
+    {
+      "epoch": 1.0084764666517956,
+      "grad_norm": 0.1609305888414383,
+      "learning_rate": 1.5060342152021757e-05,
+      "loss": 0.5136,
+      "step": 4521
+    },
+    {
+      "epoch": 1.008699531563685,
+      "grad_norm": 0.15757685899734497,
+      "learning_rate": 1.505831214766265e-05,
+      "loss": 0.4816,
+      "step": 4522
+    },
+    {
+      "epoch": 1.0089225964755744,
+      "grad_norm": 0.15898853540420532,
+      "learning_rate": 1.5056281863142394e-05,
+      "loss": 0.4704,
+      "step": 4523
+    },
+    {
+      "epoch": 1.0091456613874636,
+      "grad_norm": 0.16032522916793823,
+      "learning_rate": 1.5054251298573447e-05,
+      "loss": 0.4712,
+      "step": 4524
+    },
+    {
+      "epoch": 1.0093687262993531,
+      "grad_norm": 0.1658545881509781,
+      "learning_rate": 1.5052220454068267e-05,
+      "loss": 0.4975,
+      "step": 4525
+    },
+    {
+      "epoch": 1.0095917912112424,
+      "grad_norm": 0.15979249775409698,
+      "learning_rate": 1.5050189329739344e-05,
+      "loss": 0.4768,
+      "step": 4526
+    },
+    {
+      "epoch": 1.009814856123132,
+      "grad_norm": 0.16390205919742584,
+      "learning_rate": 1.5048157925699162e-05,
+      "loss": 0.5083,
+      "step": 4527
+    },
+    {
+      "epoch": 1.0100379210350212,
+      "grad_norm": 0.16372700035572052,
+      "learning_rate": 1.5046126242060247e-05,
+      "loss": 0.4919,
+      "step": 4528
+    },
+    {
+      "epoch": 1.0102609859469105,
+      "grad_norm": 0.16318677365779877,
+      "learning_rate": 1.5044094278935115e-05,
+      "loss": 0.4736,
+      "step": 4529
+    },
+    {
+      "epoch": 1.0104840508588,
+      "grad_norm": 0.1591407060623169,
+      "learning_rate": 1.5042062036436315e-05,
+      "loss": 0.4911,
+      "step": 4530
+    },
+    {
+      "epoch": 1.0107071157706893,
+      "grad_norm": 0.16406004130840302,
+      "learning_rate": 1.5040029514676402e-05,
+      "loss": 0.4897,
+      "step": 4531
+    },
+    {
+      "epoch": 1.0109301806825786,
+      "grad_norm": 0.15959376096725464,
+      "learning_rate": 1.5037996713767956e-05,
+      "loss": 0.4638,
+      "step": 4532
+    },
+    {
+      "epoch": 1.011153245594468,
+      "grad_norm": 0.15884444117546082,
+      "learning_rate": 1.5035963633823553e-05,
+      "loss": 0.4431,
+      "step": 4533
+    },
+    {
+      "epoch": 1.0113763105063573,
+      "grad_norm": 0.15193885564804077,
+      "learning_rate": 1.5033930274955813e-05,
+      "loss": 0.4828,
+      "step": 4534
+    },
+    {
+      "epoch": 1.0115993754182466,
+      "grad_norm": 0.16269952058792114,
+      "learning_rate": 1.503189663727735e-05,
+      "loss": 0.4977,
+      "step": 4535
+    },
+    {
+      "epoch": 1.0118224403301361,
+      "grad_norm": 0.1688687652349472,
+      "learning_rate": 1.5029862720900794e-05,
+      "loss": 0.4798,
+      "step": 4536
+    },
+    {
+      "epoch": 1.0120455052420254,
+      "grad_norm": 0.1625138372182846,
+      "learning_rate": 1.5027828525938809e-05,
+      "loss": 0.4732,
+      "step": 4537
+    },
+    {
+      "epoch": 1.0122685701539147,
+      "grad_norm": 0.162174254655838,
+      "learning_rate": 1.5025794052504048e-05,
+      "loss": 0.4912,
+      "step": 4538
+    },
+    {
+      "epoch": 1.0124916350658042,
+      "grad_norm": 0.1722794622182846,
+      "learning_rate": 1.5023759300709201e-05,
+      "loss": 0.5365,
+      "step": 4539
+    },
+    {
+      "epoch": 1.0127146999776935,
+      "grad_norm": 0.18794307112693787,
+      "learning_rate": 1.5021724270666962e-05,
+      "loss": 0.4874,
+      "step": 4540
+    },
+    {
+      "epoch": 1.0129377648895828,
+      "grad_norm": 0.16088199615478516,
+      "learning_rate": 1.5019688962490047e-05,
+      "loss": 0.4658,
+      "step": 4541
+    },
+    {
+      "epoch": 1.0131608298014723,
+      "grad_norm": 0.1688738763332367,
+      "learning_rate": 1.501765337629118e-05,
+      "loss": 0.4869,
+      "step": 4542
+    },
+    {
+      "epoch": 1.0133838947133615,
+      "grad_norm": 0.2059110552072525,
+      "learning_rate": 1.5015617512183109e-05,
+      "loss": 0.4952,
+      "step": 4543
+    },
+    {
+      "epoch": 1.013606959625251,
+      "grad_norm": 0.1552867442369461,
+      "learning_rate": 1.5013581370278587e-05,
+      "loss": 0.4761,
+      "step": 4544
+    },
+    {
+      "epoch": 1.0138300245371403,
+      "grad_norm": 0.15534162521362305,
+      "learning_rate": 1.5011544950690397e-05,
+      "loss": 0.4751,
+      "step": 4545
+    },
+    {
+      "epoch": 1.0140530894490296,
+      "grad_norm": 0.162929505109787,
+      "learning_rate": 1.5009508253531321e-05,
+      "loss": 0.5241,
+      "step": 4546
+    },
+    {
+      "epoch": 1.0142761543609191,
+      "grad_norm": 0.16057923436164856,
+      "learning_rate": 1.5007471278914167e-05,
+      "loss": 0.4944,
+      "step": 4547
+    },
+    {
+      "epoch": 1.0144992192728084,
+      "grad_norm": 0.18289563059806824,
+      "learning_rate": 1.5005434026951755e-05,
+      "loss": 0.4852,
+      "step": 4548
+    },
+    {
+      "epoch": 1.0147222841846977,
+      "grad_norm": 0.1580429971218109,
+      "learning_rate": 1.5003396497756923e-05,
+      "loss": 0.4776,
+      "step": 4549
+    },
+    {
+      "epoch": 1.0149453490965872,
+      "grad_norm": 0.1565399467945099,
+      "learning_rate": 1.5001358691442517e-05,
+      "loss": 0.4785,
+      "step": 4550
+    },
+    {
+      "epoch": 1.0151684140084765,
+      "grad_norm": 0.16028526425361633,
+      "learning_rate": 1.4999320608121411e-05,
+      "loss": 0.4797,
+      "step": 4551
+    },
+    {
+      "epoch": 1.0153914789203657,
+      "grad_norm": 0.16666477918624878,
+      "learning_rate": 1.499728224790648e-05,
+      "loss": 0.5112,
+      "step": 4552
+    },
+    {
+      "epoch": 1.0156145438322552,
+      "grad_norm": 0.164764866232872,
+      "learning_rate": 1.4995243610910625e-05,
+      "loss": 0.4843,
+      "step": 4553
+    },
+    {
+      "epoch": 1.0158376087441445,
+      "grad_norm": 0.16802887618541718,
+      "learning_rate": 1.4993204697246758e-05,
+      "loss": 0.4642,
+      "step": 4554
+    },
+    {
+      "epoch": 1.0160606736560338,
+      "grad_norm": 0.16424837708473206,
+      "learning_rate": 1.4991165507027802e-05,
+      "loss": 0.471,
+      "step": 4555
+    },
+    {
+      "epoch": 1.0162837385679233,
+      "grad_norm": 0.16172267496585846,
+      "learning_rate": 1.498912604036671e-05,
+      "loss": 0.4747,
+      "step": 4556
+    },
+    {
+      "epoch": 1.0165068034798126,
+      "grad_norm": 0.1546371579170227,
+      "learning_rate": 1.4987086297376431e-05,
+      "loss": 0.4624,
+      "step": 4557
+    },
+    {
+      "epoch": 1.0167298683917019,
+      "grad_norm": 0.15725232660770416,
+      "learning_rate": 1.4985046278169949e-05,
+      "loss": 0.4799,
+      "step": 4558
+    },
+    {
+      "epoch": 1.0169529333035914,
+      "grad_norm": 0.1614525318145752,
+      "learning_rate": 1.4983005982860241e-05,
+      "loss": 0.4652,
+      "step": 4559
+    },
+    {
+      "epoch": 1.0171759982154807,
+      "grad_norm": 0.1634308397769928,
+      "learning_rate": 1.498096541156032e-05,
+      "loss": 0.4742,
+      "step": 4560
+    },
+    {
+      "epoch": 1.0173990631273702,
+      "grad_norm": 0.16403678059577942,
+      "learning_rate": 1.4978924564383202e-05,
+      "loss": 0.4956,
+      "step": 4561
+    },
+    {
+      "epoch": 1.0176221280392594,
+      "grad_norm": 0.16014313697814941,
+      "learning_rate": 1.4976883441441924e-05,
+      "loss": 0.4956,
+      "step": 4562
+    },
+    {
+      "epoch": 1.0178451929511487,
+      "grad_norm": 0.16341127455234528,
+      "learning_rate": 1.4974842042849533e-05,
+      "loss": 0.5086,
+      "step": 4563
+    },
+    {
+      "epoch": 1.0180682578630382,
+      "grad_norm": 0.16314633190631866,
+      "learning_rate": 1.4972800368719098e-05,
+      "loss": 0.4681,
+      "step": 4564
+    },
+    {
+      "epoch": 1.0182913227749275,
+      "grad_norm": 0.16120873391628265,
+      "learning_rate": 1.4970758419163698e-05,
+      "loss": 0.4839,
+      "step": 4565
+    },
+    {
+      "epoch": 1.0185143876868168,
+      "grad_norm": 0.15808890759944916,
+      "learning_rate": 1.4968716194296429e-05,
+      "loss": 0.4711,
+      "step": 4566
+    },
+    {
+      "epoch": 1.0187374525987063,
+      "grad_norm": 0.1576492190361023,
+      "learning_rate": 1.4966673694230406e-05,
+      "loss": 0.4767,
+      "step": 4567
+    },
+    {
+      "epoch": 1.0189605175105956,
+      "grad_norm": 0.17056535184383392,
+      "learning_rate": 1.4964630919078747e-05,
+      "loss": 0.4551,
+      "step": 4568
+    },
+    {
+      "epoch": 1.0191835824224849,
+      "grad_norm": 0.15599535405635834,
+      "learning_rate": 1.4962587868954603e-05,
+      "loss": 0.462,
+      "step": 4569
+    },
+    {
+      "epoch": 1.0194066473343744,
+      "grad_norm": 0.16496673226356506,
+      "learning_rate": 1.4960544543971125e-05,
+      "loss": 0.4903,
+      "step": 4570
+    },
+    {
+      "epoch": 1.0196297122462636,
+      "grad_norm": 0.16336064040660858,
+      "learning_rate": 1.4958500944241488e-05,
+      "loss": 0.4804,
+      "step": 4571
+    },
+    {
+      "epoch": 1.019852777158153,
+      "grad_norm": 0.1574566662311554,
+      "learning_rate": 1.4956457069878875e-05,
+      "loss": 0.4405,
+      "step": 4572
+    },
+    {
+      "epoch": 1.0200758420700424,
+      "grad_norm": 0.1587173491716385,
+      "learning_rate": 1.4954412920996498e-05,
+      "loss": 0.4729,
+      "step": 4573
+    },
+    {
+      "epoch": 1.0202989069819317,
+      "grad_norm": 0.16297802329063416,
+      "learning_rate": 1.4952368497707566e-05,
+      "loss": 0.475,
+      "step": 4574
+    },
+    {
+      "epoch": 1.020521971893821,
+      "grad_norm": 0.16890546679496765,
+      "learning_rate": 1.4950323800125314e-05,
+      "loss": 0.4715,
+      "step": 4575
+    },
+    {
+      "epoch": 1.0207450368057105,
+      "grad_norm": 0.16996879875659943,
+      "learning_rate": 1.4948278828362991e-05,
+      "loss": 0.4772,
+      "step": 4576
+    },
+    {
+      "epoch": 1.0209681017175998,
+      "grad_norm": 0.16053706407546997,
+      "learning_rate": 1.4946233582533865e-05,
+      "loss": 0.4763,
+      "step": 4577
+    },
+    {
+      "epoch": 1.0211911666294893,
+      "grad_norm": 0.15968559682369232,
+      "learning_rate": 1.4944188062751207e-05,
+      "loss": 0.4593,
+      "step": 4578
+    },
+    {
+      "epoch": 1.0214142315413786,
+      "grad_norm": 0.15733575820922852,
+      "learning_rate": 1.4942142269128312e-05,
+      "loss": 0.462,
+      "step": 4579
+    },
+    {
+      "epoch": 1.0216372964532678,
+      "grad_norm": 0.1631694734096527,
+      "learning_rate": 1.4940096201778493e-05,
+      "loss": 0.4845,
+      "step": 4580
+    },
+    {
+      "epoch": 1.0218603613651573,
+      "grad_norm": 0.15469685196876526,
+      "learning_rate": 1.4938049860815072e-05,
+      "loss": 0.4726,
+      "step": 4581
+    },
+    {
+      "epoch": 1.0220834262770466,
+      "grad_norm": 0.16130223870277405,
+      "learning_rate": 1.4936003246351386e-05,
+      "loss": 0.4966,
+      "step": 4582
+    },
+    {
+      "epoch": 1.022306491188936,
+      "grad_norm": 0.19383056461811066,
+      "learning_rate": 1.4933956358500794e-05,
+      "loss": 0.5019,
+      "step": 4583
+    },
+    {
+      "epoch": 1.0225295561008254,
+      "grad_norm": 0.16322484612464905,
+      "learning_rate": 1.4931909197376664e-05,
+      "loss": 0.4344,
+      "step": 4584
+    },
+    {
+      "epoch": 1.0227526210127147,
+      "grad_norm": 0.15319214761257172,
+      "learning_rate": 1.4929861763092378e-05,
+      "loss": 0.475,
+      "step": 4585
+    },
+    {
+      "epoch": 1.022975685924604,
+      "grad_norm": 0.16148288547992706,
+      "learning_rate": 1.4927814055761336e-05,
+      "loss": 0.5031,
+      "step": 4586
+    },
+    {
+      "epoch": 1.0231987508364935,
+      "grad_norm": 0.17277774214744568,
+      "learning_rate": 1.4925766075496955e-05,
+      "loss": 0.5255,
+      "step": 4587
+    },
+    {
+      "epoch": 1.0234218157483828,
+      "grad_norm": 0.16420382261276245,
+      "learning_rate": 1.4923717822412666e-05,
+      "loss": 0.4811,
+      "step": 4588
+    },
+    {
+      "epoch": 1.023644880660272,
+      "grad_norm": 0.16079513728618622,
+      "learning_rate": 1.4921669296621912e-05,
+      "loss": 0.4742,
+      "step": 4589
+    },
+    {
+      "epoch": 1.0238679455721615,
+      "grad_norm": 0.1617536097764969,
+      "learning_rate": 1.4919620498238154e-05,
+      "loss": 0.4818,
+      "step": 4590
+    },
+    {
+      "epoch": 1.0240910104840508,
+      "grad_norm": 0.15949992835521698,
+      "learning_rate": 1.4917571427374866e-05,
+      "loss": 0.4697,
+      "step": 4591
+    },
+    {
+      "epoch": 1.0243140753959403,
+      "grad_norm": 0.1637139767408371,
+      "learning_rate": 1.4915522084145542e-05,
+      "loss": 0.4507,
+      "step": 4592
+    },
+    {
+      "epoch": 1.0245371403078296,
+      "grad_norm": 0.16350442171096802,
+      "learning_rate": 1.4913472468663681e-05,
+      "loss": 0.4676,
+      "step": 4593
+    },
+    {
+      "epoch": 1.024760205219719,
+      "grad_norm": 0.1604386270046234,
+      "learning_rate": 1.4911422581042812e-05,
+      "loss": 0.4941,
+      "step": 4594
+    },
+    {
+      "epoch": 1.0249832701316084,
+      "grad_norm": 0.16599829494953156,
+      "learning_rate": 1.4909372421396464e-05,
+      "loss": 0.486,
+      "step": 4595
+    },
+    {
+      "epoch": 1.0252063350434977,
+      "grad_norm": 0.24195000529289246,
+      "learning_rate": 1.4907321989838192e-05,
+      "loss": 0.4824,
+      "step": 4596
+    },
+    {
+      "epoch": 1.025429399955387,
+      "grad_norm": 0.1621350198984146,
+      "learning_rate": 1.4905271286481557e-05,
+      "loss": 0.4602,
+      "step": 4597
+    },
+    {
+      "epoch": 1.0256524648672765,
+      "grad_norm": 0.16584086418151855,
+      "learning_rate": 1.4903220311440147e-05,
+      "loss": 0.4581,
+      "step": 4598
+    },
+    {
+      "epoch": 1.0258755297791657,
+      "grad_norm": 0.16808323562145233,
+      "learning_rate": 1.4901169064827552e-05,
+      "loss": 0.4756,
+      "step": 4599
+    },
+    {
+      "epoch": 1.026098594691055,
+      "grad_norm": 0.15748214721679688,
+      "learning_rate": 1.4899117546757383e-05,
+      "loss": 0.4366,
+      "step": 4600
+    },
+    {
+      "epoch": 1.0263216596029445,
+      "grad_norm": 0.18983392417430878,
+      "learning_rate": 1.489706575734327e-05,
+      "loss": 0.4447,
+      "step": 4601
+    },
+    {
+      "epoch": 1.0265447245148338,
+      "grad_norm": 0.16951000690460205,
+      "learning_rate": 1.4895013696698847e-05,
+      "loss": 0.4608,
+      "step": 4602
+    },
+    {
+      "epoch": 1.026767789426723,
+      "grad_norm": 0.1630699634552002,
+      "learning_rate": 1.4892961364937779e-05,
+      "loss": 0.4773,
+      "step": 4603
+    },
+    {
+      "epoch": 1.0269908543386126,
+      "grad_norm": 0.15806427597999573,
+      "learning_rate": 1.4890908762173731e-05,
+      "loss": 0.4579,
+      "step": 4604
+    },
+    {
+      "epoch": 1.0272139192505019,
+      "grad_norm": 0.16593055427074432,
+      "learning_rate": 1.4888855888520393e-05,
+      "loss": 0.4622,
+      "step": 4605
+    },
+    {
+      "epoch": 1.0274369841623912,
+      "grad_norm": 0.16501964628696442,
+      "learning_rate": 1.488680274409146e-05,
+      "loss": 0.4793,
+      "step": 4606
+    },
+    {
+      "epoch": 1.0276600490742807,
+      "grad_norm": 0.1590343415737152,
+      "learning_rate": 1.4884749329000654e-05,
+      "loss": 0.4562,
+      "step": 4607
+    },
+    {
+      "epoch": 1.02788311398617,
+      "grad_norm": 0.1620016247034073,
+      "learning_rate": 1.4882695643361702e-05,
+      "loss": 0.4836,
+      "step": 4608
+    },
+    {
+      "epoch": 1.0281061788980594,
+      "grad_norm": 0.1608029156923294,
+      "learning_rate": 1.4880641687288356e-05,
+      "loss": 0.4819,
+      "step": 4609
+    },
+    {
+      "epoch": 1.0283292438099487,
+      "grad_norm": 0.18065601587295532,
+      "learning_rate": 1.4878587460894369e-05,
+      "loss": 0.4966,
+      "step": 4610
+    },
+    {
+      "epoch": 1.028552308721838,
+      "grad_norm": 0.16342510282993317,
+      "learning_rate": 1.4876532964293522e-05,
+      "loss": 0.504,
+      "step": 4611
+    },
+    {
+      "epoch": 1.0287753736337275,
+      "grad_norm": 0.16731272637844086,
+      "learning_rate": 1.4874478197599605e-05,
+      "loss": 0.4517,
+      "step": 4612
+    },
+    {
+      "epoch": 1.0289984385456168,
+      "grad_norm": 0.17315874993801117,
+      "learning_rate": 1.4872423160926424e-05,
+      "loss": 0.4704,
+      "step": 4613
+    },
+    {
+      "epoch": 1.029221503457506,
+      "grad_norm": 0.16523095965385437,
+      "learning_rate": 1.48703678543878e-05,
+      "loss": 0.4943,
+      "step": 4614
+    },
+    {
+      "epoch": 1.0294445683693956,
+      "grad_norm": 0.1586511880159378,
+      "learning_rate": 1.4868312278097568e-05,
+      "loss": 0.486,
+      "step": 4615
+    },
+    {
+      "epoch": 1.0296676332812849,
+      "grad_norm": 0.167024627327919,
+      "learning_rate": 1.4866256432169577e-05,
+      "loss": 0.4898,
+      "step": 4616
+    },
+    {
+      "epoch": 1.0298906981931741,
+      "grad_norm": 0.16019749641418457,
+      "learning_rate": 1.4864200316717698e-05,
+      "loss": 0.4738,
+      "step": 4617
+    },
+    {
+      "epoch": 1.0301137631050636,
+      "grad_norm": 0.16836729645729065,
+      "learning_rate": 1.4862143931855806e-05,
+      "loss": 0.4597,
+      "step": 4618
+    },
+    {
+      "epoch": 1.030336828016953,
+      "grad_norm": 0.1572437733411789,
+      "learning_rate": 1.4860087277697798e-05,
+      "loss": 0.495,
+      "step": 4619
+    },
+    {
+      "epoch": 1.0305598929288422,
+      "grad_norm": 0.1620069444179535,
+      "learning_rate": 1.4858030354357588e-05,
+      "loss": 0.4681,
+      "step": 4620
+    },
+    {
+      "epoch": 1.0307829578407317,
+      "grad_norm": 0.1601821482181549,
+      "learning_rate": 1.4855973161949097e-05,
+      "loss": 0.4803,
+      "step": 4621
+    },
+    {
+      "epoch": 1.031006022752621,
+      "grad_norm": 0.16902416944503784,
+      "learning_rate": 1.485391570058627e-05,
+      "loss": 0.479,
+      "step": 4622
+    },
+    {
+      "epoch": 1.0312290876645103,
+      "grad_norm": 0.15522709488868713,
+      "learning_rate": 1.4851857970383057e-05,
+      "loss": 0.4737,
+      "step": 4623
+    },
+    {
+      "epoch": 1.0314521525763998,
+      "grad_norm": 0.15457487106323242,
+      "learning_rate": 1.484979997145343e-05,
+      "loss": 0.4663,
+      "step": 4624
+    },
+    {
+      "epoch": 1.031675217488289,
+      "grad_norm": 0.16563165187835693,
+      "learning_rate": 1.4847741703911376e-05,
+      "loss": 0.4754,
+      "step": 4625
+    },
+    {
+      "epoch": 1.0318982824001786,
+      "grad_norm": 0.16526126861572266,
+      "learning_rate": 1.4845683167870891e-05,
+      "loss": 0.4743,
+      "step": 4626
+    },
+    {
+      "epoch": 1.0321213473120678,
+      "grad_norm": 0.1640041321516037,
+      "learning_rate": 1.4843624363445992e-05,
+      "loss": 0.4995,
+      "step": 4627
+    },
+    {
+      "epoch": 1.0323444122239571,
+      "grad_norm": 0.17919553816318512,
+      "learning_rate": 1.484156529075071e-05,
+      "loss": 0.4622,
+      "step": 4628
+    },
+    {
+      "epoch": 1.0325674771358466,
+      "grad_norm": 0.1707811802625656,
+      "learning_rate": 1.4839505949899084e-05,
+      "loss": 0.5257,
+      "step": 4629
+    },
+    {
+      "epoch": 1.032790542047736,
+      "grad_norm": 0.16861692070960999,
+      "learning_rate": 1.4837446341005179e-05,
+      "loss": 0.4661,
+      "step": 4630
+    },
+    {
+      "epoch": 1.0330136069596252,
+      "grad_norm": 0.15601199865341187,
+      "learning_rate": 1.4835386464183066e-05,
+      "loss": 0.4704,
+      "step": 4631
+    },
+    {
+      "epoch": 1.0332366718715147,
+      "grad_norm": 0.16105563938617706,
+      "learning_rate": 1.4833326319546837e-05,
+      "loss": 0.4822,
+      "step": 4632
+    },
+    {
+      "epoch": 1.033459736783404,
+      "grad_norm": 0.1516101360321045,
+      "learning_rate": 1.483126590721059e-05,
+      "loss": 0.4812,
+      "step": 4633
+    },
+    {
+      "epoch": 1.0336828016952933,
+      "grad_norm": 0.16206029057502747,
+      "learning_rate": 1.4829205227288451e-05,
+      "loss": 0.4827,
+      "step": 4634
+    },
+    {
+      "epoch": 1.0339058666071828,
+      "grad_norm": 0.16983060538768768,
+      "learning_rate": 1.4827144279894547e-05,
+      "loss": 0.4507,
+      "step": 4635
+    },
+    {
+      "epoch": 1.034128931519072,
+      "grad_norm": 0.17058612406253815,
+      "learning_rate": 1.4825083065143029e-05,
+      "loss": 0.4606,
+      "step": 4636
+    },
+    {
+      "epoch": 1.0343519964309613,
+      "grad_norm": 0.15379010140895844,
+      "learning_rate": 1.482302158314806e-05,
+      "loss": 0.4365,
+      "step": 4637
+    },
+    {
+      "epoch": 1.0345750613428508,
+      "grad_norm": 0.1577761024236679,
+      "learning_rate": 1.4820959834023821e-05,
+      "loss": 0.4678,
+      "step": 4638
+    },
+    {
+      "epoch": 1.03479812625474,
+      "grad_norm": 0.16535353660583496,
+      "learning_rate": 1.4818897817884499e-05,
+      "loss": 0.4994,
+      "step": 4639
+    },
+    {
+      "epoch": 1.0350211911666294,
+      "grad_norm": 0.16018177568912506,
+      "learning_rate": 1.48168355348443e-05,
+      "loss": 0.486,
+      "step": 4640
+    },
+    {
+      "epoch": 1.0352442560785189,
+      "grad_norm": 0.16746951639652252,
+      "learning_rate": 1.4814772985017456e-05,
+      "loss": 0.4665,
+      "step": 4641
+    },
+    {
+      "epoch": 1.0354673209904082,
+      "grad_norm": 0.1544388234615326,
+      "learning_rate": 1.4812710168518195e-05,
+      "loss": 0.4707,
+      "step": 4642
+    },
+    {
+      "epoch": 1.0356903859022977,
+      "grad_norm": 0.15881603956222534,
+      "learning_rate": 1.4810647085460771e-05,
+      "loss": 0.4727,
+      "step": 4643
+    },
+    {
+      "epoch": 1.035913450814187,
+      "grad_norm": 0.15742497146129608,
+      "learning_rate": 1.4808583735959453e-05,
+      "loss": 0.4864,
+      "step": 4644
+    },
+    {
+      "epoch": 1.0361365157260762,
+      "grad_norm": 0.15278148651123047,
+      "learning_rate": 1.4806520120128518e-05,
+      "loss": 0.4586,
+      "step": 4645
+    },
+    {
+      "epoch": 1.0363595806379657,
+      "grad_norm": 0.16195416450500488,
+      "learning_rate": 1.4804456238082266e-05,
+      "loss": 0.4532,
+      "step": 4646
+    },
+    {
+      "epoch": 1.036582645549855,
+      "grad_norm": 0.17061328887939453,
+      "learning_rate": 1.480239208993501e-05,
+      "loss": 0.4568,
+      "step": 4647
+    },
+    {
+      "epoch": 1.0368057104617443,
+      "grad_norm": 0.16437166929244995,
+      "learning_rate": 1.4800327675801065e-05,
+      "loss": 0.4751,
+      "step": 4648
+    },
+    {
+      "epoch": 1.0370287753736338,
+      "grad_norm": 0.15695516765117645,
+      "learning_rate": 1.4798262995794784e-05,
+      "loss": 0.4701,
+      "step": 4649
+    },
+    {
+      "epoch": 1.037251840285523,
+      "grad_norm": 0.16812454164028168,
+      "learning_rate": 1.479619805003051e-05,
+      "loss": 0.4543,
+      "step": 4650
+    },
+    {
+      "epoch": 1.0374749051974124,
+      "grad_norm": 0.16520899534225464,
+      "learning_rate": 1.4794132838622624e-05,
+      "loss": 0.4576,
+      "step": 4651
+    },
+    {
+      "epoch": 1.0376979701093019,
+      "grad_norm": 0.155088871717453,
+      "learning_rate": 1.4792067361685503e-05,
+      "loss": 0.4775,
+      "step": 4652
+    },
+    {
+      "epoch": 1.0379210350211912,
+      "grad_norm": 0.15887859463691711,
+      "learning_rate": 1.4790001619333547e-05,
+      "loss": 0.4642,
+      "step": 4653
+    },
+    {
+      "epoch": 1.0381440999330804,
+      "grad_norm": 0.17838329076766968,
+      "learning_rate": 1.4787935611681174e-05,
+      "loss": 0.4821,
+      "step": 4654
+    },
+    {
+      "epoch": 1.03836716484497,
+      "grad_norm": 0.15603458881378174,
+      "learning_rate": 1.4785869338842807e-05,
+      "loss": 0.4581,
+      "step": 4655
+    },
+    {
+      "epoch": 1.0385902297568592,
+      "grad_norm": 0.16940052807331085,
+      "learning_rate": 1.4783802800932894e-05,
+      "loss": 0.5005,
+      "step": 4656
+    },
+    {
+      "epoch": 1.0388132946687485,
+      "grad_norm": 0.1598014533519745,
+      "learning_rate": 1.4781735998065886e-05,
+      "loss": 0.4866,
+      "step": 4657
+    },
+    {
+      "epoch": 1.039036359580638,
+      "grad_norm": 0.1550016701221466,
+      "learning_rate": 1.4779668930356265e-05,
+      "loss": 0.4531,
+      "step": 4658
+    },
+    {
+      "epoch": 1.0392594244925273,
+      "grad_norm": 0.17037729918956757,
+      "learning_rate": 1.4777601597918511e-05,
+      "loss": 0.4867,
+      "step": 4659
+    },
+    {
+      "epoch": 1.0394824894044168,
+      "grad_norm": 0.15960454940795898,
+      "learning_rate": 1.477553400086713e-05,
+      "loss": 0.4768,
+      "step": 4660
+    },
+    {
+      "epoch": 1.039705554316306,
+      "grad_norm": 0.17182455956935883,
+      "learning_rate": 1.4773466139316634e-05,
+      "loss": 0.4833,
+      "step": 4661
+    },
+    {
+      "epoch": 1.0399286192281954,
+      "grad_norm": 0.1634978950023651,
+      "learning_rate": 1.4771398013381559e-05,
+      "loss": 0.4591,
+      "step": 4662
+    },
+    {
+      "epoch": 1.0401516841400849,
+      "grad_norm": 0.1541091501712799,
+      "learning_rate": 1.476932962317645e-05,
+      "loss": 0.4611,
+      "step": 4663
+    },
+    {
+      "epoch": 1.0403747490519741,
+      "grad_norm": 0.16876503825187683,
+      "learning_rate": 1.4767260968815864e-05,
+      "loss": 0.497,
+      "step": 4664
+    },
+    {
+      "epoch": 1.0405978139638634,
+      "grad_norm": 0.16533160209655762,
+      "learning_rate": 1.4765192050414378e-05,
+      "loss": 0.4818,
+      "step": 4665
+    },
+    {
+      "epoch": 1.040820878875753,
+      "grad_norm": 0.1572985202074051,
+      "learning_rate": 1.4763122868086584e-05,
+      "loss": 0.4748,
+      "step": 4666
+    },
+    {
+      "epoch": 1.0410439437876422,
+      "grad_norm": 0.16402465105056763,
+      "learning_rate": 1.4761053421947082e-05,
+      "loss": 0.4819,
+      "step": 4667
+    },
+    {
+      "epoch": 1.0412670086995315,
+      "grad_norm": 0.15840837359428406,
+      "learning_rate": 1.4758983712110494e-05,
+      "loss": 0.4781,
+      "step": 4668
+    },
+    {
+      "epoch": 1.041490073611421,
+      "grad_norm": 0.1562443971633911,
+      "learning_rate": 1.4756913738691451e-05,
+      "loss": 0.4784,
+      "step": 4669
+    },
+    {
+      "epoch": 1.0417131385233103,
+      "grad_norm": 0.1598626971244812,
+      "learning_rate": 1.4754843501804607e-05,
+      "loss": 0.4846,
+      "step": 4670
+    },
+    {
+      "epoch": 1.0419362034351995,
+      "grad_norm": 0.1586054414510727,
+      "learning_rate": 1.4752773001564617e-05,
+      "loss": 0.4819,
+      "step": 4671
+    },
+    {
+      "epoch": 1.042159268347089,
+      "grad_norm": 0.16802629828453064,
+      "learning_rate": 1.4750702238086164e-05,
+      "loss": 0.4851,
+      "step": 4672
+    },
+    {
+      "epoch": 1.0423823332589783,
+      "grad_norm": 0.16435806453227997,
+      "learning_rate": 1.474863121148394e-05,
+      "loss": 0.4617,
+      "step": 4673
+    },
+    {
+      "epoch": 1.0426053981708678,
+      "grad_norm": 0.1635981649160385,
+      "learning_rate": 1.4746559921872645e-05,
+      "loss": 0.4708,
+      "step": 4674
+    },
+    {
+      "epoch": 1.0428284630827571,
+      "grad_norm": 0.1575925201177597,
+      "learning_rate": 1.4744488369367007e-05,
+      "loss": 0.4662,
+      "step": 4675
+    },
+    {
+      "epoch": 1.0430515279946464,
+      "grad_norm": 0.16302283108234406,
+      "learning_rate": 1.474241655408176e-05,
+      "loss": 0.4632,
+      "step": 4676
+    },
+    {
+      "epoch": 1.043274592906536,
+      "grad_norm": 0.1619601547718048,
+      "learning_rate": 1.4740344476131652e-05,
+      "loss": 0.4489,
+      "step": 4677
+    },
+    {
+      "epoch": 1.0434976578184252,
+      "grad_norm": 0.16791771352291107,
+      "learning_rate": 1.4738272135631448e-05,
+      "loss": 0.5023,
+      "step": 4678
+    },
+    {
+      "epoch": 1.0437207227303145,
+      "grad_norm": 0.16237159073352814,
+      "learning_rate": 1.4736199532695929e-05,
+      "loss": 0.4603,
+      "step": 4679
+    },
+    {
+      "epoch": 1.043943787642204,
+      "grad_norm": 0.1556091606616974,
+      "learning_rate": 1.473412666743989e-05,
+      "loss": 0.4466,
+      "step": 4680
+    },
+    {
+      "epoch": 1.0441668525540932,
+      "grad_norm": 0.15769942104816437,
+      "learning_rate": 1.4732053539978138e-05,
+      "loss": 0.4532,
+      "step": 4681
+    },
+    {
+      "epoch": 1.0443899174659825,
+      "grad_norm": 0.15660084784030914,
+      "learning_rate": 1.472998015042549e-05,
+      "loss": 0.4862,
+      "step": 4682
+    },
+    {
+      "epoch": 1.044612982377872,
+      "grad_norm": 0.15516437590122223,
+      "learning_rate": 1.4727906498896793e-05,
+      "loss": 0.4756,
+      "step": 4683
+    },
+    {
+      "epoch": 1.0448360472897613,
+      "grad_norm": 0.16473138332366943,
+      "learning_rate": 1.4725832585506891e-05,
+      "loss": 0.4937,
+      "step": 4684
+    },
+    {
+      "epoch": 1.0450591122016506,
+      "grad_norm": 0.15583591163158417,
+      "learning_rate": 1.4723758410370654e-05,
+      "loss": 0.4732,
+      "step": 4685
+    },
+    {
+      "epoch": 1.04528217711354,
+      "grad_norm": 0.15941470861434937,
+      "learning_rate": 1.4721683973602965e-05,
+      "loss": 0.4457,
+      "step": 4686
+    },
+    {
+      "epoch": 1.0455052420254294,
+      "grad_norm": 0.1659325659275055,
+      "learning_rate": 1.4719609275318715e-05,
+      "loss": 0.4989,
+      "step": 4687
+    },
+    {
+      "epoch": 1.0457283069373187,
+      "grad_norm": 0.1708594560623169,
+      "learning_rate": 1.4717534315632817e-05,
+      "loss": 0.4833,
+      "step": 4688
+    },
+    {
+      "epoch": 1.0459513718492082,
+      "grad_norm": 0.1607704609632492,
+      "learning_rate": 1.4715459094660194e-05,
+      "loss": 0.4419,
+      "step": 4689
+    },
+    {
+      "epoch": 1.0461744367610974,
+      "grad_norm": 0.16482210159301758,
+      "learning_rate": 1.4713383612515786e-05,
+      "loss": 0.4686,
+      "step": 4690
+    },
+    {
+      "epoch": 1.046397501672987,
+      "grad_norm": 0.16981108486652374,
+      "learning_rate": 1.4711307869314544e-05,
+      "loss": 0.4961,
+      "step": 4691
+    },
+    {
+      "epoch": 1.0466205665848762,
+      "grad_norm": 0.16742509603500366,
+      "learning_rate": 1.4709231865171436e-05,
+      "loss": 0.4972,
+      "step": 4692
+    },
+    {
+      "epoch": 1.0468436314967655,
+      "grad_norm": 0.16613337397575378,
+      "learning_rate": 1.4707155600201447e-05,
+      "loss": 0.4803,
+      "step": 4693
+    },
+    {
+      "epoch": 1.047066696408655,
+      "grad_norm": 0.18546496331691742,
+      "learning_rate": 1.470507907451957e-05,
+      "loss": 0.4678,
+      "step": 4694
+    },
+    {
+      "epoch": 1.0472897613205443,
+      "grad_norm": 0.1629878729581833,
+      "learning_rate": 1.4703002288240818e-05,
+      "loss": 0.4479,
+      "step": 4695
+    },
+    {
+      "epoch": 1.0475128262324336,
+      "grad_norm": 0.1560726761817932,
+      "learning_rate": 1.4700925241480217e-05,
+      "loss": 0.451,
+      "step": 4696
+    },
+    {
+      "epoch": 1.047735891144323,
+      "grad_norm": 0.458696573972702,
+      "learning_rate": 1.4698847934352804e-05,
+      "loss": 0.493,
+      "step": 4697
+    },
+    {
+      "epoch": 1.0479589560562124,
+      "grad_norm": 0.1607166975736618,
+      "learning_rate": 1.4696770366973639e-05,
+      "loss": 0.4688,
+      "step": 4698
+    },
+    {
+      "epoch": 1.0481820209681016,
+      "grad_norm": 0.15131288766860962,
+      "learning_rate": 1.4694692539457784e-05,
+      "loss": 0.4551,
+      "step": 4699
+    },
+    {
+      "epoch": 1.0484050858799911,
+      "grad_norm": 0.16862881183624268,
+      "learning_rate": 1.4692614451920328e-05,
+      "loss": 0.4579,
+      "step": 4700
+    },
+    {
+      "epoch": 1.0486281507918804,
+      "grad_norm": 0.16511371731758118,
+      "learning_rate": 1.4690536104476364e-05,
+      "loss": 0.4887,
+      "step": 4701
+    },
+    {
+      "epoch": 1.0488512157037697,
+      "grad_norm": 0.16970515251159668,
+      "learning_rate": 1.4688457497241006e-05,
+      "loss": 0.4494,
+      "step": 4702
+    },
+    {
+      "epoch": 1.0490742806156592,
+      "grad_norm": 0.16722868382930756,
+      "learning_rate": 1.4686378630329382e-05,
+      "loss": 0.4781,
+      "step": 4703
+    },
+    {
+      "epoch": 1.0492973455275485,
+      "grad_norm": 0.16638034582138062,
+      "learning_rate": 1.4684299503856627e-05,
+      "loss": 0.4705,
+      "step": 4704
+    },
+    {
+      "epoch": 1.0495204104394378,
+      "grad_norm": 0.18224425613880157,
+      "learning_rate": 1.4682220117937904e-05,
+      "loss": 0.4902,
+      "step": 4705
+    },
+    {
+      "epoch": 1.0497434753513273,
+      "grad_norm": 0.15631026029586792,
+      "learning_rate": 1.4680140472688376e-05,
+      "loss": 0.448,
+      "step": 4706
+    },
+    {
+      "epoch": 1.0499665402632166,
+      "grad_norm": 0.15947161614894867,
+      "learning_rate": 1.4678060568223232e-05,
+      "loss": 0.453,
+      "step": 4707
+    },
+    {
+      "epoch": 1.050189605175106,
+      "grad_norm": 0.15899132192134857,
+      "learning_rate": 1.4675980404657666e-05,
+      "loss": 0.4594,
+      "step": 4708
+    },
+    {
+      "epoch": 1.0504126700869953,
+      "grad_norm": 0.1654270440340042,
+      "learning_rate": 1.4673899982106892e-05,
+      "loss": 0.482,
+      "step": 4709
+    },
+    {
+      "epoch": 1.0506357349988846,
+      "grad_norm": 0.16609270870685577,
+      "learning_rate": 1.4671819300686136e-05,
+      "loss": 0.5132,
+      "step": 4710
+    },
+    {
+      "epoch": 1.0508587999107741,
+      "grad_norm": 0.15867236256599426,
+      "learning_rate": 1.4669738360510643e-05,
+      "loss": 0.4499,
+      "step": 4711
+    },
+    {
+      "epoch": 1.0510818648226634,
+      "grad_norm": 0.1547602266073227,
+      "learning_rate": 1.4667657161695663e-05,
+      "loss": 0.4438,
+      "step": 4712
+    },
+    {
+      "epoch": 1.0513049297345527,
+      "grad_norm": 0.1628340631723404,
+      "learning_rate": 1.4665575704356472e-05,
+      "loss": 0.4922,
+      "step": 4713
+    },
+    {
+      "epoch": 1.0515279946464422,
+      "grad_norm": 0.15342377126216888,
+      "learning_rate": 1.4663493988608348e-05,
+      "loss": 0.4448,
+      "step": 4714
+    },
+    {
+      "epoch": 1.0517510595583315,
+      "grad_norm": 0.15994024276733398,
+      "learning_rate": 1.4661412014566594e-05,
+      "loss": 0.4655,
+      "step": 4715
+    },
+    {
+      "epoch": 1.0519741244702208,
+      "grad_norm": 0.17778268456459045,
+      "learning_rate": 1.4659329782346518e-05,
+      "loss": 0.5056,
+      "step": 4716
+    },
+    {
+      "epoch": 1.0521971893821103,
+      "grad_norm": 0.16450022161006927,
+      "learning_rate": 1.4657247292063455e-05,
+      "loss": 0.4914,
+      "step": 4717
+    },
+    {
+      "epoch": 1.0524202542939995,
+      "grad_norm": 0.15438216924667358,
+      "learning_rate": 1.4655164543832738e-05,
+      "loss": 0.4513,
+      "step": 4718
+    },
+    {
+      "epoch": 1.0526433192058888,
+      "grad_norm": 0.165638267993927,
+      "learning_rate": 1.4653081537769729e-05,
+      "loss": 0.487,
+      "step": 4719
+    },
+    {
+      "epoch": 1.0528663841177783,
+      "grad_norm": 0.1668267548084259,
+      "learning_rate": 1.4650998273989794e-05,
+      "loss": 0.4936,
+      "step": 4720
+    },
+    {
+      "epoch": 1.0530894490296676,
+      "grad_norm": 0.19577254354953766,
+      "learning_rate": 1.464891475260832e-05,
+      "loss": 0.4849,
+      "step": 4721
+    },
+    {
+      "epoch": 1.053312513941557,
+      "grad_norm": 0.16912342607975006,
+      "learning_rate": 1.4646830973740703e-05,
+      "loss": 0.4712,
+      "step": 4722
+    },
+    {
+      "epoch": 1.0535355788534464,
+      "grad_norm": 0.17454881966114044,
+      "learning_rate": 1.4644746937502356e-05,
+      "loss": 0.4945,
+      "step": 4723
+    },
+    {
+      "epoch": 1.0537586437653357,
+      "grad_norm": 0.16234014928340912,
+      "learning_rate": 1.464266264400871e-05,
+      "loss": 0.4644,
+      "step": 4724
+    },
+    {
+      "epoch": 1.0539817086772252,
+      "grad_norm": 0.16444821655750275,
+      "learning_rate": 1.46405780933752e-05,
+      "loss": 0.4698,
+      "step": 4725
+    },
+    {
+      "epoch": 1.0542047735891145,
+      "grad_norm": 0.15134121477603912,
+      "learning_rate": 1.4638493285717286e-05,
+      "loss": 0.4545,
+      "step": 4726
+    },
+    {
+      "epoch": 1.0544278385010037,
+      "grad_norm": 0.16439802944660187,
+      "learning_rate": 1.4636408221150436e-05,
+      "loss": 0.4881,
+      "step": 4727
+    },
+    {
+      "epoch": 1.0546509034128932,
+      "grad_norm": 0.16219298541545868,
+      "learning_rate": 1.4634322899790137e-05,
+      "loss": 0.4501,
+      "step": 4728
+    },
+    {
+      "epoch": 1.0548739683247825,
+      "grad_norm": 0.1483144313097,
+      "learning_rate": 1.463223732175188e-05,
+      "loss": 0.4461,
+      "step": 4729
+    },
+    {
+      "epoch": 1.0550970332366718,
+      "grad_norm": 0.16766944527626038,
+      "learning_rate": 1.4630151487151188e-05,
+      "loss": 0.4742,
+      "step": 4730
+    },
+    {
+      "epoch": 1.0553200981485613,
+      "grad_norm": 0.17085810005664825,
+      "learning_rate": 1.4628065396103576e-05,
+      "loss": 0.4917,
+      "step": 4731
+    },
+    {
+      "epoch": 1.0555431630604506,
+      "grad_norm": 0.18476352095603943,
+      "learning_rate": 1.4625979048724594e-05,
+      "loss": 0.4518,
+      "step": 4732
+    },
+    {
+      "epoch": 1.0557662279723399,
+      "grad_norm": 0.1589956432580948,
+      "learning_rate": 1.4623892445129792e-05,
+      "loss": 0.4732,
+      "step": 4733
+    },
+    {
+      "epoch": 1.0559892928842294,
+      "grad_norm": 0.18277813494205475,
+      "learning_rate": 1.4621805585434744e-05,
+      "loss": 0.4752,
+      "step": 4734
+    },
+    {
+      "epoch": 1.0562123577961187,
+      "grad_norm": 0.15718622505664825,
+      "learning_rate": 1.4619718469755029e-05,
+      "loss": 0.4815,
+      "step": 4735
+    },
+    {
+      "epoch": 1.056435422708008,
+      "grad_norm": 0.1619989573955536,
+      "learning_rate": 1.4617631098206244e-05,
+      "loss": 0.4595,
+      "step": 4736
+    },
+    {
+      "epoch": 1.0566584876198974,
+      "grad_norm": 0.16704005002975464,
+      "learning_rate": 1.4615543470904005e-05,
+      "loss": 0.4769,
+      "step": 4737
+    },
+    {
+      "epoch": 1.0568815525317867,
+      "grad_norm": 0.16695329546928406,
+      "learning_rate": 1.4613455587963934e-05,
+      "loss": 0.4868,
+      "step": 4738
+    },
+    {
+      "epoch": 1.057104617443676,
+      "grad_norm": 0.165815070271492,
+      "learning_rate": 1.4611367449501674e-05,
+      "loss": 0.4601,
+      "step": 4739
+    },
+    {
+      "epoch": 1.0573276823555655,
+      "grad_norm": 0.1685144156217575,
+      "learning_rate": 1.4609279055632878e-05,
+      "loss": 0.4848,
+      "step": 4740
+    },
+    {
+      "epoch": 1.0575507472674548,
+      "grad_norm": 0.16440360248088837,
+      "learning_rate": 1.4607190406473214e-05,
+      "loss": 0.4729,
+      "step": 4741
+    },
+    {
+      "epoch": 1.0577738121793443,
+      "grad_norm": 0.16086186468601227,
+      "learning_rate": 1.4605101502138363e-05,
+      "loss": 0.4719,
+      "step": 4742
+    },
+    {
+      "epoch": 1.0579968770912336,
+      "grad_norm": 0.16252481937408447,
+      "learning_rate": 1.4603012342744027e-05,
+      "loss": 0.4739,
+      "step": 4743
+    },
+    {
+      "epoch": 1.0582199420031229,
+      "grad_norm": 0.15817517042160034,
+      "learning_rate": 1.4600922928405911e-05,
+      "loss": 0.4982,
+      "step": 4744
+    },
+    {
+      "epoch": 1.0584430069150124,
+      "grad_norm": 0.15263493359088898,
+      "learning_rate": 1.4598833259239746e-05,
+      "loss": 0.4645,
+      "step": 4745
+    },
+    {
+      "epoch": 1.0586660718269016,
+      "grad_norm": 0.2680174708366394,
+      "learning_rate": 1.4596743335361263e-05,
+      "loss": 0.4876,
+      "step": 4746
+    },
+    {
+      "epoch": 1.058889136738791,
+      "grad_norm": 0.16394802927970886,
+      "learning_rate": 1.4594653156886222e-05,
+      "loss": 0.4812,
+      "step": 4747
+    },
+    {
+      "epoch": 1.0591122016506804,
+      "grad_norm": 0.1608896255493164,
+      "learning_rate": 1.4592562723930385e-05,
+      "loss": 0.4755,
+      "step": 4748
+    },
+    {
+      "epoch": 1.0593352665625697,
+      "grad_norm": 0.17238877713680267,
+      "learning_rate": 1.459047203660954e-05,
+      "loss": 0.4926,
+      "step": 4749
+    },
+    {
+      "epoch": 1.059558331474459,
+      "grad_norm": 0.16521191596984863,
+      "learning_rate": 1.4588381095039474e-05,
+      "loss": 0.4937,
+      "step": 4750
+    },
+    {
+      "epoch": 1.0597813963863485,
+      "grad_norm": 0.1620710790157318,
+      "learning_rate": 1.4586289899336003e-05,
+      "loss": 0.4339,
+      "step": 4751
+    },
+    {
+      "epoch": 1.0600044612982378,
+      "grad_norm": 0.15963472425937653,
+      "learning_rate": 1.4584198449614947e-05,
+      "loss": 0.4546,
+      "step": 4752
+    },
+    {
+      "epoch": 1.060227526210127,
+      "grad_norm": 0.17011800408363342,
+      "learning_rate": 1.4582106745992149e-05,
+      "loss": 0.4761,
+      "step": 4753
+    },
+    {
+      "epoch": 1.0604505911220166,
+      "grad_norm": 0.15755783021450043,
+      "learning_rate": 1.4580014788583452e-05,
+      "loss": 0.4621,
+      "step": 4754
+    },
+    {
+      "epoch": 1.0606736560339058,
+      "grad_norm": 0.1612185388803482,
+      "learning_rate": 1.457792257750473e-05,
+      "loss": 0.4526,
+      "step": 4755
+    },
+    {
+      "epoch": 1.0608967209457951,
+      "grad_norm": 0.17329202592372894,
+      "learning_rate": 1.4575830112871855e-05,
+      "loss": 0.472,
+      "step": 4756
+    },
+    {
+      "epoch": 1.0611197858576846,
+      "grad_norm": 0.16672907769680023,
+      "learning_rate": 1.4573737394800729e-05,
+      "loss": 0.4898,
+      "step": 4757
+    },
+    {
+      "epoch": 1.061342850769574,
+      "grad_norm": 0.17075766623020172,
+      "learning_rate": 1.4571644423407257e-05,
+      "loss": 0.474,
+      "step": 4758
+    },
+    {
+      "epoch": 1.0615659156814634,
+      "grad_norm": 0.15704765915870667,
+      "learning_rate": 1.4569551198807357e-05,
+      "loss": 0.4951,
+      "step": 4759
+    },
+    {
+      "epoch": 1.0617889805933527,
+      "grad_norm": 0.1504676192998886,
+      "learning_rate": 1.4567457721116971e-05,
+      "loss": 0.4597,
+      "step": 4760
+    },
+    {
+      "epoch": 1.062012045505242,
+      "grad_norm": 0.15645162761211395,
+      "learning_rate": 1.4565363990452046e-05,
+      "loss": 0.4548,
+      "step": 4761
+    },
+    {
+      "epoch": 1.0622351104171315,
+      "grad_norm": 0.16380253434181213,
+      "learning_rate": 1.4563270006928544e-05,
+      "loss": 0.4675,
+      "step": 4762
+    },
+    {
+      "epoch": 1.0624581753290208,
+      "grad_norm": 0.16168953478336334,
+      "learning_rate": 1.4561175770662446e-05,
+      "loss": 0.5015,
+      "step": 4763
+    },
+    {
+      "epoch": 1.06268124024091,
+      "grad_norm": 0.16128864884376526,
+      "learning_rate": 1.4559081281769742e-05,
+      "loss": 0.4827,
+      "step": 4764
+    },
+    {
+      "epoch": 1.0629043051527995,
+      "grad_norm": 0.1598096489906311,
+      "learning_rate": 1.4556986540366437e-05,
+      "loss": 0.4598,
+      "step": 4765
+    },
+    {
+      "epoch": 1.0631273700646888,
+      "grad_norm": 0.1796981692314148,
+      "learning_rate": 1.4554891546568557e-05,
+      "loss": 0.4803,
+      "step": 4766
+    },
+    {
+      "epoch": 1.063350434976578,
+      "grad_norm": 0.15178261697292328,
+      "learning_rate": 1.4552796300492129e-05,
+      "loss": 0.4228,
+      "step": 4767
+    },
+    {
+      "epoch": 1.0635734998884676,
+      "grad_norm": 0.15565361082553864,
+      "learning_rate": 1.4550700802253203e-05,
+      "loss": 0.484,
+      "step": 4768
+    },
+    {
+      "epoch": 1.063796564800357,
+      "grad_norm": 0.16248026490211487,
+      "learning_rate": 1.4548605051967843e-05,
+      "loss": 0.448,
+      "step": 4769
+    },
+    {
+      "epoch": 1.0640196297122462,
+      "grad_norm": 0.17214354872703552,
+      "learning_rate": 1.4546509049752122e-05,
+      "loss": 0.466,
+      "step": 4770
+    },
+    {
+      "epoch": 1.0642426946241357,
+      "grad_norm": 0.1753188818693161,
+      "learning_rate": 1.4544412795722135e-05,
+      "loss": 0.4861,
+      "step": 4771
+    },
+    {
+      "epoch": 1.064465759536025,
+      "grad_norm": 0.16392984986305237,
+      "learning_rate": 1.4542316289993976e-05,
+      "loss": 0.4584,
+      "step": 4772
+    },
+    {
+      "epoch": 1.0646888244479142,
+      "grad_norm": 0.16576439142227173,
+      "learning_rate": 1.4540219532683774e-05,
+      "loss": 0.456,
+      "step": 4773
+    },
+    {
+      "epoch": 1.0649118893598037,
+      "grad_norm": 0.17842838168144226,
+      "learning_rate": 1.4538122523907651e-05,
+      "loss": 0.471,
+      "step": 4774
+    },
+    {
+      "epoch": 1.065134954271693,
+      "grad_norm": 0.15883475542068481,
+      "learning_rate": 1.4536025263781762e-05,
+      "loss": 0.4868,
+      "step": 4775
+    },
+    {
+      "epoch": 1.0653580191835825,
+      "grad_norm": 0.15571808815002441,
+      "learning_rate": 1.4533927752422256e-05,
+      "loss": 0.4742,
+      "step": 4776
+    },
+    {
+      "epoch": 1.0655810840954718,
+      "grad_norm": 0.16916531324386597,
+      "learning_rate": 1.4531829989945315e-05,
+      "loss": 0.4646,
+      "step": 4777
+    },
+    {
+      "epoch": 1.065804149007361,
+      "grad_norm": 0.15617609024047852,
+      "learning_rate": 1.4529731976467119e-05,
+      "loss": 0.4569,
+      "step": 4778
+    },
+    {
+      "epoch": 1.0660272139192506,
+      "grad_norm": 0.15956276655197144,
+      "learning_rate": 1.4527633712103875e-05,
+      "loss": 0.4739,
+      "step": 4779
+    },
+    {
+      "epoch": 1.0662502788311399,
+      "grad_norm": 0.1674896627664566,
+      "learning_rate": 1.4525535196971797e-05,
+      "loss": 0.4993,
+      "step": 4780
+    },
+    {
+      "epoch": 1.0664733437430292,
+      "grad_norm": 0.14718051254749298,
+      "learning_rate": 1.4523436431187112e-05,
+      "loss": 0.4433,
+      "step": 4781
+    },
+    {
+      "epoch": 1.0666964086549187,
+      "grad_norm": 0.16965600848197937,
+      "learning_rate": 1.4521337414866064e-05,
+      "loss": 0.496,
+      "step": 4782
+    },
+    {
+      "epoch": 1.066919473566808,
+      "grad_norm": 0.16213834285736084,
+      "learning_rate": 1.451923814812491e-05,
+      "loss": 0.4385,
+      "step": 4783
+    },
+    {
+      "epoch": 1.0671425384786972,
+      "grad_norm": 0.16705723106861115,
+      "learning_rate": 1.451713863107992e-05,
+      "loss": 0.4893,
+      "step": 4784
+    },
+    {
+      "epoch": 1.0673656033905867,
+      "grad_norm": 0.16360625624656677,
+      "learning_rate": 1.451503886384738e-05,
+      "loss": 0.5065,
+      "step": 4785
+    },
+    {
+      "epoch": 1.067588668302476,
+      "grad_norm": 0.1604512631893158,
+      "learning_rate": 1.4512938846543583e-05,
+      "loss": 0.4597,
+      "step": 4786
+    },
+    {
+      "epoch": 1.0678117332143653,
+      "grad_norm": 0.16134783625602722,
+      "learning_rate": 1.4510838579284849e-05,
+      "loss": 0.467,
+      "step": 4787
+    },
+    {
+      "epoch": 1.0680347981262548,
+      "grad_norm": 0.1614135503768921,
+      "learning_rate": 1.4508738062187497e-05,
+      "loss": 0.482,
+      "step": 4788
+    },
+    {
+      "epoch": 1.068257863038144,
+      "grad_norm": 0.16068898141384125,
+      "learning_rate": 1.4506637295367872e-05,
+      "loss": 0.4557,
+      "step": 4789
+    },
+    {
+      "epoch": 1.0684809279500334,
+      "grad_norm": 0.18150994181632996,
+      "learning_rate": 1.4504536278942327e-05,
+      "loss": 0.4796,
+      "step": 4790
+    },
+    {
+      "epoch": 1.0687039928619229,
+      "grad_norm": 0.16428856551647186,
+      "learning_rate": 1.4502435013027225e-05,
+      "loss": 0.478,
+      "step": 4791
+    },
+    {
+      "epoch": 1.0689270577738121,
+      "grad_norm": 0.16501575708389282,
+      "learning_rate": 1.4500333497738955e-05,
+      "loss": 0.4575,
+      "step": 4792
+    },
+    {
+      "epoch": 1.0691501226857016,
+      "grad_norm": 0.16125361621379852,
+      "learning_rate": 1.4498231733193904e-05,
+      "loss": 0.4713,
+      "step": 4793
+    },
+    {
+      "epoch": 1.069373187597591,
+      "grad_norm": 0.15834558010101318,
+      "learning_rate": 1.4496129719508486e-05,
+      "loss": 0.4721,
+      "step": 4794
+    },
+    {
+      "epoch": 1.0695962525094802,
+      "grad_norm": 0.17870576679706573,
+      "learning_rate": 1.449402745679912e-05,
+      "loss": 0.4885,
+      "step": 4795
+    },
+    {
+      "epoch": 1.0698193174213697,
+      "grad_norm": 0.15810585021972656,
+      "learning_rate": 1.4491924945182248e-05,
+      "loss": 0.4698,
+      "step": 4796
+    },
+    {
+      "epoch": 1.070042382333259,
+      "grad_norm": 0.17196185886859894,
+      "learning_rate": 1.4489822184774317e-05,
+      "loss": 0.4946,
+      "step": 4797
+    },
+    {
+      "epoch": 1.0702654472451483,
+      "grad_norm": 0.1619616448879242,
+      "learning_rate": 1.448771917569179e-05,
+      "loss": 0.4697,
+      "step": 4798
+    },
+    {
+      "epoch": 1.0704885121570378,
+      "grad_norm": 0.1760883778333664,
+      "learning_rate": 1.4485615918051146e-05,
+      "loss": 0.4533,
+      "step": 4799
+    },
+    {
+      "epoch": 1.070711577068927,
+      "grad_norm": 0.15732388198375702,
+      "learning_rate": 1.448351241196888e-05,
+      "loss": 0.4639,
+      "step": 4800
+    },
+    {
+      "epoch": 1.0709346419808163,
+      "grad_norm": 0.1595136672258377,
+      "learning_rate": 1.448140865756149e-05,
+      "loss": 0.48,
+      "step": 4801
+    },
+    {
+      "epoch": 1.0711577068927058,
+      "grad_norm": 0.19920584559440613,
+      "learning_rate": 1.44793046549455e-05,
+      "loss": 0.4647,
+      "step": 4802
+    },
+    {
+      "epoch": 1.0713807718045951,
+      "grad_norm": 0.16085761785507202,
+      "learning_rate": 1.4477200404237446e-05,
+      "loss": 0.4788,
+      "step": 4803
+    },
+    {
+      "epoch": 1.0716038367164844,
+      "grad_norm": 0.15879055857658386,
+      "learning_rate": 1.4475095905553867e-05,
+      "loss": 0.4805,
+      "step": 4804
+    },
+    {
+      "epoch": 1.071826901628374,
+      "grad_norm": 0.16476532816886902,
+      "learning_rate": 1.4472991159011329e-05,
+      "loss": 0.4996,
+      "step": 4805
+    },
+    {
+      "epoch": 1.0720499665402632,
+      "grad_norm": 0.1692475974559784,
+      "learning_rate": 1.4470886164726403e-05,
+      "loss": 0.464,
+      "step": 4806
+    },
+    {
+      "epoch": 1.0722730314521525,
+      "grad_norm": 0.16641223430633545,
+      "learning_rate": 1.4468780922815679e-05,
+      "loss": 0.4698,
+      "step": 4807
+    },
+    {
+      "epoch": 1.072496096364042,
+      "grad_norm": 0.15899412333965302,
+      "learning_rate": 1.4466675433395758e-05,
+      "loss": 0.4688,
+      "step": 4808
+    },
+    {
+      "epoch": 1.0727191612759313,
+      "grad_norm": 0.1671571284532547,
+      "learning_rate": 1.4464569696583256e-05,
+      "loss": 0.4948,
+      "step": 4809
+    },
+    {
+      "epoch": 1.0729422261878208,
+      "grad_norm": 0.16346615552902222,
+      "learning_rate": 1.4462463712494799e-05,
+      "loss": 0.4682,
+      "step": 4810
+    },
+    {
+      "epoch": 1.07316529109971,
+      "grad_norm": 0.16411162912845612,
+      "learning_rate": 1.4460357481247035e-05,
+      "loss": 0.4704,
+      "step": 4811
+    },
+    {
+      "epoch": 1.0733883560115993,
+      "grad_norm": 0.16444380581378937,
+      "learning_rate": 1.4458251002956612e-05,
+      "loss": 0.478,
+      "step": 4812
+    },
+    {
+      "epoch": 1.0736114209234888,
+      "grad_norm": 0.15539826452732086,
+      "learning_rate": 1.4456144277740207e-05,
+      "loss": 0.4827,
+      "step": 4813
+    },
+    {
+      "epoch": 1.073834485835378,
+      "grad_norm": 0.16206075251102448,
+      "learning_rate": 1.4454037305714501e-05,
+      "loss": 0.4651,
+      "step": 4814
+    },
+    {
+      "epoch": 1.0740575507472674,
+      "grad_norm": 0.1546122431755066,
+      "learning_rate": 1.4451930086996193e-05,
+      "loss": 0.4668,
+      "step": 4815
+    },
+    {
+      "epoch": 1.074280615659157,
+      "grad_norm": 0.15941190719604492,
+      "learning_rate": 1.4449822621701992e-05,
+      "loss": 0.4674,
+      "step": 4816
+    },
+    {
+      "epoch": 1.0745036805710462,
+      "grad_norm": 0.16339150071144104,
+      "learning_rate": 1.4447714909948624e-05,
+      "loss": 0.4957,
+      "step": 4817
+    },
+    {
+      "epoch": 1.0747267454829355,
+      "grad_norm": 0.15850695967674255,
+      "learning_rate": 1.4445606951852828e-05,
+      "loss": 0.487,
+      "step": 4818
+    },
+    {
+      "epoch": 1.074949810394825,
+      "grad_norm": 0.16682858765125275,
+      "learning_rate": 1.4443498747531358e-05,
+      "loss": 0.4677,
+      "step": 4819
+    },
+    {
+      "epoch": 1.0751728753067142,
+      "grad_norm": 0.1573875993490219,
+      "learning_rate": 1.444139029710097e-05,
+      "loss": 0.4452,
+      "step": 4820
+    },
+    {
+      "epoch": 1.0753959402186035,
+      "grad_norm": 0.1620924174785614,
+      "learning_rate": 1.4439281600678455e-05,
+      "loss": 0.4767,
+      "step": 4821
+    },
+    {
+      "epoch": 1.075619005130493,
+      "grad_norm": 0.15519385039806366,
+      "learning_rate": 1.4437172658380598e-05,
+      "loss": 0.4571,
+      "step": 4822
+    },
+    {
+      "epoch": 1.0758420700423823,
+      "grad_norm": 0.17758627235889435,
+      "learning_rate": 1.443506347032421e-05,
+      "loss": 0.4767,
+      "step": 4823
+    },
+    {
+      "epoch": 1.0760651349542716,
+      "grad_norm": 0.15553906559944153,
+      "learning_rate": 1.4432954036626108e-05,
+      "loss": 0.4659,
+      "step": 4824
+    },
+    {
+      "epoch": 1.076288199866161,
+      "grad_norm": 0.15930281579494476,
+      "learning_rate": 1.4430844357403126e-05,
+      "loss": 0.4579,
+      "step": 4825
+    },
+    {
+      "epoch": 1.0765112647780504,
+      "grad_norm": 0.15948422253131866,
+      "learning_rate": 1.4428734432772115e-05,
+      "loss": 0.4682,
+      "step": 4826
+    },
+    {
+      "epoch": 1.0767343296899399,
+      "grad_norm": 0.15884706377983093,
+      "learning_rate": 1.4426624262849932e-05,
+      "loss": 0.469,
+      "step": 4827
+    },
+    {
+      "epoch": 1.0769573946018292,
+      "grad_norm": 0.15958735346794128,
+      "learning_rate": 1.4424513847753455e-05,
+      "loss": 0.4576,
+      "step": 4828
+    },
+    {
+      "epoch": 1.0771804595137184,
+      "grad_norm": 0.16710686683654785,
+      "learning_rate": 1.4422403187599565e-05,
+      "loss": 0.471,
+      "step": 4829
+    },
+    {
+      "epoch": 1.077403524425608,
+      "grad_norm": 0.2897154688835144,
+      "learning_rate": 1.4420292282505173e-05,
+      "loss": 0.4901,
+      "step": 4830
+    },
+    {
+      "epoch": 1.0776265893374972,
+      "grad_norm": 0.1648949235677719,
+      "learning_rate": 1.4418181132587185e-05,
+      "loss": 0.4867,
+      "step": 4831
+    },
+    {
+      "epoch": 1.0778496542493865,
+      "grad_norm": 0.1587132066488266,
+      "learning_rate": 1.4416069737962538e-05,
+      "loss": 0.4554,
+      "step": 4832
+    },
+    {
+      "epoch": 1.078072719161276,
+      "grad_norm": 0.3479141891002655,
+      "learning_rate": 1.4413958098748171e-05,
+      "loss": 0.4814,
+      "step": 4833
+    },
+    {
+      "epoch": 1.0782957840731653,
+      "grad_norm": 0.15964049100875854,
+      "learning_rate": 1.441184621506104e-05,
+      "loss": 0.4694,
+      "step": 4834
+    },
+    {
+      "epoch": 1.0785188489850546,
+      "grad_norm": 0.16396582126617432,
+      "learning_rate": 1.440973408701811e-05,
+      "loss": 0.4892,
+      "step": 4835
+    },
+    {
+      "epoch": 1.078741913896944,
+      "grad_norm": 0.1593598872423172,
+      "learning_rate": 1.4407621714736372e-05,
+      "loss": 0.4485,
+      "step": 4836
+    },
+    {
+      "epoch": 1.0789649788088334,
+      "grad_norm": 0.17762166261672974,
+      "learning_rate": 1.4405509098332818e-05,
+      "loss": 0.4358,
+      "step": 4837
+    },
+    {
+      "epoch": 1.0791880437207226,
+      "grad_norm": 0.15934491157531738,
+      "learning_rate": 1.4403396237924456e-05,
+      "loss": 0.4734,
+      "step": 4838
+    },
+    {
+      "epoch": 1.0794111086326121,
+      "grad_norm": 0.18883134424686432,
+      "learning_rate": 1.4401283133628315e-05,
+      "loss": 0.4488,
+      "step": 4839
+    },
+    {
+      "epoch": 1.0796341735445014,
+      "grad_norm": 0.1616184413433075,
+      "learning_rate": 1.4399169785561426e-05,
+      "loss": 0.465,
+      "step": 4840
+    },
+    {
+      "epoch": 1.0798572384563907,
+      "grad_norm": 0.16874286532402039,
+      "learning_rate": 1.4397056193840842e-05,
+      "loss": 0.5069,
+      "step": 4841
+    },
+    {
+      "epoch": 1.0800803033682802,
+      "grad_norm": 0.18162856996059418,
+      "learning_rate": 1.4394942358583627e-05,
+      "loss": 0.4929,
+      "step": 4842
+    },
+    {
+      "epoch": 1.0803033682801695,
+      "grad_norm": 0.15963618457317352,
+      "learning_rate": 1.4392828279906866e-05,
+      "loss": 0.46,
+      "step": 4843
+    },
+    {
+      "epoch": 1.080526433192059,
+      "grad_norm": 0.17380237579345703,
+      "learning_rate": 1.4390713957927636e-05,
+      "loss": 0.4929,
+      "step": 4844
+    },
+    {
+      "epoch": 1.0807494981039483,
+      "grad_norm": 0.174495667219162,
+      "learning_rate": 1.4388599392763052e-05,
+      "loss": 0.4626,
+      "step": 4845
+    },
+    {
+      "epoch": 1.0809725630158376,
+      "grad_norm": 0.17126008868217468,
+      "learning_rate": 1.4386484584530225e-05,
+      "loss": 0.47,
+      "step": 4846
+    },
+    {
+      "epoch": 1.081195627927727,
+      "grad_norm": 0.15579789876937866,
+      "learning_rate": 1.4384369533346292e-05,
+      "loss": 0.4565,
+      "step": 4847
+    },
+    {
+      "epoch": 1.0814186928396163,
+      "grad_norm": 0.16992245614528656,
+      "learning_rate": 1.4382254239328392e-05,
+      "loss": 0.4646,
+      "step": 4848
+    },
+    {
+      "epoch": 1.0816417577515056,
+      "grad_norm": 0.1682998389005661,
+      "learning_rate": 1.4380138702593691e-05,
+      "loss": 0.4807,
+      "step": 4849
+    },
+    {
+      "epoch": 1.0818648226633951,
+      "grad_norm": 0.16254080832004547,
+      "learning_rate": 1.4378022923259354e-05,
+      "loss": 0.4604,
+      "step": 4850
+    },
+    {
+      "epoch": 1.0820878875752844,
+      "grad_norm": 0.16904620826244354,
+      "learning_rate": 1.4375906901442568e-05,
+      "loss": 0.5068,
+      "step": 4851
+    },
+    {
+      "epoch": 1.0823109524871737,
+      "grad_norm": 0.15982915461063385,
+      "learning_rate": 1.4373790637260534e-05,
+      "loss": 0.4757,
+      "step": 4852
+    },
+    {
+      "epoch": 1.0825340173990632,
+      "grad_norm": 0.1598198115825653,
+      "learning_rate": 1.4371674130830462e-05,
+      "loss": 0.4596,
+      "step": 4853
+    },
+    {
+      "epoch": 1.0827570823109525,
+      "grad_norm": 0.2856222689151764,
+      "learning_rate": 1.4369557382269577e-05,
+      "loss": 0.4618,
+      "step": 4854
+    },
+    {
+      "epoch": 1.0829801472228417,
+      "grad_norm": 0.1609204113483429,
+      "learning_rate": 1.4367440391695118e-05,
+      "loss": 0.4814,
+      "step": 4855
+    },
+    {
+      "epoch": 1.0832032121347313,
+      "grad_norm": 0.1572166234254837,
+      "learning_rate": 1.4365323159224341e-05,
+      "loss": 0.4768,
+      "step": 4856
+    },
+    {
+      "epoch": 1.0834262770466205,
+      "grad_norm": 0.15718881785869598,
+      "learning_rate": 1.4363205684974504e-05,
+      "loss": 0.4594,
+      "step": 4857
+    },
+    {
+      "epoch": 1.0836493419585098,
+      "grad_norm": 0.1571321189403534,
+      "learning_rate": 1.4361087969062895e-05,
+      "loss": 0.4705,
+      "step": 4858
+    },
+    {
+      "epoch": 1.0838724068703993,
+      "grad_norm": 0.16006889939308167,
+      "learning_rate": 1.4358970011606798e-05,
+      "loss": 0.4549,
+      "step": 4859
+    },
+    {
+      "epoch": 1.0840954717822886,
+      "grad_norm": 0.1665477454662323,
+      "learning_rate": 1.4356851812723524e-05,
+      "loss": 0.4657,
+      "step": 4860
+    },
+    {
+      "epoch": 1.084318536694178,
+      "grad_norm": 0.16529056429862976,
+      "learning_rate": 1.4354733372530391e-05,
+      "loss": 0.476,
+      "step": 4861
+    },
+    {
+      "epoch": 1.0845416016060674,
+      "grad_norm": 0.1675388067960739,
+      "learning_rate": 1.4352614691144733e-05,
+      "loss": 0.4721,
+      "step": 4862
+    },
+    {
+      "epoch": 1.0847646665179567,
+      "grad_norm": 0.16424338519573212,
+      "learning_rate": 1.4350495768683893e-05,
+      "loss": 0.4856,
+      "step": 4863
+    },
+    {
+      "epoch": 1.0849877314298462,
+      "grad_norm": 0.16360746324062347,
+      "learning_rate": 1.4348376605265233e-05,
+      "loss": 0.4916,
+      "step": 4864
+    },
+    {
+      "epoch": 1.0852107963417355,
+      "grad_norm": 0.16066889464855194,
+      "learning_rate": 1.4346257201006122e-05,
+      "loss": 0.4606,
+      "step": 4865
+    },
+    {
+      "epoch": 1.0854338612536247,
+      "grad_norm": 0.15210098028182983,
+      "learning_rate": 1.434413755602395e-05,
+      "loss": 0.4684,
+      "step": 4866
+    },
+    {
+      "epoch": 1.0856569261655142,
+      "grad_norm": 0.16423378884792328,
+      "learning_rate": 1.4342017670436113e-05,
+      "loss": 0.4654,
+      "step": 4867
+    },
+    {
+      "epoch": 1.0858799910774035,
+      "grad_norm": 0.191256582736969,
+      "learning_rate": 1.4339897544360026e-05,
+      "loss": 0.4729,
+      "step": 4868
+    },
+    {
+      "epoch": 1.0861030559892928,
+      "grad_norm": 0.15907546877861023,
+      "learning_rate": 1.4337777177913113e-05,
+      "loss": 0.4545,
+      "step": 4869
+    },
+    {
+      "epoch": 1.0863261209011823,
+      "grad_norm": 0.15440797805786133,
+      "learning_rate": 1.4335656571212814e-05,
+      "loss": 0.4816,
+      "step": 4870
+    },
+    {
+      "epoch": 1.0865491858130716,
+      "grad_norm": 0.16571897268295288,
+      "learning_rate": 1.4333535724376585e-05,
+      "loss": 0.4498,
+      "step": 4871
+    },
+    {
+      "epoch": 1.0867722507249609,
+      "grad_norm": 0.16138556599617004,
+      "learning_rate": 1.4331414637521885e-05,
+      "loss": 0.4541,
+      "step": 4872
+    },
+    {
+      "epoch": 1.0869953156368504,
+      "grad_norm": 0.17464113235473633,
+      "learning_rate": 1.4329293310766196e-05,
+      "loss": 0.4997,
+      "step": 4873
+    },
+    {
+      "epoch": 1.0872183805487396,
+      "grad_norm": 0.15958289802074432,
+      "learning_rate": 1.4327171744227015e-05,
+      "loss": 0.4351,
+      "step": 4874
+    },
+    {
+      "epoch": 1.087441445460629,
+      "grad_norm": 0.17836567759513855,
+      "learning_rate": 1.432504993802184e-05,
+      "loss": 0.4626,
+      "step": 4875
+    },
+    {
+      "epoch": 1.0876645103725184,
+      "grad_norm": 0.16179955005645752,
+      "learning_rate": 1.4322927892268195e-05,
+      "loss": 0.4931,
+      "step": 4876
+    },
+    {
+      "epoch": 1.0878875752844077,
+      "grad_norm": 0.16194845736026764,
+      "learning_rate": 1.4320805607083611e-05,
+      "loss": 0.4619,
+      "step": 4877
+    },
+    {
+      "epoch": 1.0881106401962972,
+      "grad_norm": 0.16176071763038635,
+      "learning_rate": 1.4318683082585634e-05,
+      "loss": 0.4798,
+      "step": 4878
+    },
+    {
+      "epoch": 1.0883337051081865,
+      "grad_norm": 0.15933561325073242,
+      "learning_rate": 1.4316560318891823e-05,
+      "loss": 0.4635,
+      "step": 4879
+    },
+    {
+      "epoch": 1.0885567700200758,
+      "grad_norm": 0.16447357833385468,
+      "learning_rate": 1.431443731611975e-05,
+      "loss": 0.4747,
+      "step": 4880
+    },
+    {
+      "epoch": 1.0887798349319653,
+      "grad_norm": 0.1632988601922989,
+      "learning_rate": 1.4312314074386998e-05,
+      "loss": 0.489,
+      "step": 4881
+    },
+    {
+      "epoch": 1.0890028998438546,
+      "grad_norm": 0.16720198094844818,
+      "learning_rate": 1.4310190593811167e-05,
+      "loss": 0.4814,
+      "step": 4882
+    },
+    {
+      "epoch": 1.0892259647557438,
+      "grad_norm": 0.16806739568710327,
+      "learning_rate": 1.4308066874509869e-05,
+      "loss": 0.4566,
+      "step": 4883
+    },
+    {
+      "epoch": 1.0894490296676334,
+      "grad_norm": 0.1616521179676056,
+      "learning_rate": 1.430594291660073e-05,
+      "loss": 0.4875,
+      "step": 4884
+    },
+    {
+      "epoch": 1.0896720945795226,
+      "grad_norm": 0.15476809442043304,
+      "learning_rate": 1.4303818720201386e-05,
+      "loss": 0.4632,
+      "step": 4885
+    },
+    {
+      "epoch": 1.089895159491412,
+      "grad_norm": 0.15796570479869843,
+      "learning_rate": 1.4301694285429489e-05,
+      "loss": 0.4558,
+      "step": 4886
+    },
+    {
+      "epoch": 1.0901182244033014,
+      "grad_norm": 0.1628834307193756,
+      "learning_rate": 1.4299569612402701e-05,
+      "loss": 0.4578,
+      "step": 4887
+    },
+    {
+      "epoch": 1.0903412893151907,
+      "grad_norm": 0.16493961215019226,
+      "learning_rate": 1.4297444701238706e-05,
+      "loss": 0.5188,
+      "step": 4888
+    },
+    {
+      "epoch": 1.09056435422708,
+      "grad_norm": 0.17024171352386475,
+      "learning_rate": 1.4295319552055191e-05,
+      "loss": 0.4797,
+      "step": 4889
+    },
+    {
+      "epoch": 1.0907874191389695,
+      "grad_norm": 0.16179481148719788,
+      "learning_rate": 1.4293194164969859e-05,
+      "loss": 0.4894,
+      "step": 4890
+    },
+    {
+      "epoch": 1.0910104840508588,
+      "grad_norm": 0.15970705449581146,
+      "learning_rate": 1.429106854010043e-05,
+      "loss": 0.4698,
+      "step": 4891
+    },
+    {
+      "epoch": 1.091233548962748,
+      "grad_norm": 0.15802475810050964,
+      "learning_rate": 1.4288942677564634e-05,
+      "loss": 0.4585,
+      "step": 4892
+    },
+    {
+      "epoch": 1.0914566138746375,
+      "grad_norm": 0.16076651215553284,
+      "learning_rate": 1.4286816577480211e-05,
+      "loss": 0.4643,
+      "step": 4893
+    },
+    {
+      "epoch": 1.0916796787865268,
+      "grad_norm": 0.1617365926504135,
+      "learning_rate": 1.4284690239964925e-05,
+      "loss": 0.4608,
+      "step": 4894
+    },
+    {
+      "epoch": 1.0919027436984163,
+      "grad_norm": 0.1737692803144455,
+      "learning_rate": 1.428256366513654e-05,
+      "loss": 0.5016,
+      "step": 4895
+    },
+    {
+      "epoch": 1.0921258086103056,
+      "grad_norm": 0.16984876990318298,
+      "learning_rate": 1.428043685311284e-05,
+      "loss": 0.476,
+      "step": 4896
+    },
+    {
+      "epoch": 1.092348873522195,
+      "grad_norm": 0.16375143826007843,
+      "learning_rate": 1.427830980401162e-05,
+      "loss": 0.485,
+      "step": 4897
+    },
+    {
+      "epoch": 1.0925719384340844,
+      "grad_norm": 0.16099539399147034,
+      "learning_rate": 1.4276182517950696e-05,
+      "loss": 0.4743,
+      "step": 4898
+    },
+    {
+      "epoch": 1.0927950033459737,
+      "grad_norm": 0.16580551862716675,
+      "learning_rate": 1.4274054995047884e-05,
+      "loss": 0.473,
+      "step": 4899
+    },
+    {
+      "epoch": 1.093018068257863,
+      "grad_norm": 0.15540550649166107,
+      "learning_rate": 1.427192723542102e-05,
+      "loss": 0.4353,
+      "step": 4900
+    },
+    {
+      "epoch": 1.0932411331697525,
+      "grad_norm": 0.1589297354221344,
+      "learning_rate": 1.4269799239187956e-05,
+      "loss": 0.4637,
+      "step": 4901
+    },
+    {
+      "epoch": 1.0934641980816417,
+      "grad_norm": 0.16535155475139618,
+      "learning_rate": 1.4267671006466552e-05,
+      "loss": 0.4492,
+      "step": 4902
+    },
+    {
+      "epoch": 1.093687262993531,
+      "grad_norm": 0.15984977781772614,
+      "learning_rate": 1.4265542537374684e-05,
+      "loss": 0.4703,
+      "step": 4903
+    },
+    {
+      "epoch": 1.0939103279054205,
+      "grad_norm": 0.15200082957744598,
+      "learning_rate": 1.4263413832030237e-05,
+      "loss": 0.4483,
+      "step": 4904
+    },
+    {
+      "epoch": 1.0941333928173098,
+      "grad_norm": 0.21631982922554016,
+      "learning_rate": 1.4261284890551115e-05,
+      "loss": 0.4881,
+      "step": 4905
+    },
+    {
+      "epoch": 1.094356457729199,
+      "grad_norm": 0.16515322029590607,
+      "learning_rate": 1.4259155713055231e-05,
+      "loss": 0.4788,
+      "step": 4906
+    },
+    {
+      "epoch": 1.0945795226410886,
+      "grad_norm": 0.16507336497306824,
+      "learning_rate": 1.4257026299660511e-05,
+      "loss": 0.4749,
+      "step": 4907
+    },
+    {
+      "epoch": 1.0948025875529779,
+      "grad_norm": 0.17329774796962738,
+      "learning_rate": 1.4254896650484897e-05,
+      "loss": 0.4708,
+      "step": 4908
+    },
+    {
+      "epoch": 1.0950256524648672,
+      "grad_norm": 0.15935118496418,
+      "learning_rate": 1.4252766765646344e-05,
+      "loss": 0.4405,
+      "step": 4909
+    },
+    {
+      "epoch": 1.0952487173767567,
+      "grad_norm": 0.16520404815673828,
+      "learning_rate": 1.4250636645262813e-05,
+      "loss": 0.4546,
+      "step": 4910
+    },
+    {
+      "epoch": 1.095471782288646,
+      "grad_norm": 0.16655127704143524,
+      "learning_rate": 1.424850628945229e-05,
+      "loss": 0.4838,
+      "step": 4911
+    },
+    {
+      "epoch": 1.0956948472005354,
+      "grad_norm": 0.17478324472904205,
+      "learning_rate": 1.4246375698332764e-05,
+      "loss": 0.4932,
+      "step": 4912
+    },
+    {
+      "epoch": 1.0959179121124247,
+      "grad_norm": 0.15752072632312775,
+      "learning_rate": 1.4244244872022244e-05,
+      "loss": 0.4386,
+      "step": 4913
+    },
+    {
+      "epoch": 1.096140977024314,
+      "grad_norm": 0.16019755601882935,
+      "learning_rate": 1.424211381063874e-05,
+      "loss": 0.4474,
+      "step": 4914
+    },
+    {
+      "epoch": 1.0963640419362035,
+      "grad_norm": 0.1732664704322815,
+      "learning_rate": 1.4239982514300294e-05,
+      "loss": 0.4423,
+      "step": 4915
+    },
+    {
+      "epoch": 1.0965871068480928,
+      "grad_norm": 0.16661664843559265,
+      "learning_rate": 1.4237850983124943e-05,
+      "loss": 0.5198,
+      "step": 4916
+    },
+    {
+      "epoch": 1.096810171759982,
+      "grad_norm": 0.15834182500839233,
+      "learning_rate": 1.4235719217230751e-05,
+      "loss": 0.4784,
+      "step": 4917
+    },
+    {
+      "epoch": 1.0970332366718716,
+      "grad_norm": 0.16510586440563202,
+      "learning_rate": 1.423358721673578e-05,
+      "loss": 0.4843,
+      "step": 4918
+    },
+    {
+      "epoch": 1.0972563015837609,
+      "grad_norm": 0.16284050047397614,
+      "learning_rate": 1.4231454981758122e-05,
+      "loss": 0.4559,
+      "step": 4919
+    },
+    {
+      "epoch": 1.0974793664956501,
+      "grad_norm": 0.1576804369688034,
+      "learning_rate": 1.422932251241587e-05,
+      "loss": 0.4855,
+      "step": 4920
+    },
+    {
+      "epoch": 1.0977024314075396,
+      "grad_norm": 0.17963336408138275,
+      "learning_rate": 1.4227189808827131e-05,
+      "loss": 0.456,
+      "step": 4921
+    },
+    {
+      "epoch": 1.097925496319429,
+      "grad_norm": 0.16884389519691467,
+      "learning_rate": 1.4225056871110032e-05,
+      "loss": 0.4513,
+      "step": 4922
+    },
+    {
+      "epoch": 1.0981485612313182,
+      "grad_norm": 0.16066482663154602,
+      "learning_rate": 1.422292369938271e-05,
+      "loss": 0.4753,
+      "step": 4923
+    },
+    {
+      "epoch": 1.0983716261432077,
+      "grad_norm": 0.16328173875808716,
+      "learning_rate": 1.4220790293763307e-05,
+      "loss": 0.4872,
+      "step": 4924
+    },
+    {
+      "epoch": 1.098594691055097,
+      "grad_norm": 0.16611479222774506,
+      "learning_rate": 1.4218656654369987e-05,
+      "loss": 0.4599,
+      "step": 4925
+    },
+    {
+      "epoch": 1.0988177559669863,
+      "grad_norm": 0.15916502475738525,
+      "learning_rate": 1.4216522781320928e-05,
+      "loss": 0.4709,
+      "step": 4926
+    },
+    {
+      "epoch": 1.0990408208788758,
+      "grad_norm": 0.18872345983982086,
+      "learning_rate": 1.4214388674734309e-05,
+      "loss": 0.469,
+      "step": 4927
+    },
+    {
+      "epoch": 1.099263885790765,
+      "grad_norm": 0.1624644547700882,
+      "learning_rate": 1.421225433472834e-05,
+      "loss": 0.4793,
+      "step": 4928
+    },
+    {
+      "epoch": 1.0994869507026546,
+      "grad_norm": 0.16488322615623474,
+      "learning_rate": 1.4210119761421228e-05,
+      "loss": 0.4534,
+      "step": 4929
+    },
+    {
+      "epoch": 1.0997100156145438,
+      "grad_norm": 0.15583010017871857,
+      "learning_rate": 1.4207984954931204e-05,
+      "loss": 0.4851,
+      "step": 4930
+    },
+    {
+      "epoch": 1.0999330805264331,
+      "grad_norm": 0.1624946892261505,
+      "learning_rate": 1.4205849915376501e-05,
+      "loss": 0.4915,
+      "step": 4931
+    },
+    {
+      "epoch": 1.1001561454383226,
+      "grad_norm": 0.15822146832942963,
+      "learning_rate": 1.4203714642875377e-05,
+      "loss": 0.4684,
+      "step": 4932
+    },
+    {
+      "epoch": 1.100379210350212,
+      "grad_norm": 0.16068744659423828,
+      "learning_rate": 1.420157913754609e-05,
+      "loss": 0.4555,
+      "step": 4933
+    },
+    {
+      "epoch": 1.1006022752621012,
+      "grad_norm": 0.18929767608642578,
+      "learning_rate": 1.4199443399506922e-05,
+      "loss": 0.4551,
+      "step": 4934
+    },
+    {
+      "epoch": 1.1008253401739907,
+      "grad_norm": 0.1583186835050583,
+      "learning_rate": 1.4197307428876164e-05,
+      "loss": 0.4449,
+      "step": 4935
+    },
+    {
+      "epoch": 1.10104840508588,
+      "grad_norm": 0.16161227226257324,
+      "learning_rate": 1.4195171225772117e-05,
+      "loss": 0.4557,
+      "step": 4936
+    },
+    {
+      "epoch": 1.1012714699977693,
+      "grad_norm": 0.16375921666622162,
+      "learning_rate": 1.4193034790313101e-05,
+      "loss": 0.4627,
+      "step": 4937
+    },
+    {
+      "epoch": 1.1014945349096588,
+      "grad_norm": 0.17420132458209991,
+      "learning_rate": 1.4190898122617443e-05,
+      "loss": 0.4736,
+      "step": 4938
+    },
+    {
+      "epoch": 1.101717599821548,
+      "grad_norm": 0.16097956895828247,
+      "learning_rate": 1.4188761222803482e-05,
+      "loss": 0.4511,
+      "step": 4939
+    },
+    {
+      "epoch": 1.1019406647334375,
+      "grad_norm": 0.16940614581108093,
+      "learning_rate": 1.4186624090989578e-05,
+      "loss": 0.4763,
+      "step": 4940
+    },
+    {
+      "epoch": 1.1021637296453268,
+      "grad_norm": 0.1643071174621582,
+      "learning_rate": 1.4184486727294098e-05,
+      "loss": 0.4904,
+      "step": 4941
+    },
+    {
+      "epoch": 1.102386794557216,
+      "grad_norm": 0.1605340987443924,
+      "learning_rate": 1.418234913183542e-05,
+      "loss": 0.4621,
+      "step": 4942
+    },
+    {
+      "epoch": 1.1026098594691056,
+      "grad_norm": 0.15705011785030365,
+      "learning_rate": 1.4180211304731941e-05,
+      "loss": 0.4779,
+      "step": 4943
+    },
+    {
+      "epoch": 1.102832924380995,
+      "grad_norm": 0.164706289768219,
+      "learning_rate": 1.4178073246102062e-05,
+      "loss": 0.4685,
+      "step": 4944
+    },
+    {
+      "epoch": 1.1030559892928842,
+      "grad_norm": 0.1530647873878479,
+      "learning_rate": 1.417593495606421e-05,
+      "loss": 0.4643,
+      "step": 4945
+    },
+    {
+      "epoch": 1.1032790542047737,
+      "grad_norm": 0.1623847335577011,
+      "learning_rate": 1.4173796434736808e-05,
+      "loss": 0.4606,
+      "step": 4946
+    },
+    {
+      "epoch": 1.103502119116663,
+      "grad_norm": 0.17122423648834229,
+      "learning_rate": 1.4171657682238309e-05,
+      "loss": 0.4649,
+      "step": 4947
+    },
+    {
+      "epoch": 1.1037251840285522,
+      "grad_norm": 0.15987126529216766,
+      "learning_rate": 1.4169518698687164e-05,
+      "loss": 0.469,
+      "step": 4948
+    },
+    {
+      "epoch": 1.1039482489404417,
+      "grad_norm": 0.1600853055715561,
+      "learning_rate": 1.416737948420185e-05,
+      "loss": 0.4782,
+      "step": 4949
+    },
+    {
+      "epoch": 1.104171313852331,
+      "grad_norm": 0.15356458723545074,
+      "learning_rate": 1.4165240038900843e-05,
+      "loss": 0.4474,
+      "step": 4950
+    },
+    {
+      "epoch": 1.1043943787642203,
+      "grad_norm": 0.15699787437915802,
+      "learning_rate": 1.4163100362902642e-05,
+      "loss": 0.4525,
+      "step": 4951
+    },
+    {
+      "epoch": 1.1046174436761098,
+      "grad_norm": 0.161276713013649,
+      "learning_rate": 1.4160960456325757e-05,
+      "loss": 0.4644,
+      "step": 4952
+    },
+    {
+      "epoch": 1.104840508587999,
+      "grad_norm": 0.17264524102210999,
+      "learning_rate": 1.4158820319288709e-05,
+      "loss": 0.4881,
+      "step": 4953
+    },
+    {
+      "epoch": 1.1050635734998884,
+      "grad_norm": 0.15942956507205963,
+      "learning_rate": 1.4156679951910031e-05,
+      "loss": 0.4431,
+      "step": 4954
+    },
+    {
+      "epoch": 1.1052866384117779,
+      "grad_norm": 0.16725444793701172,
+      "learning_rate": 1.415453935430827e-05,
+      "loss": 0.5334,
+      "step": 4955
+    },
+    {
+      "epoch": 1.1055097033236672,
+      "grad_norm": 0.15742219984531403,
+      "learning_rate": 1.4152398526601987e-05,
+      "loss": 0.4648,
+      "step": 4956
+    },
+    {
+      "epoch": 1.1057327682355567,
+      "grad_norm": 0.16164308786392212,
+      "learning_rate": 1.4150257468909753e-05,
+      "loss": 0.4746,
+      "step": 4957
+    },
+    {
+      "epoch": 1.105955833147446,
+      "grad_norm": 0.16626529395580292,
+      "learning_rate": 1.4148116181350155e-05,
+      "loss": 0.4763,
+      "step": 4958
+    },
+    {
+      "epoch": 1.1061788980593352,
+      "grad_norm": 0.16360588371753693,
+      "learning_rate": 1.4145974664041793e-05,
+      "loss": 0.4983,
+      "step": 4959
+    },
+    {
+      "epoch": 1.1064019629712247,
+      "grad_norm": 0.16253133118152618,
+      "learning_rate": 1.4143832917103271e-05,
+      "loss": 0.4742,
+      "step": 4960
+    },
+    {
+      "epoch": 1.106625027883114,
+      "grad_norm": 0.16334842145442963,
+      "learning_rate": 1.4141690940653217e-05,
+      "loss": 0.4921,
+      "step": 4961
+    },
+    {
+      "epoch": 1.1068480927950033,
+      "grad_norm": 0.1619054228067398,
+      "learning_rate": 1.4139548734810267e-05,
+      "loss": 0.485,
+      "step": 4962
+    },
+    {
+      "epoch": 1.1070711577068928,
+      "grad_norm": 0.16352368891239166,
+      "learning_rate": 1.4137406299693068e-05,
+      "loss": 0.4702,
+      "step": 4963
+    },
+    {
+      "epoch": 1.107294222618782,
+      "grad_norm": 0.16347824037075043,
+      "learning_rate": 1.4135263635420287e-05,
+      "loss": 0.48,
+      "step": 4964
+    },
+    {
+      "epoch": 1.1075172875306714,
+      "grad_norm": 0.15994015336036682,
+      "learning_rate": 1.4133120742110591e-05,
+      "loss": 0.477,
+      "step": 4965
+    },
+    {
+      "epoch": 1.1077403524425609,
+      "grad_norm": 0.172930508852005,
+      "learning_rate": 1.4130977619882673e-05,
+      "loss": 0.4882,
+      "step": 4966
+    },
+    {
+      "epoch": 1.1079634173544501,
+      "grad_norm": 0.16305923461914062,
+      "learning_rate": 1.4128834268855224e-05,
+      "loss": 0.4737,
+      "step": 4967
+    },
+    {
+      "epoch": 1.1081864822663394,
+      "grad_norm": 0.1601942926645279,
+      "learning_rate": 1.4126690689146967e-05,
+      "loss": 0.4876,
+      "step": 4968
+    },
+    {
+      "epoch": 1.108409547178229,
+      "grad_norm": 0.16627256572246552,
+      "learning_rate": 1.4124546880876617e-05,
+      "loss": 0.4928,
+      "step": 4969
+    },
+    {
+      "epoch": 1.1086326120901182,
+      "grad_norm": 0.16845837235450745,
+      "learning_rate": 1.4122402844162921e-05,
+      "loss": 0.4756,
+      "step": 4970
+    },
+    {
+      "epoch": 1.1088556770020075,
+      "grad_norm": 0.1565047651529312,
+      "learning_rate": 1.412025857912462e-05,
+      "loss": 0.4661,
+      "step": 4971
+    },
+    {
+      "epoch": 1.109078741913897,
+      "grad_norm": 0.15977446734905243,
+      "learning_rate": 1.4118114085880484e-05,
+      "loss": 0.4466,
+      "step": 4972
+    },
+    {
+      "epoch": 1.1093018068257863,
+      "grad_norm": 0.16255097091197968,
+      "learning_rate": 1.4115969364549288e-05,
+      "loss": 0.4677,
+      "step": 4973
+    },
+    {
+      "epoch": 1.1095248717376758,
+      "grad_norm": 0.15899838507175446,
+      "learning_rate": 1.4113824415249812e-05,
+      "loss": 0.4786,
+      "step": 4974
+    },
+    {
+      "epoch": 1.109747936649565,
+      "grad_norm": 0.15630389750003815,
+      "learning_rate": 1.4111679238100868e-05,
+      "loss": 0.4657,
+      "step": 4975
+    },
+    {
+      "epoch": 1.1099710015614543,
+      "grad_norm": 0.16901795566082,
+      "learning_rate": 1.4109533833221263e-05,
+      "loss": 0.4582,
+      "step": 4976
+    },
+    {
+      "epoch": 1.1101940664733438,
+      "grad_norm": 0.15667371451854706,
+      "learning_rate": 1.4107388200729824e-05,
+      "loss": 0.4686,
+      "step": 4977
+    },
+    {
+      "epoch": 1.1104171313852331,
+      "grad_norm": 0.1538134068250656,
+      "learning_rate": 1.4105242340745388e-05,
+      "loss": 0.4591,
+      "step": 4978
+    },
+    {
+      "epoch": 1.1106401962971224,
+      "grad_norm": 0.15832480788230896,
+      "learning_rate": 1.4103096253386812e-05,
+      "loss": 0.4598,
+      "step": 4979
+    },
+    {
+      "epoch": 1.110863261209012,
+      "grad_norm": 0.16176483035087585,
+      "learning_rate": 1.4100949938772953e-05,
+      "loss": 0.4959,
+      "step": 4980
+    },
+    {
+      "epoch": 1.1110863261209012,
+      "grad_norm": 0.15992896258831024,
+      "learning_rate": 1.4098803397022694e-05,
+      "loss": 0.4693,
+      "step": 4981
+    },
+    {
+      "epoch": 1.1113093910327905,
+      "grad_norm": 0.16486084461212158,
+      "learning_rate": 1.4096656628254916e-05,
+      "loss": 0.4687,
+      "step": 4982
+    },
+    {
+      "epoch": 1.11153245594468,
+      "grad_norm": 0.16968096792697906,
+      "learning_rate": 1.4094509632588528e-05,
+      "loss": 0.4648,
+      "step": 4983
+    },
+    {
+      "epoch": 1.1117555208565693,
+      "grad_norm": 0.1609363853931427,
+      "learning_rate": 1.409236241014244e-05,
+      "loss": 0.47,
+      "step": 4984
+    },
+    {
+      "epoch": 1.1119785857684585,
+      "grad_norm": 0.16171255707740784,
+      "learning_rate": 1.409021496103558e-05,
+      "loss": 0.4538,
+      "step": 4985
+    },
+    {
+      "epoch": 1.112201650680348,
+      "grad_norm": 0.1664259135723114,
+      "learning_rate": 1.4088067285386885e-05,
+      "loss": 0.4717,
+      "step": 4986
+    },
+    {
+      "epoch": 1.1124247155922373,
+      "grad_norm": 0.16782160103321075,
+      "learning_rate": 1.4085919383315311e-05,
+      "loss": 0.472,
+      "step": 4987
+    },
+    {
+      "epoch": 1.1126477805041266,
+      "grad_norm": 0.17342595756053925,
+      "learning_rate": 1.408377125493982e-05,
+      "loss": 0.5161,
+      "step": 4988
+    },
+    {
+      "epoch": 1.112870845416016,
+      "grad_norm": 0.17580246925354004,
+      "learning_rate": 1.408162290037939e-05,
+      "loss": 0.4874,
+      "step": 4989
+    },
+    {
+      "epoch": 1.1130939103279054,
+      "grad_norm": 0.16201059520244598,
+      "learning_rate": 1.4079474319753007e-05,
+      "loss": 0.4597,
+      "step": 4990
+    },
+    {
+      "epoch": 1.113316975239795,
+      "grad_norm": 0.1731184720993042,
+      "learning_rate": 1.4077325513179676e-05,
+      "loss": 0.5251,
+      "step": 4991
+    },
+    {
+      "epoch": 1.1135400401516842,
+      "grad_norm": 0.14824289083480835,
+      "learning_rate": 1.407517648077841e-05,
+      "loss": 0.4159,
+      "step": 4992
+    },
+    {
+      "epoch": 1.1137631050635735,
+      "grad_norm": 0.1582590788602829,
+      "learning_rate": 1.4073027222668236e-05,
+      "loss": 0.4813,
+      "step": 4993
+    },
+    {
+      "epoch": 1.113986169975463,
+      "grad_norm": 0.15586505830287933,
+      "learning_rate": 1.4070877738968196e-05,
+      "loss": 0.4634,
+      "step": 4994
+    },
+    {
+      "epoch": 1.1142092348873522,
+      "grad_norm": 0.15916606783866882,
+      "learning_rate": 1.4068728029797338e-05,
+      "loss": 0.4608,
+      "step": 4995
+    },
+    {
+      "epoch": 1.1144322997992415,
+      "grad_norm": 0.15859851241111755,
+      "learning_rate": 1.4066578095274732e-05,
+      "loss": 0.4701,
+      "step": 4996
+    },
+    {
+      "epoch": 1.114655364711131,
+      "grad_norm": 0.1668933480978012,
+      "learning_rate": 1.406442793551945e-05,
+      "loss": 0.4823,
+      "step": 4997
+    },
+    {
+      "epoch": 1.1148784296230203,
+      "grad_norm": 0.15839633345603943,
+      "learning_rate": 1.406227755065058e-05,
+      "loss": 0.4411,
+      "step": 4998
+    },
+    {
+      "epoch": 1.1151014945349096,
+      "grad_norm": 0.16563774645328522,
+      "learning_rate": 1.4060126940787228e-05,
+      "loss": 0.5158,
+      "step": 4999
+    },
+    {
+      "epoch": 1.115324559446799,
+      "grad_norm": 0.16948749125003815,
+      "learning_rate": 1.4057976106048509e-05,
+      "loss": 0.4735,
+      "step": 5000
+    },
+    {
+      "epoch": 1.1155476243586884,
+      "grad_norm": 0.15439462661743164,
+      "learning_rate": 1.4055825046553544e-05,
+      "loss": 0.4646,
+      "step": 5001
+    },
+    {
+      "epoch": 1.1157706892705777,
+      "grad_norm": 0.22990640997886658,
+      "learning_rate": 1.4053673762421478e-05,
+      "loss": 0.4645,
+      "step": 5002
+    },
+    {
+      "epoch": 1.1159937541824672,
+      "grad_norm": 0.16069011390209198,
+      "learning_rate": 1.4051522253771458e-05,
+      "loss": 0.4598,
+      "step": 5003
+    },
+    {
+      "epoch": 1.1162168190943564,
+      "grad_norm": 0.15775886178016663,
+      "learning_rate": 1.4049370520722657e-05,
+      "loss": 0.468,
+      "step": 5004
+    },
+    {
+      "epoch": 1.1164398840062457,
+      "grad_norm": 0.16639457643032074,
+      "learning_rate": 1.4047218563394238e-05,
+      "loss": 0.5046,
+      "step": 5005
+    },
+    {
+      "epoch": 1.1166629489181352,
+      "grad_norm": 0.16088278591632843,
+      "learning_rate": 1.40450663819054e-05,
+      "loss": 0.4801,
+      "step": 5006
+    },
+    {
+      "epoch": 1.1168860138300245,
+      "grad_norm": 0.16275745630264282,
+      "learning_rate": 1.404291397637534e-05,
+      "loss": 0.4568,
+      "step": 5007
+    },
+    {
+      "epoch": 1.117109078741914,
+      "grad_norm": 0.16248852014541626,
+      "learning_rate": 1.4040761346923275e-05,
+      "loss": 0.4617,
+      "step": 5008
+    },
+    {
+      "epoch": 1.1173321436538033,
+      "grad_norm": 0.15882933139801025,
+      "learning_rate": 1.4038608493668428e-05,
+      "loss": 0.4553,
+      "step": 5009
+    },
+    {
+      "epoch": 1.1175552085656926,
+      "grad_norm": 0.15445846319198608,
+      "learning_rate": 1.4036455416730038e-05,
+      "loss": 0.4623,
+      "step": 5010
+    },
+    {
+      "epoch": 1.117778273477582,
+      "grad_norm": 0.16165289282798767,
+      "learning_rate": 1.4034302116227358e-05,
+      "loss": 0.4522,
+      "step": 5011
+    },
+    {
+      "epoch": 1.1180013383894714,
+      "grad_norm": 0.16482336819171906,
+      "learning_rate": 1.4032148592279649e-05,
+      "loss": 0.4897,
+      "step": 5012
+    },
+    {
+      "epoch": 1.1182244033013606,
+      "grad_norm": 0.16417334973812103,
+      "learning_rate": 1.4029994845006187e-05,
+      "loss": 0.459,
+      "step": 5013
+    },
+    {
+      "epoch": 1.1184474682132501,
+      "grad_norm": 0.16407686471939087,
+      "learning_rate": 1.4027840874526262e-05,
+      "loss": 0.4656,
+      "step": 5014
+    },
+    {
+      "epoch": 1.1186705331251394,
+      "grad_norm": 0.1576424241065979,
+      "learning_rate": 1.4025686680959174e-05,
+      "loss": 0.4284,
+      "step": 5015
+    },
+    {
+      "epoch": 1.1188935980370287,
+      "grad_norm": 0.15783734619617462,
+      "learning_rate": 1.4023532264424233e-05,
+      "loss": 0.4609,
+      "step": 5016
+    },
+    {
+      "epoch": 1.1191166629489182,
+      "grad_norm": 0.1539250910282135,
+      "learning_rate": 1.4021377625040768e-05,
+      "loss": 0.4507,
+      "step": 5017
+    },
+    {
+      "epoch": 1.1193397278608075,
+      "grad_norm": 0.16160206496715546,
+      "learning_rate": 1.4019222762928113e-05,
+      "loss": 0.456,
+      "step": 5018
+    },
+    {
+      "epoch": 1.1195627927726968,
+      "grad_norm": 0.15390300750732422,
+      "learning_rate": 1.4017067678205623e-05,
+      "loss": 0.4637,
+      "step": 5019
+    },
+    {
+      "epoch": 1.1197858576845863,
+      "grad_norm": 0.15983295440673828,
+      "learning_rate": 1.4014912370992653e-05,
+      "loss": 0.4809,
+      "step": 5020
+    },
+    {
+      "epoch": 1.1200089225964756,
+      "grad_norm": 0.16199488937854767,
+      "learning_rate": 1.4012756841408583e-05,
+      "loss": 0.4433,
+      "step": 5021
+    },
+    {
+      "epoch": 1.1202319875083648,
+      "grad_norm": 0.1602371335029602,
+      "learning_rate": 1.4010601089572794e-05,
+      "loss": 0.4472,
+      "step": 5022
+    },
+    {
+      "epoch": 1.1204550524202543,
+      "grad_norm": 0.16461622714996338,
+      "learning_rate": 1.4008445115604694e-05,
+      "loss": 0.4709,
+      "step": 5023
+    },
+    {
+      "epoch": 1.1206781173321436,
+      "grad_norm": 0.16210167109966278,
+      "learning_rate": 1.4006288919623687e-05,
+      "loss": 0.4569,
+      "step": 5024
+    },
+    {
+      "epoch": 1.1209011822440331,
+      "grad_norm": 0.15823470056056976,
+      "learning_rate": 1.4004132501749198e-05,
+      "loss": 0.4794,
+      "step": 5025
+    },
+    {
+      "epoch": 1.1211242471559224,
+      "grad_norm": 0.155581995844841,
+      "learning_rate": 1.4001975862100668e-05,
+      "loss": 0.4442,
+      "step": 5026
+    },
+    {
+      "epoch": 1.1213473120678117,
+      "grad_norm": 0.1608058661222458,
+      "learning_rate": 1.3999819000797539e-05,
+      "loss": 0.4535,
+      "step": 5027
+    },
+    {
+      "epoch": 1.1215703769797012,
+      "grad_norm": 0.16367916762828827,
+      "learning_rate": 1.3997661917959273e-05,
+      "loss": 0.4581,
+      "step": 5028
+    },
+    {
+      "epoch": 1.1217934418915905,
+      "grad_norm": 0.1618059128522873,
+      "learning_rate": 1.3995504613705344e-05,
+      "loss": 0.464,
+      "step": 5029
+    },
+    {
+      "epoch": 1.1220165068034798,
+      "grad_norm": 0.15621179342269897,
+      "learning_rate": 1.3993347088155237e-05,
+      "loss": 0.4573,
+      "step": 5030
+    },
+    {
+      "epoch": 1.1222395717153693,
+      "grad_norm": 0.1720859855413437,
+      "learning_rate": 1.399118934142845e-05,
+      "loss": 0.4854,
+      "step": 5031
+    },
+    {
+      "epoch": 1.1224626366272585,
+      "grad_norm": 0.16994187235832214,
+      "learning_rate": 1.3989031373644491e-05,
+      "loss": 0.5074,
+      "step": 5032
+    },
+    {
+      "epoch": 1.1226857015391478,
+      "grad_norm": 0.20645354688167572,
+      "learning_rate": 1.3986873184922882e-05,
+      "loss": 0.4498,
+      "step": 5033
+    },
+    {
+      "epoch": 1.1229087664510373,
+      "grad_norm": 0.1670864075422287,
+      "learning_rate": 1.3984714775383159e-05,
+      "loss": 0.4719,
+      "step": 5034
+    },
+    {
+      "epoch": 1.1231318313629266,
+      "grad_norm": 0.1660386323928833,
+      "learning_rate": 1.3982556145144866e-05,
+      "loss": 0.4869,
+      "step": 5035
+    },
+    {
+      "epoch": 1.1233548962748159,
+      "grad_norm": 0.16415423154830933,
+      "learning_rate": 1.3980397294327563e-05,
+      "loss": 0.4806,
+      "step": 5036
+    },
+    {
+      "epoch": 1.1235779611867054,
+      "grad_norm": 0.16121268272399902,
+      "learning_rate": 1.3978238223050817e-05,
+      "loss": 0.4677,
+      "step": 5037
+    },
+    {
+      "epoch": 1.1238010260985947,
+      "grad_norm": 0.16271083056926727,
+      "learning_rate": 1.3976078931434219e-05,
+      "loss": 0.4749,
+      "step": 5038
+    },
+    {
+      "epoch": 1.124024091010484,
+      "grad_norm": 0.16719701886177063,
+      "learning_rate": 1.3973919419597354e-05,
+      "loss": 0.4537,
+      "step": 5039
+    },
+    {
+      "epoch": 1.1242471559223735,
+      "grad_norm": 0.15598838031291962,
+      "learning_rate": 1.3971759687659841e-05,
+      "loss": 0.4623,
+      "step": 5040
+    },
+    {
+      "epoch": 1.1244702208342627,
+      "grad_norm": 0.1647387146949768,
+      "learning_rate": 1.3969599735741288e-05,
+      "loss": 0.4879,
+      "step": 5041
+    },
+    {
+      "epoch": 1.1246932857461522,
+      "grad_norm": 0.15755842626094818,
+      "learning_rate": 1.3967439563961334e-05,
+      "loss": 0.4564,
+      "step": 5042
+    },
+    {
+      "epoch": 1.1249163506580415,
+      "grad_norm": 0.1570240706205368,
+      "learning_rate": 1.396527917243962e-05,
+      "loss": 0.4593,
+      "step": 5043
+    },
+    {
+      "epoch": 1.1251394155699308,
+      "grad_norm": 0.6927107572555542,
+      "learning_rate": 1.3963118561295803e-05,
+      "loss": 0.4905,
+      "step": 5044
+    },
+    {
+      "epoch": 1.1253624804818203,
+      "grad_norm": 0.17146988213062286,
+      "learning_rate": 1.3960957730649551e-05,
+      "loss": 0.4754,
+      "step": 5045
+    },
+    {
+      "epoch": 1.1255855453937096,
+      "grad_norm": 0.16707734763622284,
+      "learning_rate": 1.3958796680620545e-05,
+      "loss": 0.4578,
+      "step": 5046
+    },
+    {
+      "epoch": 1.1258086103055989,
+      "grad_norm": 0.15594236552715302,
+      "learning_rate": 1.3956635411328478e-05,
+      "loss": 0.4556,
+      "step": 5047
+    },
+    {
+      "epoch": 1.1260316752174884,
+      "grad_norm": 0.15944088995456696,
+      "learning_rate": 1.395447392289305e-05,
+      "loss": 0.4478,
+      "step": 5048
+    },
+    {
+      "epoch": 1.1262547401293777,
+      "grad_norm": 0.16137385368347168,
+      "learning_rate": 1.3952312215433987e-05,
+      "loss": 0.4922,
+      "step": 5049
+    },
+    {
+      "epoch": 1.126477805041267,
+      "grad_norm": 0.18752697110176086,
+      "learning_rate": 1.3950150289071007e-05,
+      "loss": 0.4641,
+      "step": 5050
+    },
+    {
+      "epoch": 1.1267008699531564,
+      "grad_norm": 0.16448596119880676,
+      "learning_rate": 1.394798814392386e-05,
+      "loss": 0.475,
+      "step": 5051
+    },
+    {
+      "epoch": 1.1269239348650457,
+      "grad_norm": 0.16754111647605896,
+      "learning_rate": 1.3945825780112294e-05,
+      "loss": 0.4709,
+      "step": 5052
+    },
+    {
+      "epoch": 1.127146999776935,
+      "grad_norm": 0.165597066283226,
+      "learning_rate": 1.394366319775608e-05,
+      "loss": 0.4761,
+      "step": 5053
+    },
+    {
+      "epoch": 1.1273700646888245,
+      "grad_norm": 0.16928669810295105,
+      "learning_rate": 1.3941500396974984e-05,
+      "loss": 0.5111,
+      "step": 5054
+    },
+    {
+      "epoch": 1.1275931296007138,
+      "grad_norm": 0.16309335827827454,
+      "learning_rate": 1.3939337377888808e-05,
+      "loss": 0.4638,
+      "step": 5055
+    },
+    {
+      "epoch": 1.127816194512603,
+      "grad_norm": 0.16362158954143524,
+      "learning_rate": 1.3937174140617349e-05,
+      "loss": 0.4496,
+      "step": 5056
+    },
+    {
+      "epoch": 1.1280392594244926,
+      "grad_norm": 0.15970171988010406,
+      "learning_rate": 1.3935010685280417e-05,
+      "loss": 0.4583,
+      "step": 5057
+    },
+    {
+      "epoch": 1.1282623243363818,
+      "grad_norm": 0.16845880448818207,
+      "learning_rate": 1.3932847011997846e-05,
+      "loss": 0.4705,
+      "step": 5058
+    },
+    {
+      "epoch": 1.1284853892482714,
+      "grad_norm": 0.1621864140033722,
+      "learning_rate": 1.3930683120889463e-05,
+      "loss": 0.4659,
+      "step": 5059
+    },
+    {
+      "epoch": 1.1287084541601606,
+      "grad_norm": 0.16252627968788147,
+      "learning_rate": 1.3928519012075128e-05,
+      "loss": 0.4628,
+      "step": 5060
+    },
+    {
+      "epoch": 1.12893151907205,
+      "grad_norm": 0.1627832055091858,
+      "learning_rate": 1.3926354685674697e-05,
+      "loss": 0.4617,
+      "step": 5061
+    },
+    {
+      "epoch": 1.1291545839839394,
+      "grad_norm": 0.1639028936624527,
+      "learning_rate": 1.3924190141808048e-05,
+      "loss": 0.4626,
+      "step": 5062
+    },
+    {
+      "epoch": 1.1293776488958287,
+      "grad_norm": 0.1629188507795334,
+      "learning_rate": 1.3922025380595061e-05,
+      "loss": 0.4905,
+      "step": 5063
+    },
+    {
+      "epoch": 1.129600713807718,
+      "grad_norm": 0.16841430962085724,
+      "learning_rate": 1.391986040215564e-05,
+      "loss": 0.4512,
+      "step": 5064
+    },
+    {
+      "epoch": 1.1298237787196075,
+      "grad_norm": 0.1598411500453949,
+      "learning_rate": 1.3917695206609693e-05,
+      "loss": 0.4453,
+      "step": 5065
+    },
+    {
+      "epoch": 1.1300468436314968,
+      "grad_norm": 0.15308506786823273,
+      "learning_rate": 1.3915529794077142e-05,
+      "loss": 0.4563,
+      "step": 5066
+    },
+    {
+      "epoch": 1.130269908543386,
+      "grad_norm": 0.16865313053131104,
+      "learning_rate": 1.3913364164677922e-05,
+      "loss": 0.4587,
+      "step": 5067
+    },
+    {
+      "epoch": 1.1304929734552756,
+      "grad_norm": 0.1975971907377243,
+      "learning_rate": 1.3911198318531977e-05,
+      "loss": 0.4523,
+      "step": 5068
+    },
+    {
+      "epoch": 1.1307160383671648,
+      "grad_norm": 0.1655057817697525,
+      "learning_rate": 1.3909032255759267e-05,
+      "loss": 0.4446,
+      "step": 5069
+    },
+    {
+      "epoch": 1.1309391032790541,
+      "grad_norm": 0.16097813844680786,
+      "learning_rate": 1.3906865976479766e-05,
+      "loss": 0.4828,
+      "step": 5070
+    },
+    {
+      "epoch": 1.1311621681909436,
+      "grad_norm": 0.15927664935588837,
+      "learning_rate": 1.3904699480813446e-05,
+      "loss": 0.4715,
+      "step": 5071
+    },
+    {
+      "epoch": 1.131385233102833,
+      "grad_norm": 0.15957841277122498,
+      "learning_rate": 1.3902532768880313e-05,
+      "loss": 0.4638,
+      "step": 5072
+    },
+    {
+      "epoch": 1.1316082980147222,
+      "grad_norm": 0.16548919677734375,
+      "learning_rate": 1.3900365840800363e-05,
+      "loss": 0.495,
+      "step": 5073
+    },
+    {
+      "epoch": 1.1318313629266117,
+      "grad_norm": 0.1837669163942337,
+      "learning_rate": 1.3898198696693621e-05,
+      "loss": 0.4866,
+      "step": 5074
+    },
+    {
+      "epoch": 1.132054427838501,
+      "grad_norm": 0.16433456540107727,
+      "learning_rate": 1.3896031336680111e-05,
+      "loss": 0.4637,
+      "step": 5075
+    },
+    {
+      "epoch": 1.1322774927503905,
+      "grad_norm": 0.15830080211162567,
+      "learning_rate": 1.3893863760879882e-05,
+      "loss": 0.4644,
+      "step": 5076
+    },
+    {
+      "epoch": 1.1325005576622797,
+      "grad_norm": 0.16347894072532654,
+      "learning_rate": 1.3891695969412982e-05,
+      "loss": 0.4764,
+      "step": 5077
+    },
+    {
+      "epoch": 1.132723622574169,
+      "grad_norm": 0.15911467373371124,
+      "learning_rate": 1.388952796239948e-05,
+      "loss": 0.4585,
+      "step": 5078
+    },
+    {
+      "epoch": 1.1329466874860585,
+      "grad_norm": 0.16895873844623566,
+      "learning_rate": 1.3887359739959455e-05,
+      "loss": 0.4787,
+      "step": 5079
+    },
+    {
+      "epoch": 1.1331697523979478,
+      "grad_norm": 0.16661864519119263,
+      "learning_rate": 1.3885191302212993e-05,
+      "loss": 0.4987,
+      "step": 5080
+    },
+    {
+      "epoch": 1.133392817309837,
+      "grad_norm": 0.16188743710517883,
+      "learning_rate": 1.38830226492802e-05,
+      "loss": 0.4428,
+      "step": 5081
+    },
+    {
+      "epoch": 1.1336158822217266,
+      "grad_norm": 0.1691635102033615,
+      "learning_rate": 1.3880853781281187e-05,
+      "loss": 0.4713,
+      "step": 5082
+    },
+    {
+      "epoch": 1.1338389471336159,
+      "grad_norm": 0.16617491841316223,
+      "learning_rate": 1.387868469833608e-05,
+      "loss": 0.4687,
+      "step": 5083
+    },
+    {
+      "epoch": 1.1340620120455052,
+      "grad_norm": 0.17212171852588654,
+      "learning_rate": 1.3876515400565016e-05,
+      "loss": 0.4912,
+      "step": 5084
+    },
+    {
+      "epoch": 1.1342850769573947,
+      "grad_norm": 0.1617995649576187,
+      "learning_rate": 1.3874345888088145e-05,
+      "loss": 0.4807,
+      "step": 5085
+    },
+    {
+      "epoch": 1.134508141869284,
+      "grad_norm": 0.16481629014015198,
+      "learning_rate": 1.3872176161025627e-05,
+      "loss": 0.4652,
+      "step": 5086
+    },
+    {
+      "epoch": 1.1347312067811732,
+      "grad_norm": 0.16548612713813782,
+      "learning_rate": 1.3870006219497642e-05,
+      "loss": 0.4517,
+      "step": 5087
+    },
+    {
+      "epoch": 1.1349542716930627,
+      "grad_norm": 0.17424030601978302,
+      "learning_rate": 1.3867836063624363e-05,
+      "loss": 0.4804,
+      "step": 5088
+    },
+    {
+      "epoch": 1.135177336604952,
+      "grad_norm": 0.17292533814907074,
+      "learning_rate": 1.3865665693525994e-05,
+      "loss": 0.4776,
+      "step": 5089
+    },
+    {
+      "epoch": 1.1354004015168413,
+      "grad_norm": 0.1664893478155136,
+      "learning_rate": 1.3863495109322744e-05,
+      "loss": 0.4653,
+      "step": 5090
+    },
+    {
+      "epoch": 1.1356234664287308,
+      "grad_norm": 0.16548456251621246,
+      "learning_rate": 1.3861324311134832e-05,
+      "loss": 0.4709,
+      "step": 5091
+    },
+    {
+      "epoch": 1.13584653134062,
+      "grad_norm": 0.1643359214067459,
+      "learning_rate": 1.3859153299082493e-05,
+      "loss": 0.4361,
+      "step": 5092
+    },
+    {
+      "epoch": 1.1360695962525096,
+      "grad_norm": 0.17173811793327332,
+      "learning_rate": 1.3856982073285965e-05,
+      "loss": 0.4482,
+      "step": 5093
+    },
+    {
+      "epoch": 1.1362926611643989,
+      "grad_norm": 0.1580163687467575,
+      "learning_rate": 1.3854810633865512e-05,
+      "loss": 0.4395,
+      "step": 5094
+    },
+    {
+      "epoch": 1.1365157260762881,
+      "grad_norm": 0.16304674744606018,
+      "learning_rate": 1.3852638980941398e-05,
+      "loss": 0.4754,
+      "step": 5095
+    },
+    {
+      "epoch": 1.1367387909881776,
+      "grad_norm": 0.16688407957553864,
+      "learning_rate": 1.38504671146339e-05,
+      "loss": 0.5081,
+      "step": 5096
+    },
+    {
+      "epoch": 1.136961855900067,
+      "grad_norm": 0.16496771574020386,
+      "learning_rate": 1.3848295035063317e-05,
+      "loss": 0.4537,
+      "step": 5097
+    },
+    {
+      "epoch": 1.1371849208119562,
+      "grad_norm": 0.16779771447181702,
+      "learning_rate": 1.3846122742349946e-05,
+      "loss": 0.4608,
+      "step": 5098
+    },
+    {
+      "epoch": 1.1374079857238457,
+      "grad_norm": 0.1593242734670639,
+      "learning_rate": 1.3843950236614103e-05,
+      "loss": 0.4761,
+      "step": 5099
+    },
+    {
+      "epoch": 1.137631050635735,
+      "grad_norm": 0.16632351279258728,
+      "learning_rate": 1.384177751797612e-05,
+      "loss": 0.4723,
+      "step": 5100
+    },
+    {
+      "epoch": 1.1378541155476243,
+      "grad_norm": 0.16760534048080444,
+      "learning_rate": 1.383960458655633e-05,
+      "loss": 0.4509,
+      "step": 5101
+    },
+    {
+      "epoch": 1.1380771804595138,
+      "grad_norm": 0.16486233472824097,
+      "learning_rate": 1.3837431442475089e-05,
+      "loss": 0.4806,
+      "step": 5102
+    },
+    {
+      "epoch": 1.138300245371403,
+      "grad_norm": 0.1571461409330368,
+      "learning_rate": 1.3835258085852752e-05,
+      "loss": 0.4564,
+      "step": 5103
+    },
+    {
+      "epoch": 1.1385233102832923,
+      "grad_norm": 0.16537711024284363,
+      "learning_rate": 1.38330845168097e-05,
+      "loss": 0.4556,
+      "step": 5104
+    },
+    {
+      "epoch": 1.1387463751951818,
+      "grad_norm": 0.1652860939502716,
+      "learning_rate": 1.3830910735466313e-05,
+      "loss": 0.447,
+      "step": 5105
+    },
+    {
+      "epoch": 1.1389694401070711,
+      "grad_norm": 0.15692059695720673,
+      "learning_rate": 1.3828736741942998e-05,
+      "loss": 0.4566,
+      "step": 5106
+    },
+    {
+      "epoch": 1.1391925050189604,
+      "grad_norm": 0.16122755408287048,
+      "learning_rate": 1.3826562536360155e-05,
+      "loss": 0.4692,
+      "step": 5107
+    },
+    {
+      "epoch": 1.13941556993085,
+      "grad_norm": 0.1586706042289734,
+      "learning_rate": 1.382438811883821e-05,
+      "loss": 0.4316,
+      "step": 5108
+    },
+    {
+      "epoch": 1.1396386348427392,
+      "grad_norm": 0.16200071573257446,
+      "learning_rate": 1.3822213489497594e-05,
+      "loss": 0.4704,
+      "step": 5109
+    },
+    {
+      "epoch": 1.1398616997546287,
+      "grad_norm": 0.5188888311386108,
+      "learning_rate": 1.3820038648458748e-05,
+      "loss": 0.49,
+      "step": 5110
+    },
+    {
+      "epoch": 1.140084764666518,
+      "grad_norm": 0.15934044122695923,
+      "learning_rate": 1.3817863595842138e-05,
+      "loss": 0.4456,
+      "step": 5111
+    },
+    {
+      "epoch": 1.1403078295784073,
+      "grad_norm": 0.17257168889045715,
+      "learning_rate": 1.3815688331768224e-05,
+      "loss": 0.4655,
+      "step": 5112
+    },
+    {
+      "epoch": 1.1405308944902968,
+      "grad_norm": 0.16849681735038757,
+      "learning_rate": 1.3813512856357491e-05,
+      "loss": 0.4461,
+      "step": 5113
+    },
+    {
+      "epoch": 1.140753959402186,
+      "grad_norm": 0.16747885942459106,
+      "learning_rate": 1.3811337169730428e-05,
+      "loss": 0.473,
+      "step": 5114
+    },
+    {
+      "epoch": 1.1409770243140753,
+      "grad_norm": 0.1711420714855194,
+      "learning_rate": 1.3809161272007536e-05,
+      "loss": 0.4851,
+      "step": 5115
+    },
+    {
+      "epoch": 1.1412000892259648,
+      "grad_norm": 0.1751970797777176,
+      "learning_rate": 1.3806985163309334e-05,
+      "loss": 0.4955,
+      "step": 5116
+    },
+    {
+      "epoch": 1.1414231541378541,
+      "grad_norm": 0.16653122007846832,
+      "learning_rate": 1.3804808843756348e-05,
+      "loss": 0.4697,
+      "step": 5117
+    },
+    {
+      "epoch": 1.1416462190497434,
+      "grad_norm": 0.158418208360672,
+      "learning_rate": 1.3802632313469111e-05,
+      "loss": 0.433,
+      "step": 5118
+    },
+    {
+      "epoch": 1.141869283961633,
+      "grad_norm": 0.16599524021148682,
+      "learning_rate": 1.3800455572568182e-05,
+      "loss": 0.49,
+      "step": 5119
+    },
+    {
+      "epoch": 1.1420923488735222,
+      "grad_norm": 0.17085126042366028,
+      "learning_rate": 1.3798278621174113e-05,
+      "loss": 0.4914,
+      "step": 5120
+    },
+    {
+      "epoch": 1.1423154137854115,
+      "grad_norm": 0.16169953346252441,
+      "learning_rate": 1.3796101459407485e-05,
+      "loss": 0.4851,
+      "step": 5121
+    },
+    {
+      "epoch": 1.142538478697301,
+      "grad_norm": 0.19225676357746124,
+      "learning_rate": 1.3793924087388876e-05,
+      "loss": 0.4554,
+      "step": 5122
+    },
+    {
+      "epoch": 1.1427615436091902,
+      "grad_norm": 0.15610942244529724,
+      "learning_rate": 1.379174650523889e-05,
+      "loss": 0.4436,
+      "step": 5123
+    },
+    {
+      "epoch": 1.1429846085210795,
+      "grad_norm": 0.16388073563575745,
+      "learning_rate": 1.3789568713078129e-05,
+      "loss": 0.4647,
+      "step": 5124
+    },
+    {
+      "epoch": 1.143207673432969,
+      "grad_norm": 0.15931613743305206,
+      "learning_rate": 1.3787390711027217e-05,
+      "loss": 0.4707,
+      "step": 5125
+    },
+    {
+      "epoch": 1.1434307383448583,
+      "grad_norm": 0.1648622304201126,
+      "learning_rate": 1.3785212499206783e-05,
+      "loss": 0.4628,
+      "step": 5126
+    },
+    {
+      "epoch": 1.1436538032567478,
+      "grad_norm": 0.16240182518959045,
+      "learning_rate": 1.3783034077737472e-05,
+      "loss": 0.4735,
+      "step": 5127
+    },
+    {
+      "epoch": 1.143876868168637,
+      "grad_norm": 0.15895162522792816,
+      "learning_rate": 1.3780855446739937e-05,
+      "loss": 0.4587,
+      "step": 5128
+    },
+    {
+      "epoch": 1.1440999330805264,
+      "grad_norm": 0.1622496396303177,
+      "learning_rate": 1.3778676606334844e-05,
+      "loss": 0.4541,
+      "step": 5129
+    },
+    {
+      "epoch": 1.1443229979924159,
+      "grad_norm": 0.15852504968643188,
+      "learning_rate": 1.3776497556642874e-05,
+      "loss": 0.4546,
+      "step": 5130
+    },
+    {
+      "epoch": 1.1445460629043052,
+      "grad_norm": 0.16722345352172852,
+      "learning_rate": 1.377431829778471e-05,
+      "loss": 0.4754,
+      "step": 5131
+    },
+    {
+      "epoch": 1.1447691278161944,
+      "grad_norm": 0.1600501835346222,
+      "learning_rate": 1.377213882988106e-05,
+      "loss": 0.4851,
+      "step": 5132
+    },
+    {
+      "epoch": 1.144992192728084,
+      "grad_norm": 0.1717759370803833,
+      "learning_rate": 1.3769959153052634e-05,
+      "loss": 0.4581,
+      "step": 5133
+    },
+    {
+      "epoch": 1.1452152576399732,
+      "grad_norm": 0.1694609820842743,
+      "learning_rate": 1.3767779267420158e-05,
+      "loss": 0.4584,
+      "step": 5134
+    },
+    {
+      "epoch": 1.1454383225518625,
+      "grad_norm": 0.16004528105258942,
+      "learning_rate": 1.3765599173104362e-05,
+      "loss": 0.4768,
+      "step": 5135
+    },
+    {
+      "epoch": 1.145661387463752,
+      "grad_norm": 0.15713921189308167,
+      "learning_rate": 1.3763418870225999e-05,
+      "loss": 0.4515,
+      "step": 5136
+    },
+    {
+      "epoch": 1.1458844523756413,
+      "grad_norm": 0.16554328799247742,
+      "learning_rate": 1.3761238358905826e-05,
+      "loss": 0.4856,
+      "step": 5137
+    },
+    {
+      "epoch": 1.1461075172875306,
+      "grad_norm": 0.1595989465713501,
+      "learning_rate": 1.3759057639264614e-05,
+      "loss": 0.4695,
+      "step": 5138
+    },
+    {
+      "epoch": 1.14633058219942,
+      "grad_norm": 0.16599664092063904,
+      "learning_rate": 1.3756876711423143e-05,
+      "loss": 0.4821,
+      "step": 5139
+    },
+    {
+      "epoch": 1.1465536471113094,
+      "grad_norm": 0.15603865683078766,
+      "learning_rate": 1.3754695575502211e-05,
+      "loss": 0.4469,
+      "step": 5140
+    },
+    {
+      "epoch": 1.1467767120231986,
+      "grad_norm": 0.16003453731536865,
+      "learning_rate": 1.3752514231622617e-05,
+      "loss": 0.4619,
+      "step": 5141
+    },
+    {
+      "epoch": 1.1469997769350881,
+      "grad_norm": 0.15717166662216187,
+      "learning_rate": 1.375033267990518e-05,
+      "loss": 0.4585,
+      "step": 5142
+    },
+    {
+      "epoch": 1.1472228418469774,
+      "grad_norm": 0.15859000384807587,
+      "learning_rate": 1.374815092047073e-05,
+      "loss": 0.4446,
+      "step": 5143
+    },
+    {
+      "epoch": 1.147445906758867,
+      "grad_norm": 0.16028450429439545,
+      "learning_rate": 1.3745968953440105e-05,
+      "loss": 0.4512,
+      "step": 5144
+    },
+    {
+      "epoch": 1.1476689716707562,
+      "grad_norm": 0.1540936380624771,
+      "learning_rate": 1.3743786778934158e-05,
+      "loss": 0.4286,
+      "step": 5145
+    },
+    {
+      "epoch": 1.1478920365826455,
+      "grad_norm": 0.15674646198749542,
+      "learning_rate": 1.3741604397073748e-05,
+      "loss": 0.469,
+      "step": 5146
+    },
+    {
+      "epoch": 1.148115101494535,
+      "grad_norm": 0.16240215301513672,
+      "learning_rate": 1.3739421807979753e-05,
+      "loss": 0.4609,
+      "step": 5147
+    },
+    {
+      "epoch": 1.1483381664064243,
+      "grad_norm": 0.1656564623117447,
+      "learning_rate": 1.3737239011773054e-05,
+      "loss": 0.4829,
+      "step": 5148
+    },
+    {
+      "epoch": 1.1485612313183136,
+      "grad_norm": 0.16485963761806488,
+      "learning_rate": 1.3735056008574551e-05,
+      "loss": 0.4494,
+      "step": 5149
+    },
+    {
+      "epoch": 1.148784296230203,
+      "grad_norm": 0.16284818947315216,
+      "learning_rate": 1.3732872798505153e-05,
+      "loss": 0.4504,
+      "step": 5150
+    },
+    {
+      "epoch": 1.1490073611420923,
+      "grad_norm": 0.21305973827838898,
+      "learning_rate": 1.373068938168578e-05,
+      "loss": 0.4496,
+      "step": 5151
+    },
+    {
+      "epoch": 1.1492304260539816,
+      "grad_norm": 0.16297850012779236,
+      "learning_rate": 1.3728505758237358e-05,
+      "loss": 0.4799,
+      "step": 5152
+    },
+    {
+      "epoch": 1.1494534909658711,
+      "grad_norm": 0.16585685312747955,
+      "learning_rate": 1.3726321928280837e-05,
+      "loss": 0.4608,
+      "step": 5153
+    },
+    {
+      "epoch": 1.1496765558777604,
+      "grad_norm": 0.16437405347824097,
+      "learning_rate": 1.3724137891937167e-05,
+      "loss": 0.4812,
+      "step": 5154
+    },
+    {
+      "epoch": 1.14989962078965,
+      "grad_norm": 0.15935295820236206,
+      "learning_rate": 1.3721953649327316e-05,
+      "loss": 0.4633,
+      "step": 5155
+    },
+    {
+      "epoch": 1.1501226857015392,
+      "grad_norm": 0.15900611877441406,
+      "learning_rate": 1.3719769200572258e-05,
+      "loss": 0.4401,
+      "step": 5156
+    },
+    {
+      "epoch": 1.1503457506134285,
+      "grad_norm": 0.172005295753479,
+      "learning_rate": 1.3717584545792983e-05,
+      "loss": 0.4634,
+      "step": 5157
+    },
+    {
+      "epoch": 1.1505688155253178,
+      "grad_norm": 0.16333134472370148,
+      "learning_rate": 1.3715399685110492e-05,
+      "loss": 0.4391,
+      "step": 5158
+    },
+    {
+      "epoch": 1.1507918804372073,
+      "grad_norm": 0.16098088026046753,
+      "learning_rate": 1.3713214618645796e-05,
+      "loss": 0.4484,
+      "step": 5159
+    },
+    {
+      "epoch": 1.1510149453490965,
+      "grad_norm": 0.17146125435829163,
+      "learning_rate": 1.3711029346519917e-05,
+      "loss": 0.4606,
+      "step": 5160
+    },
+    {
+      "epoch": 1.151238010260986,
+      "grad_norm": 0.16574114561080933,
+      "learning_rate": 1.3708843868853889e-05,
+      "loss": 0.4862,
+      "step": 5161
+    },
+    {
+      "epoch": 1.1514610751728753,
+      "grad_norm": 0.168543741106987,
+      "learning_rate": 1.370665818576876e-05,
+      "loss": 0.4608,
+      "step": 5162
+    },
+    {
+      "epoch": 1.1516841400847646,
+      "grad_norm": 0.16033907234668732,
+      "learning_rate": 1.3704472297385583e-05,
+      "loss": 0.439,
+      "step": 5163
+    },
+    {
+      "epoch": 1.151907204996654,
+      "grad_norm": 0.16206419467926025,
+      "learning_rate": 1.3702286203825429e-05,
+      "loss": 0.4707,
+      "step": 5164
+    },
+    {
+      "epoch": 1.1521302699085434,
+      "grad_norm": 0.160027876496315,
+      "learning_rate": 1.3700099905209374e-05,
+      "loss": 0.4603,
+      "step": 5165
+    },
+    {
+      "epoch": 1.1523533348204327,
+      "grad_norm": 0.17743688821792603,
+      "learning_rate": 1.3697913401658516e-05,
+      "loss": 0.5053,
+      "step": 5166
+    },
+    {
+      "epoch": 1.1525763997323222,
+      "grad_norm": 0.15926334261894226,
+      "learning_rate": 1.3695726693293951e-05,
+      "loss": 0.4685,
+      "step": 5167
+    },
+    {
+      "epoch": 1.1527994646442115,
+      "grad_norm": 0.18017716705799103,
+      "learning_rate": 1.3693539780236798e-05,
+      "loss": 0.4864,
+      "step": 5168
+    },
+    {
+      "epoch": 1.1530225295561007,
+      "grad_norm": 0.1655382215976715,
+      "learning_rate": 1.3691352662608175e-05,
+      "loss": 0.4607,
+      "step": 5169
+    },
+    {
+      "epoch": 1.1532455944679902,
+      "grad_norm": 0.16363778710365295,
+      "learning_rate": 1.3689165340529222e-05,
+      "loss": 0.4425,
+      "step": 5170
+    },
+    {
+      "epoch": 1.1534686593798795,
+      "grad_norm": 0.1638006567955017,
+      "learning_rate": 1.3686977814121087e-05,
+      "loss": 0.4428,
+      "step": 5171
+    },
+    {
+      "epoch": 1.153691724291769,
+      "grad_norm": 0.1594667136669159,
+      "learning_rate": 1.368479008350493e-05,
+      "loss": 0.4417,
+      "step": 5172
+    },
+    {
+      "epoch": 1.1539147892036583,
+      "grad_norm": 0.1636561006307602,
+      "learning_rate": 1.3682602148801917e-05,
+      "loss": 0.4602,
+      "step": 5173
+    },
+    {
+      "epoch": 1.1541378541155476,
+      "grad_norm": 0.1705152690410614,
+      "learning_rate": 1.3680414010133237e-05,
+      "loss": 0.4494,
+      "step": 5174
+    },
+    {
+      "epoch": 1.1543609190274369,
+      "grad_norm": 0.1863928884267807,
+      "learning_rate": 1.3678225667620075e-05,
+      "loss": 0.4592,
+      "step": 5175
+    },
+    {
+      "epoch": 1.1545839839393264,
+      "grad_norm": 0.1695355474948883,
+      "learning_rate": 1.3676037121383638e-05,
+      "loss": 0.5064,
+      "step": 5176
+    },
+    {
+      "epoch": 1.1548070488512157,
+      "grad_norm": 0.16180136799812317,
+      "learning_rate": 1.3673848371545145e-05,
+      "loss": 0.473,
+      "step": 5177
+    },
+    {
+      "epoch": 1.1550301137631052,
+      "grad_norm": 0.18443214893341064,
+      "learning_rate": 1.3671659418225815e-05,
+      "loss": 0.4872,
+      "step": 5178
+    },
+    {
+      "epoch": 1.1552531786749944,
+      "grad_norm": 0.16797620058059692,
+      "learning_rate": 1.3669470261546896e-05,
+      "loss": 0.4538,
+      "step": 5179
+    },
+    {
+      "epoch": 1.1554762435868837,
+      "grad_norm": 0.16609326004981995,
+      "learning_rate": 1.3667280901629627e-05,
+      "loss": 0.5077,
+      "step": 5180
+    },
+    {
+      "epoch": 1.1556993084987732,
+      "grad_norm": 0.16129672527313232,
+      "learning_rate": 1.3665091338595277e-05,
+      "loss": 0.4544,
+      "step": 5181
+    },
+    {
+      "epoch": 1.1559223734106625,
+      "grad_norm": 0.16736097633838654,
+      "learning_rate": 1.3662901572565114e-05,
+      "loss": 0.4633,
+      "step": 5182
+    },
+    {
+      "epoch": 1.1561454383225518,
+      "grad_norm": 0.21419832110404968,
+      "learning_rate": 1.3660711603660422e-05,
+      "loss": 0.4508,
+      "step": 5183
+    },
+    {
+      "epoch": 1.1563685032344413,
+      "grad_norm": 0.16937695443630219,
+      "learning_rate": 1.3658521432002494e-05,
+      "loss": 0.4729,
+      "step": 5184
+    },
+    {
+      "epoch": 1.1565915681463306,
+      "grad_norm": 0.15548932552337646,
+      "learning_rate": 1.3656331057712637e-05,
+      "loss": 0.4829,
+      "step": 5185
+    },
+    {
+      "epoch": 1.1568146330582199,
+      "grad_norm": 0.17126090824604034,
+      "learning_rate": 1.3654140480912164e-05,
+      "loss": 0.4715,
+      "step": 5186
+    },
+    {
+      "epoch": 1.1570376979701094,
+      "grad_norm": 0.18009397387504578,
+      "learning_rate": 1.3651949701722407e-05,
+      "loss": 0.4883,
+      "step": 5187
+    },
+    {
+      "epoch": 1.1572607628819986,
+      "grad_norm": 0.171475350856781,
+      "learning_rate": 1.3649758720264705e-05,
+      "loss": 0.4411,
+      "step": 5188
+    },
+    {
+      "epoch": 1.1574838277938881,
+      "grad_norm": 0.17869600653648376,
+      "learning_rate": 1.3647567536660407e-05,
+      "loss": 0.4841,
+      "step": 5189
+    },
+    {
+      "epoch": 1.1577068927057774,
+      "grad_norm": 0.1822163164615631,
+      "learning_rate": 1.3645376151030871e-05,
+      "loss": 0.4595,
+      "step": 5190
+    },
+    {
+      "epoch": 1.1579299576176667,
+      "grad_norm": 0.17729108035564423,
+      "learning_rate": 1.3643184563497479e-05,
+      "loss": 0.4605,
+      "step": 5191
+    },
+    {
+      "epoch": 1.158153022529556,
+      "grad_norm": 0.16725416481494904,
+      "learning_rate": 1.3640992774181605e-05,
+      "loss": 0.4375,
+      "step": 5192
+    },
+    {
+      "epoch": 1.1583760874414455,
+      "grad_norm": 0.17714278399944305,
+      "learning_rate": 1.3638800783204653e-05,
+      "loss": 0.4944,
+      "step": 5193
+    },
+    {
+      "epoch": 1.1585991523533348,
+      "grad_norm": 0.159885436296463,
+      "learning_rate": 1.3636608590688019e-05,
+      "loss": 0.4793,
+      "step": 5194
+    },
+    {
+      "epoch": 1.1588222172652243,
+      "grad_norm": 0.16276371479034424,
+      "learning_rate": 1.363441619675313e-05,
+      "loss": 0.4872,
+      "step": 5195
+    },
+    {
+      "epoch": 1.1590452821771136,
+      "grad_norm": 0.16415032744407654,
+      "learning_rate": 1.3632223601521409e-05,
+      "loss": 0.4632,
+      "step": 5196
+    },
+    {
+      "epoch": 1.1592683470890028,
+      "grad_norm": 0.1709449738264084,
+      "learning_rate": 1.3630030805114297e-05,
+      "loss": 0.485,
+      "step": 5197
+    },
+    {
+      "epoch": 1.1594914120008923,
+      "grad_norm": 0.16452427208423615,
+      "learning_rate": 1.3627837807653249e-05,
+      "loss": 0.4306,
+      "step": 5198
+    },
+    {
+      "epoch": 1.1597144769127816,
+      "grad_norm": 0.18321935832500458,
+      "learning_rate": 1.3625644609259716e-05,
+      "loss": 0.4982,
+      "step": 5199
+    },
+    {
+      "epoch": 1.159937541824671,
+      "grad_norm": 0.1770065277814865,
+      "learning_rate": 1.3623451210055186e-05,
+      "loss": 0.496,
+      "step": 5200
+    },
+    {
+      "epoch": 1.1601606067365604,
+      "grad_norm": 0.16729894280433655,
+      "learning_rate": 1.3621257610161129e-05,
+      "loss": 0.4844,
+      "step": 5201
+    },
+    {
+      "epoch": 1.1603836716484497,
+      "grad_norm": 0.1584504395723343,
+      "learning_rate": 1.3619063809699054e-05,
+      "loss": 0.4801,
+      "step": 5202
+    },
+    {
+      "epoch": 1.160606736560339,
+      "grad_norm": 0.15807606279850006,
+      "learning_rate": 1.3616869808790453e-05,
+      "loss": 0.4623,
+      "step": 5203
+    },
+    {
+      "epoch": 1.1608298014722285,
+      "grad_norm": 0.15968340635299683,
+      "learning_rate": 1.3614675607556857e-05,
+      "loss": 0.4693,
+      "step": 5204
+    },
+    {
+      "epoch": 1.1610528663841178,
+      "grad_norm": 0.16744199395179749,
+      "learning_rate": 1.3612481206119786e-05,
+      "loss": 0.4821,
+      "step": 5205
+    },
+    {
+      "epoch": 1.1612759312960073,
+      "grad_norm": 0.16137133538722992,
+      "learning_rate": 1.3610286604600782e-05,
+      "loss": 0.4757,
+      "step": 5206
+    },
+    {
+      "epoch": 1.1614989962078965,
+      "grad_norm": 0.1617872714996338,
+      "learning_rate": 1.3608091803121397e-05,
+      "loss": 0.4703,
+      "step": 5207
+    },
+    {
+      "epoch": 1.1617220611197858,
+      "grad_norm": 0.16734254360198975,
+      "learning_rate": 1.360589680180319e-05,
+      "loss": 0.4706,
+      "step": 5208
+    },
+    {
+      "epoch": 1.161945126031675,
+      "grad_norm": 0.163192480802536,
+      "learning_rate": 1.3603701600767741e-05,
+      "loss": 0.4722,
+      "step": 5209
+    },
+    {
+      "epoch": 1.1621681909435646,
+      "grad_norm": 0.16038168966770172,
+      "learning_rate": 1.3601506200136624e-05,
+      "loss": 0.4443,
+      "step": 5210
+    },
+    {
+      "epoch": 1.1623912558554539,
+      "grad_norm": 0.16467641294002533,
+      "learning_rate": 1.3599310600031443e-05,
+      "loss": 0.4465,
+      "step": 5211
+    },
+    {
+      "epoch": 1.1626143207673434,
+      "grad_norm": 0.1640377789735794,
+      "learning_rate": 1.3597114800573799e-05,
+      "loss": 0.4722,
+      "step": 5212
+    },
+    {
+      "epoch": 1.1628373856792327,
+      "grad_norm": 0.16441082954406738,
+      "learning_rate": 1.359491880188531e-05,
+      "loss": 0.4617,
+      "step": 5213
+    },
+    {
+      "epoch": 1.163060450591122,
+      "grad_norm": 0.16120347380638123,
+      "learning_rate": 1.3592722604087604e-05,
+      "loss": 0.4464,
+      "step": 5214
+    },
+    {
+      "epoch": 1.1632835155030115,
+      "grad_norm": 0.17573657631874084,
+      "learning_rate": 1.3590526207302324e-05,
+      "loss": 0.5029,
+      "step": 5215
+    },
+    {
+      "epoch": 1.1635065804149007,
+      "grad_norm": 0.16203901171684265,
+      "learning_rate": 1.3588329611651117e-05,
+      "loss": 0.4654,
+      "step": 5216
+    },
+    {
+      "epoch": 1.16372964532679,
+      "grad_norm": 0.1684379279613495,
+      "learning_rate": 1.3586132817255644e-05,
+      "loss": 0.4706,
+      "step": 5217
+    },
+    {
+      "epoch": 1.1639527102386795,
+      "grad_norm": 0.15993818640708923,
+      "learning_rate": 1.3583935824237576e-05,
+      "loss": 0.4587,
+      "step": 5218
+    },
+    {
+      "epoch": 1.1641757751505688,
+      "grad_norm": 0.1687501221895218,
+      "learning_rate": 1.35817386327186e-05,
+      "loss": 0.4467,
+      "step": 5219
+    },
+    {
+      "epoch": 1.164398840062458,
+      "grad_norm": 0.16527238488197327,
+      "learning_rate": 1.3579541242820407e-05,
+      "loss": 0.4617,
+      "step": 5220
+    },
+    {
+      "epoch": 1.1646219049743476,
+      "grad_norm": 0.16829682886600494,
+      "learning_rate": 1.3577343654664705e-05,
+      "loss": 0.4582,
+      "step": 5221
+    },
+    {
+      "epoch": 1.1648449698862369,
+      "grad_norm": 0.186698317527771,
+      "learning_rate": 1.3575145868373207e-05,
+      "loss": 0.4878,
+      "step": 5222
+    },
+    {
+      "epoch": 1.1650680347981264,
+      "grad_norm": 0.19950494170188904,
+      "learning_rate": 1.3572947884067644e-05,
+      "loss": 0.4636,
+      "step": 5223
+    },
+    {
+      "epoch": 1.1652910997100157,
+      "grad_norm": 0.16333967447280884,
+      "learning_rate": 1.3570749701869751e-05,
+      "loss": 0.4772,
+      "step": 5224
+    },
+    {
+      "epoch": 1.165514164621905,
+      "grad_norm": 0.16258838772773743,
+      "learning_rate": 1.3568551321901282e-05,
+      "loss": 0.4968,
+      "step": 5225
+    },
+    {
+      "epoch": 1.1657372295337942,
+      "grad_norm": 0.18643935024738312,
+      "learning_rate": 1.356635274428399e-05,
+      "loss": 0.4937,
+      "step": 5226
+    },
+    {
+      "epoch": 1.1659602944456837,
+      "grad_norm": 0.1727738380432129,
+      "learning_rate": 1.3564153969139654e-05,
+      "loss": 0.4915,
+      "step": 5227
+    },
+    {
+      "epoch": 1.166183359357573,
+      "grad_norm": 0.16102762520313263,
+      "learning_rate": 1.3561954996590047e-05,
+      "loss": 0.4534,
+      "step": 5228
+    },
+    {
+      "epoch": 1.1664064242694625,
+      "grad_norm": 0.17399117350578308,
+      "learning_rate": 1.3559755826756968e-05,
+      "loss": 0.4775,
+      "step": 5229
+    },
+    {
+      "epoch": 1.1666294891813518,
+      "grad_norm": 0.1659899652004242,
+      "learning_rate": 1.3557556459762223e-05,
+      "loss": 0.4624,
+      "step": 5230
+    },
+    {
+      "epoch": 1.166852554093241,
+      "grad_norm": 0.17996659874916077,
+      "learning_rate": 1.3555356895727618e-05,
+      "loss": 0.4938,
+      "step": 5231
+    },
+    {
+      "epoch": 1.1670756190051306,
+      "grad_norm": 0.16586771607398987,
+      "learning_rate": 1.355315713477499e-05,
+      "loss": 0.4788,
+      "step": 5232
+    },
+    {
+      "epoch": 1.1672986839170199,
+      "grad_norm": 0.1771279126405716,
+      "learning_rate": 1.3550957177026164e-05,
+      "loss": 0.4752,
+      "step": 5233
+    },
+    {
+      "epoch": 1.1675217488289091,
+      "grad_norm": 0.17474165558815002,
+      "learning_rate": 1.3548757022602997e-05,
+      "loss": 0.4734,
+      "step": 5234
+    },
+    {
+      "epoch": 1.1677448137407986,
+      "grad_norm": 0.16958124935626984,
+      "learning_rate": 1.3546556671627341e-05,
+      "loss": 0.4713,
+      "step": 5235
+    },
+    {
+      "epoch": 1.167967878652688,
+      "grad_norm": 0.16799892485141754,
+      "learning_rate": 1.354435612422107e-05,
+      "loss": 0.4604,
+      "step": 5236
+    },
+    {
+      "epoch": 1.1681909435645772,
+      "grad_norm": 0.17070333659648895,
+      "learning_rate": 1.3542155380506059e-05,
+      "loss": 0.4706,
+      "step": 5237
+    },
+    {
+      "epoch": 1.1684140084764667,
+      "grad_norm": 0.1612580418586731,
+      "learning_rate": 1.3539954440604206e-05,
+      "loss": 0.4619,
+      "step": 5238
+    },
+    {
+      "epoch": 1.168637073388356,
+      "grad_norm": 0.16792915761470795,
+      "learning_rate": 1.3537753304637406e-05,
+      "loss": 0.4564,
+      "step": 5239
+    },
+    {
+      "epoch": 1.1688601383002455,
+      "grad_norm": 0.16356982290744781,
+      "learning_rate": 1.3535551972727577e-05,
+      "loss": 0.4843,
+      "step": 5240
+    },
+    {
+      "epoch": 1.1690832032121348,
+      "grad_norm": 0.161970853805542,
+      "learning_rate": 1.3533350444996636e-05,
+      "loss": 0.4573,
+      "step": 5241
+    },
+    {
+      "epoch": 1.169306268124024,
+      "grad_norm": 0.17388132214546204,
+      "learning_rate": 1.3531148721566525e-05,
+      "loss": 0.4739,
+      "step": 5242
+    },
+    {
+      "epoch": 1.1695293330359133,
+      "grad_norm": 0.1602068692445755,
+      "learning_rate": 1.3528946802559184e-05,
+      "loss": 0.4584,
+      "step": 5243
+    },
+    {
+      "epoch": 1.1697523979478028,
+      "grad_norm": 0.1629067063331604,
+      "learning_rate": 1.3526744688096574e-05,
+      "loss": 0.4653,
+      "step": 5244
+    },
+    {
+      "epoch": 1.1699754628596921,
+      "grad_norm": 0.16808199882507324,
+      "learning_rate": 1.3524542378300658e-05,
+      "loss": 0.4588,
+      "step": 5245
+    },
+    {
+      "epoch": 1.1701985277715816,
+      "grad_norm": 0.17957061529159546,
+      "learning_rate": 1.3522339873293416e-05,
+      "loss": 0.49,
+      "step": 5246
+    },
+    {
+      "epoch": 1.170421592683471,
+      "grad_norm": 0.15982255339622498,
+      "learning_rate": 1.3520137173196832e-05,
+      "loss": 0.4456,
+      "step": 5247
+    },
+    {
+      "epoch": 1.1706446575953602,
+      "grad_norm": 0.16950003802776337,
+      "learning_rate": 1.3517934278132909e-05,
+      "loss": 0.4802,
+      "step": 5248
+    },
+    {
+      "epoch": 1.1708677225072497,
+      "grad_norm": 0.16196782886981964,
+      "learning_rate": 1.351573118822366e-05,
+      "loss": 0.4905,
+      "step": 5249
+    },
+    {
+      "epoch": 1.171090787419139,
+      "grad_norm": 0.16096629202365875,
+      "learning_rate": 1.3513527903591101e-05,
+      "loss": 0.4829,
+      "step": 5250
+    },
+    {
+      "epoch": 1.1713138523310282,
+      "grad_norm": 0.18132293224334717,
+      "learning_rate": 1.3511324424357269e-05,
+      "loss": 0.485,
+      "step": 5251
+    },
+    {
+      "epoch": 1.1715369172429178,
+      "grad_norm": 0.18736086785793304,
+      "learning_rate": 1.3509120750644198e-05,
+      "loss": 0.4569,
+      "step": 5252
+    },
+    {
+      "epoch": 1.171759982154807,
+      "grad_norm": 0.1668289601802826,
+      "learning_rate": 1.350691688257395e-05,
+      "loss": 0.4785,
+      "step": 5253
+    },
+    {
+      "epoch": 1.1719830470666963,
+      "grad_norm": 0.16315414011478424,
+      "learning_rate": 1.3504712820268584e-05,
+      "loss": 0.4545,
+      "step": 5254
+    },
+    {
+      "epoch": 1.1722061119785858,
+      "grad_norm": 0.1799056977033615,
+      "learning_rate": 1.3502508563850179e-05,
+      "loss": 0.5051,
+      "step": 5255
+    },
+    {
+      "epoch": 1.172429176890475,
+      "grad_norm": 0.16646715998649597,
+      "learning_rate": 1.3500304113440815e-05,
+      "loss": 0.4531,
+      "step": 5256
+    },
+    {
+      "epoch": 1.1726522418023646,
+      "grad_norm": 0.16029830276966095,
+      "learning_rate": 1.3498099469162594e-05,
+      "loss": 0.4614,
+      "step": 5257
+    },
+    {
+      "epoch": 1.1728753067142539,
+      "grad_norm": 0.16910234093666077,
+      "learning_rate": 1.3495894631137618e-05,
+      "loss": 0.484,
+      "step": 5258
+    },
+    {
+      "epoch": 1.1730983716261432,
+      "grad_norm": 0.16509772837162018,
+      "learning_rate": 1.349368959948801e-05,
+      "loss": 0.4458,
+      "step": 5259
+    },
+    {
+      "epoch": 1.1733214365380324,
+      "grad_norm": 0.16832312941551208,
+      "learning_rate": 1.3491484374335893e-05,
+      "loss": 0.4574,
+      "step": 5260
+    },
+    {
+      "epoch": 1.173544501449922,
+      "grad_norm": 0.15936195850372314,
+      "learning_rate": 1.3489278955803409e-05,
+      "loss": 0.4441,
+      "step": 5261
+    },
+    {
+      "epoch": 1.1737675663618112,
+      "grad_norm": 0.17124436795711517,
+      "learning_rate": 1.348707334401271e-05,
+      "loss": 0.4525,
+      "step": 5262
+    },
+    {
+      "epoch": 1.1739906312737007,
+      "grad_norm": 0.16587725281715393,
+      "learning_rate": 1.3484867539085952e-05,
+      "loss": 0.4683,
+      "step": 5263
+    },
+    {
+      "epoch": 1.17421369618559,
+      "grad_norm": 0.16889216005802155,
+      "learning_rate": 1.3482661541145315e-05,
+      "loss": 0.4646,
+      "step": 5264
+    },
+    {
+      "epoch": 1.1744367610974793,
+      "grad_norm": 0.16621263325214386,
+      "learning_rate": 1.3480455350312968e-05,
+      "loss": 0.4614,
+      "step": 5265
+    },
+    {
+      "epoch": 1.1746598260093688,
+      "grad_norm": 0.1614251285791397,
+      "learning_rate": 1.3478248966711115e-05,
+      "loss": 0.4549,
+      "step": 5266
+    },
+    {
+      "epoch": 1.174882890921258,
+      "grad_norm": 0.16062335669994354,
+      "learning_rate": 1.3476042390461954e-05,
+      "loss": 0.4592,
+      "step": 5267
+    },
+    {
+      "epoch": 1.1751059558331474,
+      "grad_norm": 0.16389133036136627,
+      "learning_rate": 1.3473835621687701e-05,
+      "loss": 0.4582,
+      "step": 5268
+    },
+    {
+      "epoch": 1.1753290207450369,
+      "grad_norm": 0.16292013227939606,
+      "learning_rate": 1.3471628660510576e-05,
+      "loss": 0.4444,
+      "step": 5269
+    },
+    {
+      "epoch": 1.1755520856569261,
+      "grad_norm": 0.16953033208847046,
+      "learning_rate": 1.3469421507052824e-05,
+      "loss": 0.4735,
+      "step": 5270
+    },
+    {
+      "epoch": 1.1757751505688154,
+      "grad_norm": 0.1629514843225479,
+      "learning_rate": 1.346721416143668e-05,
+      "loss": 0.4534,
+      "step": 5271
+    },
+    {
+      "epoch": 1.175998215480705,
+      "grad_norm": 0.1695055365562439,
+      "learning_rate": 1.346500662378441e-05,
+      "loss": 0.4498,
+      "step": 5272
+    },
+    {
+      "epoch": 1.1762212803925942,
+      "grad_norm": 0.1747894585132599,
+      "learning_rate": 1.3462798894218278e-05,
+      "loss": 0.4773,
+      "step": 5273
+    },
+    {
+      "epoch": 1.1764443453044837,
+      "grad_norm": 0.16201353073120117,
+      "learning_rate": 1.3460590972860561e-05,
+      "loss": 0.4401,
+      "step": 5274
+    },
+    {
+      "epoch": 1.176667410216373,
+      "grad_norm": 0.16511163115501404,
+      "learning_rate": 1.3458382859833545e-05,
+      "loss": 0.4652,
+      "step": 5275
+    },
+    {
+      "epoch": 1.1768904751282623,
+      "grad_norm": 0.17246244847774506,
+      "learning_rate": 1.3456174555259535e-05,
+      "loss": 0.4619,
+      "step": 5276
+    },
+    {
+      "epoch": 1.1771135400401516,
+      "grad_norm": 0.16072800755500793,
+      "learning_rate": 1.3453966059260836e-05,
+      "loss": 0.473,
+      "step": 5277
+    },
+    {
+      "epoch": 1.177336604952041,
+      "grad_norm": 0.16221876442432404,
+      "learning_rate": 1.345175737195977e-05,
+      "loss": 0.4588,
+      "step": 5278
+    },
+    {
+      "epoch": 1.1775596698639303,
+      "grad_norm": 0.16264645755290985,
+      "learning_rate": 1.344954849347867e-05,
+      "loss": 0.4536,
+      "step": 5279
+    },
+    {
+      "epoch": 1.1777827347758199,
+      "grad_norm": 0.1796528697013855,
+      "learning_rate": 1.3447339423939876e-05,
+      "loss": 0.4775,
+      "step": 5280
+    },
+    {
+      "epoch": 1.1780057996877091,
+      "grad_norm": 0.1555919647216797,
+      "learning_rate": 1.3445130163465739e-05,
+      "loss": 0.4576,
+      "step": 5281
+    },
+    {
+      "epoch": 1.1782288645995984,
+      "grad_norm": 0.16730302572250366,
+      "learning_rate": 1.3442920712178622e-05,
+      "loss": 0.4591,
+      "step": 5282
+    },
+    {
+      "epoch": 1.178451929511488,
+      "grad_norm": 0.16881461441516876,
+      "learning_rate": 1.3440711070200903e-05,
+      "loss": 0.4516,
+      "step": 5283
+    },
+    {
+      "epoch": 1.1786749944233772,
+      "grad_norm": 0.1790027767419815,
+      "learning_rate": 1.3438501237654958e-05,
+      "loss": 0.4627,
+      "step": 5284
+    },
+    {
+      "epoch": 1.1788980593352665,
+      "grad_norm": 0.1679370403289795,
+      "learning_rate": 1.3436291214663186e-05,
+      "loss": 0.4759,
+      "step": 5285
+    },
+    {
+      "epoch": 1.179121124247156,
+      "grad_norm": 0.17167186737060547,
+      "learning_rate": 1.3434081001347992e-05,
+      "loss": 0.4687,
+      "step": 5286
+    },
+    {
+      "epoch": 1.1793441891590453,
+      "grad_norm": 0.1699121594429016,
+      "learning_rate": 1.3431870597831792e-05,
+      "loss": 0.4634,
+      "step": 5287
+    },
+    {
+      "epoch": 1.1795672540709345,
+      "grad_norm": 0.15923888981342316,
+      "learning_rate": 1.3429660004237008e-05,
+      "loss": 0.4293,
+      "step": 5288
+    },
+    {
+      "epoch": 1.179790318982824,
+      "grad_norm": 0.15723690390586853,
+      "learning_rate": 1.3427449220686085e-05,
+      "loss": 0.4476,
+      "step": 5289
+    },
+    {
+      "epoch": 1.1800133838947133,
+      "grad_norm": 0.16163286566734314,
+      "learning_rate": 1.342523824730146e-05,
+      "loss": 0.4554,
+      "step": 5290
+    },
+    {
+      "epoch": 1.1802364488066028,
+      "grad_norm": 0.16615921258926392,
+      "learning_rate": 1.3423027084205597e-05,
+      "loss": 0.4751,
+      "step": 5291
+    },
+    {
+      "epoch": 1.1804595137184921,
+      "grad_norm": 0.1613457053899765,
+      "learning_rate": 1.3420815731520965e-05,
+      "loss": 0.462,
+      "step": 5292
+    },
+    {
+      "epoch": 1.1806825786303814,
+      "grad_norm": 0.16468282043933868,
+      "learning_rate": 1.3418604189370034e-05,
+      "loss": 0.4773,
+      "step": 5293
+    },
+    {
+      "epoch": 1.1809056435422707,
+      "grad_norm": 0.1656079739332199,
+      "learning_rate": 1.3416392457875302e-05,
+      "loss": 0.4593,
+      "step": 5294
+    },
+    {
+      "epoch": 1.1811287084541602,
+      "grad_norm": 0.15908241271972656,
+      "learning_rate": 1.3414180537159265e-05,
+      "loss": 0.4527,
+      "step": 5295
+    },
+    {
+      "epoch": 1.1813517733660495,
+      "grad_norm": 0.181619331240654,
+      "learning_rate": 1.3411968427344438e-05,
+      "loss": 0.4779,
+      "step": 5296
+    },
+    {
+      "epoch": 1.181574838277939,
+      "grad_norm": 0.16857735812664032,
+      "learning_rate": 1.3409756128553331e-05,
+      "loss": 0.4994,
+      "step": 5297
+    },
+    {
+      "epoch": 1.1817979031898282,
+      "grad_norm": 0.16610047221183777,
+      "learning_rate": 1.3407543640908485e-05,
+      "loss": 0.4669,
+      "step": 5298
+    },
+    {
+      "epoch": 1.1820209681017175,
+      "grad_norm": 0.16683466732501984,
+      "learning_rate": 1.3405330964532437e-05,
+      "loss": 0.4855,
+      "step": 5299
+    },
+    {
+      "epoch": 1.182244033013607,
+      "grad_norm": 0.1574050635099411,
+      "learning_rate": 1.340311809954774e-05,
+      "loss": 0.4539,
+      "step": 5300
+    },
+    {
+      "epoch": 1.1824670979254963,
+      "grad_norm": 0.16250169277191162,
+      "learning_rate": 1.3400905046076955e-05,
+      "loss": 0.4827,
+      "step": 5301
+    },
+    {
+      "epoch": 1.1826901628373856,
+      "grad_norm": 0.179072305560112,
+      "learning_rate": 1.3398691804242658e-05,
+      "loss": 0.4821,
+      "step": 5302
+    },
+    {
+      "epoch": 1.182913227749275,
+      "grad_norm": 0.15977934002876282,
+      "learning_rate": 1.339647837416743e-05,
+      "loss": 0.436,
+      "step": 5303
+    },
+    {
+      "epoch": 1.1831362926611644,
+      "grad_norm": 0.16551627218723297,
+      "learning_rate": 1.3394264755973864e-05,
+      "loss": 0.473,
+      "step": 5304
+    },
+    {
+      "epoch": 1.1833593575730537,
+      "grad_norm": 0.16227102279663086,
+      "learning_rate": 1.3392050949784566e-05,
+      "loss": 0.4693,
+      "step": 5305
+    },
+    {
+      "epoch": 1.1835824224849432,
+      "grad_norm": 0.16538846492767334,
+      "learning_rate": 1.3389836955722149e-05,
+      "loss": 0.4523,
+      "step": 5306
+    },
+    {
+      "epoch": 1.1838054873968324,
+      "grad_norm": 0.1663748174905777,
+      "learning_rate": 1.3387622773909237e-05,
+      "loss": 0.4656,
+      "step": 5307
+    },
+    {
+      "epoch": 1.184028552308722,
+      "grad_norm": 0.1712309867143631,
+      "learning_rate": 1.338540840446847e-05,
+      "loss": 0.4611,
+      "step": 5308
+    },
+    {
+      "epoch": 1.1842516172206112,
+      "grad_norm": 0.1841256320476532,
+      "learning_rate": 1.3383193847522487e-05,
+      "loss": 0.4955,
+      "step": 5309
+    },
+    {
+      "epoch": 1.1844746821325005,
+      "grad_norm": 0.151309996843338,
+      "learning_rate": 1.338097910319395e-05,
+      "loss": 0.4323,
+      "step": 5310
+    },
+    {
+      "epoch": 1.1846977470443898,
+      "grad_norm": 0.16778849065303802,
+      "learning_rate": 1.337876417160552e-05,
+      "loss": 0.4633,
+      "step": 5311
+    },
+    {
+      "epoch": 1.1849208119562793,
+      "grad_norm": 0.16986416280269623,
+      "learning_rate": 1.3376549052879883e-05,
+      "loss": 0.4243,
+      "step": 5312
+    },
+    {
+      "epoch": 1.1851438768681686,
+      "grad_norm": 0.1710985153913498,
+      "learning_rate": 1.3374333747139714e-05,
+      "loss": 0.4569,
+      "step": 5313
+    },
+    {
+      "epoch": 1.185366941780058,
+      "grad_norm": 0.1675567626953125,
+      "learning_rate": 1.3372118254507718e-05,
+      "loss": 0.4747,
+      "step": 5314
+    },
+    {
+      "epoch": 1.1855900066919474,
+      "grad_norm": 0.19594267010688782,
+      "learning_rate": 1.3369902575106604e-05,
+      "loss": 0.4645,
+      "step": 5315
+    },
+    {
+      "epoch": 1.1858130716038366,
+      "grad_norm": 0.1636292189359665,
+      "learning_rate": 1.3367686709059084e-05,
+      "loss": 0.4467,
+      "step": 5316
+    },
+    {
+      "epoch": 1.1860361365157261,
+      "grad_norm": 0.16308671236038208,
+      "learning_rate": 1.3365470656487896e-05,
+      "loss": 0.4704,
+      "step": 5317
+    },
+    {
+      "epoch": 1.1862592014276154,
+      "grad_norm": 0.16027814149856567,
+      "learning_rate": 1.3363254417515769e-05,
+      "loss": 0.4726,
+      "step": 5318
+    },
+    {
+      "epoch": 1.1864822663395047,
+      "grad_norm": 0.16369383037090302,
+      "learning_rate": 1.336103799226546e-05,
+      "loss": 0.4467,
+      "step": 5319
+    },
+    {
+      "epoch": 1.1867053312513942,
+      "grad_norm": 0.1744934618473053,
+      "learning_rate": 1.335882138085972e-05,
+      "loss": 0.4677,
+      "step": 5320
+    },
+    {
+      "epoch": 1.1869283961632835,
+      "grad_norm": 0.17109525203704834,
+      "learning_rate": 1.3356604583421331e-05,
+      "loss": 0.4587,
+      "step": 5321
+    },
+    {
+      "epoch": 1.1871514610751728,
+      "grad_norm": 0.17207783460617065,
+      "learning_rate": 1.335438760007306e-05,
+      "loss": 0.4468,
+      "step": 5322
+    },
+    {
+      "epoch": 1.1873745259870623,
+      "grad_norm": 0.16258476674556732,
+      "learning_rate": 1.3352170430937707e-05,
+      "loss": 0.4498,
+      "step": 5323
+    },
+    {
+      "epoch": 1.1875975908989516,
+      "grad_norm": 0.1659199595451355,
+      "learning_rate": 1.334995307613807e-05,
+      "loss": 0.4492,
+      "step": 5324
+    },
+    {
+      "epoch": 1.187820655810841,
+      "grad_norm": 0.18148742616176605,
+      "learning_rate": 1.3347735535796957e-05,
+      "loss": 0.4785,
+      "step": 5325
+    },
+    {
+      "epoch": 1.1880437207227303,
+      "grad_norm": 0.16647182404994965,
+      "learning_rate": 1.3345517810037194e-05,
+      "loss": 0.4375,
+      "step": 5326
+    },
+    {
+      "epoch": 1.1882667856346196,
+      "grad_norm": 0.16515585780143738,
+      "learning_rate": 1.334329989898161e-05,
+      "loss": 0.4653,
+      "step": 5327
+    },
+    {
+      "epoch": 1.1884898505465091,
+      "grad_norm": 0.161190927028656,
+      "learning_rate": 1.3341081802753046e-05,
+      "loss": 0.4448,
+      "step": 5328
+    },
+    {
+      "epoch": 1.1887129154583984,
+      "grad_norm": 0.1584840565919876,
+      "learning_rate": 1.333886352147436e-05,
+      "loss": 0.446,
+      "step": 5329
+    },
+    {
+      "epoch": 1.1889359803702877,
+      "grad_norm": 0.16086257994174957,
+      "learning_rate": 1.3336645055268405e-05,
+      "loss": 0.461,
+      "step": 5330
+    },
+    {
+      "epoch": 1.1891590452821772,
+      "grad_norm": 0.17658580839633942,
+      "learning_rate": 1.333442640425806e-05,
+      "loss": 0.4905,
+      "step": 5331
+    },
+    {
+      "epoch": 1.1893821101940665,
+      "grad_norm": 0.17238235473632812,
+      "learning_rate": 1.3332207568566209e-05,
+      "loss": 0.4966,
+      "step": 5332
+    },
+    {
+      "epoch": 1.1896051751059558,
+      "grad_norm": 0.16364935040473938,
+      "learning_rate": 1.332998854831574e-05,
+      "loss": 0.4554,
+      "step": 5333
+    },
+    {
+      "epoch": 1.1898282400178453,
+      "grad_norm": 0.19050312042236328,
+      "learning_rate": 1.3327769343629559e-05,
+      "loss": 0.4485,
+      "step": 5334
+    },
+    {
+      "epoch": 1.1900513049297345,
+      "grad_norm": 0.17412003874778748,
+      "learning_rate": 1.3325549954630579e-05,
+      "loss": 0.4776,
+      "step": 5335
+    },
+    {
+      "epoch": 1.1902743698416238,
+      "grad_norm": 0.16760744154453278,
+      "learning_rate": 1.3323330381441723e-05,
+      "loss": 0.4481,
+      "step": 5336
+    },
+    {
+      "epoch": 1.1904974347535133,
+      "grad_norm": 0.16635169088840485,
+      "learning_rate": 1.3321110624185927e-05,
+      "loss": 0.4572,
+      "step": 5337
+    },
+    {
+      "epoch": 1.1907204996654026,
+      "grad_norm": 0.1623852401971817,
+      "learning_rate": 1.3318890682986135e-05,
+      "loss": 0.455,
+      "step": 5338
+    },
+    {
+      "epoch": 1.190943564577292,
+      "grad_norm": 0.1643582582473755,
+      "learning_rate": 1.3316670557965299e-05,
+      "loss": 0.4356,
+      "step": 5339
+    },
+    {
+      "epoch": 1.1911666294891814,
+      "grad_norm": 0.1604517102241516,
+      "learning_rate": 1.3314450249246385e-05,
+      "loss": 0.4334,
+      "step": 5340
+    },
+    {
+      "epoch": 1.1913896944010707,
+      "grad_norm": 0.16572065651416779,
+      "learning_rate": 1.3312229756952366e-05,
+      "loss": 0.467,
+      "step": 5341
+    },
+    {
+      "epoch": 1.1916127593129602,
+      "grad_norm": 0.1680530309677124,
+      "learning_rate": 1.3310009081206232e-05,
+      "loss": 0.4546,
+      "step": 5342
+    },
+    {
+      "epoch": 1.1918358242248495,
+      "grad_norm": 0.16786667704582214,
+      "learning_rate": 1.330778822213097e-05,
+      "loss": 0.4519,
+      "step": 5343
+    },
+    {
+      "epoch": 1.1920588891367387,
+      "grad_norm": 0.17252503335475922,
+      "learning_rate": 1.3305567179849594e-05,
+      "loss": 0.4748,
+      "step": 5344
+    },
+    {
+      "epoch": 1.1922819540486282,
+      "grad_norm": 0.16872383654117584,
+      "learning_rate": 1.3303345954485113e-05,
+      "loss": 0.4529,
+      "step": 5345
+    },
+    {
+      "epoch": 1.1925050189605175,
+      "grad_norm": 0.1684456765651703,
+      "learning_rate": 1.330112454616055e-05,
+      "loss": 0.4696,
+      "step": 5346
+    },
+    {
+      "epoch": 1.1927280838724068,
+      "grad_norm": 0.16428877413272858,
+      "learning_rate": 1.3298902954998951e-05,
+      "loss": 0.4595,
+      "step": 5347
+    },
+    {
+      "epoch": 1.1929511487842963,
+      "grad_norm": 0.17616313695907593,
+      "learning_rate": 1.329668118112335e-05,
+      "loss": 0.4696,
+      "step": 5348
+    },
+    {
+      "epoch": 1.1931742136961856,
+      "grad_norm": 0.18709328770637512,
+      "learning_rate": 1.3294459224656813e-05,
+      "loss": 0.4847,
+      "step": 5349
+    },
+    {
+      "epoch": 1.1933972786080749,
+      "grad_norm": 0.17944732308387756,
+      "learning_rate": 1.3292237085722396e-05,
+      "loss": 0.4338,
+      "step": 5350
+    },
+    {
+      "epoch": 1.1936203435199644,
+      "grad_norm": 0.1728639304637909,
+      "learning_rate": 1.3290014764443186e-05,
+      "loss": 0.4658,
+      "step": 5351
+    },
+    {
+      "epoch": 1.1938434084318537,
+      "grad_norm": 0.17052049934864044,
+      "learning_rate": 1.328779226094226e-05,
+      "loss": 0.4922,
+      "step": 5352
+    },
+    {
+      "epoch": 1.194066473343743,
+      "grad_norm": 0.16665810346603394,
+      "learning_rate": 1.3285569575342719e-05,
+      "loss": 0.485,
+      "step": 5353
+    },
+    {
+      "epoch": 1.1942895382556324,
+      "grad_norm": 0.17398186028003693,
+      "learning_rate": 1.3283346707767666e-05,
+      "loss": 0.4587,
+      "step": 5354
+    },
+    {
+      "epoch": 1.1945126031675217,
+      "grad_norm": 0.1634979397058487,
+      "learning_rate": 1.3281123658340222e-05,
+      "loss": 0.4309,
+      "step": 5355
+    },
+    {
+      "epoch": 1.194735668079411,
+      "grad_norm": 0.20972603559494019,
+      "learning_rate": 1.3278900427183507e-05,
+      "loss": 0.4808,
+      "step": 5356
+    },
+    {
+      "epoch": 1.1949587329913005,
+      "grad_norm": 0.17813637852668762,
+      "learning_rate": 1.3276677014420665e-05,
+      "loss": 0.4556,
+      "step": 5357
+    },
+    {
+      "epoch": 1.1951817979031898,
+      "grad_norm": 0.18965846300125122,
+      "learning_rate": 1.3274453420174835e-05,
+      "loss": 0.4716,
+      "step": 5358
+    },
+    {
+      "epoch": 1.1954048628150793,
+      "grad_norm": 0.16441969573497772,
+      "learning_rate": 1.3272229644569182e-05,
+      "loss": 0.4379,
+      "step": 5359
+    },
+    {
+      "epoch": 1.1956279277269686,
+      "grad_norm": 0.17030787467956543,
+      "learning_rate": 1.3270005687726864e-05,
+      "loss": 0.4506,
+      "step": 5360
+    },
+    {
+      "epoch": 1.1958509926388579,
+      "grad_norm": 0.1666693389415741,
+      "learning_rate": 1.3267781549771064e-05,
+      "loss": 0.471,
+      "step": 5361
+    },
+    {
+      "epoch": 1.1960740575507474,
+      "grad_norm": 0.1696898639202118,
+      "learning_rate": 1.3265557230824967e-05,
+      "loss": 0.4808,
+      "step": 5362
+    },
+    {
+      "epoch": 1.1962971224626366,
+      "grad_norm": 0.17153112590312958,
+      "learning_rate": 1.326333273101177e-05,
+      "loss": 0.4762,
+      "step": 5363
+    },
+    {
+      "epoch": 1.196520187374526,
+      "grad_norm": 0.1648848056793213,
+      "learning_rate": 1.3261108050454674e-05,
+      "loss": 0.4448,
+      "step": 5364
+    },
+    {
+      "epoch": 1.1967432522864154,
+      "grad_norm": 0.17086121439933777,
+      "learning_rate": 1.3258883189276906e-05,
+      "loss": 0.4848,
+      "step": 5365
+    },
+    {
+      "epoch": 1.1969663171983047,
+      "grad_norm": 0.16545124351978302,
+      "learning_rate": 1.3256658147601686e-05,
+      "loss": 0.4797,
+      "step": 5366
+    },
+    {
+      "epoch": 1.197189382110194,
+      "grad_norm": 0.19573906064033508,
+      "learning_rate": 1.3254432925552252e-05,
+      "loss": 0.4813,
+      "step": 5367
+    },
+    {
+      "epoch": 1.1974124470220835,
+      "grad_norm": 0.16083821654319763,
+      "learning_rate": 1.3252207523251854e-05,
+      "loss": 0.4509,
+      "step": 5368
+    },
+    {
+      "epoch": 1.1976355119339728,
+      "grad_norm": 0.1647178679704666,
+      "learning_rate": 1.3249981940823742e-05,
+      "loss": 0.4506,
+      "step": 5369
+    },
+    {
+      "epoch": 1.197858576845862,
+      "grad_norm": 0.16676411032676697,
+      "learning_rate": 1.3247756178391192e-05,
+      "loss": 0.4696,
+      "step": 5370
+    },
+    {
+      "epoch": 1.1980816417577516,
+      "grad_norm": 0.16701491177082062,
+      "learning_rate": 1.3245530236077474e-05,
+      "loss": 0.4642,
+      "step": 5371
+    },
+    {
+      "epoch": 1.1983047066696408,
+      "grad_norm": 0.16053996980190277,
+      "learning_rate": 1.3243304114005878e-05,
+      "loss": 0.449,
+      "step": 5372
+    },
+    {
+      "epoch": 1.1985277715815301,
+      "grad_norm": 0.1547413319349289,
+      "learning_rate": 1.3241077812299694e-05,
+      "loss": 0.4496,
+      "step": 5373
+    },
+    {
+      "epoch": 1.1987508364934196,
+      "grad_norm": 0.16424560546875,
+      "learning_rate": 1.3238851331082237e-05,
+      "loss": 0.4802,
+      "step": 5374
+    },
+    {
+      "epoch": 1.198973901405309,
+      "grad_norm": 0.1648808866739273,
+      "learning_rate": 1.3236624670476819e-05,
+      "loss": 0.4574,
+      "step": 5375
+    },
+    {
+      "epoch": 1.1991969663171984,
+      "grad_norm": 0.1601562201976776,
+      "learning_rate": 1.323439783060677e-05,
+      "loss": 0.4553,
+      "step": 5376
+    },
+    {
+      "epoch": 1.1994200312290877,
+      "grad_norm": 0.16455237567424774,
+      "learning_rate": 1.323217081159542e-05,
+      "loss": 0.4666,
+      "step": 5377
+    },
+    {
+      "epoch": 1.199643096140977,
+      "grad_norm": 0.16123172640800476,
+      "learning_rate": 1.3229943613566118e-05,
+      "loss": 0.4797,
+      "step": 5378
+    },
+    {
+      "epoch": 1.1998661610528665,
+      "grad_norm": 0.18106834590435028,
+      "learning_rate": 1.3227716236642226e-05,
+      "loss": 0.476,
+      "step": 5379
+    },
+    {
+      "epoch": 1.2000892259647558,
+      "grad_norm": 0.1630832701921463,
+      "learning_rate": 1.3225488680947103e-05,
+      "loss": 0.4995,
+      "step": 5380
+    },
+    {
+      "epoch": 1.200312290876645,
+      "grad_norm": 0.16625766456127167,
+      "learning_rate": 1.322326094660413e-05,
+      "loss": 0.4662,
+      "step": 5381
+    },
+    {
+      "epoch": 1.2005353557885345,
+      "grad_norm": 0.16396591067314148,
+      "learning_rate": 1.3221033033736688e-05,
+      "loss": 0.4631,
+      "step": 5382
+    },
+    {
+      "epoch": 1.2007584207004238,
+      "grad_norm": 0.161235049366951,
+      "learning_rate": 1.321880494246818e-05,
+      "loss": 0.4481,
+      "step": 5383
+    },
+    {
+      "epoch": 1.200981485612313,
+      "grad_norm": 0.16487468779087067,
+      "learning_rate": 1.3216576672922002e-05,
+      "loss": 0.463,
+      "step": 5384
+    },
+    {
+      "epoch": 1.2012045505242026,
+      "grad_norm": 0.16097383201122284,
+      "learning_rate": 1.3214348225221578e-05,
+      "loss": 0.4517,
+      "step": 5385
+    },
+    {
+      "epoch": 1.201427615436092,
+      "grad_norm": 0.1607978343963623,
+      "learning_rate": 1.3212119599490327e-05,
+      "loss": 0.4775,
+      "step": 5386
+    },
+    {
+      "epoch": 1.2016506803479812,
+      "grad_norm": 0.1695454865694046,
+      "learning_rate": 1.3209890795851693e-05,
+      "loss": 0.4667,
+      "step": 5387
+    },
+    {
+      "epoch": 1.2018737452598707,
+      "grad_norm": 0.16572801768779755,
+      "learning_rate": 1.3207661814429112e-05,
+      "loss": 0.4623,
+      "step": 5388
+    },
+    {
+      "epoch": 1.20209681017176,
+      "grad_norm": 0.1595952957868576,
+      "learning_rate": 1.3205432655346044e-05,
+      "loss": 0.4801,
+      "step": 5389
+    },
+    {
+      "epoch": 1.2023198750836492,
+      "grad_norm": 0.16465139389038086,
+      "learning_rate": 1.3203203318725951e-05,
+      "loss": 0.4801,
+      "step": 5390
+    },
+    {
+      "epoch": 1.2025429399955387,
+      "grad_norm": 0.1634162813425064,
+      "learning_rate": 1.320097380469231e-05,
+      "loss": 0.4807,
+      "step": 5391
+    },
+    {
+      "epoch": 1.202766004907428,
+      "grad_norm": 0.16548092663288116,
+      "learning_rate": 1.3198744113368604e-05,
+      "loss": 0.4569,
+      "step": 5392
+    },
+    {
+      "epoch": 1.2029890698193175,
+      "grad_norm": 0.16287754476070404,
+      "learning_rate": 1.319651424487833e-05,
+      "loss": 0.4539,
+      "step": 5393
+    },
+    {
+      "epoch": 1.2032121347312068,
+      "grad_norm": 0.16865308582782745,
+      "learning_rate": 1.3194284199344987e-05,
+      "loss": 0.4555,
+      "step": 5394
+    },
+    {
+      "epoch": 1.203435199643096,
+      "grad_norm": 0.16417443752288818,
+      "learning_rate": 1.3192053976892097e-05,
+      "loss": 0.4449,
+      "step": 5395
+    },
+    {
+      "epoch": 1.2036582645549856,
+      "grad_norm": 0.17047269642353058,
+      "learning_rate": 1.3189823577643175e-05,
+      "loss": 0.4669,
+      "step": 5396
+    },
+    {
+      "epoch": 1.2038813294668749,
+      "grad_norm": 0.15792153775691986,
+      "learning_rate": 1.3187593001721762e-05,
+      "loss": 0.4303,
+      "step": 5397
+    },
+    {
+      "epoch": 1.2041043943787642,
+      "grad_norm": 0.16310684382915497,
+      "learning_rate": 1.3185362249251395e-05,
+      "loss": 0.4373,
+      "step": 5398
+    },
+    {
+      "epoch": 1.2043274592906537,
+      "grad_norm": 0.15764662623405457,
+      "learning_rate": 1.3183131320355629e-05,
+      "loss": 0.4409,
+      "step": 5399
+    },
+    {
+      "epoch": 1.204550524202543,
+      "grad_norm": 0.17336583137512207,
+      "learning_rate": 1.3180900215158028e-05,
+      "loss": 0.4655,
+      "step": 5400
+    },
+    {
+      "epoch": 1.2047735891144322,
+      "grad_norm": 0.1825001984834671,
+      "learning_rate": 1.3178668933782166e-05,
+      "loss": 0.4676,
+      "step": 5401
+    },
+    {
+      "epoch": 1.2049966540263217,
+      "grad_norm": 0.16924050450325012,
+      "learning_rate": 1.3176437476351625e-05,
+      "loss": 0.4762,
+      "step": 5402
+    },
+    {
+      "epoch": 1.205219718938211,
+      "grad_norm": 0.163119837641716,
+      "learning_rate": 1.3174205842989993e-05,
+      "loss": 0.4078,
+      "step": 5403
+    },
+    {
+      "epoch": 1.2054427838501003,
+      "grad_norm": 0.16703581809997559,
+      "learning_rate": 1.3171974033820878e-05,
+      "loss": 0.4769,
+      "step": 5404
+    },
+    {
+      "epoch": 1.2056658487619898,
+      "grad_norm": 0.1679922491312027,
+      "learning_rate": 1.3169742048967886e-05,
+      "loss": 0.4776,
+      "step": 5405
+    },
+    {
+      "epoch": 1.205888913673879,
+      "grad_norm": 0.16275057196617126,
+      "learning_rate": 1.3167509888554641e-05,
+      "loss": 0.4646,
+      "step": 5406
+    },
+    {
+      "epoch": 1.2061119785857684,
+      "grad_norm": 0.16767756640911102,
+      "learning_rate": 1.3165277552704774e-05,
+      "loss": 0.4401,
+      "step": 5407
+    },
+    {
+      "epoch": 1.2063350434976579,
+      "grad_norm": 0.16466595232486725,
+      "learning_rate": 1.3163045041541929e-05,
+      "loss": 0.4626,
+      "step": 5408
+    },
+    {
+      "epoch": 1.2065581084095471,
+      "grad_norm": 0.17687217891216278,
+      "learning_rate": 1.316081235518975e-05,
+      "loss": 0.4553,
+      "step": 5409
+    },
+    {
+      "epoch": 1.2067811733214366,
+      "grad_norm": 0.16307489573955536,
+      "learning_rate": 1.3158579493771901e-05,
+      "loss": 0.4444,
+      "step": 5410
+    },
+    {
+      "epoch": 1.207004238233326,
+      "grad_norm": 0.21108053624629974,
+      "learning_rate": 1.3156346457412051e-05,
+      "loss": 0.4845,
+      "step": 5411
+    },
+    {
+      "epoch": 1.2072273031452152,
+      "grad_norm": 0.1662529855966568,
+      "learning_rate": 1.3154113246233881e-05,
+      "loss": 0.4673,
+      "step": 5412
+    },
+    {
+      "epoch": 1.2074503680571047,
+      "grad_norm": 0.1835554838180542,
+      "learning_rate": 1.3151879860361078e-05,
+      "loss": 0.4679,
+      "step": 5413
+    },
+    {
+      "epoch": 1.207673432968994,
+      "grad_norm": 0.16841189563274384,
+      "learning_rate": 1.3149646299917342e-05,
+      "loss": 0.4932,
+      "step": 5414
+    },
+    {
+      "epoch": 1.2078964978808833,
+      "grad_norm": 0.15623950958251953,
+      "learning_rate": 1.3147412565026385e-05,
+      "loss": 0.4396,
+      "step": 5415
+    },
+    {
+      "epoch": 1.2081195627927728,
+      "grad_norm": 0.1630774587392807,
+      "learning_rate": 1.3145178655811921e-05,
+      "loss": 0.459,
+      "step": 5416
+    },
+    {
+      "epoch": 1.208342627704662,
+      "grad_norm": 0.16008590161800385,
+      "learning_rate": 1.3142944572397677e-05,
+      "loss": 0.4662,
+      "step": 5417
+    },
+    {
+      "epoch": 1.2085656926165513,
+      "grad_norm": 0.16384388506412506,
+      "learning_rate": 1.3140710314907392e-05,
+      "loss": 0.4481,
+      "step": 5418
+    },
+    {
+      "epoch": 1.2087887575284408,
+      "grad_norm": 0.1759367287158966,
+      "learning_rate": 1.3138475883464818e-05,
+      "loss": 0.4766,
+      "step": 5419
+    },
+    {
+      "epoch": 1.2090118224403301,
+      "grad_norm": 0.1597670316696167,
+      "learning_rate": 1.3136241278193704e-05,
+      "loss": 0.4521,
+      "step": 5420
+    },
+    {
+      "epoch": 1.2092348873522194,
+      "grad_norm": 0.1790318489074707,
+      "learning_rate": 1.3134006499217824e-05,
+      "loss": 0.4404,
+      "step": 5421
+    },
+    {
+      "epoch": 1.209457952264109,
+      "grad_norm": 0.16232487559318542,
+      "learning_rate": 1.3131771546660947e-05,
+      "loss": 0.4436,
+      "step": 5422
+    },
+    {
+      "epoch": 1.2096810171759982,
+      "grad_norm": 0.16369551420211792,
+      "learning_rate": 1.3129536420646862e-05,
+      "loss": 0.4754,
+      "step": 5423
+    },
+    {
+      "epoch": 1.2099040820878875,
+      "grad_norm": 0.1614258736371994,
+      "learning_rate": 1.3127301121299367e-05,
+      "loss": 0.4436,
+      "step": 5424
+    },
+    {
+      "epoch": 1.210127146999777,
+      "grad_norm": 0.163519024848938,
+      "learning_rate": 1.3125065648742263e-05,
+      "loss": 0.4443,
+      "step": 5425
+    },
+    {
+      "epoch": 1.2103502119116663,
+      "grad_norm": 0.16275057196617126,
+      "learning_rate": 1.3122830003099364e-05,
+      "loss": 0.4632,
+      "step": 5426
+    },
+    {
+      "epoch": 1.2105732768235558,
+      "grad_norm": 0.1647331267595291,
+      "learning_rate": 1.3120594184494499e-05,
+      "loss": 0.4644,
+      "step": 5427
+    },
+    {
+      "epoch": 1.210796341735445,
+      "grad_norm": 0.16043558716773987,
+      "learning_rate": 1.3118358193051499e-05,
+      "loss": 0.4691,
+      "step": 5428
+    },
+    {
+      "epoch": 1.2110194066473343,
+      "grad_norm": 0.1672472357749939,
+      "learning_rate": 1.3116122028894206e-05,
+      "loss": 0.4663,
+      "step": 5429
+    },
+    {
+      "epoch": 1.2112424715592238,
+      "grad_norm": 0.16601301729679108,
+      "learning_rate": 1.3113885692146473e-05,
+      "loss": 0.4594,
+      "step": 5430
+    },
+    {
+      "epoch": 1.211465536471113,
+      "grad_norm": 0.16756746172904968,
+      "learning_rate": 1.3111649182932163e-05,
+      "loss": 0.4555,
+      "step": 5431
+    },
+    {
+      "epoch": 1.2116886013830024,
+      "grad_norm": 0.1739458590745926,
+      "learning_rate": 1.3109412501375149e-05,
+      "loss": 0.4703,
+      "step": 5432
+    },
+    {
+      "epoch": 1.2119116662948919,
+      "grad_norm": 0.16856876015663147,
+      "learning_rate": 1.3107175647599311e-05,
+      "loss": 0.448,
+      "step": 5433
+    },
+    {
+      "epoch": 1.2121347312067812,
+      "grad_norm": 0.16722862422466278,
+      "learning_rate": 1.3104938621728542e-05,
+      "loss": 0.4598,
+      "step": 5434
+    },
+    {
+      "epoch": 1.2123577961186704,
+      "grad_norm": 0.16905224323272705,
+      "learning_rate": 1.310270142388674e-05,
+      "loss": 0.4782,
+      "step": 5435
+    },
+    {
+      "epoch": 1.21258086103056,
+      "grad_norm": 0.16216522455215454,
+      "learning_rate": 1.3100464054197819e-05,
+      "loss": 0.4755,
+      "step": 5436
+    },
+    {
+      "epoch": 1.2128039259424492,
+      "grad_norm": 0.17112210392951965,
+      "learning_rate": 1.3098226512785695e-05,
+      "loss": 0.4774,
+      "step": 5437
+    },
+    {
+      "epoch": 1.2130269908543385,
+      "grad_norm": 0.16165569424629211,
+      "learning_rate": 1.3095988799774296e-05,
+      "loss": 0.4723,
+      "step": 5438
+    },
+    {
+      "epoch": 1.213250055766228,
+      "grad_norm": 0.16415071487426758,
+      "learning_rate": 1.3093750915287565e-05,
+      "loss": 0.4649,
+      "step": 5439
+    },
+    {
+      "epoch": 1.2134731206781173,
+      "grad_norm": 0.23170824348926544,
+      "learning_rate": 1.3091512859449447e-05,
+      "loss": 0.4754,
+      "step": 5440
+    },
+    {
+      "epoch": 1.2136961855900066,
+      "grad_norm": 0.15743504464626312,
+      "learning_rate": 1.30892746323839e-05,
+      "loss": 0.4367,
+      "step": 5441
+    },
+    {
+      "epoch": 1.213919250501896,
+      "grad_norm": 0.1775827705860138,
+      "learning_rate": 1.3087036234214892e-05,
+      "loss": 0.468,
+      "step": 5442
+    },
+    {
+      "epoch": 1.2141423154137854,
+      "grad_norm": 0.16368676722049713,
+      "learning_rate": 1.3084797665066398e-05,
+      "loss": 0.4635,
+      "step": 5443
+    },
+    {
+      "epoch": 1.2143653803256749,
+      "grad_norm": 0.15797144174575806,
+      "learning_rate": 1.3082558925062406e-05,
+      "loss": 0.4378,
+      "step": 5444
+    },
+    {
+      "epoch": 1.2145884452375642,
+      "grad_norm": 0.16582882404327393,
+      "learning_rate": 1.308032001432691e-05,
+      "loss": 0.4508,
+      "step": 5445
+    },
+    {
+      "epoch": 1.2148115101494534,
+      "grad_norm": 0.16783961653709412,
+      "learning_rate": 1.307808093298392e-05,
+      "loss": 0.478,
+      "step": 5446
+    },
+    {
+      "epoch": 1.215034575061343,
+      "grad_norm": 0.29538553953170776,
+      "learning_rate": 1.307584168115744e-05,
+      "loss": 0.4617,
+      "step": 5447
+    },
+    {
+      "epoch": 1.2152576399732322,
+      "grad_norm": 0.16198082268238068,
+      "learning_rate": 1.3073602258971503e-05,
+      "loss": 0.4605,
+      "step": 5448
+    },
+    {
+      "epoch": 1.2154807048851215,
+      "grad_norm": 0.16145728528499603,
+      "learning_rate": 1.3071362666550136e-05,
+      "loss": 0.4702,
+      "step": 5449
+    },
+    {
+      "epoch": 1.215703769797011,
+      "grad_norm": 0.16520316898822784,
+      "learning_rate": 1.3069122904017389e-05,
+      "loss": 0.4913,
+      "step": 5450
+    },
+    {
+      "epoch": 1.2159268347089003,
+      "grad_norm": 0.1647755205631256,
+      "learning_rate": 1.3066882971497308e-05,
+      "loss": 0.4538,
+      "step": 5451
+    },
+    {
+      "epoch": 1.2161498996207896,
+      "grad_norm": 0.16983206570148468,
+      "learning_rate": 1.3064642869113955e-05,
+      "loss": 0.4531,
+      "step": 5452
+    },
+    {
+      "epoch": 1.216372964532679,
+      "grad_norm": 0.16914002597332,
+      "learning_rate": 1.3062402596991407e-05,
+      "loss": 0.4643,
+      "step": 5453
+    },
+    {
+      "epoch": 1.2165960294445683,
+      "grad_norm": 0.16433201730251312,
+      "learning_rate": 1.3060162155253738e-05,
+      "loss": 0.4398,
+      "step": 5454
+    },
+    {
+      "epoch": 1.2168190943564579,
+      "grad_norm": 0.1646953821182251,
+      "learning_rate": 1.305792154402504e-05,
+      "loss": 0.4704,
+      "step": 5455
+    },
+    {
+      "epoch": 1.2170421592683471,
+      "grad_norm": 0.1791263222694397,
+      "learning_rate": 1.3055680763429411e-05,
+      "loss": 0.4338,
+      "step": 5456
+    },
+    {
+      "epoch": 1.2172652241802364,
+      "grad_norm": 0.16176985204219818,
+      "learning_rate": 1.305343981359096e-05,
+      "loss": 0.4534,
+      "step": 5457
+    },
+    {
+      "epoch": 1.2174882890921257,
+      "grad_norm": 0.16253705322742462,
+      "learning_rate": 1.3051198694633804e-05,
+      "loss": 0.4728,
+      "step": 5458
+    },
+    {
+      "epoch": 1.2177113540040152,
+      "grad_norm": 0.1789017915725708,
+      "learning_rate": 1.3048957406682074e-05,
+      "loss": 0.4333,
+      "step": 5459
+    },
+    {
+      "epoch": 1.2179344189159045,
+      "grad_norm": 0.16050660610198975,
+      "learning_rate": 1.3046715949859902e-05,
+      "loss": 0.4614,
+      "step": 5460
+    },
+    {
+      "epoch": 1.218157483827794,
+      "grad_norm": 0.15737925469875336,
+      "learning_rate": 1.3044474324291438e-05,
+      "loss": 0.4558,
+      "step": 5461
+    },
+    {
+      "epoch": 1.2183805487396833,
+      "grad_norm": 0.16488178074359894,
+      "learning_rate": 1.3042232530100833e-05,
+      "loss": 0.4804,
+      "step": 5462
+    },
+    {
+      "epoch": 1.2186036136515725,
+      "grad_norm": 0.16783639788627625,
+      "learning_rate": 1.3039990567412255e-05,
+      "loss": 0.4564,
+      "step": 5463
+    },
+    {
+      "epoch": 1.218826678563462,
+      "grad_norm": 0.16743387281894684,
+      "learning_rate": 1.303774843634988e-05,
+      "loss": 0.4568,
+      "step": 5464
+    },
+    {
+      "epoch": 1.2190497434753513,
+      "grad_norm": 0.16670465469360352,
+      "learning_rate": 1.3035506137037883e-05,
+      "loss": 0.4416,
+      "step": 5465
+    },
+    {
+      "epoch": 1.2192728083872406,
+      "grad_norm": 0.17037998139858246,
+      "learning_rate": 1.3033263669600466e-05,
+      "loss": 0.4753,
+      "step": 5466
+    },
+    {
+      "epoch": 1.2194958732991301,
+      "grad_norm": 0.17123369872570038,
+      "learning_rate": 1.3031021034161827e-05,
+      "loss": 0.4718,
+      "step": 5467
+    },
+    {
+      "epoch": 1.2197189382110194,
+      "grad_norm": 0.17543406784534454,
+      "learning_rate": 1.3028778230846178e-05,
+      "loss": 0.4564,
+      "step": 5468
+    },
+    {
+      "epoch": 1.2199420031229087,
+      "grad_norm": 0.17404866218566895,
+      "learning_rate": 1.3026535259777734e-05,
+      "loss": 0.4716,
+      "step": 5469
+    },
+    {
+      "epoch": 1.2201650680347982,
+      "grad_norm": 0.1693383902311325,
+      "learning_rate": 1.3024292121080735e-05,
+      "loss": 0.4759,
+      "step": 5470
+    },
+    {
+      "epoch": 1.2203881329466875,
+      "grad_norm": 0.18779967725276947,
+      "learning_rate": 1.3022048814879412e-05,
+      "loss": 0.4784,
+      "step": 5471
+    },
+    {
+      "epoch": 1.220611197858577,
+      "grad_norm": 0.16473285853862762,
+      "learning_rate": 1.3019805341298017e-05,
+      "loss": 0.4334,
+      "step": 5472
+    },
+    {
+      "epoch": 1.2208342627704662,
+      "grad_norm": 0.1686456799507141,
+      "learning_rate": 1.3017561700460806e-05,
+      "loss": 0.4744,
+      "step": 5473
+    },
+    {
+      "epoch": 1.2210573276823555,
+      "grad_norm": 0.16365700960159302,
+      "learning_rate": 1.301531789249205e-05,
+      "loss": 0.4711,
+      "step": 5474
+    },
+    {
+      "epoch": 1.2212803925942448,
+      "grad_norm": 0.16734614968299866,
+      "learning_rate": 1.3013073917516018e-05,
+      "loss": 0.4512,
+      "step": 5475
+    },
+    {
+      "epoch": 1.2215034575061343,
+      "grad_norm": 0.1621614247560501,
+      "learning_rate": 1.3010829775657001e-05,
+      "loss": 0.4579,
+      "step": 5476
+    },
+    {
+      "epoch": 1.2217265224180236,
+      "grad_norm": 0.16041643917560577,
+      "learning_rate": 1.3008585467039291e-05,
+      "loss": 0.4723,
+      "step": 5477
+    },
+    {
+      "epoch": 1.221949587329913,
+      "grad_norm": 0.17315927147865295,
+      "learning_rate": 1.3006340991787196e-05,
+      "loss": 0.4888,
+      "step": 5478
+    },
+    {
+      "epoch": 1.2221726522418024,
+      "grad_norm": 0.16364838182926178,
+      "learning_rate": 1.3004096350025024e-05,
+      "loss": 0.4456,
+      "step": 5479
+    },
+    {
+      "epoch": 1.2223957171536917,
+      "grad_norm": 0.17080283164978027,
+      "learning_rate": 1.30018515418771e-05,
+      "loss": 0.4545,
+      "step": 5480
+    },
+    {
+      "epoch": 1.2226187820655812,
+      "grad_norm": 0.1685846894979477,
+      "learning_rate": 1.2999606567467753e-05,
+      "loss": 0.4709,
+      "step": 5481
+    },
+    {
+      "epoch": 1.2228418469774704,
+      "grad_norm": 0.16784535348415375,
+      "learning_rate": 1.2997361426921331e-05,
+      "loss": 0.4753,
+      "step": 5482
+    },
+    {
+      "epoch": 1.2230649118893597,
+      "grad_norm": 0.17074494063854218,
+      "learning_rate": 1.2995116120362175e-05,
+      "loss": 0.4519,
+      "step": 5483
+    },
+    {
+      "epoch": 1.2232879768012492,
+      "grad_norm": 0.15669210255146027,
+      "learning_rate": 1.2992870647914648e-05,
+      "loss": 0.4205,
+      "step": 5484
+    },
+    {
+      "epoch": 1.2235110417131385,
+      "grad_norm": 0.16939307749271393,
+      "learning_rate": 1.2990625009703122e-05,
+      "loss": 0.4685,
+      "step": 5485
+    },
+    {
+      "epoch": 1.2237341066250278,
+      "grad_norm": 0.1645343005657196,
+      "learning_rate": 1.298837920585197e-05,
+      "loss": 0.4713,
+      "step": 5486
+    },
+    {
+      "epoch": 1.2239571715369173,
+      "grad_norm": 0.16123943030834198,
+      "learning_rate": 1.298613323648558e-05,
+      "loss": 0.4592,
+      "step": 5487
+    },
+    {
+      "epoch": 1.2241802364488066,
+      "grad_norm": 0.16125601530075073,
+      "learning_rate": 1.2983887101728342e-05,
+      "loss": 0.4522,
+      "step": 5488
+    },
+    {
+      "epoch": 1.224403301360696,
+      "grad_norm": 0.16092932224273682,
+      "learning_rate": 1.2981640801704676e-05,
+      "loss": 0.4469,
+      "step": 5489
+    },
+    {
+      "epoch": 1.2246263662725854,
+      "grad_norm": 0.16154557466506958,
+      "learning_rate": 1.2979394336538982e-05,
+      "loss": 0.5035,
+      "step": 5490
+    },
+    {
+      "epoch": 1.2248494311844746,
+      "grad_norm": 0.16762235760688782,
+      "learning_rate": 1.2977147706355688e-05,
+      "loss": 0.4505,
+      "step": 5491
+    },
+    {
+      "epoch": 1.225072496096364,
+      "grad_norm": 0.16036982834339142,
+      "learning_rate": 1.297490091127923e-05,
+      "loss": 0.4635,
+      "step": 5492
+    },
+    {
+      "epoch": 1.2252955610082534,
+      "grad_norm": 0.16005462408065796,
+      "learning_rate": 1.2972653951434046e-05,
+      "loss": 0.4745,
+      "step": 5493
+    },
+    {
+      "epoch": 1.2255186259201427,
+      "grad_norm": 0.16705186665058136,
+      "learning_rate": 1.2970406826944589e-05,
+      "loss": 0.4684,
+      "step": 5494
+    },
+    {
+      "epoch": 1.2257416908320322,
+      "grad_norm": 0.16716639697551727,
+      "learning_rate": 1.2968159537935312e-05,
+      "loss": 0.4612,
+      "step": 5495
+    },
+    {
+      "epoch": 1.2259647557439215,
+      "grad_norm": 0.16546767950057983,
+      "learning_rate": 1.2965912084530693e-05,
+      "loss": 0.4797,
+      "step": 5496
+    },
+    {
+      "epoch": 1.2261878206558108,
+      "grad_norm": 0.16606946289539337,
+      "learning_rate": 1.2963664466855207e-05,
+      "loss": 0.4695,
+      "step": 5497
+    },
+    {
+      "epoch": 1.2264108855677003,
+      "grad_norm": 0.16966325044631958,
+      "learning_rate": 1.2961416685033339e-05,
+      "loss": 0.4498,
+      "step": 5498
+    },
+    {
+      "epoch": 1.2266339504795896,
+      "grad_norm": 0.17357441782951355,
+      "learning_rate": 1.2959168739189587e-05,
+      "loss": 0.4576,
+      "step": 5499
+    },
+    {
+      "epoch": 1.2268570153914788,
+      "grad_norm": 0.16445501148700714,
+      "learning_rate": 1.2956920629448458e-05,
+      "loss": 0.4371,
+      "step": 5500
+    },
+    {
+      "epoch": 1.2270800803033683,
+      "grad_norm": 0.16668792068958282,
+      "learning_rate": 1.295467235593446e-05,
+      "loss": 0.4698,
+      "step": 5501
+    },
+    {
+      "epoch": 1.2273031452152576,
+      "grad_norm": 0.15939384698867798,
+      "learning_rate": 1.2952423918772128e-05,
+      "loss": 0.4261,
+      "step": 5502
+    },
+    {
+      "epoch": 1.227526210127147,
+      "grad_norm": 0.1705656349658966,
+      "learning_rate": 1.2950175318085983e-05,
+      "loss": 0.441,
+      "step": 5503
+    },
+    {
+      "epoch": 1.2277492750390364,
+      "grad_norm": 0.16186708211898804,
+      "learning_rate": 1.2947926554000574e-05,
+      "loss": 0.4795,
+      "step": 5504
+    },
+    {
+      "epoch": 1.2279723399509257,
+      "grad_norm": 0.16865457594394684,
+      "learning_rate": 1.2945677626640447e-05,
+      "loss": 0.4653,
+      "step": 5505
+    },
+    {
+      "epoch": 1.2281954048628152,
+      "grad_norm": 0.16898195445537567,
+      "learning_rate": 1.2943428536130167e-05,
+      "loss": 0.4422,
+      "step": 5506
+    },
+    {
+      "epoch": 1.2284184697747045,
+      "grad_norm": 0.15895049273967743,
+      "learning_rate": 1.2941179282594298e-05,
+      "loss": 0.4435,
+      "step": 5507
+    },
+    {
+      "epoch": 1.2286415346865938,
+      "grad_norm": 0.17526674270629883,
+      "learning_rate": 1.293892986615742e-05,
+      "loss": 0.4486,
+      "step": 5508
+    },
+    {
+      "epoch": 1.228864599598483,
+      "grad_norm": 0.16472922265529633,
+      "learning_rate": 1.2936680286944118e-05,
+      "loss": 0.4414,
+      "step": 5509
+    },
+    {
+      "epoch": 1.2290876645103725,
+      "grad_norm": 0.16342167556285858,
+      "learning_rate": 1.2934430545078991e-05,
+      "loss": 0.4708,
+      "step": 5510
+    },
+    {
+      "epoch": 1.2293107294222618,
+      "grad_norm": 0.16461318731307983,
+      "learning_rate": 1.2932180640686643e-05,
+      "loss": 0.4545,
+      "step": 5511
+    },
+    {
+      "epoch": 1.2295337943341513,
+      "grad_norm": 0.16927462816238403,
+      "learning_rate": 1.2929930573891685e-05,
+      "loss": 0.4573,
+      "step": 5512
+    },
+    {
+      "epoch": 1.2297568592460406,
+      "grad_norm": 0.1817016750574112,
+      "learning_rate": 1.2927680344818741e-05,
+      "loss": 0.442,
+      "step": 5513
+    },
+    {
+      "epoch": 1.22997992415793,
+      "grad_norm": 0.16440905630588531,
+      "learning_rate": 1.2925429953592446e-05,
+      "loss": 0.4534,
+      "step": 5514
+    },
+    {
+      "epoch": 1.2302029890698194,
+      "grad_norm": 0.16801615059375763,
+      "learning_rate": 1.2923179400337435e-05,
+      "loss": 0.447,
+      "step": 5515
+    },
+    {
+      "epoch": 1.2304260539817087,
+      "grad_norm": 0.16331814229488373,
+      "learning_rate": 1.2920928685178365e-05,
+      "loss": 0.4432,
+      "step": 5516
+    },
+    {
+      "epoch": 1.230649118893598,
+      "grad_norm": 0.17522062361240387,
+      "learning_rate": 1.291867780823989e-05,
+      "loss": 0.497,
+      "step": 5517
+    },
+    {
+      "epoch": 1.2308721838054875,
+      "grad_norm": 0.1681208312511444,
+      "learning_rate": 1.2916426769646677e-05,
+      "loss": 0.4688,
+      "step": 5518
+    },
+    {
+      "epoch": 1.2310952487173767,
+      "grad_norm": 0.16323919594287872,
+      "learning_rate": 1.2914175569523408e-05,
+      "loss": 0.4562,
+      "step": 5519
+    },
+    {
+      "epoch": 1.231318313629266,
+      "grad_norm": 0.17886684834957123,
+      "learning_rate": 1.291192420799476e-05,
+      "loss": 0.4333,
+      "step": 5520
+    },
+    {
+      "epoch": 1.2315413785411555,
+      "grad_norm": 0.17314215004444122,
+      "learning_rate": 1.2909672685185437e-05,
+      "loss": 0.4517,
+      "step": 5521
+    },
+    {
+      "epoch": 1.2317644434530448,
+      "grad_norm": 0.16686271131038666,
+      "learning_rate": 1.2907421001220138e-05,
+      "loss": 0.4842,
+      "step": 5522
+    },
+    {
+      "epoch": 1.2319875083649343,
+      "grad_norm": 0.16551433503627777,
+      "learning_rate": 1.2905169156223574e-05,
+      "loss": 0.4726,
+      "step": 5523
+    },
+    {
+      "epoch": 1.2322105732768236,
+      "grad_norm": 0.17935457825660706,
+      "learning_rate": 1.2902917150320468e-05,
+      "loss": 0.4438,
+      "step": 5524
+    },
+    {
+      "epoch": 1.2324336381887129,
+      "grad_norm": 0.1675693243741989,
+      "learning_rate": 1.2900664983635551e-05,
+      "loss": 0.4471,
+      "step": 5525
+    },
+    {
+      "epoch": 1.2326567031006022,
+      "grad_norm": 0.1640770435333252,
+      "learning_rate": 1.289841265629356e-05,
+      "loss": 0.4245,
+      "step": 5526
+    },
+    {
+      "epoch": 1.2328797680124917,
+      "grad_norm": 0.17131870985031128,
+      "learning_rate": 1.2896160168419245e-05,
+      "loss": 0.4767,
+      "step": 5527
+    },
+    {
+      "epoch": 1.233102832924381,
+      "grad_norm": 0.1778288334608078,
+      "learning_rate": 1.2893907520137364e-05,
+      "loss": 0.4754,
+      "step": 5528
+    },
+    {
+      "epoch": 1.2333258978362704,
+      "grad_norm": 0.16755534708499908,
+      "learning_rate": 1.2891654711572678e-05,
+      "loss": 0.4649,
+      "step": 5529
+    },
+    {
+      "epoch": 1.2335489627481597,
+      "grad_norm": 0.1766786277294159,
+      "learning_rate": 1.2889401742849965e-05,
+      "loss": 0.5005,
+      "step": 5530
+    },
+    {
+      "epoch": 1.233772027660049,
+      "grad_norm": 0.16578464210033417,
+      "learning_rate": 1.2887148614094012e-05,
+      "loss": 0.4371,
+      "step": 5531
+    },
+    {
+      "epoch": 1.2339950925719385,
+      "grad_norm": 0.17723850905895233,
+      "learning_rate": 1.2884895325429605e-05,
+      "loss": 0.4787,
+      "step": 5532
+    },
+    {
+      "epoch": 1.2342181574838278,
+      "grad_norm": 0.171020045876503,
+      "learning_rate": 1.288264187698155e-05,
+      "loss": 0.4576,
+      "step": 5533
+    },
+    {
+      "epoch": 1.234441222395717,
+      "grad_norm": 0.16916728019714355,
+      "learning_rate": 1.2880388268874653e-05,
+      "loss": 0.4739,
+      "step": 5534
+    },
+    {
+      "epoch": 1.2346642873076066,
+      "grad_norm": 0.16481377184391022,
+      "learning_rate": 1.2878134501233736e-05,
+      "loss": 0.4456,
+      "step": 5535
+    },
+    {
+      "epoch": 1.2348873522194959,
+      "grad_norm": 0.3519321382045746,
+      "learning_rate": 1.2875880574183628e-05,
+      "loss": 0.4754,
+      "step": 5536
+    },
+    {
+      "epoch": 1.2351104171313851,
+      "grad_norm": 0.23968984186649323,
+      "learning_rate": 1.2873626487849162e-05,
+      "loss": 0.4661,
+      "step": 5537
+    },
+    {
+      "epoch": 1.2353334820432746,
+      "grad_norm": 0.16674065589904785,
+      "learning_rate": 1.2871372242355186e-05,
+      "loss": 0.4611,
+      "step": 5538
+    },
+    {
+      "epoch": 1.235556546955164,
+      "grad_norm": 0.17423000931739807,
+      "learning_rate": 1.2869117837826553e-05,
+      "loss": 0.4456,
+      "step": 5539
+    },
+    {
+      "epoch": 1.2357796118670534,
+      "grad_norm": 0.16467906534671783,
+      "learning_rate": 1.2866863274388128e-05,
+      "loss": 0.459,
+      "step": 5540
+    },
+    {
+      "epoch": 1.2360026767789427,
+      "grad_norm": 0.1630530208349228,
+      "learning_rate": 1.2864608552164779e-05,
+      "loss": 0.4274,
+      "step": 5541
+    },
+    {
+      "epoch": 1.236225741690832,
+      "grad_norm": 0.1718243807554245,
+      "learning_rate": 1.2862353671281392e-05,
+      "loss": 0.4572,
+      "step": 5542
+    },
+    {
+      "epoch": 1.2364488066027213,
+      "grad_norm": 0.1649494469165802,
+      "learning_rate": 1.2860098631862852e-05,
+      "loss": 0.4365,
+      "step": 5543
+    },
+    {
+      "epoch": 1.2366718715146108,
+      "grad_norm": 0.16701866686344147,
+      "learning_rate": 1.2857843434034058e-05,
+      "loss": 0.4559,
+      "step": 5544
+    },
+    {
+      "epoch": 1.2368949364265,
+      "grad_norm": 0.16802479326725006,
+      "learning_rate": 1.2855588077919921e-05,
+      "loss": 0.4636,
+      "step": 5545
+    },
+    {
+      "epoch": 1.2371180013383896,
+      "grad_norm": 0.17907430231571198,
+      "learning_rate": 1.2853332563645353e-05,
+      "loss": 0.4747,
+      "step": 5546
+    },
+    {
+      "epoch": 1.2373410662502788,
+      "grad_norm": 0.16412149369716644,
+      "learning_rate": 1.2851076891335277e-05,
+      "loss": 0.4492,
+      "step": 5547
+    },
+    {
+      "epoch": 1.2375641311621681,
+      "grad_norm": 0.16632318496704102,
+      "learning_rate": 1.2848821061114629e-05,
+      "loss": 0.4628,
+      "step": 5548
+    },
+    {
+      "epoch": 1.2377871960740576,
+      "grad_norm": 0.1785164624452591,
+      "learning_rate": 1.2846565073108355e-05,
+      "loss": 0.4821,
+      "step": 5549
+    },
+    {
+      "epoch": 1.238010260985947,
+      "grad_norm": 0.1653890609741211,
+      "learning_rate": 1.2844308927441397e-05,
+      "loss": 0.435,
+      "step": 5550
+    },
+    {
+      "epoch": 1.2382333258978362,
+      "grad_norm": 0.18326538801193237,
+      "learning_rate": 1.2842052624238724e-05,
+      "loss": 0.4691,
+      "step": 5551
+    },
+    {
+      "epoch": 1.2384563908097257,
+      "grad_norm": 0.16218526661396027,
+      "learning_rate": 1.2839796163625296e-05,
+      "loss": 0.441,
+      "step": 5552
+    },
+    {
+      "epoch": 1.238679455721615,
+      "grad_norm": 0.17150120437145233,
+      "learning_rate": 1.2837539545726095e-05,
+      "loss": 0.4533,
+      "step": 5553
+    },
+    {
+      "epoch": 1.2389025206335043,
+      "grad_norm": 0.17477098107337952,
+      "learning_rate": 1.2835282770666101e-05,
+      "loss": 0.4898,
+      "step": 5554
+    },
+    {
+      "epoch": 1.2391255855453938,
+      "grad_norm": 0.16403955221176147,
+      "learning_rate": 1.2833025838570318e-05,
+      "loss": 0.4591,
+      "step": 5555
+    },
+    {
+      "epoch": 1.239348650457283,
+      "grad_norm": 0.17669036984443665,
+      "learning_rate": 1.283076874956374e-05,
+      "loss": 0.4727,
+      "step": 5556
+    },
+    {
+      "epoch": 1.2395717153691725,
+      "grad_norm": 0.17886574566364288,
+      "learning_rate": 1.2828511503771386e-05,
+      "loss": 0.4824,
+      "step": 5557
+    },
+    {
+      "epoch": 1.2397947802810618,
+      "grad_norm": 0.1615883857011795,
+      "learning_rate": 1.282625410131827e-05,
+      "loss": 0.4573,
+      "step": 5558
+    },
+    {
+      "epoch": 1.240017845192951,
+      "grad_norm": 0.16335685551166534,
+      "learning_rate": 1.2823996542329426e-05,
+      "loss": 0.4624,
+      "step": 5559
+    },
+    {
+      "epoch": 1.2402409101048404,
+      "grad_norm": 0.17322902381420135,
+      "learning_rate": 1.2821738826929888e-05,
+      "loss": 0.4607,
+      "step": 5560
+    },
+    {
+      "epoch": 1.24046397501673,
+      "grad_norm": 0.17016318440437317,
+      "learning_rate": 1.2819480955244705e-05,
+      "loss": 0.4522,
+      "step": 5561
+    },
+    {
+      "epoch": 1.2406870399286192,
+      "grad_norm": 0.16096848249435425,
+      "learning_rate": 1.2817222927398932e-05,
+      "loss": 0.4475,
+      "step": 5562
+    },
+    {
+      "epoch": 1.2409101048405087,
+      "grad_norm": 0.166842982172966,
+      "learning_rate": 1.281496474351763e-05,
+      "loss": 0.4731,
+      "step": 5563
+    },
+    {
+      "epoch": 1.241133169752398,
+      "grad_norm": 0.17023105919361115,
+      "learning_rate": 1.2812706403725876e-05,
+      "loss": 0.4626,
+      "step": 5564
+    },
+    {
+      "epoch": 1.2413562346642872,
+      "grad_norm": 0.1609751284122467,
+      "learning_rate": 1.2810447908148748e-05,
+      "loss": 0.4751,
+      "step": 5565
+    },
+    {
+      "epoch": 1.2415792995761767,
+      "grad_norm": 0.22108225524425507,
+      "learning_rate": 1.2808189256911336e-05,
+      "loss": 0.4497,
+      "step": 5566
+    },
+    {
+      "epoch": 1.241802364488066,
+      "grad_norm": 0.1703910231590271,
+      "learning_rate": 1.2805930450138742e-05,
+      "loss": 0.4731,
+      "step": 5567
+    },
+    {
+      "epoch": 1.2420254293999553,
+      "grad_norm": 0.16368292272090912,
+      "learning_rate": 1.2803671487956063e-05,
+      "loss": 0.459,
+      "step": 5568
+    },
+    {
+      "epoch": 1.2422484943118448,
+      "grad_norm": 0.1646466702222824,
+      "learning_rate": 1.2801412370488427e-05,
+      "loss": 0.4756,
+      "step": 5569
+    },
+    {
+      "epoch": 1.242471559223734,
+      "grad_norm": 0.16485071182250977,
+      "learning_rate": 1.279915309786095e-05,
+      "loss": 0.479,
+      "step": 5570
+    },
+    {
+      "epoch": 1.2426946241356234,
+      "grad_norm": 0.17181314527988434,
+      "learning_rate": 1.2796893670198767e-05,
+      "loss": 0.4802,
+      "step": 5571
+    },
+    {
+      "epoch": 1.2429176890475129,
+      "grad_norm": 0.17581303417682648,
+      "learning_rate": 1.279463408762702e-05,
+      "loss": 0.4747,
+      "step": 5572
+    },
+    {
+      "epoch": 1.2431407539594022,
+      "grad_norm": 0.16636666655540466,
+      "learning_rate": 1.2792374350270858e-05,
+      "loss": 0.443,
+      "step": 5573
+    },
+    {
+      "epoch": 1.2433638188712917,
+      "grad_norm": 0.1754438728094101,
+      "learning_rate": 1.2790114458255441e-05,
+      "loss": 0.4555,
+      "step": 5574
+    },
+    {
+      "epoch": 1.243586883783181,
+      "grad_norm": 0.16133657097816467,
+      "learning_rate": 1.2787854411705935e-05,
+      "loss": 0.4426,
+      "step": 5575
+    },
+    {
+      "epoch": 1.2438099486950702,
+      "grad_norm": 0.15894077718257904,
+      "learning_rate": 1.2785594210747513e-05,
+      "loss": 0.4511,
+      "step": 5576
+    },
+    {
+      "epoch": 1.2440330136069595,
+      "grad_norm": 0.1614711582660675,
+      "learning_rate": 1.2783333855505364e-05,
+      "loss": 0.4712,
+      "step": 5577
+    },
+    {
+      "epoch": 1.244256078518849,
+      "grad_norm": 0.16365355253219604,
+      "learning_rate": 1.2781073346104677e-05,
+      "loss": 0.4399,
+      "step": 5578
+    },
+    {
+      "epoch": 1.2444791434307383,
+      "grad_norm": 0.16568869352340698,
+      "learning_rate": 1.2778812682670654e-05,
+      "loss": 0.4695,
+      "step": 5579
+    },
+    {
+      "epoch": 1.2447022083426278,
+      "grad_norm": 0.16856862604618073,
+      "learning_rate": 1.2776551865328503e-05,
+      "loss": 0.4528,
+      "step": 5580
+    },
+    {
+      "epoch": 1.244925273254517,
+      "grad_norm": 0.17370723187923431,
+      "learning_rate": 1.277429089420345e-05,
+      "loss": 0.4822,
+      "step": 5581
+    },
+    {
+      "epoch": 1.2451483381664064,
+      "grad_norm": 0.16506825387477875,
+      "learning_rate": 1.277202976942071e-05,
+      "loss": 0.4354,
+      "step": 5582
+    },
+    {
+      "epoch": 1.2453714030782959,
+      "grad_norm": 0.16328351199626923,
+      "learning_rate": 1.276976849110553e-05,
+      "loss": 0.4509,
+      "step": 5583
+    },
+    {
+      "epoch": 1.2455944679901851,
+      "grad_norm": 0.16474324464797974,
+      "learning_rate": 1.2767507059383144e-05,
+      "loss": 0.4789,
+      "step": 5584
+    },
+    {
+      "epoch": 1.2458175329020744,
+      "grad_norm": 0.16100145876407623,
+      "learning_rate": 1.2765245474378814e-05,
+      "loss": 0.442,
+      "step": 5585
+    },
+    {
+      "epoch": 1.246040597813964,
+      "grad_norm": 0.1651872992515564,
+      "learning_rate": 1.2762983736217792e-05,
+      "loss": 0.4363,
+      "step": 5586
+    },
+    {
+      "epoch": 1.2462636627258532,
+      "grad_norm": 0.1605336219072342,
+      "learning_rate": 1.2760721845025353e-05,
+      "loss": 0.4722,
+      "step": 5587
+    },
+    {
+      "epoch": 1.2464867276377425,
+      "grad_norm": 0.16939808428287506,
+      "learning_rate": 1.2758459800926768e-05,
+      "loss": 0.4738,
+      "step": 5588
+    },
+    {
+      "epoch": 1.246709792549632,
+      "grad_norm": 0.16064578294754028,
+      "learning_rate": 1.2756197604047333e-05,
+      "loss": 0.4487,
+      "step": 5589
+    },
+    {
+      "epoch": 1.2469328574615213,
+      "grad_norm": 0.17054161429405212,
+      "learning_rate": 1.2753935254512332e-05,
+      "loss": 0.4863,
+      "step": 5590
+    },
+    {
+      "epoch": 1.2471559223734108,
+      "grad_norm": 0.16378219425678253,
+      "learning_rate": 1.2751672752447079e-05,
+      "loss": 0.4373,
+      "step": 5591
+    },
+    {
+      "epoch": 1.2473789872853,
+      "grad_norm": 0.16274695098400116,
+      "learning_rate": 1.2749410097976878e-05,
+      "loss": 0.4425,
+      "step": 5592
+    },
+    {
+      "epoch": 1.2476020521971893,
+      "grad_norm": 0.16811704635620117,
+      "learning_rate": 1.2747147291227053e-05,
+      "loss": 0.4589,
+      "step": 5593
+    },
+    {
+      "epoch": 1.2478251171090786,
+      "grad_norm": 0.1644984930753708,
+      "learning_rate": 1.2744884332322926e-05,
+      "loss": 0.4524,
+      "step": 5594
+    },
+    {
+      "epoch": 1.2480481820209681,
+      "grad_norm": 0.17755252122879028,
+      "learning_rate": 1.2742621221389846e-05,
+      "loss": 0.4453,
+      "step": 5595
+    },
+    {
+      "epoch": 1.2482712469328574,
+      "grad_norm": 0.17471030354499817,
+      "learning_rate": 1.2740357958553144e-05,
+      "loss": 0.5054,
+      "step": 5596
+    },
+    {
+      "epoch": 1.248494311844747,
+      "grad_norm": 0.16155916452407837,
+      "learning_rate": 1.2738094543938187e-05,
+      "loss": 0.4663,
+      "step": 5597
+    },
+    {
+      "epoch": 1.2487173767566362,
+      "grad_norm": 0.16668613255023956,
+      "learning_rate": 1.2735830977670325e-05,
+      "loss": 0.4664,
+      "step": 5598
+    },
+    {
+      "epoch": 1.2489404416685255,
+      "grad_norm": 0.17190469801425934,
+      "learning_rate": 1.2733567259874937e-05,
+      "loss": 0.4645,
+      "step": 5599
+    },
+    {
+      "epoch": 1.249163506580415,
+      "grad_norm": 0.17003987729549408,
+      "learning_rate": 1.2731303390677399e-05,
+      "loss": 0.4662,
+      "step": 5600
+    },
+    {
+      "epoch": 1.2493865714923043,
+      "grad_norm": 0.175228089094162,
+      "learning_rate": 1.2729039370203098e-05,
+      "loss": 0.4609,
+      "step": 5601
+    },
+    {
+      "epoch": 1.2496096364041935,
+      "grad_norm": 0.169545978307724,
+      "learning_rate": 1.2726775198577432e-05,
+      "loss": 0.4838,
+      "step": 5602
+    },
+    {
+      "epoch": 1.249832701316083,
+      "grad_norm": 0.16817337274551392,
+      "learning_rate": 1.2724510875925802e-05,
+      "loss": 0.4786,
+      "step": 5603
+    },
+    {
+      "epoch": 1.2500557662279723,
+      "grad_norm": 0.16374395787715912,
+      "learning_rate": 1.2722246402373624e-05,
+      "loss": 0.4848,
+      "step": 5604
+    },
+    {
+      "epoch": 1.2502788311398616,
+      "grad_norm": 0.16421671211719513,
+      "learning_rate": 1.2719981778046313e-05,
+      "loss": 0.4712,
+      "step": 5605
+    },
+    {
+      "epoch": 1.250501896051751,
+      "grad_norm": 0.16157956421375275,
+      "learning_rate": 1.2717717003069305e-05,
+      "loss": 0.4392,
+      "step": 5606
+    },
+    {
+      "epoch": 1.2507249609636404,
+      "grad_norm": 0.1583690643310547,
+      "learning_rate": 1.271545207756803e-05,
+      "loss": 0.4425,
+      "step": 5607
+    },
+    {
+      "epoch": 1.25094802587553,
+      "grad_norm": 0.18596141040325165,
+      "learning_rate": 1.2713187001667943e-05,
+      "loss": 0.4751,
+      "step": 5608
+    },
+    {
+      "epoch": 1.2511710907874192,
+      "grad_norm": 0.19196903705596924,
+      "learning_rate": 1.2710921775494494e-05,
+      "loss": 0.4827,
+      "step": 5609
+    },
+    {
+      "epoch": 1.2513941556993085,
+      "grad_norm": 0.16318334639072418,
+      "learning_rate": 1.270865639917314e-05,
+      "loss": 0.4548,
+      "step": 5610
+    },
+    {
+      "epoch": 1.2516172206111977,
+      "grad_norm": 0.17583735287189484,
+      "learning_rate": 1.270639087282936e-05,
+      "loss": 0.4829,
+      "step": 5611
+    },
+    {
+      "epoch": 1.2518402855230872,
+      "grad_norm": 0.19173632562160492,
+      "learning_rate": 1.2704125196588628e-05,
+      "loss": 0.4773,
+      "step": 5612
+    },
+    {
+      "epoch": 1.2520633504349765,
+      "grad_norm": 0.15936903655529022,
+      "learning_rate": 1.2701859370576432e-05,
+      "loss": 0.428,
+      "step": 5613
+    },
+    {
+      "epoch": 1.252286415346866,
+      "grad_norm": 0.16663479804992676,
+      "learning_rate": 1.2699593394918273e-05,
+      "loss": 0.4285,
+      "step": 5614
+    },
+    {
+      "epoch": 1.2525094802587553,
+      "grad_norm": 0.16250313818454742,
+      "learning_rate": 1.2697327269739646e-05,
+      "loss": 0.471,
+      "step": 5615
+    },
+    {
+      "epoch": 1.2527325451706446,
+      "grad_norm": 0.16274987161159515,
+      "learning_rate": 1.2695060995166069e-05,
+      "loss": 0.4566,
+      "step": 5616
+    },
+    {
+      "epoch": 1.252955610082534,
+      "grad_norm": 0.17902643978595734,
+      "learning_rate": 1.2692794571323064e-05,
+      "loss": 0.4694,
+      "step": 5617
+    },
+    {
+      "epoch": 1.2531786749944234,
+      "grad_norm": 0.16219644248485565,
+      "learning_rate": 1.2690527998336153e-05,
+      "loss": 0.4996,
+      "step": 5618
+    },
+    {
+      "epoch": 1.2534017399063129,
+      "grad_norm": 0.16951128840446472,
+      "learning_rate": 1.2688261276330882e-05,
+      "loss": 0.4908,
+      "step": 5619
+    },
+    {
+      "epoch": 1.2536248048182022,
+      "grad_norm": 0.1658773273229599,
+      "learning_rate": 1.2685994405432788e-05,
+      "loss": 0.4776,
+      "step": 5620
+    },
+    {
+      "epoch": 1.2538478697300914,
+      "grad_norm": 0.15743979811668396,
+      "learning_rate": 1.268372738576743e-05,
+      "loss": 0.46,
+      "step": 5621
+    },
+    {
+      "epoch": 1.2540709346419807,
+      "grad_norm": 0.16387997567653656,
+      "learning_rate": 1.2681460217460365e-05,
+      "loss": 0.4833,
+      "step": 5622
+    },
+    {
+      "epoch": 1.2542939995538702,
+      "grad_norm": 0.16733065247535706,
+      "learning_rate": 1.2679192900637172e-05,
+      "loss": 0.4621,
+      "step": 5623
+    },
+    {
+      "epoch": 1.2545170644657595,
+      "grad_norm": 0.16208809614181519,
+      "learning_rate": 1.267692543542342e-05,
+      "loss": 0.4688,
+      "step": 5624
+    },
+    {
+      "epoch": 1.254740129377649,
+      "grad_norm": 0.16228267550468445,
+      "learning_rate": 1.2674657821944699e-05,
+      "loss": 0.4499,
+      "step": 5625
+    },
+    {
+      "epoch": 1.2549631942895383,
+      "grad_norm": 0.17010217905044556,
+      "learning_rate": 1.2672390060326603e-05,
+      "loss": 0.4411,
+      "step": 5626
+    },
+    {
+      "epoch": 1.2551862592014276,
+      "grad_norm": 0.1624649167060852,
+      "learning_rate": 1.2670122150694737e-05,
+      "loss": 0.451,
+      "step": 5627
+    },
+    {
+      "epoch": 1.2554093241133168,
+      "grad_norm": 0.16551491618156433,
+      "learning_rate": 1.2667854093174707e-05,
+      "loss": 0.4502,
+      "step": 5628
+    },
+    {
+      "epoch": 1.2556323890252064,
+      "grad_norm": 0.16004303097724915,
+      "learning_rate": 1.266558588789214e-05,
+      "loss": 0.4371,
+      "step": 5629
+    },
+    {
+      "epoch": 1.2558554539370956,
+      "grad_norm": 0.1660148799419403,
+      "learning_rate": 1.2663317534972656e-05,
+      "loss": 0.4518,
+      "step": 5630
+    },
+    {
+      "epoch": 1.2560785188489851,
+      "grad_norm": 0.16411980986595154,
+      "learning_rate": 1.2661049034541897e-05,
+      "loss": 0.4644,
+      "step": 5631
+    },
+    {
+      "epoch": 1.2563015837608744,
+      "grad_norm": 0.16674858331680298,
+      "learning_rate": 1.2658780386725503e-05,
+      "loss": 0.4384,
+      "step": 5632
+    },
+    {
+      "epoch": 1.2565246486727637,
+      "grad_norm": 0.17275111377239227,
+      "learning_rate": 1.2656511591649125e-05,
+      "loss": 0.4698,
+      "step": 5633
+    },
+    {
+      "epoch": 1.2567477135846532,
+      "grad_norm": 0.16666240990161896,
+      "learning_rate": 1.2654242649438426e-05,
+      "loss": 0.4706,
+      "step": 5634
+    },
+    {
+      "epoch": 1.2569707784965425,
+      "grad_norm": 0.1748340129852295,
+      "learning_rate": 1.2651973560219073e-05,
+      "loss": 0.4442,
+      "step": 5635
+    },
+    {
+      "epoch": 1.257193843408432,
+      "grad_norm": 0.16753262281417847,
+      "learning_rate": 1.2649704324116745e-05,
+      "loss": 0.4586,
+      "step": 5636
+    },
+    {
+      "epoch": 1.2574169083203213,
+      "grad_norm": 0.1677563339471817,
+      "learning_rate": 1.264743494125712e-05,
+      "loss": 0.4699,
+      "step": 5637
+    },
+    {
+      "epoch": 1.2576399732322106,
+      "grad_norm": 0.16369731724262238,
+      "learning_rate": 1.2645165411765899e-05,
+      "loss": 0.4658,
+      "step": 5638
+    },
+    {
+      "epoch": 1.2578630381440998,
+      "grad_norm": 0.18121913075447083,
+      "learning_rate": 1.2642895735768775e-05,
+      "loss": 0.453,
+      "step": 5639
+    },
+    {
+      "epoch": 1.2580861030559893,
+      "grad_norm": 0.16915792226791382,
+      "learning_rate": 1.2640625913391464e-05,
+      "loss": 0.4496,
+      "step": 5640
+    },
+    {
+      "epoch": 1.2583091679678786,
+      "grad_norm": 0.17231988906860352,
+      "learning_rate": 1.2638355944759678e-05,
+      "loss": 0.4594,
+      "step": 5641
+    },
+    {
+      "epoch": 1.2585322328797681,
+      "grad_norm": 0.17405405640602112,
+      "learning_rate": 1.2636085829999145e-05,
+      "loss": 0.4665,
+      "step": 5642
+    },
+    {
+      "epoch": 1.2587552977916574,
+      "grad_norm": 0.17699532210826874,
+      "learning_rate": 1.2633815569235594e-05,
+      "loss": 0.4856,
+      "step": 5643
+    },
+    {
+      "epoch": 1.2589783627035467,
+      "grad_norm": 0.16303865611553192,
+      "learning_rate": 1.2631545162594773e-05,
+      "loss": 0.4475,
+      "step": 5644
+    },
+    {
+      "epoch": 1.259201427615436,
+      "grad_norm": 0.40841978788375854,
+      "learning_rate": 1.2629274610202427e-05,
+      "loss": 0.4419,
+      "step": 5645
+    },
+    {
+      "epoch": 1.2594244925273255,
+      "grad_norm": 0.17015258967876434,
+      "learning_rate": 1.2627003912184315e-05,
+      "loss": 0.4642,
+      "step": 5646
+    },
+    {
+      "epoch": 1.2596475574392147,
+      "grad_norm": 0.16850386559963226,
+      "learning_rate": 1.26247330686662e-05,
+      "loss": 0.4751,
+      "step": 5647
+    },
+    {
+      "epoch": 1.2598706223511043,
+      "grad_norm": 0.16403812170028687,
+      "learning_rate": 1.2622462079773859e-05,
+      "loss": 0.4555,
+      "step": 5648
+    },
+    {
+      "epoch": 1.2600936872629935,
+      "grad_norm": 0.16475041210651398,
+      "learning_rate": 1.2620190945633069e-05,
+      "loss": 0.4456,
+      "step": 5649
+    },
+    {
+      "epoch": 1.2603167521748828,
+      "grad_norm": 0.1695244163274765,
+      "learning_rate": 1.2617919666369627e-05,
+      "loss": 0.4489,
+      "step": 5650
+    },
+    {
+      "epoch": 1.2605398170867723,
+      "grad_norm": 0.16356348991394043,
+      "learning_rate": 1.2615648242109324e-05,
+      "loss": 0.4545,
+      "step": 5651
+    },
+    {
+      "epoch": 1.2607628819986616,
+      "grad_norm": 0.17045846581459045,
+      "learning_rate": 1.2613376672977968e-05,
+      "loss": 0.4734,
+      "step": 5652
+    },
+    {
+      "epoch": 1.260985946910551,
+      "grad_norm": 0.15884579718112946,
+      "learning_rate": 1.2611104959101374e-05,
+      "loss": 0.4357,
+      "step": 5653
+    },
+    {
+      "epoch": 1.2612090118224404,
+      "grad_norm": 0.16921289265155792,
+      "learning_rate": 1.2608833100605361e-05,
+      "loss": 0.4714,
+      "step": 5654
+    },
+    {
+      "epoch": 1.2614320767343297,
+      "grad_norm": 0.16710084676742554,
+      "learning_rate": 1.2606561097615764e-05,
+      "loss": 0.4882,
+      "step": 5655
+    },
+    {
+      "epoch": 1.261655141646219,
+      "grad_norm": 0.16774149239063263,
+      "learning_rate": 1.2604288950258414e-05,
+      "loss": 0.4501,
+      "step": 5656
+    },
+    {
+      "epoch": 1.2618782065581085,
+      "grad_norm": 0.16943363845348358,
+      "learning_rate": 1.2602016658659167e-05,
+      "loss": 0.4764,
+      "step": 5657
+    },
+    {
+      "epoch": 1.2621012714699977,
+      "grad_norm": 0.16601496934890747,
+      "learning_rate": 1.2599744222943864e-05,
+      "loss": 0.4549,
+      "step": 5658
+    },
+    {
+      "epoch": 1.2623243363818872,
+      "grad_norm": 0.1612405627965927,
+      "learning_rate": 1.2597471643238372e-05,
+      "loss": 0.4451,
+      "step": 5659
+    },
+    {
+      "epoch": 1.2625474012937765,
+      "grad_norm": 0.16680100560188293,
+      "learning_rate": 1.2595198919668566e-05,
+      "loss": 0.475,
+      "step": 5660
+    },
+    {
+      "epoch": 1.2627704662056658,
+      "grad_norm": 0.16711308062076569,
+      "learning_rate": 1.2592926052360316e-05,
+      "loss": 0.4574,
+      "step": 5661
+    },
+    {
+      "epoch": 1.262993531117555,
+      "grad_norm": 0.16317816078662872,
+      "learning_rate": 1.259065304143951e-05,
+      "loss": 0.4345,
+      "step": 5662
+    },
+    {
+      "epoch": 1.2632165960294446,
+      "grad_norm": 0.16551434993743896,
+      "learning_rate": 1.2588379887032048e-05,
+      "loss": 0.4706,
+      "step": 5663
+    },
+    {
+      "epoch": 1.2634396609413339,
+      "grad_norm": 0.16414450109004974,
+      "learning_rate": 1.2586106589263823e-05,
+      "loss": 0.4574,
+      "step": 5664
+    },
+    {
+      "epoch": 1.2636627258532234,
+      "grad_norm": 0.17107480764389038,
+      "learning_rate": 1.2583833148260749e-05,
+      "loss": 0.4698,
+      "step": 5665
+    },
+    {
+      "epoch": 1.2638857907651126,
+      "grad_norm": 0.17316211760044098,
+      "learning_rate": 1.258155956414874e-05,
+      "loss": 0.4772,
+      "step": 5666
+    },
+    {
+      "epoch": 1.264108855677002,
+      "grad_norm": 0.16382986307144165,
+      "learning_rate": 1.2579285837053722e-05,
+      "loss": 0.4463,
+      "step": 5667
+    },
+    {
+      "epoch": 1.2643319205888914,
+      "grad_norm": 0.20516705513000488,
+      "learning_rate": 1.2577011967101636e-05,
+      "loss": 0.4539,
+      "step": 5668
+    },
+    {
+      "epoch": 1.2645549855007807,
+      "grad_norm": 0.17769816517829895,
+      "learning_rate": 1.2574737954418412e-05,
+      "loss": 0.4889,
+      "step": 5669
+    },
+    {
+      "epoch": 1.2647780504126702,
+      "grad_norm": 0.16899065673351288,
+      "learning_rate": 1.2572463799130008e-05,
+      "loss": 0.476,
+      "step": 5670
+    },
+    {
+      "epoch": 1.2650011153245595,
+      "grad_norm": 0.16549617052078247,
+      "learning_rate": 1.2570189501362375e-05,
+      "loss": 0.4493,
+      "step": 5671
+    },
+    {
+      "epoch": 1.2652241802364488,
+      "grad_norm": 0.16648399829864502,
+      "learning_rate": 1.2567915061241483e-05,
+      "loss": 0.4375,
+      "step": 5672
+    },
+    {
+      "epoch": 1.265447245148338,
+      "grad_norm": 0.16941656172275543,
+      "learning_rate": 1.2565640478893299e-05,
+      "loss": 0.4543,
+      "step": 5673
+    },
+    {
+      "epoch": 1.2656703100602276,
+      "grad_norm": 0.1804044395685196,
+      "learning_rate": 1.2563365754443808e-05,
+      "loss": 0.4722,
+      "step": 5674
+    },
+    {
+      "epoch": 1.2658933749721168,
+      "grad_norm": 0.18117646872997284,
+      "learning_rate": 1.2561090888018996e-05,
+      "loss": 0.4869,
+      "step": 5675
+    },
+    {
+      "epoch": 1.2661164398840064,
+      "grad_norm": 0.16993004083633423,
+      "learning_rate": 1.2558815879744865e-05,
+      "loss": 0.4665,
+      "step": 5676
+    },
+    {
+      "epoch": 1.2663395047958956,
+      "grad_norm": 0.19133198261260986,
+      "learning_rate": 1.255654072974741e-05,
+      "loss": 0.4619,
+      "step": 5677
+    },
+    {
+      "epoch": 1.266562569707785,
+      "grad_norm": 0.17450137436389923,
+      "learning_rate": 1.2554265438152653e-05,
+      "loss": 0.4638,
+      "step": 5678
+    },
+    {
+      "epoch": 1.2667856346196742,
+      "grad_norm": 0.1640671044588089,
+      "learning_rate": 1.2551990005086604e-05,
+      "loss": 0.4364,
+      "step": 5679
+    },
+    {
+      "epoch": 1.2670086995315637,
+      "grad_norm": 0.16710147261619568,
+      "learning_rate": 1.2549714430675299e-05,
+      "loss": 0.4479,
+      "step": 5680
+    },
+    {
+      "epoch": 1.267231764443453,
+      "grad_norm": 0.17442373931407928,
+      "learning_rate": 1.2547438715044769e-05,
+      "loss": 0.4517,
+      "step": 5681
+    },
+    {
+      "epoch": 1.2674548293553425,
+      "grad_norm": 0.18162380158901215,
+      "learning_rate": 1.254516285832106e-05,
+      "loss": 0.4829,
+      "step": 5682
+    },
+    {
+      "epoch": 1.2676778942672318,
+      "grad_norm": 0.17129097878932953,
+      "learning_rate": 1.2542886860630221e-05,
+      "loss": 0.4519,
+      "step": 5683
+    },
+    {
+      "epoch": 1.267900959179121,
+      "grad_norm": 0.1645333468914032,
+      "learning_rate": 1.2540610722098314e-05,
+      "loss": 0.4544,
+      "step": 5684
+    },
+    {
+      "epoch": 1.2681240240910105,
+      "grad_norm": 0.21202479302883148,
+      "learning_rate": 1.2538334442851403e-05,
+      "loss": 0.5017,
+      "step": 5685
+    },
+    {
+      "epoch": 1.2683470890028998,
+      "grad_norm": 0.21118725836277008,
+      "learning_rate": 1.2536058023015563e-05,
+      "loss": 0.477,
+      "step": 5686
+    },
+    {
+      "epoch": 1.2685701539147893,
+      "grad_norm": 0.16621573269367218,
+      "learning_rate": 1.2533781462716879e-05,
+      "loss": 0.4427,
+      "step": 5687
+    },
+    {
+      "epoch": 1.2687932188266786,
+      "grad_norm": 0.20093737542629242,
+      "learning_rate": 1.2531504762081437e-05,
+      "loss": 0.4873,
+      "step": 5688
+    },
+    {
+      "epoch": 1.269016283738568,
+      "grad_norm": 0.1732548326253891,
+      "learning_rate": 1.2529227921235342e-05,
+      "loss": 0.4745,
+      "step": 5689
+    },
+    {
+      "epoch": 1.2692393486504572,
+      "grad_norm": 0.1718747317790985,
+      "learning_rate": 1.252695094030469e-05,
+      "loss": 0.4478,
+      "step": 5690
+    },
+    {
+      "epoch": 1.2694624135623467,
+      "grad_norm": 0.1892472505569458,
+      "learning_rate": 1.2524673819415602e-05,
+      "loss": 0.457,
+      "step": 5691
+    },
+    {
+      "epoch": 1.269685478474236,
+      "grad_norm": 0.16968894004821777,
+      "learning_rate": 1.2522396558694197e-05,
+      "loss": 0.4734,
+      "step": 5692
+    },
+    {
+      "epoch": 1.2699085433861255,
+      "grad_norm": 0.1685362160205841,
+      "learning_rate": 1.2520119158266606e-05,
+      "loss": 0.4233,
+      "step": 5693
+    },
+    {
+      "epoch": 1.2701316082980147,
+      "grad_norm": 0.16388767957687378,
+      "learning_rate": 1.2517841618258961e-05,
+      "loss": 0.4135,
+      "step": 5694
+    },
+    {
+      "epoch": 1.270354673209904,
+      "grad_norm": 0.16109982132911682,
+      "learning_rate": 1.251556393879741e-05,
+      "loss": 0.4484,
+      "step": 5695
+    },
+    {
+      "epoch": 1.2705777381217933,
+      "grad_norm": 0.18354761600494385,
+      "learning_rate": 1.2513286120008105e-05,
+      "loss": 0.4358,
+      "step": 5696
+    },
+    {
+      "epoch": 1.2708008030336828,
+      "grad_norm": 0.17148545384407043,
+      "learning_rate": 1.2511008162017209e-05,
+      "loss": 0.4646,
+      "step": 5697
+    },
+    {
+      "epoch": 1.271023867945572,
+      "grad_norm": 0.17020155489444733,
+      "learning_rate": 1.2508730064950881e-05,
+      "loss": 0.4587,
+      "step": 5698
+    },
+    {
+      "epoch": 1.2712469328574616,
+      "grad_norm": 0.16151095926761627,
+      "learning_rate": 1.2506451828935303e-05,
+      "loss": 0.4886,
+      "step": 5699
+    },
+    {
+      "epoch": 1.2714699977693509,
+      "grad_norm": 0.1669510155916214,
+      "learning_rate": 1.2504173454096658e-05,
+      "loss": 0.4432,
+      "step": 5700
+    },
+    {
+      "epoch": 1.2716930626812402,
+      "grad_norm": 0.16603168845176697,
+      "learning_rate": 1.2501894940561133e-05,
+      "loss": 0.4556,
+      "step": 5701
+    },
+    {
+      "epoch": 1.2719161275931297,
+      "grad_norm": 0.1668505221605301,
+      "learning_rate": 1.249961628845493e-05,
+      "loss": 0.477,
+      "step": 5702
+    },
+    {
+      "epoch": 1.272139192505019,
+      "grad_norm": 0.1614041030406952,
+      "learning_rate": 1.2497337497904251e-05,
+      "loss": 0.4273,
+      "step": 5703
+    },
+    {
+      "epoch": 1.2723622574169084,
+      "grad_norm": 0.1707865595817566,
+      "learning_rate": 1.2495058569035316e-05,
+      "loss": 0.4562,
+      "step": 5704
+    },
+    {
+      "epoch": 1.2725853223287977,
+      "grad_norm": 0.17685820162296295,
+      "learning_rate": 1.249277950197434e-05,
+      "loss": 0.4693,
+      "step": 5705
+    },
+    {
+      "epoch": 1.272808387240687,
+      "grad_norm": 0.16811825335025787,
+      "learning_rate": 1.2490500296847558e-05,
+      "loss": 0.4576,
+      "step": 5706
+    },
+    {
+      "epoch": 1.2730314521525763,
+      "grad_norm": 0.17165428400039673,
+      "learning_rate": 1.2488220953781201e-05,
+      "loss": 0.4851,
+      "step": 5707
+    },
+    {
+      "epoch": 1.2732545170644658,
+      "grad_norm": 0.1754395067691803,
+      "learning_rate": 1.2485941472901519e-05,
+      "loss": 0.4895,
+      "step": 5708
+    },
+    {
+      "epoch": 1.273477581976355,
+      "grad_norm": 0.1676994264125824,
+      "learning_rate": 1.2483661854334756e-05,
+      "loss": 0.4675,
+      "step": 5709
+    },
+    {
+      "epoch": 1.2737006468882446,
+      "grad_norm": 0.16332073509693146,
+      "learning_rate": 1.2481382098207181e-05,
+      "loss": 0.4444,
+      "step": 5710
+    },
+    {
+      "epoch": 1.2739237118001339,
+      "grad_norm": 0.175141841173172,
+      "learning_rate": 1.2479102204645057e-05,
+      "loss": 0.4934,
+      "step": 5711
+    },
+    {
+      "epoch": 1.2741467767120231,
+      "grad_norm": 0.17090824246406555,
+      "learning_rate": 1.247682217377466e-05,
+      "loss": 0.4588,
+      "step": 5712
+    },
+    {
+      "epoch": 1.2743698416239124,
+      "grad_norm": 0.1759280413389206,
+      "learning_rate": 1.2474542005722265e-05,
+      "loss": 0.4487,
+      "step": 5713
+    },
+    {
+      "epoch": 1.274592906535802,
+      "grad_norm": 0.16140590608119965,
+      "learning_rate": 1.2472261700614174e-05,
+      "loss": 0.451,
+      "step": 5714
+    },
+    {
+      "epoch": 1.2748159714476912,
+      "grad_norm": 0.19568322598934174,
+      "learning_rate": 1.2469981258576676e-05,
+      "loss": 0.4726,
+      "step": 5715
+    },
+    {
+      "epoch": 1.2750390363595807,
+      "grad_norm": 0.16794992983341217,
+      "learning_rate": 1.246770067973608e-05,
+      "loss": 0.465,
+      "step": 5716
+    },
+    {
+      "epoch": 1.27526210127147,
+      "grad_norm": 0.17148984968662262,
+      "learning_rate": 1.2465419964218696e-05,
+      "loss": 0.4809,
+      "step": 5717
+    },
+    {
+      "epoch": 1.2754851661833593,
+      "grad_norm": 0.17300093173980713,
+      "learning_rate": 1.2463139112150851e-05,
+      "loss": 0.4417,
+      "step": 5718
+    },
+    {
+      "epoch": 1.2757082310952488,
+      "grad_norm": 0.16525490581989288,
+      "learning_rate": 1.2460858123658863e-05,
+      "loss": 0.4744,
+      "step": 5719
+    },
+    {
+      "epoch": 1.275931296007138,
+      "grad_norm": 0.16058532893657684,
+      "learning_rate": 1.2458576998869076e-05,
+      "loss": 0.4494,
+      "step": 5720
+    },
+    {
+      "epoch": 1.2761543609190276,
+      "grad_norm": 0.16562692821025848,
+      "learning_rate": 1.2456295737907828e-05,
+      "loss": 0.3971,
+      "step": 5721
+    },
+    {
+      "epoch": 1.2763774258309168,
+      "grad_norm": 0.1694319099187851,
+      "learning_rate": 1.2454014340901472e-05,
+      "loss": 0.4956,
+      "step": 5722
+    },
+    {
+      "epoch": 1.2766004907428061,
+      "grad_norm": 0.1706772893667221,
+      "learning_rate": 1.2451732807976367e-05,
+      "loss": 0.487,
+      "step": 5723
+    },
+    {
+      "epoch": 1.2768235556546954,
+      "grad_norm": 0.16685868799686432,
+      "learning_rate": 1.2449451139258875e-05,
+      "loss": 0.4671,
+      "step": 5724
+    },
+    {
+      "epoch": 1.277046620566585,
+      "grad_norm": 0.16977933049201965,
+      "learning_rate": 1.2447169334875374e-05,
+      "loss": 0.4633,
+      "step": 5725
+    },
+    {
+      "epoch": 1.2772696854784742,
+      "grad_norm": 0.1995961219072342,
+      "learning_rate": 1.2444887394952237e-05,
+      "loss": 0.4676,
+      "step": 5726
+    },
+    {
+      "epoch": 1.2774927503903637,
+      "grad_norm": 0.17159555852413177,
+      "learning_rate": 1.2442605319615862e-05,
+      "loss": 0.4884,
+      "step": 5727
+    },
+    {
+      "epoch": 1.277715815302253,
+      "grad_norm": 0.18190377950668335,
+      "learning_rate": 1.2440323108992635e-05,
+      "loss": 0.4802,
+      "step": 5728
+    },
+    {
+      "epoch": 1.2779388802141423,
+      "grad_norm": 0.16413584351539612,
+      "learning_rate": 1.2438040763208967e-05,
+      "loss": 0.4475,
+      "step": 5729
+    },
+    {
+      "epoch": 1.2781619451260315,
+      "grad_norm": 0.17521817982196808,
+      "learning_rate": 1.2435758282391266e-05,
+      "loss": 0.4559,
+      "step": 5730
+    },
+    {
+      "epoch": 1.278385010037921,
+      "grad_norm": 0.17562466859817505,
+      "learning_rate": 1.243347566666595e-05,
+      "loss": 0.4614,
+      "step": 5731
+    },
+    {
+      "epoch": 1.2786080749498103,
+      "grad_norm": 0.18439187109470367,
+      "learning_rate": 1.2431192916159442e-05,
+      "loss": 0.4739,
+      "step": 5732
+    },
+    {
+      "epoch": 1.2788311398616998,
+      "grad_norm": 0.17046132683753967,
+      "learning_rate": 1.242891003099818e-05,
+      "loss": 0.4653,
+      "step": 5733
+    },
+    {
+      "epoch": 1.279054204773589,
+      "grad_norm": 0.17075121402740479,
+      "learning_rate": 1.24266270113086e-05,
+      "loss": 0.5025,
+      "step": 5734
+    },
+    {
+      "epoch": 1.2792772696854784,
+      "grad_norm": 0.1661718189716339,
+      "learning_rate": 1.2424343857217153e-05,
+      "loss": 0.4574,
+      "step": 5735
+    },
+    {
+      "epoch": 1.279500334597368,
+      "grad_norm": 0.1734541356563568,
+      "learning_rate": 1.2422060568850293e-05,
+      "loss": 0.4579,
+      "step": 5736
+    },
+    {
+      "epoch": 1.2797233995092572,
+      "grad_norm": 0.16650691628456116,
+      "learning_rate": 1.2419777146334486e-05,
+      "loss": 0.4793,
+      "step": 5737
+    },
+    {
+      "epoch": 1.2799464644211467,
+      "grad_norm": 0.16996720433235168,
+      "learning_rate": 1.2417493589796199e-05,
+      "loss": 0.4679,
+      "step": 5738
+    },
+    {
+      "epoch": 1.280169529333036,
+      "grad_norm": 0.1680038571357727,
+      "learning_rate": 1.2415209899361908e-05,
+      "loss": 0.4842,
+      "step": 5739
+    },
+    {
+      "epoch": 1.2803925942449252,
+      "grad_norm": 0.21281017363071442,
+      "learning_rate": 1.2412926075158103e-05,
+      "loss": 0.4534,
+      "step": 5740
+    },
+    {
+      "epoch": 1.2806156591568145,
+      "grad_norm": 0.16885453462600708,
+      "learning_rate": 1.241064211731127e-05,
+      "loss": 0.4758,
+      "step": 5741
+    },
+    {
+      "epoch": 1.280838724068704,
+      "grad_norm": 0.19396549463272095,
+      "learning_rate": 1.2408358025947917e-05,
+      "loss": 0.4729,
+      "step": 5742
+    },
+    {
+      "epoch": 1.2810617889805933,
+      "grad_norm": 0.1660171002149582,
+      "learning_rate": 1.2406073801194546e-05,
+      "loss": 0.4574,
+      "step": 5743
+    },
+    {
+      "epoch": 1.2812848538924828,
+      "grad_norm": 0.17650169134140015,
+      "learning_rate": 1.2403789443177672e-05,
+      "loss": 0.452,
+      "step": 5744
+    },
+    {
+      "epoch": 1.281507918804372,
+      "grad_norm": 0.16995863616466522,
+      "learning_rate": 1.240150495202382e-05,
+      "loss": 0.4556,
+      "step": 5745
+    },
+    {
+      "epoch": 1.2817309837162614,
+      "grad_norm": 0.17916269600391388,
+      "learning_rate": 1.2399220327859516e-05,
+      "loss": 0.4595,
+      "step": 5746
+    },
+    {
+      "epoch": 1.2819540486281507,
+      "grad_norm": 0.16572338342666626,
+      "learning_rate": 1.2396935570811299e-05,
+      "loss": 0.4408,
+      "step": 5747
+    },
+    {
+      "epoch": 1.2821771135400402,
+      "grad_norm": 0.1570359319448471,
+      "learning_rate": 1.2394650681005713e-05,
+      "loss": 0.4357,
+      "step": 5748
+    },
+    {
+      "epoch": 1.2824001784519294,
+      "grad_norm": 0.1629701554775238,
+      "learning_rate": 1.239236565856931e-05,
+      "loss": 0.4481,
+      "step": 5749
+    },
+    {
+      "epoch": 1.282623243363819,
+      "grad_norm": 0.17372570931911469,
+      "learning_rate": 1.2390080503628647e-05,
+      "loss": 0.4717,
+      "step": 5750
+    },
+    {
+      "epoch": 1.2828463082757082,
+      "grad_norm": 0.1692957580089569,
+      "learning_rate": 1.2387795216310292e-05,
+      "loss": 0.4665,
+      "step": 5751
+    },
+    {
+      "epoch": 1.2830693731875975,
+      "grad_norm": 0.16883119940757751,
+      "learning_rate": 1.2385509796740818e-05,
+      "loss": 0.4566,
+      "step": 5752
+    },
+    {
+      "epoch": 1.283292438099487,
+      "grad_norm": 0.1655278503894806,
+      "learning_rate": 1.2383224245046805e-05,
+      "loss": 0.4482,
+      "step": 5753
+    },
+    {
+      "epoch": 1.2835155030113763,
+      "grad_norm": 0.17099110782146454,
+      "learning_rate": 1.2380938561354846e-05,
+      "loss": 0.4733,
+      "step": 5754
+    },
+    {
+      "epoch": 1.2837385679232658,
+      "grad_norm": 0.16379962861537933,
+      "learning_rate": 1.2378652745791528e-05,
+      "loss": 0.4337,
+      "step": 5755
+    },
+    {
+      "epoch": 1.283961632835155,
+      "grad_norm": 0.1781388372182846,
+      "learning_rate": 1.237636679848346e-05,
+      "loss": 0.4642,
+      "step": 5756
+    },
+    {
+      "epoch": 1.2841846977470444,
+      "grad_norm": 0.16018837690353394,
+      "learning_rate": 1.2374080719557253e-05,
+      "loss": 0.4488,
+      "step": 5757
+    },
+    {
+      "epoch": 1.2844077626589336,
+      "grad_norm": 0.17385733127593994,
+      "learning_rate": 1.237179450913952e-05,
+      "loss": 0.4448,
+      "step": 5758
+    },
+    {
+      "epoch": 1.2846308275708231,
+      "grad_norm": 0.1600496470928192,
+      "learning_rate": 1.236950816735689e-05,
+      "loss": 0.4523,
+      "step": 5759
+    },
+    {
+      "epoch": 1.2848538924827124,
+      "grad_norm": 0.17114561796188354,
+      "learning_rate": 1.2367221694335992e-05,
+      "loss": 0.4717,
+      "step": 5760
+    },
+    {
+      "epoch": 1.285076957394602,
+      "grad_norm": 0.17317497730255127,
+      "learning_rate": 1.2364935090203464e-05,
+      "loss": 0.4571,
+      "step": 5761
+    },
+    {
+      "epoch": 1.2853000223064912,
+      "grad_norm": 0.18465475738048553,
+      "learning_rate": 1.2362648355085958e-05,
+      "loss": 0.4622,
+      "step": 5762
+    },
+    {
+      "epoch": 1.2855230872183805,
+      "grad_norm": 0.16968859732151031,
+      "learning_rate": 1.2360361489110123e-05,
+      "loss": 0.445,
+      "step": 5763
+    },
+    {
+      "epoch": 1.2857461521302698,
+      "grad_norm": 0.16804584860801697,
+      "learning_rate": 1.235807449240262e-05,
+      "loss": 0.4482,
+      "step": 5764
+    },
+    {
+      "epoch": 1.2859692170421593,
+      "grad_norm": 0.170896977186203,
+      "learning_rate": 1.2355787365090122e-05,
+      "loss": 0.4517,
+      "step": 5765
+    },
+    {
+      "epoch": 1.2861922819540486,
+      "grad_norm": 0.1710241287946701,
+      "learning_rate": 1.2353500107299299e-05,
+      "loss": 0.4884,
+      "step": 5766
+    },
+    {
+      "epoch": 1.286415346865938,
+      "grad_norm": 0.16201786696910858,
+      "learning_rate": 1.2351212719156835e-05,
+      "loss": 0.4542,
+      "step": 5767
+    },
+    {
+      "epoch": 1.2866384117778273,
+      "grad_norm": 0.16651326417922974,
+      "learning_rate": 1.234892520078942e-05,
+      "loss": 0.4649,
+      "step": 5768
+    },
+    {
+      "epoch": 1.2868614766897166,
+      "grad_norm": 0.16634796559810638,
+      "learning_rate": 1.2346637552323757e-05,
+      "loss": 0.4793,
+      "step": 5769
+    },
+    {
+      "epoch": 1.2870845416016061,
+      "grad_norm": 0.16899818181991577,
+      "learning_rate": 1.2344349773886542e-05,
+      "loss": 0.4633,
+      "step": 5770
+    },
+    {
+      "epoch": 1.2873076065134954,
+      "grad_norm": 0.17024937272071838,
+      "learning_rate": 1.2342061865604492e-05,
+      "loss": 0.4471,
+      "step": 5771
+    },
+    {
+      "epoch": 1.287530671425385,
+      "grad_norm": 0.1707090586423874,
+      "learning_rate": 1.2339773827604322e-05,
+      "loss": 0.4758,
+      "step": 5772
+    },
+    {
+      "epoch": 1.2877537363372742,
+      "grad_norm": 0.17313364148139954,
+      "learning_rate": 1.2337485660012757e-05,
+      "loss": 0.4858,
+      "step": 5773
+    },
+    {
+      "epoch": 1.2879768012491635,
+      "grad_norm": 0.1700849086046219,
+      "learning_rate": 1.2335197362956537e-05,
+      "loss": 0.4505,
+      "step": 5774
+    },
+    {
+      "epoch": 1.2881998661610528,
+      "grad_norm": 0.17922791838645935,
+      "learning_rate": 1.2332908936562395e-05,
+      "loss": 0.4559,
+      "step": 5775
+    },
+    {
+      "epoch": 1.2884229310729423,
+      "grad_norm": 0.1638413369655609,
+      "learning_rate": 1.2330620380957086e-05,
+      "loss": 0.4526,
+      "step": 5776
+    },
+    {
+      "epoch": 1.2886459959848315,
+      "grad_norm": 0.15971173346042633,
+      "learning_rate": 1.2328331696267357e-05,
+      "loss": 0.4386,
+      "step": 5777
+    },
+    {
+      "epoch": 1.288869060896721,
+      "grad_norm": 0.17769648134708405,
+      "learning_rate": 1.2326042882619973e-05,
+      "loss": 0.4659,
+      "step": 5778
+    },
+    {
+      "epoch": 1.2890921258086103,
+      "grad_norm": 0.16623780131340027,
+      "learning_rate": 1.2323753940141704e-05,
+      "loss": 0.4593,
+      "step": 5779
+    },
+    {
+      "epoch": 1.2893151907204996,
+      "grad_norm": 0.16500690579414368,
+      "learning_rate": 1.2321464868959326e-05,
+      "loss": 0.45,
+      "step": 5780
+    },
+    {
+      "epoch": 1.289538255632389,
+      "grad_norm": 0.17217601835727692,
+      "learning_rate": 1.2319175669199619e-05,
+      "loss": 0.4747,
+      "step": 5781
+    },
+    {
+      "epoch": 1.2897613205442784,
+      "grad_norm": 0.16614340245723724,
+      "learning_rate": 1.2316886340989375e-05,
+      "loss": 0.4685,
+      "step": 5782
+    },
+    {
+      "epoch": 1.2899843854561677,
+      "grad_norm": 0.1841314285993576,
+      "learning_rate": 1.231459688445539e-05,
+      "loss": 0.4537,
+      "step": 5783
+    },
+    {
+      "epoch": 1.2902074503680572,
+      "grad_norm": 0.1670687049627304,
+      "learning_rate": 1.231230729972447e-05,
+      "loss": 0.4763,
+      "step": 5784
+    },
+    {
+      "epoch": 1.2904305152799465,
+      "grad_norm": 0.1700257658958435,
+      "learning_rate": 1.2310017586923431e-05,
+      "loss": 0.4698,
+      "step": 5785
+    },
+    {
+      "epoch": 1.2906535801918357,
+      "grad_norm": 0.17322078347206116,
+      "learning_rate": 1.2307727746179085e-05,
+      "loss": 0.4422,
+      "step": 5786
+    },
+    {
+      "epoch": 1.2908766451037252,
+      "grad_norm": 0.18490198254585266,
+      "learning_rate": 1.230543777761826e-05,
+      "loss": 0.451,
+      "step": 5787
+    },
+    {
+      "epoch": 1.2910997100156145,
+      "grad_norm": 0.17593975365161896,
+      "learning_rate": 1.2303147681367788e-05,
+      "loss": 0.4806,
+      "step": 5788
+    },
+    {
+      "epoch": 1.291322774927504,
+      "grad_norm": 0.16735313832759857,
+      "learning_rate": 1.2300857457554513e-05,
+      "loss": 0.4551,
+      "step": 5789
+    },
+    {
+      "epoch": 1.2915458398393933,
+      "grad_norm": 0.18818075954914093,
+      "learning_rate": 1.2298567106305277e-05,
+      "loss": 0.4722,
+      "step": 5790
+    },
+    {
+      "epoch": 1.2917689047512826,
+      "grad_norm": 0.33603474497795105,
+      "learning_rate": 1.2296276627746938e-05,
+      "loss": 0.4789,
+      "step": 5791
+    },
+    {
+      "epoch": 1.2919919696631719,
+      "grad_norm": 0.16247035562992096,
+      "learning_rate": 1.2293986022006353e-05,
+      "loss": 0.4493,
+      "step": 5792
+    },
+    {
+      "epoch": 1.2922150345750614,
+      "grad_norm": 0.17141158878803253,
+      "learning_rate": 1.2291695289210395e-05,
+      "loss": 0.4718,
+      "step": 5793
+    },
+    {
+      "epoch": 1.2924380994869507,
+      "grad_norm": 0.15925908088684082,
+      "learning_rate": 1.2289404429485932e-05,
+      "loss": 0.4805,
+      "step": 5794
+    },
+    {
+      "epoch": 1.2926611643988402,
+      "grad_norm": 0.2104969471693039,
+      "learning_rate": 1.2287113442959854e-05,
+      "loss": 0.4718,
+      "step": 5795
+    },
+    {
+      "epoch": 1.2928842293107294,
+      "grad_norm": 0.1778918355703354,
+      "learning_rate": 1.2284822329759047e-05,
+      "loss": 0.4708,
+      "step": 5796
+    },
+    {
+      "epoch": 1.2931072942226187,
+      "grad_norm": 0.1711834967136383,
+      "learning_rate": 1.2282531090010408e-05,
+      "loss": 0.4613,
+      "step": 5797
+    },
+    {
+      "epoch": 1.2933303591345082,
+      "grad_norm": 0.17444384098052979,
+      "learning_rate": 1.2280239723840836e-05,
+      "loss": 0.4477,
+      "step": 5798
+    },
+    {
+      "epoch": 1.2935534240463975,
+      "grad_norm": 0.16771963238716125,
+      "learning_rate": 1.2277948231377247e-05,
+      "loss": 0.4698,
+      "step": 5799
+    },
+    {
+      "epoch": 1.2937764889582868,
+      "grad_norm": 0.1844215989112854,
+      "learning_rate": 1.2275656612746556e-05,
+      "loss": 0.4681,
+      "step": 5800
+    },
+    {
+      "epoch": 1.2939995538701763,
+      "grad_norm": 0.1627340018749237,
+      "learning_rate": 1.227336486807569e-05,
+      "loss": 0.4758,
+      "step": 5801
+    },
+    {
+      "epoch": 1.2942226187820656,
+      "grad_norm": 0.17234186828136444,
+      "learning_rate": 1.2271072997491573e-05,
+      "loss": 0.4646,
+      "step": 5802
+    },
+    {
+      "epoch": 1.2944456836939549,
+      "grad_norm": 0.169833242893219,
+      "learning_rate": 1.2268781001121151e-05,
+      "loss": 0.4709,
+      "step": 5803
+    },
+    {
+      "epoch": 1.2946687486058444,
+      "grad_norm": 0.16857881844043732,
+      "learning_rate": 1.2266488879091365e-05,
+      "loss": 0.484,
+      "step": 5804
+    },
+    {
+      "epoch": 1.2948918135177336,
+      "grad_norm": 0.16517959535121918,
+      "learning_rate": 1.2264196631529166e-05,
+      "loss": 0.4608,
+      "step": 5805
+    },
+    {
+      "epoch": 1.2951148784296231,
+      "grad_norm": 0.16706417500972748,
+      "learning_rate": 1.226190425856152e-05,
+      "loss": 0.4541,
+      "step": 5806
+    },
+    {
+      "epoch": 1.2953379433415124,
+      "grad_norm": 0.16569514572620392,
+      "learning_rate": 1.2259611760315381e-05,
+      "loss": 0.4393,
+      "step": 5807
+    },
+    {
+      "epoch": 1.2955610082534017,
+      "grad_norm": 0.17674359679222107,
+      "learning_rate": 1.2257319136917735e-05,
+      "loss": 0.4869,
+      "step": 5808
+    },
+    {
+      "epoch": 1.295784073165291,
+      "grad_norm": 0.1684064269065857,
+      "learning_rate": 1.2255026388495554e-05,
+      "loss": 0.488,
+      "step": 5809
+    },
+    {
+      "epoch": 1.2960071380771805,
+      "grad_norm": 0.16627109050750732,
+      "learning_rate": 1.2252733515175829e-05,
+      "loss": 0.4577,
+      "step": 5810
+    },
+    {
+      "epoch": 1.2962302029890698,
+      "grad_norm": 0.18663783371448517,
+      "learning_rate": 1.2250440517085549e-05,
+      "loss": 0.4512,
+      "step": 5811
+    },
+    {
+      "epoch": 1.2964532679009593,
+      "grad_norm": 0.17597997188568115,
+      "learning_rate": 1.2248147394351719e-05,
+      "loss": 0.4482,
+      "step": 5812
+    },
+    {
+      "epoch": 1.2966763328128486,
+      "grad_norm": 0.1624067723751068,
+      "learning_rate": 1.2245854147101344e-05,
+      "loss": 0.4809,
+      "step": 5813
+    },
+    {
+      "epoch": 1.2968993977247378,
+      "grad_norm": 0.1702115684747696,
+      "learning_rate": 1.2243560775461441e-05,
+      "loss": 0.4572,
+      "step": 5814
+    },
+    {
+      "epoch": 1.2971224626366273,
+      "grad_norm": 0.17795614898204803,
+      "learning_rate": 1.2241267279559029e-05,
+      "loss": 0.4668,
+      "step": 5815
+    },
+    {
+      "epoch": 1.2973455275485166,
+      "grad_norm": 0.17128905653953552,
+      "learning_rate": 1.2238973659521136e-05,
+      "loss": 0.4649,
+      "step": 5816
+    },
+    {
+      "epoch": 1.297568592460406,
+      "grad_norm": 0.1639898121356964,
+      "learning_rate": 1.2236679915474799e-05,
+      "loss": 0.4741,
+      "step": 5817
+    },
+    {
+      "epoch": 1.2977916573722954,
+      "grad_norm": 0.17181125283241272,
+      "learning_rate": 1.2234386047547057e-05,
+      "loss": 0.4569,
+      "step": 5818
+    },
+    {
+      "epoch": 1.2980147222841847,
+      "grad_norm": 0.17386600375175476,
+      "learning_rate": 1.2232092055864961e-05,
+      "loss": 0.4536,
+      "step": 5819
+    },
+    {
+      "epoch": 1.298237787196074,
+      "grad_norm": 0.1769663542509079,
+      "learning_rate": 1.222979794055557e-05,
+      "loss": 0.461,
+      "step": 5820
+    },
+    {
+      "epoch": 1.2984608521079635,
+      "grad_norm": 0.17835234105587006,
+      "learning_rate": 1.2227503701745942e-05,
+      "loss": 0.4535,
+      "step": 5821
+    },
+    {
+      "epoch": 1.2986839170198528,
+      "grad_norm": 0.1780080497264862,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.4532,
+      "step": 5822
+    },
+    {
+      "epoch": 1.2989069819317423,
+      "grad_norm": 0.1722138375043869,
+      "learning_rate": 1.2222914854134261e-05,
+      "loss": 0.4626,
+      "step": 5823
+    },
+    {
+      "epoch": 1.2991300468436315,
+      "grad_norm": 0.1659894585609436,
+      "learning_rate": 1.2220620245586365e-05,
+      "loss": 0.4466,
+      "step": 5824
+    },
+    {
+      "epoch": 1.2993531117555208,
+      "grad_norm": 0.17937776446342468,
+      "learning_rate": 1.2218325514046557e-05,
+      "loss": 0.4572,
+      "step": 5825
+    },
+    {
+      "epoch": 1.29957617666741,
+      "grad_norm": 0.18704891204833984,
+      "learning_rate": 1.2216030659641924e-05,
+      "loss": 0.4745,
+      "step": 5826
+    },
+    {
+      "epoch": 1.2997992415792996,
+      "grad_norm": 0.17715615034103394,
+      "learning_rate": 1.2213735682499578e-05,
+      "loss": 0.4536,
+      "step": 5827
+    },
+    {
+      "epoch": 1.3000223064911889,
+      "grad_norm": 0.1589982509613037,
+      "learning_rate": 1.2211440582746619e-05,
+      "loss": 0.4619,
+      "step": 5828
+    },
+    {
+      "epoch": 1.3002453714030784,
+      "grad_norm": 0.1834399253129959,
+      "learning_rate": 1.2209145360510175e-05,
+      "loss": 0.457,
+      "step": 5829
+    },
+    {
+      "epoch": 1.3004684363149677,
+      "grad_norm": 0.16908270120620728,
+      "learning_rate": 1.2206850015917362e-05,
+      "loss": 0.4432,
+      "step": 5830
+    },
+    {
+      "epoch": 1.300691501226857,
+      "grad_norm": 0.16872212290763855,
+      "learning_rate": 1.2204554549095316e-05,
+      "loss": 0.4378,
+      "step": 5831
+    },
+    {
+      "epoch": 1.3009145661387465,
+      "grad_norm": 0.1816396862268448,
+      "learning_rate": 1.2202258960171167e-05,
+      "loss": 0.4645,
+      "step": 5832
+    },
+    {
+      "epoch": 1.3011376310506357,
+      "grad_norm": 0.17055541276931763,
+      "learning_rate": 1.219996324927207e-05,
+      "loss": 0.4567,
+      "step": 5833
+    },
+    {
+      "epoch": 1.301360695962525,
+      "grad_norm": 0.17526254057884216,
+      "learning_rate": 1.2197667416525165e-05,
+      "loss": 0.4335,
+      "step": 5834
+    },
+    {
+      "epoch": 1.3015837608744145,
+      "grad_norm": 0.17466293275356293,
+      "learning_rate": 1.2195371462057619e-05,
+      "loss": 0.4645,
+      "step": 5835
+    },
+    {
+      "epoch": 1.3018068257863038,
+      "grad_norm": 0.1705324649810791,
+      "learning_rate": 1.2193075385996589e-05,
+      "loss": 0.4782,
+      "step": 5836
+    },
+    {
+      "epoch": 1.302029890698193,
+      "grad_norm": 0.1765318214893341,
+      "learning_rate": 1.2190779188469248e-05,
+      "loss": 0.4651,
+      "step": 5837
+    },
+    {
+      "epoch": 1.3022529556100826,
+      "grad_norm": 0.18065306544303894,
+      "learning_rate": 1.2188482869602778e-05,
+      "loss": 0.4507,
+      "step": 5838
+    },
+    {
+      "epoch": 1.3024760205219719,
+      "grad_norm": 0.17382916808128357,
+      "learning_rate": 1.2186186429524358e-05,
+      "loss": 0.4621,
+      "step": 5839
+    },
+    {
+      "epoch": 1.3026990854338614,
+      "grad_norm": 0.17227432131767273,
+      "learning_rate": 1.2183889868361185e-05,
+      "loss": 0.4493,
+      "step": 5840
+    },
+    {
+      "epoch": 1.3029221503457507,
+      "grad_norm": 0.16566970944404602,
+      "learning_rate": 1.218159318624045e-05,
+      "loss": 0.4613,
+      "step": 5841
+    },
+    {
+      "epoch": 1.30314521525764,
+      "grad_norm": 0.16630716621875763,
+      "learning_rate": 1.2179296383289366e-05,
+      "loss": 0.4703,
+      "step": 5842
+    },
+    {
+      "epoch": 1.3033682801695292,
+      "grad_norm": 0.16445283591747284,
+      "learning_rate": 1.2176999459635137e-05,
+      "loss": 0.471,
+      "step": 5843
+    },
+    {
+      "epoch": 1.3035913450814187,
+      "grad_norm": 0.1610761433839798,
+      "learning_rate": 1.2174702415404987e-05,
+      "loss": 0.4669,
+      "step": 5844
+    },
+    {
+      "epoch": 1.303814409993308,
+      "grad_norm": 0.16646409034729004,
+      "learning_rate": 1.2172405250726134e-05,
+      "loss": 0.4691,
+      "step": 5845
+    },
+    {
+      "epoch": 1.3040374749051975,
+      "grad_norm": 0.173641636967659,
+      "learning_rate": 1.2170107965725815e-05,
+      "loss": 0.4525,
+      "step": 5846
+    },
+    {
+      "epoch": 1.3042605398170868,
+      "grad_norm": 0.16962803900241852,
+      "learning_rate": 1.2167810560531266e-05,
+      "loss": 0.4491,
+      "step": 5847
+    },
+    {
+      "epoch": 1.304483604728976,
+      "grad_norm": 0.1649630218744278,
+      "learning_rate": 1.2165513035269733e-05,
+      "loss": 0.4409,
+      "step": 5848
+    },
+    {
+      "epoch": 1.3047066696408656,
+      "grad_norm": 0.1694600135087967,
+      "learning_rate": 1.2163215390068466e-05,
+      "loss": 0.47,
+      "step": 5849
+    },
+    {
+      "epoch": 1.3049297345527548,
+      "grad_norm": 0.1770717054605484,
+      "learning_rate": 1.2160917625054721e-05,
+      "loss": 0.4647,
+      "step": 5850
+    },
+    {
+      "epoch": 1.3051527994646441,
+      "grad_norm": 0.15664999186992645,
+      "learning_rate": 1.2158619740355767e-05,
+      "loss": 0.4371,
+      "step": 5851
+    },
+    {
+      "epoch": 1.3053758643765336,
+      "grad_norm": 0.16437092423439026,
+      "learning_rate": 1.2156321736098877e-05,
+      "loss": 0.436,
+      "step": 5852
+    },
+    {
+      "epoch": 1.305598929288423,
+      "grad_norm": 0.17147205770015717,
+      "learning_rate": 1.2154023612411321e-05,
+      "loss": 0.4696,
+      "step": 5853
+    },
+    {
+      "epoch": 1.3058219942003122,
+      "grad_norm": 0.1643034666776657,
+      "learning_rate": 1.215172536942039e-05,
+      "loss": 0.472,
+      "step": 5854
+    },
+    {
+      "epoch": 1.3060450591122017,
+      "grad_norm": 0.1705443561077118,
+      "learning_rate": 1.2149427007253372e-05,
+      "loss": 0.4422,
+      "step": 5855
+    },
+    {
+      "epoch": 1.306268124024091,
+      "grad_norm": 0.1692255437374115,
+      "learning_rate": 1.2147128526037568e-05,
+      "loss": 0.4583,
+      "step": 5856
+    },
+    {
+      "epoch": 1.3064911889359805,
+      "grad_norm": 0.17892007529735565,
+      "learning_rate": 1.2144829925900278e-05,
+      "loss": 0.4834,
+      "step": 5857
+    },
+    {
+      "epoch": 1.3067142538478698,
+      "grad_norm": 0.1718159317970276,
+      "learning_rate": 1.2142531206968815e-05,
+      "loss": 0.4576,
+      "step": 5858
+    },
+    {
+      "epoch": 1.306937318759759,
+      "grad_norm": 0.16335871815681458,
+      "learning_rate": 1.21402323693705e-05,
+      "loss": 0.4709,
+      "step": 5859
+    },
+    {
+      "epoch": 1.3071603836716483,
+      "grad_norm": 0.1720331460237503,
+      "learning_rate": 1.2137933413232651e-05,
+      "loss": 0.4536,
+      "step": 5860
+    },
+    {
+      "epoch": 1.3073834485835378,
+      "grad_norm": 0.17654761672019958,
+      "learning_rate": 1.2135634338682605e-05,
+      "loss": 0.4706,
+      "step": 5861
+    },
+    {
+      "epoch": 1.3076065134954271,
+      "grad_norm": 0.16412043571472168,
+      "learning_rate": 1.2133335145847691e-05,
+      "loss": 0.4914,
+      "step": 5862
+    },
+    {
+      "epoch": 1.3078295784073166,
+      "grad_norm": 0.17055629193782806,
+      "learning_rate": 1.213103583485526e-05,
+      "loss": 0.4666,
+      "step": 5863
+    },
+    {
+      "epoch": 1.308052643319206,
+      "grad_norm": 0.174026221036911,
+      "learning_rate": 1.2128736405832657e-05,
+      "loss": 0.46,
+      "step": 5864
+    },
+    {
+      "epoch": 1.3082757082310952,
+      "grad_norm": 0.17043223977088928,
+      "learning_rate": 1.2126436858907244e-05,
+      "loss": 0.4658,
+      "step": 5865
+    },
+    {
+      "epoch": 1.3084987731429847,
+      "grad_norm": 0.1644662320613861,
+      "learning_rate": 1.212413719420638e-05,
+      "loss": 0.4527,
+      "step": 5866
+    },
+    {
+      "epoch": 1.308721838054874,
+      "grad_norm": 0.1671665757894516,
+      "learning_rate": 1.212183741185744e-05,
+      "loss": 0.4617,
+      "step": 5867
+    },
+    {
+      "epoch": 1.3089449029667632,
+      "grad_norm": 0.2064938247203827,
+      "learning_rate": 1.2119537511987794e-05,
+      "loss": 0.4554,
+      "step": 5868
+    },
+    {
+      "epoch": 1.3091679678786527,
+      "grad_norm": 0.1725645214319229,
+      "learning_rate": 1.211723749472483e-05,
+      "loss": 0.4534,
+      "step": 5869
+    },
+    {
+      "epoch": 1.309391032790542,
+      "grad_norm": 0.16695371270179749,
+      "learning_rate": 1.2114937360195935e-05,
+      "loss": 0.4517,
+      "step": 5870
+    },
+    {
+      "epoch": 1.3096140977024313,
+      "grad_norm": 0.18011373281478882,
+      "learning_rate": 1.2112637108528505e-05,
+      "loss": 0.4576,
+      "step": 5871
+    },
+    {
+      "epoch": 1.3098371626143208,
+      "grad_norm": 0.17157688736915588,
+      "learning_rate": 1.2110336739849944e-05,
+      "loss": 0.438,
+      "step": 5872
+    },
+    {
+      "epoch": 1.31006022752621,
+      "grad_norm": 0.17350567877292633,
+      "learning_rate": 1.2108036254287658e-05,
+      "loss": 0.4647,
+      "step": 5873
+    },
+    {
+      "epoch": 1.3102832924380996,
+      "grad_norm": 0.16434237360954285,
+      "learning_rate": 1.2105735651969066e-05,
+      "loss": 0.4603,
+      "step": 5874
+    },
+    {
+      "epoch": 1.3105063573499889,
+      "grad_norm": 0.16745516657829285,
+      "learning_rate": 1.2103434933021587e-05,
+      "loss": 0.4585,
+      "step": 5875
+    },
+    {
+      "epoch": 1.3107294222618782,
+      "grad_norm": 0.16241510212421417,
+      "learning_rate": 1.2101134097572654e-05,
+      "loss": 0.4415,
+      "step": 5876
+    },
+    {
+      "epoch": 1.3109524871737674,
+      "grad_norm": 0.16632577776908875,
+      "learning_rate": 1.2098833145749691e-05,
+      "loss": 0.4715,
+      "step": 5877
+    },
+    {
+      "epoch": 1.311175552085657,
+      "grad_norm": 0.17647215723991394,
+      "learning_rate": 1.2096532077680153e-05,
+      "loss": 0.491,
+      "step": 5878
+    },
+    {
+      "epoch": 1.3113986169975462,
+      "grad_norm": 0.1738002449274063,
+      "learning_rate": 1.2094230893491475e-05,
+      "loss": 0.4731,
+      "step": 5879
+    },
+    {
+      "epoch": 1.3116216819094357,
+      "grad_norm": 0.16479554772377014,
+      "learning_rate": 1.2091929593311122e-05,
+      "loss": 0.4787,
+      "step": 5880
+    },
+    {
+      "epoch": 1.311844746821325,
+      "grad_norm": 0.16516366600990295,
+      "learning_rate": 1.2089628177266545e-05,
+      "loss": 0.4551,
+      "step": 5881
+    },
+    {
+      "epoch": 1.3120678117332143,
+      "grad_norm": 0.1664283722639084,
+      "learning_rate": 1.2087326645485218e-05,
+      "loss": 0.4516,
+      "step": 5882
+    },
+    {
+      "epoch": 1.3122908766451038,
+      "grad_norm": 0.23761501908302307,
+      "learning_rate": 1.208502499809461e-05,
+      "loss": 0.4679,
+      "step": 5883
+    },
+    {
+      "epoch": 1.312513941556993,
+      "grad_norm": 0.17012745141983032,
+      "learning_rate": 1.2082723235222205e-05,
+      "loss": 0.4618,
+      "step": 5884
+    },
+    {
+      "epoch": 1.3127370064688826,
+      "grad_norm": 0.1753227859735489,
+      "learning_rate": 1.2080421356995484e-05,
+      "loss": 0.4671,
+      "step": 5885
+    },
+    {
+      "epoch": 1.3129600713807719,
+      "grad_norm": 0.17254190146923065,
+      "learning_rate": 1.2078119363541942e-05,
+      "loss": 0.4379,
+      "step": 5886
+    },
+    {
+      "epoch": 1.3131831362926611,
+      "grad_norm": 0.1624683141708374,
+      "learning_rate": 1.2075817254989078e-05,
+      "loss": 0.4361,
+      "step": 5887
+    },
+    {
+      "epoch": 1.3134062012045504,
+      "grad_norm": 0.16935352981090546,
+      "learning_rate": 1.2073515031464397e-05,
+      "loss": 0.4528,
+      "step": 5888
+    },
+    {
+      "epoch": 1.31362926611644,
+      "grad_norm": 0.1754865050315857,
+      "learning_rate": 1.207121269309541e-05,
+      "loss": 0.4503,
+      "step": 5889
+    },
+    {
+      "epoch": 1.3138523310283292,
+      "grad_norm": 0.1585618406534195,
+      "learning_rate": 1.2068910240009636e-05,
+      "loss": 0.4401,
+      "step": 5890
+    },
+    {
+      "epoch": 1.3140753959402187,
+      "grad_norm": 0.18017402291297913,
+      "learning_rate": 1.20666076723346e-05,
+      "loss": 0.4595,
+      "step": 5891
+    },
+    {
+      "epoch": 1.314298460852108,
+      "grad_norm": 0.1608768105506897,
+      "learning_rate": 1.206430499019783e-05,
+      "loss": 0.4414,
+      "step": 5892
+    },
+    {
+      "epoch": 1.3145215257639973,
+      "grad_norm": 0.1631273776292801,
+      "learning_rate": 1.2062002193726867e-05,
+      "loss": 0.4567,
+      "step": 5893
+    },
+    {
+      "epoch": 1.3147445906758866,
+      "grad_norm": 0.16995559632778168,
+      "learning_rate": 1.2059699283049249e-05,
+      "loss": 0.4611,
+      "step": 5894
+    },
+    {
+      "epoch": 1.314967655587776,
+      "grad_norm": 0.1610070914030075,
+      "learning_rate": 1.2057396258292533e-05,
+      "loss": 0.4271,
+      "step": 5895
+    },
+    {
+      "epoch": 1.3151907204996653,
+      "grad_norm": 0.16440489888191223,
+      "learning_rate": 1.2055093119584264e-05,
+      "loss": 0.4531,
+      "step": 5896
+    },
+    {
+      "epoch": 1.3154137854115548,
+      "grad_norm": 0.16929922997951508,
+      "learning_rate": 1.2052789867052018e-05,
+      "loss": 0.4426,
+      "step": 5897
+    },
+    {
+      "epoch": 1.3156368503234441,
+      "grad_norm": 0.17286795377731323,
+      "learning_rate": 1.2050486500823352e-05,
+      "loss": 0.4584,
+      "step": 5898
+    },
+    {
+      "epoch": 1.3158599152353334,
+      "grad_norm": 0.1717979460954666,
+      "learning_rate": 1.2048183021025847e-05,
+      "loss": 0.461,
+      "step": 5899
+    },
+    {
+      "epoch": 1.316082980147223,
+      "grad_norm": 0.15612578392028809,
+      "learning_rate": 1.2045879427787084e-05,
+      "loss": 0.4466,
+      "step": 5900
+    },
+    {
+      "epoch": 1.3163060450591122,
+      "grad_norm": 0.17022433876991272,
+      "learning_rate": 1.2043575721234649e-05,
+      "loss": 0.4376,
+      "step": 5901
+    },
+    {
+      "epoch": 1.3165291099710017,
+      "grad_norm": 0.16973432898521423,
+      "learning_rate": 1.2041271901496136e-05,
+      "loss": 0.4606,
+      "step": 5902
+    },
+    {
+      "epoch": 1.316752174882891,
+      "grad_norm": 0.1685478538274765,
+      "learning_rate": 1.2038967968699143e-05,
+      "loss": 0.4621,
+      "step": 5903
+    },
+    {
+      "epoch": 1.3169752397947803,
+      "grad_norm": 0.16581124067306519,
+      "learning_rate": 1.2036663922971279e-05,
+      "loss": 0.4652,
+      "step": 5904
+    },
+    {
+      "epoch": 1.3171983047066695,
+      "grad_norm": 0.17322222888469696,
+      "learning_rate": 1.2034359764440156e-05,
+      "loss": 0.5025,
+      "step": 5905
+    },
+    {
+      "epoch": 1.317421369618559,
+      "grad_norm": 0.16943678259849548,
+      "learning_rate": 1.2032055493233394e-05,
+      "loss": 0.4508,
+      "step": 5906
+    },
+    {
+      "epoch": 1.3176444345304483,
+      "grad_norm": 0.17897944152355194,
+      "learning_rate": 1.2029751109478614e-05,
+      "loss": 0.4623,
+      "step": 5907
+    },
+    {
+      "epoch": 1.3178674994423378,
+      "grad_norm": 0.1740075647830963,
+      "learning_rate": 1.2027446613303454e-05,
+      "loss": 0.4835,
+      "step": 5908
+    },
+    {
+      "epoch": 1.3180905643542271,
+      "grad_norm": 0.173087015748024,
+      "learning_rate": 1.2025142004835541e-05,
+      "loss": 0.4713,
+      "step": 5909
+    },
+    {
+      "epoch": 1.3183136292661164,
+      "grad_norm": 0.18074947595596313,
+      "learning_rate": 1.2022837284202531e-05,
+      "loss": 0.4621,
+      "step": 5910
+    },
+    {
+      "epoch": 1.3185366941780057,
+      "grad_norm": 0.16880744695663452,
+      "learning_rate": 1.2020532451532063e-05,
+      "loss": 0.4486,
+      "step": 5911
+    },
+    {
+      "epoch": 1.3187597590898952,
+      "grad_norm": 0.16226506233215332,
+      "learning_rate": 1.2018227506951802e-05,
+      "loss": 0.4502,
+      "step": 5912
+    },
+    {
+      "epoch": 1.3189828240017845,
+      "grad_norm": 0.16763652861118317,
+      "learning_rate": 1.2015922450589405e-05,
+      "loss": 0.4588,
+      "step": 5913
+    },
+    {
+      "epoch": 1.319205888913674,
+      "grad_norm": 0.1670703887939453,
+      "learning_rate": 1.2013617282572545e-05,
+      "loss": 0.4398,
+      "step": 5914
+    },
+    {
+      "epoch": 1.3194289538255632,
+      "grad_norm": 0.18015524744987488,
+      "learning_rate": 1.201131200302889e-05,
+      "loss": 0.4625,
+      "step": 5915
+    },
+    {
+      "epoch": 1.3196520187374525,
+      "grad_norm": 0.16982007026672363,
+      "learning_rate": 1.2009006612086128e-05,
+      "loss": 0.4619,
+      "step": 5916
+    },
+    {
+      "epoch": 1.319875083649342,
+      "grad_norm": 0.16795918345451355,
+      "learning_rate": 1.2006701109871936e-05,
+      "loss": 0.4577,
+      "step": 5917
+    },
+    {
+      "epoch": 1.3200981485612313,
+      "grad_norm": 0.17438872158527374,
+      "learning_rate": 1.2004395496514021e-05,
+      "loss": 0.4668,
+      "step": 5918
+    },
+    {
+      "epoch": 1.3203212134731208,
+      "grad_norm": 0.17492419481277466,
+      "learning_rate": 1.2002089772140071e-05,
+      "loss": 0.4592,
+      "step": 5919
+    },
+    {
+      "epoch": 1.32054427838501,
+      "grad_norm": 0.17449866235256195,
+      "learning_rate": 1.19997839368778e-05,
+      "loss": 0.4632,
+      "step": 5920
+    },
+    {
+      "epoch": 1.3207673432968994,
+      "grad_norm": 0.16483739018440247,
+      "learning_rate": 1.199747799085491e-05,
+      "loss": 0.4496,
+      "step": 5921
+    },
+    {
+      "epoch": 1.3209904082087887,
+      "grad_norm": 0.19091255962848663,
+      "learning_rate": 1.1995171934199128e-05,
+      "loss": 0.4647,
+      "step": 5922
+    },
+    {
+      "epoch": 1.3212134731206782,
+      "grad_norm": 0.17579780519008636,
+      "learning_rate": 1.1992865767038172e-05,
+      "loss": 0.4545,
+      "step": 5923
+    },
+    {
+      "epoch": 1.3214365380325674,
+      "grad_norm": 0.17818951606750488,
+      "learning_rate": 1.1990559489499776e-05,
+      "loss": 0.4806,
+      "step": 5924
+    },
+    {
+      "epoch": 1.321659602944457,
+      "grad_norm": 0.17060208320617676,
+      "learning_rate": 1.1988253101711675e-05,
+      "loss": 0.4435,
+      "step": 5925
+    },
+    {
+      "epoch": 1.3218826678563462,
+      "grad_norm": 0.17483805119991302,
+      "learning_rate": 1.1985946603801608e-05,
+      "loss": 0.4901,
+      "step": 5926
+    },
+    {
+      "epoch": 1.3221057327682355,
+      "grad_norm": 0.17470861971378326,
+      "learning_rate": 1.198363999589733e-05,
+      "loss": 0.462,
+      "step": 5927
+    },
+    {
+      "epoch": 1.3223287976801248,
+      "grad_norm": 0.16998319327831268,
+      "learning_rate": 1.1981333278126585e-05,
+      "loss": 0.4469,
+      "step": 5928
+    },
+    {
+      "epoch": 1.3225518625920143,
+      "grad_norm": 0.1669076532125473,
+      "learning_rate": 1.1979026450617147e-05,
+      "loss": 0.4422,
+      "step": 5929
+    },
+    {
+      "epoch": 1.3227749275039036,
+      "grad_norm": 0.17433854937553406,
+      "learning_rate": 1.197671951349677e-05,
+      "loss": 0.4572,
+      "step": 5930
+    },
+    {
+      "epoch": 1.322997992415793,
+      "grad_norm": 0.16878628730773926,
+      "learning_rate": 1.1974412466893237e-05,
+      "loss": 0.4354,
+      "step": 5931
+    },
+    {
+      "epoch": 1.3232210573276824,
+      "grad_norm": 0.17208078503608704,
+      "learning_rate": 1.1972105310934318e-05,
+      "loss": 0.4149,
+      "step": 5932
+    },
+    {
+      "epoch": 1.3234441222395716,
+      "grad_norm": 0.1662035584449768,
+      "learning_rate": 1.1969798045747805e-05,
+      "loss": 0.4224,
+      "step": 5933
+    },
+    {
+      "epoch": 1.3236671871514611,
+      "grad_norm": 0.16633039712905884,
+      "learning_rate": 1.1967490671461484e-05,
+      "loss": 0.4621,
+      "step": 5934
+    },
+    {
+      "epoch": 1.3238902520633504,
+      "grad_norm": 0.17780262231826782,
+      "learning_rate": 1.1965183188203154e-05,
+      "loss": 0.4969,
+      "step": 5935
+    },
+    {
+      "epoch": 1.32411331697524,
+      "grad_norm": 0.16385456919670105,
+      "learning_rate": 1.1962875596100616e-05,
+      "loss": 0.4476,
+      "step": 5936
+    },
+    {
+      "epoch": 1.3243363818871292,
+      "grad_norm": 0.17453844845294952,
+      "learning_rate": 1.1960567895281682e-05,
+      "loss": 0.4784,
+      "step": 5937
+    },
+    {
+      "epoch": 1.3245594467990185,
+      "grad_norm": 0.17324872314929962,
+      "learning_rate": 1.1958260085874165e-05,
+      "loss": 0.4734,
+      "step": 5938
+    },
+    {
+      "epoch": 1.3247825117109078,
+      "grad_norm": 0.1686837375164032,
+      "learning_rate": 1.1955952168005889e-05,
+      "loss": 0.4497,
+      "step": 5939
+    },
+    {
+      "epoch": 1.3250055766227973,
+      "grad_norm": 0.17261843383312225,
+      "learning_rate": 1.1953644141804675e-05,
+      "loss": 0.4805,
+      "step": 5940
+    },
+    {
+      "epoch": 1.3252286415346866,
+      "grad_norm": 0.16724753379821777,
+      "learning_rate": 1.1951336007398362e-05,
+      "loss": 0.4568,
+      "step": 5941
+    },
+    {
+      "epoch": 1.325451706446576,
+      "grad_norm": 0.17275574803352356,
+      "learning_rate": 1.1949027764914786e-05,
+      "loss": 0.4567,
+      "step": 5942
+    },
+    {
+      "epoch": 1.3256747713584653,
+      "grad_norm": 0.16790451109409332,
+      "learning_rate": 1.194671941448179e-05,
+      "loss": 0.4821,
+      "step": 5943
+    },
+    {
+      "epoch": 1.3258978362703546,
+      "grad_norm": 0.16740332543849945,
+      "learning_rate": 1.1944410956227233e-05,
+      "loss": 0.4712,
+      "step": 5944
+    },
+    {
+      "epoch": 1.326120901182244,
+      "grad_norm": 0.185842826962471,
+      "learning_rate": 1.1942102390278961e-05,
+      "loss": 0.4639,
+      "step": 5945
+    },
+    {
+      "epoch": 1.3263439660941334,
+      "grad_norm": 0.1872578263282776,
+      "learning_rate": 1.1939793716764845e-05,
+      "loss": 0.4601,
+      "step": 5946
+    },
+    {
+      "epoch": 1.3265670310060227,
+      "grad_norm": 0.17680978775024414,
+      "learning_rate": 1.1937484935812749e-05,
+      "loss": 0.4536,
+      "step": 5947
+    },
+    {
+      "epoch": 1.3267900959179122,
+      "grad_norm": 0.1672886312007904,
+      "learning_rate": 1.1935176047550552e-05,
+      "loss": 0.4573,
+      "step": 5948
+    },
+    {
+      "epoch": 1.3270131608298015,
+      "grad_norm": 0.16590574383735657,
+      "learning_rate": 1.1932867052106132e-05,
+      "loss": 0.4702,
+      "step": 5949
+    },
+    {
+      "epoch": 1.3272362257416908,
+      "grad_norm": 0.15358559787273407,
+      "learning_rate": 1.1930557949607378e-05,
+      "loss": 0.4202,
+      "step": 5950
+    },
+    {
+      "epoch": 1.3274592906535803,
+      "grad_norm": 0.16293473541736603,
+      "learning_rate": 1.1928248740182177e-05,
+      "loss": 0.4528,
+      "step": 5951
+    },
+    {
+      "epoch": 1.3276823555654695,
+      "grad_norm": 0.16601158678531647,
+      "learning_rate": 1.1925939423958437e-05,
+      "loss": 0.4641,
+      "step": 5952
+    },
+    {
+      "epoch": 1.327905420477359,
+      "grad_norm": 0.16637404263019562,
+      "learning_rate": 1.1923630001064052e-05,
+      "loss": 0.4606,
+      "step": 5953
+    },
+    {
+      "epoch": 1.3281284853892483,
+      "grad_norm": 0.16997739672660828,
+      "learning_rate": 1.1921320471626939e-05,
+      "loss": 0.4688,
+      "step": 5954
+    },
+    {
+      "epoch": 1.3283515503011376,
+      "grad_norm": 0.1634935885667801,
+      "learning_rate": 1.1919010835775015e-05,
+      "loss": 0.4367,
+      "step": 5955
+    },
+    {
+      "epoch": 1.3285746152130269,
+      "grad_norm": 0.16118724644184113,
+      "learning_rate": 1.1916701093636196e-05,
+      "loss": 0.4367,
+      "step": 5956
+    },
+    {
+      "epoch": 1.3287976801249164,
+      "grad_norm": 0.16823935508728027,
+      "learning_rate": 1.1914391245338417e-05,
+      "loss": 0.4566,
+      "step": 5957
+    },
+    {
+      "epoch": 1.3290207450368057,
+      "grad_norm": 0.16638781130313873,
+      "learning_rate": 1.1912081291009608e-05,
+      "loss": 0.4533,
+      "step": 5958
+    },
+    {
+      "epoch": 1.3292438099486952,
+      "grad_norm": 0.20548874139785767,
+      "learning_rate": 1.1909771230777709e-05,
+      "loss": 0.4482,
+      "step": 5959
+    },
+    {
+      "epoch": 1.3294668748605845,
+      "grad_norm": 0.16751278936862946,
+      "learning_rate": 1.1907461064770667e-05,
+      "loss": 0.4552,
+      "step": 5960
+    },
+    {
+      "epoch": 1.3296899397724737,
+      "grad_norm": 0.17895007133483887,
+      "learning_rate": 1.1905150793116433e-05,
+      "loss": 0.4638,
+      "step": 5961
+    },
+    {
+      "epoch": 1.329913004684363,
+      "grad_norm": 0.17109131813049316,
+      "learning_rate": 1.1902840415942964e-05,
+      "loss": 0.4386,
+      "step": 5962
+    },
+    {
+      "epoch": 1.3301360695962525,
+      "grad_norm": 0.1729149967432022,
+      "learning_rate": 1.1900529933378224e-05,
+      "loss": 0.4626,
+      "step": 5963
+    },
+    {
+      "epoch": 1.3303591345081418,
+      "grad_norm": 0.16538968682289124,
+      "learning_rate": 1.189821934555018e-05,
+      "loss": 0.4514,
+      "step": 5964
+    },
+    {
+      "epoch": 1.3305821994200313,
+      "grad_norm": 0.1783122420310974,
+      "learning_rate": 1.189590865258681e-05,
+      "loss": 0.4652,
+      "step": 5965
+    },
+    {
+      "epoch": 1.3308052643319206,
+      "grad_norm": 0.15844860672950745,
+      "learning_rate": 1.1893597854616092e-05,
+      "loss": 0.4402,
+      "step": 5966
+    },
+    {
+      "epoch": 1.3310283292438099,
+      "grad_norm": 0.1638982594013214,
+      "learning_rate": 1.1891286951766014e-05,
+      "loss": 0.4608,
+      "step": 5967
+    },
+    {
+      "epoch": 1.3312513941556994,
+      "grad_norm": 0.1644224226474762,
+      "learning_rate": 1.1888975944164567e-05,
+      "loss": 0.4526,
+      "step": 5968
+    },
+    {
+      "epoch": 1.3314744590675887,
+      "grad_norm": 0.20200717449188232,
+      "learning_rate": 1.1886664831939751e-05,
+      "loss": 0.4611,
+      "step": 5969
+    },
+    {
+      "epoch": 1.3316975239794782,
+      "grad_norm": 0.1918911337852478,
+      "learning_rate": 1.188435361521957e-05,
+      "loss": 0.4704,
+      "step": 5970
+    },
+    {
+      "epoch": 1.3319205888913674,
+      "grad_norm": 0.16691647469997406,
+      "learning_rate": 1.1882042294132032e-05,
+      "loss": 0.4787,
+      "step": 5971
+    },
+    {
+      "epoch": 1.3321436538032567,
+      "grad_norm": 0.16211196780204773,
+      "learning_rate": 1.1879730868805153e-05,
+      "loss": 0.4535,
+      "step": 5972
+    },
+    {
+      "epoch": 1.332366718715146,
+      "grad_norm": 0.16488417983055115,
+      "learning_rate": 1.1877419339366953e-05,
+      "loss": 0.4461,
+      "step": 5973
+    },
+    {
+      "epoch": 1.3325897836270355,
+      "grad_norm": 0.16110685467720032,
+      "learning_rate": 1.1875107705945461e-05,
+      "loss": 0.444,
+      "step": 5974
+    },
+    {
+      "epoch": 1.3328128485389248,
+      "grad_norm": 0.16213123500347137,
+      "learning_rate": 1.187279596866871e-05,
+      "loss": 0.4324,
+      "step": 5975
+    },
+    {
+      "epoch": 1.3330359134508143,
+      "grad_norm": 0.17590108513832092,
+      "learning_rate": 1.1870484127664739e-05,
+      "loss": 0.4416,
+      "step": 5976
+    },
+    {
+      "epoch": 1.3332589783627036,
+      "grad_norm": 0.20271827280521393,
+      "learning_rate": 1.186817218306159e-05,
+      "loss": 0.4514,
+      "step": 5977
+    },
+    {
+      "epoch": 1.3334820432745929,
+      "grad_norm": 0.17544665932655334,
+      "learning_rate": 1.1865860134987317e-05,
+      "loss": 0.4788,
+      "step": 5978
+    },
+    {
+      "epoch": 1.3337051081864821,
+      "grad_norm": 0.2024955302476883,
+      "learning_rate": 1.1863547983569967e-05,
+      "loss": 0.4804,
+      "step": 5979
+    },
+    {
+      "epoch": 1.3339281730983716,
+      "grad_norm": 0.17373384535312653,
+      "learning_rate": 1.1861235728937613e-05,
+      "loss": 0.4485,
+      "step": 5980
+    },
+    {
+      "epoch": 1.334151238010261,
+      "grad_norm": 0.17016762495040894,
+      "learning_rate": 1.1858923371218314e-05,
+      "loss": 0.4399,
+      "step": 5981
+    },
+    {
+      "epoch": 1.3343743029221504,
+      "grad_norm": 0.167991504073143,
+      "learning_rate": 1.1856610910540149e-05,
+      "loss": 0.4612,
+      "step": 5982
+    },
+    {
+      "epoch": 1.3345973678340397,
+      "grad_norm": 0.1790834218263626,
+      "learning_rate": 1.185429834703119e-05,
+      "loss": 0.456,
+      "step": 5983
+    },
+    {
+      "epoch": 1.334820432745929,
+      "grad_norm": 0.16934290528297424,
+      "learning_rate": 1.1851985680819526e-05,
+      "loss": 0.4717,
+      "step": 5984
+    },
+    {
+      "epoch": 1.3350434976578185,
+      "grad_norm": 0.1672845035791397,
+      "learning_rate": 1.1849672912033245e-05,
+      "loss": 0.4748,
+      "step": 5985
+    },
+    {
+      "epoch": 1.3352665625697078,
+      "grad_norm": 0.16489911079406738,
+      "learning_rate": 1.1847360040800442e-05,
+      "loss": 0.416,
+      "step": 5986
+    },
+    {
+      "epoch": 1.3354896274815973,
+      "grad_norm": 0.18953464925289154,
+      "learning_rate": 1.184504706724922e-05,
+      "loss": 0.4601,
+      "step": 5987
+    },
+    {
+      "epoch": 1.3357126923934866,
+      "grad_norm": 0.16453106701374054,
+      "learning_rate": 1.1842733991507687e-05,
+      "loss": 0.4417,
+      "step": 5988
+    },
+    {
+      "epoch": 1.3359357573053758,
+      "grad_norm": 0.1774827390909195,
+      "learning_rate": 1.1840420813703955e-05,
+      "loss": 0.4459,
+      "step": 5989
+    },
+    {
+      "epoch": 1.3361588222172651,
+      "grad_norm": 0.1909363716840744,
+      "learning_rate": 1.183810753396614e-05,
+      "loss": 0.4695,
+      "step": 5990
+    },
+    {
+      "epoch": 1.3363818871291546,
+      "grad_norm": 0.18591512739658356,
+      "learning_rate": 1.1835794152422366e-05,
+      "loss": 0.4579,
+      "step": 5991
+    },
+    {
+      "epoch": 1.336604952041044,
+      "grad_norm": 0.1766308695077896,
+      "learning_rate": 1.1833480669200765e-05,
+      "loss": 0.4595,
+      "step": 5992
+    },
+    {
+      "epoch": 1.3368280169529334,
+      "grad_norm": 0.16689802706241608,
+      "learning_rate": 1.1831167084429474e-05,
+      "loss": 0.4371,
+      "step": 5993
+    },
+    {
+      "epoch": 1.3370510818648227,
+      "grad_norm": 0.180791437625885,
+      "learning_rate": 1.1828853398236629e-05,
+      "loss": 0.4773,
+      "step": 5994
+    },
+    {
+      "epoch": 1.337274146776712,
+      "grad_norm": 0.18029746413230896,
+      "learning_rate": 1.182653961075038e-05,
+      "loss": 0.4281,
+      "step": 5995
+    },
+    {
+      "epoch": 1.3374972116886013,
+      "grad_norm": 0.16131596267223358,
+      "learning_rate": 1.1824225722098877e-05,
+      "loss": 0.4479,
+      "step": 5996
+    },
+    {
+      "epoch": 1.3377202766004908,
+      "grad_norm": 0.16809871792793274,
+      "learning_rate": 1.182191173241028e-05,
+      "loss": 0.4473,
+      "step": 5997
+    },
+    {
+      "epoch": 1.33794334151238,
+      "grad_norm": 0.17205321788787842,
+      "learning_rate": 1.181959764181275e-05,
+      "loss": 0.4357,
+      "step": 5998
+    },
+    {
+      "epoch": 1.3381664064242695,
+      "grad_norm": 0.17919965088367462,
+      "learning_rate": 1.1817283450434459e-05,
+      "loss": 0.4892,
+      "step": 5999
+    },
+    {
+      "epoch": 1.3383894713361588,
+      "grad_norm": 0.16656439006328583,
+      "learning_rate": 1.1814969158403575e-05,
+      "loss": 0.4686,
+      "step": 6000
+    },
+    {
+      "epoch": 1.338612536248048,
+      "grad_norm": 0.16457730531692505,
+      "learning_rate": 1.1812654765848285e-05,
+      "loss": 0.4346,
+      "step": 6001
+    },
+    {
+      "epoch": 1.3388356011599376,
+      "grad_norm": 0.17078131437301636,
+      "learning_rate": 1.1810340272896772e-05,
+      "loss": 0.4567,
+      "step": 6002
+    },
+    {
+      "epoch": 1.3390586660718269,
+      "grad_norm": 0.16368982195854187,
+      "learning_rate": 1.1808025679677229e-05,
+      "loss": 0.451,
+      "step": 6003
+    },
+    {
+      "epoch": 1.3392817309837164,
+      "grad_norm": 0.16397084295749664,
+      "learning_rate": 1.1805710986317846e-05,
+      "loss": 0.4541,
+      "step": 6004
+    },
+    {
+      "epoch": 1.3395047958956057,
+      "grad_norm": 0.16752174496650696,
+      "learning_rate": 1.1803396192946835e-05,
+      "loss": 0.4714,
+      "step": 6005
+    },
+    {
+      "epoch": 1.339727860807495,
+      "grad_norm": 0.17410169541835785,
+      "learning_rate": 1.1801081299692396e-05,
+      "loss": 0.4782,
+      "step": 6006
+    },
+    {
+      "epoch": 1.3399509257193842,
+      "grad_norm": 0.16793891787528992,
+      "learning_rate": 1.1798766306682746e-05,
+      "loss": 0.4733,
+      "step": 6007
+    },
+    {
+      "epoch": 1.3401739906312737,
+      "grad_norm": 0.17633196711540222,
+      "learning_rate": 1.1796451214046106e-05,
+      "loss": 0.4668,
+      "step": 6008
+    },
+    {
+      "epoch": 1.340397055543163,
+      "grad_norm": 0.17554394900798798,
+      "learning_rate": 1.1794136021910694e-05,
+      "loss": 0.4375,
+      "step": 6009
+    },
+    {
+      "epoch": 1.3406201204550525,
+      "grad_norm": 0.1725650429725647,
+      "learning_rate": 1.1791820730404746e-05,
+      "loss": 0.46,
+      "step": 6010
+    },
+    {
+      "epoch": 1.3408431853669418,
+      "grad_norm": 0.17213909327983856,
+      "learning_rate": 1.1789505339656493e-05,
+      "loss": 0.463,
+      "step": 6011
+    },
+    {
+      "epoch": 1.341066250278831,
+      "grad_norm": 0.17620781064033508,
+      "learning_rate": 1.1787189849794178e-05,
+      "loss": 0.4817,
+      "step": 6012
+    },
+    {
+      "epoch": 1.3412893151907204,
+      "grad_norm": 0.1637295037508011,
+      "learning_rate": 1.1784874260946048e-05,
+      "loss": 0.4883,
+      "step": 6013
+    },
+    {
+      "epoch": 1.3415123801026099,
+      "grad_norm": 0.17048433423042297,
+      "learning_rate": 1.1782558573240355e-05,
+      "loss": 0.4333,
+      "step": 6014
+    },
+    {
+      "epoch": 1.3417354450144992,
+      "grad_norm": 0.17557942867279053,
+      "learning_rate": 1.1780242786805353e-05,
+      "loss": 0.4704,
+      "step": 6015
+    },
+    {
+      "epoch": 1.3419585099263887,
+      "grad_norm": 0.17155835032463074,
+      "learning_rate": 1.177792690176931e-05,
+      "loss": 0.4627,
+      "step": 6016
+    },
+    {
+      "epoch": 1.342181574838278,
+      "grad_norm": 0.1626436561346054,
+      "learning_rate": 1.177561091826049e-05,
+      "loss": 0.4703,
+      "step": 6017
+    },
+    {
+      "epoch": 1.3424046397501672,
+      "grad_norm": 0.18032924830913544,
+      "learning_rate": 1.177329483640717e-05,
+      "loss": 0.4727,
+      "step": 6018
+    },
+    {
+      "epoch": 1.3426277046620567,
+      "grad_norm": 0.18225564062595367,
+      "learning_rate": 1.1770978656337626e-05,
+      "loss": 0.4737,
+      "step": 6019
+    },
+    {
+      "epoch": 1.342850769573946,
+      "grad_norm": 0.22427327930927277,
+      "learning_rate": 1.1768662378180144e-05,
+      "loss": 0.4353,
+      "step": 6020
+    },
+    {
+      "epoch": 1.3430738344858355,
+      "grad_norm": 0.16713419556617737,
+      "learning_rate": 1.1766346002063017e-05,
+      "loss": 0.4674,
+      "step": 6021
+    },
+    {
+      "epoch": 1.3432968993977248,
+      "grad_norm": 0.17510871589183807,
+      "learning_rate": 1.1764029528114533e-05,
+      "loss": 0.4907,
+      "step": 6022
+    },
+    {
+      "epoch": 1.343519964309614,
+      "grad_norm": 0.20499172806739807,
+      "learning_rate": 1.1761712956463003e-05,
+      "loss": 0.4798,
+      "step": 6023
+    },
+    {
+      "epoch": 1.3437430292215033,
+      "grad_norm": 0.17217415571212769,
+      "learning_rate": 1.1759396287236721e-05,
+      "loss": 0.4642,
+      "step": 6024
+    },
+    {
+      "epoch": 1.3439660941333929,
+      "grad_norm": 0.17448726296424866,
+      "learning_rate": 1.1757079520564012e-05,
+      "loss": 0.4569,
+      "step": 6025
+    },
+    {
+      "epoch": 1.3441891590452821,
+      "grad_norm": 0.16537372767925262,
+      "learning_rate": 1.1754762656573182e-05,
+      "loss": 0.4348,
+      "step": 6026
+    },
+    {
+      "epoch": 1.3444122239571716,
+      "grad_norm": 0.16384288668632507,
+      "learning_rate": 1.1752445695392563e-05,
+      "loss": 0.444,
+      "step": 6027
+    },
+    {
+      "epoch": 1.344635288869061,
+      "grad_norm": 0.25420817732810974,
+      "learning_rate": 1.1750128637150473e-05,
+      "loss": 0.4489,
+      "step": 6028
+    },
+    {
+      "epoch": 1.3448583537809502,
+      "grad_norm": 0.17444314062595367,
+      "learning_rate": 1.1747811481975253e-05,
+      "loss": 0.4641,
+      "step": 6029
+    },
+    {
+      "epoch": 1.3450814186928395,
+      "grad_norm": 0.17278996109962463,
+      "learning_rate": 1.1745494229995237e-05,
+      "loss": 0.4622,
+      "step": 6030
+    },
+    {
+      "epoch": 1.345304483604729,
+      "grad_norm": 0.1730216145515442,
+      "learning_rate": 1.1743176881338773e-05,
+      "loss": 0.4413,
+      "step": 6031
+    },
+    {
+      "epoch": 1.3455275485166183,
+      "grad_norm": 0.1769610494375229,
+      "learning_rate": 1.174085943613421e-05,
+      "loss": 0.4739,
+      "step": 6032
+    },
+    {
+      "epoch": 1.3457506134285078,
+      "grad_norm": 0.17104336619377136,
+      "learning_rate": 1.1738541894509898e-05,
+      "loss": 0.4513,
+      "step": 6033
+    },
+    {
+      "epoch": 1.345973678340397,
+      "grad_norm": 0.1742696762084961,
+      "learning_rate": 1.17362242565942e-05,
+      "loss": 0.4549,
+      "step": 6034
+    },
+    {
+      "epoch": 1.3461967432522863,
+      "grad_norm": 0.16949966549873352,
+      "learning_rate": 1.1733906522515483e-05,
+      "loss": 0.4773,
+      "step": 6035
+    },
+    {
+      "epoch": 1.3464198081641758,
+      "grad_norm": 0.17838764190673828,
+      "learning_rate": 1.1731588692402114e-05,
+      "loss": 0.47,
+      "step": 6036
+    },
+    {
+      "epoch": 1.3466428730760651,
+      "grad_norm": 0.1655292510986328,
+      "learning_rate": 1.1729270766382474e-05,
+      "loss": 0.4577,
+      "step": 6037
+    },
+    {
+      "epoch": 1.3468659379879546,
+      "grad_norm": 0.1620989590883255,
+      "learning_rate": 1.172695274458494e-05,
+      "loss": 0.427,
+      "step": 6038
+    },
+    {
+      "epoch": 1.347089002899844,
+      "grad_norm": 0.15767668187618256,
+      "learning_rate": 1.1724634627137896e-05,
+      "loss": 0.4419,
+      "step": 6039
+    },
+    {
+      "epoch": 1.3473120678117332,
+      "grad_norm": 0.17484217882156372,
+      "learning_rate": 1.1722316414169743e-05,
+      "loss": 0.4221,
+      "step": 6040
+    },
+    {
+      "epoch": 1.3475351327236225,
+      "grad_norm": 0.16977369785308838,
+      "learning_rate": 1.1719998105808871e-05,
+      "loss": 0.4639,
+      "step": 6041
+    },
+    {
+      "epoch": 1.347758197635512,
+      "grad_norm": 0.16105391085147858,
+      "learning_rate": 1.1717679702183687e-05,
+      "loss": 0.4707,
+      "step": 6042
+    },
+    {
+      "epoch": 1.3479812625474012,
+      "grad_norm": 0.16509099304676056,
+      "learning_rate": 1.1715361203422595e-05,
+      "loss": 0.4378,
+      "step": 6043
+    },
+    {
+      "epoch": 1.3482043274592908,
+      "grad_norm": 0.18078747391700745,
+      "learning_rate": 1.1713042609654008e-05,
+      "loss": 0.4565,
+      "step": 6044
+    },
+    {
+      "epoch": 1.34842739237118,
+      "grad_norm": 0.17105768620967865,
+      "learning_rate": 1.1710723921006348e-05,
+      "loss": 0.4574,
+      "step": 6045
+    },
+    {
+      "epoch": 1.3486504572830693,
+      "grad_norm": 0.17831863462924957,
+      "learning_rate": 1.1708405137608036e-05,
+      "loss": 0.4789,
+      "step": 6046
+    },
+    {
+      "epoch": 1.3488735221949586,
+      "grad_norm": 0.17648382484912872,
+      "learning_rate": 1.1706086259587499e-05,
+      "loss": 0.4757,
+      "step": 6047
+    },
+    {
+      "epoch": 1.349096587106848,
+      "grad_norm": 0.1652226746082306,
+      "learning_rate": 1.1703767287073177e-05,
+      "loss": 0.4559,
+      "step": 6048
+    },
+    {
+      "epoch": 1.3493196520187374,
+      "grad_norm": 0.16818653047084808,
+      "learning_rate": 1.1701448220193503e-05,
+      "loss": 0.4739,
+      "step": 6049
+    },
+    {
+      "epoch": 1.3495427169306269,
+      "grad_norm": 0.17042328417301178,
+      "learning_rate": 1.1699129059076925e-05,
+      "loss": 0.4512,
+      "step": 6050
+    },
+    {
+      "epoch": 1.3497657818425162,
+      "grad_norm": 0.16678601503372192,
+      "learning_rate": 1.1696809803851891e-05,
+      "loss": 0.4454,
+      "step": 6051
+    },
+    {
+      "epoch": 1.3499888467544054,
+      "grad_norm": 0.1808800995349884,
+      "learning_rate": 1.1694490454646858e-05,
+      "loss": 0.4656,
+      "step": 6052
+    },
+    {
+      "epoch": 1.350211911666295,
+      "grad_norm": 0.17242035269737244,
+      "learning_rate": 1.1692171011590282e-05,
+      "loss": 0.4635,
+      "step": 6053
+    },
+    {
+      "epoch": 1.3504349765781842,
+      "grad_norm": 0.16750234365463257,
+      "learning_rate": 1.1689851474810636e-05,
+      "loss": 0.4627,
+      "step": 6054
+    },
+    {
+      "epoch": 1.3506580414900737,
+      "grad_norm": 0.1674799621105194,
+      "learning_rate": 1.1687531844436382e-05,
+      "loss": 0.4452,
+      "step": 6055
+    },
+    {
+      "epoch": 1.350881106401963,
+      "grad_norm": 0.17459098994731903,
+      "learning_rate": 1.1685212120596001e-05,
+      "loss": 0.467,
+      "step": 6056
+    },
+    {
+      "epoch": 1.3511041713138523,
+      "grad_norm": 0.17721140384674072,
+      "learning_rate": 1.168289230341797e-05,
+      "loss": 0.4788,
+      "step": 6057
+    },
+    {
+      "epoch": 1.3513272362257416,
+      "grad_norm": 0.167550191283226,
+      "learning_rate": 1.168057239303078e-05,
+      "loss": 0.4488,
+      "step": 6058
+    },
+    {
+      "epoch": 1.351550301137631,
+      "grad_norm": 0.17352843284606934,
+      "learning_rate": 1.1678252389562919e-05,
+      "loss": 0.4582,
+      "step": 6059
+    },
+    {
+      "epoch": 1.3517733660495204,
+      "grad_norm": 0.1822769194841385,
+      "learning_rate": 1.1675932293142882e-05,
+      "loss": 0.4597,
+      "step": 6060
+    },
+    {
+      "epoch": 1.3519964309614099,
+      "grad_norm": 0.16711196303367615,
+      "learning_rate": 1.1673612103899176e-05,
+      "loss": 0.4624,
+      "step": 6061
+    },
+    {
+      "epoch": 1.3522194958732991,
+      "grad_norm": 0.16815169155597687,
+      "learning_rate": 1.16712918219603e-05,
+      "loss": 0.4416,
+      "step": 6062
+    },
+    {
+      "epoch": 1.3524425607851884,
+      "grad_norm": 0.17060011625289917,
+      "learning_rate": 1.1668971447454775e-05,
+      "loss": 0.4472,
+      "step": 6063
+    },
+    {
+      "epoch": 1.3526656256970777,
+      "grad_norm": 0.1819947212934494,
+      "learning_rate": 1.1666650980511112e-05,
+      "loss": 0.4791,
+      "step": 6064
+    },
+    {
+      "epoch": 1.3528886906089672,
+      "grad_norm": 0.16703422367572784,
+      "learning_rate": 1.1664330421257835e-05,
+      "loss": 0.4596,
+      "step": 6065
+    },
+    {
+      "epoch": 1.3531117555208565,
+      "grad_norm": 0.16398312151432037,
+      "learning_rate": 1.1662009769823466e-05,
+      "loss": 0.46,
+      "step": 6066
+    },
+    {
+      "epoch": 1.353334820432746,
+      "grad_norm": 0.177035391330719,
+      "learning_rate": 1.1659689026336544e-05,
+      "loss": 0.4566,
+      "step": 6067
+    },
+    {
+      "epoch": 1.3535578853446353,
+      "grad_norm": 0.17809878289699554,
+      "learning_rate": 1.1657368190925602e-05,
+      "loss": 0.4687,
+      "step": 6068
+    },
+    {
+      "epoch": 1.3537809502565246,
+      "grad_norm": 0.17178352177143097,
+      "learning_rate": 1.1655047263719188e-05,
+      "loss": 0.4609,
+      "step": 6069
+    },
+    {
+      "epoch": 1.354004015168414,
+      "grad_norm": 0.16717985272407532,
+      "learning_rate": 1.1652726244845843e-05,
+      "loss": 0.4341,
+      "step": 6070
+    },
+    {
+      "epoch": 1.3542270800803033,
+      "grad_norm": 0.1680830717086792,
+      "learning_rate": 1.1650405134434122e-05,
+      "loss": 0.4286,
+      "step": 6071
+    },
+    {
+      "epoch": 1.3544501449921929,
+      "grad_norm": 0.1840554177761078,
+      "learning_rate": 1.1648083932612584e-05,
+      "loss": 0.4781,
+      "step": 6072
+    },
+    {
+      "epoch": 1.3546732099040821,
+      "grad_norm": 0.16866901516914368,
+      "learning_rate": 1.164576263950979e-05,
+      "loss": 0.4588,
+      "step": 6073
+    },
+    {
+      "epoch": 1.3548962748159714,
+      "grad_norm": 0.18350256979465485,
+      "learning_rate": 1.164344125525431e-05,
+      "loss": 0.4828,
+      "step": 6074
+    },
+    {
+      "epoch": 1.3551193397278607,
+      "grad_norm": 0.18490628898143768,
+      "learning_rate": 1.1641119779974717e-05,
+      "loss": 0.4127,
+      "step": 6075
+    },
+    {
+      "epoch": 1.3553424046397502,
+      "grad_norm": 0.1628909409046173,
+      "learning_rate": 1.1638798213799585e-05,
+      "loss": 0.4403,
+      "step": 6076
+    },
+    {
+      "epoch": 1.3555654695516395,
+      "grad_norm": 0.17124617099761963,
+      "learning_rate": 1.1636476556857502e-05,
+      "loss": 0.4638,
+      "step": 6077
+    },
+    {
+      "epoch": 1.355788534463529,
+      "grad_norm": 0.18702422082424164,
+      "learning_rate": 1.1634154809277052e-05,
+      "loss": 0.4675,
+      "step": 6078
+    },
+    {
+      "epoch": 1.3560115993754183,
+      "grad_norm": 0.174610435962677,
+      "learning_rate": 1.1631832971186827e-05,
+      "loss": 0.448,
+      "step": 6079
+    },
+    {
+      "epoch": 1.3562346642873075,
+      "grad_norm": 0.165745347738266,
+      "learning_rate": 1.162951104271543e-05,
+      "loss": 0.4722,
+      "step": 6080
+    },
+    {
+      "epoch": 1.356457729199197,
+      "grad_norm": 0.1741677075624466,
+      "learning_rate": 1.162718902399146e-05,
+      "loss": 0.4478,
+      "step": 6081
+    },
+    {
+      "epoch": 1.3566807941110863,
+      "grad_norm": 0.17158982157707214,
+      "learning_rate": 1.162486691514353e-05,
+      "loss": 0.4594,
+      "step": 6082
+    },
+    {
+      "epoch": 1.3569038590229756,
+      "grad_norm": 0.16993384063243866,
+      "learning_rate": 1.1622544716300245e-05,
+      "loss": 0.4304,
+      "step": 6083
+    },
+    {
+      "epoch": 1.3571269239348651,
+      "grad_norm": 0.1746159791946411,
+      "learning_rate": 1.1620222427590232e-05,
+      "loss": 0.4395,
+      "step": 6084
+    },
+    {
+      "epoch": 1.3573499888467544,
+      "grad_norm": 0.18195772171020508,
+      "learning_rate": 1.1617900049142105e-05,
+      "loss": 0.449,
+      "step": 6085
+    },
+    {
+      "epoch": 1.3575730537586437,
+      "grad_norm": 0.19646844267845154,
+      "learning_rate": 1.16155775810845e-05,
+      "loss": 0.4391,
+      "step": 6086
+    },
+    {
+      "epoch": 1.3577961186705332,
+      "grad_norm": 0.17097879946231842,
+      "learning_rate": 1.1613255023546043e-05,
+      "loss": 0.4744,
+      "step": 6087
+    },
+    {
+      "epoch": 1.3580191835824225,
+      "grad_norm": 0.18478180468082428,
+      "learning_rate": 1.1610932376655377e-05,
+      "loss": 0.5241,
+      "step": 6088
+    },
+    {
+      "epoch": 1.358242248494312,
+      "grad_norm": 0.16517724096775055,
+      "learning_rate": 1.1608609640541142e-05,
+      "loss": 0.465,
+      "step": 6089
+    },
+    {
+      "epoch": 1.3584653134062012,
+      "grad_norm": 0.17419113218784332,
+      "learning_rate": 1.1606286815331988e-05,
+      "loss": 0.4501,
+      "step": 6090
+    },
+    {
+      "epoch": 1.3586883783180905,
+      "grad_norm": 0.18028919398784637,
+      "learning_rate": 1.1603963901156563e-05,
+      "loss": 0.5103,
+      "step": 6091
+    },
+    {
+      "epoch": 1.3589114432299798,
+      "grad_norm": 0.17037071287631989,
+      "learning_rate": 1.1601640898143529e-05,
+      "loss": 0.48,
+      "step": 6092
+    },
+    {
+      "epoch": 1.3591345081418693,
+      "grad_norm": 0.1665482372045517,
+      "learning_rate": 1.1599317806421548e-05,
+      "loss": 0.4826,
+      "step": 6093
+    },
+    {
+      "epoch": 1.3593575730537586,
+      "grad_norm": 0.18023498356342316,
+      "learning_rate": 1.1596994626119287e-05,
+      "loss": 0.4615,
+      "step": 6094
+    },
+    {
+      "epoch": 1.359580637965648,
+      "grad_norm": 0.16920949518680573,
+      "learning_rate": 1.159467135736542e-05,
+      "loss": 0.4777,
+      "step": 6095
+    },
+    {
+      "epoch": 1.3598037028775374,
+      "grad_norm": 0.17280833423137665,
+      "learning_rate": 1.1592348000288618e-05,
+      "loss": 0.4691,
+      "step": 6096
+    },
+    {
+      "epoch": 1.3600267677894267,
+      "grad_norm": 0.17000465095043182,
+      "learning_rate": 1.1590024555017571e-05,
+      "loss": 0.44,
+      "step": 6097
+    },
+    {
+      "epoch": 1.3602498327013162,
+      "grad_norm": 0.1656549870967865,
+      "learning_rate": 1.1587701021680959e-05,
+      "loss": 0.4266,
+      "step": 6098
+    },
+    {
+      "epoch": 1.3604728976132054,
+      "grad_norm": 0.1633346974849701,
+      "learning_rate": 1.1585377400407483e-05,
+      "loss": 0.3994,
+      "step": 6099
+    },
+    {
+      "epoch": 1.3606959625250947,
+      "grad_norm": 0.1661108434200287,
+      "learning_rate": 1.1583053691325829e-05,
+      "loss": 0.4668,
+      "step": 6100
+    },
+    {
+      "epoch": 1.3609190274369842,
+      "grad_norm": 0.17027215659618378,
+      "learning_rate": 1.1580729894564706e-05,
+      "loss": 0.4536,
+      "step": 6101
+    },
+    {
+      "epoch": 1.3611420923488735,
+      "grad_norm": 0.17399293184280396,
+      "learning_rate": 1.1578406010252818e-05,
+      "loss": 0.4718,
+      "step": 6102
+    },
+    {
+      "epoch": 1.3613651572607628,
+      "grad_norm": 0.16790112853050232,
+      "learning_rate": 1.1576082038518876e-05,
+      "loss": 0.4487,
+      "step": 6103
+    },
+    {
+      "epoch": 1.3615882221726523,
+      "grad_norm": 0.16512347757816315,
+      "learning_rate": 1.15737579794916e-05,
+      "loss": 0.4497,
+      "step": 6104
+    },
+    {
+      "epoch": 1.3618112870845416,
+      "grad_norm": 0.16861163079738617,
+      "learning_rate": 1.1571433833299703e-05,
+      "loss": 0.4552,
+      "step": 6105
+    },
+    {
+      "epoch": 1.362034351996431,
+      "grad_norm": 0.16993777453899384,
+      "learning_rate": 1.156910960007192e-05,
+      "loss": 0.4651,
+      "step": 6106
+    },
+    {
+      "epoch": 1.3622574169083204,
+      "grad_norm": 0.16158044338226318,
+      "learning_rate": 1.1566785279936972e-05,
+      "loss": 0.4361,
+      "step": 6107
+    },
+    {
+      "epoch": 1.3624804818202096,
+      "grad_norm": 0.17330992221832275,
+      "learning_rate": 1.1564460873023604e-05,
+      "loss": 0.4651,
+      "step": 6108
+    },
+    {
+      "epoch": 1.362703546732099,
+      "grad_norm": 0.16832031309604645,
+      "learning_rate": 1.156213637946055e-05,
+      "loss": 0.4802,
+      "step": 6109
+    },
+    {
+      "epoch": 1.3629266116439884,
+      "grad_norm": 0.17535291612148285,
+      "learning_rate": 1.1559811799376557e-05,
+      "loss": 0.4598,
+      "step": 6110
+    },
+    {
+      "epoch": 1.3631496765558777,
+      "grad_norm": 0.17599309980869293,
+      "learning_rate": 1.1557487132900376e-05,
+      "loss": 0.4568,
+      "step": 6111
+    },
+    {
+      "epoch": 1.3633727414677672,
+      "grad_norm": 0.18158310651779175,
+      "learning_rate": 1.155516238016076e-05,
+      "loss": 0.4597,
+      "step": 6112
+    },
+    {
+      "epoch": 1.3635958063796565,
+      "grad_norm": 0.1722884625196457,
+      "learning_rate": 1.1552837541286468e-05,
+      "loss": 0.4703,
+      "step": 6113
+    },
+    {
+      "epoch": 1.3638188712915458,
+      "grad_norm": 0.16803938150405884,
+      "learning_rate": 1.1550512616406269e-05,
+      "loss": 0.4524,
+      "step": 6114
+    },
+    {
+      "epoch": 1.3640419362034353,
+      "grad_norm": 0.16398029029369354,
+      "learning_rate": 1.1548187605648923e-05,
+      "loss": 0.476,
+      "step": 6115
+    },
+    {
+      "epoch": 1.3642650011153246,
+      "grad_norm": 0.17081895470619202,
+      "learning_rate": 1.1545862509143212e-05,
+      "loss": 0.4794,
+      "step": 6116
+    },
+    {
+      "epoch": 1.3644880660272138,
+      "grad_norm": 0.16784706711769104,
+      "learning_rate": 1.1543537327017911e-05,
+      "loss": 0.4522,
+      "step": 6117
+    },
+    {
+      "epoch": 1.3647111309391033,
+      "grad_norm": 0.1711539328098297,
+      "learning_rate": 1.1541212059401806e-05,
+      "loss": 0.4607,
+      "step": 6118
+    },
+    {
+      "epoch": 1.3649341958509926,
+      "grad_norm": 0.17196914553642273,
+      "learning_rate": 1.1538886706423678e-05,
+      "loss": 0.4689,
+      "step": 6119
+    },
+    {
+      "epoch": 1.365157260762882,
+      "grad_norm": 0.17912612855434418,
+      "learning_rate": 1.1536561268212328e-05,
+      "loss": 0.4616,
+      "step": 6120
+    },
+    {
+      "epoch": 1.3653803256747714,
+      "grad_norm": 0.16970542073249817,
+      "learning_rate": 1.1534235744896547e-05,
+      "loss": 0.4481,
+      "step": 6121
+    },
+    {
+      "epoch": 1.3656033905866607,
+      "grad_norm": 0.17101642489433289,
+      "learning_rate": 1.153191013660514e-05,
+      "loss": 0.4507,
+      "step": 6122
+    },
+    {
+      "epoch": 1.3658264554985502,
+      "grad_norm": 0.16778020560741425,
+      "learning_rate": 1.1529584443466915e-05,
+      "loss": 0.4473,
+      "step": 6123
+    },
+    {
+      "epoch": 1.3660495204104395,
+      "grad_norm": 0.16892312467098236,
+      "learning_rate": 1.1527258665610681e-05,
+      "loss": 0.4837,
+      "step": 6124
+    },
+    {
+      "epoch": 1.3662725853223288,
+      "grad_norm": 0.1817074865102768,
+      "learning_rate": 1.1524932803165254e-05,
+      "loss": 0.4611,
+      "step": 6125
+    },
+    {
+      "epoch": 1.366495650234218,
+      "grad_norm": 0.1750434935092926,
+      "learning_rate": 1.1522606856259457e-05,
+      "loss": 0.4877,
+      "step": 6126
+    },
+    {
+      "epoch": 1.3667187151461075,
+      "grad_norm": 0.16866067051887512,
+      "learning_rate": 1.1520280825022116e-05,
+      "loss": 0.4593,
+      "step": 6127
+    },
+    {
+      "epoch": 1.3669417800579968,
+      "grad_norm": 0.16479606926441193,
+      "learning_rate": 1.1517954709582058e-05,
+      "loss": 0.4685,
+      "step": 6128
+    },
+    {
+      "epoch": 1.3671648449698863,
+      "grad_norm": 0.16955013573169708,
+      "learning_rate": 1.1515628510068122e-05,
+      "loss": 0.4905,
+      "step": 6129
+    },
+    {
+      "epoch": 1.3673879098817756,
+      "grad_norm": 0.1888909786939621,
+      "learning_rate": 1.1513302226609144e-05,
+      "loss": 0.4382,
+      "step": 6130
+    },
+    {
+      "epoch": 1.367610974793665,
+      "grad_norm": 0.179367333650589,
+      "learning_rate": 1.151097585933397e-05,
+      "loss": 0.4522,
+      "step": 6131
+    },
+    {
+      "epoch": 1.3678340397055544,
+      "grad_norm": 0.16989760100841522,
+      "learning_rate": 1.1508649408371448e-05,
+      "loss": 0.4614,
+      "step": 6132
+    },
+    {
+      "epoch": 1.3680571046174437,
+      "grad_norm": 0.17462711036205292,
+      "learning_rate": 1.1506322873850434e-05,
+      "loss": 0.4512,
+      "step": 6133
+    },
+    {
+      "epoch": 1.368280169529333,
+      "grad_norm": 0.18676964938640594,
+      "learning_rate": 1.1503996255899783e-05,
+      "loss": 0.4778,
+      "step": 6134
+    },
+    {
+      "epoch": 1.3685032344412225,
+      "grad_norm": 0.1641889065504074,
+      "learning_rate": 1.1501669554648359e-05,
+      "loss": 0.4562,
+      "step": 6135
+    },
+    {
+      "epoch": 1.3687262993531117,
+      "grad_norm": 0.1635216772556305,
+      "learning_rate": 1.1499342770225028e-05,
+      "loss": 0.456,
+      "step": 6136
+    },
+    {
+      "epoch": 1.368949364265001,
+      "grad_norm": 0.20571352541446686,
+      "learning_rate": 1.1497015902758663e-05,
+      "loss": 0.4625,
+      "step": 6137
+    },
+    {
+      "epoch": 1.3691724291768905,
+      "grad_norm": 0.1723473221063614,
+      "learning_rate": 1.1494688952378141e-05,
+      "loss": 0.4592,
+      "step": 6138
+    },
+    {
+      "epoch": 1.3693954940887798,
+      "grad_norm": 0.167448028922081,
+      "learning_rate": 1.1492361919212345e-05,
+      "loss": 0.454,
+      "step": 6139
+    },
+    {
+      "epoch": 1.3696185590006693,
+      "grad_norm": 0.16840708255767822,
+      "learning_rate": 1.1490034803390157e-05,
+      "loss": 0.4703,
+      "step": 6140
+    },
+    {
+      "epoch": 1.3698416239125586,
+      "grad_norm": 0.17156104743480682,
+      "learning_rate": 1.148770760504047e-05,
+      "loss": 0.4603,
+      "step": 6141
+    },
+    {
+      "epoch": 1.3700646888244479,
+      "grad_norm": 0.16438238322734833,
+      "learning_rate": 1.1485380324292175e-05,
+      "loss": 0.4521,
+      "step": 6142
+    },
+    {
+      "epoch": 1.3702877537363372,
+      "grad_norm": 0.17488059401512146,
+      "learning_rate": 1.1483052961274177e-05,
+      "loss": 0.4845,
+      "step": 6143
+    },
+    {
+      "epoch": 1.3705108186482267,
+      "grad_norm": 0.16744007170200348,
+      "learning_rate": 1.1480725516115374e-05,
+      "loss": 0.441,
+      "step": 6144
+    },
+    {
+      "epoch": 1.370733883560116,
+      "grad_norm": 0.16667571663856506,
+      "learning_rate": 1.1478397988944683e-05,
+      "loss": 0.4253,
+      "step": 6145
+    },
+    {
+      "epoch": 1.3709569484720054,
+      "grad_norm": 0.1638866513967514,
+      "learning_rate": 1.1476070379891009e-05,
+      "loss": 0.4539,
+      "step": 6146
+    },
+    {
+      "epoch": 1.3711800133838947,
+      "grad_norm": 0.2004345804452896,
+      "learning_rate": 1.1473742689083271e-05,
+      "loss": 0.4615,
+      "step": 6147
+    },
+    {
+      "epoch": 1.371403078295784,
+      "grad_norm": 0.17425598204135895,
+      "learning_rate": 1.1471414916650397e-05,
+      "loss": 0.4428,
+      "step": 6148
+    },
+    {
+      "epoch": 1.3716261432076735,
+      "grad_norm": 0.1777925044298172,
+      "learning_rate": 1.1469087062721305e-05,
+      "loss": 0.4741,
+      "step": 6149
+    },
+    {
+      "epoch": 1.3718492081195628,
+      "grad_norm": 0.17382198572158813,
+      "learning_rate": 1.146675912742493e-05,
+      "loss": 0.4744,
+      "step": 6150
+    },
+    {
+      "epoch": 1.372072273031452,
+      "grad_norm": 0.18798965215682983,
+      "learning_rate": 1.146443111089021e-05,
+      "loss": 0.4216,
+      "step": 6151
+    },
+    {
+      "epoch": 1.3722953379433416,
+      "grad_norm": 0.1665268987417221,
+      "learning_rate": 1.1462103013246086e-05,
+      "loss": 0.4669,
+      "step": 6152
+    },
+    {
+      "epoch": 1.3725184028552309,
+      "grad_norm": 0.1658693253993988,
+      "learning_rate": 1.1459774834621498e-05,
+      "loss": 0.444,
+      "step": 6153
+    },
+    {
+      "epoch": 1.3727414677671201,
+      "grad_norm": 0.16875915229320526,
+      "learning_rate": 1.1457446575145397e-05,
+      "loss": 0.46,
+      "step": 6154
+    },
+    {
+      "epoch": 1.3729645326790096,
+      "grad_norm": 0.17175830900669098,
+      "learning_rate": 1.1455118234946737e-05,
+      "loss": 0.4879,
+      "step": 6155
+    },
+    {
+      "epoch": 1.373187597590899,
+      "grad_norm": 0.17912839353084564,
+      "learning_rate": 1.1452789814154475e-05,
+      "loss": 0.4488,
+      "step": 6156
+    },
+    {
+      "epoch": 1.3734106625027884,
+      "grad_norm": 0.16743507981300354,
+      "learning_rate": 1.1450461312897576e-05,
+      "loss": 0.4565,
+      "step": 6157
+    },
+    {
+      "epoch": 1.3736337274146777,
+      "grad_norm": 0.1700303554534912,
+      "learning_rate": 1.1448132731305005e-05,
+      "loss": 0.4278,
+      "step": 6158
+    },
+    {
+      "epoch": 1.373856792326567,
+      "grad_norm": 0.17981837689876556,
+      "learning_rate": 1.1445804069505735e-05,
+      "loss": 0.4379,
+      "step": 6159
+    },
+    {
+      "epoch": 1.3740798572384563,
+      "grad_norm": 0.172749325633049,
+      "learning_rate": 1.1443475327628739e-05,
+      "loss": 0.4344,
+      "step": 6160
+    },
+    {
+      "epoch": 1.3743029221503458,
+      "grad_norm": 0.16664694249629974,
+      "learning_rate": 1.1441146505803003e-05,
+      "loss": 0.4574,
+      "step": 6161
+    },
+    {
+      "epoch": 1.374525987062235,
+      "grad_norm": 0.1706552803516388,
+      "learning_rate": 1.1438817604157506e-05,
+      "loss": 0.4525,
+      "step": 6162
+    },
+    {
+      "epoch": 1.3747490519741246,
+      "grad_norm": 0.17497579753398895,
+      "learning_rate": 1.1436488622821243e-05,
+      "loss": 0.4882,
+      "step": 6163
+    },
+    {
+      "epoch": 1.3749721168860138,
+      "grad_norm": 0.17119908332824707,
+      "learning_rate": 1.14341595619232e-05,
+      "loss": 0.4728,
+      "step": 6164
+    },
+    {
+      "epoch": 1.3751951817979031,
+      "grad_norm": 0.1758640706539154,
+      "learning_rate": 1.143183042159238e-05,
+      "loss": 0.4205,
+      "step": 6165
+    },
+    {
+      "epoch": 1.3754182467097926,
+      "grad_norm": 0.17280808091163635,
+      "learning_rate": 1.1429501201957785e-05,
+      "loss": 0.4964,
+      "step": 6166
+    },
+    {
+      "epoch": 1.375641311621682,
+      "grad_norm": 0.17379097640514374,
+      "learning_rate": 1.1427171903148425e-05,
+      "loss": 0.4624,
+      "step": 6167
+    },
+    {
+      "epoch": 1.3758643765335712,
+      "grad_norm": 0.17071013152599335,
+      "learning_rate": 1.14248425252933e-05,
+      "loss": 0.4503,
+      "step": 6168
+    },
+    {
+      "epoch": 1.3760874414454607,
+      "grad_norm": 0.18627150356769562,
+      "learning_rate": 1.1422513068521442e-05,
+      "loss": 0.4476,
+      "step": 6169
+    },
+    {
+      "epoch": 1.37631050635735,
+      "grad_norm": 0.1711094081401825,
+      "learning_rate": 1.1420183532961855e-05,
+      "loss": 0.445,
+      "step": 6170
+    },
+    {
+      "epoch": 1.3765335712692393,
+      "grad_norm": 0.16976723074913025,
+      "learning_rate": 1.1417853918743576e-05,
+      "loss": 0.4638,
+      "step": 6171
+    },
+    {
+      "epoch": 1.3767566361811288,
+      "grad_norm": 0.15716324746608734,
+      "learning_rate": 1.1415524225995624e-05,
+      "loss": 0.4555,
+      "step": 6172
+    },
+    {
+      "epoch": 1.376979701093018,
+      "grad_norm": 0.17595480382442474,
+      "learning_rate": 1.1413194454847041e-05,
+      "loss": 0.4777,
+      "step": 6173
+    },
+    {
+      "epoch": 1.3772027660049075,
+      "grad_norm": 0.1665569245815277,
+      "learning_rate": 1.1410864605426856e-05,
+      "loss": 0.4478,
+      "step": 6174
+    },
+    {
+      "epoch": 1.3774258309167968,
+      "grad_norm": 0.16966520249843597,
+      "learning_rate": 1.1408534677864119e-05,
+      "loss": 0.4626,
+      "step": 6175
+    },
+    {
+      "epoch": 1.377648895828686,
+      "grad_norm": 0.17505811154842377,
+      "learning_rate": 1.1406204672287867e-05,
+      "loss": 0.4595,
+      "step": 6176
+    },
+    {
+      "epoch": 1.3778719607405754,
+      "grad_norm": 0.1676858365535736,
+      "learning_rate": 1.1403874588827156e-05,
+      "loss": 0.4451,
+      "step": 6177
+    },
+    {
+      "epoch": 1.3780950256524649,
+      "grad_norm": 0.1657474786043167,
+      "learning_rate": 1.1401544427611037e-05,
+      "loss": 0.4268,
+      "step": 6178
+    },
+    {
+      "epoch": 1.3783180905643542,
+      "grad_norm": 0.16758085787296295,
+      "learning_rate": 1.1399214188768574e-05,
+      "loss": 0.4783,
+      "step": 6179
+    },
+    {
+      "epoch": 1.3785411554762437,
+      "grad_norm": 0.17600210011005402,
+      "learning_rate": 1.139688387242883e-05,
+      "loss": 0.4546,
+      "step": 6180
+    },
+    {
+      "epoch": 1.378764220388133,
+      "grad_norm": 0.17735408246517181,
+      "learning_rate": 1.1394553478720868e-05,
+      "loss": 0.4663,
+      "step": 6181
+    },
+    {
+      "epoch": 1.3789872853000222,
+      "grad_norm": 0.17437157034873962,
+      "learning_rate": 1.1392223007773764e-05,
+      "loss": 0.4914,
+      "step": 6182
+    },
+    {
+      "epoch": 1.3792103502119117,
+      "grad_norm": 0.16675308346748352,
+      "learning_rate": 1.1389892459716589e-05,
+      "loss": 0.463,
+      "step": 6183
+    },
+    {
+      "epoch": 1.379433415123801,
+      "grad_norm": 0.17255185544490814,
+      "learning_rate": 1.138756183467843e-05,
+      "loss": 0.4716,
+      "step": 6184
+    },
+    {
+      "epoch": 1.3796564800356905,
+      "grad_norm": 0.16526752710342407,
+      "learning_rate": 1.1385231132788368e-05,
+      "loss": 0.427,
+      "step": 6185
+    },
+    {
+      "epoch": 1.3798795449475798,
+      "grad_norm": 0.16648603975772858,
+      "learning_rate": 1.1382900354175494e-05,
+      "loss": 0.4393,
+      "step": 6186
+    },
+    {
+      "epoch": 1.380102609859469,
+      "grad_norm": 0.16615265607833862,
+      "learning_rate": 1.1380569498968896e-05,
+      "loss": 0.4493,
+      "step": 6187
+    },
+    {
+      "epoch": 1.3803256747713584,
+      "grad_norm": 0.19119910895824432,
+      "learning_rate": 1.1378238567297677e-05,
+      "loss": 0.4583,
+      "step": 6188
+    },
+    {
+      "epoch": 1.3805487396832479,
+      "grad_norm": 0.17033104598522186,
+      "learning_rate": 1.1375907559290935e-05,
+      "loss": 0.4919,
+      "step": 6189
+    },
+    {
+      "epoch": 1.3807718045951372,
+      "grad_norm": 0.16820746660232544,
+      "learning_rate": 1.1373576475077778e-05,
+      "loss": 0.4316,
+      "step": 6190
+    },
+    {
+      "epoch": 1.3809948695070267,
+      "grad_norm": 0.17084772884845734,
+      "learning_rate": 1.1371245314787318e-05,
+      "loss": 0.4225,
+      "step": 6191
+    },
+    {
+      "epoch": 1.381217934418916,
+      "grad_norm": 0.18454383313655853,
+      "learning_rate": 1.1368914078548666e-05,
+      "loss": 0.4675,
+      "step": 6192
+    },
+    {
+      "epoch": 1.3814409993308052,
+      "grad_norm": 0.3558218777179718,
+      "learning_rate": 1.136658276649094e-05,
+      "loss": 0.4507,
+      "step": 6193
+    },
+    {
+      "epoch": 1.3816640642426945,
+      "grad_norm": 0.18921050429344177,
+      "learning_rate": 1.1364251378743266e-05,
+      "loss": 0.4656,
+      "step": 6194
+    },
+    {
+      "epoch": 1.381887129154584,
+      "grad_norm": 0.1709054410457611,
+      "learning_rate": 1.136191991543477e-05,
+      "loss": 0.468,
+      "step": 6195
+    },
+    {
+      "epoch": 1.3821101940664733,
+      "grad_norm": 0.17301930487155914,
+      "learning_rate": 1.1359588376694577e-05,
+      "loss": 0.4486,
+      "step": 6196
+    },
+    {
+      "epoch": 1.3823332589783628,
+      "grad_norm": 0.17260199785232544,
+      "learning_rate": 1.1357256762651834e-05,
+      "loss": 0.4671,
+      "step": 6197
+    },
+    {
+      "epoch": 1.382556323890252,
+      "grad_norm": 0.16965550184249878,
+      "learning_rate": 1.135492507343567e-05,
+      "loss": 0.4657,
+      "step": 6198
+    },
+    {
+      "epoch": 1.3827793888021414,
+      "grad_norm": 0.17701901495456696,
+      "learning_rate": 1.1352593309175233e-05,
+      "loss": 0.4698,
+      "step": 6199
+    },
+    {
+      "epoch": 1.3830024537140309,
+      "grad_norm": 0.16689814627170563,
+      "learning_rate": 1.135026146999967e-05,
+      "loss": 0.4606,
+      "step": 6200
+    },
+    {
+      "epoch": 1.3832255186259201,
+      "grad_norm": 0.1691475808620453,
+      "learning_rate": 1.1347929556038135e-05,
+      "loss": 0.4287,
+      "step": 6201
+    },
+    {
+      "epoch": 1.3834485835378096,
+      "grad_norm": 0.17036226391792297,
+      "learning_rate": 1.1345597567419782e-05,
+      "loss": 0.44,
+      "step": 6202
+    },
+    {
+      "epoch": 1.383671648449699,
+      "grad_norm": 0.1688498854637146,
+      "learning_rate": 1.1343265504273773e-05,
+      "loss": 0.4621,
+      "step": 6203
+    },
+    {
+      "epoch": 1.3838947133615882,
+      "grad_norm": 0.17137567698955536,
+      "learning_rate": 1.1340933366729268e-05,
+      "loss": 0.4415,
+      "step": 6204
+    },
+    {
+      "epoch": 1.3841177782734775,
+      "grad_norm": 0.17283524572849274,
+      "learning_rate": 1.1338601154915441e-05,
+      "loss": 0.4303,
+      "step": 6205
+    },
+    {
+      "epoch": 1.384340843185367,
+      "grad_norm": 0.1761488914489746,
+      "learning_rate": 1.1336268868961459e-05,
+      "loss": 0.4892,
+      "step": 6206
+    },
+    {
+      "epoch": 1.3845639080972563,
+      "grad_norm": 0.1702503263950348,
+      "learning_rate": 1.1333936508996503e-05,
+      "loss": 0.4523,
+      "step": 6207
+    },
+    {
+      "epoch": 1.3847869730091458,
+      "grad_norm": 0.17347446084022522,
+      "learning_rate": 1.1331604075149753e-05,
+      "loss": 0.4681,
+      "step": 6208
+    },
+    {
+      "epoch": 1.385010037921035,
+      "grad_norm": 0.18203677237033844,
+      "learning_rate": 1.1329271567550394e-05,
+      "loss": 0.4697,
+      "step": 6209
+    },
+    {
+      "epoch": 1.3852331028329243,
+      "grad_norm": 0.168579563498497,
+      "learning_rate": 1.132693898632761e-05,
+      "loss": 0.4579,
+      "step": 6210
+    },
+    {
+      "epoch": 1.3854561677448136,
+      "grad_norm": 0.17098036408424377,
+      "learning_rate": 1.1324606331610602e-05,
+      "loss": 0.4451,
+      "step": 6211
+    },
+    {
+      "epoch": 1.3856792326567031,
+      "grad_norm": 0.1700953096151352,
+      "learning_rate": 1.1322273603528562e-05,
+      "loss": 0.4446,
+      "step": 6212
+    },
+    {
+      "epoch": 1.3859022975685924,
+      "grad_norm": 0.18041159212589264,
+      "learning_rate": 1.1319940802210692e-05,
+      "loss": 0.4943,
+      "step": 6213
+    },
+    {
+      "epoch": 1.386125362480482,
+      "grad_norm": 0.17734865844249725,
+      "learning_rate": 1.1317607927786201e-05,
+      "loss": 0.4507,
+      "step": 6214
+    },
+    {
+      "epoch": 1.3863484273923712,
+      "grad_norm": 0.18012134730815887,
+      "learning_rate": 1.131527498038429e-05,
+      "loss": 0.4572,
+      "step": 6215
+    },
+    {
+      "epoch": 1.3865714923042605,
+      "grad_norm": 0.17000438272953033,
+      "learning_rate": 1.131294196013418e-05,
+      "loss": 0.4896,
+      "step": 6216
+    },
+    {
+      "epoch": 1.38679455721615,
+      "grad_norm": 0.16048485040664673,
+      "learning_rate": 1.1310608867165082e-05,
+      "loss": 0.4398,
+      "step": 6217
+    },
+    {
+      "epoch": 1.3870176221280393,
+      "grad_norm": 0.17833036184310913,
+      "learning_rate": 1.1308275701606226e-05,
+      "loss": 0.4706,
+      "step": 6218
+    },
+    {
+      "epoch": 1.3872406870399288,
+      "grad_norm": 0.17137986421585083,
+      "learning_rate": 1.130594246358683e-05,
+      "loss": 0.4862,
+      "step": 6219
+    },
+    {
+      "epoch": 1.387463751951818,
+      "grad_norm": 0.1729787439107895,
+      "learning_rate": 1.1303609153236127e-05,
+      "loss": 0.4607,
+      "step": 6220
+    },
+    {
+      "epoch": 1.3876868168637073,
+      "grad_norm": 0.16867367923259735,
+      "learning_rate": 1.1301275770683344e-05,
+      "loss": 0.4671,
+      "step": 6221
+    },
+    {
+      "epoch": 1.3879098817755966,
+      "grad_norm": 0.17126032710075378,
+      "learning_rate": 1.1298942316057731e-05,
+      "loss": 0.4162,
+      "step": 6222
+    },
+    {
+      "epoch": 1.388132946687486,
+      "grad_norm": 0.16363003849983215,
+      "learning_rate": 1.1296608789488515e-05,
+      "loss": 0.4292,
+      "step": 6223
+    },
+    {
+      "epoch": 1.3883560115993754,
+      "grad_norm": 0.16574527323246002,
+      "learning_rate": 1.1294275191104952e-05,
+      "loss": 0.4622,
+      "step": 6224
+    },
+    {
+      "epoch": 1.3885790765112649,
+      "grad_norm": 0.16721156239509583,
+      "learning_rate": 1.1291941521036286e-05,
+      "loss": 0.4569,
+      "step": 6225
+    },
+    {
+      "epoch": 1.3888021414231542,
+      "grad_norm": 0.16934387385845184,
+      "learning_rate": 1.1289607779411775e-05,
+      "loss": 0.4488,
+      "step": 6226
+    },
+    {
+      "epoch": 1.3890252063350434,
+      "grad_norm": 0.16877540946006775,
+      "learning_rate": 1.1287273966360673e-05,
+      "loss": 0.456,
+      "step": 6227
+    },
+    {
+      "epoch": 1.3892482712469327,
+      "grad_norm": 0.17397183179855347,
+      "learning_rate": 1.1284940082012238e-05,
+      "loss": 0.4727,
+      "step": 6228
+    },
+    {
+      "epoch": 1.3894713361588222,
+      "grad_norm": 0.1647840440273285,
+      "learning_rate": 1.128260612649574e-05,
+      "loss": 0.4497,
+      "step": 6229
+    },
+    {
+      "epoch": 1.3896944010707115,
+      "grad_norm": 0.1644977629184723,
+      "learning_rate": 1.1280272099940446e-05,
+      "loss": 0.4395,
+      "step": 6230
+    },
+    {
+      "epoch": 1.389917465982601,
+      "grad_norm": 0.17503903806209564,
+      "learning_rate": 1.1277938002475633e-05,
+      "loss": 0.4515,
+      "step": 6231
+    },
+    {
+      "epoch": 1.3901405308944903,
+      "grad_norm": 0.16988563537597656,
+      "learning_rate": 1.127560383423057e-05,
+      "loss": 0.4365,
+      "step": 6232
+    },
+    {
+      "epoch": 1.3903635958063796,
+      "grad_norm": 0.1652899980545044,
+      "learning_rate": 1.1273269595334547e-05,
+      "loss": 0.4484,
+      "step": 6233
+    },
+    {
+      "epoch": 1.390586660718269,
+      "grad_norm": 0.1692708283662796,
+      "learning_rate": 1.1270935285916842e-05,
+      "loss": 0.4472,
+      "step": 6234
+    },
+    {
+      "epoch": 1.3908097256301584,
+      "grad_norm": 0.16943638026714325,
+      "learning_rate": 1.1268600906106749e-05,
+      "loss": 0.4454,
+      "step": 6235
+    },
+    {
+      "epoch": 1.3910327905420479,
+      "grad_norm": 0.1718982458114624,
+      "learning_rate": 1.1266266456033555e-05,
+      "loss": 0.4305,
+      "step": 6236
+    },
+    {
+      "epoch": 1.3912558554539372,
+      "grad_norm": 0.17046236991882324,
+      "learning_rate": 1.1263931935826561e-05,
+      "loss": 0.4383,
+      "step": 6237
+    },
+    {
+      "epoch": 1.3914789203658264,
+      "grad_norm": 0.17139141261577606,
+      "learning_rate": 1.1261597345615064e-05,
+      "loss": 0.4437,
+      "step": 6238
+    },
+    {
+      "epoch": 1.3917019852777157,
+      "grad_norm": 0.1709447205066681,
+      "learning_rate": 1.1259262685528376e-05,
+      "loss": 0.4938,
+      "step": 6239
+    },
+    {
+      "epoch": 1.3919250501896052,
+      "grad_norm": 0.16337740421295166,
+      "learning_rate": 1.1256927955695793e-05,
+      "loss": 0.4651,
+      "step": 6240
+    },
+    {
+      "epoch": 1.3921481151014945,
+      "grad_norm": 0.17332367599010468,
+      "learning_rate": 1.1254593156246638e-05,
+      "loss": 0.4803,
+      "step": 6241
+    },
+    {
+      "epoch": 1.392371180013384,
+      "grad_norm": 0.17104537785053253,
+      "learning_rate": 1.1252258287310219e-05,
+      "loss": 0.4717,
+      "step": 6242
+    },
+    {
+      "epoch": 1.3925942449252733,
+      "grad_norm": 0.16978146135807037,
+      "learning_rate": 1.1249923349015859e-05,
+      "loss": 0.4999,
+      "step": 6243
+    },
+    {
+      "epoch": 1.3928173098371626,
+      "grad_norm": 0.17217592895030975,
+      "learning_rate": 1.1247588341492884e-05,
+      "loss": 0.4738,
+      "step": 6244
+    },
+    {
+      "epoch": 1.3930403747490518,
+      "grad_norm": 0.17651988565921783,
+      "learning_rate": 1.1245253264870616e-05,
+      "loss": 0.441,
+      "step": 6245
+    },
+    {
+      "epoch": 1.3932634396609413,
+      "grad_norm": 0.16616612672805786,
+      "learning_rate": 1.1242918119278395e-05,
+      "loss": 0.4334,
+      "step": 6246
+    },
+    {
+      "epoch": 1.3934865045728306,
+      "grad_norm": 0.17374052107334137,
+      "learning_rate": 1.1240582904845542e-05,
+      "loss": 0.4464,
+      "step": 6247
+    },
+    {
+      "epoch": 1.3937095694847201,
+      "grad_norm": 0.17393873631954193,
+      "learning_rate": 1.1238247621701413e-05,
+      "loss": 0.4438,
+      "step": 6248
+    },
+    {
+      "epoch": 1.3939326343966094,
+      "grad_norm": 0.17229264974594116,
+      "learning_rate": 1.1235912269975335e-05,
+      "loss": 0.4775,
+      "step": 6249
+    },
+    {
+      "epoch": 1.3941556993084987,
+      "grad_norm": 0.17282813787460327,
+      "learning_rate": 1.1233576849796666e-05,
+      "loss": 0.454,
+      "step": 6250
+    },
+    {
+      "epoch": 1.3943787642203882,
+      "grad_norm": 0.18351981043815613,
+      "learning_rate": 1.1231241361294747e-05,
+      "loss": 0.4702,
+      "step": 6251
+    },
+    {
+      "epoch": 1.3946018291322775,
+      "grad_norm": 0.17423385381698608,
+      "learning_rate": 1.122890580459894e-05,
+      "loss": 0.4506,
+      "step": 6252
+    },
+    {
+      "epoch": 1.394824894044167,
+      "grad_norm": 0.17099998891353607,
+      "learning_rate": 1.1226570179838596e-05,
+      "loss": 0.4509,
+      "step": 6253
+    },
+    {
+      "epoch": 1.3950479589560563,
+      "grad_norm": 0.16235356032848358,
+      "learning_rate": 1.1224234487143085e-05,
+      "loss": 0.4243,
+      "step": 6254
+    },
+    {
+      "epoch": 1.3952710238679455,
+      "grad_norm": 0.17439930140972137,
+      "learning_rate": 1.1221898726641762e-05,
+      "loss": 0.4701,
+      "step": 6255
+    },
+    {
+      "epoch": 1.3954940887798348,
+      "grad_norm": 0.16977347433567047,
+      "learning_rate": 1.1219562898464006e-05,
+      "loss": 0.4516,
+      "step": 6256
+    },
+    {
+      "epoch": 1.3957171536917243,
+      "grad_norm": 0.16299669444561005,
+      "learning_rate": 1.1217227002739181e-05,
+      "loss": 0.4579,
+      "step": 6257
+    },
+    {
+      "epoch": 1.3959402186036136,
+      "grad_norm": 0.17073673009872437,
+      "learning_rate": 1.1214891039596673e-05,
+      "loss": 0.4989,
+      "step": 6258
+    },
+    {
+      "epoch": 1.3961632835155031,
+      "grad_norm": 0.16990354657173157,
+      "learning_rate": 1.1212555009165852e-05,
+      "loss": 0.436,
+      "step": 6259
+    },
+    {
+      "epoch": 1.3963863484273924,
+      "grad_norm": 0.16700886189937592,
+      "learning_rate": 1.1210218911576112e-05,
+      "loss": 0.4324,
+      "step": 6260
+    },
+    {
+      "epoch": 1.3966094133392817,
+      "grad_norm": 0.16864052414894104,
+      "learning_rate": 1.1207882746956834e-05,
+      "loss": 0.4834,
+      "step": 6261
+    },
+    {
+      "epoch": 1.396832478251171,
+      "grad_norm": 0.16238147020339966,
+      "learning_rate": 1.1205546515437413e-05,
+      "loss": 0.4447,
+      "step": 6262
+    },
+    {
+      "epoch": 1.3970555431630605,
+      "grad_norm": 0.1640692502260208,
+      "learning_rate": 1.120321021714724e-05,
+      "loss": 0.4539,
+      "step": 6263
+    },
+    {
+      "epoch": 1.3972786080749497,
+      "grad_norm": 0.16666129231452942,
+      "learning_rate": 1.1200873852215717e-05,
+      "loss": 0.4434,
+      "step": 6264
+    },
+    {
+      "epoch": 1.3975016729868392,
+      "grad_norm": 0.16352206468582153,
+      "learning_rate": 1.1198537420772249e-05,
+      "loss": 0.4345,
+      "step": 6265
+    },
+    {
+      "epoch": 1.3977247378987285,
+      "grad_norm": 0.16906216740608215,
+      "learning_rate": 1.1196200922946237e-05,
+      "loss": 0.4753,
+      "step": 6266
+    },
+    {
+      "epoch": 1.3979478028106178,
+      "grad_norm": 0.16937586665153503,
+      "learning_rate": 1.1193864358867097e-05,
+      "loss": 0.4545,
+      "step": 6267
+    },
+    {
+      "epoch": 1.3981708677225073,
+      "grad_norm": 0.17257355153560638,
+      "learning_rate": 1.1191527728664235e-05,
+      "loss": 0.4722,
+      "step": 6268
+    },
+    {
+      "epoch": 1.3983939326343966,
+      "grad_norm": 0.1705016791820526,
+      "learning_rate": 1.1189191032467074e-05,
+      "loss": 0.4647,
+      "step": 6269
+    },
+    {
+      "epoch": 1.398616997546286,
+      "grad_norm": 0.23135609924793243,
+      "learning_rate": 1.1186854270405035e-05,
+      "loss": 0.4725,
+      "step": 6270
+    },
+    {
+      "epoch": 1.3988400624581754,
+      "grad_norm": 0.17285971343517303,
+      "learning_rate": 1.1184517442607538e-05,
+      "loss": 0.4284,
+      "step": 6271
+    },
+    {
+      "epoch": 1.3990631273700647,
+      "grad_norm": 0.17679868638515472,
+      "learning_rate": 1.1182180549204013e-05,
+      "loss": 0.4412,
+      "step": 6272
+    },
+    {
+      "epoch": 1.399286192281954,
+      "grad_norm": 0.16253626346588135,
+      "learning_rate": 1.1179843590323897e-05,
+      "loss": 0.4321,
+      "step": 6273
+    },
+    {
+      "epoch": 1.3995092571938434,
+      "grad_norm": 0.1678762137889862,
+      "learning_rate": 1.1177506566096619e-05,
+      "loss": 0.483,
+      "step": 6274
+    },
+    {
+      "epoch": 1.3997323221057327,
+      "grad_norm": 0.16577322781085968,
+      "learning_rate": 1.1175169476651622e-05,
+      "loss": 0.4419,
+      "step": 6275
+    },
+    {
+      "epoch": 1.3999553870176222,
+      "grad_norm": 0.16397406160831451,
+      "learning_rate": 1.1172832322118346e-05,
+      "loss": 0.4529,
+      "step": 6276
+    },
+    {
+      "epoch": 1.4001784519295115,
+      "grad_norm": 0.17259271442890167,
+      "learning_rate": 1.1170495102626238e-05,
+      "loss": 0.4763,
+      "step": 6277
+    },
+    {
+      "epoch": 1.4004015168414008,
+      "grad_norm": 0.16409693658351898,
+      "learning_rate": 1.116815781830475e-05,
+      "loss": 0.4737,
+      "step": 6278
+    },
+    {
+      "epoch": 1.40062458175329,
+      "grad_norm": 0.17240269482135773,
+      "learning_rate": 1.1165820469283333e-05,
+      "loss": 0.4708,
+      "step": 6279
+    },
+    {
+      "epoch": 1.4008476466651796,
+      "grad_norm": 0.16498403251171112,
+      "learning_rate": 1.1163483055691447e-05,
+      "loss": 0.4387,
+      "step": 6280
+    },
+    {
+      "epoch": 1.4010707115770689,
+      "grad_norm": 0.16875067353248596,
+      "learning_rate": 1.116114557765855e-05,
+      "loss": 0.4768,
+      "step": 6281
+    },
+    {
+      "epoch": 1.4012937764889584,
+      "grad_norm": 0.16360358893871307,
+      "learning_rate": 1.1158808035314105e-05,
+      "loss": 0.4598,
+      "step": 6282
+    },
+    {
+      "epoch": 1.4015168414008476,
+      "grad_norm": 0.1615404486656189,
+      "learning_rate": 1.1156470428787582e-05,
+      "loss": 0.4529,
+      "step": 6283
+    },
+    {
+      "epoch": 1.401739906312737,
+      "grad_norm": 0.16733388602733612,
+      "learning_rate": 1.1154132758208456e-05,
+      "loss": 0.4564,
+      "step": 6284
+    },
+    {
+      "epoch": 1.4019629712246264,
+      "grad_norm": 0.176735982298851,
+      "learning_rate": 1.1151795023706194e-05,
+      "loss": 0.4371,
+      "step": 6285
+    },
+    {
+      "epoch": 1.4021860361365157,
+      "grad_norm": 0.17406433820724487,
+      "learning_rate": 1.1149457225410281e-05,
+      "loss": 0.4602,
+      "step": 6286
+    },
+    {
+      "epoch": 1.4024091010484052,
+      "grad_norm": 0.18209131062030792,
+      "learning_rate": 1.1147119363450197e-05,
+      "loss": 0.4559,
+      "step": 6287
+    },
+    {
+      "epoch": 1.4026321659602945,
+      "grad_norm": 0.17238366603851318,
+      "learning_rate": 1.1144781437955426e-05,
+      "loss": 0.4725,
+      "step": 6288
+    },
+    {
+      "epoch": 1.4028552308721838,
+      "grad_norm": 0.16568414866924286,
+      "learning_rate": 1.1142443449055455e-05,
+      "loss": 0.4291,
+      "step": 6289
+    },
+    {
+      "epoch": 1.403078295784073,
+      "grad_norm": 0.17432232201099396,
+      "learning_rate": 1.1140105396879783e-05,
+      "loss": 0.4827,
+      "step": 6290
+    },
+    {
+      "epoch": 1.4033013606959626,
+      "grad_norm": 0.17041410505771637,
+      "learning_rate": 1.11377672815579e-05,
+      "loss": 0.4523,
+      "step": 6291
+    },
+    {
+      "epoch": 1.4035244256078518,
+      "grad_norm": 0.1681598722934723,
+      "learning_rate": 1.113542910321931e-05,
+      "loss": 0.4588,
+      "step": 6292
+    },
+    {
+      "epoch": 1.4037474905197413,
+      "grad_norm": 0.16634954512119293,
+      "learning_rate": 1.1133090861993514e-05,
+      "loss": 0.4306,
+      "step": 6293
+    },
+    {
+      "epoch": 1.4039705554316306,
+      "grad_norm": 0.1695157140493393,
+      "learning_rate": 1.113075255801002e-05,
+      "loss": 0.4532,
+      "step": 6294
+    },
+    {
+      "epoch": 1.40419362034352,
+      "grad_norm": 0.16797398030757904,
+      "learning_rate": 1.1128414191398333e-05,
+      "loss": 0.4452,
+      "step": 6295
+    },
+    {
+      "epoch": 1.4044166852554092,
+      "grad_norm": 0.1665017306804657,
+      "learning_rate": 1.1126075762287972e-05,
+      "loss": 0.4523,
+      "step": 6296
+    },
+    {
+      "epoch": 1.4046397501672987,
+      "grad_norm": 0.17001719772815704,
+      "learning_rate": 1.1123737270808452e-05,
+      "loss": 0.4472,
+      "step": 6297
+    },
+    {
+      "epoch": 1.404862815079188,
+      "grad_norm": 0.16899511218070984,
+      "learning_rate": 1.1121398717089294e-05,
+      "loss": 0.4753,
+      "step": 6298
+    },
+    {
+      "epoch": 1.4050858799910775,
+      "grad_norm": 0.16389401257038116,
+      "learning_rate": 1.1119060101260021e-05,
+      "loss": 0.4494,
+      "step": 6299
+    },
+    {
+      "epoch": 1.4053089449029668,
+      "grad_norm": 0.16568145155906677,
+      "learning_rate": 1.1116721423450158e-05,
+      "loss": 0.4509,
+      "step": 6300
+    },
+    {
+      "epoch": 1.405532009814856,
+      "grad_norm": 0.16839419305324554,
+      "learning_rate": 1.1114382683789241e-05,
+      "loss": 0.4425,
+      "step": 6301
+    },
+    {
+      "epoch": 1.4057550747267455,
+      "grad_norm": 0.17624156177043915,
+      "learning_rate": 1.1112043882406802e-05,
+      "loss": 0.4668,
+      "step": 6302
+    },
+    {
+      "epoch": 1.4059781396386348,
+      "grad_norm": 0.2432551085948944,
+      "learning_rate": 1.1109705019432378e-05,
+      "loss": 0.4504,
+      "step": 6303
+    },
+    {
+      "epoch": 1.4062012045505243,
+      "grad_norm": 0.17527322471141815,
+      "learning_rate": 1.1107366094995506e-05,
+      "loss": 0.4734,
+      "step": 6304
+    },
+    {
+      "epoch": 1.4064242694624136,
+      "grad_norm": 0.16600951552391052,
+      "learning_rate": 1.1105027109225737e-05,
+      "loss": 0.4789,
+      "step": 6305
+    },
+    {
+      "epoch": 1.406647334374303,
+      "grad_norm": 0.20400017499923706,
+      "learning_rate": 1.1102688062252614e-05,
+      "loss": 0.4688,
+      "step": 6306
+    },
+    {
+      "epoch": 1.4068703992861922,
+      "grad_norm": 0.17171353101730347,
+      "learning_rate": 1.110034895420569e-05,
+      "loss": 0.4374,
+      "step": 6307
+    },
+    {
+      "epoch": 1.4070934641980817,
+      "grad_norm": 0.1854102462530136,
+      "learning_rate": 1.1098009785214523e-05,
+      "loss": 0.4255,
+      "step": 6308
+    },
+    {
+      "epoch": 1.407316529109971,
+      "grad_norm": 0.16316324472427368,
+      "learning_rate": 1.1095670555408662e-05,
+      "loss": 0.4676,
+      "step": 6309
+    },
+    {
+      "epoch": 1.4075395940218605,
+      "grad_norm": 0.18751972913742065,
+      "learning_rate": 1.1093331264917676e-05,
+      "loss": 0.4734,
+      "step": 6310
+    },
+    {
+      "epoch": 1.4077626589337497,
+      "grad_norm": 0.22429490089416504,
+      "learning_rate": 1.1090991913871128e-05,
+      "loss": 0.444,
+      "step": 6311
+    },
+    {
+      "epoch": 1.407985723845639,
+      "grad_norm": 0.16278603672981262,
+      "learning_rate": 1.1088652502398585e-05,
+      "loss": 0.4528,
+      "step": 6312
+    },
+    {
+      "epoch": 1.4082087887575283,
+      "grad_norm": 0.1643955558538437,
+      "learning_rate": 1.108631303062962e-05,
+      "loss": 0.4646,
+      "step": 6313
+    },
+    {
+      "epoch": 1.4084318536694178,
+      "grad_norm": 0.17145782709121704,
+      "learning_rate": 1.1083973498693802e-05,
+      "loss": 0.4592,
+      "step": 6314
+    },
+    {
+      "epoch": 1.408654918581307,
+      "grad_norm": 0.18193364143371582,
+      "learning_rate": 1.1081633906720714e-05,
+      "loss": 0.4932,
+      "step": 6315
+    },
+    {
+      "epoch": 1.4088779834931966,
+      "grad_norm": 0.17122429609298706,
+      "learning_rate": 1.1079294254839941e-05,
+      "loss": 0.4708,
+      "step": 6316
+    },
+    {
+      "epoch": 1.4091010484050859,
+      "grad_norm": 0.1604917347431183,
+      "learning_rate": 1.1076954543181058e-05,
+      "loss": 0.4659,
+      "step": 6317
+    },
+    {
+      "epoch": 1.4093241133169752,
+      "grad_norm": 0.1739075481891632,
+      "learning_rate": 1.1074614771873661e-05,
+      "loss": 0.4714,
+      "step": 6318
+    },
+    {
+      "epoch": 1.4095471782288647,
+      "grad_norm": 0.16061022877693176,
+      "learning_rate": 1.1072274941047336e-05,
+      "loss": 0.4414,
+      "step": 6319
+    },
+    {
+      "epoch": 1.409770243140754,
+      "grad_norm": 0.1809171438217163,
+      "learning_rate": 1.1069935050831683e-05,
+      "loss": 0.4467,
+      "step": 6320
+    },
+    {
+      "epoch": 1.4099933080526434,
+      "grad_norm": 0.1696557253599167,
+      "learning_rate": 1.1067595101356295e-05,
+      "loss": 0.462,
+      "step": 6321
+    },
+    {
+      "epoch": 1.4102163729645327,
+      "grad_norm": 0.1734270453453064,
+      "learning_rate": 1.1065255092750774e-05,
+      "loss": 0.4859,
+      "step": 6322
+    },
+    {
+      "epoch": 1.410439437876422,
+      "grad_norm": 0.1748056411743164,
+      "learning_rate": 1.1062915025144727e-05,
+      "loss": 0.4709,
+      "step": 6323
+    },
+    {
+      "epoch": 1.4106625027883113,
+      "grad_norm": 0.17737674713134766,
+      "learning_rate": 1.106057489866776e-05,
+      "loss": 0.4875,
+      "step": 6324
+    },
+    {
+      "epoch": 1.4108855677002008,
+      "grad_norm": 0.16855676472187042,
+      "learning_rate": 1.105823471344948e-05,
+      "loss": 0.4713,
+      "step": 6325
+    },
+    {
+      "epoch": 1.41110863261209,
+      "grad_norm": 0.17489537596702576,
+      "learning_rate": 1.105589446961951e-05,
+      "loss": 0.4678,
+      "step": 6326
+    },
+    {
+      "epoch": 1.4113316975239796,
+      "grad_norm": 0.19364696741104126,
+      "learning_rate": 1.1053554167307458e-05,
+      "loss": 0.4732,
+      "step": 6327
+    },
+    {
+      "epoch": 1.4115547624358689,
+      "grad_norm": 0.17721161246299744,
+      "learning_rate": 1.1051213806642951e-05,
+      "loss": 0.4753,
+      "step": 6328
+    },
+    {
+      "epoch": 1.4117778273477581,
+      "grad_norm": 0.17709079384803772,
+      "learning_rate": 1.1048873387755615e-05,
+      "loss": 0.4453,
+      "step": 6329
+    },
+    {
+      "epoch": 1.4120008922596474,
+      "grad_norm": 0.1669922173023224,
+      "learning_rate": 1.1046532910775068e-05,
+      "loss": 0.4639,
+      "step": 6330
+    },
+    {
+      "epoch": 1.412223957171537,
+      "grad_norm": 0.17192141711711884,
+      "learning_rate": 1.1044192375830946e-05,
+      "loss": 0.4559,
+      "step": 6331
+    },
+    {
+      "epoch": 1.4124470220834262,
+      "grad_norm": 0.16646385192871094,
+      "learning_rate": 1.1041851783052882e-05,
+      "loss": 0.4393,
+      "step": 6332
+    },
+    {
+      "epoch": 1.4126700869953157,
+      "grad_norm": 0.1777653992176056,
+      "learning_rate": 1.1039511132570516e-05,
+      "loss": 0.4681,
+      "step": 6333
+    },
+    {
+      "epoch": 1.412893151907205,
+      "grad_norm": 0.17389416694641113,
+      "learning_rate": 1.1037170424513482e-05,
+      "loss": 0.4644,
+      "step": 6334
+    },
+    {
+      "epoch": 1.4131162168190943,
+      "grad_norm": 0.17686106264591217,
+      "learning_rate": 1.1034829659011426e-05,
+      "loss": 0.4946,
+      "step": 6335
+    },
+    {
+      "epoch": 1.4133392817309838,
+      "grad_norm": 0.17192257940769196,
+      "learning_rate": 1.1032488836193994e-05,
+      "loss": 0.4583,
+      "step": 6336
+    },
+    {
+      "epoch": 1.413562346642873,
+      "grad_norm": 0.17798608541488647,
+      "learning_rate": 1.1030147956190835e-05,
+      "loss": 0.453,
+      "step": 6337
+    },
+    {
+      "epoch": 1.4137854115547626,
+      "grad_norm": 0.18982993066310883,
+      "learning_rate": 1.1027807019131605e-05,
+      "loss": 0.448,
+      "step": 6338
+    },
+    {
+      "epoch": 1.4140084764666518,
+      "grad_norm": 0.17006950080394745,
+      "learning_rate": 1.1025466025145955e-05,
+      "loss": 0.4625,
+      "step": 6339
+    },
+    {
+      "epoch": 1.4142315413785411,
+      "grad_norm": 0.16911807656288147,
+      "learning_rate": 1.1023124974363546e-05,
+      "loss": 0.4514,
+      "step": 6340
+    },
+    {
+      "epoch": 1.4144546062904304,
+      "grad_norm": 0.16757598519325256,
+      "learning_rate": 1.1020783866914042e-05,
+      "loss": 0.4415,
+      "step": 6341
+    },
+    {
+      "epoch": 1.41467767120232,
+      "grad_norm": 0.18296495079994202,
+      "learning_rate": 1.1018442702927104e-05,
+      "loss": 0.4786,
+      "step": 6342
+    },
+    {
+      "epoch": 1.4149007361142092,
+      "grad_norm": 0.16517338156700134,
+      "learning_rate": 1.1016101482532404e-05,
+      "loss": 0.432,
+      "step": 6343
+    },
+    {
+      "epoch": 1.4151238010260987,
+      "grad_norm": 0.16575907170772552,
+      "learning_rate": 1.1013760205859611e-05,
+      "loss": 0.4495,
+      "step": 6344
+    },
+    {
+      "epoch": 1.415346865937988,
+      "grad_norm": 0.17554599046707153,
+      "learning_rate": 1.1011418873038404e-05,
+      "loss": 0.4354,
+      "step": 6345
+    },
+    {
+      "epoch": 1.4155699308498773,
+      "grad_norm": 0.17527008056640625,
+      "learning_rate": 1.1009077484198456e-05,
+      "loss": 0.4729,
+      "step": 6346
+    },
+    {
+      "epoch": 1.4157929957617665,
+      "grad_norm": 0.1661120504140854,
+      "learning_rate": 1.100673603946945e-05,
+      "loss": 0.456,
+      "step": 6347
+    },
+    {
+      "epoch": 1.416016060673656,
+      "grad_norm": 0.16820774972438812,
+      "learning_rate": 1.1004394538981069e-05,
+      "loss": 0.4675,
+      "step": 6348
+    },
+    {
+      "epoch": 1.4162391255855453,
+      "grad_norm": 0.16874277591705322,
+      "learning_rate": 1.1002052982863001e-05,
+      "loss": 0.4414,
+      "step": 6349
+    },
+    {
+      "epoch": 1.4164621904974348,
+      "grad_norm": 0.17940542101860046,
+      "learning_rate": 1.0999711371244936e-05,
+      "loss": 0.4615,
+      "step": 6350
+    },
+    {
+      "epoch": 1.416685255409324,
+      "grad_norm": 0.1683221310377121,
+      "learning_rate": 1.0997369704256566e-05,
+      "loss": 0.4633,
+      "step": 6351
+    },
+    {
+      "epoch": 1.4169083203212134,
+      "grad_norm": 0.17308497428894043,
+      "learning_rate": 1.0995027982027588e-05,
+      "loss": 0.4399,
+      "step": 6352
+    },
+    {
+      "epoch": 1.417131385233103,
+      "grad_norm": 0.17420804500579834,
+      "learning_rate": 1.0992686204687701e-05,
+      "loss": 0.4701,
+      "step": 6353
+    },
+    {
+      "epoch": 1.4173544501449922,
+      "grad_norm": 0.1695612072944641,
+      "learning_rate": 1.0990344372366611e-05,
+      "loss": 0.4307,
+      "step": 6354
+    },
+    {
+      "epoch": 1.4175775150568817,
+      "grad_norm": 0.17334359884262085,
+      "learning_rate": 1.0988002485194016e-05,
+      "loss": 0.4673,
+      "step": 6355
+    },
+    {
+      "epoch": 1.417800579968771,
+      "grad_norm": 0.1705588549375534,
+      "learning_rate": 1.0985660543299632e-05,
+      "loss": 0.4589,
+      "step": 6356
+    },
+    {
+      "epoch": 1.4180236448806602,
+      "grad_norm": 0.1716470867395401,
+      "learning_rate": 1.0983318546813164e-05,
+      "loss": 0.4645,
+      "step": 6357
+    },
+    {
+      "epoch": 1.4182467097925495,
+      "grad_norm": 0.17242403328418732,
+      "learning_rate": 1.098097649586433e-05,
+      "loss": 0.4401,
+      "step": 6358
+    },
+    {
+      "epoch": 1.418469774704439,
+      "grad_norm": 0.1679268628358841,
+      "learning_rate": 1.0978634390582847e-05,
+      "loss": 0.46,
+      "step": 6359
+    },
+    {
+      "epoch": 1.4186928396163283,
+      "grad_norm": 0.17091570794582367,
+      "learning_rate": 1.0976292231098435e-05,
+      "loss": 0.4361,
+      "step": 6360
+    },
+    {
+      "epoch": 1.4189159045282178,
+      "grad_norm": 0.17040546238422394,
+      "learning_rate": 1.0973950017540823e-05,
+      "loss": 0.4295,
+      "step": 6361
+    },
+    {
+      "epoch": 1.419138969440107,
+      "grad_norm": 0.17108914256095886,
+      "learning_rate": 1.0971607750039727e-05,
+      "loss": 0.4581,
+      "step": 6362
+    },
+    {
+      "epoch": 1.4193620343519964,
+      "grad_norm": 0.16708308458328247,
+      "learning_rate": 1.0969265428724887e-05,
+      "loss": 0.455,
+      "step": 6363
+    },
+    {
+      "epoch": 1.4195850992638857,
+      "grad_norm": 0.16952864825725555,
+      "learning_rate": 1.0966923053726025e-05,
+      "loss": 0.4411,
+      "step": 6364
+    },
+    {
+      "epoch": 1.4198081641757752,
+      "grad_norm": 0.16820164024829865,
+      "learning_rate": 1.0964580625172887e-05,
+      "loss": 0.4328,
+      "step": 6365
+    },
+    {
+      "epoch": 1.4200312290876644,
+      "grad_norm": 0.16284674406051636,
+      "learning_rate": 1.0962238143195203e-05,
+      "loss": 0.4227,
+      "step": 6366
+    },
+    {
+      "epoch": 1.420254293999554,
+      "grad_norm": 0.16655023396015167,
+      "learning_rate": 1.0959895607922722e-05,
+      "loss": 0.4568,
+      "step": 6367
+    },
+    {
+      "epoch": 1.4204773589114432,
+      "grad_norm": 0.17234466969966888,
+      "learning_rate": 1.095755301948518e-05,
+      "loss": 0.4756,
+      "step": 6368
+    },
+    {
+      "epoch": 1.4207004238233325,
+      "grad_norm": 0.1787555068731308,
+      "learning_rate": 1.0955210378012331e-05,
+      "loss": 0.4822,
+      "step": 6369
+    },
+    {
+      "epoch": 1.420923488735222,
+      "grad_norm": 0.1660604625940323,
+      "learning_rate": 1.0952867683633922e-05,
+      "loss": 0.4712,
+      "step": 6370
+    },
+    {
+      "epoch": 1.4211465536471113,
+      "grad_norm": 0.19277963042259216,
+      "learning_rate": 1.0950524936479708e-05,
+      "loss": 0.4813,
+      "step": 6371
+    },
+    {
+      "epoch": 1.4213696185590008,
+      "grad_norm": 0.16839838027954102,
+      "learning_rate": 1.0948182136679442e-05,
+      "loss": 0.411,
+      "step": 6372
+    },
+    {
+      "epoch": 1.42159268347089,
+      "grad_norm": 0.18013301491737366,
+      "learning_rate": 1.0945839284362885e-05,
+      "loss": 0.4256,
+      "step": 6373
+    },
+    {
+      "epoch": 1.4218157483827794,
+      "grad_norm": 0.16745387017726898,
+      "learning_rate": 1.09434963796598e-05,
+      "loss": 0.4408,
+      "step": 6374
+    },
+    {
+      "epoch": 1.4220388132946686,
+      "grad_norm": 0.16723443567752838,
+      "learning_rate": 1.094115342269995e-05,
+      "loss": 0.4702,
+      "step": 6375
+    },
+    {
+      "epoch": 1.4222618782065581,
+      "grad_norm": 0.17772428691387177,
+      "learning_rate": 1.0938810413613103e-05,
+      "loss": 0.4401,
+      "step": 6376
+    },
+    {
+      "epoch": 1.4224849431184474,
+      "grad_norm": 0.17869889736175537,
+      "learning_rate": 1.0936467352529032e-05,
+      "loss": 0.4536,
+      "step": 6377
+    },
+    {
+      "epoch": 1.422708008030337,
+      "grad_norm": 0.18365752696990967,
+      "learning_rate": 1.0934124239577506e-05,
+      "loss": 0.4603,
+      "step": 6378
+    },
+    {
+      "epoch": 1.4229310729422262,
+      "grad_norm": 0.18287013471126556,
+      "learning_rate": 1.0931781074888306e-05,
+      "loss": 0.4501,
+      "step": 6379
+    },
+    {
+      "epoch": 1.4231541378541155,
+      "grad_norm": 0.21559575200080872,
+      "learning_rate": 1.0929437858591207e-05,
+      "loss": 0.4447,
+      "step": 6380
+    },
+    {
+      "epoch": 1.423377202766005,
+      "grad_norm": 0.189785435795784,
+      "learning_rate": 1.0927094590815992e-05,
+      "loss": 0.4608,
+      "step": 6381
+    },
+    {
+      "epoch": 1.4236002676778943,
+      "grad_norm": 0.17252789437770844,
+      "learning_rate": 1.092475127169245e-05,
+      "loss": 0.4806,
+      "step": 6382
+    },
+    {
+      "epoch": 1.4238233325897836,
+      "grad_norm": 0.16337281465530396,
+      "learning_rate": 1.0922407901350365e-05,
+      "loss": 0.4523,
+      "step": 6383
+    },
+    {
+      "epoch": 1.424046397501673,
+      "grad_norm": 0.17701326310634613,
+      "learning_rate": 1.0920064479919527e-05,
+      "loss": 0.4629,
+      "step": 6384
+    },
+    {
+      "epoch": 1.4242694624135623,
+      "grad_norm": 0.17202292382717133,
+      "learning_rate": 1.0917721007529731e-05,
+      "loss": 0.4646,
+      "step": 6385
+    },
+    {
+      "epoch": 1.4244925273254516,
+      "grad_norm": 0.18118809163570404,
+      "learning_rate": 1.0915377484310774e-05,
+      "loss": 0.4814,
+      "step": 6386
+    },
+    {
+      "epoch": 1.4247155922373411,
+      "grad_norm": 0.1883123368024826,
+      "learning_rate": 1.0913033910392452e-05,
+      "loss": 0.4767,
+      "step": 6387
+    },
+    {
+      "epoch": 1.4249386571492304,
+      "grad_norm": 0.16664288938045502,
+      "learning_rate": 1.0910690285904573e-05,
+      "loss": 0.429,
+      "step": 6388
+    },
+    {
+      "epoch": 1.42516172206112,
+      "grad_norm": 0.17694927752017975,
+      "learning_rate": 1.0908346610976934e-05,
+      "loss": 0.4506,
+      "step": 6389
+    },
+    {
+      "epoch": 1.4253847869730092,
+      "grad_norm": 0.17796604335308075,
+      "learning_rate": 1.0906002885739348e-05,
+      "loss": 0.4564,
+      "step": 6390
+    },
+    {
+      "epoch": 1.4256078518848985,
+      "grad_norm": 0.1945108026266098,
+      "learning_rate": 1.090365911032162e-05,
+      "loss": 0.4479,
+      "step": 6391
+    },
+    {
+      "epoch": 1.4258309167967878,
+      "grad_norm": 0.1725914031267166,
+      "learning_rate": 1.0901315284853566e-05,
+      "loss": 0.4606,
+      "step": 6392
+    },
+    {
+      "epoch": 1.4260539817086773,
+      "grad_norm": 0.16722865402698517,
+      "learning_rate": 1.0898971409465006e-05,
+      "loss": 0.4595,
+      "step": 6393
+    },
+    {
+      "epoch": 1.4262770466205665,
+      "grad_norm": 0.1750846952199936,
+      "learning_rate": 1.0896627484285752e-05,
+      "loss": 0.4485,
+      "step": 6394
+    },
+    {
+      "epoch": 1.426500111532456,
+      "grad_norm": 0.1740426868200302,
+      "learning_rate": 1.0894283509445629e-05,
+      "loss": 0.4627,
+      "step": 6395
+    },
+    {
+      "epoch": 1.4267231764443453,
+      "grad_norm": 0.17253148555755615,
+      "learning_rate": 1.0891939485074459e-05,
+      "loss": 0.4481,
+      "step": 6396
+    },
+    {
+      "epoch": 1.4269462413562346,
+      "grad_norm": 0.17242968082427979,
+      "learning_rate": 1.088959541130207e-05,
+      "loss": 0.4573,
+      "step": 6397
+    },
+    {
+      "epoch": 1.427169306268124,
+      "grad_norm": 0.1693577617406845,
+      "learning_rate": 1.0887251288258291e-05,
+      "loss": 0.461,
+      "step": 6398
+    },
+    {
+      "epoch": 1.4273923711800134,
+      "grad_norm": 0.17257948219776154,
+      "learning_rate": 1.0884907116072956e-05,
+      "loss": 0.4735,
+      "step": 6399
+    },
+    {
+      "epoch": 1.4276154360919027,
+      "grad_norm": 0.15882723033428192,
+      "learning_rate": 1.0882562894875897e-05,
+      "loss": 0.4495,
+      "step": 6400
+    },
+    {
+      "epoch": 1.4278385010037922,
+      "grad_norm": 0.16930413246154785,
+      "learning_rate": 1.0880218624796954e-05,
+      "loss": 0.4321,
+      "step": 6401
+    },
+    {
+      "epoch": 1.4280615659156815,
+      "grad_norm": 0.1744142770767212,
+      "learning_rate": 1.0877874305965968e-05,
+      "loss": 0.4472,
+      "step": 6402
+    },
+    {
+      "epoch": 1.4282846308275707,
+      "grad_norm": 0.1769794076681137,
+      "learning_rate": 1.0875529938512779e-05,
+      "loss": 0.475,
+      "step": 6403
+    },
+    {
+      "epoch": 1.4285076957394602,
+      "grad_norm": 0.1765398234128952,
+      "learning_rate": 1.0873185522567236e-05,
+      "loss": 0.4595,
+      "step": 6404
+    },
+    {
+      "epoch": 1.4287307606513495,
+      "grad_norm": 0.1690848469734192,
+      "learning_rate": 1.0870841058259185e-05,
+      "loss": 0.4647,
+      "step": 6405
+    },
+    {
+      "epoch": 1.428953825563239,
+      "grad_norm": 0.17777849733829498,
+      "learning_rate": 1.086849654571848e-05,
+      "loss": 0.4644,
+      "step": 6406
+    },
+    {
+      "epoch": 1.4291768904751283,
+      "grad_norm": 0.17765332758426666,
+      "learning_rate": 1.0866151985074973e-05,
+      "loss": 0.4528,
+      "step": 6407
+    },
+    {
+      "epoch": 1.4293999553870176,
+      "grad_norm": 0.1740536242723465,
+      "learning_rate": 1.0863807376458516e-05,
+      "loss": 0.4675,
+      "step": 6408
+    },
+    {
+      "epoch": 1.4296230202989069,
+      "grad_norm": 0.16334398090839386,
+      "learning_rate": 1.0861462719998981e-05,
+      "loss": 0.4409,
+      "step": 6409
+    },
+    {
+      "epoch": 1.4298460852107964,
+      "grad_norm": 0.17333824932575226,
+      "learning_rate": 1.0859118015826216e-05,
+      "loss": 0.4712,
+      "step": 6410
+    },
+    {
+      "epoch": 1.4300691501226857,
+      "grad_norm": 0.17034272849559784,
+      "learning_rate": 1.0856773264070092e-05,
+      "loss": 0.4635,
+      "step": 6411
+    },
+    {
+      "epoch": 1.4302922150345752,
+      "grad_norm": 0.16784878075122833,
+      "learning_rate": 1.0854428464860476e-05,
+      "loss": 0.4438,
+      "step": 6412
+    },
+    {
+      "epoch": 1.4305152799464644,
+      "grad_norm": 0.17363719642162323,
+      "learning_rate": 1.0852083618327239e-05,
+      "loss": 0.4789,
+      "step": 6413
+    },
+    {
+      "epoch": 1.4307383448583537,
+      "grad_norm": 0.16398392617702484,
+      "learning_rate": 1.084973872460025e-05,
+      "loss": 0.4559,
+      "step": 6414
+    },
+    {
+      "epoch": 1.4309614097702432,
+      "grad_norm": 0.17007873952388763,
+      "learning_rate": 1.0847393783809383e-05,
+      "loss": 0.468,
+      "step": 6415
+    },
+    {
+      "epoch": 1.4311844746821325,
+      "grad_norm": 0.1761980801820755,
+      "learning_rate": 1.084504879608452e-05,
+      "loss": 0.4525,
+      "step": 6416
+    },
+    {
+      "epoch": 1.4314075395940218,
+      "grad_norm": 0.17042046785354614,
+      "learning_rate": 1.084270376155554e-05,
+      "loss": 0.4498,
+      "step": 6417
+    },
+    {
+      "epoch": 1.4316306045059113,
+      "grad_norm": 0.1663273125886917,
+      "learning_rate": 1.0840358680352324e-05,
+      "loss": 0.4453,
+      "step": 6418
+    },
+    {
+      "epoch": 1.4318536694178006,
+      "grad_norm": 0.1743524819612503,
+      "learning_rate": 1.0838013552604758e-05,
+      "loss": 0.481,
+      "step": 6419
+    },
+    {
+      "epoch": 1.4320767343296898,
+      "grad_norm": 0.16921429336071014,
+      "learning_rate": 1.083566837844273e-05,
+      "loss": 0.4284,
+      "step": 6420
+    },
+    {
+      "epoch": 1.4322997992415794,
+      "grad_norm": 0.1793074607849121,
+      "learning_rate": 1.083332315799613e-05,
+      "loss": 0.4842,
+      "step": 6421
+    },
+    {
+      "epoch": 1.4325228641534686,
+      "grad_norm": 0.16839145123958588,
+      "learning_rate": 1.0830977891394853e-05,
+      "loss": 0.4509,
+      "step": 6422
+    },
+    {
+      "epoch": 1.4327459290653581,
+      "grad_norm": 0.1707042008638382,
+      "learning_rate": 1.0828632578768794e-05,
+      "loss": 0.446,
+      "step": 6423
+    },
+    {
+      "epoch": 1.4329689939772474,
+      "grad_norm": 0.16985002160072327,
+      "learning_rate": 1.0826287220247851e-05,
+      "loss": 0.4856,
+      "step": 6424
+    },
+    {
+      "epoch": 1.4331920588891367,
+      "grad_norm": 0.17573486268520355,
+      "learning_rate": 1.082394181596192e-05,
+      "loss": 0.4552,
+      "step": 6425
+    },
+    {
+      "epoch": 1.433415123801026,
+      "grad_norm": 0.20836280286312103,
+      "learning_rate": 1.0821596366040911e-05,
+      "loss": 0.4532,
+      "step": 6426
+    },
+    {
+      "epoch": 1.4336381887129155,
+      "grad_norm": 0.17805512249469757,
+      "learning_rate": 1.0819250870614729e-05,
+      "loss": 0.462,
+      "step": 6427
+    },
+    {
+      "epoch": 1.4338612536248048,
+      "grad_norm": 0.17473873496055603,
+      "learning_rate": 1.081690532981328e-05,
+      "loss": 0.444,
+      "step": 6428
+    },
+    {
+      "epoch": 1.4340843185366943,
+      "grad_norm": 0.17013612389564514,
+      "learning_rate": 1.081455974376647e-05,
+      "loss": 0.4578,
+      "step": 6429
+    },
+    {
+      "epoch": 1.4343073834485836,
+      "grad_norm": 0.16665178537368774,
+      "learning_rate": 1.0812214112604224e-05,
+      "loss": 0.432,
+      "step": 6430
+    },
+    {
+      "epoch": 1.4345304483604728,
+      "grad_norm": 0.1708722859621048,
+      "learning_rate": 1.080986843645645e-05,
+      "loss": 0.4473,
+      "step": 6431
+    },
+    {
+      "epoch": 1.4347535132723623,
+      "grad_norm": 0.17868392169475555,
+      "learning_rate": 1.0807522715453067e-05,
+      "loss": 0.4696,
+      "step": 6432
+    },
+    {
+      "epoch": 1.4349765781842516,
+      "grad_norm": 0.17720571160316467,
+      "learning_rate": 1.0805176949723997e-05,
+      "loss": 0.4708,
+      "step": 6433
+    },
+    {
+      "epoch": 1.435199643096141,
+      "grad_norm": 0.1707199364900589,
+      "learning_rate": 1.080283113939916e-05,
+      "loss": 0.4668,
+      "step": 6434
+    },
+    {
+      "epoch": 1.4354227080080304,
+      "grad_norm": 0.1737671196460724,
+      "learning_rate": 1.0800485284608488e-05,
+      "loss": 0.4539,
+      "step": 6435
+    },
+    {
+      "epoch": 1.4356457729199197,
+      "grad_norm": 0.1636429876089096,
+      "learning_rate": 1.0798139385481903e-05,
+      "loss": 0.4321,
+      "step": 6436
+    },
+    {
+      "epoch": 1.435868837831809,
+      "grad_norm": 0.1818714439868927,
+      "learning_rate": 1.079579344214934e-05,
+      "loss": 0.4698,
+      "step": 6437
+    },
+    {
+      "epoch": 1.4360919027436985,
+      "grad_norm": 0.17777089774608612,
+      "learning_rate": 1.0793447454740731e-05,
+      "loss": 0.4674,
+      "step": 6438
+    },
+    {
+      "epoch": 1.4363149676555877,
+      "grad_norm": 0.16908809542655945,
+      "learning_rate": 1.079110142338601e-05,
+      "loss": 0.4584,
+      "step": 6439
+    },
+    {
+      "epoch": 1.4365380325674773,
+      "grad_norm": 0.1798996478319168,
+      "learning_rate": 1.0788755348215114e-05,
+      "loss": 0.4565,
+      "step": 6440
+    },
+    {
+      "epoch": 1.4367610974793665,
+      "grad_norm": 0.1786443144083023,
+      "learning_rate": 1.0786409229357991e-05,
+      "loss": 0.4672,
+      "step": 6441
+    },
+    {
+      "epoch": 1.4369841623912558,
+      "grad_norm": 0.1697261929512024,
+      "learning_rate": 1.0784063066944572e-05,
+      "loss": 0.4342,
+      "step": 6442
+    },
+    {
+      "epoch": 1.437207227303145,
+      "grad_norm": 0.1702648103237152,
+      "learning_rate": 1.0781716861104812e-05,
+      "loss": 0.4487,
+      "step": 6443
+    },
+    {
+      "epoch": 1.4374302922150346,
+      "grad_norm": 0.17532885074615479,
+      "learning_rate": 1.0779370611968652e-05,
+      "loss": 0.4431,
+      "step": 6444
+    },
+    {
+      "epoch": 1.4376533571269239,
+      "grad_norm": 0.17340442538261414,
+      "learning_rate": 1.0777024319666048e-05,
+      "loss": 0.4404,
+      "step": 6445
+    },
+    {
+      "epoch": 1.4378764220388134,
+      "grad_norm": 0.1587846279144287,
+      "learning_rate": 1.0774677984326946e-05,
+      "loss": 0.4275,
+      "step": 6446
+    },
+    {
+      "epoch": 1.4380994869507027,
+      "grad_norm": 0.1758798211812973,
+      "learning_rate": 1.0772331606081308e-05,
+      "loss": 0.462,
+      "step": 6447
+    },
+    {
+      "epoch": 1.438322551862592,
+      "grad_norm": 0.16627365350723267,
+      "learning_rate": 1.0769985185059087e-05,
+      "loss": 0.4244,
+      "step": 6448
+    },
+    {
+      "epoch": 1.4385456167744815,
+      "grad_norm": 0.17225994169712067,
+      "learning_rate": 1.0767638721390242e-05,
+      "loss": 0.4572,
+      "step": 6449
+    },
+    {
+      "epoch": 1.4387686816863707,
+      "grad_norm": 0.1741783767938614,
+      "learning_rate": 1.0765292215204738e-05,
+      "loss": 0.45,
+      "step": 6450
+    },
+    {
+      "epoch": 1.43899174659826,
+      "grad_norm": 0.16373707354068756,
+      "learning_rate": 1.0762945666632534e-05,
+      "loss": 0.4508,
+      "step": 6451
+    },
+    {
+      "epoch": 1.4392148115101495,
+      "grad_norm": 0.18814869225025177,
+      "learning_rate": 1.0760599075803601e-05,
+      "loss": 0.5056,
+      "step": 6452
+    },
+    {
+      "epoch": 1.4394378764220388,
+      "grad_norm": 0.17694085836410522,
+      "learning_rate": 1.0758252442847907e-05,
+      "loss": 0.4423,
+      "step": 6453
+    },
+    {
+      "epoch": 1.439660941333928,
+      "grad_norm": 0.17024989426136017,
+      "learning_rate": 1.0755905767895425e-05,
+      "loss": 0.4288,
+      "step": 6454
+    },
+    {
+      "epoch": 1.4398840062458176,
+      "grad_norm": 0.1731920838356018,
+      "learning_rate": 1.0753559051076123e-05,
+      "loss": 0.4909,
+      "step": 6455
+    },
+    {
+      "epoch": 1.4401070711577069,
+      "grad_norm": 0.17619867622852325,
+      "learning_rate": 1.0751212292519983e-05,
+      "loss": 0.4567,
+      "step": 6456
+    },
+    {
+      "epoch": 1.4403301360695964,
+      "grad_norm": 0.512635350227356,
+      "learning_rate": 1.0748865492356981e-05,
+      "loss": 0.4643,
+      "step": 6457
+    },
+    {
+      "epoch": 1.4405532009814856,
+      "grad_norm": 0.17129187285900116,
+      "learning_rate": 1.0746518650717097e-05,
+      "loss": 0.4453,
+      "step": 6458
+    },
+    {
+      "epoch": 1.440776265893375,
+      "grad_norm": 0.16431821882724762,
+      "learning_rate": 1.0744171767730315e-05,
+      "loss": 0.4289,
+      "step": 6459
+    },
+    {
+      "epoch": 1.4409993308052642,
+      "grad_norm": 0.18094895780086517,
+      "learning_rate": 1.0741824843526619e-05,
+      "loss": 0.4711,
+      "step": 6460
+    },
+    {
+      "epoch": 1.4412223957171537,
+      "grad_norm": 0.19894392788410187,
+      "learning_rate": 1.0739477878235996e-05,
+      "loss": 0.4643,
+      "step": 6461
+    },
+    {
+      "epoch": 1.441445460629043,
+      "grad_norm": 0.17579229176044464,
+      "learning_rate": 1.073713087198844e-05,
+      "loss": 0.4529,
+      "step": 6462
+    },
+    {
+      "epoch": 1.4416685255409325,
+      "grad_norm": 0.1695978045463562,
+      "learning_rate": 1.0734783824913935e-05,
+      "loss": 0.4772,
+      "step": 6463
+    },
+    {
+      "epoch": 1.4418915904528218,
+      "grad_norm": 0.17056889832019806,
+      "learning_rate": 1.0732436737142482e-05,
+      "loss": 0.4648,
+      "step": 6464
+    },
+    {
+      "epoch": 1.442114655364711,
+      "grad_norm": 0.18249176442623138,
+      "learning_rate": 1.0730089608804074e-05,
+      "loss": 0.4555,
+      "step": 6465
+    },
+    {
+      "epoch": 1.4423377202766006,
+      "grad_norm": 0.2158237099647522,
+      "learning_rate": 1.0727742440028712e-05,
+      "loss": 0.4276,
+      "step": 6466
+    },
+    {
+      "epoch": 1.4425607851884898,
+      "grad_norm": 0.17240867018699646,
+      "learning_rate": 1.0725395230946396e-05,
+      "loss": 0.4629,
+      "step": 6467
+    },
+    {
+      "epoch": 1.4427838501003791,
+      "grad_norm": 0.16894541680812836,
+      "learning_rate": 1.072304798168713e-05,
+      "loss": 0.4522,
+      "step": 6468
+    },
+    {
+      "epoch": 1.4430069150122686,
+      "grad_norm": 0.17741340398788452,
+      "learning_rate": 1.0720700692380918e-05,
+      "loss": 0.4489,
+      "step": 6469
+    },
+    {
+      "epoch": 1.443229979924158,
+      "grad_norm": 0.16370341181755066,
+      "learning_rate": 1.0718353363157767e-05,
+      "loss": 0.4731,
+      "step": 6470
+    },
+    {
+      "epoch": 1.4434530448360472,
+      "grad_norm": 0.1746392399072647,
+      "learning_rate": 1.0716005994147694e-05,
+      "loss": 0.4832,
+      "step": 6471
+    },
+    {
+      "epoch": 1.4436761097479367,
+      "grad_norm": 0.16757871210575104,
+      "learning_rate": 1.0713658585480697e-05,
+      "loss": 0.4523,
+      "step": 6472
+    },
+    {
+      "epoch": 1.443899174659826,
+      "grad_norm": 0.18170872330665588,
+      "learning_rate": 1.0711311137286804e-05,
+      "loss": 0.4357,
+      "step": 6473
+    },
+    {
+      "epoch": 1.4441222395717155,
+      "grad_norm": 0.17265670001506805,
+      "learning_rate": 1.0708963649696023e-05,
+      "loss": 0.4685,
+      "step": 6474
+    },
+    {
+      "epoch": 1.4443453044836048,
+      "grad_norm": 0.15775421261787415,
+      "learning_rate": 1.0706616122838379e-05,
+      "loss": 0.427,
+      "step": 6475
+    },
+    {
+      "epoch": 1.444568369395494,
+      "grad_norm": 0.18243922293186188,
+      "learning_rate": 1.0704268556843884e-05,
+      "loss": 0.4614,
+      "step": 6476
+    },
+    {
+      "epoch": 1.4447914343073833,
+      "grad_norm": 0.17497234046459198,
+      "learning_rate": 1.0701920951842568e-05,
+      "loss": 0.4814,
+      "step": 6477
+    },
+    {
+      "epoch": 1.4450144992192728,
+      "grad_norm": 0.1762491762638092,
+      "learning_rate": 1.0699573307964457e-05,
+      "loss": 0.458,
+      "step": 6478
+    },
+    {
+      "epoch": 1.4452375641311621,
+      "grad_norm": 0.1725545972585678,
+      "learning_rate": 1.0697225625339573e-05,
+      "loss": 0.4371,
+      "step": 6479
+    },
+    {
+      "epoch": 1.4454606290430516,
+      "grad_norm": 0.9475487470626831,
+      "learning_rate": 1.0694877904097952e-05,
+      "loss": 0.4609,
+      "step": 6480
+    },
+    {
+      "epoch": 1.445683693954941,
+      "grad_norm": 0.21081972122192383,
+      "learning_rate": 1.0692530144369615e-05,
+      "loss": 0.4515,
+      "step": 6481
+    },
+    {
+      "epoch": 1.4459067588668302,
+      "grad_norm": 0.1702289879322052,
+      "learning_rate": 1.0690182346284608e-05,
+      "loss": 0.4586,
+      "step": 6482
+    },
+    {
+      "epoch": 1.4461298237787197,
+      "grad_norm": 0.18808738887310028,
+      "learning_rate": 1.0687834509972958e-05,
+      "loss": 0.4688,
+      "step": 6483
+    },
+    {
+      "epoch": 1.446352888690609,
+      "grad_norm": 0.17759791016578674,
+      "learning_rate": 1.068548663556471e-05,
+      "loss": 0.4544,
+      "step": 6484
+    },
+    {
+      "epoch": 1.4465759536024985,
+      "grad_norm": 0.1792079657316208,
+      "learning_rate": 1.0683138723189897e-05,
+      "loss": 0.4475,
+      "step": 6485
+    },
+    {
+      "epoch": 1.4467990185143877,
+      "grad_norm": 0.17630550265312195,
+      "learning_rate": 1.0680790772978566e-05,
+      "loss": 0.4504,
+      "step": 6486
+    },
+    {
+      "epoch": 1.447022083426277,
+      "grad_norm": 0.18616117537021637,
+      "learning_rate": 1.0678442785060758e-05,
+      "loss": 0.4455,
+      "step": 6487
+    },
+    {
+      "epoch": 1.4472451483381663,
+      "grad_norm": 0.18077711760997772,
+      "learning_rate": 1.0676094759566524e-05,
+      "loss": 0.4937,
+      "step": 6488
+    },
+    {
+      "epoch": 1.4474682132500558,
+      "grad_norm": 0.1756681501865387,
+      "learning_rate": 1.0673746696625906e-05,
+      "loss": 0.4585,
+      "step": 6489
+    },
+    {
+      "epoch": 1.447691278161945,
+      "grad_norm": 0.17320303618907928,
+      "learning_rate": 1.0671398596368961e-05,
+      "loss": 0.4564,
+      "step": 6490
+    },
+    {
+      "epoch": 1.4479143430738346,
+      "grad_norm": 0.17476066946983337,
+      "learning_rate": 1.0669050458925736e-05,
+      "loss": 0.4621,
+      "step": 6491
+    },
+    {
+      "epoch": 1.4481374079857239,
+      "grad_norm": 0.1720237284898758,
+      "learning_rate": 1.0666702284426289e-05,
+      "loss": 0.4483,
+      "step": 6492
+    },
+    {
+      "epoch": 1.4483604728976132,
+      "grad_norm": 0.1799459308385849,
+      "learning_rate": 1.0664354073000676e-05,
+      "loss": 0.4825,
+      "step": 6493
+    },
+    {
+      "epoch": 1.4485835378095024,
+      "grad_norm": 0.173868328332901,
+      "learning_rate": 1.0662005824778957e-05,
+      "loss": 0.4586,
+      "step": 6494
+    },
+    {
+      "epoch": 1.448806602721392,
+      "grad_norm": 0.17507266998291016,
+      "learning_rate": 1.0659657539891189e-05,
+      "loss": 0.453,
+      "step": 6495
+    },
+    {
+      "epoch": 1.4490296676332812,
+      "grad_norm": 0.17579588294029236,
+      "learning_rate": 1.0657309218467437e-05,
+      "loss": 0.4606,
+      "step": 6496
+    },
+    {
+      "epoch": 1.4492527325451707,
+      "grad_norm": 0.18230777978897095,
+      "learning_rate": 1.0654960860637766e-05,
+      "loss": 0.4721,
+      "step": 6497
+    },
+    {
+      "epoch": 1.44947579745706,
+      "grad_norm": 0.1772591918706894,
+      "learning_rate": 1.0652612466532242e-05,
+      "loss": 0.4642,
+      "step": 6498
+    },
+    {
+      "epoch": 1.4496988623689493,
+      "grad_norm": 0.17671100795269012,
+      "learning_rate": 1.0650264036280935e-05,
+      "loss": 0.4526,
+      "step": 6499
+    },
+    {
+      "epoch": 1.4499219272808388,
+      "grad_norm": 0.1718316227197647,
+      "learning_rate": 1.0647915570013916e-05,
+      "loss": 0.4592,
+      "step": 6500
+    },
+    {
+      "epoch": 1.450144992192728,
+      "grad_norm": 0.17270374298095703,
+      "learning_rate": 1.0645567067861257e-05,
+      "loss": 0.4722,
+      "step": 6501
+    },
+    {
+      "epoch": 1.4503680571046176,
+      "grad_norm": 0.1727593094110489,
+      "learning_rate": 1.0643218529953032e-05,
+      "loss": 0.4519,
+      "step": 6502
+    },
+    {
+      "epoch": 1.4505911220165069,
+      "grad_norm": 0.18215766549110413,
+      "learning_rate": 1.0640869956419321e-05,
+      "loss": 0.4584,
+      "step": 6503
+    },
+    {
+      "epoch": 1.4508141869283961,
+      "grad_norm": 0.17123596370220184,
+      "learning_rate": 1.0638521347390198e-05,
+      "loss": 0.4829,
+      "step": 6504
+    },
+    {
+      "epoch": 1.4510372518402854,
+      "grad_norm": 0.17143850028514862,
+      "learning_rate": 1.063617270299575e-05,
+      "loss": 0.4703,
+      "step": 6505
+    },
+    {
+      "epoch": 1.451260316752175,
+      "grad_norm": 0.17444929480552673,
+      "learning_rate": 1.0633824023366053e-05,
+      "loss": 0.4424,
+      "step": 6506
+    },
+    {
+      "epoch": 1.4514833816640642,
+      "grad_norm": 0.17457221448421478,
+      "learning_rate": 1.0631475308631196e-05,
+      "loss": 0.4759,
+      "step": 6507
+    },
+    {
+      "epoch": 1.4517064465759537,
+      "grad_norm": 0.16844333708286285,
+      "learning_rate": 1.0629126558921264e-05,
+      "loss": 0.4622,
+      "step": 6508
+    },
+    {
+      "epoch": 1.451929511487843,
+      "grad_norm": 0.17886194586753845,
+      "learning_rate": 1.0626777774366347e-05,
+      "loss": 0.4502,
+      "step": 6509
+    },
+    {
+      "epoch": 1.4521525763997323,
+      "grad_norm": 0.17846544086933136,
+      "learning_rate": 1.062442895509653e-05,
+      "loss": 0.4472,
+      "step": 6510
+    },
+    {
+      "epoch": 1.4523756413116216,
+      "grad_norm": 0.19240468740463257,
+      "learning_rate": 1.0622080101241914e-05,
+      "loss": 0.4895,
+      "step": 6511
+    },
+    {
+      "epoch": 1.452598706223511,
+      "grad_norm": 0.17172129452228546,
+      "learning_rate": 1.061973121293259e-05,
+      "loss": 0.4443,
+      "step": 6512
+    },
+    {
+      "epoch": 1.4528217711354003,
+      "grad_norm": 0.16445989906787872,
+      "learning_rate": 1.0617382290298649e-05,
+      "loss": 0.4089,
+      "step": 6513
+    },
+    {
+      "epoch": 1.4530448360472898,
+      "grad_norm": 0.1770135909318924,
+      "learning_rate": 1.0615033333470194e-05,
+      "loss": 0.4726,
+      "step": 6514
+    },
+    {
+      "epoch": 1.4532679009591791,
+      "grad_norm": 0.16954918205738068,
+      "learning_rate": 1.0612684342577326e-05,
+      "loss": 0.4364,
+      "step": 6515
+    },
+    {
+      "epoch": 1.4534909658710684,
+      "grad_norm": 0.16754356026649475,
+      "learning_rate": 1.0610335317750144e-05,
+      "loss": 0.433,
+      "step": 6516
+    },
+    {
+      "epoch": 1.453714030782958,
+      "grad_norm": 0.17707470059394836,
+      "learning_rate": 1.0607986259118752e-05,
+      "loss": 0.4817,
+      "step": 6517
+    },
+    {
+      "epoch": 1.4539370956948472,
+      "grad_norm": 0.16972193121910095,
+      "learning_rate": 1.0605637166813261e-05,
+      "loss": 0.4376,
+      "step": 6518
+    },
+    {
+      "epoch": 1.4541601606067367,
+      "grad_norm": 0.1755881905555725,
+      "learning_rate": 1.0603288040963768e-05,
+      "loss": 0.4354,
+      "step": 6519
+    },
+    {
+      "epoch": 1.454383225518626,
+      "grad_norm": 0.17727094888687134,
+      "learning_rate": 1.0600938881700394e-05,
+      "loss": 0.4519,
+      "step": 6520
+    },
+    {
+      "epoch": 1.4546062904305153,
+      "grad_norm": 0.17010506987571716,
+      "learning_rate": 1.059858968915324e-05,
+      "loss": 0.4579,
+      "step": 6521
+    },
+    {
+      "epoch": 1.4548293553424045,
+      "grad_norm": 0.1777375340461731,
+      "learning_rate": 1.0596240463452427e-05,
+      "loss": 0.4646,
+      "step": 6522
+    },
+    {
+      "epoch": 1.455052420254294,
+      "grad_norm": 0.17490847408771515,
+      "learning_rate": 1.0593891204728064e-05,
+      "loss": 0.4906,
+      "step": 6523
+    },
+    {
+      "epoch": 1.4552754851661833,
+      "grad_norm": 0.16298337280750275,
+      "learning_rate": 1.0591541913110273e-05,
+      "loss": 0.4104,
+      "step": 6524
+    },
+    {
+      "epoch": 1.4554985500780728,
+      "grad_norm": 0.1771763265132904,
+      "learning_rate": 1.0589192588729167e-05,
+      "loss": 0.4829,
+      "step": 6525
+    },
+    {
+      "epoch": 1.455721614989962,
+      "grad_norm": 0.16941291093826294,
+      "learning_rate": 1.0586843231714874e-05,
+      "loss": 0.4469,
+      "step": 6526
+    },
+    {
+      "epoch": 1.4559446799018514,
+      "grad_norm": 0.16429558396339417,
+      "learning_rate": 1.0584493842197505e-05,
+      "loss": 0.4358,
+      "step": 6527
+    },
+    {
+      "epoch": 1.4561677448137407,
+      "grad_norm": 0.18524731695652008,
+      "learning_rate": 1.0582144420307196e-05,
+      "loss": 0.4476,
+      "step": 6528
+    },
+    {
+      "epoch": 1.4563908097256302,
+      "grad_norm": 0.18424685299396515,
+      "learning_rate": 1.0579794966174064e-05,
+      "loss": 0.4621,
+      "step": 6529
+    },
+    {
+      "epoch": 1.4566138746375195,
+      "grad_norm": 0.17425791919231415,
+      "learning_rate": 1.0577445479928239e-05,
+      "loss": 0.4229,
+      "step": 6530
+    },
+    {
+      "epoch": 1.456836939549409,
+      "grad_norm": 0.1773643046617508,
+      "learning_rate": 1.0575095961699856e-05,
+      "loss": 0.4825,
+      "step": 6531
+    },
+    {
+      "epoch": 1.4570600044612982,
+      "grad_norm": 0.1699049323797226,
+      "learning_rate": 1.0572746411619034e-05,
+      "loss": 0.448,
+      "step": 6532
+    },
+    {
+      "epoch": 1.4572830693731875,
+      "grad_norm": 0.1901503950357437,
+      "learning_rate": 1.0570396829815919e-05,
+      "loss": 0.4498,
+      "step": 6533
+    },
+    {
+      "epoch": 1.457506134285077,
+      "grad_norm": 0.1757887601852417,
+      "learning_rate": 1.0568047216420636e-05,
+      "loss": 0.4576,
+      "step": 6534
+    },
+    {
+      "epoch": 1.4577291991969663,
+      "grad_norm": 0.17276814579963684,
+      "learning_rate": 1.056569757156333e-05,
+      "loss": 0.4685,
+      "step": 6535
+    },
+    {
+      "epoch": 1.4579522641088558,
+      "grad_norm": 0.1706877201795578,
+      "learning_rate": 1.0563347895374127e-05,
+      "loss": 0.4404,
+      "step": 6536
+    },
+    {
+      "epoch": 1.458175329020745,
+      "grad_norm": 0.1756119430065155,
+      "learning_rate": 1.056099818798318e-05,
+      "loss": 0.485,
+      "step": 6537
+    },
+    {
+      "epoch": 1.4583983939326344,
+      "grad_norm": 0.17187528312206268,
+      "learning_rate": 1.0558648449520621e-05,
+      "loss": 0.4528,
+      "step": 6538
+    },
+    {
+      "epoch": 1.4586214588445237,
+      "grad_norm": 0.17733196914196014,
+      "learning_rate": 1.05562986801166e-05,
+      "loss": 0.4607,
+      "step": 6539
+    },
+    {
+      "epoch": 1.4588445237564132,
+      "grad_norm": 0.1726623773574829,
+      "learning_rate": 1.0553948879901255e-05,
+      "loss": 0.4593,
+      "step": 6540
+    },
+    {
+      "epoch": 1.4590675886683024,
+      "grad_norm": 0.1699986606836319,
+      "learning_rate": 1.0551599049004738e-05,
+      "loss": 0.4614,
+      "step": 6541
+    },
+    {
+      "epoch": 1.459290653580192,
+      "grad_norm": 0.17713840305805206,
+      "learning_rate": 1.0549249187557196e-05,
+      "loss": 0.4911,
+      "step": 6542
+    },
+    {
+      "epoch": 1.4595137184920812,
+      "grad_norm": 0.17454893887043,
+      "learning_rate": 1.054689929568878e-05,
+      "loss": 0.469,
+      "step": 6543
+    },
+    {
+      "epoch": 1.4597367834039705,
+      "grad_norm": 0.17392531037330627,
+      "learning_rate": 1.054454937352964e-05,
+      "loss": 0.4355,
+      "step": 6544
+    },
+    {
+      "epoch": 1.4599598483158598,
+      "grad_norm": 0.16687026619911194,
+      "learning_rate": 1.0542199421209931e-05,
+      "loss": 0.456,
+      "step": 6545
+    },
+    {
+      "epoch": 1.4601829132277493,
+      "grad_norm": 0.18140020966529846,
+      "learning_rate": 1.0539849438859806e-05,
+      "loss": 0.4495,
+      "step": 6546
+    },
+    {
+      "epoch": 1.4604059781396386,
+      "grad_norm": 0.1847231388092041,
+      "learning_rate": 1.0537499426609426e-05,
+      "loss": 0.4382,
+      "step": 6547
+    },
+    {
+      "epoch": 1.460629043051528,
+      "grad_norm": 0.17485037446022034,
+      "learning_rate": 1.0535149384588943e-05,
+      "loss": 0.428,
+      "step": 6548
+    },
+    {
+      "epoch": 1.4608521079634174,
+      "grad_norm": 0.17149090766906738,
+      "learning_rate": 1.0532799312928525e-05,
+      "loss": 0.4624,
+      "step": 6549
+    },
+    {
+      "epoch": 1.4610751728753066,
+      "grad_norm": 0.17815500497817993,
+      "learning_rate": 1.0530449211758327e-05,
+      "loss": 0.4581,
+      "step": 6550
+    },
+    {
+      "epoch": 1.4612982377871961,
+      "grad_norm": 0.16746407747268677,
+      "learning_rate": 1.0528099081208514e-05,
+      "loss": 0.4839,
+      "step": 6551
+    },
+    {
+      "epoch": 1.4615213026990854,
+      "grad_norm": 0.17022733390331268,
+      "learning_rate": 1.0525748921409256e-05,
+      "loss": 0.4636,
+      "step": 6552
+    },
+    {
+      "epoch": 1.461744367610975,
+      "grad_norm": 0.1741877943277359,
+      "learning_rate": 1.0523398732490712e-05,
+      "loss": 0.4602,
+      "step": 6553
+    },
+    {
+      "epoch": 1.4619674325228642,
+      "grad_norm": 0.16914743185043335,
+      "learning_rate": 1.0521048514583057e-05,
+      "loss": 0.4616,
+      "step": 6554
+    },
+    {
+      "epoch": 1.4621904974347535,
+      "grad_norm": 0.16964775323867798,
+      "learning_rate": 1.0518698267816454e-05,
+      "loss": 0.4738,
+      "step": 6555
+    },
+    {
+      "epoch": 1.4624135623466428,
+      "grad_norm": 0.17116379737854004,
+      "learning_rate": 1.0516347992321081e-05,
+      "loss": 0.4588,
+      "step": 6556
+    },
+    {
+      "epoch": 1.4626366272585323,
+      "grad_norm": 0.22291259467601776,
+      "learning_rate": 1.0513997688227107e-05,
+      "loss": 0.4697,
+      "step": 6557
+    },
+    {
+      "epoch": 1.4628596921704216,
+      "grad_norm": 0.18003083765506744,
+      "learning_rate": 1.051164735566471e-05,
+      "loss": 0.4628,
+      "step": 6558
+    },
+    {
+      "epoch": 1.463082757082311,
+      "grad_norm": 0.1763797551393509,
+      "learning_rate": 1.050929699476406e-05,
+      "loss": 0.4565,
+      "step": 6559
+    },
+    {
+      "epoch": 1.4633058219942003,
+      "grad_norm": 0.1819460391998291,
+      "learning_rate": 1.0506946605655342e-05,
+      "loss": 0.483,
+      "step": 6560
+    },
+    {
+      "epoch": 1.4635288869060896,
+      "grad_norm": 0.18002116680145264,
+      "learning_rate": 1.050459618846873e-05,
+      "loss": 0.4485,
+      "step": 6561
+    },
+    {
+      "epoch": 1.463751951817979,
+      "grad_norm": 0.1870141327381134,
+      "learning_rate": 1.0502245743334409e-05,
+      "loss": 0.4761,
+      "step": 6562
+    },
+    {
+      "epoch": 1.4639750167298684,
+      "grad_norm": 0.16674816608428955,
+      "learning_rate": 1.0499895270382558e-05,
+      "loss": 0.4615,
+      "step": 6563
+    },
+    {
+      "epoch": 1.4641980816417577,
+      "grad_norm": 0.18010953068733215,
+      "learning_rate": 1.0497544769743362e-05,
+      "loss": 0.4677,
+      "step": 6564
+    },
+    {
+      "epoch": 1.4644211465536472,
+      "grad_norm": 0.17383818328380585,
+      "learning_rate": 1.049519424154701e-05,
+      "loss": 0.4165,
+      "step": 6565
+    },
+    {
+      "epoch": 1.4646442114655365,
+      "grad_norm": 0.17807599902153015,
+      "learning_rate": 1.0492843685923684e-05,
+      "loss": 0.4671,
+      "step": 6566
+    },
+    {
+      "epoch": 1.4648672763774258,
+      "grad_norm": 0.1714266836643219,
+      "learning_rate": 1.0490493103003573e-05,
+      "loss": 0.4445,
+      "step": 6567
+    },
+    {
+      "epoch": 1.4650903412893153,
+      "grad_norm": 0.18563859164714813,
+      "learning_rate": 1.048814249291687e-05,
+      "loss": 0.4694,
+      "step": 6568
+    },
+    {
+      "epoch": 1.4653134062012045,
+      "grad_norm": 0.1768602728843689,
+      "learning_rate": 1.0485791855793768e-05,
+      "loss": 0.4529,
+      "step": 6569
+    },
+    {
+      "epoch": 1.465536471113094,
+      "grad_norm": 0.16756857931613922,
+      "learning_rate": 1.0483441191764453e-05,
+      "loss": 0.4739,
+      "step": 6570
+    },
+    {
+      "epoch": 1.4657595360249833,
+      "grad_norm": 0.16969642043113708,
+      "learning_rate": 1.0481090500959125e-05,
+      "loss": 0.4815,
+      "step": 6571
+    },
+    {
+      "epoch": 1.4659826009368726,
+      "grad_norm": 0.17080920934677124,
+      "learning_rate": 1.047873978350798e-05,
+      "loss": 0.476,
+      "step": 6572
+    },
+    {
+      "epoch": 1.4662056658487619,
+      "grad_norm": 0.17539383471012115,
+      "learning_rate": 1.0476389039541214e-05,
+      "loss": 0.4739,
+      "step": 6573
+    },
+    {
+      "epoch": 1.4664287307606514,
+      "grad_norm": 0.1617891639471054,
+      "learning_rate": 1.0474038269189026e-05,
+      "loss": 0.4419,
+      "step": 6574
+    },
+    {
+      "epoch": 1.4666517956725407,
+      "grad_norm": 0.1763634830713272,
+      "learning_rate": 1.0471687472581617e-05,
+      "loss": 0.4513,
+      "step": 6575
+    },
+    {
+      "epoch": 1.4668748605844302,
+      "grad_norm": 0.16819307208061218,
+      "learning_rate": 1.046933664984919e-05,
+      "loss": 0.4533,
+      "step": 6576
+    },
+    {
+      "epoch": 1.4670979254963195,
+      "grad_norm": 0.16723337769508362,
+      "learning_rate": 1.0466985801121948e-05,
+      "loss": 0.4588,
+      "step": 6577
+    },
+    {
+      "epoch": 1.4673209904082087,
+      "grad_norm": 0.16584204137325287,
+      "learning_rate": 1.046463492653009e-05,
+      "loss": 0.432,
+      "step": 6578
+    },
+    {
+      "epoch": 1.467544055320098,
+      "grad_norm": 0.1695747673511505,
+      "learning_rate": 1.0462284026203831e-05,
+      "loss": 0.4584,
+      "step": 6579
+    },
+    {
+      "epoch": 1.4677671202319875,
+      "grad_norm": 0.17204824090003967,
+      "learning_rate": 1.0459933100273371e-05,
+      "loss": 0.4653,
+      "step": 6580
+    },
+    {
+      "epoch": 1.4679901851438768,
+      "grad_norm": 0.18302413821220398,
+      "learning_rate": 1.0457582148868928e-05,
+      "loss": 0.4231,
+      "step": 6581
+    },
+    {
+      "epoch": 1.4682132500557663,
+      "grad_norm": 0.17187362909317017,
+      "learning_rate": 1.0455231172120704e-05,
+      "loss": 0.4444,
+      "step": 6582
+    },
+    {
+      "epoch": 1.4684363149676556,
+      "grad_norm": 0.16787424683570862,
+      "learning_rate": 1.0452880170158914e-05,
+      "loss": 0.4393,
+      "step": 6583
+    },
+    {
+      "epoch": 1.4686593798795449,
+      "grad_norm": 0.16989830136299133,
+      "learning_rate": 1.0450529143113772e-05,
+      "loss": 0.4475,
+      "step": 6584
+    },
+    {
+      "epoch": 1.4688824447914344,
+      "grad_norm": 0.17312617599964142,
+      "learning_rate": 1.0448178091115493e-05,
+      "loss": 0.4664,
+      "step": 6585
+    },
+    {
+      "epoch": 1.4691055097033237,
+      "grad_norm": 0.16683322191238403,
+      "learning_rate": 1.0445827014294292e-05,
+      "loss": 0.4654,
+      "step": 6586
+    },
+    {
+      "epoch": 1.4693285746152132,
+      "grad_norm": 0.18341569602489471,
+      "learning_rate": 1.0443475912780386e-05,
+      "loss": 0.4659,
+      "step": 6587
+    },
+    {
+      "epoch": 1.4695516395271024,
+      "grad_norm": 0.18148088455200195,
+      "learning_rate": 1.0441124786703996e-05,
+      "loss": 0.4771,
+      "step": 6588
+    },
+    {
+      "epoch": 1.4697747044389917,
+      "grad_norm": 0.1746566742658615,
+      "learning_rate": 1.0438773636195336e-05,
+      "loss": 0.4601,
+      "step": 6589
+    },
+    {
+      "epoch": 1.469997769350881,
+      "grad_norm": 0.1989150047302246,
+      "learning_rate": 1.0436422461384636e-05,
+      "loss": 0.4603,
+      "step": 6590
+    },
+    {
+      "epoch": 1.4702208342627705,
+      "grad_norm": 0.1698281466960907,
+      "learning_rate": 1.0434071262402114e-05,
+      "loss": 0.4584,
+      "step": 6591
+    },
+    {
+      "epoch": 1.4704438991746598,
+      "grad_norm": 0.1744074821472168,
+      "learning_rate": 1.0431720039377998e-05,
+      "loss": 0.4839,
+      "step": 6592
+    },
+    {
+      "epoch": 1.4706669640865493,
+      "grad_norm": 0.1633480042219162,
+      "learning_rate": 1.0429368792442507e-05,
+      "loss": 0.4426,
+      "step": 6593
+    },
+    {
+      "epoch": 1.4708900289984386,
+      "grad_norm": 0.17857415974140167,
+      "learning_rate": 1.0427017521725873e-05,
+      "loss": 0.4595,
+      "step": 6594
+    },
+    {
+      "epoch": 1.4711130939103279,
+      "grad_norm": 0.1769183874130249,
+      "learning_rate": 1.0424666227358323e-05,
+      "loss": 0.4862,
+      "step": 6595
+    },
+    {
+      "epoch": 1.4713361588222171,
+      "grad_norm": 0.16599570214748383,
+      "learning_rate": 1.0422314909470082e-05,
+      "loss": 0.4185,
+      "step": 6596
+    },
+    {
+      "epoch": 1.4715592237341066,
+      "grad_norm": 0.16501422226428986,
+      "learning_rate": 1.0419963568191389e-05,
+      "loss": 0.4314,
+      "step": 6597
+    },
+    {
+      "epoch": 1.471782288645996,
+      "grad_norm": 0.17242339253425598,
+      "learning_rate": 1.041761220365247e-05,
+      "loss": 0.4411,
+      "step": 6598
+    },
+    {
+      "epoch": 1.4720053535578854,
+      "grad_norm": 0.1672823429107666,
+      "learning_rate": 1.0415260815983561e-05,
+      "loss": 0.4531,
+      "step": 6599
+    },
+    {
+      "epoch": 1.4722284184697747,
+      "grad_norm": 0.1693340539932251,
+      "learning_rate": 1.0412909405314896e-05,
+      "loss": 0.4494,
+      "step": 6600
+    },
+    {
+      "epoch": 1.472451483381664,
+      "grad_norm": 0.17069166898727417,
+      "learning_rate": 1.0410557971776711e-05,
+      "loss": 0.4587,
+      "step": 6601
+    },
+    {
+      "epoch": 1.4726745482935535,
+      "grad_norm": 0.1749400794506073,
+      "learning_rate": 1.0408206515499242e-05,
+      "loss": 0.4258,
+      "step": 6602
+    },
+    {
+      "epoch": 1.4728976132054428,
+      "grad_norm": 0.17911866307258606,
+      "learning_rate": 1.0405855036612728e-05,
+      "loss": 0.4718,
+      "step": 6603
+    },
+    {
+      "epoch": 1.4731206781173323,
+      "grad_norm": 0.1690555214881897,
+      "learning_rate": 1.040350353524741e-05,
+      "loss": 0.4594,
+      "step": 6604
+    },
+    {
+      "epoch": 1.4733437430292216,
+      "grad_norm": 0.1977461576461792,
+      "learning_rate": 1.0401152011533531e-05,
+      "loss": 0.4651,
+      "step": 6605
+    },
+    {
+      "epoch": 1.4735668079411108,
+      "grad_norm": 0.17112137377262115,
+      "learning_rate": 1.0398800465601327e-05,
+      "loss": 0.47,
+      "step": 6606
+    },
+    {
+      "epoch": 1.4737898728530001,
+      "grad_norm": 0.20035414397716522,
+      "learning_rate": 1.0396448897581043e-05,
+      "loss": 0.4539,
+      "step": 6607
+    },
+    {
+      "epoch": 1.4740129377648896,
+      "grad_norm": 0.1666097790002823,
+      "learning_rate": 1.0394097307602928e-05,
+      "loss": 0.4328,
+      "step": 6608
+    },
+    {
+      "epoch": 1.474236002676779,
+      "grad_norm": 0.17964059114456177,
+      "learning_rate": 1.0391745695797226e-05,
+      "loss": 0.4684,
+      "step": 6609
+    },
+    {
+      "epoch": 1.4744590675886684,
+      "grad_norm": 0.1727062463760376,
+      "learning_rate": 1.038939406229418e-05,
+      "loss": 0.435,
+      "step": 6610
+    },
+    {
+      "epoch": 1.4746821325005577,
+      "grad_norm": 0.18273860216140747,
+      "learning_rate": 1.0387042407224046e-05,
+      "loss": 0.4723,
+      "step": 6611
+    },
+    {
+      "epoch": 1.474905197412447,
+      "grad_norm": 0.1780747026205063,
+      "learning_rate": 1.0384690730717065e-05,
+      "loss": 0.4544,
+      "step": 6612
+    },
+    {
+      "epoch": 1.4751282623243362,
+      "grad_norm": 0.17232657968997955,
+      "learning_rate": 1.0382339032903492e-05,
+      "loss": 0.4507,
+      "step": 6613
+    },
+    {
+      "epoch": 1.4753513272362258,
+      "grad_norm": 0.17787334322929382,
+      "learning_rate": 1.037998731391358e-05,
+      "loss": 0.4654,
+      "step": 6614
+    },
+    {
+      "epoch": 1.475574392148115,
+      "grad_norm": 0.17883604764938354,
+      "learning_rate": 1.0377635573877581e-05,
+      "loss": 0.4569,
+      "step": 6615
+    },
+    {
+      "epoch": 1.4757974570600045,
+      "grad_norm": 0.172362819314003,
+      "learning_rate": 1.0375283812925745e-05,
+      "loss": 0.4433,
+      "step": 6616
+    },
+    {
+      "epoch": 1.4760205219718938,
+      "grad_norm": 0.17884163558483124,
+      "learning_rate": 1.037293203118833e-05,
+      "loss": 0.437,
+      "step": 6617
+    },
+    {
+      "epoch": 1.476243586883783,
+      "grad_norm": 0.1667267084121704,
+      "learning_rate": 1.0370580228795597e-05,
+      "loss": 0.4556,
+      "step": 6618
+    },
+    {
+      "epoch": 1.4764666517956726,
+      "grad_norm": 0.17479191720485687,
+      "learning_rate": 1.0368228405877799e-05,
+      "loss": 0.4462,
+      "step": 6619
+    },
+    {
+      "epoch": 1.4766897167075619,
+      "grad_norm": 0.1751541793346405,
+      "learning_rate": 1.0365876562565195e-05,
+      "loss": 0.4504,
+      "step": 6620
+    },
+    {
+      "epoch": 1.4769127816194514,
+      "grad_norm": 0.20716704428195953,
+      "learning_rate": 1.0363524698988045e-05,
+      "loss": 0.4626,
+      "step": 6621
+    },
+    {
+      "epoch": 1.4771358465313407,
+      "grad_norm": 0.16398191452026367,
+      "learning_rate": 1.0361172815276615e-05,
+      "loss": 0.4581,
+      "step": 6622
+    },
+    {
+      "epoch": 1.47735891144323,
+      "grad_norm": 0.17340189218521118,
+      "learning_rate": 1.0358820911561157e-05,
+      "loss": 0.4472,
+      "step": 6623
+    },
+    {
+      "epoch": 1.4775819763551192,
+      "grad_norm": 0.18412639200687408,
+      "learning_rate": 1.0356468987971944e-05,
+      "loss": 0.4323,
+      "step": 6624
+    },
+    {
+      "epoch": 1.4778050412670087,
+      "grad_norm": 0.16925156116485596,
+      "learning_rate": 1.0354117044639232e-05,
+      "loss": 0.4324,
+      "step": 6625
+    },
+    {
+      "epoch": 1.478028106178898,
+      "grad_norm": 0.1772768348455429,
+      "learning_rate": 1.0351765081693294e-05,
+      "loss": 0.4574,
+      "step": 6626
+    },
+    {
+      "epoch": 1.4782511710907875,
+      "grad_norm": 0.1698581427335739,
+      "learning_rate": 1.0349413099264391e-05,
+      "loss": 0.4748,
+      "step": 6627
+    },
+    {
+      "epoch": 1.4784742360026768,
+      "grad_norm": 0.1687650978565216,
+      "learning_rate": 1.0347061097482794e-05,
+      "loss": 0.4226,
+      "step": 6628
+    },
+    {
+      "epoch": 1.478697300914566,
+      "grad_norm": 0.18025682866573334,
+      "learning_rate": 1.034470907647877e-05,
+      "loss": 0.4615,
+      "step": 6629
+    },
+    {
+      "epoch": 1.4789203658264554,
+      "grad_norm": 0.18218332529067993,
+      "learning_rate": 1.034235703638259e-05,
+      "loss": 0.4393,
+      "step": 6630
+    },
+    {
+      "epoch": 1.4791434307383449,
+      "grad_norm": 0.17049945890903473,
+      "learning_rate": 1.0340004977324521e-05,
+      "loss": 0.4613,
+      "step": 6631
+    },
+    {
+      "epoch": 1.4793664956502341,
+      "grad_norm": 0.18157726526260376,
+      "learning_rate": 1.033765289943484e-05,
+      "loss": 0.4811,
+      "step": 6632
+    },
+    {
+      "epoch": 1.4795895605621237,
+      "grad_norm": 0.17979174852371216,
+      "learning_rate": 1.0335300802843815e-05,
+      "loss": 0.4478,
+      "step": 6633
+    },
+    {
+      "epoch": 1.479812625474013,
+      "grad_norm": 0.17775587737560272,
+      "learning_rate": 1.0332948687681725e-05,
+      "loss": 0.4669,
+      "step": 6634
+    },
+    {
+      "epoch": 1.4800356903859022,
+      "grad_norm": 0.1723344624042511,
+      "learning_rate": 1.0330596554078842e-05,
+      "loss": 0.4577,
+      "step": 6635
+    },
+    {
+      "epoch": 1.4802587552977917,
+      "grad_norm": 0.1759558767080307,
+      "learning_rate": 1.0328244402165442e-05,
+      "loss": 0.476,
+      "step": 6636
+    },
+    {
+      "epoch": 1.480481820209681,
+      "grad_norm": 0.1862598955631256,
+      "learning_rate": 1.0325892232071803e-05,
+      "loss": 0.464,
+      "step": 6637
+    },
+    {
+      "epoch": 1.4807048851215705,
+      "grad_norm": 0.18910667300224304,
+      "learning_rate": 1.0323540043928199e-05,
+      "loss": 0.4197,
+      "step": 6638
+    },
+    {
+      "epoch": 1.4809279500334598,
+      "grad_norm": 0.16868862509727478,
+      "learning_rate": 1.0321187837864917e-05,
+      "loss": 0.4334,
+      "step": 6639
+    },
+    {
+      "epoch": 1.481151014945349,
+      "grad_norm": 0.17516660690307617,
+      "learning_rate": 1.0318835614012228e-05,
+      "loss": 0.4503,
+      "step": 6640
+    },
+    {
+      "epoch": 1.4813740798572383,
+      "grad_norm": 0.1748761534690857,
+      "learning_rate": 1.0316483372500422e-05,
+      "loss": 0.4656,
+      "step": 6641
+    },
+    {
+      "epoch": 1.4815971447691278,
+      "grad_norm": 0.17153500020503998,
+      "learning_rate": 1.0314131113459772e-05,
+      "loss": 0.4646,
+      "step": 6642
+    },
+    {
+      "epoch": 1.4818202096810171,
+      "grad_norm": 0.1690903753042221,
+      "learning_rate": 1.0311778837020565e-05,
+      "loss": 0.4384,
+      "step": 6643
+    },
+    {
+      "epoch": 1.4820432745929066,
+      "grad_norm": 0.17170549929141998,
+      "learning_rate": 1.0309426543313086e-05,
+      "loss": 0.4606,
+      "step": 6644
+    },
+    {
+      "epoch": 1.482266339504796,
+      "grad_norm": 0.20019586384296417,
+      "learning_rate": 1.030707423246762e-05,
+      "loss": 0.4458,
+      "step": 6645
+    },
+    {
+      "epoch": 1.4824894044166852,
+      "grad_norm": 0.17187346518039703,
+      "learning_rate": 1.0304721904614447e-05,
+      "loss": 0.4777,
+      "step": 6646
+    },
+    {
+      "epoch": 1.4827124693285745,
+      "grad_norm": 0.17592963576316833,
+      "learning_rate": 1.0302369559883862e-05,
+      "loss": 0.462,
+      "step": 6647
+    },
+    {
+      "epoch": 1.482935534240464,
+      "grad_norm": 0.18362174928188324,
+      "learning_rate": 1.0300017198406148e-05,
+      "loss": 0.4851,
+      "step": 6648
+    },
+    {
+      "epoch": 1.4831585991523533,
+      "grad_norm": 0.17138110101222992,
+      "learning_rate": 1.0297664820311593e-05,
+      "loss": 0.5046,
+      "step": 6649
+    },
+    {
+      "epoch": 1.4833816640642428,
+      "grad_norm": 0.16969837248325348,
+      "learning_rate": 1.029531242573049e-05,
+      "loss": 0.4376,
+      "step": 6650
+    },
+    {
+      "epoch": 1.483604728976132,
+      "grad_norm": 0.1832205057144165,
+      "learning_rate": 1.0292960014793126e-05,
+      "loss": 0.4376,
+      "step": 6651
+    },
+    {
+      "epoch": 1.4838277938880213,
+      "grad_norm": 0.1777479201555252,
+      "learning_rate": 1.0290607587629795e-05,
+      "loss": 0.4616,
+      "step": 6652
+    },
+    {
+      "epoch": 1.4840508587999108,
+      "grad_norm": 0.18245477974414825,
+      "learning_rate": 1.0288255144370784e-05,
+      "loss": 0.4699,
+      "step": 6653
+    },
+    {
+      "epoch": 1.4842739237118001,
+      "grad_norm": 0.16267704963684082,
+      "learning_rate": 1.0285902685146394e-05,
+      "loss": 0.4358,
+      "step": 6654
+    },
+    {
+      "epoch": 1.4844969886236896,
+      "grad_norm": 0.1723002940416336,
+      "learning_rate": 1.0283550210086913e-05,
+      "loss": 0.4621,
+      "step": 6655
+    },
+    {
+      "epoch": 1.484720053535579,
+      "grad_norm": 0.17394044995307922,
+      "learning_rate": 1.028119771932264e-05,
+      "loss": 0.4456,
+      "step": 6656
+    },
+    {
+      "epoch": 1.4849431184474682,
+      "grad_norm": 0.1774415820837021,
+      "learning_rate": 1.0278845212983865e-05,
+      "loss": 0.4761,
+      "step": 6657
+    },
+    {
+      "epoch": 1.4851661833593575,
+      "grad_norm": 0.16867241263389587,
+      "learning_rate": 1.0276492691200893e-05,
+      "loss": 0.482,
+      "step": 6658
+    },
+    {
+      "epoch": 1.485389248271247,
+      "grad_norm": 0.1685202419757843,
+      "learning_rate": 1.027414015410401e-05,
+      "loss": 0.45,
+      "step": 6659
+    },
+    {
+      "epoch": 1.4856123131831362,
+      "grad_norm": 0.1738312542438507,
+      "learning_rate": 1.0271787601823526e-05,
+      "loss": 0.4591,
+      "step": 6660
+    },
+    {
+      "epoch": 1.4858353780950257,
+      "grad_norm": 0.17828047275543213,
+      "learning_rate": 1.0269435034489733e-05,
+      "loss": 0.4261,
+      "step": 6661
+    },
+    {
+      "epoch": 1.486058443006915,
+      "grad_norm": 0.164927139878273,
+      "learning_rate": 1.0267082452232935e-05,
+      "loss": 0.4383,
+      "step": 6662
+    },
+    {
+      "epoch": 1.4862815079188043,
+      "grad_norm": 0.17446818947792053,
+      "learning_rate": 1.0264729855183426e-05,
+      "loss": 0.461,
+      "step": 6663
+    },
+    {
+      "epoch": 1.4865045728306936,
+      "grad_norm": 0.1646289974451065,
+      "learning_rate": 1.0262377243471517e-05,
+      "loss": 0.4335,
+      "step": 6664
+    },
+    {
+      "epoch": 1.486727637742583,
+      "grad_norm": 0.19351711869239807,
+      "learning_rate": 1.0260024617227504e-05,
+      "loss": 0.4697,
+      "step": 6665
+    },
+    {
+      "epoch": 1.4869507026544724,
+      "grad_norm": 0.18421606719493866,
+      "learning_rate": 1.0257671976581694e-05,
+      "loss": 0.4531,
+      "step": 6666
+    },
+    {
+      "epoch": 1.4871737675663619,
+      "grad_norm": 0.17079128324985504,
+      "learning_rate": 1.0255319321664386e-05,
+      "loss": 0.4531,
+      "step": 6667
+    },
+    {
+      "epoch": 1.4873968324782512,
+      "grad_norm": 0.16973423957824707,
+      "learning_rate": 1.0252966652605889e-05,
+      "loss": 0.4358,
+      "step": 6668
+    },
+    {
+      "epoch": 1.4876198973901404,
+      "grad_norm": 0.17064902186393738,
+      "learning_rate": 1.0250613969536507e-05,
+      "loss": 0.4843,
+      "step": 6669
+    },
+    {
+      "epoch": 1.48784296230203,
+      "grad_norm": 0.1614328771829605,
+      "learning_rate": 1.0248261272586549e-05,
+      "loss": 0.4197,
+      "step": 6670
+    },
+    {
+      "epoch": 1.4880660272139192,
+      "grad_norm": 0.1676885336637497,
+      "learning_rate": 1.0245908561886319e-05,
+      "loss": 0.464,
+      "step": 6671
+    },
+    {
+      "epoch": 1.4882890921258087,
+      "grad_norm": 0.17245353758335114,
+      "learning_rate": 1.0243555837566124e-05,
+      "loss": 0.4368,
+      "step": 6672
+    },
+    {
+      "epoch": 1.488512157037698,
+      "grad_norm": 0.184434175491333,
+      "learning_rate": 1.0241203099756279e-05,
+      "loss": 0.4419,
+      "step": 6673
+    },
+    {
+      "epoch": 1.4887352219495873,
+      "grad_norm": 0.17373715341091156,
+      "learning_rate": 1.0238850348587088e-05,
+      "loss": 0.471,
+      "step": 6674
+    },
+    {
+      "epoch": 1.4889582868614766,
+      "grad_norm": 0.1708296537399292,
+      "learning_rate": 1.0236497584188862e-05,
+      "loss": 0.4468,
+      "step": 6675
+    },
+    {
+      "epoch": 1.489181351773366,
+      "grad_norm": 0.18406598269939423,
+      "learning_rate": 1.023414480669191e-05,
+      "loss": 0.4622,
+      "step": 6676
+    },
+    {
+      "epoch": 1.4894044166852554,
+      "grad_norm": 0.17563220858573914,
+      "learning_rate": 1.0231792016226546e-05,
+      "loss": 0.4529,
+      "step": 6677
+    },
+    {
+      "epoch": 1.4896274815971449,
+      "grad_norm": 0.21065136790275574,
+      "learning_rate": 1.0229439212923084e-05,
+      "loss": 0.4668,
+      "step": 6678
+    },
+    {
+      "epoch": 1.4898505465090341,
+      "grad_norm": 0.18098492920398712,
+      "learning_rate": 1.0227086396911837e-05,
+      "loss": 0.4669,
+      "step": 6679
+    },
+    {
+      "epoch": 1.4900736114209234,
+      "grad_norm": 0.18423466384410858,
+      "learning_rate": 1.0224733568323111e-05,
+      "loss": 0.4815,
+      "step": 6680
+    },
+    {
+      "epoch": 1.490296676332813,
+      "grad_norm": 0.17648786306381226,
+      "learning_rate": 1.022238072728723e-05,
+      "loss": 0.4566,
+      "step": 6681
+    },
+    {
+      "epoch": 1.4905197412447022,
+      "grad_norm": 0.17524605989456177,
+      "learning_rate": 1.0220027873934506e-05,
+      "loss": 0.4466,
+      "step": 6682
+    },
+    {
+      "epoch": 1.4907428061565915,
+      "grad_norm": 0.18514925241470337,
+      "learning_rate": 1.021767500839525e-05,
+      "loss": 0.4545,
+      "step": 6683
+    },
+    {
+      "epoch": 1.490965871068481,
+      "grad_norm": 0.17596104741096497,
+      "learning_rate": 1.021532213079979e-05,
+      "loss": 0.4637,
+      "step": 6684
+    },
+    {
+      "epoch": 1.4911889359803703,
+      "grad_norm": 0.1722065955400467,
+      "learning_rate": 1.021296924127843e-05,
+      "loss": 0.4471,
+      "step": 6685
+    },
+    {
+      "epoch": 1.4914120008922596,
+      "grad_norm": 0.1769905835390091,
+      "learning_rate": 1.0210616339961497e-05,
+      "loss": 0.4665,
+      "step": 6686
+    },
+    {
+      "epoch": 1.491635065804149,
+      "grad_norm": 0.1704041212797165,
+      "learning_rate": 1.0208263426979304e-05,
+      "loss": 0.437,
+      "step": 6687
+    },
+    {
+      "epoch": 1.4918581307160383,
+      "grad_norm": 0.1706121265888214,
+      "learning_rate": 1.0205910502462174e-05,
+      "loss": 0.4533,
+      "step": 6688
+    },
+    {
+      "epoch": 1.4920811956279278,
+      "grad_norm": 0.1680435687303543,
+      "learning_rate": 1.0203557566540425e-05,
+      "loss": 0.4465,
+      "step": 6689
+    },
+    {
+      "epoch": 1.4923042605398171,
+      "grad_norm": 0.17137907445430756,
+      "learning_rate": 1.0201204619344378e-05,
+      "loss": 0.4712,
+      "step": 6690
+    },
+    {
+      "epoch": 1.4925273254517064,
+      "grad_norm": 0.17346073687076569,
+      "learning_rate": 1.0198851661004352e-05,
+      "loss": 0.4396,
+      "step": 6691
+    },
+    {
+      "epoch": 1.4927503903635957,
+      "grad_norm": 0.1783093512058258,
+      "learning_rate": 1.0196498691650671e-05,
+      "loss": 0.453,
+      "step": 6692
+    },
+    {
+      "epoch": 1.4929734552754852,
+      "grad_norm": 0.16575607657432556,
+      "learning_rate": 1.0194145711413656e-05,
+      "loss": 0.4234,
+      "step": 6693
+    },
+    {
+      "epoch": 1.4931965201873745,
+      "grad_norm": 0.17617058753967285,
+      "learning_rate": 1.0191792720423632e-05,
+      "loss": 0.4275,
+      "step": 6694
+    },
+    {
+      "epoch": 1.493419585099264,
+      "grad_norm": 0.17400598526000977,
+      "learning_rate": 1.0189439718810919e-05,
+      "loss": 0.4667,
+      "step": 6695
+    },
+    {
+      "epoch": 1.4936426500111533,
+      "grad_norm": 0.17919796705245972,
+      "learning_rate": 1.0187086706705844e-05,
+      "loss": 0.4669,
+      "step": 6696
+    },
+    {
+      "epoch": 1.4938657149230425,
+      "grad_norm": 0.17682035267353058,
+      "learning_rate": 1.0184733684238728e-05,
+      "loss": 0.4918,
+      "step": 6697
+    },
+    {
+      "epoch": 1.494088779834932,
+      "grad_norm": 0.16982607543468475,
+      "learning_rate": 1.01823806515399e-05,
+      "loss": 0.4626,
+      "step": 6698
+    },
+    {
+      "epoch": 1.4943118447468213,
+      "grad_norm": 0.17292605340480804,
+      "learning_rate": 1.0180027608739684e-05,
+      "loss": 0.4648,
+      "step": 6699
+    },
+    {
+      "epoch": 1.4945349096587106,
+      "grad_norm": 0.16893380880355835,
+      "learning_rate": 1.017767455596841e-05,
+      "loss": 0.4285,
+      "step": 6700
+    },
+    {
+      "epoch": 1.4947579745706001,
+      "grad_norm": 0.17039963603019714,
+      "learning_rate": 1.0175321493356396e-05,
+      "loss": 0.4589,
+      "step": 6701
+    },
+    {
+      "epoch": 1.4949810394824894,
+      "grad_norm": 0.1750868260860443,
+      "learning_rate": 1.0172968421033977e-05,
+      "loss": 0.4703,
+      "step": 6702
+    },
+    {
+      "epoch": 1.4952041043943787,
+      "grad_norm": 0.17685000598430634,
+      "learning_rate": 1.017061533913148e-05,
+      "loss": 0.4585,
+      "step": 6703
+    },
+    {
+      "epoch": 1.4954271693062682,
+      "grad_norm": 0.1752578467130661,
+      "learning_rate": 1.0168262247779231e-05,
+      "loss": 0.4583,
+      "step": 6704
+    },
+    {
+      "epoch": 1.4956502342181575,
+      "grad_norm": 0.16632848978042603,
+      "learning_rate": 1.0165909147107563e-05,
+      "loss": 0.4684,
+      "step": 6705
+    },
+    {
+      "epoch": 1.495873299130047,
+      "grad_norm": 0.17605751752853394,
+      "learning_rate": 1.0163556037246798e-05,
+      "loss": 0.4747,
+      "step": 6706
+    },
+    {
+      "epoch": 1.4960963640419362,
+      "grad_norm": 0.16805213689804077,
+      "learning_rate": 1.0161202918327276e-05,
+      "loss": 0.4272,
+      "step": 6707
+    },
+    {
+      "epoch": 1.4963194289538255,
+      "grad_norm": 0.25550636649131775,
+      "learning_rate": 1.0158849790479318e-05,
+      "loss": 0.4756,
+      "step": 6708
+    },
+    {
+      "epoch": 1.4965424938657148,
+      "grad_norm": 0.1749117523431778,
+      "learning_rate": 1.015649665383326e-05,
+      "loss": 0.4746,
+      "step": 6709
+    },
+    {
+      "epoch": 1.4967655587776043,
+      "grad_norm": 0.17432990670204163,
+      "learning_rate": 1.0154143508519434e-05,
+      "loss": 0.4707,
+      "step": 6710
+    },
+    {
+      "epoch": 1.4969886236894936,
+      "grad_norm": 0.16568134725093842,
+      "learning_rate": 1.0151790354668171e-05,
+      "loss": 0.4359,
+      "step": 6711
+    },
+    {
+      "epoch": 1.497211688601383,
+      "grad_norm": 0.17027510702610016,
+      "learning_rate": 1.0149437192409803e-05,
+      "loss": 0.4352,
+      "step": 6712
+    },
+    {
+      "epoch": 1.4974347535132724,
+      "grad_norm": 0.17918628454208374,
+      "learning_rate": 1.0147084021874664e-05,
+      "loss": 0.4749,
+      "step": 6713
+    },
+    {
+      "epoch": 1.4976578184251617,
+      "grad_norm": 0.1729065179824829,
+      "learning_rate": 1.0144730843193086e-05,
+      "loss": 0.4465,
+      "step": 6714
+    },
+    {
+      "epoch": 1.4978808833370512,
+      "grad_norm": 0.17687968909740448,
+      "learning_rate": 1.0142377656495405e-05,
+      "loss": 0.4861,
+      "step": 6715
+    },
+    {
+      "epoch": 1.4981039482489404,
+      "grad_norm": 0.17104387283325195,
+      "learning_rate": 1.0140024461911955e-05,
+      "loss": 0.4529,
+      "step": 6716
+    },
+    {
+      "epoch": 1.4983270131608297,
+      "grad_norm": 0.1739518642425537,
+      "learning_rate": 1.0137671259573066e-05,
+      "loss": 0.4493,
+      "step": 6717
+    },
+    {
+      "epoch": 1.4985500780727192,
+      "grad_norm": 0.17664003372192383,
+      "learning_rate": 1.013531804960908e-05,
+      "loss": 0.4567,
+      "step": 6718
+    },
+    {
+      "epoch": 1.4987731429846085,
+      "grad_norm": 0.16971804201602936,
+      "learning_rate": 1.0132964832150325e-05,
+      "loss": 0.4546,
+      "step": 6719
+    },
+    {
+      "epoch": 1.4989962078964978,
+      "grad_norm": 0.17096221446990967,
+      "learning_rate": 1.0130611607327144e-05,
+      "loss": 0.4399,
+      "step": 6720
+    },
+    {
+      "epoch": 1.4992192728083873,
+      "grad_norm": 0.16578656435012817,
+      "learning_rate": 1.0128258375269868e-05,
+      "loss": 0.4558,
+      "step": 6721
+    },
+    {
+      "epoch": 1.4994423377202766,
+      "grad_norm": 0.17407840490341187,
+      "learning_rate": 1.012590513610884e-05,
+      "loss": 0.4345,
+      "step": 6722
+    },
+    {
+      "epoch": 1.499665402632166,
+      "grad_norm": 0.16804394125938416,
+      "learning_rate": 1.012355188997439e-05,
+      "loss": 0.4497,
+      "step": 6723
+    },
+    {
+      "epoch": 1.4998884675440554,
+      "grad_norm": 0.1964014172554016,
+      "learning_rate": 1.0121198636996862e-05,
+      "loss": 0.4412,
+      "step": 6724
+    },
+    {
+      "epoch": 1.5001115324559446,
+      "grad_norm": 0.17465969920158386,
+      "learning_rate": 1.011884537730659e-05,
+      "loss": 0.4605,
+      "step": 6725
+    },
+    {
+      "epoch": 1.500334597367834,
+      "grad_norm": 0.17849335074424744,
+      "learning_rate": 1.0116492111033916e-05,
+      "loss": 0.4766,
+      "step": 6726
+    },
+    {
+      "epoch": 1.5005576622797234,
+      "grad_norm": 0.1719314008951187,
+      "learning_rate": 1.0114138838309171e-05,
+      "loss": 0.4745,
+      "step": 6727
+    },
+    {
+      "epoch": 1.5007807271916127,
+      "grad_norm": 0.17707185447216034,
+      "learning_rate": 1.0111785559262703e-05,
+      "loss": 0.4544,
+      "step": 6728
+    },
+    {
+      "epoch": 1.5010037921035022,
+      "grad_norm": 0.18315830826759338,
+      "learning_rate": 1.0109432274024846e-05,
+      "loss": 0.4607,
+      "step": 6729
+    },
+    {
+      "epoch": 1.5012268570153915,
+      "grad_norm": 0.18333736062049866,
+      "learning_rate": 1.0107078982725942e-05,
+      "loss": 0.4835,
+      "step": 6730
+    },
+    {
+      "epoch": 1.5014499219272808,
+      "grad_norm": 0.1704811453819275,
+      "learning_rate": 1.010472568549633e-05,
+      "loss": 0.4337,
+      "step": 6731
+    },
+    {
+      "epoch": 1.50167298683917,
+      "grad_norm": 0.1818595677614212,
+      "learning_rate": 1.0102372382466352e-05,
+      "loss": 0.46,
+      "step": 6732
+    },
+    {
+      "epoch": 1.5018960517510596,
+      "grad_norm": 0.16688323020935059,
+      "learning_rate": 1.0100019073766344e-05,
+      "loss": 0.4584,
+      "step": 6733
+    },
+    {
+      "epoch": 1.502119116662949,
+      "grad_norm": 0.17898161709308624,
+      "learning_rate": 1.0097665759526654e-05,
+      "loss": 0.4605,
+      "step": 6734
+    },
+    {
+      "epoch": 1.5023421815748383,
+      "grad_norm": 0.16883964836597443,
+      "learning_rate": 1.009531243987762e-05,
+      "loss": 0.4334,
+      "step": 6735
+    },
+    {
+      "epoch": 1.5025652464867276,
+      "grad_norm": 0.16763442754745483,
+      "learning_rate": 1.009295911494958e-05,
+      "loss": 0.4353,
+      "step": 6736
+    },
+    {
+      "epoch": 1.502788311398617,
+      "grad_norm": 0.1733693778514862,
+      "learning_rate": 1.0090605784872884e-05,
+      "loss": 0.4385,
+      "step": 6737
+    },
+    {
+      "epoch": 1.5030113763105064,
+      "grad_norm": 0.17827312648296356,
+      "learning_rate": 1.0088252449777866e-05,
+      "loss": 0.4436,
+      "step": 6738
+    },
+    {
+      "epoch": 1.5032344412223957,
+      "grad_norm": 0.17768245935440063,
+      "learning_rate": 1.0085899109794874e-05,
+      "loss": 0.4745,
+      "step": 6739
+    },
+    {
+      "epoch": 1.5034575061342852,
+      "grad_norm": 0.1763843595981598,
+      "learning_rate": 1.0083545765054248e-05,
+      "loss": 0.4604,
+      "step": 6740
+    },
+    {
+      "epoch": 1.5036805710461745,
+      "grad_norm": 0.17313605546951294,
+      "learning_rate": 1.0081192415686334e-05,
+      "loss": 0.4513,
+      "step": 6741
+    },
+    {
+      "epoch": 1.5039036359580638,
+      "grad_norm": 0.17623473703861237,
+      "learning_rate": 1.0078839061821473e-05,
+      "loss": 0.4541,
+      "step": 6742
+    },
+    {
+      "epoch": 1.504126700869953,
+      "grad_norm": 0.17384295165538788,
+      "learning_rate": 1.007648570359001e-05,
+      "loss": 0.454,
+      "step": 6743
+    },
+    {
+      "epoch": 1.5043497657818425,
+      "grad_norm": 0.1829373836517334,
+      "learning_rate": 1.0074132341122282e-05,
+      "loss": 0.4691,
+      "step": 6744
+    },
+    {
+      "epoch": 1.5045728306937318,
+      "grad_norm": 0.17881907522678375,
+      "learning_rate": 1.0071778974548642e-05,
+      "loss": 0.4705,
+      "step": 6745
+    },
+    {
+      "epoch": 1.5047958956056213,
+      "grad_norm": 0.17440833151340485,
+      "learning_rate": 1.0069425603999432e-05,
+      "loss": 0.45,
+      "step": 6746
+    },
+    {
+      "epoch": 1.5050189605175106,
+      "grad_norm": 0.17206740379333496,
+      "learning_rate": 1.0067072229604995e-05,
+      "loss": 0.4508,
+      "step": 6747
+    },
+    {
+      "epoch": 1.5052420254293999,
+      "grad_norm": 0.18030904233455658,
+      "learning_rate": 1.0064718851495674e-05,
+      "loss": 0.4786,
+      "step": 6748
+    },
+    {
+      "epoch": 1.5054650903412892,
+      "grad_norm": 0.17595484852790833,
+      "learning_rate": 1.0062365469801818e-05,
+      "loss": 0.4772,
+      "step": 6749
+    },
+    {
+      "epoch": 1.5056881552531787,
+      "grad_norm": 0.1742616444826126,
+      "learning_rate": 1.006001208465377e-05,
+      "loss": 0.4564,
+      "step": 6750
+    },
+    {
+      "epoch": 1.5059112201650682,
+      "grad_norm": 0.16961626708507538,
+      "learning_rate": 1.0057658696181875e-05,
+      "loss": 0.4514,
+      "step": 6751
+    },
+    {
+      "epoch": 1.5061342850769575,
+      "grad_norm": 0.16909094154834747,
+      "learning_rate": 1.0055305304516477e-05,
+      "loss": 0.4647,
+      "step": 6752
+    },
+    {
+      "epoch": 1.5063573499888467,
+      "grad_norm": 0.168659508228302,
+      "learning_rate": 1.0052951909787927e-05,
+      "loss": 0.4883,
+      "step": 6753
+    },
+    {
+      "epoch": 1.506580414900736,
+      "grad_norm": 0.16884328424930573,
+      "learning_rate": 1.0050598512126562e-05,
+      "loss": 0.4683,
+      "step": 6754
+    },
+    {
+      "epoch": 1.5068034798126255,
+      "grad_norm": 0.17461951076984406,
+      "learning_rate": 1.0048245111662735e-05,
+      "loss": 0.455,
+      "step": 6755
+    },
+    {
+      "epoch": 1.5070265447245148,
+      "grad_norm": 0.1754961460828781,
+      "learning_rate": 1.0045891708526796e-05,
+      "loss": 0.4833,
+      "step": 6756
+    },
+    {
+      "epoch": 1.5072496096364043,
+      "grad_norm": 0.16815195977687836,
+      "learning_rate": 1.0043538302849078e-05,
+      "loss": 0.4338,
+      "step": 6757
+    },
+    {
+      "epoch": 1.5074726745482936,
+      "grad_norm": 0.1777064949274063,
+      "learning_rate": 1.004118489475994e-05,
+      "loss": 0.4582,
+      "step": 6758
+    },
+    {
+      "epoch": 1.5076957394601829,
+      "grad_norm": 0.17062366008758545,
+      "learning_rate": 1.0038831484389719e-05,
+      "loss": 0.4427,
+      "step": 6759
+    },
+    {
+      "epoch": 1.5079188043720722,
+      "grad_norm": 0.16902603209018707,
+      "learning_rate": 1.003647807186877e-05,
+      "loss": 0.464,
+      "step": 6760
+    },
+    {
+      "epoch": 1.5081418692839617,
+      "grad_norm": 0.4289887547492981,
+      "learning_rate": 1.0034124657327433e-05,
+      "loss": 0.4714,
+      "step": 6761
+    },
+    {
+      "epoch": 1.5083649341958512,
+      "grad_norm": 0.17374208569526672,
+      "learning_rate": 1.003177124089606e-05,
+      "loss": 0.4568,
+      "step": 6762
+    },
+    {
+      "epoch": 1.5085879991077404,
+      "grad_norm": 0.2121153473854065,
+      "learning_rate": 1.0029417822704995e-05,
+      "loss": 0.4444,
+      "step": 6763
+    },
+    {
+      "epoch": 1.5088110640196297,
+      "grad_norm": 0.17344814538955688,
+      "learning_rate": 1.0027064402884585e-05,
+      "loss": 0.4663,
+      "step": 6764
+    },
+    {
+      "epoch": 1.509034128931519,
+      "grad_norm": 0.18848566710948944,
+      "learning_rate": 1.0024710981565177e-05,
+      "loss": 0.4632,
+      "step": 6765
+    },
+    {
+      "epoch": 1.5092571938434083,
+      "grad_norm": 0.16708870232105255,
+      "learning_rate": 1.002235755887712e-05,
+      "loss": 0.4484,
+      "step": 6766
+    },
+    {
+      "epoch": 1.5094802587552978,
+      "grad_norm": 0.17515549063682556,
+      "learning_rate": 1.0020004134950765e-05,
+      "loss": 0.4862,
+      "step": 6767
+    },
+    {
+      "epoch": 1.5097033236671873,
+      "grad_norm": 0.17192357778549194,
+      "learning_rate": 1.001765070991645e-05,
+      "loss": 0.4601,
+      "step": 6768
+    },
+    {
+      "epoch": 1.5099263885790766,
+      "grad_norm": 0.1708642542362213,
+      "learning_rate": 1.001529728390453e-05,
+      "loss": 0.4686,
+      "step": 6769
+    },
+    {
+      "epoch": 1.5101494534909659,
+      "grad_norm": 0.17151105403900146,
+      "learning_rate": 1.0012943857045349e-05,
+      "loss": 0.459,
+      "step": 6770
+    },
+    {
+      "epoch": 1.5103725184028551,
+      "grad_norm": 0.173813134431839,
+      "learning_rate": 1.001059042946926e-05,
+      "loss": 0.4437,
+      "step": 6771
+    },
+    {
+      "epoch": 1.5105955833147446,
+      "grad_norm": 0.17130842804908752,
+      "learning_rate": 1.0008237001306602e-05,
+      "loss": 0.457,
+      "step": 6772
+    },
+    {
+      "epoch": 1.510818648226634,
+      "grad_norm": 0.17134280502796173,
+      "learning_rate": 1.000588357268773e-05,
+      "loss": 0.4546,
+      "step": 6773
+    },
+    {
+      "epoch": 1.5110417131385234,
+      "grad_norm": 0.17797818779945374,
+      "learning_rate": 1.0003530143742986e-05,
+      "loss": 0.4637,
+      "step": 6774
+    },
+    {
+      "epoch": 1.5112647780504127,
+      "grad_norm": 0.16602249443531036,
+      "learning_rate": 1.000117671460272e-05,
+      "loss": 0.4494,
+      "step": 6775
+    },
+    {
+      "epoch": 1.511487842962302,
+      "grad_norm": 0.16833782196044922,
+      "learning_rate": 9.998823285397282e-06,
+      "loss": 0.4414,
+      "step": 6776
+    },
+    {
+      "epoch": 1.5117109078741913,
+      "grad_norm": 0.1781986951828003,
+      "learning_rate": 9.996469856257018e-06,
+      "loss": 0.444,
+      "step": 6777
+    },
+    {
+      "epoch": 1.5119339727860808,
+      "grad_norm": 0.16520027816295624,
+      "learning_rate": 9.994116427312274e-06,
+      "loss": 0.4289,
+      "step": 6778
+    },
+    {
+      "epoch": 1.5121570376979703,
+      "grad_norm": 0.17205528914928436,
+      "learning_rate": 9.991762998693401e-06,
+      "loss": 0.4745,
+      "step": 6779
+    },
+    {
+      "epoch": 1.5123801026098596,
+      "grad_norm": 0.17844898998737335,
+      "learning_rate": 9.989409570530743e-06,
+      "loss": 0.4685,
+      "step": 6780
+    },
+    {
+      "epoch": 1.5126031675217488,
+      "grad_norm": 0.1733841598033905,
+      "learning_rate": 9.987056142954653e-06,
+      "loss": 0.4505,
+      "step": 6781
+    },
+    {
+      "epoch": 1.5128262324336381,
+      "grad_norm": 0.16828158497810364,
+      "learning_rate": 9.984702716095472e-06,
+      "loss": 0.4303,
+      "step": 6782
+    },
+    {
+      "epoch": 1.5130492973455274,
+      "grad_norm": 0.18432192504405975,
+      "learning_rate": 9.982349290083553e-06,
+      "loss": 0.4657,
+      "step": 6783
+    },
+    {
+      "epoch": 1.513272362257417,
+      "grad_norm": 0.17612218856811523,
+      "learning_rate": 9.979995865049239e-06,
+      "loss": 0.4531,
+      "step": 6784
+    },
+    {
+      "epoch": 1.5134954271693064,
+      "grad_norm": 0.1782546043395996,
+      "learning_rate": 9.977642441122881e-06,
+      "loss": 0.4442,
+      "step": 6785
+    },
+    {
+      "epoch": 1.5137184920811957,
+      "grad_norm": 0.17371070384979248,
+      "learning_rate": 9.975289018434826e-06,
+      "loss": 0.4636,
+      "step": 6786
+    },
+    {
+      "epoch": 1.513941556993085,
+      "grad_norm": 0.16479544341564178,
+      "learning_rate": 9.972935597115417e-06,
+      "loss": 0.4454,
+      "step": 6787
+    },
+    {
+      "epoch": 1.5141646219049743,
+      "grad_norm": 0.18058109283447266,
+      "learning_rate": 9.970582177295009e-06,
+      "loss": 0.4914,
+      "step": 6788
+    },
+    {
+      "epoch": 1.5143876868168638,
+      "grad_norm": 0.17538684606552124,
+      "learning_rate": 9.968228759103943e-06,
+      "loss": 0.4258,
+      "step": 6789
+    },
+    {
+      "epoch": 1.514610751728753,
+      "grad_norm": 0.17010897397994995,
+      "learning_rate": 9.965875342672572e-06,
+      "loss": 0.4544,
+      "step": 6790
+    },
+    {
+      "epoch": 1.5148338166406425,
+      "grad_norm": 0.1833350658416748,
+      "learning_rate": 9.963521928131234e-06,
+      "loss": 0.4608,
+      "step": 6791
+    },
+    {
+      "epoch": 1.5150568815525318,
+      "grad_norm": 0.1769610494375229,
+      "learning_rate": 9.961168515610283e-06,
+      "loss": 0.4679,
+      "step": 6792
+    },
+    {
+      "epoch": 1.515279946464421,
+      "grad_norm": 0.17435282468795776,
+      "learning_rate": 9.958815105240064e-06,
+      "loss": 0.4737,
+      "step": 6793
+    },
+    {
+      "epoch": 1.5155030113763104,
+      "grad_norm": 0.1680775284767151,
+      "learning_rate": 9.956461697150925e-06,
+      "loss": 0.4635,
+      "step": 6794
+    },
+    {
+      "epoch": 1.5157260762881999,
+      "grad_norm": 0.17000484466552734,
+      "learning_rate": 9.954108291473207e-06,
+      "loss": 0.4558,
+      "step": 6795
+    },
+    {
+      "epoch": 1.5159491412000894,
+      "grad_norm": 0.17273864150047302,
+      "learning_rate": 9.951754888337265e-06,
+      "loss": 0.4472,
+      "step": 6796
+    },
+    {
+      "epoch": 1.5161722061119787,
+      "grad_norm": 0.18519042432308197,
+      "learning_rate": 9.949401487873438e-06,
+      "loss": 0.455,
+      "step": 6797
+    },
+    {
+      "epoch": 1.516395271023868,
+      "grad_norm": 0.17533376812934875,
+      "learning_rate": 9.947048090212076e-06,
+      "loss": 0.4684,
+      "step": 6798
+    },
+    {
+      "epoch": 1.5166183359357572,
+      "grad_norm": 0.1752437800168991,
+      "learning_rate": 9.944694695483523e-06,
+      "loss": 0.4625,
+      "step": 6799
+    },
+    {
+      "epoch": 1.5168414008476465,
+      "grad_norm": 0.17211464047431946,
+      "learning_rate": 9.942341303818128e-06,
+      "loss": 0.443,
+      "step": 6800
+    },
+    {
+      "epoch": 1.517064465759536,
+      "grad_norm": 0.170721635222435,
+      "learning_rate": 9.939987915346232e-06,
+      "loss": 0.4211,
+      "step": 6801
+    },
+    {
+      "epoch": 1.5172875306714255,
+      "grad_norm": 0.17440038919448853,
+      "learning_rate": 9.937634530198184e-06,
+      "loss": 0.456,
+      "step": 6802
+    },
+    {
+      "epoch": 1.5175105955833148,
+      "grad_norm": 0.17179426550865173,
+      "learning_rate": 9.935281148504326e-06,
+      "loss": 0.4487,
+      "step": 6803
+    },
+    {
+      "epoch": 1.517733660495204,
+      "grad_norm": 0.17573867738246918,
+      "learning_rate": 9.932927770395007e-06,
+      "loss": 0.4663,
+      "step": 6804
+    },
+    {
+      "epoch": 1.5179567254070934,
+      "grad_norm": 0.18278174102306366,
+      "learning_rate": 9.93057439600057e-06,
+      "loss": 0.4471,
+      "step": 6805
+    },
+    {
+      "epoch": 1.5181797903189829,
+      "grad_norm": 0.17256884276866913,
+      "learning_rate": 9.92822102545136e-06,
+      "loss": 0.4675,
+      "step": 6806
+    },
+    {
+      "epoch": 1.5184028552308722,
+      "grad_norm": 0.19855371117591858,
+      "learning_rate": 9.925867658877718e-06,
+      "loss": 0.4661,
+      "step": 6807
+    },
+    {
+      "epoch": 1.5186259201427617,
+      "grad_norm": 0.1643659472465515,
+      "learning_rate": 9.923514296409992e-06,
+      "loss": 0.4457,
+      "step": 6808
+    },
+    {
+      "epoch": 1.518848985054651,
+      "grad_norm": 0.16748277842998505,
+      "learning_rate": 9.921160938178529e-06,
+      "loss": 0.4558,
+      "step": 6809
+    },
+    {
+      "epoch": 1.5190720499665402,
+      "grad_norm": 0.16805437207221985,
+      "learning_rate": 9.918807584313666e-06,
+      "loss": 0.4283,
+      "step": 6810
+    },
+    {
+      "epoch": 1.5192951148784295,
+      "grad_norm": 0.18906736373901367,
+      "learning_rate": 9.916454234945752e-06,
+      "loss": 0.4675,
+      "step": 6811
+    },
+    {
+      "epoch": 1.519518179790319,
+      "grad_norm": 0.19310304522514343,
+      "learning_rate": 9.914100890205124e-06,
+      "loss": 0.4689,
+      "step": 6812
+    },
+    {
+      "epoch": 1.5197412447022085,
+      "grad_norm": 0.17019645869731903,
+      "learning_rate": 9.911747550222134e-06,
+      "loss": 0.4309,
+      "step": 6813
+    },
+    {
+      "epoch": 1.5199643096140978,
+      "grad_norm": 0.16500838100910187,
+      "learning_rate": 9.909394215127118e-06,
+      "loss": 0.4364,
+      "step": 6814
+    },
+    {
+      "epoch": 1.520187374525987,
+      "grad_norm": 0.1711840033531189,
+      "learning_rate": 9.90704088505042e-06,
+      "loss": 0.4439,
+      "step": 6815
+    },
+    {
+      "epoch": 1.5204104394378763,
+      "grad_norm": 0.17483307421207428,
+      "learning_rate": 9.904687560122381e-06,
+      "loss": 0.493,
+      "step": 6816
+    },
+    {
+      "epoch": 1.5206335043497656,
+      "grad_norm": 0.17699526250362396,
+      "learning_rate": 9.90233424047335e-06,
+      "loss": 0.4446,
+      "step": 6817
+    },
+    {
+      "epoch": 1.5208565692616551,
+      "grad_norm": 0.18029628694057465,
+      "learning_rate": 9.89998092623366e-06,
+      "loss": 0.4676,
+      "step": 6818
+    },
+    {
+      "epoch": 1.5210796341735446,
+      "grad_norm": 0.17293018102645874,
+      "learning_rate": 9.897627617533653e-06,
+      "loss": 0.4501,
+      "step": 6819
+    },
+    {
+      "epoch": 1.521302699085434,
+      "grad_norm": 0.17556585371494293,
+      "learning_rate": 9.895274314503674e-06,
+      "loss": 0.4454,
+      "step": 6820
+    },
+    {
+      "epoch": 1.5215257639973232,
+      "grad_norm": 0.1644999384880066,
+      "learning_rate": 9.892921017274062e-06,
+      "loss": 0.4197,
+      "step": 6821
+    },
+    {
+      "epoch": 1.5217488289092125,
+      "grad_norm": 0.17591945827007294,
+      "learning_rate": 9.890567725975159e-06,
+      "loss": 0.4593,
+      "step": 6822
+    },
+    {
+      "epoch": 1.521971893821102,
+      "grad_norm": 0.1673215925693512,
+      "learning_rate": 9.888214440737302e-06,
+      "loss": 0.4442,
+      "step": 6823
+    },
+    {
+      "epoch": 1.5221949587329913,
+      "grad_norm": 0.16372933983802795,
+      "learning_rate": 9.885861161690832e-06,
+      "loss": 0.4252,
+      "step": 6824
+    },
+    {
+      "epoch": 1.5224180236448808,
+      "grad_norm": 0.17661777138710022,
+      "learning_rate": 9.88350788896609e-06,
+      "loss": 0.4593,
+      "step": 6825
+    },
+    {
+      "epoch": 1.52264108855677,
+      "grad_norm": 0.17217707633972168,
+      "learning_rate": 9.881154622693415e-06,
+      "loss": 0.4418,
+      "step": 6826
+    },
+    {
+      "epoch": 1.5228641534686593,
+      "grad_norm": 0.1641775518655777,
+      "learning_rate": 9.878801363003143e-06,
+      "loss": 0.4624,
+      "step": 6827
+    },
+    {
+      "epoch": 1.5230872183805486,
+      "grad_norm": 0.17330925166606903,
+      "learning_rate": 9.876448110025615e-06,
+      "loss": 0.4916,
+      "step": 6828
+    },
+    {
+      "epoch": 1.5233102832924381,
+      "grad_norm": 0.17335113883018494,
+      "learning_rate": 9.874094863891166e-06,
+      "loss": 0.4581,
+      "step": 6829
+    },
+    {
+      "epoch": 1.5235333482043276,
+      "grad_norm": 0.18039341270923615,
+      "learning_rate": 9.871741624730134e-06,
+      "loss": 0.4441,
+      "step": 6830
+    },
+    {
+      "epoch": 1.523756413116217,
+      "grad_norm": 0.18012328445911407,
+      "learning_rate": 9.869388392672861e-06,
+      "loss": 0.4618,
+      "step": 6831
+    },
+    {
+      "epoch": 1.5239794780281062,
+      "grad_norm": 0.17947714030742645,
+      "learning_rate": 9.867035167849678e-06,
+      "loss": 0.4905,
+      "step": 6832
+    },
+    {
+      "epoch": 1.5242025429399955,
+      "grad_norm": 0.18116816878318787,
+      "learning_rate": 9.864681950390927e-06,
+      "loss": 0.4649,
+      "step": 6833
+    },
+    {
+      "epoch": 1.5244256078518847,
+      "grad_norm": 0.1693393439054489,
+      "learning_rate": 9.862328740426938e-06,
+      "loss": 0.458,
+      "step": 6834
+    },
+    {
+      "epoch": 1.5246486727637742,
+      "grad_norm": 0.17401763796806335,
+      "learning_rate": 9.85997553808805e-06,
+      "loss": 0.4719,
+      "step": 6835
+    },
+    {
+      "epoch": 1.5248717376756638,
+      "grad_norm": 0.17534461617469788,
+      "learning_rate": 9.857622343504597e-06,
+      "loss": 0.4504,
+      "step": 6836
+    },
+    {
+      "epoch": 1.525094802587553,
+      "grad_norm": 0.17880570888519287,
+      "learning_rate": 9.855269156806916e-06,
+      "loss": 0.4832,
+      "step": 6837
+    },
+    {
+      "epoch": 1.5253178674994423,
+      "grad_norm": 0.16782738268375397,
+      "learning_rate": 9.852915978125337e-06,
+      "loss": 0.46,
+      "step": 6838
+    },
+    {
+      "epoch": 1.5255409324113316,
+      "grad_norm": 0.18301822245121002,
+      "learning_rate": 9.850562807590199e-06,
+      "loss": 0.4459,
+      "step": 6839
+    },
+    {
+      "epoch": 1.525763997323221,
+      "grad_norm": 0.1692187637090683,
+      "learning_rate": 9.84820964533183e-06,
+      "loss": 0.4465,
+      "step": 6840
+    },
+    {
+      "epoch": 1.5259870622351104,
+      "grad_norm": 0.17732323706150055,
+      "learning_rate": 9.845856491480569e-06,
+      "loss": 0.4374,
+      "step": 6841
+    },
+    {
+      "epoch": 1.5262101271469999,
+      "grad_norm": 0.17886339128017426,
+      "learning_rate": 9.843503346166741e-06,
+      "loss": 0.445,
+      "step": 6842
+    },
+    {
+      "epoch": 1.5264331920588892,
+      "grad_norm": 0.17064787447452545,
+      "learning_rate": 9.841150209520686e-06,
+      "loss": 0.4343,
+      "step": 6843
+    },
+    {
+      "epoch": 1.5266562569707784,
+      "grad_norm": 0.18215104937553406,
+      "learning_rate": 9.83879708167273e-06,
+      "loss": 0.4515,
+      "step": 6844
+    },
+    {
+      "epoch": 1.5268793218826677,
+      "grad_norm": 0.17810726165771484,
+      "learning_rate": 9.836443962753205e-06,
+      "loss": 0.4746,
+      "step": 6845
+    },
+    {
+      "epoch": 1.5271023867945572,
+      "grad_norm": 0.17213033139705658,
+      "learning_rate": 9.834090852892442e-06,
+      "loss": 0.4731,
+      "step": 6846
+    },
+    {
+      "epoch": 1.5273254517064467,
+      "grad_norm": 0.18000519275665283,
+      "learning_rate": 9.831737752220772e-06,
+      "loss": 0.4725,
+      "step": 6847
+    },
+    {
+      "epoch": 1.527548516618336,
+      "grad_norm": 0.17720220983028412,
+      "learning_rate": 9.829384660868523e-06,
+      "loss": 0.449,
+      "step": 6848
+    },
+    {
+      "epoch": 1.5277715815302253,
+      "grad_norm": 0.1757756471633911,
+      "learning_rate": 9.827031578966026e-06,
+      "loss": 0.4492,
+      "step": 6849
+    },
+    {
+      "epoch": 1.5279946464421146,
+      "grad_norm": 0.18418079614639282,
+      "learning_rate": 9.824678506643607e-06,
+      "loss": 0.4514,
+      "step": 6850
+    },
+    {
+      "epoch": 1.5282177113540039,
+      "grad_norm": 0.16821002960205078,
+      "learning_rate": 9.822325444031594e-06,
+      "loss": 0.4331,
+      "step": 6851
+    },
+    {
+      "epoch": 1.5284407762658934,
+      "grad_norm": 0.16792722046375275,
+      "learning_rate": 9.819972391260318e-06,
+      "loss": 0.4373,
+      "step": 6852
+    },
+    {
+      "epoch": 1.5286638411777829,
+      "grad_norm": 0.18326304852962494,
+      "learning_rate": 9.8176193484601e-06,
+      "loss": 0.4577,
+      "step": 6853
+    },
+    {
+      "epoch": 1.5288869060896721,
+      "grad_norm": 0.1714770793914795,
+      "learning_rate": 9.815266315761275e-06,
+      "loss": 0.437,
+      "step": 6854
+    },
+    {
+      "epoch": 1.5291099710015614,
+      "grad_norm": 0.16261571645736694,
+      "learning_rate": 9.81291329329416e-06,
+      "loss": 0.406,
+      "step": 6855
+    },
+    {
+      "epoch": 1.5293330359134507,
+      "grad_norm": 0.17615249752998352,
+      "learning_rate": 9.810560281189085e-06,
+      "loss": 0.4704,
+      "step": 6856
+    },
+    {
+      "epoch": 1.5295561008253402,
+      "grad_norm": 0.1748465746641159,
+      "learning_rate": 9.80820727957637e-06,
+      "loss": 0.4763,
+      "step": 6857
+    },
+    {
+      "epoch": 1.5297791657372295,
+      "grad_norm": 0.1718483865261078,
+      "learning_rate": 9.805854288586346e-06,
+      "loss": 0.4574,
+      "step": 6858
+    },
+    {
+      "epoch": 1.530002230649119,
+      "grad_norm": 0.17926479876041412,
+      "learning_rate": 9.803501308349332e-06,
+      "loss": 0.4556,
+      "step": 6859
+    },
+    {
+      "epoch": 1.5302252955610083,
+      "grad_norm": 0.17003634572029114,
+      "learning_rate": 9.801148338995652e-06,
+      "loss": 0.4553,
+      "step": 6860
+    },
+    {
+      "epoch": 1.5304483604728976,
+      "grad_norm": 0.17237353324890137,
+      "learning_rate": 9.798795380655625e-06,
+      "loss": 0.4521,
+      "step": 6861
+    },
+    {
+      "epoch": 1.5306714253847868,
+      "grad_norm": 0.18084143102169037,
+      "learning_rate": 9.796442433459578e-06,
+      "loss": 0.4592,
+      "step": 6862
+    },
+    {
+      "epoch": 1.5308944902966763,
+      "grad_norm": 0.18167221546173096,
+      "learning_rate": 9.794089497537827e-06,
+      "loss": 0.4735,
+      "step": 6863
+    },
+    {
+      "epoch": 1.5311175552085659,
+      "grad_norm": 0.18819572031497955,
+      "learning_rate": 9.791736573020697e-06,
+      "loss": 0.4906,
+      "step": 6864
+    },
+    {
+      "epoch": 1.5313406201204551,
+      "grad_norm": 0.17070408165454865,
+      "learning_rate": 9.789383660038506e-06,
+      "loss": 0.4242,
+      "step": 6865
+    },
+    {
+      "epoch": 1.5315636850323444,
+      "grad_norm": 0.17565836012363434,
+      "learning_rate": 9.787030758721573e-06,
+      "loss": 0.4431,
+      "step": 6866
+    },
+    {
+      "epoch": 1.5317867499442337,
+      "grad_norm": 0.174460306763649,
+      "learning_rate": 9.784677869200215e-06,
+      "loss": 0.4487,
+      "step": 6867
+    },
+    {
+      "epoch": 1.532009814856123,
+      "grad_norm": 0.17642372846603394,
+      "learning_rate": 9.782324991604751e-06,
+      "loss": 0.469,
+      "step": 6868
+    },
+    {
+      "epoch": 1.5322328797680125,
+      "grad_norm": 0.17292480170726776,
+      "learning_rate": 9.779972126065497e-06,
+      "loss": 0.4408,
+      "step": 6869
+    },
+    {
+      "epoch": 1.532455944679902,
+      "grad_norm": 0.17120680212974548,
+      "learning_rate": 9.777619272712774e-06,
+      "loss": 0.474,
+      "step": 6870
+    },
+    {
+      "epoch": 1.5326790095917913,
+      "grad_norm": 0.16610749065876007,
+      "learning_rate": 9.77526643167689e-06,
+      "loss": 0.4486,
+      "step": 6871
+    },
+    {
+      "epoch": 1.5329020745036805,
+      "grad_norm": 0.17615914344787598,
+      "learning_rate": 9.772913603088166e-06,
+      "loss": 0.4654,
+      "step": 6872
+    },
+    {
+      "epoch": 1.5331251394155698,
+      "grad_norm": 0.1731085181236267,
+      "learning_rate": 9.770560787076918e-06,
+      "loss": 0.4459,
+      "step": 6873
+    },
+    {
+      "epoch": 1.5333482043274593,
+      "grad_norm": 0.17174161970615387,
+      "learning_rate": 9.768207983773455e-06,
+      "loss": 0.4443,
+      "step": 6874
+    },
+    {
+      "epoch": 1.5335712692393486,
+      "grad_norm": 0.17707736790180206,
+      "learning_rate": 9.765855193308092e-06,
+      "loss": 0.4707,
+      "step": 6875
+    },
+    {
+      "epoch": 1.5337943341512381,
+      "grad_norm": 0.17120012640953064,
+      "learning_rate": 9.763502415811141e-06,
+      "loss": 0.4599,
+      "step": 6876
+    },
+    {
+      "epoch": 1.5340173990631274,
+      "grad_norm": 0.18111848831176758,
+      "learning_rate": 9.761149651412915e-06,
+      "loss": 0.4354,
+      "step": 6877
+    },
+    {
+      "epoch": 1.5342404639750167,
+      "grad_norm": 0.17889343202114105,
+      "learning_rate": 9.758796900243723e-06,
+      "loss": 0.4521,
+      "step": 6878
+    },
+    {
+      "epoch": 1.534463528886906,
+      "grad_norm": 0.1871626377105713,
+      "learning_rate": 9.756444162433876e-06,
+      "loss": 0.4711,
+      "step": 6879
+    },
+    {
+      "epoch": 1.5346865937987955,
+      "grad_norm": 0.17596866190433502,
+      "learning_rate": 9.754091438113683e-06,
+      "loss": 0.4566,
+      "step": 6880
+    },
+    {
+      "epoch": 1.534909658710685,
+      "grad_norm": 0.17852075397968292,
+      "learning_rate": 9.751738727413453e-06,
+      "loss": 0.4763,
+      "step": 6881
+    },
+    {
+      "epoch": 1.5351327236225742,
+      "grad_norm": 0.191118523478508,
+      "learning_rate": 9.749386030463493e-06,
+      "loss": 0.4396,
+      "step": 6882
+    },
+    {
+      "epoch": 1.5353557885344635,
+      "grad_norm": 0.19689539074897766,
+      "learning_rate": 9.747033347394113e-06,
+      "loss": 0.4421,
+      "step": 6883
+    },
+    {
+      "epoch": 1.5355788534463528,
+      "grad_norm": 0.16877755522727966,
+      "learning_rate": 9.744680678335614e-06,
+      "loss": 0.4185,
+      "step": 6884
+    },
+    {
+      "epoch": 1.535801918358242,
+      "grad_norm": 0.1971082091331482,
+      "learning_rate": 9.74232802341831e-06,
+      "loss": 0.4537,
+      "step": 6885
+    },
+    {
+      "epoch": 1.5360249832701316,
+      "grad_norm": 0.1687203049659729,
+      "learning_rate": 9.739975382772498e-06,
+      "loss": 0.4524,
+      "step": 6886
+    },
+    {
+      "epoch": 1.536248048182021,
+      "grad_norm": 0.169671893119812,
+      "learning_rate": 9.737622756528485e-06,
+      "loss": 0.4467,
+      "step": 6887
+    },
+    {
+      "epoch": 1.5364711130939104,
+      "grad_norm": 0.1722356528043747,
+      "learning_rate": 9.735270144816574e-06,
+      "loss": 0.4771,
+      "step": 6888
+    },
+    {
+      "epoch": 1.5366941780057997,
+      "grad_norm": 0.17518217861652374,
+      "learning_rate": 9.732917547767069e-06,
+      "loss": 0.4543,
+      "step": 6889
+    },
+    {
+      "epoch": 1.536917242917689,
+      "grad_norm": 0.17109224200248718,
+      "learning_rate": 9.730564965510269e-06,
+      "loss": 0.4401,
+      "step": 6890
+    },
+    {
+      "epoch": 1.5371403078295784,
+      "grad_norm": 0.1737949699163437,
+      "learning_rate": 9.728212398176476e-06,
+      "loss": 0.4415,
+      "step": 6891
+    },
+    {
+      "epoch": 1.5373633727414677,
+      "grad_norm": 0.16947439312934875,
+      "learning_rate": 9.72585984589599e-06,
+      "loss": 0.4723,
+      "step": 6892
+    },
+    {
+      "epoch": 1.5375864376533572,
+      "grad_norm": 0.1718679517507553,
+      "learning_rate": 9.723507308799109e-06,
+      "loss": 0.4842,
+      "step": 6893
+    },
+    {
+      "epoch": 1.5378095025652465,
+      "grad_norm": 0.1752775013446808,
+      "learning_rate": 9.721154787016135e-06,
+      "loss": 0.4461,
+      "step": 6894
+    },
+    {
+      "epoch": 1.5380325674771358,
+      "grad_norm": 0.17731130123138428,
+      "learning_rate": 9.71880228067736e-06,
+      "loss": 0.4566,
+      "step": 6895
+    },
+    {
+      "epoch": 1.538255632389025,
+      "grad_norm": 0.22506393492221832,
+      "learning_rate": 9.716449789913087e-06,
+      "loss": 0.4486,
+      "step": 6896
+    },
+    {
+      "epoch": 1.5384786973009146,
+      "grad_norm": 0.17493079602718353,
+      "learning_rate": 9.714097314853606e-06,
+      "loss": 0.439,
+      "step": 6897
+    },
+    {
+      "epoch": 1.538701762212804,
+      "grad_norm": 0.16993321478366852,
+      "learning_rate": 9.711744855629218e-06,
+      "loss": 0.4549,
+      "step": 6898
+    },
+    {
+      "epoch": 1.5389248271246934,
+      "grad_norm": 0.17897512018680573,
+      "learning_rate": 9.709392412370212e-06,
+      "loss": 0.4452,
+      "step": 6899
+    },
+    {
+      "epoch": 1.5391478920365826,
+      "grad_norm": 0.16903938353061676,
+      "learning_rate": 9.707039985206878e-06,
+      "loss": 0.4665,
+      "step": 6900
+    },
+    {
+      "epoch": 1.539370956948472,
+      "grad_norm": 0.17775236070156097,
+      "learning_rate": 9.704687574269515e-06,
+      "loss": 0.4702,
+      "step": 6901
+    },
+    {
+      "epoch": 1.5395940218603612,
+      "grad_norm": 0.17940306663513184,
+      "learning_rate": 9.70233517968841e-06,
+      "loss": 0.4671,
+      "step": 6902
+    },
+    {
+      "epoch": 1.5398170867722507,
+      "grad_norm": 0.163747638463974,
+      "learning_rate": 9.699982801593857e-06,
+      "loss": 0.4533,
+      "step": 6903
+    },
+    {
+      "epoch": 1.5400401516841402,
+      "grad_norm": 0.1675698310136795,
+      "learning_rate": 9.697630440116141e-06,
+      "loss": 0.4422,
+      "step": 6904
+    },
+    {
+      "epoch": 1.5402632165960295,
+      "grad_norm": 0.1684923619031906,
+      "learning_rate": 9.695278095385558e-06,
+      "loss": 0.4439,
+      "step": 6905
+    },
+    {
+      "epoch": 1.5404862815079188,
+      "grad_norm": 0.16581715643405914,
+      "learning_rate": 9.692925767532386e-06,
+      "loss": 0.4288,
+      "step": 6906
+    },
+    {
+      "epoch": 1.540709346419808,
+      "grad_norm": 0.1779187023639679,
+      "learning_rate": 9.690573456686919e-06,
+      "loss": 0.4589,
+      "step": 6907
+    },
+    {
+      "epoch": 1.5409324113316976,
+      "grad_norm": 0.17237865924835205,
+      "learning_rate": 9.688221162979438e-06,
+      "loss": 0.4595,
+      "step": 6908
+    },
+    {
+      "epoch": 1.5411554762435868,
+      "grad_norm": 0.17078691720962524,
+      "learning_rate": 9.685868886540233e-06,
+      "loss": 0.4417,
+      "step": 6909
+    },
+    {
+      "epoch": 1.5413785411554763,
+      "grad_norm": 0.16685281693935394,
+      "learning_rate": 9.683516627499583e-06,
+      "loss": 0.4311,
+      "step": 6910
+    },
+    {
+      "epoch": 1.5416016060673656,
+      "grad_norm": 0.17877590656280518,
+      "learning_rate": 9.681164385987776e-06,
+      "loss": 0.4475,
+      "step": 6911
+    },
+    {
+      "epoch": 1.541824670979255,
+      "grad_norm": 0.17893154919147491,
+      "learning_rate": 9.678812162135087e-06,
+      "loss": 0.4585,
+      "step": 6912
+    },
+    {
+      "epoch": 1.5420477358911442,
+      "grad_norm": 0.17963162064552307,
+      "learning_rate": 9.676459956071804e-06,
+      "loss": 0.4643,
+      "step": 6913
+    },
+    {
+      "epoch": 1.5422708008030337,
+      "grad_norm": 0.16504304111003876,
+      "learning_rate": 9.674107767928202e-06,
+      "loss": 0.4181,
+      "step": 6914
+    },
+    {
+      "epoch": 1.5424938657149232,
+      "grad_norm": 0.17094488441944122,
+      "learning_rate": 9.671755597834563e-06,
+      "loss": 0.4739,
+      "step": 6915
+    },
+    {
+      "epoch": 1.5427169306268125,
+      "grad_norm": 0.1811668872833252,
+      "learning_rate": 9.669403445921162e-06,
+      "loss": 0.4549,
+      "step": 6916
+    },
+    {
+      "epoch": 1.5429399955387018,
+      "grad_norm": 0.17800799012184143,
+      "learning_rate": 9.667051312318277e-06,
+      "loss": 0.4655,
+      "step": 6917
+    },
+    {
+      "epoch": 1.543163060450591,
+      "grad_norm": 0.1688285917043686,
+      "learning_rate": 9.664699197156188e-06,
+      "loss": 0.4442,
+      "step": 6918
+    },
+    {
+      "epoch": 1.5433861253624803,
+      "grad_norm": 0.17787446081638336,
+      "learning_rate": 9.662347100565163e-06,
+      "loss": 0.4689,
+      "step": 6919
+    },
+    {
+      "epoch": 1.5436091902743698,
+      "grad_norm": 0.1820237636566162,
+      "learning_rate": 9.659995022675482e-06,
+      "loss": 0.453,
+      "step": 6920
+    },
+    {
+      "epoch": 1.5438322551862593,
+      "grad_norm": 0.18955709040164948,
+      "learning_rate": 9.657642963617414e-06,
+      "loss": 0.4653,
+      "step": 6921
+    },
+    {
+      "epoch": 1.5440553200981486,
+      "grad_norm": 0.17111030220985413,
+      "learning_rate": 9.655290923521235e-06,
+      "loss": 0.4436,
+      "step": 6922
+    },
+    {
+      "epoch": 1.544278385010038,
+      "grad_norm": 0.17172935605049133,
+      "learning_rate": 9.65293890251721e-06,
+      "loss": 0.4361,
+      "step": 6923
+    },
+    {
+      "epoch": 1.5445014499219272,
+      "grad_norm": 0.18299220502376556,
+      "learning_rate": 9.650586900735612e-06,
+      "loss": 0.4775,
+      "step": 6924
+    },
+    {
+      "epoch": 1.5447245148338167,
+      "grad_norm": 0.18452848494052887,
+      "learning_rate": 9.648234918306709e-06,
+      "loss": 0.4646,
+      "step": 6925
+    },
+    {
+      "epoch": 1.544947579745706,
+      "grad_norm": 0.17104355990886688,
+      "learning_rate": 9.645882955360771e-06,
+      "loss": 0.454,
+      "step": 6926
+    },
+    {
+      "epoch": 1.5451706446575955,
+      "grad_norm": 0.18376219272613525,
+      "learning_rate": 9.643531012028061e-06,
+      "loss": 0.4683,
+      "step": 6927
+    },
+    {
+      "epoch": 1.5453937095694847,
+      "grad_norm": 0.17010129988193512,
+      "learning_rate": 9.641179088438845e-06,
+      "loss": 0.4417,
+      "step": 6928
+    },
+    {
+      "epoch": 1.545616774481374,
+      "grad_norm": 0.167456716299057,
+      "learning_rate": 9.63882718472339e-06,
+      "loss": 0.4114,
+      "step": 6929
+    },
+    {
+      "epoch": 1.5458398393932633,
+      "grad_norm": 0.17169076204299927,
+      "learning_rate": 9.636475301011957e-06,
+      "loss": 0.4526,
+      "step": 6930
+    },
+    {
+      "epoch": 1.5460629043051528,
+      "grad_norm": 0.16744768619537354,
+      "learning_rate": 9.634123437434807e-06,
+      "loss": 0.4615,
+      "step": 6931
+    },
+    {
+      "epoch": 1.5462859692170423,
+      "grad_norm": 0.1830861121416092,
+      "learning_rate": 9.631771594122205e-06,
+      "loss": 0.4554,
+      "step": 6932
+    },
+    {
+      "epoch": 1.5465090341289316,
+      "grad_norm": 0.17061203718185425,
+      "learning_rate": 9.629419771204405e-06,
+      "loss": 0.4317,
+      "step": 6933
+    },
+    {
+      "epoch": 1.5467320990408209,
+      "grad_norm": 0.1750810146331787,
+      "learning_rate": 9.627067968811671e-06,
+      "loss": 0.4648,
+      "step": 6934
+    },
+    {
+      "epoch": 1.5469551639527102,
+      "grad_norm": 0.16664697229862213,
+      "learning_rate": 9.624716187074258e-06,
+      "loss": 0.4555,
+      "step": 6935
+    },
+    {
+      "epoch": 1.5471782288645994,
+      "grad_norm": 0.17463943362236023,
+      "learning_rate": 9.622364426122424e-06,
+      "loss": 0.4334,
+      "step": 6936
+    },
+    {
+      "epoch": 1.547401293776489,
+      "grad_norm": 0.182994544506073,
+      "learning_rate": 9.620012686086423e-06,
+      "loss": 0.4657,
+      "step": 6937
+    },
+    {
+      "epoch": 1.5476243586883784,
+      "grad_norm": 0.18054361641407013,
+      "learning_rate": 9.61766096709651e-06,
+      "loss": 0.4595,
+      "step": 6938
+    },
+    {
+      "epoch": 1.5478474236002677,
+      "grad_norm": 0.17179542779922485,
+      "learning_rate": 9.615309269282938e-06,
+      "loss": 0.473,
+      "step": 6939
+    },
+    {
+      "epoch": 1.548070488512157,
+      "grad_norm": 0.17718704044818878,
+      "learning_rate": 9.612957592775957e-06,
+      "loss": 0.4708,
+      "step": 6940
+    },
+    {
+      "epoch": 1.5482935534240463,
+      "grad_norm": 0.17120683193206787,
+      "learning_rate": 9.610605937705823e-06,
+      "loss": 0.4432,
+      "step": 6941
+    },
+    {
+      "epoch": 1.5485166183359358,
+      "grad_norm": 0.17965561151504517,
+      "learning_rate": 9.608254304202776e-06,
+      "loss": 0.439,
+      "step": 6942
+    },
+    {
+      "epoch": 1.548739683247825,
+      "grad_norm": 0.17135334014892578,
+      "learning_rate": 9.605902692397074e-06,
+      "loss": 0.4597,
+      "step": 6943
+    },
+    {
+      "epoch": 1.5489627481597146,
+      "grad_norm": 0.1764613837003708,
+      "learning_rate": 9.603551102418958e-06,
+      "loss": 0.4478,
+      "step": 6944
+    },
+    {
+      "epoch": 1.5491858130716039,
+      "grad_norm": 0.16766582429409027,
+      "learning_rate": 9.601199534398678e-06,
+      "loss": 0.425,
+      "step": 6945
+    },
+    {
+      "epoch": 1.5494088779834931,
+      "grad_norm": 0.1771164983510971,
+      "learning_rate": 9.598847988466472e-06,
+      "loss": 0.457,
+      "step": 6946
+    },
+    {
+      "epoch": 1.5496319428953824,
+      "grad_norm": 0.1649523228406906,
+      "learning_rate": 9.596496464752593e-06,
+      "loss": 0.4563,
+      "step": 6947
+    },
+    {
+      "epoch": 1.549855007807272,
+      "grad_norm": 0.17230331897735596,
+      "learning_rate": 9.594144963387274e-06,
+      "loss": 0.4587,
+      "step": 6948
+    },
+    {
+      "epoch": 1.5500780727191614,
+      "grad_norm": 0.17549771070480347,
+      "learning_rate": 9.591793484500762e-06,
+      "loss": 0.4719,
+      "step": 6949
+    },
+    {
+      "epoch": 1.5503011376310507,
+      "grad_norm": 0.17885257303714752,
+      "learning_rate": 9.589442028223292e-06,
+      "loss": 0.4479,
+      "step": 6950
+    },
+    {
+      "epoch": 1.55052420254294,
+      "grad_norm": 0.1839398443698883,
+      "learning_rate": 9.587090594685108e-06,
+      "loss": 0.4631,
+      "step": 6951
+    },
+    {
+      "epoch": 1.5507472674548293,
+      "grad_norm": 0.17899379134178162,
+      "learning_rate": 9.584739184016442e-06,
+      "loss": 0.4705,
+      "step": 6952
+    },
+    {
+      "epoch": 1.5509703323667186,
+      "grad_norm": 0.17247112095355988,
+      "learning_rate": 9.582387796347534e-06,
+      "loss": 0.4653,
+      "step": 6953
+    },
+    {
+      "epoch": 1.551193397278608,
+      "grad_norm": 0.17476417124271393,
+      "learning_rate": 9.580036431808614e-06,
+      "loss": 0.447,
+      "step": 6954
+    },
+    {
+      "epoch": 1.5514164621904976,
+      "grad_norm": 0.17558352649211884,
+      "learning_rate": 9.57768509052992e-06,
+      "loss": 0.4607,
+      "step": 6955
+    },
+    {
+      "epoch": 1.5516395271023868,
+      "grad_norm": 0.1909688413143158,
+      "learning_rate": 9.575333772641682e-06,
+      "loss": 0.4619,
+      "step": 6956
+    },
+    {
+      "epoch": 1.5518625920142761,
+      "grad_norm": 0.1743474304676056,
+      "learning_rate": 9.572982478274128e-06,
+      "loss": 0.443,
+      "step": 6957
+    },
+    {
+      "epoch": 1.5520856569261654,
+      "grad_norm": 0.18644817173480988,
+      "learning_rate": 9.570631207557495e-06,
+      "loss": 0.4533,
+      "step": 6958
+    },
+    {
+      "epoch": 1.552308721838055,
+      "grad_norm": 0.17602680623531342,
+      "learning_rate": 9.568279960622004e-06,
+      "loss": 0.4394,
+      "step": 6959
+    },
+    {
+      "epoch": 1.5525317867499442,
+      "grad_norm": 0.17043206095695496,
+      "learning_rate": 9.565928737597886e-06,
+      "loss": 0.4551,
+      "step": 6960
+    },
+    {
+      "epoch": 1.5527548516618337,
+      "grad_norm": 0.18989066779613495,
+      "learning_rate": 9.563577538615363e-06,
+      "loss": 0.471,
+      "step": 6961
+    },
+    {
+      "epoch": 1.552977916573723,
+      "grad_norm": 0.19321848452091217,
+      "learning_rate": 9.561226363804663e-06,
+      "loss": 0.4739,
+      "step": 6962
+    },
+    {
+      "epoch": 1.5532009814856123,
+      "grad_norm": 0.17869152128696442,
+      "learning_rate": 9.558875213296006e-06,
+      "loss": 0.4409,
+      "step": 6963
+    },
+    {
+      "epoch": 1.5534240463975015,
+      "grad_norm": 0.17526346445083618,
+      "learning_rate": 9.556524087219616e-06,
+      "loss": 0.4514,
+      "step": 6964
+    },
+    {
+      "epoch": 1.553647111309391,
+      "grad_norm": 0.17142999172210693,
+      "learning_rate": 9.55417298570571e-06,
+      "loss": 0.4427,
+      "step": 6965
+    },
+    {
+      "epoch": 1.5538701762212805,
+      "grad_norm": 0.17005646228790283,
+      "learning_rate": 9.551821908884507e-06,
+      "loss": 0.4434,
+      "step": 6966
+    },
+    {
+      "epoch": 1.5540932411331698,
+      "grad_norm": 0.1687459945678711,
+      "learning_rate": 9.549470856886228e-06,
+      "loss": 0.4651,
+      "step": 6967
+    },
+    {
+      "epoch": 1.554316306045059,
+      "grad_norm": 0.1726951152086258,
+      "learning_rate": 9.547119829841088e-06,
+      "loss": 0.4582,
+      "step": 6968
+    },
+    {
+      "epoch": 1.5545393709569484,
+      "grad_norm": 0.17231762409210205,
+      "learning_rate": 9.544768827879297e-06,
+      "loss": 0.4472,
+      "step": 6969
+    },
+    {
+      "epoch": 1.5547624358688377,
+      "grad_norm": 0.179609015583992,
+      "learning_rate": 9.542417851131076e-06,
+      "loss": 0.4563,
+      "step": 6970
+    },
+    {
+      "epoch": 1.5549855007807272,
+      "grad_norm": 0.17300112545490265,
+      "learning_rate": 9.540066899726627e-06,
+      "loss": 0.4245,
+      "step": 6971
+    },
+    {
+      "epoch": 1.5552085656926167,
+      "grad_norm": 0.19064456224441528,
+      "learning_rate": 9.537715973796172e-06,
+      "loss": 0.4519,
+      "step": 6972
+    },
+    {
+      "epoch": 1.555431630604506,
+      "grad_norm": 0.1847352236509323,
+      "learning_rate": 9.535365073469911e-06,
+      "loss": 0.4784,
+      "step": 6973
+    },
+    {
+      "epoch": 1.5556546955163952,
+      "grad_norm": 0.18160337209701538,
+      "learning_rate": 9.533014198878057e-06,
+      "loss": 0.4533,
+      "step": 6974
+    },
+    {
+      "epoch": 1.5558777604282845,
+      "grad_norm": 0.1749866008758545,
+      "learning_rate": 9.530663350150812e-06,
+      "loss": 0.4607,
+      "step": 6975
+    },
+    {
+      "epoch": 1.556100825340174,
+      "grad_norm": 0.1792685091495514,
+      "learning_rate": 9.528312527418384e-06,
+      "loss": 0.4598,
+      "step": 6976
+    },
+    {
+      "epoch": 1.5563238902520633,
+      "grad_norm": 0.16535772383213043,
+      "learning_rate": 9.525961730810974e-06,
+      "loss": 0.4733,
+      "step": 6977
+    },
+    {
+      "epoch": 1.5565469551639528,
+      "grad_norm": 0.17856188118457794,
+      "learning_rate": 9.523610960458786e-06,
+      "loss": 0.4724,
+      "step": 6978
+    },
+    {
+      "epoch": 1.556770020075842,
+      "grad_norm": 0.1673804074525833,
+      "learning_rate": 9.521260216492025e-06,
+      "loss": 0.4281,
+      "step": 6979
+    },
+    {
+      "epoch": 1.5569930849877314,
+      "grad_norm": 0.17213106155395508,
+      "learning_rate": 9.518909499040878e-06,
+      "loss": 0.4652,
+      "step": 6980
+    },
+    {
+      "epoch": 1.5572161498996206,
+      "grad_norm": 0.1793312132358551,
+      "learning_rate": 9.51655880823555e-06,
+      "loss": 0.4645,
+      "step": 6981
+    },
+    {
+      "epoch": 1.5574392148115102,
+      "grad_norm": 0.17523737251758575,
+      "learning_rate": 9.514208144206237e-06,
+      "loss": 0.4459,
+      "step": 6982
+    },
+    {
+      "epoch": 1.5576622797233997,
+      "grad_norm": 0.17327353358268738,
+      "learning_rate": 9.511857507083132e-06,
+      "loss": 0.4531,
+      "step": 6983
+    },
+    {
+      "epoch": 1.557885344635289,
+      "grad_norm": 0.3019886016845703,
+      "learning_rate": 9.509506896996432e-06,
+      "loss": 0.4868,
+      "step": 6984
+    },
+    {
+      "epoch": 1.5581084095471782,
+      "grad_norm": 0.16475124657154083,
+      "learning_rate": 9.50715631407632e-06,
+      "loss": 0.4522,
+      "step": 6985
+    },
+    {
+      "epoch": 1.5583314744590675,
+      "grad_norm": 0.17215771973133087,
+      "learning_rate": 9.504805758452996e-06,
+      "loss": 0.4741,
+      "step": 6986
+    },
+    {
+      "epoch": 1.558554539370957,
+      "grad_norm": 0.17033734917640686,
+      "learning_rate": 9.502455230256641e-06,
+      "loss": 0.4406,
+      "step": 6987
+    },
+    {
+      "epoch": 1.5587776042828463,
+      "grad_norm": 0.17442695796489716,
+      "learning_rate": 9.500104729617447e-06,
+      "loss": 0.4473,
+      "step": 6988
+    },
+    {
+      "epoch": 1.5590006691947358,
+      "grad_norm": 0.17998719215393066,
+      "learning_rate": 9.497754256665596e-06,
+      "loss": 0.4747,
+      "step": 6989
+    },
+    {
+      "epoch": 1.559223734106625,
+      "grad_norm": 0.18134324252605438,
+      "learning_rate": 9.495403811531274e-06,
+      "loss": 0.457,
+      "step": 6990
+    },
+    {
+      "epoch": 1.5594467990185144,
+      "grad_norm": 0.16917896270751953,
+      "learning_rate": 9.493053394344663e-06,
+      "loss": 0.4701,
+      "step": 6991
+    },
+    {
+      "epoch": 1.5596698639304036,
+      "grad_norm": 0.18709082901477814,
+      "learning_rate": 9.490703005235945e-06,
+      "loss": 0.4577,
+      "step": 6992
+    },
+    {
+      "epoch": 1.5598929288422931,
+      "grad_norm": 0.17913807928562164,
+      "learning_rate": 9.488352644335296e-06,
+      "loss": 0.4391,
+      "step": 6993
+    },
+    {
+      "epoch": 1.5601159937541824,
+      "grad_norm": 0.24650318920612335,
+      "learning_rate": 9.486002311772898e-06,
+      "loss": 0.4355,
+      "step": 6994
+    },
+    {
+      "epoch": 1.560339058666072,
+      "grad_norm": 0.17876996099948883,
+      "learning_rate": 9.483652007678924e-06,
+      "loss": 0.4741,
+      "step": 6995
+    },
+    {
+      "epoch": 1.5605621235779612,
+      "grad_norm": 0.1874217391014099,
+      "learning_rate": 9.48130173218355e-06,
+      "loss": 0.4516,
+      "step": 6996
+    },
+    {
+      "epoch": 1.5607851884898505,
+      "grad_norm": 0.1787550449371338,
+      "learning_rate": 9.478951485416948e-06,
+      "loss": 0.4784,
+      "step": 6997
+    },
+    {
+      "epoch": 1.5610082534017398,
+      "grad_norm": 0.17699971795082092,
+      "learning_rate": 9.476601267509293e-06,
+      "loss": 0.4708,
+      "step": 6998
+    },
+    {
+      "epoch": 1.5612313183136293,
+      "grad_norm": 0.17713846266269684,
+      "learning_rate": 9.474251078590749e-06,
+      "loss": 0.4404,
+      "step": 6999
+    },
+    {
+      "epoch": 1.5614543832255188,
+      "grad_norm": 0.1823769062757492,
+      "learning_rate": 9.47190091879149e-06,
+      "loss": 0.428,
+      "step": 7000
+    },
+    {
+      "epoch": 1.561677448137408,
+      "grad_norm": 0.17455317080020905,
+      "learning_rate": 9.469550788241678e-06,
+      "loss": 0.4416,
+      "step": 7001
+    },
+    {
+      "epoch": 1.5619005130492973,
+      "grad_norm": 0.164621502161026,
+      "learning_rate": 9.467200687071478e-06,
+      "loss": 0.4429,
+      "step": 7002
+    },
+    {
+      "epoch": 1.5621235779611866,
+      "grad_norm": 0.1848084181547165,
+      "learning_rate": 9.464850615411059e-06,
+      "loss": 0.5004,
+      "step": 7003
+    },
+    {
+      "epoch": 1.5623466428730761,
+      "grad_norm": 0.20320458710193634,
+      "learning_rate": 9.462500573390578e-06,
+      "loss": 0.4704,
+      "step": 7004
+    },
+    {
+      "epoch": 1.5625697077849654,
+      "grad_norm": 0.16972646117210388,
+      "learning_rate": 9.460150561140196e-06,
+      "loss": 0.451,
+      "step": 7005
+    },
+    {
+      "epoch": 1.562792772696855,
+      "grad_norm": 0.16521897912025452,
+      "learning_rate": 9.457800578790072e-06,
+      "loss": 0.4457,
+      "step": 7006
+    },
+    {
+      "epoch": 1.5630158376087442,
+      "grad_norm": 0.18121238052845,
+      "learning_rate": 9.455450626470363e-06,
+      "loss": 0.4887,
+      "step": 7007
+    },
+    {
+      "epoch": 1.5632389025206335,
+      "grad_norm": 0.17698867619037628,
+      "learning_rate": 9.453100704311222e-06,
+      "loss": 0.4559,
+      "step": 7008
+    },
+    {
+      "epoch": 1.5634619674325227,
+      "grad_norm": 0.18639512360095978,
+      "learning_rate": 9.450750812442808e-06,
+      "loss": 0.4504,
+      "step": 7009
+    },
+    {
+      "epoch": 1.5636850323444123,
+      "grad_norm": 0.17181695997714996,
+      "learning_rate": 9.448400950995265e-06,
+      "loss": 0.4383,
+      "step": 7010
+    },
+    {
+      "epoch": 1.5639080972563015,
+      "grad_norm": 0.1778135597705841,
+      "learning_rate": 9.446051120098749e-06,
+      "loss": 0.4787,
+      "step": 7011
+    },
+    {
+      "epoch": 1.564131162168191,
+      "grad_norm": 0.17463596165180206,
+      "learning_rate": 9.443701319883404e-06,
+      "loss": 0.4513,
+      "step": 7012
+    },
+    {
+      "epoch": 1.5643542270800803,
+      "grad_norm": 0.1755073219537735,
+      "learning_rate": 9.441351550479382e-06,
+      "loss": 0.472,
+      "step": 7013
+    },
+    {
+      "epoch": 1.5645772919919696,
+      "grad_norm": 0.17314688861370087,
+      "learning_rate": 9.439001812016824e-06,
+      "loss": 0.4824,
+      "step": 7014
+    },
+    {
+      "epoch": 1.5648003569038589,
+      "grad_norm": 0.17931319773197174,
+      "learning_rate": 9.436652104625875e-06,
+      "loss": 0.487,
+      "step": 7015
+    },
+    {
+      "epoch": 1.5650234218157484,
+      "grad_norm": 0.1748175173997879,
+      "learning_rate": 9.434302428436674e-06,
+      "loss": 0.4322,
+      "step": 7016
+    },
+    {
+      "epoch": 1.5652464867276379,
+      "grad_norm": 0.1719309538602829,
+      "learning_rate": 9.431952783579365e-06,
+      "loss": 0.4584,
+      "step": 7017
+    },
+    {
+      "epoch": 1.5654695516395272,
+      "grad_norm": 0.1643352508544922,
+      "learning_rate": 9.429603170184085e-06,
+      "loss": 0.4204,
+      "step": 7018
+    },
+    {
+      "epoch": 1.5656926165514164,
+      "grad_norm": 0.17811031639575958,
+      "learning_rate": 9.427253588380968e-06,
+      "loss": 0.4827,
+      "step": 7019
+    },
+    {
+      "epoch": 1.5659156814633057,
+      "grad_norm": 0.18726174533367157,
+      "learning_rate": 9.42490403830015e-06,
+      "loss": 0.4466,
+      "step": 7020
+    },
+    {
+      "epoch": 1.5661387463751952,
+      "grad_norm": 0.1743079274892807,
+      "learning_rate": 9.422554520071763e-06,
+      "loss": 0.4595,
+      "step": 7021
+    },
+    {
+      "epoch": 1.5663618112870845,
+      "grad_norm": 0.16648052632808685,
+      "learning_rate": 9.42020503382594e-06,
+      "loss": 0.4399,
+      "step": 7022
+    },
+    {
+      "epoch": 1.566584876198974,
+      "grad_norm": 0.1800967901945114,
+      "learning_rate": 9.417855579692807e-06,
+      "loss": 0.4602,
+      "step": 7023
+    },
+    {
+      "epoch": 1.5668079411108633,
+      "grad_norm": 0.17495980858802795,
+      "learning_rate": 9.415506157802497e-06,
+      "loss": 0.45,
+      "step": 7024
+    },
+    {
+      "epoch": 1.5670310060227526,
+      "grad_norm": 0.1846977323293686,
+      "learning_rate": 9.41315676828513e-06,
+      "loss": 0.4677,
+      "step": 7025
+    },
+    {
+      "epoch": 1.5672540709346419,
+      "grad_norm": 0.1642792671918869,
+      "learning_rate": 9.410807411270834e-06,
+      "loss": 0.452,
+      "step": 7026
+    },
+    {
+      "epoch": 1.5674771358465314,
+      "grad_norm": 0.18164919316768646,
+      "learning_rate": 9.40845808688973e-06,
+      "loss": 0.457,
+      "step": 7027
+    },
+    {
+      "epoch": 1.5677002007584206,
+      "grad_norm": 0.18115800619125366,
+      "learning_rate": 9.40610879527194e-06,
+      "loss": 0.4561,
+      "step": 7028
+    },
+    {
+      "epoch": 1.5679232656703102,
+      "grad_norm": 0.17622902989387512,
+      "learning_rate": 9.403759536547577e-06,
+      "loss": 0.4436,
+      "step": 7029
+    },
+    {
+      "epoch": 1.5681463305821994,
+      "grad_norm": 0.17321671545505524,
+      "learning_rate": 9.401410310846762e-06,
+      "loss": 0.4635,
+      "step": 7030
+    },
+    {
+      "epoch": 1.5683693954940887,
+      "grad_norm": 0.16843074560165405,
+      "learning_rate": 9.39906111829961e-06,
+      "loss": 0.4266,
+      "step": 7031
+    },
+    {
+      "epoch": 1.568592460405978,
+      "grad_norm": 0.17982402443885803,
+      "learning_rate": 9.396711959036234e-06,
+      "loss": 0.4572,
+      "step": 7032
+    },
+    {
+      "epoch": 1.5688155253178675,
+      "grad_norm": 0.16617652773857117,
+      "learning_rate": 9.394362833186742e-06,
+      "loss": 0.4548,
+      "step": 7033
+    },
+    {
+      "epoch": 1.569038590229757,
+      "grad_norm": 0.18477341532707214,
+      "learning_rate": 9.39201374088125e-06,
+      "loss": 0.4708,
+      "step": 7034
+    },
+    {
+      "epoch": 1.5692616551416463,
+      "grad_norm": 0.1808331459760666,
+      "learning_rate": 9.389664682249859e-06,
+      "loss": 0.4721,
+      "step": 7035
+    },
+    {
+      "epoch": 1.5694847200535356,
+      "grad_norm": 0.16547119617462158,
+      "learning_rate": 9.387315657422677e-06,
+      "loss": 0.4477,
+      "step": 7036
+    },
+    {
+      "epoch": 1.5697077849654248,
+      "grad_norm": 0.19338512420654297,
+      "learning_rate": 9.384966666529807e-06,
+      "loss": 0.4559,
+      "step": 7037
+    },
+    {
+      "epoch": 1.5699308498773143,
+      "grad_norm": 0.1708751618862152,
+      "learning_rate": 9.382617709701355e-06,
+      "loss": 0.4593,
+      "step": 7038
+    },
+    {
+      "epoch": 1.5701539147892036,
+      "grad_norm": 0.17655450105667114,
+      "learning_rate": 9.380268787067416e-06,
+      "loss": 0.452,
+      "step": 7039
+    },
+    {
+      "epoch": 1.5703769797010931,
+      "grad_norm": 0.1719549298286438,
+      "learning_rate": 9.377919898758089e-06,
+      "loss": 0.452,
+      "step": 7040
+    },
+    {
+      "epoch": 1.5706000446129824,
+      "grad_norm": 0.17467504739761353,
+      "learning_rate": 9.375571044903469e-06,
+      "loss": 0.4699,
+      "step": 7041
+    },
+    {
+      "epoch": 1.5708231095248717,
+      "grad_norm": 0.17232942581176758,
+      "learning_rate": 9.373222225633655e-06,
+      "loss": 0.4696,
+      "step": 7042
+    },
+    {
+      "epoch": 1.571046174436761,
+      "grad_norm": 0.1764511615037918,
+      "learning_rate": 9.370873441078738e-06,
+      "loss": 0.4655,
+      "step": 7043
+    },
+    {
+      "epoch": 1.5712692393486505,
+      "grad_norm": 0.16992317140102386,
+      "learning_rate": 9.368524691368804e-06,
+      "loss": 0.4342,
+      "step": 7044
+    },
+    {
+      "epoch": 1.5714923042605398,
+      "grad_norm": 0.1786929816007614,
+      "learning_rate": 9.366175976633949e-06,
+      "loss": 0.4513,
+      "step": 7045
+    },
+    {
+      "epoch": 1.5717153691724293,
+      "grad_norm": 0.1684352606534958,
+      "learning_rate": 9.363827297004252e-06,
+      "loss": 0.4698,
+      "step": 7046
+    },
+    {
+      "epoch": 1.5719384340843185,
+      "grad_norm": 0.17656061053276062,
+      "learning_rate": 9.3614786526098e-06,
+      "loss": 0.4762,
+      "step": 7047
+    },
+    {
+      "epoch": 1.5721614989962078,
+      "grad_norm": 0.1755649596452713,
+      "learning_rate": 9.359130043580679e-06,
+      "loss": 0.471,
+      "step": 7048
+    },
+    {
+      "epoch": 1.572384563908097,
+      "grad_norm": 0.1698499619960785,
+      "learning_rate": 9.356781470046968e-06,
+      "loss": 0.4455,
+      "step": 7049
+    },
+    {
+      "epoch": 1.5726076288199866,
+      "grad_norm": 0.1733773797750473,
+      "learning_rate": 9.354432932138743e-06,
+      "loss": 0.4295,
+      "step": 7050
+    },
+    {
+      "epoch": 1.5728306937318761,
+      "grad_norm": 0.1755225956439972,
+      "learning_rate": 9.352084429986085e-06,
+      "loss": 0.4365,
+      "step": 7051
+    },
+    {
+      "epoch": 1.5730537586437654,
+      "grad_norm": 0.16997897624969482,
+      "learning_rate": 9.349735963719065e-06,
+      "loss": 0.439,
+      "step": 7052
+    },
+    {
+      "epoch": 1.5732768235556547,
+      "grad_norm": 0.17118623852729797,
+      "learning_rate": 9.34738753346776e-06,
+      "loss": 0.4616,
+      "step": 7053
+    },
+    {
+      "epoch": 1.573499888467544,
+      "grad_norm": 0.1747768670320511,
+      "learning_rate": 9.345039139362235e-06,
+      "loss": 0.4594,
+      "step": 7054
+    },
+    {
+      "epoch": 1.5737229533794335,
+      "grad_norm": 0.17374879121780396,
+      "learning_rate": 9.342690781532566e-06,
+      "loss": 0.461,
+      "step": 7055
+    },
+    {
+      "epoch": 1.5739460182913227,
+      "grad_norm": 0.18289963901042938,
+      "learning_rate": 9.340342460108813e-06,
+      "loss": 0.4668,
+      "step": 7056
+    },
+    {
+      "epoch": 1.5741690832032122,
+      "grad_norm": 0.17302124202251434,
+      "learning_rate": 9.337994175221047e-06,
+      "loss": 0.4441,
+      "step": 7057
+    },
+    {
+      "epoch": 1.5743921481151015,
+      "grad_norm": 0.17973129451274872,
+      "learning_rate": 9.335645926999324e-06,
+      "loss": 0.4488,
+      "step": 7058
+    },
+    {
+      "epoch": 1.5746152130269908,
+      "grad_norm": 0.17324143648147583,
+      "learning_rate": 9.333297715573713e-06,
+      "loss": 0.4394,
+      "step": 7059
+    },
+    {
+      "epoch": 1.57483827793888,
+      "grad_norm": 0.17785955965518951,
+      "learning_rate": 9.330949541074265e-06,
+      "loss": 0.4533,
+      "step": 7060
+    },
+    {
+      "epoch": 1.5750613428507696,
+      "grad_norm": 0.17939788103103638,
+      "learning_rate": 9.328601403631044e-06,
+      "loss": 0.4592,
+      "step": 7061
+    },
+    {
+      "epoch": 1.575284407762659,
+      "grad_norm": 0.18555541336536407,
+      "learning_rate": 9.326253303374099e-06,
+      "loss": 0.4838,
+      "step": 7062
+    },
+    {
+      "epoch": 1.5755074726745484,
+      "grad_norm": 0.17354725301265717,
+      "learning_rate": 9.323905240433481e-06,
+      "loss": 0.4543,
+      "step": 7063
+    },
+    {
+      "epoch": 1.5757305375864377,
+      "grad_norm": 0.18157054483890533,
+      "learning_rate": 9.321557214939247e-06,
+      "loss": 0.444,
+      "step": 7064
+    },
+    {
+      "epoch": 1.575953602498327,
+      "grad_norm": 0.1770789921283722,
+      "learning_rate": 9.319209227021439e-06,
+      "loss": 0.4877,
+      "step": 7065
+    },
+    {
+      "epoch": 1.5761766674102162,
+      "grad_norm": 0.18698160350322723,
+      "learning_rate": 9.316861276810105e-06,
+      "loss": 0.453,
+      "step": 7066
+    },
+    {
+      "epoch": 1.5763997323221057,
+      "grad_norm": 0.1865304410457611,
+      "learning_rate": 9.314513364435295e-06,
+      "loss": 0.4649,
+      "step": 7067
+    },
+    {
+      "epoch": 1.5766227972339952,
+      "grad_norm": 0.18025344610214233,
+      "learning_rate": 9.312165490027044e-06,
+      "loss": 0.4874,
+      "step": 7068
+    },
+    {
+      "epoch": 1.5768458621458845,
+      "grad_norm": 0.17585685849189758,
+      "learning_rate": 9.309817653715395e-06,
+      "loss": 0.4716,
+      "step": 7069
+    },
+    {
+      "epoch": 1.5770689270577738,
+      "grad_norm": 0.17630046606063843,
+      "learning_rate": 9.307469855630386e-06,
+      "loss": 0.4836,
+      "step": 7070
+    },
+    {
+      "epoch": 1.577291991969663,
+      "grad_norm": 0.17560921609401703,
+      "learning_rate": 9.305122095902055e-06,
+      "loss": 0.4392,
+      "step": 7071
+    },
+    {
+      "epoch": 1.5775150568815526,
+      "grad_norm": 0.17391790449619293,
+      "learning_rate": 9.30277437466043e-06,
+      "loss": 0.4587,
+      "step": 7072
+    },
+    {
+      "epoch": 1.5777381217934419,
+      "grad_norm": 0.17293858528137207,
+      "learning_rate": 9.30042669203555e-06,
+      "loss": 0.4473,
+      "step": 7073
+    },
+    {
+      "epoch": 1.5779611867053314,
+      "grad_norm": 0.17845462262630463,
+      "learning_rate": 9.298079048157434e-06,
+      "loss": 0.4684,
+      "step": 7074
+    },
+    {
+      "epoch": 1.5781842516172206,
+      "grad_norm": 0.18190234899520874,
+      "learning_rate": 9.295731443156121e-06,
+      "loss": 0.4561,
+      "step": 7075
+    },
+    {
+      "epoch": 1.57840731652911,
+      "grad_norm": 0.1686115711927414,
+      "learning_rate": 9.293383877161628e-06,
+      "loss": 0.4418,
+      "step": 7076
+    },
+    {
+      "epoch": 1.5786303814409992,
+      "grad_norm": 0.17196792364120483,
+      "learning_rate": 9.291036350303982e-06,
+      "loss": 0.4569,
+      "step": 7077
+    },
+    {
+      "epoch": 1.5788534463528887,
+      "grad_norm": 0.17236146330833435,
+      "learning_rate": 9.288688862713201e-06,
+      "loss": 0.4466,
+      "step": 7078
+    },
+    {
+      "epoch": 1.5790765112647782,
+      "grad_norm": 0.1745366007089615,
+      "learning_rate": 9.286341414519307e-06,
+      "loss": 0.4356,
+      "step": 7079
+    },
+    {
+      "epoch": 1.5792995761766675,
+      "grad_norm": 0.1981036514043808,
+      "learning_rate": 9.283994005852313e-06,
+      "loss": 0.4698,
+      "step": 7080
+    },
+    {
+      "epoch": 1.5795226410885568,
+      "grad_norm": 0.17840775847434998,
+      "learning_rate": 9.281646636842235e-06,
+      "loss": 0.4553,
+      "step": 7081
+    },
+    {
+      "epoch": 1.579745706000446,
+      "grad_norm": 0.17592334747314453,
+      "learning_rate": 9.279299307619085e-06,
+      "loss": 0.4468,
+      "step": 7082
+    },
+    {
+      "epoch": 1.5799687709123353,
+      "grad_norm": 0.17644402384757996,
+      "learning_rate": 9.276952018312874e-06,
+      "loss": 0.4373,
+      "step": 7083
+    },
+    {
+      "epoch": 1.5801918358242248,
+      "grad_norm": 0.18005765974521637,
+      "learning_rate": 9.274604769053605e-06,
+      "loss": 0.4825,
+      "step": 7084
+    },
+    {
+      "epoch": 1.5804149007361143,
+      "grad_norm": 0.17240746319293976,
+      "learning_rate": 9.272257559971291e-06,
+      "loss": 0.4562,
+      "step": 7085
+    },
+    {
+      "epoch": 1.5806379656480036,
+      "grad_norm": 0.171810120344162,
+      "learning_rate": 9.269910391195929e-06,
+      "loss": 0.4314,
+      "step": 7086
+    },
+    {
+      "epoch": 1.580861030559893,
+      "grad_norm": 0.33604612946510315,
+      "learning_rate": 9.26756326285752e-06,
+      "loss": 0.4709,
+      "step": 7087
+    },
+    {
+      "epoch": 1.5810840954717822,
+      "grad_norm": 0.18000468611717224,
+      "learning_rate": 9.265216175086068e-06,
+      "loss": 0.4858,
+      "step": 7088
+    },
+    {
+      "epoch": 1.5813071603836717,
+      "grad_norm": 0.17121154069900513,
+      "learning_rate": 9.262869128011565e-06,
+      "loss": 0.4492,
+      "step": 7089
+    },
+    {
+      "epoch": 1.581530225295561,
+      "grad_norm": 0.1873648315668106,
+      "learning_rate": 9.260522121764007e-06,
+      "loss": 0.4501,
+      "step": 7090
+    },
+    {
+      "epoch": 1.5817532902074505,
+      "grad_norm": 0.17154662311077118,
+      "learning_rate": 9.258175156473383e-06,
+      "loss": 0.4237,
+      "step": 7091
+    },
+    {
+      "epoch": 1.5819763551193398,
+      "grad_norm": 0.17864510416984558,
+      "learning_rate": 9.255828232269689e-06,
+      "loss": 0.4674,
+      "step": 7092
+    },
+    {
+      "epoch": 1.582199420031229,
+      "grad_norm": 0.18545754253864288,
+      "learning_rate": 9.253481349282906e-06,
+      "loss": 0.4657,
+      "step": 7093
+    },
+    {
+      "epoch": 1.5824224849431183,
+      "grad_norm": 0.18193696439266205,
+      "learning_rate": 9.251134507643022e-06,
+      "loss": 0.4538,
+      "step": 7094
+    },
+    {
+      "epoch": 1.5826455498550078,
+      "grad_norm": 0.17298151552677155,
+      "learning_rate": 9.248787707480018e-06,
+      "loss": 0.4696,
+      "step": 7095
+    },
+    {
+      "epoch": 1.5828686147668973,
+      "grad_norm": 0.17081047594547272,
+      "learning_rate": 9.24644094892388e-06,
+      "loss": 0.4593,
+      "step": 7096
+    },
+    {
+      "epoch": 1.5830916796787866,
+      "grad_norm": 0.18091976642608643,
+      "learning_rate": 9.244094232104578e-06,
+      "loss": 0.4611,
+      "step": 7097
+    },
+    {
+      "epoch": 1.583314744590676,
+      "grad_norm": 0.29800283908843994,
+      "learning_rate": 9.241747557152096e-06,
+      "loss": 0.4455,
+      "step": 7098
+    },
+    {
+      "epoch": 1.5835378095025652,
+      "grad_norm": 0.16630631685256958,
+      "learning_rate": 9.239400924196402e-06,
+      "loss": 0.4664,
+      "step": 7099
+    },
+    {
+      "epoch": 1.5837608744144545,
+      "grad_norm": 0.17668049037456512,
+      "learning_rate": 9.23705433336747e-06,
+      "loss": 0.4666,
+      "step": 7100
+    },
+    {
+      "epoch": 1.583983939326344,
+      "grad_norm": 0.17897532880306244,
+      "learning_rate": 9.234707784795266e-06,
+      "loss": 0.4586,
+      "step": 7101
+    },
+    {
+      "epoch": 1.5842070042382335,
+      "grad_norm": 0.18220888078212738,
+      "learning_rate": 9.232361278609761e-06,
+      "loss": 0.459,
+      "step": 7102
+    },
+    {
+      "epoch": 1.5844300691501227,
+      "grad_norm": 0.17005829513072968,
+      "learning_rate": 9.230014814940917e-06,
+      "loss": 0.4506,
+      "step": 7103
+    },
+    {
+      "epoch": 1.584653134062012,
+      "grad_norm": 0.17794610559940338,
+      "learning_rate": 9.227668393918695e-06,
+      "loss": 0.4894,
+      "step": 7104
+    },
+    {
+      "epoch": 1.5848761989739013,
+      "grad_norm": 0.1713520735502243,
+      "learning_rate": 9.225322015673055e-06,
+      "loss": 0.4762,
+      "step": 7105
+    },
+    {
+      "epoch": 1.5850992638857908,
+      "grad_norm": 0.17363131046295166,
+      "learning_rate": 9.222975680333956e-06,
+      "loss": 0.453,
+      "step": 7106
+    },
+    {
+      "epoch": 1.58532232879768,
+      "grad_norm": 0.1799049973487854,
+      "learning_rate": 9.22062938803135e-06,
+      "loss": 0.4692,
+      "step": 7107
+    },
+    {
+      "epoch": 1.5855453937095696,
+      "grad_norm": 0.1882503479719162,
+      "learning_rate": 9.218283138895192e-06,
+      "loss": 0.4721,
+      "step": 7108
+    },
+    {
+      "epoch": 1.5857684586214589,
+      "grad_norm": 0.19091439247131348,
+      "learning_rate": 9.215936933055431e-06,
+      "loss": 0.471,
+      "step": 7109
+    },
+    {
+      "epoch": 1.5859915235333482,
+      "grad_norm": 0.1782108098268509,
+      "learning_rate": 9.213590770642014e-06,
+      "loss": 0.4753,
+      "step": 7110
+    },
+    {
+      "epoch": 1.5862145884452374,
+      "grad_norm": 0.17361801862716675,
+      "learning_rate": 9.211244651784888e-06,
+      "loss": 0.4706,
+      "step": 7111
+    },
+    {
+      "epoch": 1.586437653357127,
+      "grad_norm": 0.18161118030548096,
+      "learning_rate": 9.208898576613993e-06,
+      "loss": 0.457,
+      "step": 7112
+    },
+    {
+      "epoch": 1.5866607182690164,
+      "grad_norm": 0.16798540949821472,
+      "learning_rate": 9.206552545259274e-06,
+      "loss": 0.4232,
+      "step": 7113
+    },
+    {
+      "epoch": 1.5868837831809057,
+      "grad_norm": 0.1804846078157425,
+      "learning_rate": 9.204206557850661e-06,
+      "loss": 0.4703,
+      "step": 7114
+    },
+    {
+      "epoch": 1.587106848092795,
+      "grad_norm": 0.17677569389343262,
+      "learning_rate": 9.201860614518098e-06,
+      "loss": 0.4451,
+      "step": 7115
+    },
+    {
+      "epoch": 1.5873299130046843,
+      "grad_norm": 0.18691909313201904,
+      "learning_rate": 9.199514715391514e-06,
+      "loss": 0.4525,
+      "step": 7116
+    },
+    {
+      "epoch": 1.5875529779165736,
+      "grad_norm": 0.1880151480436325,
+      "learning_rate": 9.197168860600843e-06,
+      "loss": 0.4887,
+      "step": 7117
+    },
+    {
+      "epoch": 1.587776042828463,
+      "grad_norm": 0.18004381656646729,
+      "learning_rate": 9.194823050276007e-06,
+      "loss": 0.4763,
+      "step": 7118
+    },
+    {
+      "epoch": 1.5879991077403526,
+      "grad_norm": 0.17934048175811768,
+      "learning_rate": 9.192477284546937e-06,
+      "loss": 0.4896,
+      "step": 7119
+    },
+    {
+      "epoch": 1.5882221726522419,
+      "grad_norm": 0.17603585124015808,
+      "learning_rate": 9.190131563543551e-06,
+      "loss": 0.4514,
+      "step": 7120
+    },
+    {
+      "epoch": 1.5884452375641311,
+      "grad_norm": 0.1763220578432083,
+      "learning_rate": 9.187785887395778e-06,
+      "loss": 0.4501,
+      "step": 7121
+    },
+    {
+      "epoch": 1.5886683024760204,
+      "grad_norm": 0.16422604024410248,
+      "learning_rate": 9.18544025623353e-06,
+      "loss": 0.4382,
+      "step": 7122
+    },
+    {
+      "epoch": 1.58889136738791,
+      "grad_norm": 0.17664875090122223,
+      "learning_rate": 9.183094670186724e-06,
+      "loss": 0.4802,
+      "step": 7123
+    },
+    {
+      "epoch": 1.5891144322997992,
+      "grad_norm": 0.1742987334728241,
+      "learning_rate": 9.180749129385273e-06,
+      "loss": 0.4307,
+      "step": 7124
+    },
+    {
+      "epoch": 1.5893374972116887,
+      "grad_norm": 0.16630713641643524,
+      "learning_rate": 9.17840363395909e-06,
+      "loss": 0.4245,
+      "step": 7125
+    },
+    {
+      "epoch": 1.589560562123578,
+      "grad_norm": 0.17244166135787964,
+      "learning_rate": 9.176058184038081e-06,
+      "loss": 0.4488,
+      "step": 7126
+    },
+    {
+      "epoch": 1.5897836270354673,
+      "grad_norm": 0.1764003038406372,
+      "learning_rate": 9.17371277975215e-06,
+      "loss": 0.4711,
+      "step": 7127
+    },
+    {
+      "epoch": 1.5900066919473566,
+      "grad_norm": 0.18569590151309967,
+      "learning_rate": 9.171367421231207e-06,
+      "loss": 0.4765,
+      "step": 7128
+    },
+    {
+      "epoch": 1.590229756859246,
+      "grad_norm": 0.19138166308403015,
+      "learning_rate": 9.169022108605147e-06,
+      "loss": 0.4742,
+      "step": 7129
+    },
+    {
+      "epoch": 1.5904528217711356,
+      "grad_norm": 0.17266172170639038,
+      "learning_rate": 9.16667684200387e-06,
+      "loss": 0.4525,
+      "step": 7130
+    },
+    {
+      "epoch": 1.5906758866830248,
+      "grad_norm": 0.17976059019565582,
+      "learning_rate": 9.16433162155727e-06,
+      "loss": 0.4669,
+      "step": 7131
+    },
+    {
+      "epoch": 1.5908989515949141,
+      "grad_norm": 0.16930918395519257,
+      "learning_rate": 9.161986447395244e-06,
+      "loss": 0.4463,
+      "step": 7132
+    },
+    {
+      "epoch": 1.5911220165068034,
+      "grad_norm": 0.185069739818573,
+      "learning_rate": 9.159641319647676e-06,
+      "loss": 0.4553,
+      "step": 7133
+    },
+    {
+      "epoch": 1.5913450814186927,
+      "grad_norm": 0.17120788991451263,
+      "learning_rate": 9.157296238444462e-06,
+      "loss": 0.4564,
+      "step": 7134
+    },
+    {
+      "epoch": 1.5915681463305822,
+      "grad_norm": 0.18028856813907623,
+      "learning_rate": 9.15495120391548e-06,
+      "loss": 0.4903,
+      "step": 7135
+    },
+    {
+      "epoch": 1.5917912112424717,
+      "grad_norm": 0.18500107526779175,
+      "learning_rate": 9.152606216190619e-06,
+      "loss": 0.4839,
+      "step": 7136
+    },
+    {
+      "epoch": 1.592014276154361,
+      "grad_norm": 0.17206281423568726,
+      "learning_rate": 9.150261275399752e-06,
+      "loss": 0.4445,
+      "step": 7137
+    },
+    {
+      "epoch": 1.5922373410662503,
+      "grad_norm": 0.19422174990177155,
+      "learning_rate": 9.147916381672763e-06,
+      "loss": 0.4412,
+      "step": 7138
+    },
+    {
+      "epoch": 1.5924604059781395,
+      "grad_norm": 0.17535541951656342,
+      "learning_rate": 9.145571535139524e-06,
+      "loss": 0.4682,
+      "step": 7139
+    },
+    {
+      "epoch": 1.592683470890029,
+      "grad_norm": 0.17184211313724518,
+      "learning_rate": 9.14322673592991e-06,
+      "loss": 0.4383,
+      "step": 7140
+    },
+    {
+      "epoch": 1.5929065358019183,
+      "grad_norm": 0.1792580634355545,
+      "learning_rate": 9.140881984173786e-06,
+      "loss": 0.4486,
+      "step": 7141
+    },
+    {
+      "epoch": 1.5931296007138078,
+      "grad_norm": 0.17893850803375244,
+      "learning_rate": 9.138537280001025e-06,
+      "loss": 0.4494,
+      "step": 7142
+    },
+    {
+      "epoch": 1.593352665625697,
+      "grad_norm": 0.17815670371055603,
+      "learning_rate": 9.136192623541487e-06,
+      "loss": 0.46,
+      "step": 7143
+    },
+    {
+      "epoch": 1.5935757305375864,
+      "grad_norm": 0.17325416207313538,
+      "learning_rate": 9.133848014925032e-06,
+      "loss": 0.4582,
+      "step": 7144
+    },
+    {
+      "epoch": 1.5937987954494757,
+      "grad_norm": 0.1770549863576889,
+      "learning_rate": 9.131503454281526e-06,
+      "loss": 0.4393,
+      "step": 7145
+    },
+    {
+      "epoch": 1.5940218603613652,
+      "grad_norm": 0.1701727956533432,
+      "learning_rate": 9.129158941740818e-06,
+      "loss": 0.4728,
+      "step": 7146
+    },
+    {
+      "epoch": 1.5942449252732547,
+      "grad_norm": 0.1813841164112091,
+      "learning_rate": 9.12681447743277e-06,
+      "loss": 0.4432,
+      "step": 7147
+    },
+    {
+      "epoch": 1.594467990185144,
+      "grad_norm": 0.17347143590450287,
+      "learning_rate": 9.124470061487225e-06,
+      "loss": 0.4405,
+      "step": 7148
+    },
+    {
+      "epoch": 1.5946910550970332,
+      "grad_norm": 0.18103331327438354,
+      "learning_rate": 9.122125694034039e-06,
+      "loss": 0.4811,
+      "step": 7149
+    },
+    {
+      "epoch": 1.5949141200089225,
+      "grad_norm": 0.17634287476539612,
+      "learning_rate": 9.11978137520305e-06,
+      "loss": 0.4625,
+      "step": 7150
+    },
+    {
+      "epoch": 1.5951371849208118,
+      "grad_norm": 0.16531561315059662,
+      "learning_rate": 9.117437105124107e-06,
+      "loss": 0.4206,
+      "step": 7151
+    },
+    {
+      "epoch": 1.5953602498327013,
+      "grad_norm": 0.1829378753900528,
+      "learning_rate": 9.11509288392705e-06,
+      "loss": 0.4665,
+      "step": 7152
+    },
+    {
+      "epoch": 1.5955833147445908,
+      "grad_norm": 0.17331163585186005,
+      "learning_rate": 9.112748711741712e-06,
+      "loss": 0.4393,
+      "step": 7153
+    },
+    {
+      "epoch": 1.59580637965648,
+      "grad_norm": 0.1706630289554596,
+      "learning_rate": 9.110404588697935e-06,
+      "loss": 0.46,
+      "step": 7154
+    },
+    {
+      "epoch": 1.5960294445683694,
+      "grad_norm": 0.16561025381088257,
+      "learning_rate": 9.108060514925546e-06,
+      "loss": 0.4287,
+      "step": 7155
+    },
+    {
+      "epoch": 1.5962525094802587,
+      "grad_norm": 0.17239882051944733,
+      "learning_rate": 9.105716490554377e-06,
+      "loss": 0.466,
+      "step": 7156
+    },
+    {
+      "epoch": 1.5964755743921482,
+      "grad_norm": 0.16943053901195526,
+      "learning_rate": 9.103372515714252e-06,
+      "loss": 0.4275,
+      "step": 7157
+    },
+    {
+      "epoch": 1.5966986393040374,
+      "grad_norm": 0.1701379269361496,
+      "learning_rate": 9.101028590534999e-06,
+      "loss": 0.4459,
+      "step": 7158
+    },
+    {
+      "epoch": 1.596921704215927,
+      "grad_norm": 0.18113826215267181,
+      "learning_rate": 9.098684715146435e-06,
+      "loss": 0.454,
+      "step": 7159
+    },
+    {
+      "epoch": 1.5971447691278162,
+      "grad_norm": 0.19422270357608795,
+      "learning_rate": 9.096340889678384e-06,
+      "loss": 0.4594,
+      "step": 7160
+    },
+    {
+      "epoch": 1.5973678340397055,
+      "grad_norm": 0.18085762858390808,
+      "learning_rate": 9.093997114260658e-06,
+      "loss": 0.4853,
+      "step": 7161
+    },
+    {
+      "epoch": 1.5975908989515948,
+      "grad_norm": 0.18114502727985382,
+      "learning_rate": 9.09165338902307e-06,
+      "loss": 0.4688,
+      "step": 7162
+    },
+    {
+      "epoch": 1.5978139638634843,
+      "grad_norm": 0.18122133612632751,
+      "learning_rate": 9.089309714095432e-06,
+      "loss": 0.4679,
+      "step": 7163
+    },
+    {
+      "epoch": 1.5980370287753738,
+      "grad_norm": 0.18571965396404266,
+      "learning_rate": 9.086966089607551e-06,
+      "loss": 0.48,
+      "step": 7164
+    },
+    {
+      "epoch": 1.598260093687263,
+      "grad_norm": 0.1749122440814972,
+      "learning_rate": 9.084622515689228e-06,
+      "loss": 0.4706,
+      "step": 7165
+    },
+    {
+      "epoch": 1.5984831585991524,
+      "grad_norm": 0.17465496063232422,
+      "learning_rate": 9.082278992470272e-06,
+      "loss": 0.4586,
+      "step": 7166
+    },
+    {
+      "epoch": 1.5987062235110416,
+      "grad_norm": 0.1802026480436325,
+      "learning_rate": 9.079935520080474e-06,
+      "loss": 0.4348,
+      "step": 7167
+    },
+    {
+      "epoch": 1.598929288422931,
+      "grad_norm": 0.17339572310447693,
+      "learning_rate": 9.077592098649639e-06,
+      "loss": 0.4688,
+      "step": 7168
+    },
+    {
+      "epoch": 1.5991523533348204,
+      "grad_norm": 0.17421960830688477,
+      "learning_rate": 9.075248728307551e-06,
+      "loss": 0.4779,
+      "step": 7169
+    },
+    {
+      "epoch": 1.59937541824671,
+      "grad_norm": 0.17133331298828125,
+      "learning_rate": 9.07290540918401e-06,
+      "loss": 0.451,
+      "step": 7170
+    },
+    {
+      "epoch": 1.5995984831585992,
+      "grad_norm": 0.1657906174659729,
+      "learning_rate": 9.070562141408795e-06,
+      "loss": 0.4296,
+      "step": 7171
+    },
+    {
+      "epoch": 1.5998215480704885,
+      "grad_norm": 0.17014503479003906,
+      "learning_rate": 9.068218925111695e-06,
+      "loss": 0.4302,
+      "step": 7172
+    },
+    {
+      "epoch": 1.6000446129823778,
+      "grad_norm": 0.1789960116147995,
+      "learning_rate": 9.065875760422496e-06,
+      "loss": 0.4808,
+      "step": 7173
+    },
+    {
+      "epoch": 1.6002676778942673,
+      "grad_norm": 0.16898800432682037,
+      "learning_rate": 9.06353264747097e-06,
+      "loss": 0.4614,
+      "step": 7174
+    },
+    {
+      "epoch": 1.6004907428061566,
+      "grad_norm": 0.1819867193698883,
+      "learning_rate": 9.0611895863869e-06,
+      "loss": 0.4687,
+      "step": 7175
+    },
+    {
+      "epoch": 1.600713807718046,
+      "grad_norm": 0.17777277529239655,
+      "learning_rate": 9.058846577300052e-06,
+      "loss": 0.4452,
+      "step": 7176
+    },
+    {
+      "epoch": 1.6009368726299353,
+      "grad_norm": 0.1771070510149002,
+      "learning_rate": 9.056503620340204e-06,
+      "loss": 0.4647,
+      "step": 7177
+    },
+    {
+      "epoch": 1.6011599375418246,
+      "grad_norm": 0.17987768352031708,
+      "learning_rate": 9.054160715637117e-06,
+      "loss": 0.465,
+      "step": 7178
+    },
+    {
+      "epoch": 1.601383002453714,
+      "grad_norm": 0.17189104855060577,
+      "learning_rate": 9.051817863320563e-06,
+      "loss": 0.46,
+      "step": 7179
+    },
+    {
+      "epoch": 1.6016060673656034,
+      "grad_norm": 0.19388031959533691,
+      "learning_rate": 9.049475063520295e-06,
+      "loss": 0.4347,
+      "step": 7180
+    },
+    {
+      "epoch": 1.601829132277493,
+      "grad_norm": 0.1758044809103012,
+      "learning_rate": 9.047132316366082e-06,
+      "loss": 0.4477,
+      "step": 7181
+    },
+    {
+      "epoch": 1.6020521971893822,
+      "grad_norm": 0.19193986058235168,
+      "learning_rate": 9.044789621987672e-06,
+      "loss": 0.4698,
+      "step": 7182
+    },
+    {
+      "epoch": 1.6022752621012715,
+      "grad_norm": 0.1996065378189087,
+      "learning_rate": 9.042446980514822e-06,
+      "loss": 0.4733,
+      "step": 7183
+    },
+    {
+      "epoch": 1.6024983270131608,
+      "grad_norm": 0.1759694218635559,
+      "learning_rate": 9.040104392077283e-06,
+      "loss": 0.4534,
+      "step": 7184
+    },
+    {
+      "epoch": 1.60272139192505,
+      "grad_norm": 0.1798923909664154,
+      "learning_rate": 9.0377618568048e-06,
+      "loss": 0.4792,
+      "step": 7185
+    },
+    {
+      "epoch": 1.6029444568369395,
+      "grad_norm": 0.17134787142276764,
+      "learning_rate": 9.035419374827116e-06,
+      "loss": 0.4889,
+      "step": 7186
+    },
+    {
+      "epoch": 1.603167521748829,
+      "grad_norm": 0.2020411342382431,
+      "learning_rate": 9.033076946273978e-06,
+      "loss": 0.4891,
+      "step": 7187
+    },
+    {
+      "epoch": 1.6033905866607183,
+      "grad_norm": 0.1749039590358734,
+      "learning_rate": 9.030734571275118e-06,
+      "loss": 0.4535,
+      "step": 7188
+    },
+    {
+      "epoch": 1.6036136515726076,
+      "grad_norm": 0.1836901307106018,
+      "learning_rate": 9.028392249960276e-06,
+      "loss": 0.4793,
+      "step": 7189
+    },
+    {
+      "epoch": 1.6038367164844969,
+      "grad_norm": 0.19300808012485504,
+      "learning_rate": 9.02604998245918e-06,
+      "loss": 0.4434,
+      "step": 7190
+    },
+    {
+      "epoch": 1.6040597813963864,
+      "grad_norm": 0.18210625648498535,
+      "learning_rate": 9.023707768901567e-06,
+      "loss": 0.4591,
+      "step": 7191
+    },
+    {
+      "epoch": 1.6042828463082757,
+      "grad_norm": 0.17498520016670227,
+      "learning_rate": 9.021365609417155e-06,
+      "loss": 0.4658,
+      "step": 7192
+    },
+    {
+      "epoch": 1.6045059112201652,
+      "grad_norm": 0.17925825715065002,
+      "learning_rate": 9.019023504135671e-06,
+      "loss": 0.4617,
+      "step": 7193
+    },
+    {
+      "epoch": 1.6047289761320545,
+      "grad_norm": 0.18746887147426605,
+      "learning_rate": 9.01668145318684e-06,
+      "loss": 0.4819,
+      "step": 7194
+    },
+    {
+      "epoch": 1.6049520410439437,
+      "grad_norm": 0.16884252429008484,
+      "learning_rate": 9.014339456700371e-06,
+      "loss": 0.4457,
+      "step": 7195
+    },
+    {
+      "epoch": 1.605175105955833,
+      "grad_norm": 0.1717928946018219,
+      "learning_rate": 9.011997514805986e-06,
+      "loss": 0.4507,
+      "step": 7196
+    },
+    {
+      "epoch": 1.6053981708677225,
+      "grad_norm": 0.18760330975055695,
+      "learning_rate": 9.009655627633392e-06,
+      "loss": 0.4447,
+      "step": 7197
+    },
+    {
+      "epoch": 1.605621235779612,
+      "grad_norm": 0.16959816217422485,
+      "learning_rate": 9.0073137953123e-06,
+      "loss": 0.4185,
+      "step": 7198
+    },
+    {
+      "epoch": 1.6058443006915013,
+      "grad_norm": 0.17206130921840668,
+      "learning_rate": 9.004972017972414e-06,
+      "loss": 0.4572,
+      "step": 7199
+    },
+    {
+      "epoch": 1.6060673656033906,
+      "grad_norm": 0.1713947355747223,
+      "learning_rate": 9.002630295743437e-06,
+      "loss": 0.4656,
+      "step": 7200
+    },
+    {
+      "epoch": 1.6062904305152799,
+      "grad_norm": 0.1734878569841385,
+      "learning_rate": 9.000288628755065e-06,
+      "loss": 0.4543,
+      "step": 7201
+    },
+    {
+      "epoch": 1.6065134954271691,
+      "grad_norm": 0.17771008610725403,
+      "learning_rate": 8.997947017137e-06,
+      "loss": 0.4749,
+      "step": 7202
+    },
+    {
+      "epoch": 1.6067365603390587,
+      "grad_norm": 0.17664694786071777,
+      "learning_rate": 8.995605461018933e-06,
+      "loss": 0.4472,
+      "step": 7203
+    },
+    {
+      "epoch": 1.6069596252509482,
+      "grad_norm": 0.18092039227485657,
+      "learning_rate": 8.993263960530552e-06,
+      "loss": 0.4716,
+      "step": 7204
+    },
+    {
+      "epoch": 1.6071826901628374,
+      "grad_norm": 0.17921680212020874,
+      "learning_rate": 8.990922515801546e-06,
+      "loss": 0.4736,
+      "step": 7205
+    },
+    {
+      "epoch": 1.6074057550747267,
+      "grad_norm": 0.17247629165649414,
+      "learning_rate": 8.9885811269616e-06,
+      "loss": 0.4575,
+      "step": 7206
+    },
+    {
+      "epoch": 1.607628819986616,
+      "grad_norm": 0.1867935061454773,
+      "learning_rate": 8.986239794140389e-06,
+      "loss": 0.4509,
+      "step": 7207
+    },
+    {
+      "epoch": 1.6078518848985055,
+      "grad_norm": 0.18020030856132507,
+      "learning_rate": 8.983898517467598e-06,
+      "loss": 0.4526,
+      "step": 7208
+    },
+    {
+      "epoch": 1.6080749498103948,
+      "grad_norm": 0.17290444672107697,
+      "learning_rate": 8.981557297072897e-06,
+      "loss": 0.4358,
+      "step": 7209
+    },
+    {
+      "epoch": 1.6082980147222843,
+      "grad_norm": 0.17369163036346436,
+      "learning_rate": 8.979216133085961e-06,
+      "loss": 0.4485,
+      "step": 7210
+    },
+    {
+      "epoch": 1.6085210796341736,
+      "grad_norm": 0.17620067298412323,
+      "learning_rate": 8.976875025636455e-06,
+      "loss": 0.4515,
+      "step": 7211
+    },
+    {
+      "epoch": 1.6087441445460628,
+      "grad_norm": 0.17683733999729156,
+      "learning_rate": 8.974533974854046e-06,
+      "loss": 0.4434,
+      "step": 7212
+    },
+    {
+      "epoch": 1.6089672094579521,
+      "grad_norm": 0.16844935715198517,
+      "learning_rate": 8.972192980868397e-06,
+      "loss": 0.4471,
+      "step": 7213
+    },
+    {
+      "epoch": 1.6091902743698416,
+      "grad_norm": 0.17559292912483215,
+      "learning_rate": 8.969852043809165e-06,
+      "loss": 0.4755,
+      "step": 7214
+    },
+    {
+      "epoch": 1.6094133392817311,
+      "grad_norm": 0.17093756794929504,
+      "learning_rate": 8.967511163806008e-06,
+      "loss": 0.4571,
+      "step": 7215
+    },
+    {
+      "epoch": 1.6096364041936204,
+      "grad_norm": 0.17171628773212433,
+      "learning_rate": 8.965170340988574e-06,
+      "loss": 0.4461,
+      "step": 7216
+    },
+    {
+      "epoch": 1.6098594691055097,
+      "grad_norm": 0.17218218743801117,
+      "learning_rate": 8.96282957548652e-06,
+      "loss": 0.4595,
+      "step": 7217
+    },
+    {
+      "epoch": 1.610082534017399,
+      "grad_norm": 0.17221598327159882,
+      "learning_rate": 8.960488867429486e-06,
+      "loss": 0.4469,
+      "step": 7218
+    },
+    {
+      "epoch": 1.6103055989292883,
+      "grad_norm": 0.19169965386390686,
+      "learning_rate": 8.958148216947118e-06,
+      "loss": 0.4863,
+      "step": 7219
+    },
+    {
+      "epoch": 1.6105286638411778,
+      "grad_norm": 0.19341188669204712,
+      "learning_rate": 8.955807624169054e-06,
+      "loss": 0.482,
+      "step": 7220
+    },
+    {
+      "epoch": 1.6107517287530673,
+      "grad_norm": 0.17554320394992828,
+      "learning_rate": 8.953467089224934e-06,
+      "loss": 0.4396,
+      "step": 7221
+    },
+    {
+      "epoch": 1.6109747936649566,
+      "grad_norm": 0.17117850482463837,
+      "learning_rate": 8.951126612244387e-06,
+      "loss": 0.4319,
+      "step": 7222
+    },
+    {
+      "epoch": 1.6111978585768458,
+      "grad_norm": 0.17589138448238373,
+      "learning_rate": 8.948786193357049e-06,
+      "loss": 0.4553,
+      "step": 7223
+    },
+    {
+      "epoch": 1.6114209234887351,
+      "grad_norm": 0.265455961227417,
+      "learning_rate": 8.946445832692545e-06,
+      "loss": 0.481,
+      "step": 7224
+    },
+    {
+      "epoch": 1.6116439884006246,
+      "grad_norm": 0.17754699289798737,
+      "learning_rate": 8.944105530380495e-06,
+      "loss": 0.4489,
+      "step": 7225
+    },
+    {
+      "epoch": 1.611867053312514,
+      "grad_norm": 0.19480863213539124,
+      "learning_rate": 8.941765286550523e-06,
+      "loss": 0.4515,
+      "step": 7226
+    },
+    {
+      "epoch": 1.6120901182244034,
+      "grad_norm": 0.16930724680423737,
+      "learning_rate": 8.939425101332245e-06,
+      "loss": 0.4376,
+      "step": 7227
+    },
+    {
+      "epoch": 1.6123131831362927,
+      "grad_norm": 0.1783130019903183,
+      "learning_rate": 8.937084974855278e-06,
+      "loss": 0.4679,
+      "step": 7228
+    },
+    {
+      "epoch": 1.612536248048182,
+      "grad_norm": 0.19291211664676666,
+      "learning_rate": 8.934744907249229e-06,
+      "loss": 0.481,
+      "step": 7229
+    },
+    {
+      "epoch": 1.6127593129600712,
+      "grad_norm": 0.1795196533203125,
+      "learning_rate": 8.93240489864371e-06,
+      "loss": 0.4683,
+      "step": 7230
+    },
+    {
+      "epoch": 1.6129823778719607,
+      "grad_norm": 0.17331638932228088,
+      "learning_rate": 8.930064949168322e-06,
+      "loss": 0.4553,
+      "step": 7231
+    },
+    {
+      "epoch": 1.6132054427838503,
+      "grad_norm": 0.15906356275081635,
+      "learning_rate": 8.927725058952669e-06,
+      "loss": 0.4416,
+      "step": 7232
+    },
+    {
+      "epoch": 1.6134285076957395,
+      "grad_norm": 0.17841282486915588,
+      "learning_rate": 8.925385228126344e-06,
+      "loss": 0.4671,
+      "step": 7233
+    },
+    {
+      "epoch": 1.6136515726076288,
+      "grad_norm": 0.17484787106513977,
+      "learning_rate": 8.923045456818947e-06,
+      "loss": 0.4748,
+      "step": 7234
+    },
+    {
+      "epoch": 1.613874637519518,
+      "grad_norm": 0.17049816250801086,
+      "learning_rate": 8.920705745160064e-06,
+      "loss": 0.4464,
+      "step": 7235
+    },
+    {
+      "epoch": 1.6140977024314074,
+      "grad_norm": 0.17688851058483124,
+      "learning_rate": 8.918366093279288e-06,
+      "loss": 0.4538,
+      "step": 7236
+    },
+    {
+      "epoch": 1.6143207673432969,
+      "grad_norm": 0.19764314591884613,
+      "learning_rate": 8.916026501306203e-06,
+      "loss": 0.464,
+      "step": 7237
+    },
+    {
+      "epoch": 1.6145438322551864,
+      "grad_norm": 0.16911235451698303,
+      "learning_rate": 8.913686969370386e-06,
+      "loss": 0.4505,
+      "step": 7238
+    },
+    {
+      "epoch": 1.6147668971670757,
+      "grad_norm": 0.1755746304988861,
+      "learning_rate": 8.911347497601418e-06,
+      "loss": 0.4726,
+      "step": 7239
+    },
+    {
+      "epoch": 1.614989962078965,
+      "grad_norm": 0.17547662556171417,
+      "learning_rate": 8.909008086128874e-06,
+      "loss": 0.4596,
+      "step": 7240
+    },
+    {
+      "epoch": 1.6152130269908542,
+      "grad_norm": 0.18209370970726013,
+      "learning_rate": 8.906668735082327e-06,
+      "loss": 0.4692,
+      "step": 7241
+    },
+    {
+      "epoch": 1.6154360919027437,
+      "grad_norm": 0.17402005195617676,
+      "learning_rate": 8.90432944459134e-06,
+      "loss": 0.4477,
+      "step": 7242
+    },
+    {
+      "epoch": 1.615659156814633,
+      "grad_norm": 0.1721334308385849,
+      "learning_rate": 8.901990214785483e-06,
+      "loss": 0.4463,
+      "step": 7243
+    },
+    {
+      "epoch": 1.6158822217265225,
+      "grad_norm": 0.17670831084251404,
+      "learning_rate": 8.899651045794313e-06,
+      "loss": 0.4583,
+      "step": 7244
+    },
+    {
+      "epoch": 1.6161052866384118,
+      "grad_norm": 0.1660810261964798,
+      "learning_rate": 8.89731193774739e-06,
+      "loss": 0.4553,
+      "step": 7245
+    },
+    {
+      "epoch": 1.616328351550301,
+      "grad_norm": 0.16802763938903809,
+      "learning_rate": 8.894972890774266e-06,
+      "loss": 0.4426,
+      "step": 7246
+    },
+    {
+      "epoch": 1.6165514164621904,
+      "grad_norm": 0.18203771114349365,
+      "learning_rate": 8.892633905004498e-06,
+      "loss": 0.4281,
+      "step": 7247
+    },
+    {
+      "epoch": 1.6167744813740799,
+      "grad_norm": 0.17793646454811096,
+      "learning_rate": 8.890294980567626e-06,
+      "loss": 0.4463,
+      "step": 7248
+    },
+    {
+      "epoch": 1.6169975462859694,
+      "grad_norm": 0.1768350899219513,
+      "learning_rate": 8.887956117593201e-06,
+      "loss": 0.4461,
+      "step": 7249
+    },
+    {
+      "epoch": 1.6172206111978586,
+      "grad_norm": 0.16968591511249542,
+      "learning_rate": 8.88561731621076e-06,
+      "loss": 0.4256,
+      "step": 7250
+    },
+    {
+      "epoch": 1.617443676109748,
+      "grad_norm": 0.17799051105976105,
+      "learning_rate": 8.883278576549843e-06,
+      "loss": 0.4425,
+      "step": 7251
+    },
+    {
+      "epoch": 1.6176667410216372,
+      "grad_norm": 0.1699870377779007,
+      "learning_rate": 8.88093989873998e-06,
+      "loss": 0.4579,
+      "step": 7252
+    },
+    {
+      "epoch": 1.6178898059335265,
+      "grad_norm": 0.17788825929164886,
+      "learning_rate": 8.87860128291071e-06,
+      "loss": 0.4816,
+      "step": 7253
+    },
+    {
+      "epoch": 1.618112870845416,
+      "grad_norm": 0.1790701150894165,
+      "learning_rate": 8.87626272919155e-06,
+      "loss": 0.4347,
+      "step": 7254
+    },
+    {
+      "epoch": 1.6183359357573055,
+      "grad_norm": 0.1901506632566452,
+      "learning_rate": 8.87392423771203e-06,
+      "loss": 0.4522,
+      "step": 7255
+    },
+    {
+      "epoch": 1.6185590006691948,
+      "grad_norm": 0.19295074045658112,
+      "learning_rate": 8.871585808601669e-06,
+      "loss": 0.4651,
+      "step": 7256
+    },
+    {
+      "epoch": 1.618782065581084,
+      "grad_norm": 0.1766098290681839,
+      "learning_rate": 8.869247441989983e-06,
+      "loss": 0.4137,
+      "step": 7257
+    },
+    {
+      "epoch": 1.6190051304929733,
+      "grad_norm": 0.17081737518310547,
+      "learning_rate": 8.866909138006488e-06,
+      "loss": 0.4277,
+      "step": 7258
+    },
+    {
+      "epoch": 1.6192281954048628,
+      "grad_norm": 0.17560650408267975,
+      "learning_rate": 8.864570896780691e-06,
+      "loss": 0.4675,
+      "step": 7259
+    },
+    {
+      "epoch": 1.6194512603167521,
+      "grad_norm": 0.1765887439250946,
+      "learning_rate": 8.862232718442101e-06,
+      "loss": 0.4536,
+      "step": 7260
+    },
+    {
+      "epoch": 1.6196743252286416,
+      "grad_norm": 0.18248333036899567,
+      "learning_rate": 8.85989460312022e-06,
+      "loss": 0.4778,
+      "step": 7261
+    },
+    {
+      "epoch": 1.619897390140531,
+      "grad_norm": 0.17278814315795898,
+      "learning_rate": 8.857556550944548e-06,
+      "loss": 0.4378,
+      "step": 7262
+    },
+    {
+      "epoch": 1.6201204550524202,
+      "grad_norm": 0.18780216574668884,
+      "learning_rate": 8.855218562044579e-06,
+      "loss": 0.4768,
+      "step": 7263
+    },
+    {
+      "epoch": 1.6203435199643095,
+      "grad_norm": 0.16610752046108246,
+      "learning_rate": 8.852880636549808e-06,
+      "loss": 0.4362,
+      "step": 7264
+    },
+    {
+      "epoch": 1.620566584876199,
+      "grad_norm": 0.16919295489788055,
+      "learning_rate": 8.85054277458972e-06,
+      "loss": 0.4394,
+      "step": 7265
+    },
+    {
+      "epoch": 1.6207896497880885,
+      "grad_norm": 0.16546227037906647,
+      "learning_rate": 8.848204976293807e-06,
+      "loss": 0.4484,
+      "step": 7266
+    },
+    {
+      "epoch": 1.6210127146999778,
+      "grad_norm": 0.1768079698085785,
+      "learning_rate": 8.845867241791548e-06,
+      "loss": 0.445,
+      "step": 7267
+    },
+    {
+      "epoch": 1.621235779611867,
+      "grad_norm": 0.17911723256111145,
+      "learning_rate": 8.84352957121242e-06,
+      "loss": 0.4643,
+      "step": 7268
+    },
+    {
+      "epoch": 1.6214588445237563,
+      "grad_norm": 0.17884330451488495,
+      "learning_rate": 8.841191964685896e-06,
+      "loss": 0.4687,
+      "step": 7269
+    },
+    {
+      "epoch": 1.6216819094356456,
+      "grad_norm": 0.17899781465530396,
+      "learning_rate": 8.838854422341454e-06,
+      "loss": 0.4641,
+      "step": 7270
+    },
+    {
+      "epoch": 1.6219049743475351,
+      "grad_norm": 0.17130422592163086,
+      "learning_rate": 8.836516944308555e-06,
+      "loss": 0.4413,
+      "step": 7271
+    },
+    {
+      "epoch": 1.6221280392594246,
+      "grad_norm": 0.17263583838939667,
+      "learning_rate": 8.834179530716669e-06,
+      "loss": 0.4521,
+      "step": 7272
+    },
+    {
+      "epoch": 1.622351104171314,
+      "grad_norm": 0.17574861645698547,
+      "learning_rate": 8.831842181695251e-06,
+      "loss": 0.4695,
+      "step": 7273
+    },
+    {
+      "epoch": 1.6225741690832032,
+      "grad_norm": 0.1733783781528473,
+      "learning_rate": 8.829504897373764e-06,
+      "loss": 0.4458,
+      "step": 7274
+    },
+    {
+      "epoch": 1.6227972339950925,
+      "grad_norm": 0.18253105878829956,
+      "learning_rate": 8.827167677881656e-06,
+      "loss": 0.4412,
+      "step": 7275
+    },
+    {
+      "epoch": 1.623020298906982,
+      "grad_norm": 0.1744835376739502,
+      "learning_rate": 8.824830523348383e-06,
+      "loss": 0.441,
+      "step": 7276
+    },
+    {
+      "epoch": 1.6232433638188712,
+      "grad_norm": 0.17317108809947968,
+      "learning_rate": 8.822493433903383e-06,
+      "loss": 0.4819,
+      "step": 7277
+    },
+    {
+      "epoch": 1.6234664287307607,
+      "grad_norm": 0.17538447678089142,
+      "learning_rate": 8.820156409676105e-06,
+      "loss": 0.4303,
+      "step": 7278
+    },
+    {
+      "epoch": 1.62368949364265,
+      "grad_norm": 0.18321259319782257,
+      "learning_rate": 8.817819450795989e-06,
+      "loss": 0.4616,
+      "step": 7279
+    },
+    {
+      "epoch": 1.6239125585545393,
+      "grad_norm": 0.17948691546916962,
+      "learning_rate": 8.815482557392463e-06,
+      "loss": 0.4734,
+      "step": 7280
+    },
+    {
+      "epoch": 1.6241356234664286,
+      "grad_norm": 0.17915484309196472,
+      "learning_rate": 8.81314572959497e-06,
+      "loss": 0.4497,
+      "step": 7281
+    },
+    {
+      "epoch": 1.624358688378318,
+      "grad_norm": 0.18481531739234924,
+      "learning_rate": 8.810808967532927e-06,
+      "loss": 0.4729,
+      "step": 7282
+    },
+    {
+      "epoch": 1.6245817532902076,
+      "grad_norm": 0.18924424052238464,
+      "learning_rate": 8.808472271335767e-06,
+      "loss": 0.4762,
+      "step": 7283
+    },
+    {
+      "epoch": 1.6248048182020969,
+      "grad_norm": 0.1721714437007904,
+      "learning_rate": 8.806135641132907e-06,
+      "loss": 0.4682,
+      "step": 7284
+    },
+    {
+      "epoch": 1.6250278831139862,
+      "grad_norm": 0.1784234195947647,
+      "learning_rate": 8.803799077053765e-06,
+      "loss": 0.465,
+      "step": 7285
+    },
+    {
+      "epoch": 1.6252509480258754,
+      "grad_norm": 0.18713349103927612,
+      "learning_rate": 8.801462579227751e-06,
+      "loss": 0.4549,
+      "step": 7286
+    },
+    {
+      "epoch": 1.625474012937765,
+      "grad_norm": 0.1756950169801712,
+      "learning_rate": 8.799126147784284e-06,
+      "loss": 0.4389,
+      "step": 7287
+    },
+    {
+      "epoch": 1.6256970778496542,
+      "grad_norm": 0.16715659201145172,
+      "learning_rate": 8.796789782852761e-06,
+      "loss": 0.4456,
+      "step": 7288
+    },
+    {
+      "epoch": 1.6259201427615437,
+      "grad_norm": 0.1801193356513977,
+      "learning_rate": 8.79445348456259e-06,
+      "loss": 0.4601,
+      "step": 7289
+    },
+    {
+      "epoch": 1.626143207673433,
+      "grad_norm": 0.18265050649642944,
+      "learning_rate": 8.792117253043166e-06,
+      "loss": 0.4617,
+      "step": 7290
+    },
+    {
+      "epoch": 1.6263662725853223,
+      "grad_norm": 0.17601865530014038,
+      "learning_rate": 8.78978108842389e-06,
+      "loss": 0.4572,
+      "step": 7291
+    },
+    {
+      "epoch": 1.6265893374972116,
+      "grad_norm": 0.1800207793712616,
+      "learning_rate": 8.787444990834146e-06,
+      "loss": 0.4626,
+      "step": 7292
+    },
+    {
+      "epoch": 1.626812402409101,
+      "grad_norm": 0.18082986772060394,
+      "learning_rate": 8.78510896040333e-06,
+      "loss": 0.4583,
+      "step": 7293
+    },
+    {
+      "epoch": 1.6270354673209904,
+      "grad_norm": 0.1825123131275177,
+      "learning_rate": 8.782772997260819e-06,
+      "loss": 0.4604,
+      "step": 7294
+    },
+    {
+      "epoch": 1.6272585322328799,
+      "grad_norm": 0.17616558074951172,
+      "learning_rate": 8.780437101535997e-06,
+      "loss": 0.442,
+      "step": 7295
+    },
+    {
+      "epoch": 1.6274815971447691,
+      "grad_norm": 0.17222437262535095,
+      "learning_rate": 8.778101273358238e-06,
+      "loss": 0.4374,
+      "step": 7296
+    },
+    {
+      "epoch": 1.6277046620566584,
+      "grad_norm": 0.1752903312444687,
+      "learning_rate": 8.775765512856919e-06,
+      "loss": 0.4703,
+      "step": 7297
+    },
+    {
+      "epoch": 1.6279277269685477,
+      "grad_norm": 0.17624272406101227,
+      "learning_rate": 8.773429820161404e-06,
+      "loss": 0.4692,
+      "step": 7298
+    },
+    {
+      "epoch": 1.6281507918804372,
+      "grad_norm": 0.17555177211761475,
+      "learning_rate": 8.771094195401062e-06,
+      "loss": 0.4517,
+      "step": 7299
+    },
+    {
+      "epoch": 1.6283738567923267,
+      "grad_norm": 0.1727064996957779,
+      "learning_rate": 8.768758638705253e-06,
+      "loss": 0.4617,
+      "step": 7300
+    },
+    {
+      "epoch": 1.628596921704216,
+      "grad_norm": 0.17664147913455963,
+      "learning_rate": 8.766423150203335e-06,
+      "loss": 0.4565,
+      "step": 7301
+    },
+    {
+      "epoch": 1.6288199866161053,
+      "grad_norm": 0.17871947586536407,
+      "learning_rate": 8.764087730024667e-06,
+      "loss": 0.464,
+      "step": 7302
+    },
+    {
+      "epoch": 1.6290430515279946,
+      "grad_norm": 0.18316639959812164,
+      "learning_rate": 8.761752378298589e-06,
+      "loss": 0.4474,
+      "step": 7303
+    },
+    {
+      "epoch": 1.629266116439884,
+      "grad_norm": 0.19442254304885864,
+      "learning_rate": 8.759417095154456e-06,
+      "loss": 0.4353,
+      "step": 7304
+    },
+    {
+      "epoch": 1.6294891813517733,
+      "grad_norm": 0.17314781248569489,
+      "learning_rate": 8.757081880721612e-06,
+      "loss": 0.4544,
+      "step": 7305
+    },
+    {
+      "epoch": 1.6297122462636628,
+      "grad_norm": 0.1879189908504486,
+      "learning_rate": 8.754746735129385e-06,
+      "loss": 0.4588,
+      "step": 7306
+    },
+    {
+      "epoch": 1.6299353111755521,
+      "grad_norm": 0.17803287506103516,
+      "learning_rate": 8.752411658507121e-06,
+      "loss": 0.4526,
+      "step": 7307
+    },
+    {
+      "epoch": 1.6301583760874414,
+      "grad_norm": 0.18903419375419617,
+      "learning_rate": 8.750076650984143e-06,
+      "loss": 0.4412,
+      "step": 7308
+    },
+    {
+      "epoch": 1.6303814409993307,
+      "grad_norm": 0.17319492995738983,
+      "learning_rate": 8.747741712689786e-06,
+      "loss": 0.4506,
+      "step": 7309
+    },
+    {
+      "epoch": 1.6306045059112202,
+      "grad_norm": 0.1812223494052887,
+      "learning_rate": 8.745406843753369e-06,
+      "loss": 0.4256,
+      "step": 7310
+    },
+    {
+      "epoch": 1.6308275708231095,
+      "grad_norm": 0.1832951158285141,
+      "learning_rate": 8.743072044304212e-06,
+      "loss": 0.4721,
+      "step": 7311
+    },
+    {
+      "epoch": 1.631050635734999,
+      "grad_norm": 0.183165043592453,
+      "learning_rate": 8.74073731447163e-06,
+      "loss": 0.4675,
+      "step": 7312
+    },
+    {
+      "epoch": 1.6312737006468883,
+      "grad_norm": 0.18091800808906555,
+      "learning_rate": 8.738402654384939e-06,
+      "loss": 0.4438,
+      "step": 7313
+    },
+    {
+      "epoch": 1.6314967655587775,
+      "grad_norm": 0.1746785044670105,
+      "learning_rate": 8.736068064173444e-06,
+      "loss": 0.4539,
+      "step": 7314
+    },
+    {
+      "epoch": 1.6317198304706668,
+      "grad_norm": 0.18262261152267456,
+      "learning_rate": 8.733733543966449e-06,
+      "loss": 0.4514,
+      "step": 7315
+    },
+    {
+      "epoch": 1.6319428953825563,
+      "grad_norm": 0.18014779686927795,
+      "learning_rate": 8.731399093893256e-06,
+      "loss": 0.4688,
+      "step": 7316
+    },
+    {
+      "epoch": 1.6321659602944458,
+      "grad_norm": 0.1781376600265503,
+      "learning_rate": 8.729064714083163e-06,
+      "loss": 0.4552,
+      "step": 7317
+    },
+    {
+      "epoch": 1.632389025206335,
+      "grad_norm": 0.17916658520698547,
+      "learning_rate": 8.726730404665458e-06,
+      "loss": 0.4463,
+      "step": 7318
+    },
+    {
+      "epoch": 1.6326120901182244,
+      "grad_norm": 0.17721021175384521,
+      "learning_rate": 8.724396165769435e-06,
+      "loss": 0.4552,
+      "step": 7319
+    },
+    {
+      "epoch": 1.6328351550301137,
+      "grad_norm": 0.18291765451431274,
+      "learning_rate": 8.722061997524374e-06,
+      "loss": 0.4492,
+      "step": 7320
+    },
+    {
+      "epoch": 1.6330582199420032,
+      "grad_norm": 0.19074088335037231,
+      "learning_rate": 8.719727900059559e-06,
+      "loss": 0.4784,
+      "step": 7321
+    },
+    {
+      "epoch": 1.6332812848538925,
+      "grad_norm": 0.18184469640254974,
+      "learning_rate": 8.717393873504265e-06,
+      "loss": 0.4602,
+      "step": 7322
+    },
+    {
+      "epoch": 1.633504349765782,
+      "grad_norm": 0.17274357378482819,
+      "learning_rate": 8.715059917987766e-06,
+      "loss": 0.4394,
+      "step": 7323
+    },
+    {
+      "epoch": 1.6337274146776712,
+      "grad_norm": 0.1714925616979599,
+      "learning_rate": 8.712726033639334e-06,
+      "loss": 0.4277,
+      "step": 7324
+    },
+    {
+      "epoch": 1.6339504795895605,
+      "grad_norm": 0.17422372102737427,
+      "learning_rate": 8.710392220588229e-06,
+      "loss": 0.4585,
+      "step": 7325
+    },
+    {
+      "epoch": 1.6341735445014498,
+      "grad_norm": 0.17235322296619415,
+      "learning_rate": 8.708058478963717e-06,
+      "loss": 0.4489,
+      "step": 7326
+    },
+    {
+      "epoch": 1.6343966094133393,
+      "grad_norm": 0.17422306537628174,
+      "learning_rate": 8.70572480889505e-06,
+      "loss": 0.4624,
+      "step": 7327
+    },
+    {
+      "epoch": 1.6346196743252286,
+      "grad_norm": 0.1853763461112976,
+      "learning_rate": 8.703391210511486e-06,
+      "loss": 0.4723,
+      "step": 7328
+    },
+    {
+      "epoch": 1.634842739237118,
+      "grad_norm": 0.17920438945293427,
+      "learning_rate": 8.701057683942274e-06,
+      "loss": 0.4596,
+      "step": 7329
+    },
+    {
+      "epoch": 1.6350658041490074,
+      "grad_norm": 0.17522959411144257,
+      "learning_rate": 8.698724229316658e-06,
+      "loss": 0.4501,
+      "step": 7330
+    },
+    {
+      "epoch": 1.6352888690608967,
+      "grad_norm": 0.17898008227348328,
+      "learning_rate": 8.696390846763877e-06,
+      "loss": 0.4416,
+      "step": 7331
+    },
+    {
+      "epoch": 1.635511933972786,
+      "grad_norm": 0.1786876916885376,
+      "learning_rate": 8.694057536413175e-06,
+      "loss": 0.4541,
+      "step": 7332
+    },
+    {
+      "epoch": 1.6357349988846754,
+      "grad_norm": 0.18402035534381866,
+      "learning_rate": 8.691724298393777e-06,
+      "loss": 0.4781,
+      "step": 7333
+    },
+    {
+      "epoch": 1.635958063796565,
+      "grad_norm": 0.18904945254325867,
+      "learning_rate": 8.68939113283492e-06,
+      "loss": 0.4541,
+      "step": 7334
+    },
+    {
+      "epoch": 1.6361811287084542,
+      "grad_norm": 0.18246184289455414,
+      "learning_rate": 8.687058039865823e-06,
+      "loss": 0.4711,
+      "step": 7335
+    },
+    {
+      "epoch": 1.6364041936203435,
+      "grad_norm": 0.1694568693637848,
+      "learning_rate": 8.684725019615714e-06,
+      "loss": 0.4648,
+      "step": 7336
+    },
+    {
+      "epoch": 1.6366272585322328,
+      "grad_norm": 0.17736688256263733,
+      "learning_rate": 8.682392072213804e-06,
+      "loss": 0.4192,
+      "step": 7337
+    },
+    {
+      "epoch": 1.6368503234441223,
+      "grad_norm": 0.16740672290325165,
+      "learning_rate": 8.680059197789311e-06,
+      "loss": 0.4591,
+      "step": 7338
+    },
+    {
+      "epoch": 1.6370733883560116,
+      "grad_norm": 0.17872267961502075,
+      "learning_rate": 8.67772639647144e-06,
+      "loss": 0.4453,
+      "step": 7339
+    },
+    {
+      "epoch": 1.637296453267901,
+      "grad_norm": 0.18077607452869415,
+      "learning_rate": 8.675393668389402e-06,
+      "loss": 0.459,
+      "step": 7340
+    },
+    {
+      "epoch": 1.6375195181797904,
+      "grad_norm": 0.17609618604183197,
+      "learning_rate": 8.67306101367239e-06,
+      "loss": 0.4774,
+      "step": 7341
+    },
+    {
+      "epoch": 1.6377425830916796,
+      "grad_norm": 0.17567932605743408,
+      "learning_rate": 8.670728432449608e-06,
+      "loss": 0.4577,
+      "step": 7342
+    },
+    {
+      "epoch": 1.637965648003569,
+      "grad_norm": 0.18669098615646362,
+      "learning_rate": 8.66839592485025e-06,
+      "loss": 0.4321,
+      "step": 7343
+    },
+    {
+      "epoch": 1.6381887129154584,
+      "grad_norm": 0.1745385229587555,
+      "learning_rate": 8.666063491003499e-06,
+      "loss": 0.4595,
+      "step": 7344
+    },
+    {
+      "epoch": 1.6384117778273477,
+      "grad_norm": 0.16913901269435883,
+      "learning_rate": 8.663731131038544e-06,
+      "loss": 0.4179,
+      "step": 7345
+    },
+    {
+      "epoch": 1.6386348427392372,
+      "grad_norm": 0.16662278771400452,
+      "learning_rate": 8.661398845084562e-06,
+      "loss": 0.4333,
+      "step": 7346
+    },
+    {
+      "epoch": 1.6388579076511265,
+      "grad_norm": 0.17218153178691864,
+      "learning_rate": 8.659066633270736e-06,
+      "loss": 0.4324,
+      "step": 7347
+    },
+    {
+      "epoch": 1.6390809725630158,
+      "grad_norm": 0.1764010488986969,
+      "learning_rate": 8.65673449572623e-06,
+      "loss": 0.44,
+      "step": 7348
+    },
+    {
+      "epoch": 1.639304037474905,
+      "grad_norm": 0.17202350497245789,
+      "learning_rate": 8.65440243258022e-06,
+      "loss": 0.4611,
+      "step": 7349
+    },
+    {
+      "epoch": 1.6395271023867946,
+      "grad_norm": 0.17550015449523926,
+      "learning_rate": 8.652070443961866e-06,
+      "loss": 0.4641,
+      "step": 7350
+    },
+    {
+      "epoch": 1.639750167298684,
+      "grad_norm": 0.18475106358528137,
+      "learning_rate": 8.649738530000333e-06,
+      "loss": 0.4537,
+      "step": 7351
+    },
+    {
+      "epoch": 1.6399732322105733,
+      "grad_norm": 0.1760861724615097,
+      "learning_rate": 8.647406690824769e-06,
+      "loss": 0.4216,
+      "step": 7352
+    },
+    {
+      "epoch": 1.6401962971224626,
+      "grad_norm": 0.20794068276882172,
+      "learning_rate": 8.645074926564334e-06,
+      "loss": 0.4492,
+      "step": 7353
+    },
+    {
+      "epoch": 1.640419362034352,
+      "grad_norm": 0.1916939616203308,
+      "learning_rate": 8.64274323734817e-06,
+      "loss": 0.4492,
+      "step": 7354
+    },
+    {
+      "epoch": 1.6406424269462414,
+      "grad_norm": 0.17501096427440643,
+      "learning_rate": 8.640411623305425e-06,
+      "loss": 0.4365,
+      "step": 7355
+    },
+    {
+      "epoch": 1.6408654918581307,
+      "grad_norm": 0.1728835254907608,
+      "learning_rate": 8.638080084565235e-06,
+      "loss": 0.444,
+      "step": 7356
+    },
+    {
+      "epoch": 1.6410885567700202,
+      "grad_norm": 0.17348189651966095,
+      "learning_rate": 8.63574862125674e-06,
+      "loss": 0.4614,
+      "step": 7357
+    },
+    {
+      "epoch": 1.6413116216819095,
+      "grad_norm": 0.17011727392673492,
+      "learning_rate": 8.633417233509063e-06,
+      "loss": 0.4539,
+      "step": 7358
+    },
+    {
+      "epoch": 1.6415346865937988,
+      "grad_norm": 0.17248336970806122,
+      "learning_rate": 8.63108592145134e-06,
+      "loss": 0.4545,
+      "step": 7359
+    },
+    {
+      "epoch": 1.641757751505688,
+      "grad_norm": 0.17574246227741241,
+      "learning_rate": 8.628754685212685e-06,
+      "loss": 0.458,
+      "step": 7360
+    },
+    {
+      "epoch": 1.6419808164175775,
+      "grad_norm": 0.19359445571899414,
+      "learning_rate": 8.626423524922224e-06,
+      "loss": 0.4388,
+      "step": 7361
+    },
+    {
+      "epoch": 1.642203881329467,
+      "grad_norm": 0.17245686054229736,
+      "learning_rate": 8.624092440709066e-06,
+      "loss": 0.4345,
+      "step": 7362
+    },
+    {
+      "epoch": 1.6424269462413563,
+      "grad_norm": 0.180564284324646,
+      "learning_rate": 8.621761432702325e-06,
+      "loss": 0.4665,
+      "step": 7363
+    },
+    {
+      "epoch": 1.6426500111532456,
+      "grad_norm": 0.18473738431930542,
+      "learning_rate": 8.619430501031106e-06,
+      "loss": 0.4364,
+      "step": 7364
+    },
+    {
+      "epoch": 1.6428730760651349,
+      "grad_norm": 0.18806539475917816,
+      "learning_rate": 8.617099645824509e-06,
+      "loss": 0.4498,
+      "step": 7365
+    },
+    {
+      "epoch": 1.6430961409770242,
+      "grad_norm": 0.1734282672405243,
+      "learning_rate": 8.614768867211634e-06,
+      "loss": 0.4782,
+      "step": 7366
+    },
+    {
+      "epoch": 1.6433192058889137,
+      "grad_norm": 0.183370441198349,
+      "learning_rate": 8.612438165321571e-06,
+      "loss": 0.4892,
+      "step": 7367
+    },
+    {
+      "epoch": 1.6435422708008032,
+      "grad_norm": 0.1682674139738083,
+      "learning_rate": 8.610107540283411e-06,
+      "loss": 0.4256,
+      "step": 7368
+    },
+    {
+      "epoch": 1.6437653357126925,
+      "grad_norm": 0.17661243677139282,
+      "learning_rate": 8.607776992226238e-06,
+      "loss": 0.4693,
+      "step": 7369
+    },
+    {
+      "epoch": 1.6439884006245817,
+      "grad_norm": 0.2739085257053375,
+      "learning_rate": 8.605446521279132e-06,
+      "loss": 0.461,
+      "step": 7370
+    },
+    {
+      "epoch": 1.644211465536471,
+      "grad_norm": 0.1764136105775833,
+      "learning_rate": 8.60311612757117e-06,
+      "loss": 0.4833,
+      "step": 7371
+    },
+    {
+      "epoch": 1.6444345304483605,
+      "grad_norm": 0.17037895321846008,
+      "learning_rate": 8.600785811231426e-06,
+      "loss": 0.474,
+      "step": 7372
+    },
+    {
+      "epoch": 1.6446575953602498,
+      "grad_norm": 0.17914576828479767,
+      "learning_rate": 8.598455572388961e-06,
+      "loss": 0.4404,
+      "step": 7373
+    },
+    {
+      "epoch": 1.6448806602721393,
+      "grad_norm": 0.1834004670381546,
+      "learning_rate": 8.596125411172846e-06,
+      "loss": 0.4738,
+      "step": 7374
+    },
+    {
+      "epoch": 1.6451037251840286,
+      "grad_norm": 0.542524516582489,
+      "learning_rate": 8.593795327712135e-06,
+      "loss": 0.4519,
+      "step": 7375
+    },
+    {
+      "epoch": 1.6453267900959179,
+      "grad_norm": 0.17706745862960815,
+      "learning_rate": 8.591465322135886e-06,
+      "loss": 0.47,
+      "step": 7376
+    },
+    {
+      "epoch": 1.6455498550078071,
+      "grad_norm": 0.18335482478141785,
+      "learning_rate": 8.589135394573146e-06,
+      "loss": 0.4763,
+      "step": 7377
+    },
+    {
+      "epoch": 1.6457729199196967,
+      "grad_norm": 0.19544199109077454,
+      "learning_rate": 8.586805545152962e-06,
+      "loss": 0.4459,
+      "step": 7378
+    },
+    {
+      "epoch": 1.6459959848315862,
+      "grad_norm": 0.19434663653373718,
+      "learning_rate": 8.584475774004374e-06,
+      "loss": 0.4611,
+      "step": 7379
+    },
+    {
+      "epoch": 1.6462190497434754,
+      "grad_norm": 0.17542743682861328,
+      "learning_rate": 8.582146081256428e-06,
+      "loss": 0.4435,
+      "step": 7380
+    },
+    {
+      "epoch": 1.6464421146553647,
+      "grad_norm": 0.17821446061134338,
+      "learning_rate": 8.579816467038144e-06,
+      "loss": 0.4674,
+      "step": 7381
+    },
+    {
+      "epoch": 1.646665179567254,
+      "grad_norm": 0.16441945731639862,
+      "learning_rate": 8.577486931478563e-06,
+      "loss": 0.4263,
+      "step": 7382
+    },
+    {
+      "epoch": 1.6468882444791433,
+      "grad_norm": 0.17425528168678284,
+      "learning_rate": 8.5751574747067e-06,
+      "loss": 0.4491,
+      "step": 7383
+    },
+    {
+      "epoch": 1.6471113093910328,
+      "grad_norm": 0.17746862769126892,
+      "learning_rate": 8.572828096851577e-06,
+      "loss": 0.4633,
+      "step": 7384
+    },
+    {
+      "epoch": 1.6473343743029223,
+      "grad_norm": 0.16490298509597778,
+      "learning_rate": 8.570498798042217e-06,
+      "loss": 0.4116,
+      "step": 7385
+    },
+    {
+      "epoch": 1.6475574392148116,
+      "grad_norm": 0.17692579329013824,
+      "learning_rate": 8.568169578407624e-06,
+      "loss": 0.4585,
+      "step": 7386
+    },
+    {
+      "epoch": 1.6477805041267009,
+      "grad_norm": 0.18186244368553162,
+      "learning_rate": 8.565840438076805e-06,
+      "loss": 0.4792,
+      "step": 7387
+    },
+    {
+      "epoch": 1.6480035690385901,
+      "grad_norm": 0.16964338719844818,
+      "learning_rate": 8.563511377178764e-06,
+      "loss": 0.4254,
+      "step": 7388
+    },
+    {
+      "epoch": 1.6482266339504796,
+      "grad_norm": 0.17917568981647491,
+      "learning_rate": 8.561182395842497e-06,
+      "loss": 0.4823,
+      "step": 7389
+    },
+    {
+      "epoch": 1.648449698862369,
+      "grad_norm": 0.1826728731393814,
+      "learning_rate": 8.558853494197e-06,
+      "loss": 0.4297,
+      "step": 7390
+    },
+    {
+      "epoch": 1.6486727637742584,
+      "grad_norm": 0.17334018647670746,
+      "learning_rate": 8.556524672371263e-06,
+      "loss": 0.4342,
+      "step": 7391
+    },
+    {
+      "epoch": 1.6488958286861477,
+      "grad_norm": 0.17298614978790283,
+      "learning_rate": 8.55419593049427e-06,
+      "loss": 0.4605,
+      "step": 7392
+    },
+    {
+      "epoch": 1.649118893598037,
+      "grad_norm": 0.1751953512430191,
+      "learning_rate": 8.551867268694999e-06,
+      "loss": 0.4426,
+      "step": 7393
+    },
+    {
+      "epoch": 1.6493419585099263,
+      "grad_norm": 0.17890918254852295,
+      "learning_rate": 8.54953868710243e-06,
+      "loss": 0.4479,
+      "step": 7394
+    },
+    {
+      "epoch": 1.6495650234218158,
+      "grad_norm": 0.17455795407295227,
+      "learning_rate": 8.54721018584553e-06,
+      "loss": 0.4452,
+      "step": 7395
+    },
+    {
+      "epoch": 1.6497880883337053,
+      "grad_norm": 0.1724436730146408,
+      "learning_rate": 8.54488176505327e-06,
+      "loss": 0.4326,
+      "step": 7396
+    },
+    {
+      "epoch": 1.6500111532455946,
+      "grad_norm": 0.17377914488315582,
+      "learning_rate": 8.542553424854608e-06,
+      "loss": 0.4589,
+      "step": 7397
+    },
+    {
+      "epoch": 1.6502342181574838,
+      "grad_norm": 0.17500680685043335,
+      "learning_rate": 8.540225165378509e-06,
+      "loss": 0.432,
+      "step": 7398
+    },
+    {
+      "epoch": 1.6504572830693731,
+      "grad_norm": 0.17203043401241302,
+      "learning_rate": 8.53789698675392e-06,
+      "loss": 0.4801,
+      "step": 7399
+    },
+    {
+      "epoch": 1.6506803479812624,
+      "grad_norm": 0.17373989522457123,
+      "learning_rate": 8.535568889109794e-06,
+      "loss": 0.4431,
+      "step": 7400
+    },
+    {
+      "epoch": 1.650903412893152,
+      "grad_norm": 0.1792883723974228,
+      "learning_rate": 8.533240872575073e-06,
+      "loss": 0.4531,
+      "step": 7401
+    },
+    {
+      "epoch": 1.6511264778050414,
+      "grad_norm": 0.17690402269363403,
+      "learning_rate": 8.530912937278702e-06,
+      "loss": 0.4548,
+      "step": 7402
+    },
+    {
+      "epoch": 1.6513495427169307,
+      "grad_norm": 0.17443451285362244,
+      "learning_rate": 8.52858508334961e-06,
+      "loss": 0.4478,
+      "step": 7403
+    },
+    {
+      "epoch": 1.65157260762882,
+      "grad_norm": 0.1743832528591156,
+      "learning_rate": 8.526257310916734e-06,
+      "loss": 0.4732,
+      "step": 7404
+    },
+    {
+      "epoch": 1.6517956725407092,
+      "grad_norm": 0.1756555140018463,
+      "learning_rate": 8.523929620108996e-06,
+      "loss": 0.4439,
+      "step": 7405
+    },
+    {
+      "epoch": 1.6520187374525988,
+      "grad_norm": 0.19035564363002777,
+      "learning_rate": 8.521602011055324e-06,
+      "loss": 0.4677,
+      "step": 7406
+    },
+    {
+      "epoch": 1.652241802364488,
+      "grad_norm": 0.18924179673194885,
+      "learning_rate": 8.519274483884627e-06,
+      "loss": 0.463,
+      "step": 7407
+    },
+    {
+      "epoch": 1.6524648672763775,
+      "grad_norm": 0.18173234164714813,
+      "learning_rate": 8.516947038725826e-06,
+      "loss": 0.4318,
+      "step": 7408
+    },
+    {
+      "epoch": 1.6526879321882668,
+      "grad_norm": 0.175716370344162,
+      "learning_rate": 8.514619675707828e-06,
+      "loss": 0.4292,
+      "step": 7409
+    },
+    {
+      "epoch": 1.652910997100156,
+      "grad_norm": 0.17146100103855133,
+      "learning_rate": 8.512292394959533e-06,
+      "loss": 0.4241,
+      "step": 7410
+    },
+    {
+      "epoch": 1.6531340620120454,
+      "grad_norm": 0.17284603416919708,
+      "learning_rate": 8.509965196609846e-06,
+      "loss": 0.4442,
+      "step": 7411
+    },
+    {
+      "epoch": 1.6533571269239349,
+      "grad_norm": 0.18271994590759277,
+      "learning_rate": 8.507638080787657e-06,
+      "loss": 0.451,
+      "step": 7412
+    },
+    {
+      "epoch": 1.6535801918358244,
+      "grad_norm": 0.17891302704811096,
+      "learning_rate": 8.50531104762186e-06,
+      "loss": 0.4496,
+      "step": 7413
+    },
+    {
+      "epoch": 1.6538032567477137,
+      "grad_norm": 0.19851596653461456,
+      "learning_rate": 8.502984097241338e-06,
+      "loss": 0.4715,
+      "step": 7414
+    },
+    {
+      "epoch": 1.654026321659603,
+      "grad_norm": 0.18337850272655487,
+      "learning_rate": 8.500657229774975e-06,
+      "loss": 0.439,
+      "step": 7415
+    },
+    {
+      "epoch": 1.6542493865714922,
+      "grad_norm": 0.18898166716098785,
+      "learning_rate": 8.498330445351643e-06,
+      "loss": 0.4641,
+      "step": 7416
+    },
+    {
+      "epoch": 1.6544724514833815,
+      "grad_norm": 0.18804045021533966,
+      "learning_rate": 8.49600374410022e-06,
+      "loss": 0.4488,
+      "step": 7417
+    },
+    {
+      "epoch": 1.654695516395271,
+      "grad_norm": 0.18279941380023956,
+      "learning_rate": 8.493677126149569e-06,
+      "loss": 0.4395,
+      "step": 7418
+    },
+    {
+      "epoch": 1.6549185813071605,
+      "grad_norm": 0.18200235068798065,
+      "learning_rate": 8.491350591628554e-06,
+      "loss": 0.4475,
+      "step": 7419
+    },
+    {
+      "epoch": 1.6551416462190498,
+      "grad_norm": 0.1727295070886612,
+      "learning_rate": 8.489024140666032e-06,
+      "loss": 0.4627,
+      "step": 7420
+    },
+    {
+      "epoch": 1.655364711130939,
+      "grad_norm": 0.17785604298114777,
+      "learning_rate": 8.486697773390859e-06,
+      "loss": 0.4605,
+      "step": 7421
+    },
+    {
+      "epoch": 1.6555877760428284,
+      "grad_norm": 0.18455497920513153,
+      "learning_rate": 8.48437148993188e-06,
+      "loss": 0.4463,
+      "step": 7422
+    },
+    {
+      "epoch": 1.6558108409547179,
+      "grad_norm": 0.17284157872200012,
+      "learning_rate": 8.482045290417946e-06,
+      "loss": 0.4735,
+      "step": 7423
+    },
+    {
+      "epoch": 1.6560339058666071,
+      "grad_norm": 0.1757441759109497,
+      "learning_rate": 8.479719174977887e-06,
+      "loss": 0.4296,
+      "step": 7424
+    },
+    {
+      "epoch": 1.6562569707784967,
+      "grad_norm": 0.17652331292629242,
+      "learning_rate": 8.477393143740546e-06,
+      "loss": 0.4289,
+      "step": 7425
+    },
+    {
+      "epoch": 1.656480035690386,
+      "grad_norm": 0.17844471335411072,
+      "learning_rate": 8.475067196834749e-06,
+      "loss": 0.4624,
+      "step": 7426
+    },
+    {
+      "epoch": 1.6567031006022752,
+      "grad_norm": 0.1853104829788208,
+      "learning_rate": 8.472741334389322e-06,
+      "loss": 0.444,
+      "step": 7427
+    },
+    {
+      "epoch": 1.6569261655141645,
+      "grad_norm": 0.20148655772209167,
+      "learning_rate": 8.47041555653309e-06,
+      "loss": 0.4412,
+      "step": 7428
+    },
+    {
+      "epoch": 1.657149230426054,
+      "grad_norm": 0.17949336767196655,
+      "learning_rate": 8.468089863394864e-06,
+      "loss": 0.4799,
+      "step": 7429
+    },
+    {
+      "epoch": 1.6573722953379435,
+      "grad_norm": 0.1833929866552353,
+      "learning_rate": 8.465764255103457e-06,
+      "loss": 0.459,
+      "step": 7430
+    },
+    {
+      "epoch": 1.6575953602498328,
+      "grad_norm": 0.16851405799388885,
+      "learning_rate": 8.463438731787677e-06,
+      "loss": 0.4509,
+      "step": 7431
+    },
+    {
+      "epoch": 1.657818425161722,
+      "grad_norm": 0.17039094865322113,
+      "learning_rate": 8.461113293576325e-06,
+      "loss": 0.4591,
+      "step": 7432
+    },
+    {
+      "epoch": 1.6580414900736113,
+      "grad_norm": 0.1713404804468155,
+      "learning_rate": 8.4587879405982e-06,
+      "loss": 0.4469,
+      "step": 7433
+    },
+    {
+      "epoch": 1.6582645549855006,
+      "grad_norm": 0.1776912361383438,
+      "learning_rate": 8.456462672982092e-06,
+      "loss": 0.4526,
+      "step": 7434
+    },
+    {
+      "epoch": 1.6584876198973901,
+      "grad_norm": 0.1807934045791626,
+      "learning_rate": 8.45413749085679e-06,
+      "loss": 0.4689,
+      "step": 7435
+    },
+    {
+      "epoch": 1.6587106848092796,
+      "grad_norm": 0.17492172122001648,
+      "learning_rate": 8.451812394351078e-06,
+      "loss": 0.4605,
+      "step": 7436
+    },
+    {
+      "epoch": 1.658933749721169,
+      "grad_norm": 0.1770438402891159,
+      "learning_rate": 8.449487383593734e-06,
+      "loss": 0.4603,
+      "step": 7437
+    },
+    {
+      "epoch": 1.6591568146330582,
+      "grad_norm": 0.18041646480560303,
+      "learning_rate": 8.447162458713534e-06,
+      "loss": 0.4534,
+      "step": 7438
+    },
+    {
+      "epoch": 1.6593798795449475,
+      "grad_norm": 0.1849108189344406,
+      "learning_rate": 8.444837619839243e-06,
+      "loss": 0.4547,
+      "step": 7439
+    },
+    {
+      "epoch": 1.659602944456837,
+      "grad_norm": 0.18359196186065674,
+      "learning_rate": 8.442512867099627e-06,
+      "loss": 0.4545,
+      "step": 7440
+    },
+    {
+      "epoch": 1.6598260093687263,
+      "grad_norm": 0.19102846086025238,
+      "learning_rate": 8.440188200623445e-06,
+      "loss": 0.4257,
+      "step": 7441
+    },
+    {
+      "epoch": 1.6600490742806158,
+      "grad_norm": 0.1707092970609665,
+      "learning_rate": 8.437863620539454e-06,
+      "loss": 0.4307,
+      "step": 7442
+    },
+    {
+      "epoch": 1.660272139192505,
+      "grad_norm": 0.18710510432720184,
+      "learning_rate": 8.4355391269764e-06,
+      "loss": 0.4453,
+      "step": 7443
+    },
+    {
+      "epoch": 1.6604952041043943,
+      "grad_norm": 0.18393000960350037,
+      "learning_rate": 8.43321472006303e-06,
+      "loss": 0.448,
+      "step": 7444
+    },
+    {
+      "epoch": 1.6607182690162836,
+      "grad_norm": 0.16592559218406677,
+      "learning_rate": 8.430890399928085e-06,
+      "loss": 0.4243,
+      "step": 7445
+    },
+    {
+      "epoch": 1.6609413339281731,
+      "grad_norm": 0.16804459691047668,
+      "learning_rate": 8.4285661667003e-06,
+      "loss": 0.4357,
+      "step": 7446
+    },
+    {
+      "epoch": 1.6611643988400626,
+      "grad_norm": 0.1712721884250641,
+      "learning_rate": 8.426242020508405e-06,
+      "loss": 0.4403,
+      "step": 7447
+    },
+    {
+      "epoch": 1.661387463751952,
+      "grad_norm": 0.17568182945251465,
+      "learning_rate": 8.423917961481124e-06,
+      "loss": 0.4489,
+      "step": 7448
+    },
+    {
+      "epoch": 1.6616105286638412,
+      "grad_norm": 0.16803157329559326,
+      "learning_rate": 8.421593989747184e-06,
+      "loss": 0.4293,
+      "step": 7449
+    },
+    {
+      "epoch": 1.6618335935757305,
+      "grad_norm": 0.18901732563972473,
+      "learning_rate": 8.419270105435294e-06,
+      "loss": 0.4637,
+      "step": 7450
+    },
+    {
+      "epoch": 1.6620566584876197,
+      "grad_norm": 0.18891791999340057,
+      "learning_rate": 8.416946308674173e-06,
+      "loss": 0.4602,
+      "step": 7451
+    },
+    {
+      "epoch": 1.6622797233995092,
+      "grad_norm": 0.17820779979228973,
+      "learning_rate": 8.414622599592518e-06,
+      "loss": 0.4555,
+      "step": 7452
+    },
+    {
+      "epoch": 1.6625027883113987,
+      "grad_norm": 0.18127723038196564,
+      "learning_rate": 8.41229897831904e-06,
+      "loss": 0.4512,
+      "step": 7453
+    },
+    {
+      "epoch": 1.662725853223288,
+      "grad_norm": 0.18046575784683228,
+      "learning_rate": 8.40997544498243e-06,
+      "loss": 0.4635,
+      "step": 7454
+    },
+    {
+      "epoch": 1.6629489181351773,
+      "grad_norm": 0.1841399371623993,
+      "learning_rate": 8.407651999711383e-06,
+      "loss": 0.4605,
+      "step": 7455
+    },
+    {
+      "epoch": 1.6631719830470666,
+      "grad_norm": 0.17511752247810364,
+      "learning_rate": 8.405328642634582e-06,
+      "loss": 0.4308,
+      "step": 7456
+    },
+    {
+      "epoch": 1.663395047958956,
+      "grad_norm": 0.17428848147392273,
+      "learning_rate": 8.403005373880713e-06,
+      "loss": 0.4426,
+      "step": 7457
+    },
+    {
+      "epoch": 1.6636181128708454,
+      "grad_norm": 0.18597403168678284,
+      "learning_rate": 8.400682193578451e-06,
+      "loss": 0.4476,
+      "step": 7458
+    },
+    {
+      "epoch": 1.6638411777827349,
+      "grad_norm": 0.17569150030612946,
+      "learning_rate": 8.398359101856471e-06,
+      "loss": 0.4161,
+      "step": 7459
+    },
+    {
+      "epoch": 1.6640642426946242,
+      "grad_norm": 0.17648978531360626,
+      "learning_rate": 8.396036098843438e-06,
+      "loss": 0.4596,
+      "step": 7460
+    },
+    {
+      "epoch": 1.6642873076065134,
+      "grad_norm": 0.1786419153213501,
+      "learning_rate": 8.393713184668015e-06,
+      "loss": 0.4623,
+      "step": 7461
+    },
+    {
+      "epoch": 1.6645103725184027,
+      "grad_norm": 0.18259434401988983,
+      "learning_rate": 8.391390359458858e-06,
+      "loss": 0.4703,
+      "step": 7462
+    },
+    {
+      "epoch": 1.6647334374302922,
+      "grad_norm": 0.18008796870708466,
+      "learning_rate": 8.389067623344625e-06,
+      "loss": 0.464,
+      "step": 7463
+    },
+    {
+      "epoch": 1.6649565023421817,
+      "grad_norm": 0.20690587162971497,
+      "learning_rate": 8.386744976453958e-06,
+      "loss": 0.457,
+      "step": 7464
+    },
+    {
+      "epoch": 1.665179567254071,
+      "grad_norm": 0.21626955270767212,
+      "learning_rate": 8.384422418915503e-06,
+      "loss": 0.4171,
+      "step": 7465
+    },
+    {
+      "epoch": 1.6654026321659603,
+      "grad_norm": 0.16983038187026978,
+      "learning_rate": 8.382099950857895e-06,
+      "loss": 0.434,
+      "step": 7466
+    },
+    {
+      "epoch": 1.6656256970778496,
+      "grad_norm": 0.18319158256053925,
+      "learning_rate": 8.379777572409771e-06,
+      "loss": 0.4808,
+      "step": 7467
+    },
+    {
+      "epoch": 1.6658487619897389,
+      "grad_norm": 0.2683338522911072,
+      "learning_rate": 8.377455283699758e-06,
+      "loss": 0.4772,
+      "step": 7468
+    },
+    {
+      "epoch": 1.6660718269016284,
+      "grad_norm": 0.18146146833896637,
+      "learning_rate": 8.375133084856475e-06,
+      "loss": 0.4226,
+      "step": 7469
+    },
+    {
+      "epoch": 1.6662948918135179,
+      "grad_norm": 0.17817704379558563,
+      "learning_rate": 8.372810976008543e-06,
+      "loss": 0.4272,
+      "step": 7470
+    },
+    {
+      "epoch": 1.6665179567254071,
+      "grad_norm": 0.18885549902915955,
+      "learning_rate": 8.370488957284574e-06,
+      "loss": 0.4722,
+      "step": 7471
+    },
+    {
+      "epoch": 1.6667410216372964,
+      "grad_norm": 0.18117155134677887,
+      "learning_rate": 8.368167028813176e-06,
+      "loss": 0.4377,
+      "step": 7472
+    },
+    {
+      "epoch": 1.6669640865491857,
+      "grad_norm": 0.19117340445518494,
+      "learning_rate": 8.365845190722955e-06,
+      "loss": 0.4563,
+      "step": 7473
+    },
+    {
+      "epoch": 1.6671871514610752,
+      "grad_norm": 0.1751892864704132,
+      "learning_rate": 8.363523443142503e-06,
+      "loss": 0.4433,
+      "step": 7474
+    },
+    {
+      "epoch": 1.6674102163729645,
+      "grad_norm": 0.18443763256072998,
+      "learning_rate": 8.36120178620042e-06,
+      "loss": 0.439,
+      "step": 7475
+    },
+    {
+      "epoch": 1.667633281284854,
+      "grad_norm": 0.18026113510131836,
+      "learning_rate": 8.358880220025288e-06,
+      "loss": 0.4275,
+      "step": 7476
+    },
+    {
+      "epoch": 1.6678563461967433,
+      "grad_norm": 0.17562150955200195,
+      "learning_rate": 8.356558744745695e-06,
+      "loss": 0.4685,
+      "step": 7477
+    },
+    {
+      "epoch": 1.6680794111086326,
+      "grad_norm": 0.1727527379989624,
+      "learning_rate": 8.354237360490212e-06,
+      "loss": 0.4602,
+      "step": 7478
+    },
+    {
+      "epoch": 1.6683024760205218,
+      "grad_norm": 0.1706780195236206,
+      "learning_rate": 8.351916067387421e-06,
+      "loss": 0.4431,
+      "step": 7479
+    },
+    {
+      "epoch": 1.6685255409324113,
+      "grad_norm": 0.19286386668682098,
+      "learning_rate": 8.349594865565882e-06,
+      "loss": 0.4496,
+      "step": 7480
+    },
+    {
+      "epoch": 1.6687486058443008,
+      "grad_norm": 0.2684648931026459,
+      "learning_rate": 8.347273755154164e-06,
+      "loss": 0.4839,
+      "step": 7481
+    },
+    {
+      "epoch": 1.6689716707561901,
+      "grad_norm": 0.1720086932182312,
+      "learning_rate": 8.344952736280819e-06,
+      "loss": 0.4253,
+      "step": 7482
+    },
+    {
+      "epoch": 1.6691947356680794,
+      "grad_norm": 0.17519108951091766,
+      "learning_rate": 8.342631809074403e-06,
+      "loss": 0.4363,
+      "step": 7483
+    },
+    {
+      "epoch": 1.6694178005799687,
+      "grad_norm": 0.18655794858932495,
+      "learning_rate": 8.340310973663461e-06,
+      "loss": 0.4811,
+      "step": 7484
+    },
+    {
+      "epoch": 1.669640865491858,
+      "grad_norm": 0.16531935334205627,
+      "learning_rate": 8.33799023017654e-06,
+      "loss": 0.431,
+      "step": 7485
+    },
+    {
+      "epoch": 1.6698639304037475,
+      "grad_norm": 0.2258535474538803,
+      "learning_rate": 8.335669578742172e-06,
+      "loss": 0.4407,
+      "step": 7486
+    },
+    {
+      "epoch": 1.670086995315637,
+      "grad_norm": 0.17526793479919434,
+      "learning_rate": 8.333349019488893e-06,
+      "loss": 0.4531,
+      "step": 7487
+    },
+    {
+      "epoch": 1.6703100602275263,
+      "grad_norm": 0.17840759456157684,
+      "learning_rate": 8.331028552545228e-06,
+      "loss": 0.4374,
+      "step": 7488
+    },
+    {
+      "epoch": 1.6705331251394155,
+      "grad_norm": 0.1826542466878891,
+      "learning_rate": 8.328708178039702e-06,
+      "loss": 0.4881,
+      "step": 7489
+    },
+    {
+      "epoch": 1.6707561900513048,
+      "grad_norm": 0.17568182945251465,
+      "learning_rate": 8.326387896100827e-06,
+      "loss": 0.4607,
+      "step": 7490
+    },
+    {
+      "epoch": 1.6709792549631943,
+      "grad_norm": 0.1761264204978943,
+      "learning_rate": 8.324067706857121e-06,
+      "loss": 0.4399,
+      "step": 7491
+    },
+    {
+      "epoch": 1.6712023198750836,
+      "grad_norm": 0.16690705716609955,
+      "learning_rate": 8.321747610437084e-06,
+      "loss": 0.4417,
+      "step": 7492
+    },
+    {
+      "epoch": 1.6714253847869731,
+      "grad_norm": 0.17830541729927063,
+      "learning_rate": 8.319427606969223e-06,
+      "loss": 0.4437,
+      "step": 7493
+    },
+    {
+      "epoch": 1.6716484496988624,
+      "grad_norm": 0.271932452917099,
+      "learning_rate": 8.317107696582031e-06,
+      "loss": 0.4486,
+      "step": 7494
+    },
+    {
+      "epoch": 1.6718715146107517,
+      "grad_norm": 0.19442439079284668,
+      "learning_rate": 8.314787879404002e-06,
+      "loss": 0.474,
+      "step": 7495
+    },
+    {
+      "epoch": 1.672094579522641,
+      "grad_norm": 0.1841454654932022,
+      "learning_rate": 8.312468155563623e-06,
+      "loss": 0.4689,
+      "step": 7496
+    },
+    {
+      "epoch": 1.6723176444345305,
+      "grad_norm": 0.18820422887802124,
+      "learning_rate": 8.310148525189367e-06,
+      "loss": 0.4349,
+      "step": 7497
+    },
+    {
+      "epoch": 1.67254070934642,
+      "grad_norm": 0.17644047737121582,
+      "learning_rate": 8.30782898840972e-06,
+      "loss": 0.4665,
+      "step": 7498
+    },
+    {
+      "epoch": 1.6727637742583092,
+      "grad_norm": 0.17768198251724243,
+      "learning_rate": 8.305509545353144e-06,
+      "loss": 0.4551,
+      "step": 7499
+    },
+    {
+      "epoch": 1.6729868391701985,
+      "grad_norm": 0.18405990302562714,
+      "learning_rate": 8.303190196148112e-06,
+      "loss": 0.4434,
+      "step": 7500
+    },
+    {
+      "epoch": 1.6732099040820878,
+      "grad_norm": 0.17325492203235626,
+      "learning_rate": 8.300870940923077e-06,
+      "loss": 0.4501,
+      "step": 7501
+    },
+    {
+      "epoch": 1.673432968993977,
+      "grad_norm": 0.18856076896190643,
+      "learning_rate": 8.2985517798065e-06,
+      "loss": 0.4431,
+      "step": 7502
+    },
+    {
+      "epoch": 1.6736560339058666,
+      "grad_norm": 0.1935485303401947,
+      "learning_rate": 8.296232712926826e-06,
+      "loss": 0.4648,
+      "step": 7503
+    },
+    {
+      "epoch": 1.673879098817756,
+      "grad_norm": 0.17944234609603882,
+      "learning_rate": 8.293913740412503e-06,
+      "loss": 0.4561,
+      "step": 7504
+    },
+    {
+      "epoch": 1.6741021637296454,
+      "grad_norm": 0.17433197796344757,
+      "learning_rate": 8.291594862391966e-06,
+      "loss": 0.4551,
+      "step": 7505
+    },
+    {
+      "epoch": 1.6743252286415347,
+      "grad_norm": 0.18297721445560455,
+      "learning_rate": 8.289276078993655e-06,
+      "loss": 0.4609,
+      "step": 7506
+    },
+    {
+      "epoch": 1.674548293553424,
+      "grad_norm": 0.17488986253738403,
+      "learning_rate": 8.286957390345994e-06,
+      "loss": 0.4294,
+      "step": 7507
+    },
+    {
+      "epoch": 1.6747713584653134,
+      "grad_norm": 0.1750224232673645,
+      "learning_rate": 8.28463879657741e-06,
+      "loss": 0.4609,
+      "step": 7508
+    },
+    {
+      "epoch": 1.6749944233772027,
+      "grad_norm": 0.18751104176044464,
+      "learning_rate": 8.282320297816315e-06,
+      "loss": 0.4519,
+      "step": 7509
+    },
+    {
+      "epoch": 1.6752174882890922,
+      "grad_norm": 0.17194952070713043,
+      "learning_rate": 8.280001894191132e-06,
+      "loss": 0.4437,
+      "step": 7510
+    },
+    {
+      "epoch": 1.6754405532009815,
+      "grad_norm": 0.17177167534828186,
+      "learning_rate": 8.277683585830259e-06,
+      "loss": 0.4455,
+      "step": 7511
+    },
+    {
+      "epoch": 1.6756636181128708,
+      "grad_norm": 0.17539672553539276,
+      "learning_rate": 8.275365372862106e-06,
+      "loss": 0.4598,
+      "step": 7512
+    },
+    {
+      "epoch": 1.67588668302476,
+      "grad_norm": 0.17903223633766174,
+      "learning_rate": 8.273047255415066e-06,
+      "loss": 0.4473,
+      "step": 7513
+    },
+    {
+      "epoch": 1.6761097479366496,
+      "grad_norm": 0.17606431245803833,
+      "learning_rate": 8.27072923361753e-06,
+      "loss": 0.4592,
+      "step": 7514
+    },
+    {
+      "epoch": 1.676332812848539,
+      "grad_norm": 0.1900903880596161,
+      "learning_rate": 8.26841130759789e-06,
+      "loss": 0.4735,
+      "step": 7515
+    },
+    {
+      "epoch": 1.6765558777604284,
+      "grad_norm": 0.18751583993434906,
+      "learning_rate": 8.26609347748452e-06,
+      "loss": 0.4795,
+      "step": 7516
+    },
+    {
+      "epoch": 1.6767789426723176,
+      "grad_norm": 0.18636786937713623,
+      "learning_rate": 8.263775743405804e-06,
+      "loss": 0.4474,
+      "step": 7517
+    },
+    {
+      "epoch": 1.677002007584207,
+      "grad_norm": 0.1759779453277588,
+      "learning_rate": 8.261458105490104e-06,
+      "loss": 0.4613,
+      "step": 7518
+    },
+    {
+      "epoch": 1.6772250724960962,
+      "grad_norm": 0.16955047845840454,
+      "learning_rate": 8.259140563865796e-06,
+      "loss": 0.4526,
+      "step": 7519
+    },
+    {
+      "epoch": 1.6774481374079857,
+      "grad_norm": 0.18776053190231323,
+      "learning_rate": 8.256823118661228e-06,
+      "loss": 0.4265,
+      "step": 7520
+    },
+    {
+      "epoch": 1.6776712023198752,
+      "grad_norm": 0.17667481303215027,
+      "learning_rate": 8.254505770004764e-06,
+      "loss": 0.4782,
+      "step": 7521
+    },
+    {
+      "epoch": 1.6778942672317645,
+      "grad_norm": 0.17100000381469727,
+      "learning_rate": 8.252188518024748e-06,
+      "loss": 0.464,
+      "step": 7522
+    },
+    {
+      "epoch": 1.6781173321436538,
+      "grad_norm": 0.16992557048797607,
+      "learning_rate": 8.24987136284953e-06,
+      "loss": 0.4451,
+      "step": 7523
+    },
+    {
+      "epoch": 1.678340397055543,
+      "grad_norm": 0.18691998720169067,
+      "learning_rate": 8.247554304607442e-06,
+      "loss": 0.4701,
+      "step": 7524
+    },
+    {
+      "epoch": 1.6785634619674326,
+      "grad_norm": 0.16606706380844116,
+      "learning_rate": 8.24523734342682e-06,
+      "loss": 0.4341,
+      "step": 7525
+    },
+    {
+      "epoch": 1.6787865268793218,
+      "grad_norm": 0.16912147402763367,
+      "learning_rate": 8.242920479435991e-06,
+      "loss": 0.4381,
+      "step": 7526
+    },
+    {
+      "epoch": 1.6790095917912113,
+      "grad_norm": 0.1704743504524231,
+      "learning_rate": 8.24060371276328e-06,
+      "loss": 0.4458,
+      "step": 7527
+    },
+    {
+      "epoch": 1.6792326567031006,
+      "grad_norm": 0.18589746952056885,
+      "learning_rate": 8.238287043537e-06,
+      "loss": 0.4349,
+      "step": 7528
+    },
+    {
+      "epoch": 1.67945572161499,
+      "grad_norm": 0.17729128897190094,
+      "learning_rate": 8.235970471885468e-06,
+      "loss": 0.446,
+      "step": 7529
+    },
+    {
+      "epoch": 1.6796787865268792,
+      "grad_norm": 0.17116542160511017,
+      "learning_rate": 8.233653997936985e-06,
+      "loss": 0.42,
+      "step": 7530
+    },
+    {
+      "epoch": 1.6799018514387687,
+      "grad_norm": 0.1848667562007904,
+      "learning_rate": 8.231337621819858e-06,
+      "loss": 0.4622,
+      "step": 7531
+    },
+    {
+      "epoch": 1.6801249163506582,
+      "grad_norm": 0.16781508922576904,
+      "learning_rate": 8.229021343662376e-06,
+      "loss": 0.4326,
+      "step": 7532
+    },
+    {
+      "epoch": 1.6803479812625475,
+      "grad_norm": 0.1825597584247589,
+      "learning_rate": 8.22670516359283e-06,
+      "loss": 0.4471,
+      "step": 7533
+    },
+    {
+      "epoch": 1.6805710461744368,
+      "grad_norm": 0.17909115552902222,
+      "learning_rate": 8.22438908173951e-06,
+      "loss": 0.464,
+      "step": 7534
+    },
+    {
+      "epoch": 1.680794111086326,
+      "grad_norm": 0.16868308186531067,
+      "learning_rate": 8.22207309823069e-06,
+      "loss": 0.4446,
+      "step": 7535
+    },
+    {
+      "epoch": 1.6810171759982153,
+      "grad_norm": 0.17775742709636688,
+      "learning_rate": 8.219757213194647e-06,
+      "loss": 0.4638,
+      "step": 7536
+    },
+    {
+      "epoch": 1.6812402409101048,
+      "grad_norm": 0.1721707284450531,
+      "learning_rate": 8.217441426759645e-06,
+      "loss": 0.4598,
+      "step": 7537
+    },
+    {
+      "epoch": 1.6814633058219943,
+      "grad_norm": 0.17508956789970398,
+      "learning_rate": 8.215125739053953e-06,
+      "loss": 0.4804,
+      "step": 7538
+    },
+    {
+      "epoch": 1.6816863707338836,
+      "grad_norm": 0.18540501594543457,
+      "learning_rate": 8.21281015020582e-06,
+      "loss": 0.457,
+      "step": 7539
+    },
+    {
+      "epoch": 1.6819094356457729,
+      "grad_norm": 0.19119501113891602,
+      "learning_rate": 8.210494660343508e-06,
+      "loss": 0.4486,
+      "step": 7540
+    },
+    {
+      "epoch": 1.6821325005576622,
+      "grad_norm": 0.18080313503742218,
+      "learning_rate": 8.208179269595255e-06,
+      "loss": 0.4727,
+      "step": 7541
+    },
+    {
+      "epoch": 1.6823555654695517,
+      "grad_norm": 0.192534402012825,
+      "learning_rate": 8.205863978089308e-06,
+      "loss": 0.4712,
+      "step": 7542
+    },
+    {
+      "epoch": 1.682578630381441,
+      "grad_norm": 0.16961976885795593,
+      "learning_rate": 8.203548785953896e-06,
+      "loss": 0.4169,
+      "step": 7543
+    },
+    {
+      "epoch": 1.6828016952933305,
+      "grad_norm": 0.17554304003715515,
+      "learning_rate": 8.201233693317254e-06,
+      "loss": 0.4448,
+      "step": 7544
+    },
+    {
+      "epoch": 1.6830247602052197,
+      "grad_norm": 0.1828324943780899,
+      "learning_rate": 8.198918700307604e-06,
+      "loss": 0.4758,
+      "step": 7545
+    },
+    {
+      "epoch": 1.683247825117109,
+      "grad_norm": 0.1798938661813736,
+      "learning_rate": 8.196603807053167e-06,
+      "loss": 0.4567,
+      "step": 7546
+    },
+    {
+      "epoch": 1.6834708900289983,
+      "grad_norm": 0.178616002202034,
+      "learning_rate": 8.194289013682154e-06,
+      "loss": 0.4547,
+      "step": 7547
+    },
+    {
+      "epoch": 1.6836939549408878,
+      "grad_norm": 0.20149093866348267,
+      "learning_rate": 8.191974320322776e-06,
+      "loss": 0.4427,
+      "step": 7548
+    },
+    {
+      "epoch": 1.6839170198527773,
+      "grad_norm": 0.17913025617599487,
+      "learning_rate": 8.189659727103233e-06,
+      "loss": 0.4371,
+      "step": 7549
+    },
+    {
+      "epoch": 1.6841400847646666,
+      "grad_norm": 0.18823112547397614,
+      "learning_rate": 8.18734523415172e-06,
+      "loss": 0.4548,
+      "step": 7550
+    },
+    {
+      "epoch": 1.6843631496765559,
+      "grad_norm": 0.17050491273403168,
+      "learning_rate": 8.185030841596431e-06,
+      "loss": 0.4481,
+      "step": 7551
+    },
+    {
+      "epoch": 1.6845862145884452,
+      "grad_norm": 0.1750423163175583,
+      "learning_rate": 8.182716549565548e-06,
+      "loss": 0.4527,
+      "step": 7552
+    },
+    {
+      "epoch": 1.6848092795003344,
+      "grad_norm": 0.17691361904144287,
+      "learning_rate": 8.180402358187256e-06,
+      "loss": 0.4564,
+      "step": 7553
+    },
+    {
+      "epoch": 1.685032344412224,
+      "grad_norm": 0.17147858440876007,
+      "learning_rate": 8.178088267589725e-06,
+      "loss": 0.4572,
+      "step": 7554
+    },
+    {
+      "epoch": 1.6852554093241134,
+      "grad_norm": 0.18109926581382751,
+      "learning_rate": 8.175774277901128e-06,
+      "loss": 0.4387,
+      "step": 7555
+    },
+    {
+      "epoch": 1.6854784742360027,
+      "grad_norm": 0.17862330377101898,
+      "learning_rate": 8.173460389249625e-06,
+      "loss": 0.4453,
+      "step": 7556
+    },
+    {
+      "epoch": 1.685701539147892,
+      "grad_norm": 0.18046163022518158,
+      "learning_rate": 8.171146601763374e-06,
+      "loss": 0.4471,
+      "step": 7557
+    },
+    {
+      "epoch": 1.6859246040597813,
+      "grad_norm": 0.172270730137825,
+      "learning_rate": 8.168832915570531e-06,
+      "loss": 0.4167,
+      "step": 7558
+    },
+    {
+      "epoch": 1.6861476689716708,
+      "grad_norm": 0.17891794443130493,
+      "learning_rate": 8.166519330799237e-06,
+      "loss": 0.4455,
+      "step": 7559
+    },
+    {
+      "epoch": 1.68637073388356,
+      "grad_norm": 0.1716044843196869,
+      "learning_rate": 8.16420584757764e-06,
+      "loss": 0.4451,
+      "step": 7560
+    },
+    {
+      "epoch": 1.6865937987954496,
+      "grad_norm": 0.17391957342624664,
+      "learning_rate": 8.161892466033865e-06,
+      "loss": 0.4543,
+      "step": 7561
+    },
+    {
+      "epoch": 1.6868168637073389,
+      "grad_norm": 0.17183732986450195,
+      "learning_rate": 8.159579186296052e-06,
+      "loss": 0.4653,
+      "step": 7562
+    },
+    {
+      "epoch": 1.6870399286192281,
+      "grad_norm": 0.19677892327308655,
+      "learning_rate": 8.157266008492318e-06,
+      "loss": 0.4543,
+      "step": 7563
+    },
+    {
+      "epoch": 1.6872629935311174,
+      "grad_norm": 0.18332447111606598,
+      "learning_rate": 8.154952932750784e-06,
+      "loss": 0.4407,
+      "step": 7564
+    },
+    {
+      "epoch": 1.687486058443007,
+      "grad_norm": 0.1903562992811203,
+      "learning_rate": 8.152639959199561e-06,
+      "loss": 0.4464,
+      "step": 7565
+    },
+    {
+      "epoch": 1.6877091233548964,
+      "grad_norm": 0.22077789902687073,
+      "learning_rate": 8.150327087966761e-06,
+      "loss": 0.4941,
+      "step": 7566
+    },
+    {
+      "epoch": 1.6879321882667857,
+      "grad_norm": 0.172964408993721,
+      "learning_rate": 8.148014319180479e-06,
+      "loss": 0.4584,
+      "step": 7567
+    },
+    {
+      "epoch": 1.688155253178675,
+      "grad_norm": 0.16977094113826752,
+      "learning_rate": 8.145701652968814e-06,
+      "loss": 0.4347,
+      "step": 7568
+    },
+    {
+      "epoch": 1.6883783180905643,
+      "grad_norm": 0.1772548258304596,
+      "learning_rate": 8.143389089459855e-06,
+      "loss": 0.4701,
+      "step": 7569
+    },
+    {
+      "epoch": 1.6886013830024535,
+      "grad_norm": 0.16908633708953857,
+      "learning_rate": 8.14107662878169e-06,
+      "loss": 0.4595,
+      "step": 7570
+    },
+    {
+      "epoch": 1.688824447914343,
+      "grad_norm": 0.18083757162094116,
+      "learning_rate": 8.138764271062389e-06,
+      "loss": 0.4919,
+      "step": 7571
+    },
+    {
+      "epoch": 1.6890475128262326,
+      "grad_norm": 0.18248751759529114,
+      "learning_rate": 8.136452016430035e-06,
+      "loss": 0.4767,
+      "step": 7572
+    },
+    {
+      "epoch": 1.6892705777381218,
+      "grad_norm": 0.1837497055530548,
+      "learning_rate": 8.134139865012688e-06,
+      "loss": 0.4462,
+      "step": 7573
+    },
+    {
+      "epoch": 1.6894936426500111,
+      "grad_norm": 0.1747952401638031,
+      "learning_rate": 8.131827816938412e-06,
+      "loss": 0.4586,
+      "step": 7574
+    },
+    {
+      "epoch": 1.6897167075619004,
+      "grad_norm": 0.16300569474697113,
+      "learning_rate": 8.129515872335263e-06,
+      "loss": 0.4319,
+      "step": 7575
+    },
+    {
+      "epoch": 1.68993977247379,
+      "grad_norm": 0.17720209062099457,
+      "learning_rate": 8.127204031331293e-06,
+      "loss": 0.4447,
+      "step": 7576
+    },
+    {
+      "epoch": 1.6901628373856792,
+      "grad_norm": 0.1850651055574417,
+      "learning_rate": 8.12489229405454e-06,
+      "loss": 0.4473,
+      "step": 7577
+    },
+    {
+      "epoch": 1.6903859022975687,
+      "grad_norm": 0.17101094126701355,
+      "learning_rate": 8.122580660633048e-06,
+      "loss": 0.4259,
+      "step": 7578
+    },
+    {
+      "epoch": 1.690608967209458,
+      "grad_norm": 0.17117485404014587,
+      "learning_rate": 8.12026913119485e-06,
+      "loss": 0.4467,
+      "step": 7579
+    },
+    {
+      "epoch": 1.6908320321213473,
+      "grad_norm": 0.18159452080726624,
+      "learning_rate": 8.117957705867971e-06,
+      "loss": 0.4606,
+      "step": 7580
+    },
+    {
+      "epoch": 1.6910550970332365,
+      "grad_norm": 0.17214803397655487,
+      "learning_rate": 8.115646384780434e-06,
+      "loss": 0.4454,
+      "step": 7581
+    },
+    {
+      "epoch": 1.691278161945126,
+      "grad_norm": 0.17703603208065033,
+      "learning_rate": 8.11333516806025e-06,
+      "loss": 0.4455,
+      "step": 7582
+    },
+    {
+      "epoch": 1.6915012268570155,
+      "grad_norm": 0.18035492300987244,
+      "learning_rate": 8.111024055835436e-06,
+      "loss": 0.4668,
+      "step": 7583
+    },
+    {
+      "epoch": 1.6917242917689048,
+      "grad_norm": 0.17585453391075134,
+      "learning_rate": 8.108713048233988e-06,
+      "loss": 0.4267,
+      "step": 7584
+    },
+    {
+      "epoch": 1.691947356680794,
+      "grad_norm": 0.184658944606781,
+      "learning_rate": 8.106402145383911e-06,
+      "loss": 0.4665,
+      "step": 7585
+    },
+    {
+      "epoch": 1.6921704215926834,
+      "grad_norm": 0.19270674884319305,
+      "learning_rate": 8.104091347413192e-06,
+      "loss": 0.4386,
+      "step": 7586
+    },
+    {
+      "epoch": 1.6923934865045729,
+      "grad_norm": 0.1800229251384735,
+      "learning_rate": 8.101780654449822e-06,
+      "loss": 0.4618,
+      "step": 7587
+    },
+    {
+      "epoch": 1.6926165514164622,
+      "grad_norm": 0.18103908002376556,
+      "learning_rate": 8.099470066621778e-06,
+      "loss": 0.4513,
+      "step": 7588
+    },
+    {
+      "epoch": 1.6928396163283517,
+      "grad_norm": 0.16606682538986206,
+      "learning_rate": 8.09715958405704e-06,
+      "loss": 0.4251,
+      "step": 7589
+    },
+    {
+      "epoch": 1.693062681240241,
+      "grad_norm": 0.17416086792945862,
+      "learning_rate": 8.094849206883569e-06,
+      "loss": 0.4473,
+      "step": 7590
+    },
+    {
+      "epoch": 1.6932857461521302,
+      "grad_norm": 0.21117763221263885,
+      "learning_rate": 8.092538935229336e-06,
+      "loss": 0.4578,
+      "step": 7591
+    },
+    {
+      "epoch": 1.6935088110640195,
+      "grad_norm": 0.17628377676010132,
+      "learning_rate": 8.090228769222292e-06,
+      "loss": 0.4716,
+      "step": 7592
+    },
+    {
+      "epoch": 1.693731875975909,
+      "grad_norm": 0.1700354665517807,
+      "learning_rate": 8.087918708990396e-06,
+      "loss": 0.4444,
+      "step": 7593
+    },
+    {
+      "epoch": 1.6939549408877983,
+      "grad_norm": 0.18636515736579895,
+      "learning_rate": 8.085608754661585e-06,
+      "loss": 0.4843,
+      "step": 7594
+    },
+    {
+      "epoch": 1.6941780057996878,
+      "grad_norm": 0.1767621785402298,
+      "learning_rate": 8.083298906363806e-06,
+      "loss": 0.4422,
+      "step": 7595
+    },
+    {
+      "epoch": 1.694401070711577,
+      "grad_norm": 0.1970623880624771,
+      "learning_rate": 8.080989164224988e-06,
+      "loss": 0.4432,
+      "step": 7596
+    },
+    {
+      "epoch": 1.6946241356234664,
+      "grad_norm": 0.1670975536108017,
+      "learning_rate": 8.078679528373063e-06,
+      "loss": 0.4353,
+      "step": 7597
+    },
+    {
+      "epoch": 1.6948472005353556,
+      "grad_norm": 0.18546359241008759,
+      "learning_rate": 8.076369998935951e-06,
+      "loss": 0.4501,
+      "step": 7598
+    },
+    {
+      "epoch": 1.6950702654472452,
+      "grad_norm": 0.17603303492069244,
+      "learning_rate": 8.074060576041566e-06,
+      "loss": 0.4451,
+      "step": 7599
+    },
+    {
+      "epoch": 1.6952933303591347,
+      "grad_norm": 0.17194673418998718,
+      "learning_rate": 8.071751259817825e-06,
+      "loss": 0.431,
+      "step": 7600
+    },
+    {
+      "epoch": 1.695516395271024,
+      "grad_norm": 0.1741739809513092,
+      "learning_rate": 8.069442050392625e-06,
+      "loss": 0.4715,
+      "step": 7601
+    },
+    {
+      "epoch": 1.6957394601829132,
+      "grad_norm": 0.24975387752056122,
+      "learning_rate": 8.067132947893872e-06,
+      "loss": 0.4369,
+      "step": 7602
+    },
+    {
+      "epoch": 1.6959625250948025,
+      "grad_norm": 0.18625226616859436,
+      "learning_rate": 8.064823952449449e-06,
+      "loss": 0.4464,
+      "step": 7603
+    },
+    {
+      "epoch": 1.696185590006692,
+      "grad_norm": 0.1777094006538391,
+      "learning_rate": 8.062515064187253e-06,
+      "loss": 0.4496,
+      "step": 7604
+    },
+    {
+      "epoch": 1.6964086549185813,
+      "grad_norm": 0.1841573417186737,
+      "learning_rate": 8.060206283235159e-06,
+      "loss": 0.4731,
+      "step": 7605
+    },
+    {
+      "epoch": 1.6966317198304708,
+      "grad_norm": 0.17774668335914612,
+      "learning_rate": 8.05789760972104e-06,
+      "loss": 0.4905,
+      "step": 7606
+    },
+    {
+      "epoch": 1.69685478474236,
+      "grad_norm": 0.16449688374996185,
+      "learning_rate": 8.055589043772772e-06,
+      "loss": 0.4542,
+      "step": 7607
+    },
+    {
+      "epoch": 1.6970778496542493,
+      "grad_norm": 0.1715865582227707,
+      "learning_rate": 8.053280585518211e-06,
+      "loss": 0.4448,
+      "step": 7608
+    },
+    {
+      "epoch": 1.6973009145661386,
+      "grad_norm": 0.18522848188877106,
+      "learning_rate": 8.050972235085217e-06,
+      "loss": 0.4821,
+      "step": 7609
+    },
+    {
+      "epoch": 1.6975239794780281,
+      "grad_norm": 0.17819975316524506,
+      "learning_rate": 8.048663992601641e-06,
+      "loss": 0.4306,
+      "step": 7610
+    },
+    {
+      "epoch": 1.6977470443899174,
+      "grad_norm": 0.1718638390302658,
+      "learning_rate": 8.046355858195325e-06,
+      "loss": 0.4349,
+      "step": 7611
+    },
+    {
+      "epoch": 1.697970109301807,
+      "grad_norm": 0.16903117299079895,
+      "learning_rate": 8.044047831994114e-06,
+      "loss": 0.4348,
+      "step": 7612
+    },
+    {
+      "epoch": 1.6981931742136962,
+      "grad_norm": 0.1886761337518692,
+      "learning_rate": 8.041739914125835e-06,
+      "loss": 0.4732,
+      "step": 7613
+    },
+    {
+      "epoch": 1.6984162391255855,
+      "grad_norm": 0.20157945156097412,
+      "learning_rate": 8.03943210471832e-06,
+      "loss": 0.4639,
+      "step": 7614
+    },
+    {
+      "epoch": 1.6986393040374748,
+      "grad_norm": 0.1816253513097763,
+      "learning_rate": 8.037124403899384e-06,
+      "loss": 0.4429,
+      "step": 7615
+    },
+    {
+      "epoch": 1.6988623689493643,
+      "grad_norm": 0.17475511133670807,
+      "learning_rate": 8.03481681179685e-06,
+      "loss": 0.4565,
+      "step": 7616
+    },
+    {
+      "epoch": 1.6990854338612538,
+      "grad_norm": 0.17711593210697174,
+      "learning_rate": 8.032509328538518e-06,
+      "loss": 0.4488,
+      "step": 7617
+    },
+    {
+      "epoch": 1.699308498773143,
+      "grad_norm": 0.18735013902187347,
+      "learning_rate": 8.030201954252198e-06,
+      "loss": 0.4554,
+      "step": 7618
+    },
+    {
+      "epoch": 1.6995315636850323,
+      "grad_norm": 0.5531375408172607,
+      "learning_rate": 8.027894689065684e-06,
+      "loss": 0.4486,
+      "step": 7619
+    },
+    {
+      "epoch": 1.6997546285969216,
+      "grad_norm": 0.16994529962539673,
+      "learning_rate": 8.025587533106765e-06,
+      "loss": 0.4126,
+      "step": 7620
+    },
+    {
+      "epoch": 1.6999776935088111,
+      "grad_norm": 0.18847712874412537,
+      "learning_rate": 8.02328048650323e-06,
+      "loss": 0.4653,
+      "step": 7621
+    },
+    {
+      "epoch": 1.7002007584207004,
+      "grad_norm": 0.17005674540996552,
+      "learning_rate": 8.020973549382855e-06,
+      "loss": 0.4099,
+      "step": 7622
+    },
+    {
+      "epoch": 1.70042382333259,
+      "grad_norm": 0.18046452105045319,
+      "learning_rate": 8.018666721873414e-06,
+      "loss": 0.423,
+      "step": 7623
+    },
+    {
+      "epoch": 1.7006468882444792,
+      "grad_norm": 0.17824363708496094,
+      "learning_rate": 8.016360004102672e-06,
+      "loss": 0.4659,
+      "step": 7624
+    },
+    {
+      "epoch": 1.7008699531563685,
+      "grad_norm": 0.16878539323806763,
+      "learning_rate": 8.014053396198392e-06,
+      "loss": 0.4464,
+      "step": 7625
+    },
+    {
+      "epoch": 1.7010930180682577,
+      "grad_norm": 0.1796082705259323,
+      "learning_rate": 8.011746898288326e-06,
+      "loss": 0.4493,
+      "step": 7626
+    },
+    {
+      "epoch": 1.7013160829801472,
+      "grad_norm": 0.1814991682767868,
+      "learning_rate": 8.009440510500224e-06,
+      "loss": 0.4588,
+      "step": 7627
+    },
+    {
+      "epoch": 1.7015391478920365,
+      "grad_norm": 0.1700638234615326,
+      "learning_rate": 8.007134232961828e-06,
+      "loss": 0.4207,
+      "step": 7628
+    },
+    {
+      "epoch": 1.701762212803926,
+      "grad_norm": 0.1950269490480423,
+      "learning_rate": 8.004828065800874e-06,
+      "loss": 0.4338,
+      "step": 7629
+    },
+    {
+      "epoch": 1.7019852777158153,
+      "grad_norm": 0.17464493215084076,
+      "learning_rate": 8.00252200914509e-06,
+      "loss": 0.4465,
+      "step": 7630
+    },
+    {
+      "epoch": 1.7022083426277046,
+      "grad_norm": 0.19278262555599213,
+      "learning_rate": 8.000216063122205e-06,
+      "loss": 0.4549,
+      "step": 7631
+    },
+    {
+      "epoch": 1.7024314075395939,
+      "grad_norm": 0.1833004504442215,
+      "learning_rate": 7.997910227859932e-06,
+      "loss": 0.4433,
+      "step": 7632
+    },
+    {
+      "epoch": 1.7026544724514834,
+      "grad_norm": 0.17740319669246674,
+      "learning_rate": 7.995604503485984e-06,
+      "loss": 0.4509,
+      "step": 7633
+    },
+    {
+      "epoch": 1.7028775373633729,
+      "grad_norm": 0.17956921458244324,
+      "learning_rate": 7.993298890128067e-06,
+      "loss": 0.4465,
+      "step": 7634
+    },
+    {
+      "epoch": 1.7031006022752622,
+      "grad_norm": 0.18017272651195526,
+      "learning_rate": 7.99099338791388e-06,
+      "loss": 0.4817,
+      "step": 7635
+    },
+    {
+      "epoch": 1.7033236671871514,
+      "grad_norm": 0.17207284271717072,
+      "learning_rate": 7.988687996971116e-06,
+      "loss": 0.4545,
+      "step": 7636
+    },
+    {
+      "epoch": 1.7035467320990407,
+      "grad_norm": 0.19489426910877228,
+      "learning_rate": 7.986382717427461e-06,
+      "loss": 0.4867,
+      "step": 7637
+    },
+    {
+      "epoch": 1.7037697970109302,
+      "grad_norm": 0.1766097992658615,
+      "learning_rate": 7.984077549410598e-06,
+      "loss": 0.4792,
+      "step": 7638
+    },
+    {
+      "epoch": 1.7039928619228195,
+      "grad_norm": 0.17337632179260254,
+      "learning_rate": 7.981772493048203e-06,
+      "loss": 0.4619,
+      "step": 7639
+    },
+    {
+      "epoch": 1.704215926834709,
+      "grad_norm": 0.1748560070991516,
+      "learning_rate": 7.97946754846794e-06,
+      "loss": 0.449,
+      "step": 7640
+    },
+    {
+      "epoch": 1.7044389917465983,
+      "grad_norm": 0.18279384076595306,
+      "learning_rate": 7.977162715797476e-06,
+      "loss": 0.4498,
+      "step": 7641
+    },
+    {
+      "epoch": 1.7046620566584876,
+      "grad_norm": 0.1735672801733017,
+      "learning_rate": 7.97485799516446e-06,
+      "loss": 0.4588,
+      "step": 7642
+    },
+    {
+      "epoch": 1.7048851215703769,
+      "grad_norm": 0.17070254683494568,
+      "learning_rate": 7.972553386696553e-06,
+      "loss": 0.466,
+      "step": 7643
+    },
+    {
+      "epoch": 1.7051081864822664,
+      "grad_norm": 0.19621379673480988,
+      "learning_rate": 7.970248890521389e-06,
+      "loss": 0.4496,
+      "step": 7644
+    },
+    {
+      "epoch": 1.7053312513941556,
+      "grad_norm": 0.18287599086761475,
+      "learning_rate": 7.967944506766611e-06,
+      "loss": 0.449,
+      "step": 7645
+    },
+    {
+      "epoch": 1.7055543163060451,
+      "grad_norm": 0.1747177392244339,
+      "learning_rate": 7.965640235559847e-06,
+      "loss": 0.4491,
+      "step": 7646
+    },
+    {
+      "epoch": 1.7057773812179344,
+      "grad_norm": 0.1790647655725479,
+      "learning_rate": 7.963336077028725e-06,
+      "loss": 0.4558,
+      "step": 7647
+    },
+    {
+      "epoch": 1.7060004461298237,
+      "grad_norm": 0.1670437902212143,
+      "learning_rate": 7.96103203130086e-06,
+      "loss": 0.4331,
+      "step": 7648
+    },
+    {
+      "epoch": 1.706223511041713,
+      "grad_norm": 0.18212758004665375,
+      "learning_rate": 7.95872809850387e-06,
+      "loss": 0.4953,
+      "step": 7649
+    },
+    {
+      "epoch": 1.7064465759536025,
+      "grad_norm": 0.17667363584041595,
+      "learning_rate": 7.956424278765354e-06,
+      "loss": 0.4552,
+      "step": 7650
+    },
+    {
+      "epoch": 1.706669640865492,
+      "grad_norm": 0.18635283410549164,
+      "learning_rate": 7.95412057221292e-06,
+      "loss": 0.4677,
+      "step": 7651
+    },
+    {
+      "epoch": 1.7068927057773813,
+      "grad_norm": 0.1755317598581314,
+      "learning_rate": 7.951816978974154e-06,
+      "loss": 0.4372,
+      "step": 7652
+    },
+    {
+      "epoch": 1.7071157706892706,
+      "grad_norm": 0.1793239861726761,
+      "learning_rate": 7.949513499176651e-06,
+      "loss": 0.4509,
+      "step": 7653
+    },
+    {
+      "epoch": 1.7073388356011598,
+      "grad_norm": 0.17480173707008362,
+      "learning_rate": 7.947210132947984e-06,
+      "loss": 0.4487,
+      "step": 7654
+    },
+    {
+      "epoch": 1.7075619005130493,
+      "grad_norm": 0.19050490856170654,
+      "learning_rate": 7.944906880415738e-06,
+      "loss": 0.4449,
+      "step": 7655
+    },
+    {
+      "epoch": 1.7077849654249386,
+      "grad_norm": 0.17552480101585388,
+      "learning_rate": 7.94260374170747e-06,
+      "loss": 0.4457,
+      "step": 7656
+    },
+    {
+      "epoch": 1.7080080303368281,
+      "grad_norm": 0.17692866921424866,
+      "learning_rate": 7.940300716950753e-06,
+      "loss": 0.4179,
+      "step": 7657
+    },
+    {
+      "epoch": 1.7082310952487174,
+      "grad_norm": 0.18498113751411438,
+      "learning_rate": 7.937997806273135e-06,
+      "loss": 0.4221,
+      "step": 7658
+    },
+    {
+      "epoch": 1.7084541601606067,
+      "grad_norm": 0.18404224514961243,
+      "learning_rate": 7.935695009802172e-06,
+      "loss": 0.4682,
+      "step": 7659
+    },
+    {
+      "epoch": 1.708677225072496,
+      "grad_norm": 0.18323245644569397,
+      "learning_rate": 7.933392327665403e-06,
+      "loss": 0.4541,
+      "step": 7660
+    },
+    {
+      "epoch": 1.7089002899843855,
+      "grad_norm": 0.17359468340873718,
+      "learning_rate": 7.931089759990367e-06,
+      "loss": 0.4504,
+      "step": 7661
+    },
+    {
+      "epoch": 1.709123354896275,
+      "grad_norm": 0.1712869256734848,
+      "learning_rate": 7.928787306904593e-06,
+      "loss": 0.45,
+      "step": 7662
+    },
+    {
+      "epoch": 1.7093464198081643,
+      "grad_norm": 0.17846348881721497,
+      "learning_rate": 7.926484968535604e-06,
+      "loss": 0.4644,
+      "step": 7663
+    },
+    {
+      "epoch": 1.7095694847200535,
+      "grad_norm": 0.18150021135807037,
+      "learning_rate": 7.924182745010926e-06,
+      "loss": 0.4206,
+      "step": 7664
+    },
+    {
+      "epoch": 1.7097925496319428,
+      "grad_norm": 0.17358383536338806,
+      "learning_rate": 7.921880636458061e-06,
+      "loss": 0.4492,
+      "step": 7665
+    },
+    {
+      "epoch": 1.710015614543832,
+      "grad_norm": 0.18115226924419403,
+      "learning_rate": 7.919578643004519e-06,
+      "loss": 0.4601,
+      "step": 7666
+    },
+    {
+      "epoch": 1.7102386794557216,
+      "grad_norm": 0.17906229197978973,
+      "learning_rate": 7.917276764777799e-06,
+      "loss": 0.4428,
+      "step": 7667
+    },
+    {
+      "epoch": 1.7104617443676111,
+      "grad_norm": 0.1732749044895172,
+      "learning_rate": 7.914975001905393e-06,
+      "loss": 0.4274,
+      "step": 7668
+    },
+    {
+      "epoch": 1.7106848092795004,
+      "grad_norm": 0.1772816926240921,
+      "learning_rate": 7.912673354514784e-06,
+      "loss": 0.4574,
+      "step": 7669
+    },
+    {
+      "epoch": 1.7109078741913897,
+      "grad_norm": 0.18031185865402222,
+      "learning_rate": 7.910371822733458e-06,
+      "loss": 0.4498,
+      "step": 7670
+    },
+    {
+      "epoch": 1.711130939103279,
+      "grad_norm": 0.18245866894721985,
+      "learning_rate": 7.908070406688881e-06,
+      "loss": 0.4271,
+      "step": 7671
+    },
+    {
+      "epoch": 1.7113540040151685,
+      "grad_norm": 0.17715051770210266,
+      "learning_rate": 7.905769106508527e-06,
+      "loss": 0.4714,
+      "step": 7672
+    },
+    {
+      "epoch": 1.7115770689270577,
+      "grad_norm": 0.18403951823711395,
+      "learning_rate": 7.90346792231985e-06,
+      "loss": 0.4349,
+      "step": 7673
+    },
+    {
+      "epoch": 1.7118001338389472,
+      "grad_norm": 0.1747589409351349,
+      "learning_rate": 7.90116685425031e-06,
+      "loss": 0.4318,
+      "step": 7674
+    },
+    {
+      "epoch": 1.7120231987508365,
+      "grad_norm": 0.18269647657871246,
+      "learning_rate": 7.898865902427351e-06,
+      "loss": 0.4687,
+      "step": 7675
+    },
+    {
+      "epoch": 1.7122462636627258,
+      "grad_norm": 0.18082202970981598,
+      "learning_rate": 7.896565066978416e-06,
+      "loss": 0.4601,
+      "step": 7676
+    },
+    {
+      "epoch": 1.712469328574615,
+      "grad_norm": 0.16931264102458954,
+      "learning_rate": 7.894264348030935e-06,
+      "loss": 0.4478,
+      "step": 7677
+    },
+    {
+      "epoch": 1.7126923934865046,
+      "grad_norm": 0.1926901787519455,
+      "learning_rate": 7.891963745712344e-06,
+      "loss": 0.4334,
+      "step": 7678
+    },
+    {
+      "epoch": 1.712915458398394,
+      "grad_norm": 0.18817317485809326,
+      "learning_rate": 7.88966326015006e-06,
+      "loss": 0.4705,
+      "step": 7679
+    },
+    {
+      "epoch": 1.7131385233102834,
+      "grad_norm": 0.18068066239356995,
+      "learning_rate": 7.887362891471499e-06,
+      "loss": 0.4796,
+      "step": 7680
+    },
+    {
+      "epoch": 1.7133615882221727,
+      "grad_norm": 0.17936988174915314,
+      "learning_rate": 7.885062639804068e-06,
+      "loss": 0.4743,
+      "step": 7681
+    },
+    {
+      "epoch": 1.713584653134062,
+      "grad_norm": 0.17455992102622986,
+      "learning_rate": 7.882762505275175e-06,
+      "loss": 0.4528,
+      "step": 7682
+    },
+    {
+      "epoch": 1.7138077180459512,
+      "grad_norm": 0.17338865995407104,
+      "learning_rate": 7.880462488012208e-06,
+      "loss": 0.4324,
+      "step": 7683
+    },
+    {
+      "epoch": 1.7140307829578407,
+      "grad_norm": 0.18230746686458588,
+      "learning_rate": 7.878162588142563e-06,
+      "loss": 0.4634,
+      "step": 7684
+    },
+    {
+      "epoch": 1.7142538478697302,
+      "grad_norm": 0.21422216296195984,
+      "learning_rate": 7.875862805793622e-06,
+      "loss": 0.4554,
+      "step": 7685
+    },
+    {
+      "epoch": 1.7144769127816195,
+      "grad_norm": 0.17937451601028442,
+      "learning_rate": 7.873563141092759e-06,
+      "loss": 0.4646,
+      "step": 7686
+    },
+    {
+      "epoch": 1.7146999776935088,
+      "grad_norm": 0.17571154236793518,
+      "learning_rate": 7.871263594167345e-06,
+      "loss": 0.4408,
+      "step": 7687
+    },
+    {
+      "epoch": 1.714923042605398,
+      "grad_norm": 0.18515697121620178,
+      "learning_rate": 7.868964165144743e-06,
+      "loss": 0.4778,
+      "step": 7688
+    },
+    {
+      "epoch": 1.7151461075172876,
+      "grad_norm": 0.17536167800426483,
+      "learning_rate": 7.866664854152312e-06,
+      "loss": 0.4284,
+      "step": 7689
+    },
+    {
+      "epoch": 1.7153691724291769,
+      "grad_norm": 0.17925989627838135,
+      "learning_rate": 7.864365661317399e-06,
+      "loss": 0.4425,
+      "step": 7690
+    },
+    {
+      "epoch": 1.7155922373410664,
+      "grad_norm": 0.17760370671749115,
+      "learning_rate": 7.86206658676735e-06,
+      "loss": 0.4521,
+      "step": 7691
+    },
+    {
+      "epoch": 1.7158153022529556,
+      "grad_norm": 0.1762177050113678,
+      "learning_rate": 7.8597676306295e-06,
+      "loss": 0.4415,
+      "step": 7692
+    },
+    {
+      "epoch": 1.716038367164845,
+      "grad_norm": 0.17917610704898834,
+      "learning_rate": 7.857468793031185e-06,
+      "loss": 0.4356,
+      "step": 7693
+    },
+    {
+      "epoch": 1.7162614320767342,
+      "grad_norm": 0.17529255151748657,
+      "learning_rate": 7.855170074099723e-06,
+      "loss": 0.4345,
+      "step": 7694
+    },
+    {
+      "epoch": 1.7164844969886237,
+      "grad_norm": 0.17721855640411377,
+      "learning_rate": 7.852871473962435e-06,
+      "loss": 0.4607,
+      "step": 7695
+    },
+    {
+      "epoch": 1.7167075619005132,
+      "grad_norm": 0.18101619184017181,
+      "learning_rate": 7.850572992746628e-06,
+      "loss": 0.4568,
+      "step": 7696
+    },
+    {
+      "epoch": 1.7169306268124025,
+      "grad_norm": 0.1923847496509552,
+      "learning_rate": 7.848274630579611e-06,
+      "loss": 0.4492,
+      "step": 7697
+    },
+    {
+      "epoch": 1.7171536917242918,
+      "grad_norm": 0.1815698742866516,
+      "learning_rate": 7.845976387588679e-06,
+      "loss": 0.4323,
+      "step": 7698
+    },
+    {
+      "epoch": 1.717376756636181,
+      "grad_norm": 0.1835760623216629,
+      "learning_rate": 7.843678263901125e-06,
+      "loss": 0.4363,
+      "step": 7699
+    },
+    {
+      "epoch": 1.7175998215480703,
+      "grad_norm": 0.1698540896177292,
+      "learning_rate": 7.841380259644231e-06,
+      "loss": 0.4399,
+      "step": 7700
+    },
+    {
+      "epoch": 1.7178228864599598,
+      "grad_norm": 0.1790151596069336,
+      "learning_rate": 7.83908237494528e-06,
+      "loss": 0.4504,
+      "step": 7701
+    },
+    {
+      "epoch": 1.7180459513718493,
+      "grad_norm": 0.18497934937477112,
+      "learning_rate": 7.836784609931536e-06,
+      "loss": 0.4655,
+      "step": 7702
+    },
+    {
+      "epoch": 1.7182690162837386,
+      "grad_norm": 0.19526800513267517,
+      "learning_rate": 7.83448696473027e-06,
+      "loss": 0.4419,
+      "step": 7703
+    },
+    {
+      "epoch": 1.718492081195628,
+      "grad_norm": 0.1786162108182907,
+      "learning_rate": 7.832189439468734e-06,
+      "loss": 0.4286,
+      "step": 7704
+    },
+    {
+      "epoch": 1.7187151461075172,
+      "grad_norm": 0.18116159737110138,
+      "learning_rate": 7.829892034274184e-06,
+      "loss": 0.4799,
+      "step": 7705
+    },
+    {
+      "epoch": 1.7189382110194067,
+      "grad_norm": 0.1766580492258072,
+      "learning_rate": 7.827594749273867e-06,
+      "loss": 0.4692,
+      "step": 7706
+    },
+    {
+      "epoch": 1.719161275931296,
+      "grad_norm": 0.1848170906305313,
+      "learning_rate": 7.825297584595014e-06,
+      "loss": 0.4547,
+      "step": 7707
+    },
+    {
+      "epoch": 1.7193843408431855,
+      "grad_norm": 0.17963701486587524,
+      "learning_rate": 7.823000540364865e-06,
+      "loss": 0.4674,
+      "step": 7708
+    },
+    {
+      "epoch": 1.7196074057550748,
+      "grad_norm": 0.183467298746109,
+      "learning_rate": 7.820703616710634e-06,
+      "loss": 0.4579,
+      "step": 7709
+    },
+    {
+      "epoch": 1.719830470666964,
+      "grad_norm": 0.17056439816951752,
+      "learning_rate": 7.81840681375955e-06,
+      "loss": 0.4638,
+      "step": 7710
+    },
+    {
+      "epoch": 1.7200535355788533,
+      "grad_norm": 0.18178224563598633,
+      "learning_rate": 7.816110131638815e-06,
+      "loss": 0.486,
+      "step": 7711
+    },
+    {
+      "epoch": 1.7202766004907428,
+      "grad_norm": 0.20546191930770874,
+      "learning_rate": 7.813813570475645e-06,
+      "loss": 0.4659,
+      "step": 7712
+    },
+    {
+      "epoch": 1.7204996654026323,
+      "grad_norm": 0.18711180984973907,
+      "learning_rate": 7.811517130397227e-06,
+      "loss": 0.4554,
+      "step": 7713
+    },
+    {
+      "epoch": 1.7207227303145216,
+      "grad_norm": 0.17499269545078278,
+      "learning_rate": 7.809220811530755e-06,
+      "loss": 0.434,
+      "step": 7714
+    },
+    {
+      "epoch": 1.720945795226411,
+      "grad_norm": 0.1984056979417801,
+      "learning_rate": 7.806924614003416e-06,
+      "loss": 0.4382,
+      "step": 7715
+    },
+    {
+      "epoch": 1.7211688601383002,
+      "grad_norm": 0.20512497425079346,
+      "learning_rate": 7.804628537942386e-06,
+      "loss": 0.4726,
+      "step": 7716
+    },
+    {
+      "epoch": 1.7213919250501895,
+      "grad_norm": 0.18379102647304535,
+      "learning_rate": 7.80233258347484e-06,
+      "loss": 0.4779,
+      "step": 7717
+    },
+    {
+      "epoch": 1.721614989962079,
+      "grad_norm": 0.17394806444644928,
+      "learning_rate": 7.800036750727935e-06,
+      "loss": 0.4678,
+      "step": 7718
+    },
+    {
+      "epoch": 1.7218380548739685,
+      "grad_norm": 0.1689169555902481,
+      "learning_rate": 7.797741039828836e-06,
+      "loss": 0.4272,
+      "step": 7719
+    },
+    {
+      "epoch": 1.7220611197858577,
+      "grad_norm": 0.17023077607154846,
+      "learning_rate": 7.79544545090469e-06,
+      "loss": 0.4591,
+      "step": 7720
+    },
+    {
+      "epoch": 1.722284184697747,
+      "grad_norm": 0.18222182989120483,
+      "learning_rate": 7.793149984082643e-06,
+      "loss": 0.4452,
+      "step": 7721
+    },
+    {
+      "epoch": 1.7225072496096363,
+      "grad_norm": 0.18358302116394043,
+      "learning_rate": 7.79085463948983e-06,
+      "loss": 0.4537,
+      "step": 7722
+    },
+    {
+      "epoch": 1.7227303145215258,
+      "grad_norm": 0.1760125756263733,
+      "learning_rate": 7.788559417253386e-06,
+      "loss": 0.4384,
+      "step": 7723
+    },
+    {
+      "epoch": 1.722953379433415,
+      "grad_norm": 0.1962590217590332,
+      "learning_rate": 7.786264317500429e-06,
+      "loss": 0.4778,
+      "step": 7724
+    },
+    {
+      "epoch": 1.7231764443453046,
+      "grad_norm": 0.18565121293067932,
+      "learning_rate": 7.783969340358081e-06,
+      "loss": 0.4379,
+      "step": 7725
+    },
+    {
+      "epoch": 1.7233995092571939,
+      "grad_norm": 0.19673305749893188,
+      "learning_rate": 7.781674485953448e-06,
+      "loss": 0.4361,
+      "step": 7726
+    },
+    {
+      "epoch": 1.7236225741690832,
+      "grad_norm": 0.17665378749370575,
+      "learning_rate": 7.779379754413636e-06,
+      "loss": 0.4323,
+      "step": 7727
+    },
+    {
+      "epoch": 1.7238456390809724,
+      "grad_norm": 0.16950669884681702,
+      "learning_rate": 7.777085145865744e-06,
+      "loss": 0.4348,
+      "step": 7728
+    },
+    {
+      "epoch": 1.724068703992862,
+      "grad_norm": 0.17674678564071655,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.4745,
+      "step": 7729
+    },
+    {
+      "epoch": 1.7242917689047514,
+      "grad_norm": 0.183528333902359,
+      "learning_rate": 7.772496298254063e-06,
+      "loss": 0.4611,
+      "step": 7730
+    },
+    {
+      "epoch": 1.7245148338166407,
+      "grad_norm": 0.18080073595046997,
+      "learning_rate": 7.770202059444433e-06,
+      "loss": 0.4605,
+      "step": 7731
+    },
+    {
+      "epoch": 1.72473789872853,
+      "grad_norm": 0.18171045184135437,
+      "learning_rate": 7.76790794413504e-06,
+      "loss": 0.4527,
+      "step": 7732
+    },
+    {
+      "epoch": 1.7249609636404193,
+      "grad_norm": 0.18763373792171478,
+      "learning_rate": 7.765613952452945e-06,
+      "loss": 0.4559,
+      "step": 7733
+    },
+    {
+      "epoch": 1.7251840285523086,
+      "grad_norm": 0.1811000555753708,
+      "learning_rate": 7.763320084525205e-06,
+      "loss": 0.4536,
+      "step": 7734
+    },
+    {
+      "epoch": 1.725407093464198,
+      "grad_norm": 0.1753537356853485,
+      "learning_rate": 7.761026340478866e-06,
+      "loss": 0.4383,
+      "step": 7735
+    },
+    {
+      "epoch": 1.7256301583760876,
+      "grad_norm": 0.1794290989637375,
+      "learning_rate": 7.758732720440976e-06,
+      "loss": 0.4593,
+      "step": 7736
+    },
+    {
+      "epoch": 1.7258532232879769,
+      "grad_norm": 0.18469299376010895,
+      "learning_rate": 7.75643922453856e-06,
+      "loss": 0.4239,
+      "step": 7737
+    },
+    {
+      "epoch": 1.7260762881998661,
+      "grad_norm": 0.17508934438228607,
+      "learning_rate": 7.754145852898658e-06,
+      "loss": 0.4473,
+      "step": 7738
+    },
+    {
+      "epoch": 1.7262993531117554,
+      "grad_norm": 0.18345895409584045,
+      "learning_rate": 7.751852605648283e-06,
+      "loss": 0.45,
+      "step": 7739
+    },
+    {
+      "epoch": 1.726522418023645,
+      "grad_norm": 0.18025921285152435,
+      "learning_rate": 7.749559482914453e-06,
+      "loss": 0.4645,
+      "step": 7740
+    },
+    {
+      "epoch": 1.7267454829355342,
+      "grad_norm": 0.1723143756389618,
+      "learning_rate": 7.747266484824174e-06,
+      "loss": 0.4087,
+      "step": 7741
+    },
+    {
+      "epoch": 1.7269685478474237,
+      "grad_norm": 0.17820636928081512,
+      "learning_rate": 7.744973611504448e-06,
+      "loss": 0.451,
+      "step": 7742
+    },
+    {
+      "epoch": 1.727191612759313,
+      "grad_norm": 0.17735761404037476,
+      "learning_rate": 7.742680863082267e-06,
+      "loss": 0.435,
+      "step": 7743
+    },
+    {
+      "epoch": 1.7274146776712023,
+      "grad_norm": 0.169492706656456,
+      "learning_rate": 7.74038823968462e-06,
+      "loss": 0.4232,
+      "step": 7744
+    },
+    {
+      "epoch": 1.7276377425830916,
+      "grad_norm": 0.17863723635673523,
+      "learning_rate": 7.738095741438485e-06,
+      "loss": 0.4407,
+      "step": 7745
+    },
+    {
+      "epoch": 1.727860807494981,
+      "grad_norm": 0.17826594412326813,
+      "learning_rate": 7.735803368470836e-06,
+      "loss": 0.4568,
+      "step": 7746
+    },
+    {
+      "epoch": 1.7280838724068706,
+      "grad_norm": 0.18301434814929962,
+      "learning_rate": 7.733511120908639e-06,
+      "loss": 0.471,
+      "step": 7747
+    },
+    {
+      "epoch": 1.7283069373187598,
+      "grad_norm": 0.17796590924263,
+      "learning_rate": 7.73121899887885e-06,
+      "loss": 0.4517,
+      "step": 7748
+    },
+    {
+      "epoch": 1.7285300022306491,
+      "grad_norm": 0.17701533436775208,
+      "learning_rate": 7.728927002508429e-06,
+      "loss": 0.4672,
+      "step": 7749
+    },
+    {
+      "epoch": 1.7287530671425384,
+      "grad_norm": 0.17282108962535858,
+      "learning_rate": 7.726635131924313e-06,
+      "loss": 0.4205,
+      "step": 7750
+    },
+    {
+      "epoch": 1.7289761320544277,
+      "grad_norm": 0.17632122337818146,
+      "learning_rate": 7.724343387253446e-06,
+      "loss": 0.4561,
+      "step": 7751
+    },
+    {
+      "epoch": 1.7291991969663172,
+      "grad_norm": 0.18215633928775787,
+      "learning_rate": 7.722051768622754e-06,
+      "loss": 0.4757,
+      "step": 7752
+    },
+    {
+      "epoch": 1.7294222618782067,
+      "grad_norm": 0.1942843198776245,
+      "learning_rate": 7.719760276159166e-06,
+      "loss": 0.4407,
+      "step": 7753
+    },
+    {
+      "epoch": 1.729645326790096,
+      "grad_norm": 0.18691153824329376,
+      "learning_rate": 7.717468909989595e-06,
+      "loss": 0.4524,
+      "step": 7754
+    },
+    {
+      "epoch": 1.7298683917019853,
+      "grad_norm": 0.16908486187458038,
+      "learning_rate": 7.715177670240956e-06,
+      "loss": 0.4627,
+      "step": 7755
+    },
+    {
+      "epoch": 1.7300914566138745,
+      "grad_norm": 0.1892176866531372,
+      "learning_rate": 7.712886557040147e-06,
+      "loss": 0.4367,
+      "step": 7756
+    },
+    {
+      "epoch": 1.730314521525764,
+      "grad_norm": 0.17933063209056854,
+      "learning_rate": 7.71059557051407e-06,
+      "loss": 0.4912,
+      "step": 7757
+    },
+    {
+      "epoch": 1.7305375864376533,
+      "grad_norm": 0.17910033464431763,
+      "learning_rate": 7.70830471078961e-06,
+      "loss": 0.4412,
+      "step": 7758
+    },
+    {
+      "epoch": 1.7307606513495428,
+      "grad_norm": 0.17961618304252625,
+      "learning_rate": 7.70601397799365e-06,
+      "loss": 0.441,
+      "step": 7759
+    },
+    {
+      "epoch": 1.730983716261432,
+      "grad_norm": 0.17368407547473907,
+      "learning_rate": 7.703723372253064e-06,
+      "loss": 0.4236,
+      "step": 7760
+    },
+    {
+      "epoch": 1.7312067811733214,
+      "grad_norm": 0.16743780672550201,
+      "learning_rate": 7.701432893694727e-06,
+      "loss": 0.4032,
+      "step": 7761
+    },
+    {
+      "epoch": 1.7314298460852107,
+      "grad_norm": 0.17270828783512115,
+      "learning_rate": 7.699142542445489e-06,
+      "loss": 0.4428,
+      "step": 7762
+    },
+    {
+      "epoch": 1.7316529109971002,
+      "grad_norm": 0.18314076960086823,
+      "learning_rate": 7.696852318632214e-06,
+      "loss": 0.4728,
+      "step": 7763
+    },
+    {
+      "epoch": 1.7318759759089897,
+      "grad_norm": 0.17854903638362885,
+      "learning_rate": 7.694562222381741e-06,
+      "loss": 0.4628,
+      "step": 7764
+    },
+    {
+      "epoch": 1.732099040820879,
+      "grad_norm": 0.17335215210914612,
+      "learning_rate": 7.692272253820918e-06,
+      "loss": 0.4492,
+      "step": 7765
+    },
+    {
+      "epoch": 1.7323221057327682,
+      "grad_norm": 0.19753512740135193,
+      "learning_rate": 7.68998241307657e-06,
+      "loss": 0.4499,
+      "step": 7766
+    },
+    {
+      "epoch": 1.7325451706446575,
+      "grad_norm": 0.17499345541000366,
+      "learning_rate": 7.687692700275532e-06,
+      "loss": 0.4394,
+      "step": 7767
+    },
+    {
+      "epoch": 1.7327682355565468,
+      "grad_norm": 0.17762510478496552,
+      "learning_rate": 7.685403115544612e-06,
+      "loss": 0.4656,
+      "step": 7768
+    },
+    {
+      "epoch": 1.7329913004684363,
+      "grad_norm": 0.1747395098209381,
+      "learning_rate": 7.683113659010628e-06,
+      "loss": 0.4431,
+      "step": 7769
+    },
+    {
+      "epoch": 1.7332143653803258,
+      "grad_norm": 0.1724051684141159,
+      "learning_rate": 7.680824330800384e-06,
+      "loss": 0.4266,
+      "step": 7770
+    },
+    {
+      "epoch": 1.733437430292215,
+      "grad_norm": 0.1789834052324295,
+      "learning_rate": 7.678535131040676e-06,
+      "loss": 0.4342,
+      "step": 7771
+    },
+    {
+      "epoch": 1.7336604952041044,
+      "grad_norm": 0.18412499129772186,
+      "learning_rate": 7.676246059858297e-06,
+      "loss": 0.4515,
+      "step": 7772
+    },
+    {
+      "epoch": 1.7338835601159936,
+      "grad_norm": 0.18302331864833832,
+      "learning_rate": 7.673957117380027e-06,
+      "loss": 0.4316,
+      "step": 7773
+    },
+    {
+      "epoch": 1.7341066250278832,
+      "grad_norm": 0.18772312998771667,
+      "learning_rate": 7.671668303732644e-06,
+      "loss": 0.4918,
+      "step": 7774
+    },
+    {
+      "epoch": 1.7343296899397724,
+      "grad_norm": 0.21190685033798218,
+      "learning_rate": 7.669379619042915e-06,
+      "loss": 0.4785,
+      "step": 7775
+    },
+    {
+      "epoch": 1.734552754851662,
+      "grad_norm": 0.18342965841293335,
+      "learning_rate": 7.667091063437604e-06,
+      "loss": 0.4516,
+      "step": 7776
+    },
+    {
+      "epoch": 1.7347758197635512,
+      "grad_norm": 0.1735561341047287,
+      "learning_rate": 7.664802637043463e-06,
+      "loss": 0.426,
+      "step": 7777
+    },
+    {
+      "epoch": 1.7349988846754405,
+      "grad_norm": 0.18314534425735474,
+      "learning_rate": 7.662514339987243e-06,
+      "loss": 0.4523,
+      "step": 7778
+    },
+    {
+      "epoch": 1.7352219495873298,
+      "grad_norm": 0.18044976890087128,
+      "learning_rate": 7.66022617239568e-06,
+      "loss": 0.4398,
+      "step": 7779
+    },
+    {
+      "epoch": 1.7354450144992193,
+      "grad_norm": 0.18999162316322327,
+      "learning_rate": 7.65793813439551e-06,
+      "loss": 0.4582,
+      "step": 7780
+    },
+    {
+      "epoch": 1.7356680794111088,
+      "grad_norm": 0.1812235713005066,
+      "learning_rate": 7.655650226113458e-06,
+      "loss": 0.4576,
+      "step": 7781
+    },
+    {
+      "epoch": 1.735891144322998,
+      "grad_norm": 0.1805853396654129,
+      "learning_rate": 7.653362447676245e-06,
+      "loss": 0.4331,
+      "step": 7782
+    },
+    {
+      "epoch": 1.7361142092348874,
+      "grad_norm": 0.19444364309310913,
+      "learning_rate": 7.651074799210578e-06,
+      "loss": 0.4718,
+      "step": 7783
+    },
+    {
+      "epoch": 1.7363372741467766,
+      "grad_norm": 0.17560544610023499,
+      "learning_rate": 7.648787280843167e-06,
+      "loss": 0.4434,
+      "step": 7784
+    },
+    {
+      "epoch": 1.736560339058666,
+      "grad_norm": 0.17922593653202057,
+      "learning_rate": 7.646499892700703e-06,
+      "loss": 0.4171,
+      "step": 7785
+    },
+    {
+      "epoch": 1.7367834039705554,
+      "grad_norm": 0.17291361093521118,
+      "learning_rate": 7.644212634909881e-06,
+      "loss": 0.4309,
+      "step": 7786
+    },
+    {
+      "epoch": 1.737006468882445,
+      "grad_norm": 0.2205720990896225,
+      "learning_rate": 7.641925507597381e-06,
+      "loss": 0.4471,
+      "step": 7787
+    },
+    {
+      "epoch": 1.7372295337943342,
+      "grad_norm": 0.2528192698955536,
+      "learning_rate": 7.63963851088988e-06,
+      "loss": 0.463,
+      "step": 7788
+    },
+    {
+      "epoch": 1.7374525987062235,
+      "grad_norm": 0.1895085573196411,
+      "learning_rate": 7.637351644914044e-06,
+      "loss": 0.4486,
+      "step": 7789
+    },
+    {
+      "epoch": 1.7376756636181128,
+      "grad_norm": 0.17160922288894653,
+      "learning_rate": 7.635064909796534e-06,
+      "loss": 0.4531,
+      "step": 7790
+    },
+    {
+      "epoch": 1.7378987285300023,
+      "grad_norm": 0.17404474318027496,
+      "learning_rate": 7.63277830566401e-06,
+      "loss": 0.4446,
+      "step": 7791
+    },
+    {
+      "epoch": 1.7381217934418915,
+      "grad_norm": 0.17817366123199463,
+      "learning_rate": 7.63049183264311e-06,
+      "loss": 0.4627,
+      "step": 7792
+    },
+    {
+      "epoch": 1.738344858353781,
+      "grad_norm": 0.1933346390724182,
+      "learning_rate": 7.628205490860482e-06,
+      "loss": 0.4744,
+      "step": 7793
+    },
+    {
+      "epoch": 1.7385679232656703,
+      "grad_norm": 0.1809963434934616,
+      "learning_rate": 7.625919280442751e-06,
+      "loss": 0.4791,
+      "step": 7794
+    },
+    {
+      "epoch": 1.7387909881775596,
+      "grad_norm": 0.31423863768577576,
+      "learning_rate": 7.623633201516542e-06,
+      "loss": 0.4465,
+      "step": 7795
+    },
+    {
+      "epoch": 1.739014053089449,
+      "grad_norm": 0.19363394379615784,
+      "learning_rate": 7.621347254208475e-06,
+      "loss": 0.4696,
+      "step": 7796
+    },
+    {
+      "epoch": 1.7392371180013384,
+      "grad_norm": 0.18826474249362946,
+      "learning_rate": 7.619061438645159e-06,
+      "loss": 0.4746,
+      "step": 7797
+    },
+    {
+      "epoch": 1.739460182913228,
+      "grad_norm": 0.16698677837848663,
+      "learning_rate": 7.616775754953199e-06,
+      "loss": 0.4349,
+      "step": 7798
+    },
+    {
+      "epoch": 1.7396832478251172,
+      "grad_norm": 0.17490437626838684,
+      "learning_rate": 7.614490203259186e-06,
+      "loss": 0.4237,
+      "step": 7799
+    },
+    {
+      "epoch": 1.7399063127370065,
+      "grad_norm": 0.17656764388084412,
+      "learning_rate": 7.6122047836897125e-06,
+      "loss": 0.4583,
+      "step": 7800
+    },
+    {
+      "epoch": 1.7401293776488957,
+      "grad_norm": 0.17693257331848145,
+      "learning_rate": 7.609919496371357e-06,
+      "loss": 0.454,
+      "step": 7801
+    },
+    {
+      "epoch": 1.740352442560785,
+      "grad_norm": 0.1721879243850708,
+      "learning_rate": 7.6076343414306965e-06,
+      "loss": 0.4373,
+      "step": 7802
+    },
+    {
+      "epoch": 1.7405755074726745,
+      "grad_norm": 0.17377860844135284,
+      "learning_rate": 7.605349318994291e-06,
+      "loss": 0.4456,
+      "step": 7803
+    },
+    {
+      "epoch": 1.740798572384564,
+      "grad_norm": 0.17918775975704193,
+      "learning_rate": 7.603064429188707e-06,
+      "loss": 0.4473,
+      "step": 7804
+    },
+    {
+      "epoch": 1.7410216372964533,
+      "grad_norm": 0.17350159585475922,
+      "learning_rate": 7.600779672140488e-06,
+      "loss": 0.4505,
+      "step": 7805
+    },
+    {
+      "epoch": 1.7412447022083426,
+      "grad_norm": 0.1735851764678955,
+      "learning_rate": 7.5984950479761865e-06,
+      "loss": 0.432,
+      "step": 7806
+    },
+    {
+      "epoch": 1.7414677671202319,
+      "grad_norm": 0.19026067852973938,
+      "learning_rate": 7.596210556822332e-06,
+      "loss": 0.4668,
+      "step": 7807
+    },
+    {
+      "epoch": 1.7416908320321214,
+      "grad_norm": 0.17454437911510468,
+      "learning_rate": 7.59392619880546e-06,
+      "loss": 0.4583,
+      "step": 7808
+    },
+    {
+      "epoch": 1.7419138969440107,
+      "grad_norm": 0.17719118297100067,
+      "learning_rate": 7.591641974052089e-06,
+      "loss": 0.4214,
+      "step": 7809
+    },
+    {
+      "epoch": 1.7421369618559002,
+      "grad_norm": 0.17559269070625305,
+      "learning_rate": 7.589357882688735e-06,
+      "loss": 0.4234,
+      "step": 7810
+    },
+    {
+      "epoch": 1.7423600267677894,
+      "grad_norm": 0.17429721355438232,
+      "learning_rate": 7.587073924841904e-06,
+      "loss": 0.4205,
+      "step": 7811
+    },
+    {
+      "epoch": 1.7425830916796787,
+      "grad_norm": 0.18059833347797394,
+      "learning_rate": 7.584790100638098e-06,
+      "loss": 0.4475,
+      "step": 7812
+    },
+    {
+      "epoch": 1.742806156591568,
+      "grad_norm": 0.1845857799053192,
+      "learning_rate": 7.582506410203806e-06,
+      "loss": 0.4408,
+      "step": 7813
+    },
+    {
+      "epoch": 1.7430292215034575,
+      "grad_norm": 0.18028376996517181,
+      "learning_rate": 7.580222853665517e-06,
+      "loss": 0.4213,
+      "step": 7814
+    },
+    {
+      "epoch": 1.743252286415347,
+      "grad_norm": 0.1828741878271103,
+      "learning_rate": 7.577939431149709e-06,
+      "loss": 0.4559,
+      "step": 7815
+    },
+    {
+      "epoch": 1.7434753513272363,
+      "grad_norm": 0.19069290161132812,
+      "learning_rate": 7.575656142782848e-06,
+      "loss": 0.4372,
+      "step": 7816
+    },
+    {
+      "epoch": 1.7436984162391256,
+      "grad_norm": 0.1698480099439621,
+      "learning_rate": 7.573372988691402e-06,
+      "loss": 0.431,
+      "step": 7817
+    },
+    {
+      "epoch": 1.7439214811510149,
+      "grad_norm": 0.17554740607738495,
+      "learning_rate": 7.5710899690018226e-06,
+      "loss": 0.4536,
+      "step": 7818
+    },
+    {
+      "epoch": 1.7441445460629041,
+      "grad_norm": 0.17417636513710022,
+      "learning_rate": 7.568807083840561e-06,
+      "loss": 0.4561,
+      "step": 7819
+    },
+    {
+      "epoch": 1.7443676109747936,
+      "grad_norm": 0.1864258497953415,
+      "learning_rate": 7.566524333334053e-06,
+      "loss": 0.4933,
+      "step": 7820
+    },
+    {
+      "epoch": 1.7445906758866832,
+      "grad_norm": 0.20237578451633453,
+      "learning_rate": 7.564241717608737e-06,
+      "loss": 0.4429,
+      "step": 7821
+    },
+    {
+      "epoch": 1.7448137407985724,
+      "grad_norm": 0.18540610373020172,
+      "learning_rate": 7.5619592367910345e-06,
+      "loss": 0.4572,
+      "step": 7822
+    },
+    {
+      "epoch": 1.7450368057104617,
+      "grad_norm": 0.17512571811676025,
+      "learning_rate": 7.5596768910073674e-06,
+      "loss": 0.4408,
+      "step": 7823
+    },
+    {
+      "epoch": 1.745259870622351,
+      "grad_norm": 0.19032084941864014,
+      "learning_rate": 7.557394680384142e-06,
+      "loss": 0.4466,
+      "step": 7824
+    },
+    {
+      "epoch": 1.7454829355342405,
+      "grad_norm": 0.18703311681747437,
+      "learning_rate": 7.555112605047766e-06,
+      "loss": 0.4681,
+      "step": 7825
+    },
+    {
+      "epoch": 1.7457060004461298,
+      "grad_norm": 0.17146390676498413,
+      "learning_rate": 7.55283066512463e-06,
+      "loss": 0.4518,
+      "step": 7826
+    },
+    {
+      "epoch": 1.7459290653580193,
+      "grad_norm": 0.17544762790203094,
+      "learning_rate": 7.550548860741128e-06,
+      "loss": 0.4459,
+      "step": 7827
+    },
+    {
+      "epoch": 1.7461521302699086,
+      "grad_norm": 0.16931942105293274,
+      "learning_rate": 7.548267192023636e-06,
+      "loss": 0.4317,
+      "step": 7828
+    },
+    {
+      "epoch": 1.7463751951817978,
+      "grad_norm": 0.17575553059577942,
+      "learning_rate": 7.545985659098531e-06,
+      "loss": 0.4239,
+      "step": 7829
+    },
+    {
+      "epoch": 1.7465982600936871,
+      "grad_norm": 0.1667536497116089,
+      "learning_rate": 7.543704262092174e-06,
+      "loss": 0.4351,
+      "step": 7830
+    },
+    {
+      "epoch": 1.7468213250055766,
+      "grad_norm": 0.17519286274909973,
+      "learning_rate": 7.541423001130928e-06,
+      "loss": 0.4405,
+      "step": 7831
+    },
+    {
+      "epoch": 1.7470443899174661,
+      "grad_norm": 0.7807138562202454,
+      "learning_rate": 7.5391418763411385e-06,
+      "loss": 0.4395,
+      "step": 7832
+    },
+    {
+      "epoch": 1.7472674548293554,
+      "grad_norm": 0.17985686659812927,
+      "learning_rate": 7.536860887849152e-06,
+      "loss": 0.4564,
+      "step": 7833
+    },
+    {
+      "epoch": 1.7474905197412447,
+      "grad_norm": 0.18918108940124512,
+      "learning_rate": 7.534580035781304e-06,
+      "loss": 0.4898,
+      "step": 7834
+    },
+    {
+      "epoch": 1.747713584653134,
+      "grad_norm": 0.19184869527816772,
+      "learning_rate": 7.532299320263922e-06,
+      "loss": 0.4714,
+      "step": 7835
+    },
+    {
+      "epoch": 1.7479366495650233,
+      "grad_norm": 0.18925781548023224,
+      "learning_rate": 7.530018741423328e-06,
+      "loss": 0.4548,
+      "step": 7836
+    },
+    {
+      "epoch": 1.7481597144769128,
+      "grad_norm": 0.188811257481575,
+      "learning_rate": 7.5277382993858295e-06,
+      "loss": 0.4444,
+      "step": 7837
+    },
+    {
+      "epoch": 1.7483827793888023,
+      "grad_norm": 0.18189990520477295,
+      "learning_rate": 7.525457994277737e-06,
+      "loss": 0.4754,
+      "step": 7838
+    },
+    {
+      "epoch": 1.7486058443006915,
+      "grad_norm": 0.179085835814476,
+      "learning_rate": 7.5231778262253456e-06,
+      "loss": 0.4511,
+      "step": 7839
+    },
+    {
+      "epoch": 1.7488289092125808,
+      "grad_norm": 0.17552562057971954,
+      "learning_rate": 7.520897795354947e-06,
+      "loss": 0.446,
+      "step": 7840
+    },
+    {
+      "epoch": 1.74905197412447,
+      "grad_norm": 0.1869988739490509,
+      "learning_rate": 7.518617901792819e-06,
+      "loss": 0.4591,
+      "step": 7841
+    },
+    {
+      "epoch": 1.7492750390363596,
+      "grad_norm": 0.17976555228233337,
+      "learning_rate": 7.516338145665245e-06,
+      "loss": 0.4489,
+      "step": 7842
+    },
+    {
+      "epoch": 1.749498103948249,
+      "grad_norm": 0.18475496768951416,
+      "learning_rate": 7.514058527098484e-06,
+      "loss": 0.4781,
+      "step": 7843
+    },
+    {
+      "epoch": 1.7497211688601384,
+      "grad_norm": 0.17996063828468323,
+      "learning_rate": 7.5117790462188016e-06,
+      "loss": 0.435,
+      "step": 7844
+    },
+    {
+      "epoch": 1.7499442337720277,
+      "grad_norm": 0.1795787513256073,
+      "learning_rate": 7.509499703152445e-06,
+      "loss": 0.4901,
+      "step": 7845
+    },
+    {
+      "epoch": 1.750167298683917,
+      "grad_norm": 0.16901874542236328,
+      "learning_rate": 7.507220498025662e-06,
+      "loss": 0.4694,
+      "step": 7846
+    },
+    {
+      "epoch": 1.7503903635958062,
+      "grad_norm": 0.16905230283737183,
+      "learning_rate": 7.504941430964687e-06,
+      "loss": 0.4333,
+      "step": 7847
+    },
+    {
+      "epoch": 1.7506134285076957,
+      "grad_norm": 0.17756043374538422,
+      "learning_rate": 7.502662502095752e-06,
+      "loss": 0.4535,
+      "step": 7848
+    },
+    {
+      "epoch": 1.7508364934195852,
+      "grad_norm": 0.17667444050312042,
+      "learning_rate": 7.500383711545074e-06,
+      "loss": 0.4636,
+      "step": 7849
+    },
+    {
+      "epoch": 1.7510595583314745,
+      "grad_norm": 0.18121804296970367,
+      "learning_rate": 7.4981050594388716e-06,
+      "loss": 0.4796,
+      "step": 7850
+    },
+    {
+      "epoch": 1.7512826232433638,
+      "grad_norm": 0.18036022782325745,
+      "learning_rate": 7.495826545903347e-06,
+      "loss": 0.4667,
+      "step": 7851
+    },
+    {
+      "epoch": 1.751505688155253,
+      "grad_norm": 0.24715840816497803,
+      "learning_rate": 7.4935481710647005e-06,
+      "loss": 0.4713,
+      "step": 7852
+    },
+    {
+      "epoch": 1.7517287530671424,
+      "grad_norm": 0.19006216526031494,
+      "learning_rate": 7.491269935049122e-06,
+      "loss": 0.4464,
+      "step": 7853
+    },
+    {
+      "epoch": 1.7519518179790319,
+      "grad_norm": 0.18384769558906555,
+      "learning_rate": 7.488991837982794e-06,
+      "loss": 0.4707,
+      "step": 7854
+    },
+    {
+      "epoch": 1.7521748828909214,
+      "grad_norm": 0.18586494028568268,
+      "learning_rate": 7.486713879991896e-06,
+      "loss": 0.4731,
+      "step": 7855
+    },
+    {
+      "epoch": 1.7523979478028107,
+      "grad_norm": 0.19443517923355103,
+      "learning_rate": 7.484436061202589e-06,
+      "loss": 0.4883,
+      "step": 7856
+    },
+    {
+      "epoch": 1.7526210127147,
+      "grad_norm": 0.17276805639266968,
+      "learning_rate": 7.48215838174104e-06,
+      "loss": 0.4449,
+      "step": 7857
+    },
+    {
+      "epoch": 1.7528440776265892,
+      "grad_norm": 0.1835545003414154,
+      "learning_rate": 7.4798808417333954e-06,
+      "loss": 0.4748,
+      "step": 7858
+    },
+    {
+      "epoch": 1.7530671425384787,
+      "grad_norm": 0.1789509505033493,
+      "learning_rate": 7.477603441305803e-06,
+      "loss": 0.4639,
+      "step": 7859
+    },
+    {
+      "epoch": 1.753290207450368,
+      "grad_norm": 0.193350687623024,
+      "learning_rate": 7.475326180584398e-06,
+      "loss": 0.4897,
+      "step": 7860
+    },
+    {
+      "epoch": 1.7535132723622575,
+      "grad_norm": 0.17808929085731506,
+      "learning_rate": 7.473049059695311e-06,
+      "loss": 0.4593,
+      "step": 7861
+    },
+    {
+      "epoch": 1.7537363372741468,
+      "grad_norm": 0.18136462569236755,
+      "learning_rate": 7.470772078764661e-06,
+      "loss": 0.4693,
+      "step": 7862
+    },
+    {
+      "epoch": 1.753959402186036,
+      "grad_norm": 0.1805723011493683,
+      "learning_rate": 7.4684952379185625e-06,
+      "loss": 0.4663,
+      "step": 7863
+    },
+    {
+      "epoch": 1.7541824670979254,
+      "grad_norm": 0.1832210123538971,
+      "learning_rate": 7.466218537283122e-06,
+      "loss": 0.4541,
+      "step": 7864
+    },
+    {
+      "epoch": 1.7544055320098149,
+      "grad_norm": 0.1848025768995285,
+      "learning_rate": 7.463941976984439e-06,
+      "loss": 0.4507,
+      "step": 7865
+    },
+    {
+      "epoch": 1.7546285969217044,
+      "grad_norm": 0.1897166222333908,
+      "learning_rate": 7.461665557148599e-06,
+      "loss": 0.4422,
+      "step": 7866
+    },
+    {
+      "epoch": 1.7548516618335936,
+      "grad_norm": 0.1763024479150772,
+      "learning_rate": 7.459389277901689e-06,
+      "loss": 0.4611,
+      "step": 7867
+    },
+    {
+      "epoch": 1.755074726745483,
+      "grad_norm": 0.1759800910949707,
+      "learning_rate": 7.4571131393697795e-06,
+      "loss": 0.4406,
+      "step": 7868
+    },
+    {
+      "epoch": 1.7552977916573722,
+      "grad_norm": 0.17623275518417358,
+      "learning_rate": 7.454837141678942e-06,
+      "loss": 0.4643,
+      "step": 7869
+    },
+    {
+      "epoch": 1.7555208565692615,
+      "grad_norm": 0.1801615208387375,
+      "learning_rate": 7.452561284955231e-06,
+      "loss": 0.4649,
+      "step": 7870
+    },
+    {
+      "epoch": 1.755743921481151,
+      "grad_norm": 0.1817091703414917,
+      "learning_rate": 7.450285569324703e-06,
+      "loss": 0.4531,
+      "step": 7871
+    },
+    {
+      "epoch": 1.7559669863930405,
+      "grad_norm": 0.18977338075637817,
+      "learning_rate": 7.448009994913396e-06,
+      "loss": 0.4476,
+      "step": 7872
+    },
+    {
+      "epoch": 1.7561900513049298,
+      "grad_norm": 0.17964302003383636,
+      "learning_rate": 7.445734561847351e-06,
+      "loss": 0.4365,
+      "step": 7873
+    },
+    {
+      "epoch": 1.756413116216819,
+      "grad_norm": 0.17655393481254578,
+      "learning_rate": 7.44345927025259e-06,
+      "loss": 0.4412,
+      "step": 7874
+    },
+    {
+      "epoch": 1.7566361811287083,
+      "grad_norm": 0.1832321286201477,
+      "learning_rate": 7.441184120255141e-06,
+      "loss": 0.4589,
+      "step": 7875
+    },
+    {
+      "epoch": 1.7568592460405978,
+      "grad_norm": 0.18122322857379913,
+      "learning_rate": 7.438909111981008e-06,
+      "loss": 0.4363,
+      "step": 7876
+    },
+    {
+      "epoch": 1.7570823109524871,
+      "grad_norm": 0.19032621383666992,
+      "learning_rate": 7.436634245556195e-06,
+      "loss": 0.454,
+      "step": 7877
+    },
+    {
+      "epoch": 1.7573053758643766,
+      "grad_norm": 0.1851622313261032,
+      "learning_rate": 7.4343595211067045e-06,
+      "loss": 0.4839,
+      "step": 7878
+    },
+    {
+      "epoch": 1.757528440776266,
+      "grad_norm": 0.2012101113796234,
+      "learning_rate": 7.432084938758521e-06,
+      "loss": 0.4918,
+      "step": 7879
+    },
+    {
+      "epoch": 1.7577515056881552,
+      "grad_norm": 0.18474964797496796,
+      "learning_rate": 7.429810498637627e-06,
+      "loss": 0.4552,
+      "step": 7880
+    },
+    {
+      "epoch": 1.7579745706000445,
+      "grad_norm": 0.17664113640785217,
+      "learning_rate": 7.427536200869996e-06,
+      "loss": 0.4456,
+      "step": 7881
+    },
+    {
+      "epoch": 1.758197635511934,
+      "grad_norm": 0.18062026798725128,
+      "learning_rate": 7.425262045581589e-06,
+      "loss": 0.456,
+      "step": 7882
+    },
+    {
+      "epoch": 1.7584207004238235,
+      "grad_norm": 0.18045106530189514,
+      "learning_rate": 7.42298803289837e-06,
+      "loss": 0.4886,
+      "step": 7883
+    },
+    {
+      "epoch": 1.7586437653357128,
+      "grad_norm": 0.17931503057479858,
+      "learning_rate": 7.420714162946278e-06,
+      "loss": 0.4453,
+      "step": 7884
+    },
+    {
+      "epoch": 1.758866830247602,
+      "grad_norm": 0.1748034656047821,
+      "learning_rate": 7.418440435851265e-06,
+      "loss": 0.4626,
+      "step": 7885
+    },
+    {
+      "epoch": 1.7590898951594913,
+      "grad_norm": 0.17426064610481262,
+      "learning_rate": 7.416166851739257e-06,
+      "loss": 0.4369,
+      "step": 7886
+    },
+    {
+      "epoch": 1.7593129600713808,
+      "grad_norm": 0.17878130078315735,
+      "learning_rate": 7.4138934107361825e-06,
+      "loss": 0.4624,
+      "step": 7887
+    },
+    {
+      "epoch": 1.75953602498327,
+      "grad_norm": 0.1841915249824524,
+      "learning_rate": 7.411620112967957e-06,
+      "loss": 0.4625,
+      "step": 7888
+    },
+    {
+      "epoch": 1.7597590898951596,
+      "grad_norm": 0.18047453463077545,
+      "learning_rate": 7.4093469585604925e-06,
+      "loss": 0.4383,
+      "step": 7889
+    },
+    {
+      "epoch": 1.759982154807049,
+      "grad_norm": 0.1770285815000534,
+      "learning_rate": 7.407073947639688e-06,
+      "loss": 0.4267,
+      "step": 7890
+    },
+    {
+      "epoch": 1.7602052197189382,
+      "grad_norm": 0.17349712550640106,
+      "learning_rate": 7.404801080331441e-06,
+      "loss": 0.4404,
+      "step": 7891
+    },
+    {
+      "epoch": 1.7604282846308275,
+      "grad_norm": 0.17680755257606506,
+      "learning_rate": 7.4025283567616315e-06,
+      "loss": 0.4272,
+      "step": 7892
+    },
+    {
+      "epoch": 1.760651349542717,
+      "grad_norm": 0.18388980627059937,
+      "learning_rate": 7.400255777056142e-06,
+      "loss": 0.4677,
+      "step": 7893
+    },
+    {
+      "epoch": 1.7608744144546062,
+      "grad_norm": 0.18694987893104553,
+      "learning_rate": 7.39798334134084e-06,
+      "loss": 0.4669,
+      "step": 7894
+    },
+    {
+      "epoch": 1.7610974793664957,
+      "grad_norm": 0.16978652775287628,
+      "learning_rate": 7.395711049741588e-06,
+      "loss": 0.4435,
+      "step": 7895
+    },
+    {
+      "epoch": 1.761320544278385,
+      "grad_norm": 0.17169198393821716,
+      "learning_rate": 7.3934389023842405e-06,
+      "loss": 0.4306,
+      "step": 7896
+    },
+    {
+      "epoch": 1.7615436091902743,
+      "grad_norm": 0.18931961059570312,
+      "learning_rate": 7.391166899394642e-06,
+      "loss": 0.4597,
+      "step": 7897
+    },
+    {
+      "epoch": 1.7617666741021636,
+      "grad_norm": 0.17674033343791962,
+      "learning_rate": 7.3888950408986294e-06,
+      "loss": 0.4626,
+      "step": 7898
+    },
+    {
+      "epoch": 1.761989739014053,
+      "grad_norm": 0.18383964896202087,
+      "learning_rate": 7.386623327022034e-06,
+      "loss": 0.4635,
+      "step": 7899
+    },
+    {
+      "epoch": 1.7622128039259426,
+      "grad_norm": 0.1803615838289261,
+      "learning_rate": 7.38435175789068e-06,
+      "loss": 0.4811,
+      "step": 7900
+    },
+    {
+      "epoch": 1.7624358688378319,
+      "grad_norm": 0.19607679545879364,
+      "learning_rate": 7.382080333630377e-06,
+      "loss": 0.4715,
+      "step": 7901
+    },
+    {
+      "epoch": 1.7626589337497212,
+      "grad_norm": 0.17395561933517456,
+      "learning_rate": 7.379809054366933e-06,
+      "loss": 0.4515,
+      "step": 7902
+    },
+    {
+      "epoch": 1.7628819986616104,
+      "grad_norm": 0.1800202876329422,
+      "learning_rate": 7.3775379202261455e-06,
+      "loss": 0.4933,
+      "step": 7903
+    },
+    {
+      "epoch": 1.7631050635735,
+      "grad_norm": 0.1730235368013382,
+      "learning_rate": 7.375266931333805e-06,
+      "loss": 0.4634,
+      "step": 7904
+    },
+    {
+      "epoch": 1.7633281284853892,
+      "grad_norm": 0.1844315528869629,
+      "learning_rate": 7.37299608781569e-06,
+      "loss": 0.456,
+      "step": 7905
+    },
+    {
+      "epoch": 1.7635511933972787,
+      "grad_norm": 0.17907127737998962,
+      "learning_rate": 7.370725389797577e-06,
+      "loss": 0.451,
+      "step": 7906
+    },
+    {
+      "epoch": 1.763774258309168,
+      "grad_norm": 0.18444164097309113,
+      "learning_rate": 7.368454837405229e-06,
+      "loss": 0.4647,
+      "step": 7907
+    },
+    {
+      "epoch": 1.7639973232210573,
+      "grad_norm": 0.1789025217294693,
+      "learning_rate": 7.366184430764407e-06,
+      "loss": 0.4542,
+      "step": 7908
+    },
+    {
+      "epoch": 1.7642203881329466,
+      "grad_norm": 0.17742502689361572,
+      "learning_rate": 7.363914170000858e-06,
+      "loss": 0.4419,
+      "step": 7909
+    },
+    {
+      "epoch": 1.764443453044836,
+      "grad_norm": 0.18832780420780182,
+      "learning_rate": 7.361644055240325e-06,
+      "loss": 0.4763,
+      "step": 7910
+    },
+    {
+      "epoch": 1.7646665179567254,
+      "grad_norm": 0.2054174244403839,
+      "learning_rate": 7.359374086608539e-06,
+      "loss": 0.4543,
+      "step": 7911
+    },
+    {
+      "epoch": 1.7648895828686149,
+      "grad_norm": 0.17511680722236633,
+      "learning_rate": 7.357104264231228e-06,
+      "loss": 0.4377,
+      "step": 7912
+    },
+    {
+      "epoch": 1.7651126477805041,
+      "grad_norm": 0.1764538437128067,
+      "learning_rate": 7.354834588234105e-06,
+      "loss": 0.4675,
+      "step": 7913
+    },
+    {
+      "epoch": 1.7653357126923934,
+      "grad_norm": 0.17360688745975494,
+      "learning_rate": 7.352565058742882e-06,
+      "loss": 0.4409,
+      "step": 7914
+    },
+    {
+      "epoch": 1.7655587776042827,
+      "grad_norm": 0.1800130158662796,
+      "learning_rate": 7.350295675883259e-06,
+      "loss": 0.466,
+      "step": 7915
+    },
+    {
+      "epoch": 1.7657818425161722,
+      "grad_norm": 0.1828777939081192,
+      "learning_rate": 7.34802643978093e-06,
+      "loss": 0.4603,
+      "step": 7916
+    },
+    {
+      "epoch": 1.7660049074280617,
+      "grad_norm": 0.180416077375412,
+      "learning_rate": 7.345757350561577e-06,
+      "loss": 0.4442,
+      "step": 7917
+    },
+    {
+      "epoch": 1.766227972339951,
+      "grad_norm": 0.18395313620567322,
+      "learning_rate": 7.343488408350879e-06,
+      "loss": 0.4801,
+      "step": 7918
+    },
+    {
+      "epoch": 1.7664510372518403,
+      "grad_norm": 0.1887856274843216,
+      "learning_rate": 7.3412196132745005e-06,
+      "loss": 0.4421,
+      "step": 7919
+    },
+    {
+      "epoch": 1.7666741021637296,
+      "grad_norm": 0.17162606120109558,
+      "learning_rate": 7.3389509654581045e-06,
+      "loss": 0.4293,
+      "step": 7920
+    },
+    {
+      "epoch": 1.766897167075619,
+      "grad_norm": 0.18713156878948212,
+      "learning_rate": 7.336682465027346e-06,
+      "loss": 0.4521,
+      "step": 7921
+    },
+    {
+      "epoch": 1.7671202319875083,
+      "grad_norm": 0.1808861941099167,
+      "learning_rate": 7.334414112107863e-06,
+      "loss": 0.4461,
+      "step": 7922
+    },
+    {
+      "epoch": 1.7673432968993978,
+      "grad_norm": 0.1840147227048874,
+      "learning_rate": 7.3321459068252955e-06,
+      "loss": 0.4606,
+      "step": 7923
+    },
+    {
+      "epoch": 1.7675663618112871,
+      "grad_norm": 0.22466281056404114,
+      "learning_rate": 7.329877849305267e-06,
+      "loss": 0.4342,
+      "step": 7924
+    },
+    {
+      "epoch": 1.7677894267231764,
+      "grad_norm": 0.18192514777183533,
+      "learning_rate": 7.327609939673401e-06,
+      "loss": 0.4638,
+      "step": 7925
+    },
+    {
+      "epoch": 1.7680124916350657,
+      "grad_norm": 0.1804400235414505,
+      "learning_rate": 7.325342178055304e-06,
+      "loss": 0.4558,
+      "step": 7926
+    },
+    {
+      "epoch": 1.7682355565469552,
+      "grad_norm": 0.1800002157688141,
+      "learning_rate": 7.3230745645765845e-06,
+      "loss": 0.4303,
+      "step": 7927
+    },
+    {
+      "epoch": 1.7684586214588445,
+      "grad_norm": 0.17886100709438324,
+      "learning_rate": 7.320807099362831e-06,
+      "loss": 0.4507,
+      "step": 7928
+    },
+    {
+      "epoch": 1.768681686370734,
+      "grad_norm": 0.1851029098033905,
+      "learning_rate": 7.318539782539635e-06,
+      "loss": 0.4693,
+      "step": 7929
+    },
+    {
+      "epoch": 1.7689047512826233,
+      "grad_norm": 0.19121414422988892,
+      "learning_rate": 7.316272614232572e-06,
+      "loss": 0.4465,
+      "step": 7930
+    },
+    {
+      "epoch": 1.7691278161945125,
+      "grad_norm": 0.18079820275306702,
+      "learning_rate": 7.314005594567215e-06,
+      "loss": 0.4475,
+      "step": 7931
+    },
+    {
+      "epoch": 1.7693508811064018,
+      "grad_norm": 0.1832534670829773,
+      "learning_rate": 7.311738723669122e-06,
+      "loss": 0.4585,
+      "step": 7932
+    },
+    {
+      "epoch": 1.7695739460182913,
+      "grad_norm": 0.18712462484836578,
+      "learning_rate": 7.309472001663849e-06,
+      "loss": 0.4578,
+      "step": 7933
+    },
+    {
+      "epoch": 1.7697970109301808,
+      "grad_norm": 0.17979365587234497,
+      "learning_rate": 7.307205428676939e-06,
+      "loss": 0.4711,
+      "step": 7934
+    },
+    {
+      "epoch": 1.77002007584207,
+      "grad_norm": 0.19163386523723602,
+      "learning_rate": 7.304939004833933e-06,
+      "loss": 0.432,
+      "step": 7935
+    },
+    {
+      "epoch": 1.7702431407539594,
+      "grad_norm": 0.20467031002044678,
+      "learning_rate": 7.302672730260356e-06,
+      "loss": 0.4606,
+      "step": 7936
+    },
+    {
+      "epoch": 1.7704662056658487,
+      "grad_norm": 0.1881827414035797,
+      "learning_rate": 7.300406605081732e-06,
+      "loss": 0.459,
+      "step": 7937
+    },
+    {
+      "epoch": 1.7706892705777382,
+      "grad_norm": 0.18381071090698242,
+      "learning_rate": 7.298140629423568e-06,
+      "loss": 0.4804,
+      "step": 7938
+    },
+    {
+      "epoch": 1.7709123354896275,
+      "grad_norm": 0.18901976943016052,
+      "learning_rate": 7.295874803411373e-06,
+      "loss": 0.4568,
+      "step": 7939
+    },
+    {
+      "epoch": 1.771135400401517,
+      "grad_norm": 0.17536242306232452,
+      "learning_rate": 7.293609127170642e-06,
+      "loss": 0.4217,
+      "step": 7940
+    },
+    {
+      "epoch": 1.7713584653134062,
+      "grad_norm": 0.17019738256931305,
+      "learning_rate": 7.291343600826859e-06,
+      "loss": 0.4405,
+      "step": 7941
+    },
+    {
+      "epoch": 1.7715815302252955,
+      "grad_norm": 0.1769004762172699,
+      "learning_rate": 7.2890782245055084e-06,
+      "loss": 0.461,
+      "step": 7942
+    },
+    {
+      "epoch": 1.7718045951371848,
+      "grad_norm": 0.18277987837791443,
+      "learning_rate": 7.286812998332056e-06,
+      "loss": 0.4691,
+      "step": 7943
+    },
+    {
+      "epoch": 1.7720276600490743,
+      "grad_norm": 0.18873664736747742,
+      "learning_rate": 7.2845479224319685e-06,
+      "loss": 0.4606,
+      "step": 7944
+    },
+    {
+      "epoch": 1.7722507249609636,
+      "grad_norm": 0.1791076958179474,
+      "learning_rate": 7.2822829969306955e-06,
+      "loss": 0.4369,
+      "step": 7945
+    },
+    {
+      "epoch": 1.772473789872853,
+      "grad_norm": 0.17617996037006378,
+      "learning_rate": 7.280018221953686e-06,
+      "loss": 0.4538,
+      "step": 7946
+    },
+    {
+      "epoch": 1.7726968547847424,
+      "grad_norm": 0.1820133924484253,
+      "learning_rate": 7.277753597626378e-06,
+      "loss": 0.4444,
+      "step": 7947
+    },
+    {
+      "epoch": 1.7729199196966317,
+      "grad_norm": 0.16976666450500488,
+      "learning_rate": 7.275489124074198e-06,
+      "loss": 0.4415,
+      "step": 7948
+    },
+    {
+      "epoch": 1.773142984608521,
+      "grad_norm": 0.19467420876026154,
+      "learning_rate": 7.273224801422568e-06,
+      "loss": 0.4527,
+      "step": 7949
+    },
+    {
+      "epoch": 1.7733660495204104,
+      "grad_norm": 0.18623554706573486,
+      "learning_rate": 7.270960629796903e-06,
+      "loss": 0.4552,
+      "step": 7950
+    },
+    {
+      "epoch": 1.7735891144323,
+      "grad_norm": 0.1663247048854828,
+      "learning_rate": 7.2686966093226e-06,
+      "loss": 0.4154,
+      "step": 7951
+    },
+    {
+      "epoch": 1.7738121793441892,
+      "grad_norm": 0.18195995688438416,
+      "learning_rate": 7.266432740125065e-06,
+      "loss": 0.4501,
+      "step": 7952
+    },
+    {
+      "epoch": 1.7740352442560785,
+      "grad_norm": 0.18873853981494904,
+      "learning_rate": 7.264169022329676e-06,
+      "loss": 0.4529,
+      "step": 7953
+    },
+    {
+      "epoch": 1.7742583091679678,
+      "grad_norm": 0.1879783272743225,
+      "learning_rate": 7.2619054560618175e-06,
+      "loss": 0.4949,
+      "step": 7954
+    },
+    {
+      "epoch": 1.7744813740798573,
+      "grad_norm": 0.17232050001621246,
+      "learning_rate": 7.259642041446856e-06,
+      "loss": 0.4307,
+      "step": 7955
+    },
+    {
+      "epoch": 1.7747044389917466,
+      "grad_norm": 0.18015244603157043,
+      "learning_rate": 7.25737877861016e-06,
+      "loss": 0.4438,
+      "step": 7956
+    },
+    {
+      "epoch": 1.774927503903636,
+      "grad_norm": 0.17779506742954254,
+      "learning_rate": 7.255115667677077e-06,
+      "loss": 0.4488,
+      "step": 7957
+    },
+    {
+      "epoch": 1.7751505688155254,
+      "grad_norm": 0.17906124889850616,
+      "learning_rate": 7.252852708772953e-06,
+      "loss": 0.4606,
+      "step": 7958
+    },
+    {
+      "epoch": 1.7753736337274146,
+      "grad_norm": 0.18236404657363892,
+      "learning_rate": 7.250589902023127e-06,
+      "loss": 0.452,
+      "step": 7959
+    },
+    {
+      "epoch": 1.775596698639304,
+      "grad_norm": 0.17182929813861847,
+      "learning_rate": 7.248327247552925e-06,
+      "loss": 0.4517,
+      "step": 7960
+    },
+    {
+      "epoch": 1.7758197635511934,
+      "grad_norm": 0.18052874505519867,
+      "learning_rate": 7.246064745487671e-06,
+      "loss": 0.4603,
+      "step": 7961
+    },
+    {
+      "epoch": 1.776042828463083,
+      "grad_norm": 0.18056614696979523,
+      "learning_rate": 7.243802395952673e-06,
+      "loss": 0.4641,
+      "step": 7962
+    },
+    {
+      "epoch": 1.7762658933749722,
+      "grad_norm": 0.17786100506782532,
+      "learning_rate": 7.241540199073235e-06,
+      "loss": 0.4782,
+      "step": 7963
+    },
+    {
+      "epoch": 1.7764889582868615,
+      "grad_norm": 0.17451272904872894,
+      "learning_rate": 7.239278154974654e-06,
+      "loss": 0.4126,
+      "step": 7964
+    },
+    {
+      "epoch": 1.7767120231987508,
+      "grad_norm": 0.1730719357728958,
+      "learning_rate": 7.237016263782213e-06,
+      "loss": 0.4368,
+      "step": 7965
+    },
+    {
+      "epoch": 1.77693508811064,
+      "grad_norm": 0.173664852976799,
+      "learning_rate": 7.234754525621192e-06,
+      "loss": 0.4113,
+      "step": 7966
+    },
+    {
+      "epoch": 1.7771581530225296,
+      "grad_norm": 0.17867377400398254,
+      "learning_rate": 7.232492940616858e-06,
+      "loss": 0.4555,
+      "step": 7967
+    },
+    {
+      "epoch": 1.777381217934419,
+      "grad_norm": 0.1810644119977951,
+      "learning_rate": 7.2302315088944755e-06,
+      "loss": 0.458,
+      "step": 7968
+    },
+    {
+      "epoch": 1.7776042828463083,
+      "grad_norm": 0.18765312433242798,
+      "learning_rate": 7.227970230579292e-06,
+      "loss": 0.4574,
+      "step": 7969
+    },
+    {
+      "epoch": 1.7778273477581976,
+      "grad_norm": 0.1742631047964096,
+      "learning_rate": 7.225709105796557e-06,
+      "loss": 0.4651,
+      "step": 7970
+    },
+    {
+      "epoch": 1.778050412670087,
+      "grad_norm": 0.22400738298892975,
+      "learning_rate": 7.223448134671499e-06,
+      "loss": 0.4561,
+      "step": 7971
+    },
+    {
+      "epoch": 1.7782734775819764,
+      "grad_norm": 0.18479710817337036,
+      "learning_rate": 7.221187317329353e-06,
+      "loss": 0.4447,
+      "step": 7972
+    },
+    {
+      "epoch": 1.7784965424938657,
+      "grad_norm": 0.17623472213745117,
+      "learning_rate": 7.218926653895329e-06,
+      "loss": 0.44,
+      "step": 7973
+    },
+    {
+      "epoch": 1.7787196074057552,
+      "grad_norm": 0.18266162276268005,
+      "learning_rate": 7.216666144494642e-06,
+      "loss": 0.4708,
+      "step": 7974
+    },
+    {
+      "epoch": 1.7789426723176445,
+      "grad_norm": 0.1808854639530182,
+      "learning_rate": 7.21440578925249e-06,
+      "loss": 0.4403,
+      "step": 7975
+    },
+    {
+      "epoch": 1.7791657372295338,
+      "grad_norm": 0.1751621812582016,
+      "learning_rate": 7.212145588294071e-06,
+      "loss": 0.4669,
+      "step": 7976
+    },
+    {
+      "epoch": 1.779388802141423,
+      "grad_norm": 0.18863114714622498,
+      "learning_rate": 7.2098855417445625e-06,
+      "loss": 0.4674,
+      "step": 7977
+    },
+    {
+      "epoch": 1.7796118670533125,
+      "grad_norm": 0.17852959036827087,
+      "learning_rate": 7.207625649729145e-06,
+      "loss": 0.4389,
+      "step": 7978
+    },
+    {
+      "epoch": 1.779834931965202,
+      "grad_norm": 0.18438269197940826,
+      "learning_rate": 7.205365912372982e-06,
+      "loss": 0.4467,
+      "step": 7979
+    },
+    {
+      "epoch": 1.7800579968770913,
+      "grad_norm": 0.179841086268425,
+      "learning_rate": 7.203106329801236e-06,
+      "loss": 0.4554,
+      "step": 7980
+    },
+    {
+      "epoch": 1.7802810617889806,
+      "grad_norm": 0.1838013380765915,
+      "learning_rate": 7.200846902139053e-06,
+      "loss": 0.4643,
+      "step": 7981
+    },
+    {
+      "epoch": 1.7805041267008699,
+      "grad_norm": 0.17783623933792114,
+      "learning_rate": 7.198587629511577e-06,
+      "loss": 0.4615,
+      "step": 7982
+    },
+    {
+      "epoch": 1.7807271916127592,
+      "grad_norm": 0.18178191781044006,
+      "learning_rate": 7.196328512043938e-06,
+      "loss": 0.4285,
+      "step": 7983
+    },
+    {
+      "epoch": 1.7809502565246487,
+      "grad_norm": 0.1845524162054062,
+      "learning_rate": 7.194069549861262e-06,
+      "loss": 0.4441,
+      "step": 7984
+    },
+    {
+      "epoch": 1.7811733214365382,
+      "grad_norm": 0.17717888951301575,
+      "learning_rate": 7.191810743088667e-06,
+      "loss": 0.465,
+      "step": 7985
+    },
+    {
+      "epoch": 1.7813963863484275,
+      "grad_norm": 0.1790652871131897,
+      "learning_rate": 7.189552091851254e-06,
+      "loss": 0.4652,
+      "step": 7986
+    },
+    {
+      "epoch": 1.7816194512603167,
+      "grad_norm": 0.17189590632915497,
+      "learning_rate": 7.187293596274128e-06,
+      "loss": 0.4241,
+      "step": 7987
+    },
+    {
+      "epoch": 1.781842516172206,
+      "grad_norm": 0.16961240768432617,
+      "learning_rate": 7.1850352564823715e-06,
+      "loss": 0.4425,
+      "step": 7988
+    },
+    {
+      "epoch": 1.7820655810840955,
+      "grad_norm": 0.17109279334545135,
+      "learning_rate": 7.182777072601072e-06,
+      "loss": 0.4479,
+      "step": 7989
+    },
+    {
+      "epoch": 1.7822886459959848,
+      "grad_norm": 0.17438296973705292,
+      "learning_rate": 7.180519044755298e-06,
+      "loss": 0.4439,
+      "step": 7990
+    },
+    {
+      "epoch": 1.7825117109078743,
+      "grad_norm": 0.17713730037212372,
+      "learning_rate": 7.178261173070116e-06,
+      "loss": 0.4643,
+      "step": 7991
+    },
+    {
+      "epoch": 1.7827347758197636,
+      "grad_norm": 0.18717728555202484,
+      "learning_rate": 7.176003457670577e-06,
+      "loss": 0.4673,
+      "step": 7992
+    },
+    {
+      "epoch": 1.7829578407316529,
+      "grad_norm": 0.18204233050346375,
+      "learning_rate": 7.173745898681732e-06,
+      "loss": 0.4605,
+      "step": 7993
+    },
+    {
+      "epoch": 1.7831809056435421,
+      "grad_norm": 0.18667717278003693,
+      "learning_rate": 7.171488496228617e-06,
+      "loss": 0.484,
+      "step": 7994
+    },
+    {
+      "epoch": 1.7834039705554317,
+      "grad_norm": 0.18115851283073425,
+      "learning_rate": 7.169231250436261e-06,
+      "loss": 0.4749,
+      "step": 7995
+    },
+    {
+      "epoch": 1.7836270354673212,
+      "grad_norm": 0.17957910895347595,
+      "learning_rate": 7.166974161429684e-06,
+      "loss": 0.4673,
+      "step": 7996
+    },
+    {
+      "epoch": 1.7838501003792104,
+      "grad_norm": 0.17228113114833832,
+      "learning_rate": 7.1647172293338995e-06,
+      "loss": 0.4658,
+      "step": 7997
+    },
+    {
+      "epoch": 1.7840731652910997,
+      "grad_norm": 0.17893186211585999,
+      "learning_rate": 7.162460454273909e-06,
+      "loss": 0.4639,
+      "step": 7998
+    },
+    {
+      "epoch": 1.784296230202989,
+      "grad_norm": 0.17566993832588196,
+      "learning_rate": 7.160203836374708e-06,
+      "loss": 0.4567,
+      "step": 7999
+    },
+    {
+      "epoch": 1.7845192951148783,
+      "grad_norm": 0.17004132270812988,
+      "learning_rate": 7.1579473757612795e-06,
+      "loss": 0.4419,
+      "step": 8000
+    },
+    {
+      "epoch": 1.7847423600267678,
+      "grad_norm": 0.17692558467388153,
+      "learning_rate": 7.155691072558604e-06,
+      "loss": 0.4546,
+      "step": 8001
+    },
+    {
+      "epoch": 1.7849654249386573,
+      "grad_norm": 0.1739339381456375,
+      "learning_rate": 7.153434926891648e-06,
+      "loss": 0.4636,
+      "step": 8002
+    },
+    {
+      "epoch": 1.7851884898505466,
+      "grad_norm": 0.17398999631404877,
+      "learning_rate": 7.151178938885373e-06,
+      "loss": 0.4291,
+      "step": 8003
+    },
+    {
+      "epoch": 1.7854115547624358,
+      "grad_norm": 0.1721528023481369,
+      "learning_rate": 7.148923108664725e-06,
+      "loss": 0.4325,
+      "step": 8004
+    },
+    {
+      "epoch": 1.7856346196743251,
+      "grad_norm": 0.18906241655349731,
+      "learning_rate": 7.146667436354651e-06,
+      "loss": 0.4211,
+      "step": 8005
+    },
+    {
+      "epoch": 1.7858576845862146,
+      "grad_norm": 0.18445979058742523,
+      "learning_rate": 7.144411922080082e-06,
+      "loss": 0.4536,
+      "step": 8006
+    },
+    {
+      "epoch": 1.786080749498104,
+      "grad_norm": 0.18223892152309418,
+      "learning_rate": 7.142156565965943e-06,
+      "loss": 0.4554,
+      "step": 8007
+    },
+    {
+      "epoch": 1.7863038144099934,
+      "grad_norm": 0.18098188936710358,
+      "learning_rate": 7.139901368137152e-06,
+      "loss": 0.4453,
+      "step": 8008
+    },
+    {
+      "epoch": 1.7865268793218827,
+      "grad_norm": 0.17738774418830872,
+      "learning_rate": 7.1376463287186105e-06,
+      "loss": 0.4302,
+      "step": 8009
+    },
+    {
+      "epoch": 1.786749944233772,
+      "grad_norm": 0.18761973083019257,
+      "learning_rate": 7.135391447835224e-06,
+      "loss": 0.4673,
+      "step": 8010
+    },
+    {
+      "epoch": 1.7869730091456613,
+      "grad_norm": 0.17969244718551636,
+      "learning_rate": 7.133136725611876e-06,
+      "loss": 0.4278,
+      "step": 8011
+    },
+    {
+      "epoch": 1.7871960740575508,
+      "grad_norm": 0.19217950105667114,
+      "learning_rate": 7.13088216217345e-06,
+      "loss": 0.4448,
+      "step": 8012
+    },
+    {
+      "epoch": 1.7874191389694403,
+      "grad_norm": 0.1748211681842804,
+      "learning_rate": 7.128627757644816e-06,
+      "loss": 0.4345,
+      "step": 8013
+    },
+    {
+      "epoch": 1.7876422038813296,
+      "grad_norm": 0.1840067058801651,
+      "learning_rate": 7.126373512150842e-06,
+      "loss": 0.4478,
+      "step": 8014
+    },
+    {
+      "epoch": 1.7878652687932188,
+      "grad_norm": 0.1767268180847168,
+      "learning_rate": 7.1241194258163735e-06,
+      "loss": 0.4493,
+      "step": 8015
+    },
+    {
+      "epoch": 1.7880883337051081,
+      "grad_norm": 0.17844226956367493,
+      "learning_rate": 7.121865498766265e-06,
+      "loss": 0.4131,
+      "step": 8016
+    },
+    {
+      "epoch": 1.7883113986169974,
+      "grad_norm": 0.22129550576210022,
+      "learning_rate": 7.119611731125349e-06,
+      "loss": 0.4441,
+      "step": 8017
+    },
+    {
+      "epoch": 1.788534463528887,
+      "grad_norm": 0.18443824350833893,
+      "learning_rate": 7.117358123018454e-06,
+      "loss": 0.4344,
+      "step": 8018
+    },
+    {
+      "epoch": 1.7887575284407764,
+      "grad_norm": 0.17921538650989532,
+      "learning_rate": 7.115104674570397e-06,
+      "loss": 0.4646,
+      "step": 8019
+    },
+    {
+      "epoch": 1.7889805933526657,
+      "grad_norm": 0.1854918748140335,
+      "learning_rate": 7.11285138590599e-06,
+      "loss": 0.4692,
+      "step": 8020
+    },
+    {
+      "epoch": 1.789203658264555,
+      "grad_norm": 0.1828143298625946,
+      "learning_rate": 7.110598257150034e-06,
+      "loss": 0.444,
+      "step": 8021
+    },
+    {
+      "epoch": 1.7894267231764442,
+      "grad_norm": 0.18005596101284027,
+      "learning_rate": 7.108345288427324e-06,
+      "loss": 0.4447,
+      "step": 8022
+    },
+    {
+      "epoch": 1.7896497880883337,
+      "grad_norm": 0.18779170513153076,
+      "learning_rate": 7.106092479862639e-06,
+      "loss": 0.4516,
+      "step": 8023
+    },
+    {
+      "epoch": 1.789872853000223,
+      "grad_norm": 0.17747437953948975,
+      "learning_rate": 7.103839831580754e-06,
+      "loss": 0.4436,
+      "step": 8024
+    },
+    {
+      "epoch": 1.7900959179121125,
+      "grad_norm": 0.1890641450881958,
+      "learning_rate": 7.101587343706441e-06,
+      "loss": 0.4624,
+      "step": 8025
+    },
+    {
+      "epoch": 1.7903189828240018,
+      "grad_norm": 0.18202100694179535,
+      "learning_rate": 7.099335016364449e-06,
+      "loss": 0.4821,
+      "step": 8026
+    },
+    {
+      "epoch": 1.790542047735891,
+      "grad_norm": 0.17150594294071198,
+      "learning_rate": 7.097082849679533e-06,
+      "loss": 0.444,
+      "step": 8027
+    },
+    {
+      "epoch": 1.7907651126477804,
+      "grad_norm": 0.18223629891872406,
+      "learning_rate": 7.0948308437764255e-06,
+      "loss": 0.4514,
+      "step": 8028
+    },
+    {
+      "epoch": 1.7909881775596699,
+      "grad_norm": 0.1873582899570465,
+      "learning_rate": 7.092578998779864e-06,
+      "loss": 0.4614,
+      "step": 8029
+    },
+    {
+      "epoch": 1.7912112424715594,
+      "grad_norm": 0.17026644945144653,
+      "learning_rate": 7.090327314814562e-06,
+      "loss": 0.4616,
+      "step": 8030
+    },
+    {
+      "epoch": 1.7914343073834487,
+      "grad_norm": 0.1856604516506195,
+      "learning_rate": 7.088075792005239e-06,
+      "loss": 0.4666,
+      "step": 8031
+    },
+    {
+      "epoch": 1.791657372295338,
+      "grad_norm": 0.18316306173801422,
+      "learning_rate": 7.085824430476593e-06,
+      "loss": 0.4768,
+      "step": 8032
+    },
+    {
+      "epoch": 1.7918804372072272,
+      "grad_norm": 0.16916830837726593,
+      "learning_rate": 7.083573230353324e-06,
+      "loss": 0.4437,
+      "step": 8033
+    },
+    {
+      "epoch": 1.7921035021191165,
+      "grad_norm": 0.17978118360042572,
+      "learning_rate": 7.0813221917601104e-06,
+      "loss": 0.4525,
+      "step": 8034
+    },
+    {
+      "epoch": 1.792326567031006,
+      "grad_norm": 0.19063124060630798,
+      "learning_rate": 7.079071314821637e-06,
+      "loss": 0.4366,
+      "step": 8035
+    },
+    {
+      "epoch": 1.7925496319428955,
+      "grad_norm": 0.1843474954366684,
+      "learning_rate": 7.076820599662564e-06,
+      "loss": 0.4635,
+      "step": 8036
+    },
+    {
+      "epoch": 1.7927726968547848,
+      "grad_norm": 0.17682790756225586,
+      "learning_rate": 7.074570046407556e-06,
+      "loss": 0.4645,
+      "step": 8037
+    },
+    {
+      "epoch": 1.792995761766674,
+      "grad_norm": 0.19309797883033752,
+      "learning_rate": 7.072319655181263e-06,
+      "loss": 0.4527,
+      "step": 8038
+    },
+    {
+      "epoch": 1.7932188266785634,
+      "grad_norm": 0.17107698321342468,
+      "learning_rate": 7.07006942610832e-06,
+      "loss": 0.4349,
+      "step": 8039
+    },
+    {
+      "epoch": 1.7934418915904529,
+      "grad_norm": 0.18476001918315887,
+      "learning_rate": 7.067819359313364e-06,
+      "loss": 0.4645,
+      "step": 8040
+    },
+    {
+      "epoch": 1.7936649565023421,
+      "grad_norm": 0.17997145652770996,
+      "learning_rate": 7.065569454921013e-06,
+      "loss": 0.4634,
+      "step": 8041
+    },
+    {
+      "epoch": 1.7938880214142316,
+      "grad_norm": 0.17692801356315613,
+      "learning_rate": 7.063319713055887e-06,
+      "loss": 0.4336,
+      "step": 8042
+    },
+    {
+      "epoch": 1.794111086326121,
+      "grad_norm": 0.17240415513515472,
+      "learning_rate": 7.061070133842584e-06,
+      "loss": 0.4452,
+      "step": 8043
+    },
+    {
+      "epoch": 1.7943341512380102,
+      "grad_norm": 0.16578581929206848,
+      "learning_rate": 7.058820717405707e-06,
+      "loss": 0.4297,
+      "step": 8044
+    },
+    {
+      "epoch": 1.7945572161498995,
+      "grad_norm": 0.1810625195503235,
+      "learning_rate": 7.056571463869838e-06,
+      "loss": 0.4628,
+      "step": 8045
+    },
+    {
+      "epoch": 1.794780281061789,
+      "grad_norm": 0.1759740114212036,
+      "learning_rate": 7.054322373359557e-06,
+      "loss": 0.4383,
+      "step": 8046
+    },
+    {
+      "epoch": 1.7950033459736785,
+      "grad_norm": 0.18540038168430328,
+      "learning_rate": 7.052073445999431e-06,
+      "loss": 0.4655,
+      "step": 8047
+    },
+    {
+      "epoch": 1.7952264108855678,
+      "grad_norm": 0.17017890512943268,
+      "learning_rate": 7.04982468191402e-06,
+      "loss": 0.4216,
+      "step": 8048
+    },
+    {
+      "epoch": 1.795449475797457,
+      "grad_norm": 0.17637605965137482,
+      "learning_rate": 7.047576081227878e-06,
+      "loss": 0.463,
+      "step": 8049
+    },
+    {
+      "epoch": 1.7956725407093463,
+      "grad_norm": 0.1885131597518921,
+      "learning_rate": 7.045327644065542e-06,
+      "loss": 0.4563,
+      "step": 8050
+    },
+    {
+      "epoch": 1.7958956056212356,
+      "grad_norm": 0.18475908041000366,
+      "learning_rate": 7.043079370551549e-06,
+      "loss": 0.4511,
+      "step": 8051
+    },
+    {
+      "epoch": 1.7961186705331251,
+      "grad_norm": 0.19659362733364105,
+      "learning_rate": 7.0408312608104164e-06,
+      "loss": 0.4794,
+      "step": 8052
+    },
+    {
+      "epoch": 1.7963417354450146,
+      "grad_norm": 0.1826995611190796,
+      "learning_rate": 7.038583314966666e-06,
+      "loss": 0.4628,
+      "step": 8053
+    },
+    {
+      "epoch": 1.796564800356904,
+      "grad_norm": 0.1809930056333542,
+      "learning_rate": 7.036335533144798e-06,
+      "loss": 0.4272,
+      "step": 8054
+    },
+    {
+      "epoch": 1.7967878652687932,
+      "grad_norm": 0.177769273519516,
+      "learning_rate": 7.034087915469312e-06,
+      "loss": 0.4509,
+      "step": 8055
+    },
+    {
+      "epoch": 1.7970109301806825,
+      "grad_norm": 0.17363698780536652,
+      "learning_rate": 7.031840462064691e-06,
+      "loss": 0.4564,
+      "step": 8056
+    },
+    {
+      "epoch": 1.797233995092572,
+      "grad_norm": 0.17677956819534302,
+      "learning_rate": 7.029593173055418e-06,
+      "loss": 0.4327,
+      "step": 8057
+    },
+    {
+      "epoch": 1.7974570600044613,
+      "grad_norm": 0.1727108657360077,
+      "learning_rate": 7.0273460485659576e-06,
+      "loss": 0.4229,
+      "step": 8058
+    },
+    {
+      "epoch": 1.7976801249163508,
+      "grad_norm": 0.1822299361228943,
+      "learning_rate": 7.025099088720773e-06,
+      "loss": 0.4397,
+      "step": 8059
+    },
+    {
+      "epoch": 1.79790318982824,
+      "grad_norm": 0.18957071006298065,
+      "learning_rate": 7.022852293644313e-06,
+      "loss": 0.4679,
+      "step": 8060
+    },
+    {
+      "epoch": 1.7981262547401293,
+      "grad_norm": 0.17512580752372742,
+      "learning_rate": 7.020605663461022e-06,
+      "loss": 0.4267,
+      "step": 8061
+    },
+    {
+      "epoch": 1.7983493196520186,
+      "grad_norm": 0.1905566304922104,
+      "learning_rate": 7.018359198295328e-06,
+      "loss": 0.4542,
+      "step": 8062
+    },
+    {
+      "epoch": 1.798572384563908,
+      "grad_norm": 0.1788644641637802,
+      "learning_rate": 7.016112898271658e-06,
+      "loss": 0.4609,
+      "step": 8063
+    },
+    {
+      "epoch": 1.7987954494757976,
+      "grad_norm": 0.1741524040699005,
+      "learning_rate": 7.013866763514426e-06,
+      "loss": 0.439,
+      "step": 8064
+    },
+    {
+      "epoch": 1.799018514387687,
+      "grad_norm": 0.169100821018219,
+      "learning_rate": 7.011620794148035e-06,
+      "loss": 0.4511,
+      "step": 8065
+    },
+    {
+      "epoch": 1.7992415792995762,
+      "grad_norm": 0.1780511438846588,
+      "learning_rate": 7.009374990296882e-06,
+      "loss": 0.4391,
+      "step": 8066
+    },
+    {
+      "epoch": 1.7994646442114655,
+      "grad_norm": 0.17674995958805084,
+      "learning_rate": 7.0071293520853555e-06,
+      "loss": 0.4662,
+      "step": 8067
+    },
+    {
+      "epoch": 1.7996877091233547,
+      "grad_norm": 0.17258092761039734,
+      "learning_rate": 7.004883879637829e-06,
+      "loss": 0.4235,
+      "step": 8068
+    },
+    {
+      "epoch": 1.7999107740352442,
+      "grad_norm": 0.17394836246967316,
+      "learning_rate": 7.002638573078673e-06,
+      "loss": 0.4728,
+      "step": 8069
+    },
+    {
+      "epoch": 1.8001338389471337,
+      "grad_norm": 0.17654062807559967,
+      "learning_rate": 7.000393432532249e-06,
+      "loss": 0.4386,
+      "step": 8070
+    },
+    {
+      "epoch": 1.800356903859023,
+      "grad_norm": 0.17876291275024414,
+      "learning_rate": 6.998148458122903e-06,
+      "loss": 0.4621,
+      "step": 8071
+    },
+    {
+      "epoch": 1.8005799687709123,
+      "grad_norm": 0.18045787513256073,
+      "learning_rate": 6.99590364997498e-06,
+      "loss": 0.4521,
+      "step": 8072
+    },
+    {
+      "epoch": 1.8008030336828016,
+      "grad_norm": 0.1908344030380249,
+      "learning_rate": 6.993659008212808e-06,
+      "loss": 0.4671,
+      "step": 8073
+    },
+    {
+      "epoch": 1.801026098594691,
+      "grad_norm": 0.18054337799549103,
+      "learning_rate": 6.991414532960712e-06,
+      "loss": 0.4476,
+      "step": 8074
+    },
+    {
+      "epoch": 1.8012491635065804,
+      "grad_norm": 0.18316704034805298,
+      "learning_rate": 6.989170224343003e-06,
+      "loss": 0.4653,
+      "step": 8075
+    },
+    {
+      "epoch": 1.8014722284184699,
+      "grad_norm": 0.17818012833595276,
+      "learning_rate": 6.986926082483986e-06,
+      "loss": 0.4487,
+      "step": 8076
+    },
+    {
+      "epoch": 1.8016952933303592,
+      "grad_norm": 0.17305941879749298,
+      "learning_rate": 6.9846821075079544e-06,
+      "loss": 0.4331,
+      "step": 8077
+    },
+    {
+      "epoch": 1.8019183582422484,
+      "grad_norm": 0.17837919294834137,
+      "learning_rate": 6.982438299539197e-06,
+      "loss": 0.433,
+      "step": 8078
+    },
+    {
+      "epoch": 1.8021414231541377,
+      "grad_norm": 0.198192298412323,
+      "learning_rate": 6.980194658701985e-06,
+      "loss": 0.4623,
+      "step": 8079
+    },
+    {
+      "epoch": 1.8023644880660272,
+      "grad_norm": 0.18458157777786255,
+      "learning_rate": 6.977951185120591e-06,
+      "loss": 0.4642,
+      "step": 8080
+    },
+    {
+      "epoch": 1.8025875529779167,
+      "grad_norm": 0.17324891686439514,
+      "learning_rate": 6.975707878919268e-06,
+      "loss": 0.4549,
+      "step": 8081
+    },
+    {
+      "epoch": 1.802810617889806,
+      "grad_norm": 0.1909075230360031,
+      "learning_rate": 6.973464740222268e-06,
+      "loss": 0.4647,
+      "step": 8082
+    },
+    {
+      "epoch": 1.8030336828016953,
+      "grad_norm": 0.17792390286922455,
+      "learning_rate": 6.971221769153826e-06,
+      "loss": 0.4563,
+      "step": 8083
+    },
+    {
+      "epoch": 1.8032567477135846,
+      "grad_norm": 0.17481090128421783,
+      "learning_rate": 6.968978965838176e-06,
+      "loss": 0.4537,
+      "step": 8084
+    },
+    {
+      "epoch": 1.8034798126254739,
+      "grad_norm": 0.19158580899238586,
+      "learning_rate": 6.966736330399535e-06,
+      "loss": 0.4493,
+      "step": 8085
+    },
+    {
+      "epoch": 1.8037028775373634,
+      "grad_norm": 0.18006612360477448,
+      "learning_rate": 6.964493862962118e-06,
+      "loss": 0.4838,
+      "step": 8086
+    },
+    {
+      "epoch": 1.8039259424492529,
+      "grad_norm": 0.18528711795806885,
+      "learning_rate": 6.962251563650124e-06,
+      "loss": 0.4397,
+      "step": 8087
+    },
+    {
+      "epoch": 1.8041490073611421,
+      "grad_norm": 0.17403841018676758,
+      "learning_rate": 6.960009432587747e-06,
+      "loss": 0.4219,
+      "step": 8088
+    },
+    {
+      "epoch": 1.8043720722730314,
+      "grad_norm": 0.17622390389442444,
+      "learning_rate": 6.957767469899168e-06,
+      "loss": 0.4645,
+      "step": 8089
+    },
+    {
+      "epoch": 1.8045951371849207,
+      "grad_norm": 0.17650552093982697,
+      "learning_rate": 6.955525675708564e-06,
+      "loss": 0.4427,
+      "step": 8090
+    },
+    {
+      "epoch": 1.8048182020968102,
+      "grad_norm": 0.1795479953289032,
+      "learning_rate": 6.953284050140101e-06,
+      "loss": 0.4268,
+      "step": 8091
+    },
+    {
+      "epoch": 1.8050412670086995,
+      "grad_norm": 0.18363459408283234,
+      "learning_rate": 6.951042593317929e-06,
+      "loss": 0.4403,
+      "step": 8092
+    },
+    {
+      "epoch": 1.805264331920589,
+      "grad_norm": 0.18579071760177612,
+      "learning_rate": 6.948801305366199e-06,
+      "loss": 0.4556,
+      "step": 8093
+    },
+    {
+      "epoch": 1.8054873968324783,
+      "grad_norm": 0.1805351972579956,
+      "learning_rate": 6.946560186409042e-06,
+      "loss": 0.4356,
+      "step": 8094
+    },
+    {
+      "epoch": 1.8057104617443676,
+      "grad_norm": 0.18544495105743408,
+      "learning_rate": 6.944319236570593e-06,
+      "loss": 0.46,
+      "step": 8095
+    },
+    {
+      "epoch": 1.8059335266562568,
+      "grad_norm": 0.1721157431602478,
+      "learning_rate": 6.942078455974963e-06,
+      "loss": 0.4562,
+      "step": 8096
+    },
+    {
+      "epoch": 1.8061565915681463,
+      "grad_norm": 0.1732361763715744,
+      "learning_rate": 6.9398378447462645e-06,
+      "loss": 0.4537,
+      "step": 8097
+    },
+    {
+      "epoch": 1.8063796564800358,
+      "grad_norm": 0.1871383637189865,
+      "learning_rate": 6.937597403008594e-06,
+      "loss": 0.4808,
+      "step": 8098
+    },
+    {
+      "epoch": 1.8066027213919251,
+      "grad_norm": 0.17318753898143768,
+      "learning_rate": 6.9353571308860446e-06,
+      "loss": 0.4451,
+      "step": 8099
+    },
+    {
+      "epoch": 1.8068257863038144,
+      "grad_norm": 0.17707256972789764,
+      "learning_rate": 6.933117028502693e-06,
+      "loss": 0.4374,
+      "step": 8100
+    },
+    {
+      "epoch": 1.8070488512157037,
+      "grad_norm": 0.18298032879829407,
+      "learning_rate": 6.930877095982613e-06,
+      "loss": 0.4588,
+      "step": 8101
+    },
+    {
+      "epoch": 1.807271916127593,
+      "grad_norm": 0.18046192824840546,
+      "learning_rate": 6.928637333449863e-06,
+      "loss": 0.4731,
+      "step": 8102
+    },
+    {
+      "epoch": 1.8074949810394825,
+      "grad_norm": 0.16831883788108826,
+      "learning_rate": 6.9263977410285014e-06,
+      "loss": 0.4308,
+      "step": 8103
+    },
+    {
+      "epoch": 1.807718045951372,
+      "grad_norm": 0.17589864134788513,
+      "learning_rate": 6.924158318842562e-06,
+      "loss": 0.4372,
+      "step": 8104
+    },
+    {
+      "epoch": 1.8079411108632613,
+      "grad_norm": 0.1834905594587326,
+      "learning_rate": 6.9219190670160856e-06,
+      "loss": 0.477,
+      "step": 8105
+    },
+    {
+      "epoch": 1.8081641757751505,
+      "grad_norm": 0.18303707242012024,
+      "learning_rate": 6.919679985673091e-06,
+      "loss": 0.4603,
+      "step": 8106
+    },
+    {
+      "epoch": 1.8083872406870398,
+      "grad_norm": 0.1804521679878235,
+      "learning_rate": 6.917441074937597e-06,
+      "loss": 0.4504,
+      "step": 8107
+    },
+    {
+      "epoch": 1.8086103055989293,
+      "grad_norm": 0.19304148852825165,
+      "learning_rate": 6.915202334933604e-06,
+      "loss": 0.4768,
+      "step": 8108
+    },
+    {
+      "epoch": 1.8088333705108186,
+      "grad_norm": 0.1749470978975296,
+      "learning_rate": 6.912963765785111e-06,
+      "loss": 0.449,
+      "step": 8109
+    },
+    {
+      "epoch": 1.809056435422708,
+      "grad_norm": 0.18221113085746765,
+      "learning_rate": 6.910725367616102e-06,
+      "loss": 0.454,
+      "step": 8110
+    },
+    {
+      "epoch": 1.8092795003345974,
+      "grad_norm": 0.1730881929397583,
+      "learning_rate": 6.908487140550555e-06,
+      "loss": 0.4506,
+      "step": 8111
+    },
+    {
+      "epoch": 1.8095025652464867,
+      "grad_norm": 0.17381657660007477,
+      "learning_rate": 6.9062490847124375e-06,
+      "loss": 0.4499,
+      "step": 8112
+    },
+    {
+      "epoch": 1.809725630158376,
+      "grad_norm": 0.17647477984428406,
+      "learning_rate": 6.904011200225704e-06,
+      "loss": 0.4394,
+      "step": 8113
+    },
+    {
+      "epoch": 1.8099486950702655,
+      "grad_norm": 0.17345494031906128,
+      "learning_rate": 6.901773487214308e-06,
+      "loss": 0.463,
+      "step": 8114
+    },
+    {
+      "epoch": 1.810171759982155,
+      "grad_norm": 0.178610160946846,
+      "learning_rate": 6.899535945802182e-06,
+      "loss": 0.4519,
+      "step": 8115
+    },
+    {
+      "epoch": 1.8103948248940442,
+      "grad_norm": 0.17938841879367828,
+      "learning_rate": 6.89729857611326e-06,
+      "loss": 0.4326,
+      "step": 8116
+    },
+    {
+      "epoch": 1.8106178898059335,
+      "grad_norm": 0.19250169396400452,
+      "learning_rate": 6.895061378271457e-06,
+      "loss": 0.4666,
+      "step": 8117
+    },
+    {
+      "epoch": 1.8108409547178228,
+      "grad_norm": 0.17775999009609222,
+      "learning_rate": 6.892824352400689e-06,
+      "loss": 0.4465,
+      "step": 8118
+    },
+    {
+      "epoch": 1.811064019629712,
+      "grad_norm": 0.1746588945388794,
+      "learning_rate": 6.890587498624854e-06,
+      "loss": 0.4471,
+      "step": 8119
+    },
+    {
+      "epoch": 1.8112870845416016,
+      "grad_norm": 0.2063416689634323,
+      "learning_rate": 6.88835081706784e-06,
+      "loss": 0.4341,
+      "step": 8120
+    },
+    {
+      "epoch": 1.811510149453491,
+      "grad_norm": 0.1776486337184906,
+      "learning_rate": 6.8861143078535334e-06,
+      "loss": 0.4288,
+      "step": 8121
+    },
+    {
+      "epoch": 1.8117332143653804,
+      "grad_norm": 0.1845918595790863,
+      "learning_rate": 6.883877971105799e-06,
+      "loss": 0.4622,
+      "step": 8122
+    },
+    {
+      "epoch": 1.8119562792772697,
+      "grad_norm": 0.174449160695076,
+      "learning_rate": 6.881641806948507e-06,
+      "loss": 0.4518,
+      "step": 8123
+    },
+    {
+      "epoch": 1.812179344189159,
+      "grad_norm": 0.16784648597240448,
+      "learning_rate": 6.879405815505504e-06,
+      "loss": 0.4349,
+      "step": 8124
+    },
+    {
+      "epoch": 1.8124024091010484,
+      "grad_norm": 0.19099272787570953,
+      "learning_rate": 6.877169996900639e-06,
+      "loss": 0.4694,
+      "step": 8125
+    },
+    {
+      "epoch": 1.8126254740129377,
+      "grad_norm": 0.17356911301612854,
+      "learning_rate": 6.874934351257742e-06,
+      "loss": 0.4486,
+      "step": 8126
+    },
+    {
+      "epoch": 1.8128485389248272,
+      "grad_norm": 0.1857564002275467,
+      "learning_rate": 6.872698878700639e-06,
+      "loss": 0.4687,
+      "step": 8127
+    },
+    {
+      "epoch": 1.8130716038367165,
+      "grad_norm": 0.18476644158363342,
+      "learning_rate": 6.870463579353141e-06,
+      "loss": 0.4673,
+      "step": 8128
+    },
+    {
+      "epoch": 1.8132946687486058,
+      "grad_norm": 0.17950499057769775,
+      "learning_rate": 6.868228453339058e-06,
+      "loss": 0.439,
+      "step": 8129
+    },
+    {
+      "epoch": 1.813517733660495,
+      "grad_norm": 0.17879453301429749,
+      "learning_rate": 6.865993500782182e-06,
+      "loss": 0.461,
+      "step": 8130
+    },
+    {
+      "epoch": 1.8137407985723846,
+      "grad_norm": 0.1743541806936264,
+      "learning_rate": 6.8637587218063e-06,
+      "loss": 0.4458,
+      "step": 8131
+    },
+    {
+      "epoch": 1.813963863484274,
+      "grad_norm": 0.18763816356658936,
+      "learning_rate": 6.861524116535187e-06,
+      "loss": 0.4482,
+      "step": 8132
+    },
+    {
+      "epoch": 1.8141869283961634,
+      "grad_norm": 0.17765846848487854,
+      "learning_rate": 6.85928968509261e-06,
+      "loss": 0.4387,
+      "step": 8133
+    },
+    {
+      "epoch": 1.8144099933080526,
+      "grad_norm": 0.1837596446275711,
+      "learning_rate": 6.857055427602327e-06,
+      "loss": 0.4789,
+      "step": 8134
+    },
+    {
+      "epoch": 1.814633058219942,
+      "grad_norm": 0.18249496817588806,
+      "learning_rate": 6.854821344188083e-06,
+      "loss": 0.4511,
+      "step": 8135
+    },
+    {
+      "epoch": 1.8148561231318312,
+      "grad_norm": 0.17417484521865845,
+      "learning_rate": 6.852587434973619e-06,
+      "loss": 0.4265,
+      "step": 8136
+    },
+    {
+      "epoch": 1.8150791880437207,
+      "grad_norm": 0.17669354379177094,
+      "learning_rate": 6.850353700082659e-06,
+      "loss": 0.4448,
+      "step": 8137
+    },
+    {
+      "epoch": 1.8153022529556102,
+      "grad_norm": 0.17180798947811127,
+      "learning_rate": 6.8481201396389254e-06,
+      "loss": 0.4516,
+      "step": 8138
+    },
+    {
+      "epoch": 1.8155253178674995,
+      "grad_norm": 0.17706137895584106,
+      "learning_rate": 6.845886753766122e-06,
+      "loss": 0.4411,
+      "step": 8139
+    },
+    {
+      "epoch": 1.8157483827793888,
+      "grad_norm": 0.18320493400096893,
+      "learning_rate": 6.843653542587952e-06,
+      "loss": 0.4633,
+      "step": 8140
+    },
+    {
+      "epoch": 1.815971447691278,
+      "grad_norm": 0.1703965961933136,
+      "learning_rate": 6.8414205062281025e-06,
+      "loss": 0.435,
+      "step": 8141
+    },
+    {
+      "epoch": 1.8161945126031676,
+      "grad_norm": 0.2011895328760147,
+      "learning_rate": 6.839187644810255e-06,
+      "loss": 0.4653,
+      "step": 8142
+    },
+    {
+      "epoch": 1.8164175775150568,
+      "grad_norm": 0.17818677425384521,
+      "learning_rate": 6.836954958458075e-06,
+      "loss": 0.4564,
+      "step": 8143
+    },
+    {
+      "epoch": 1.8166406424269463,
+      "grad_norm": 0.17748653888702393,
+      "learning_rate": 6.834722447295228e-06,
+      "loss": 0.4357,
+      "step": 8144
+    },
+    {
+      "epoch": 1.8168637073388356,
+      "grad_norm": 0.1691962629556656,
+      "learning_rate": 6.832490111445361e-06,
+      "loss": 0.4411,
+      "step": 8145
+    },
+    {
+      "epoch": 1.817086772250725,
+      "grad_norm": 0.1772109568119049,
+      "learning_rate": 6.830257951032118e-06,
+      "loss": 0.4365,
+      "step": 8146
+    },
+    {
+      "epoch": 1.8173098371626142,
+      "grad_norm": 0.17656628787517548,
+      "learning_rate": 6.828025966179126e-06,
+      "loss": 0.4387,
+      "step": 8147
+    },
+    {
+      "epoch": 1.8175329020745037,
+      "grad_norm": 0.19161885976791382,
+      "learning_rate": 6.825794157010011e-06,
+      "loss": 0.4518,
+      "step": 8148
+    },
+    {
+      "epoch": 1.8177559669863932,
+      "grad_norm": 0.18089298903942108,
+      "learning_rate": 6.823562523648379e-06,
+      "loss": 0.4414,
+      "step": 8149
+    },
+    {
+      "epoch": 1.8179790318982825,
+      "grad_norm": 0.1826842874288559,
+      "learning_rate": 6.8213310662178375e-06,
+      "loss": 0.4651,
+      "step": 8150
+    },
+    {
+      "epoch": 1.8182020968101718,
+      "grad_norm": 0.1827784776687622,
+      "learning_rate": 6.819099784841974e-06,
+      "loss": 0.4793,
+      "step": 8151
+    },
+    {
+      "epoch": 1.818425161722061,
+      "grad_norm": 0.193526491522789,
+      "learning_rate": 6.816868679644375e-06,
+      "loss": 0.4758,
+      "step": 8152
+    },
+    {
+      "epoch": 1.8186482266339503,
+      "grad_norm": 0.18154491484165192,
+      "learning_rate": 6.814637750748609e-06,
+      "loss": 0.4674,
+      "step": 8153
+    },
+    {
+      "epoch": 1.8188712915458398,
+      "grad_norm": 0.18000057339668274,
+      "learning_rate": 6.812406998278242e-06,
+      "loss": 0.4513,
+      "step": 8154
+    },
+    {
+      "epoch": 1.8190943564577293,
+      "grad_norm": 0.17618589103221893,
+      "learning_rate": 6.8101764223568286e-06,
+      "loss": 0.4684,
+      "step": 8155
+    },
+    {
+      "epoch": 1.8193174213696186,
+      "grad_norm": 0.1722210943698883,
+      "learning_rate": 6.807946023107906e-06,
+      "loss": 0.4327,
+      "step": 8156
+    },
+    {
+      "epoch": 1.8195404862815079,
+      "grad_norm": 0.18736113607883453,
+      "learning_rate": 6.805715800655015e-06,
+      "loss": 0.4647,
+      "step": 8157
+    },
+    {
+      "epoch": 1.8197635511933972,
+      "grad_norm": 0.17818038165569305,
+      "learning_rate": 6.803485755121672e-06,
+      "loss": 0.4543,
+      "step": 8158
+    },
+    {
+      "epoch": 1.8199866161052867,
+      "grad_norm": 0.18235214054584503,
+      "learning_rate": 6.801255886631399e-06,
+      "loss": 0.4679,
+      "step": 8159
+    },
+    {
+      "epoch": 1.820209681017176,
+      "grad_norm": 0.1716981828212738,
+      "learning_rate": 6.799026195307693e-06,
+      "loss": 0.4515,
+      "step": 8160
+    },
+    {
+      "epoch": 1.8204327459290655,
+      "grad_norm": 0.17725342512130737,
+      "learning_rate": 6.796796681274054e-06,
+      "loss": 0.4501,
+      "step": 8161
+    },
+    {
+      "epoch": 1.8206558108409547,
+      "grad_norm": 0.18533247709274292,
+      "learning_rate": 6.79456734465396e-06,
+      "loss": 0.4674,
+      "step": 8162
+    },
+    {
+      "epoch": 1.820878875752844,
+      "grad_norm": 0.17556942999362946,
+      "learning_rate": 6.792338185570893e-06,
+      "loss": 0.4552,
+      "step": 8163
+    },
+    {
+      "epoch": 1.8211019406647333,
+      "grad_norm": 0.1827983409166336,
+      "learning_rate": 6.790109204148311e-06,
+      "loss": 0.435,
+      "step": 8164
+    },
+    {
+      "epoch": 1.8213250055766228,
+      "grad_norm": 0.18367482721805573,
+      "learning_rate": 6.787880400509674e-06,
+      "loss": 0.4447,
+      "step": 8165
+    },
+    {
+      "epoch": 1.8215480704885123,
+      "grad_norm": 0.17707587778568268,
+      "learning_rate": 6.785651774778425e-06,
+      "loss": 0.4555,
+      "step": 8166
+    },
+    {
+      "epoch": 1.8217711354004016,
+      "grad_norm": 0.17791809141635895,
+      "learning_rate": 6.7834233270780005e-06,
+      "loss": 0.4766,
+      "step": 8167
+    },
+    {
+      "epoch": 1.8219942003122909,
+      "grad_norm": 0.1814098060131073,
+      "learning_rate": 6.781195057531824e-06,
+      "loss": 0.4634,
+      "step": 8168
+    },
+    {
+      "epoch": 1.8222172652241801,
+      "grad_norm": 0.1738249808549881,
+      "learning_rate": 6.778966966263314e-06,
+      "loss": 0.4515,
+      "step": 8169
+    },
+    {
+      "epoch": 1.8224403301360694,
+      "grad_norm": 0.17479370534420013,
+      "learning_rate": 6.7767390533958735e-06,
+      "loss": 0.4376,
+      "step": 8170
+    },
+    {
+      "epoch": 1.822663395047959,
+      "grad_norm": 0.17913582921028137,
+      "learning_rate": 6.774511319052899e-06,
+      "loss": 0.4571,
+      "step": 8171
+    },
+    {
+      "epoch": 1.8228864599598484,
+      "grad_norm": 0.1840064972639084,
+      "learning_rate": 6.7722837633577766e-06,
+      "loss": 0.4191,
+      "step": 8172
+    },
+    {
+      "epoch": 1.8231095248717377,
+      "grad_norm": 0.1903713196516037,
+      "learning_rate": 6.770056386433884e-06,
+      "loss": 0.4352,
+      "step": 8173
+    },
+    {
+      "epoch": 1.823332589783627,
+      "grad_norm": 0.18376924097537994,
+      "learning_rate": 6.767829188404585e-06,
+      "loss": 0.4649,
+      "step": 8174
+    },
+    {
+      "epoch": 1.8235556546955163,
+      "grad_norm": 0.17748649418354034,
+      "learning_rate": 6.765602169393235e-06,
+      "loss": 0.4321,
+      "step": 8175
+    },
+    {
+      "epoch": 1.8237787196074058,
+      "grad_norm": 0.19591280817985535,
+      "learning_rate": 6.763375329523185e-06,
+      "loss": 0.4656,
+      "step": 8176
+    },
+    {
+      "epoch": 1.824001784519295,
+      "grad_norm": 0.1707315593957901,
+      "learning_rate": 6.761148668917766e-06,
+      "loss": 0.4413,
+      "step": 8177
+    },
+    {
+      "epoch": 1.8242248494311846,
+      "grad_norm": 0.17590689659118652,
+      "learning_rate": 6.758922187700309e-06,
+      "loss": 0.4682,
+      "step": 8178
+    },
+    {
+      "epoch": 1.8244479143430739,
+      "grad_norm": 0.17584408819675446,
+      "learning_rate": 6.756695885994126e-06,
+      "loss": 0.4626,
+      "step": 8179
+    },
+    {
+      "epoch": 1.8246709792549631,
+      "grad_norm": 0.1750144511461258,
+      "learning_rate": 6.754469763922529e-06,
+      "loss": 0.4217,
+      "step": 8180
+    },
+    {
+      "epoch": 1.8248940441668524,
+      "grad_norm": 0.18190471827983856,
+      "learning_rate": 6.752243821608808e-06,
+      "loss": 0.448,
+      "step": 8181
+    },
+    {
+      "epoch": 1.825117109078742,
+      "grad_norm": 0.1814090758562088,
+      "learning_rate": 6.7500180591762575e-06,
+      "loss": 0.442,
+      "step": 8182
+    },
+    {
+      "epoch": 1.8253401739906314,
+      "grad_norm": 0.18113404512405396,
+      "learning_rate": 6.747792476748147e-06,
+      "loss": 0.4522,
+      "step": 8183
+    },
+    {
+      "epoch": 1.8255632389025207,
+      "grad_norm": 0.1728588193655014,
+      "learning_rate": 6.7455670744477484e-06,
+      "loss": 0.4394,
+      "step": 8184
+    },
+    {
+      "epoch": 1.82578630381441,
+      "grad_norm": 0.17083577811717987,
+      "learning_rate": 6.743341852398315e-06,
+      "loss": 0.4432,
+      "step": 8185
+    },
+    {
+      "epoch": 1.8260093687262993,
+      "grad_norm": 0.17789053916931152,
+      "learning_rate": 6.741116810723096e-06,
+      "loss": 0.4453,
+      "step": 8186
+    },
+    {
+      "epoch": 1.8262324336381888,
+      "grad_norm": 0.18853406608104706,
+      "learning_rate": 6.738891949545325e-06,
+      "loss": 0.4398,
+      "step": 8187
+    },
+    {
+      "epoch": 1.826455498550078,
+      "grad_norm": 0.1743546426296234,
+      "learning_rate": 6.736667268988235e-06,
+      "loss": 0.4292,
+      "step": 8188
+    },
+    {
+      "epoch": 1.8266785634619676,
+      "grad_norm": 0.18301378190517426,
+      "learning_rate": 6.734442769175036e-06,
+      "loss": 0.4604,
+      "step": 8189
+    },
+    {
+      "epoch": 1.8269016283738568,
+      "grad_norm": 0.18097054958343506,
+      "learning_rate": 6.732218450228938e-06,
+      "loss": 0.4585,
+      "step": 8190
+    },
+    {
+      "epoch": 1.8271246932857461,
+      "grad_norm": 0.17369534075260162,
+      "learning_rate": 6.729994312273137e-06,
+      "loss": 0.4206,
+      "step": 8191
+    },
+    {
+      "epoch": 1.8273477581976354,
+      "grad_norm": 0.19749490916728973,
+      "learning_rate": 6.727770355430822e-06,
+      "loss": 0.4724,
+      "step": 8192
+    },
+    {
+      "epoch": 1.827570823109525,
+      "grad_norm": 0.17227788269519806,
+      "learning_rate": 6.725546579825165e-06,
+      "loss": 0.4397,
+      "step": 8193
+    },
+    {
+      "epoch": 1.8277938880214142,
+      "grad_norm": 0.17980121076107025,
+      "learning_rate": 6.723322985579338e-06,
+      "loss": 0.4568,
+      "step": 8194
+    },
+    {
+      "epoch": 1.8280169529333037,
+      "grad_norm": 0.1813877820968628,
+      "learning_rate": 6.721099572816494e-06,
+      "loss": 0.4348,
+      "step": 8195
+    },
+    {
+      "epoch": 1.828240017845193,
+      "grad_norm": 0.1828600913286209,
+      "learning_rate": 6.718876341659779e-06,
+      "loss": 0.4685,
+      "step": 8196
+    },
+    {
+      "epoch": 1.8284630827570822,
+      "grad_norm": 0.175664484500885,
+      "learning_rate": 6.716653292232334e-06,
+      "loss": 0.4359,
+      "step": 8197
+    },
+    {
+      "epoch": 1.8286861476689715,
+      "grad_norm": 0.18827195465564728,
+      "learning_rate": 6.714430424657281e-06,
+      "loss": 0.4545,
+      "step": 8198
+    },
+    {
+      "epoch": 1.828909212580861,
+      "grad_norm": 0.18291041254997253,
+      "learning_rate": 6.7122077390577415e-06,
+      "loss": 0.455,
+      "step": 8199
+    },
+    {
+      "epoch": 1.8291322774927505,
+      "grad_norm": 0.17457644641399384,
+      "learning_rate": 6.709985235556819e-06,
+      "loss": 0.4355,
+      "step": 8200
+    },
+    {
+      "epoch": 1.8293553424046398,
+      "grad_norm": 0.19204942882061005,
+      "learning_rate": 6.707762914277604e-06,
+      "loss": 0.4779,
+      "step": 8201
+    },
+    {
+      "epoch": 1.829578407316529,
+      "grad_norm": 0.17864084243774414,
+      "learning_rate": 6.705540775343192e-06,
+      "loss": 0.4301,
+      "step": 8202
+    },
+    {
+      "epoch": 1.8298014722284184,
+      "grad_norm": 0.17506073415279388,
+      "learning_rate": 6.703318818876652e-06,
+      "loss": 0.4389,
+      "step": 8203
+    },
+    {
+      "epoch": 1.8300245371403079,
+      "grad_norm": 0.1823255568742752,
+      "learning_rate": 6.701097045001055e-06,
+      "loss": 0.4644,
+      "step": 8204
+    },
+    {
+      "epoch": 1.8302476020521972,
+      "grad_norm": 0.192154660820961,
+      "learning_rate": 6.698875453839453e-06,
+      "loss": 0.447,
+      "step": 8205
+    },
+    {
+      "epoch": 1.8304706669640867,
+      "grad_norm": 0.18675589561462402,
+      "learning_rate": 6.696654045514894e-06,
+      "loss": 0.4493,
+      "step": 8206
+    },
+    {
+      "epoch": 1.830693731875976,
+      "grad_norm": 0.193002849817276,
+      "learning_rate": 6.694432820150412e-06,
+      "loss": 0.4124,
+      "step": 8207
+    },
+    {
+      "epoch": 1.8309167967878652,
+      "grad_norm": 0.1812351495027542,
+      "learning_rate": 6.692211777869033e-06,
+      "loss": 0.4555,
+      "step": 8208
+    },
+    {
+      "epoch": 1.8311398616997545,
+      "grad_norm": 0.18915650248527527,
+      "learning_rate": 6.689990918793773e-06,
+      "loss": 0.4678,
+      "step": 8209
+    },
+    {
+      "epoch": 1.831362926611644,
+      "grad_norm": 0.18118512630462646,
+      "learning_rate": 6.6877702430476374e-06,
+      "loss": 0.4425,
+      "step": 8210
+    },
+    {
+      "epoch": 1.8315859915235333,
+      "grad_norm": 0.17849402129650116,
+      "learning_rate": 6.68554975075362e-06,
+      "loss": 0.453,
+      "step": 8211
+    },
+    {
+      "epoch": 1.8318090564354228,
+      "grad_norm": 0.17353329062461853,
+      "learning_rate": 6.683329442034707e-06,
+      "loss": 0.4514,
+      "step": 8212
+    },
+    {
+      "epoch": 1.832032121347312,
+      "grad_norm": 0.1685749590396881,
+      "learning_rate": 6.681109317013869e-06,
+      "loss": 0.4358,
+      "step": 8213
+    },
+    {
+      "epoch": 1.8322551862592014,
+      "grad_norm": 0.17803844809532166,
+      "learning_rate": 6.678889375814077e-06,
+      "loss": 0.4495,
+      "step": 8214
+    },
+    {
+      "epoch": 1.8324782511710906,
+      "grad_norm": 0.18683882057666779,
+      "learning_rate": 6.67666961855828e-06,
+      "loss": 0.4501,
+      "step": 8215
+    },
+    {
+      "epoch": 1.8327013160829801,
+      "grad_norm": 0.18386538326740265,
+      "learning_rate": 6.674450045369427e-06,
+      "loss": 0.4349,
+      "step": 8216
+    },
+    {
+      "epoch": 1.8329243809948697,
+      "grad_norm": 0.20062412321567535,
+      "learning_rate": 6.672230656370446e-06,
+      "loss": 0.4646,
+      "step": 8217
+    },
+    {
+      "epoch": 1.833147445906759,
+      "grad_norm": 0.17725883424282074,
+      "learning_rate": 6.670011451684266e-06,
+      "loss": 0.452,
+      "step": 8218
+    },
+    {
+      "epoch": 1.8333705108186482,
+      "grad_norm": 0.16746068000793457,
+      "learning_rate": 6.667792431433796e-06,
+      "loss": 0.4277,
+      "step": 8219
+    },
+    {
+      "epoch": 1.8335935757305375,
+      "grad_norm": 0.18574824929237366,
+      "learning_rate": 6.665573595741941e-06,
+      "loss": 0.4381,
+      "step": 8220
+    },
+    {
+      "epoch": 1.833816640642427,
+      "grad_norm": 0.18854008615016937,
+      "learning_rate": 6.663354944731598e-06,
+      "loss": 0.4505,
+      "step": 8221
+    },
+    {
+      "epoch": 1.8340397055543163,
+      "grad_norm": 0.18151527643203735,
+      "learning_rate": 6.661136478525644e-06,
+      "loss": 0.4744,
+      "step": 8222
+    },
+    {
+      "epoch": 1.8342627704662058,
+      "grad_norm": 0.17463651299476624,
+      "learning_rate": 6.6589181972469556e-06,
+      "loss": 0.4667,
+      "step": 8223
+    },
+    {
+      "epoch": 1.834485835378095,
+      "grad_norm": 0.17236928641796112,
+      "learning_rate": 6.656700101018392e-06,
+      "loss": 0.4546,
+      "step": 8224
+    },
+    {
+      "epoch": 1.8347089002899843,
+      "grad_norm": 0.1812690645456314,
+      "learning_rate": 6.65448218996281e-06,
+      "loss": 0.4311,
+      "step": 8225
+    },
+    {
+      "epoch": 1.8349319652018736,
+      "grad_norm": 0.2300224006175995,
+      "learning_rate": 6.6522644642030445e-06,
+      "loss": 0.4551,
+      "step": 8226
+    },
+    {
+      "epoch": 1.8351550301137631,
+      "grad_norm": 0.17556323111057281,
+      "learning_rate": 6.650046923861935e-06,
+      "loss": 0.4333,
+      "step": 8227
+    },
+    {
+      "epoch": 1.8353780950256524,
+      "grad_norm": 0.17821857333183289,
+      "learning_rate": 6.647829569062295e-06,
+      "loss": 0.4486,
+      "step": 8228
+    },
+    {
+      "epoch": 1.835601159937542,
+      "grad_norm": 0.23370222747325897,
+      "learning_rate": 6.645612399926941e-06,
+      "loss": 0.4446,
+      "step": 8229
+    },
+    {
+      "epoch": 1.8358242248494312,
+      "grad_norm": 0.1826113909482956,
+      "learning_rate": 6.643395416578673e-06,
+      "loss": 0.4362,
+      "step": 8230
+    },
+    {
+      "epoch": 1.8360472897613205,
+      "grad_norm": 0.1840929388999939,
+      "learning_rate": 6.641178619140282e-06,
+      "loss": 0.4662,
+      "step": 8231
+    },
+    {
+      "epoch": 1.8362703546732098,
+      "grad_norm": 0.19260503351688385,
+      "learning_rate": 6.638962007734544e-06,
+      "loss": 0.4709,
+      "step": 8232
+    },
+    {
+      "epoch": 1.8364934195850993,
+      "grad_norm": 0.17979590594768524,
+      "learning_rate": 6.636745582484234e-06,
+      "loss": 0.4509,
+      "step": 8233
+    },
+    {
+      "epoch": 1.8367164844969888,
+      "grad_norm": 0.17919979989528656,
+      "learning_rate": 6.634529343512107e-06,
+      "loss": 0.4582,
+      "step": 8234
+    },
+    {
+      "epoch": 1.836939549408878,
+      "grad_norm": 0.1846793293952942,
+      "learning_rate": 6.632313290940917e-06,
+      "loss": 0.4356,
+      "step": 8235
+    },
+    {
+      "epoch": 1.8371626143207673,
+      "grad_norm": 0.18610712885856628,
+      "learning_rate": 6.630097424893398e-06,
+      "loss": 0.4555,
+      "step": 8236
+    },
+    {
+      "epoch": 1.8373856792326566,
+      "grad_norm": 0.17965011298656464,
+      "learning_rate": 6.6278817454922835e-06,
+      "loss": 0.4494,
+      "step": 8237
+    },
+    {
+      "epoch": 1.8376087441445461,
+      "grad_norm": 0.1982518583536148,
+      "learning_rate": 6.625666252860288e-06,
+      "loss": 0.4644,
+      "step": 8238
+    },
+    {
+      "epoch": 1.8378318090564354,
+      "grad_norm": 0.18403403460979462,
+      "learning_rate": 6.6234509471201205e-06,
+      "loss": 0.4501,
+      "step": 8239
+    },
+    {
+      "epoch": 1.838054873968325,
+      "grad_norm": 0.17158350348472595,
+      "learning_rate": 6.621235828394481e-06,
+      "loss": 0.4386,
+      "step": 8240
+    },
+    {
+      "epoch": 1.8382779388802142,
+      "grad_norm": 0.1909009963274002,
+      "learning_rate": 6.619020896806052e-06,
+      "loss": 0.4464,
+      "step": 8241
+    },
+    {
+      "epoch": 1.8385010037921035,
+      "grad_norm": 0.18062900006771088,
+      "learning_rate": 6.616806152477515e-06,
+      "loss": 0.4657,
+      "step": 8242
+    },
+    {
+      "epoch": 1.8387240687039927,
+      "grad_norm": 0.1820714771747589,
+      "learning_rate": 6.614591595531533e-06,
+      "loss": 0.4723,
+      "step": 8243
+    },
+    {
+      "epoch": 1.8389471336158822,
+      "grad_norm": 0.18476656079292297,
+      "learning_rate": 6.612377226090764e-06,
+      "loss": 0.4457,
+      "step": 8244
+    },
+    {
+      "epoch": 1.8391701985277715,
+      "grad_norm": 0.17424450814723969,
+      "learning_rate": 6.610163044277853e-06,
+      "loss": 0.444,
+      "step": 8245
+    },
+    {
+      "epoch": 1.839393263439661,
+      "grad_norm": 0.19445692002773285,
+      "learning_rate": 6.607949050215438e-06,
+      "loss": 0.4446,
+      "step": 8246
+    },
+    {
+      "epoch": 1.8396163283515503,
+      "grad_norm": 0.18586407601833344,
+      "learning_rate": 6.605735244026138e-06,
+      "loss": 0.4495,
+      "step": 8247
+    },
+    {
+      "epoch": 1.8398393932634396,
+      "grad_norm": 0.17516399919986725,
+      "learning_rate": 6.6035216258325745e-06,
+      "loss": 0.4261,
+      "step": 8248
+    },
+    {
+      "epoch": 1.8400624581753289,
+      "grad_norm": 0.17480303347110748,
+      "learning_rate": 6.601308195757343e-06,
+      "loss": 0.4283,
+      "step": 8249
+    },
+    {
+      "epoch": 1.8402855230872184,
+      "grad_norm": 0.1745281219482422,
+      "learning_rate": 6.599094953923046e-06,
+      "loss": 0.4384,
+      "step": 8250
+    },
+    {
+      "epoch": 1.8405085879991079,
+      "grad_norm": 0.18595942854881287,
+      "learning_rate": 6.5968819004522625e-06,
+      "loss": 0.4519,
+      "step": 8251
+    },
+    {
+      "epoch": 1.8407316529109972,
+      "grad_norm": 0.17765408754348755,
+      "learning_rate": 6.594669035467567e-06,
+      "loss": 0.4463,
+      "step": 8252
+    },
+    {
+      "epoch": 1.8409547178228864,
+      "grad_norm": 0.17620229721069336,
+      "learning_rate": 6.592456359091517e-06,
+      "loss": 0.4545,
+      "step": 8253
+    },
+    {
+      "epoch": 1.8411777827347757,
+      "grad_norm": 0.18198491632938385,
+      "learning_rate": 6.5902438714466706e-06,
+      "loss": 0.4706,
+      "step": 8254
+    },
+    {
+      "epoch": 1.8414008476466652,
+      "grad_norm": 0.1813773214817047,
+      "learning_rate": 6.588031572655566e-06,
+      "loss": 0.4435,
+      "step": 8255
+    },
+    {
+      "epoch": 1.8416239125585545,
+      "grad_norm": 0.1799456626176834,
+      "learning_rate": 6.585819462840737e-06,
+      "loss": 0.4108,
+      "step": 8256
+    },
+    {
+      "epoch": 1.841846977470444,
+      "grad_norm": 0.18853728473186493,
+      "learning_rate": 6.583607542124699e-06,
+      "loss": 0.4527,
+      "step": 8257
+    },
+    {
+      "epoch": 1.8420700423823333,
+      "grad_norm": 0.1802626997232437,
+      "learning_rate": 6.581395810629969e-06,
+      "loss": 0.4413,
+      "step": 8258
+    },
+    {
+      "epoch": 1.8422931072942226,
+      "grad_norm": 0.18371166288852692,
+      "learning_rate": 6.57918426847904e-06,
+      "loss": 0.4408,
+      "step": 8259
+    },
+    {
+      "epoch": 1.8425161722061119,
+      "grad_norm": 0.17911924421787262,
+      "learning_rate": 6.576972915794404e-06,
+      "loss": 0.444,
+      "step": 8260
+    },
+    {
+      "epoch": 1.8427392371180014,
+      "grad_norm": 0.1887340396642685,
+      "learning_rate": 6.574761752698542e-06,
+      "loss": 0.4522,
+      "step": 8261
+    },
+    {
+      "epoch": 1.8429623020298909,
+      "grad_norm": 0.2037588208913803,
+      "learning_rate": 6.572550779313916e-06,
+      "loss": 0.4419,
+      "step": 8262
+    },
+    {
+      "epoch": 1.8431853669417801,
+      "grad_norm": 0.17989180982112885,
+      "learning_rate": 6.570339995762991e-06,
+      "loss": 0.4453,
+      "step": 8263
+    },
+    {
+      "epoch": 1.8434084318536694,
+      "grad_norm": 0.17799720168113708,
+      "learning_rate": 6.5681294021682084e-06,
+      "loss": 0.4333,
+      "step": 8264
+    },
+    {
+      "epoch": 1.8436314967655587,
+      "grad_norm": 0.17597989737987518,
+      "learning_rate": 6.5659189986520085e-06,
+      "loss": 0.4441,
+      "step": 8265
+    },
+    {
+      "epoch": 1.843854561677448,
+      "grad_norm": 0.1881999969482422,
+      "learning_rate": 6.563708785336813e-06,
+      "loss": 0.4674,
+      "step": 8266
+    },
+    {
+      "epoch": 1.8440776265893375,
+      "grad_norm": 0.18646754324436188,
+      "learning_rate": 6.561498762345044e-06,
+      "loss": 0.4533,
+      "step": 8267
+    },
+    {
+      "epoch": 1.844300691501227,
+      "grad_norm": 0.1898210197687149,
+      "learning_rate": 6.559288929799099e-06,
+      "loss": 0.4494,
+      "step": 8268
+    },
+    {
+      "epoch": 1.8445237564131163,
+      "grad_norm": 0.1805214136838913,
+      "learning_rate": 6.557079287821378e-06,
+      "loss": 0.4533,
+      "step": 8269
+    },
+    {
+      "epoch": 1.8447468213250056,
+      "grad_norm": 0.17660482227802277,
+      "learning_rate": 6.554869836534261e-06,
+      "loss": 0.4237,
+      "step": 8270
+    },
+    {
+      "epoch": 1.8449698862368948,
+      "grad_norm": 0.1824430525302887,
+      "learning_rate": 6.552660576060126e-06,
+      "loss": 0.4767,
+      "step": 8271
+    },
+    {
+      "epoch": 1.8451929511487843,
+      "grad_norm": 0.1871567666530609,
+      "learning_rate": 6.55045150652133e-06,
+      "loss": 0.4576,
+      "step": 8272
+    },
+    {
+      "epoch": 1.8454160160606736,
+      "grad_norm": 0.17943458259105682,
+      "learning_rate": 6.548242628040231e-06,
+      "loss": 0.4475,
+      "step": 8273
+    },
+    {
+      "epoch": 1.8456390809725631,
+      "grad_norm": 0.1877557635307312,
+      "learning_rate": 6.5460339407391645e-06,
+      "loss": 0.4419,
+      "step": 8274
+    },
+    {
+      "epoch": 1.8458621458844524,
+      "grad_norm": 0.17813464999198914,
+      "learning_rate": 6.543825444740468e-06,
+      "loss": 0.4647,
+      "step": 8275
+    },
+    {
+      "epoch": 1.8460852107963417,
+      "grad_norm": 0.18218432366847992,
+      "learning_rate": 6.5416171401664565e-06,
+      "loss": 0.4735,
+      "step": 8276
+    },
+    {
+      "epoch": 1.846308275708231,
+      "grad_norm": 0.1764039546251297,
+      "learning_rate": 6.539409027139443e-06,
+      "loss": 0.4525,
+      "step": 8277
+    },
+    {
+      "epoch": 1.8465313406201205,
+      "grad_norm": 0.1932554543018341,
+      "learning_rate": 6.537201105781723e-06,
+      "loss": 0.4911,
+      "step": 8278
+    },
+    {
+      "epoch": 1.84675440553201,
+      "grad_norm": 0.1814078986644745,
+      "learning_rate": 6.53499337621559e-06,
+      "loss": 0.4461,
+      "step": 8279
+    },
+    {
+      "epoch": 1.8469774704438993,
+      "grad_norm": 0.173579141497612,
+      "learning_rate": 6.5327858385633184e-06,
+      "loss": 0.4262,
+      "step": 8280
+    },
+    {
+      "epoch": 1.8472005353557885,
+      "grad_norm": 0.18026985228061676,
+      "learning_rate": 6.530578492947177e-06,
+      "loss": 0.4385,
+      "step": 8281
+    },
+    {
+      "epoch": 1.8474236002676778,
+      "grad_norm": 0.18310683965682983,
+      "learning_rate": 6.528371339489428e-06,
+      "loss": 0.4574,
+      "step": 8282
+    },
+    {
+      "epoch": 1.847646665179567,
+      "grad_norm": 0.18035286664962769,
+      "learning_rate": 6.526164378312305e-06,
+      "loss": 0.4579,
+      "step": 8283
+    },
+    {
+      "epoch": 1.8478697300914566,
+      "grad_norm": 0.20232632756233215,
+      "learning_rate": 6.523957609538049e-06,
+      "loss": 0.4492,
+      "step": 8284
+    },
+    {
+      "epoch": 1.8480927950033461,
+      "grad_norm": 0.180943563580513,
+      "learning_rate": 6.5217510332888904e-06,
+      "loss": 0.4339,
+      "step": 8285
+    },
+    {
+      "epoch": 1.8483158599152354,
+      "grad_norm": 0.17553845047950745,
+      "learning_rate": 6.519544649687034e-06,
+      "loss": 0.4313,
+      "step": 8286
+    },
+    {
+      "epoch": 1.8485389248271247,
+      "grad_norm": 0.17904485762119293,
+      "learning_rate": 6.517338458854692e-06,
+      "loss": 0.4484,
+      "step": 8287
+    },
+    {
+      "epoch": 1.848761989739014,
+      "grad_norm": 0.17891845107078552,
+      "learning_rate": 6.515132460914049e-06,
+      "loss": 0.4655,
+      "step": 8288
+    },
+    {
+      "epoch": 1.8489850546509035,
+      "grad_norm": 0.19273404777050018,
+      "learning_rate": 6.512926655987295e-06,
+      "loss": 0.4718,
+      "step": 8289
+    },
+    {
+      "epoch": 1.8492081195627927,
+      "grad_norm": 0.21862982213497162,
+      "learning_rate": 6.510721044196593e-06,
+      "loss": 0.4678,
+      "step": 8290
+    },
+    {
+      "epoch": 1.8494311844746822,
+      "grad_norm": 0.17714935541152954,
+      "learning_rate": 6.508515625664111e-06,
+      "loss": 0.4626,
+      "step": 8291
+    },
+    {
+      "epoch": 1.8496542493865715,
+      "grad_norm": 0.1834956407546997,
+      "learning_rate": 6.5063104005119945e-06,
+      "loss": 0.4463,
+      "step": 8292
+    },
+    {
+      "epoch": 1.8498773142984608,
+      "grad_norm": 0.17443230748176575,
+      "learning_rate": 6.504105368862386e-06,
+      "loss": 0.4563,
+      "step": 8293
+    },
+    {
+      "epoch": 1.85010037921035,
+      "grad_norm": 0.17853355407714844,
+      "learning_rate": 6.50190053083741e-06,
+      "loss": 0.4588,
+      "step": 8294
+    },
+    {
+      "epoch": 1.8503234441222396,
+      "grad_norm": 0.17415203154087067,
+      "learning_rate": 6.4996958865591895e-06,
+      "loss": 0.4303,
+      "step": 8295
+    },
+    {
+      "epoch": 1.850546509034129,
+      "grad_norm": 0.169075608253479,
+      "learning_rate": 6.497491436149826e-06,
+      "loss": 0.4292,
+      "step": 8296
+    },
+    {
+      "epoch": 1.8507695739460184,
+      "grad_norm": 0.1862117052078247,
+      "learning_rate": 6.49528717973142e-06,
+      "loss": 0.4308,
+      "step": 8297
+    },
+    {
+      "epoch": 1.8509926388579077,
+      "grad_norm": 0.1834726333618164,
+      "learning_rate": 6.4930831174260535e-06,
+      "loss": 0.4493,
+      "step": 8298
+    },
+    {
+      "epoch": 1.851215703769797,
+      "grad_norm": 0.18994340300559998,
+      "learning_rate": 6.490879249355806e-06,
+      "loss": 0.4688,
+      "step": 8299
+    },
+    {
+      "epoch": 1.8514387686816862,
+      "grad_norm": 0.18322892487049103,
+      "learning_rate": 6.4886755756427365e-06,
+      "loss": 0.4473,
+      "step": 8300
+    },
+    {
+      "epoch": 1.8516618335935757,
+      "grad_norm": 0.1817985475063324,
+      "learning_rate": 6.486472096408903e-06,
+      "loss": 0.4633,
+      "step": 8301
+    },
+    {
+      "epoch": 1.8518848985054652,
+      "grad_norm": 0.17432516813278198,
+      "learning_rate": 6.484268811776343e-06,
+      "loss": 0.4574,
+      "step": 8302
+    },
+    {
+      "epoch": 1.8521079634173545,
+      "grad_norm": 0.17498242855072021,
+      "learning_rate": 6.482065721867095e-06,
+      "loss": 0.4379,
+      "step": 8303
+    },
+    {
+      "epoch": 1.8523310283292438,
+      "grad_norm": 0.18785923719406128,
+      "learning_rate": 6.479862826803172e-06,
+      "loss": 0.4572,
+      "step": 8304
+    },
+    {
+      "epoch": 1.852554093241133,
+      "grad_norm": 0.1780933290719986,
+      "learning_rate": 6.477660126706588e-06,
+      "loss": 0.4161,
+      "step": 8305
+    },
+    {
+      "epoch": 1.8527771581530226,
+      "grad_norm": 0.17613643407821655,
+      "learning_rate": 6.4754576216993455e-06,
+      "loss": 0.4323,
+      "step": 8306
+    },
+    {
+      "epoch": 1.8530002230649119,
+      "grad_norm": 0.17086435854434967,
+      "learning_rate": 6.473255311903428e-06,
+      "loss": 0.4497,
+      "step": 8307
+    },
+    {
+      "epoch": 1.8532232879768014,
+      "grad_norm": 0.18054986000061035,
+      "learning_rate": 6.471053197440817e-06,
+      "loss": 0.4697,
+      "step": 8308
+    },
+    {
+      "epoch": 1.8534463528886906,
+      "grad_norm": 0.17696300148963928,
+      "learning_rate": 6.468851278433476e-06,
+      "loss": 0.4704,
+      "step": 8309
+    },
+    {
+      "epoch": 1.85366941780058,
+      "grad_norm": 0.18155822157859802,
+      "learning_rate": 6.466649555003366e-06,
+      "loss": 0.4576,
+      "step": 8310
+    },
+    {
+      "epoch": 1.8538924827124692,
+      "grad_norm": 0.18807131052017212,
+      "learning_rate": 6.464448027272427e-06,
+      "loss": 0.4651,
+      "step": 8311
+    },
+    {
+      "epoch": 1.8541155476243587,
+      "grad_norm": 0.1846548169851303,
+      "learning_rate": 6.462246695362597e-06,
+      "loss": 0.4697,
+      "step": 8312
+    },
+    {
+      "epoch": 1.8543386125362482,
+      "grad_norm": 0.18467946350574493,
+      "learning_rate": 6.460045559395797e-06,
+      "loss": 0.4474,
+      "step": 8313
+    },
+    {
+      "epoch": 1.8545616774481375,
+      "grad_norm": 0.18795856833457947,
+      "learning_rate": 6.457844619493943e-06,
+      "loss": 0.461,
+      "step": 8314
+    },
+    {
+      "epoch": 1.8547847423600268,
+      "grad_norm": 0.1844158172607422,
+      "learning_rate": 6.455643875778932e-06,
+      "loss": 0.4456,
+      "step": 8315
+    },
+    {
+      "epoch": 1.855007807271916,
+      "grad_norm": 0.19735929369926453,
+      "learning_rate": 6.453443328372662e-06,
+      "loss": 0.4785,
+      "step": 8316
+    },
+    {
+      "epoch": 1.8552308721838053,
+      "grad_norm": 0.187907874584198,
+      "learning_rate": 6.451242977397006e-06,
+      "loss": 0.4334,
+      "step": 8317
+    },
+    {
+      "epoch": 1.8554539370956948,
+      "grad_norm": 0.18026615679264069,
+      "learning_rate": 6.449042822973839e-06,
+      "loss": 0.4454,
+      "step": 8318
+    },
+    {
+      "epoch": 1.8556770020075843,
+      "grad_norm": 0.17745928466320038,
+      "learning_rate": 6.446842865225014e-06,
+      "loss": 0.4621,
+      "step": 8319
+    },
+    {
+      "epoch": 1.8559000669194736,
+      "grad_norm": 0.17665286362171173,
+      "learning_rate": 6.4446431042723835e-06,
+      "loss": 0.4459,
+      "step": 8320
+    },
+    {
+      "epoch": 1.856123131831363,
+      "grad_norm": 0.18266570568084717,
+      "learning_rate": 6.44244354023778e-06,
+      "loss": 0.466,
+      "step": 8321
+    },
+    {
+      "epoch": 1.8563461967432522,
+      "grad_norm": 0.27143698930740356,
+      "learning_rate": 6.440244173243033e-06,
+      "loss": 0.4728,
+      "step": 8322
+    },
+    {
+      "epoch": 1.8565692616551417,
+      "grad_norm": 0.18305523693561554,
+      "learning_rate": 6.438045003409954e-06,
+      "loss": 0.4272,
+      "step": 8323
+    },
+    {
+      "epoch": 1.856792326567031,
+      "grad_norm": 0.17935436964035034,
+      "learning_rate": 6.435846030860349e-06,
+      "loss": 0.4616,
+      "step": 8324
+    },
+    {
+      "epoch": 1.8570153914789205,
+      "grad_norm": 0.17304593324661255,
+      "learning_rate": 6.433647255716011e-06,
+      "loss": 0.4485,
+      "step": 8325
+    },
+    {
+      "epoch": 1.8572384563908098,
+      "grad_norm": 0.1776076853275299,
+      "learning_rate": 6.43144867809872e-06,
+      "loss": 0.4431,
+      "step": 8326
+    },
+    {
+      "epoch": 1.857461521302699,
+      "grad_norm": 0.18469847738742828,
+      "learning_rate": 6.4292502981302494e-06,
+      "loss": 0.4633,
+      "step": 8327
+    },
+    {
+      "epoch": 1.8576845862145883,
+      "grad_norm": 0.18084385991096497,
+      "learning_rate": 6.427052115932357e-06,
+      "loss": 0.4558,
+      "step": 8328
+    },
+    {
+      "epoch": 1.8579076511264778,
+      "grad_norm": 0.1821753978729248,
+      "learning_rate": 6.424854131626796e-06,
+      "loss": 0.4565,
+      "step": 8329
+    },
+    {
+      "epoch": 1.8581307160383673,
+      "grad_norm": 0.17873086035251617,
+      "learning_rate": 6.422656345335299e-06,
+      "loss": 0.4521,
+      "step": 8330
+    },
+    {
+      "epoch": 1.8583537809502566,
+      "grad_norm": 0.2390872985124588,
+      "learning_rate": 6.420458757179597e-06,
+      "loss": 0.4443,
+      "step": 8331
+    },
+    {
+      "epoch": 1.8585768458621459,
+      "grad_norm": 0.1892087161540985,
+      "learning_rate": 6.418261367281404e-06,
+      "loss": 0.4496,
+      "step": 8332
+    },
+    {
+      "epoch": 1.8587999107740352,
+      "grad_norm": 0.1815430074930191,
+      "learning_rate": 6.416064175762428e-06,
+      "loss": 0.4725,
+      "step": 8333
+    },
+    {
+      "epoch": 1.8590229756859245,
+      "grad_norm": 0.17400890588760376,
+      "learning_rate": 6.41386718274436e-06,
+      "loss": 0.4264,
+      "step": 8334
+    },
+    {
+      "epoch": 1.859246040597814,
+      "grad_norm": 0.17337702214717865,
+      "learning_rate": 6.4116703883488875e-06,
+      "loss": 0.4363,
+      "step": 8335
+    },
+    {
+      "epoch": 1.8594691055097035,
+      "grad_norm": 0.1814270168542862,
+      "learning_rate": 6.409473792697678e-06,
+      "loss": 0.4413,
+      "step": 8336
+    },
+    {
+      "epoch": 1.8596921704215927,
+      "grad_norm": 0.17769312858581543,
+      "learning_rate": 6.407277395912397e-06,
+      "loss": 0.4651,
+      "step": 8337
+    },
+    {
+      "epoch": 1.859915235333482,
+      "grad_norm": 0.18968099355697632,
+      "learning_rate": 6.405081198114692e-06,
+      "loss": 0.4984,
+      "step": 8338
+    },
+    {
+      "epoch": 1.8601383002453713,
+      "grad_norm": 0.17772215604782104,
+      "learning_rate": 6.4028851994262045e-06,
+      "loss": 0.4267,
+      "step": 8339
+    },
+    {
+      "epoch": 1.8603613651572608,
+      "grad_norm": 0.18239617347717285,
+      "learning_rate": 6.400689399968559e-06,
+      "loss": 0.47,
+      "step": 8340
+    },
+    {
+      "epoch": 1.86058443006915,
+      "grad_norm": 0.1779201775789261,
+      "learning_rate": 6.398493799863378e-06,
+      "loss": 0.4684,
+      "step": 8341
+    },
+    {
+      "epoch": 1.8608074949810396,
+      "grad_norm": 0.18368324637413025,
+      "learning_rate": 6.396298399232261e-06,
+      "loss": 0.4599,
+      "step": 8342
+    },
+    {
+      "epoch": 1.8610305598929289,
+      "grad_norm": 0.1744937300682068,
+      "learning_rate": 6.39410319819681e-06,
+      "loss": 0.4278,
+      "step": 8343
+    },
+    {
+      "epoch": 1.8612536248048182,
+      "grad_norm": 0.17699581384658813,
+      "learning_rate": 6.391908196878605e-06,
+      "loss": 0.4393,
+      "step": 8344
+    },
+    {
+      "epoch": 1.8614766897167074,
+      "grad_norm": 0.17797012627124786,
+      "learning_rate": 6.389713395399219e-06,
+      "loss": 0.4653,
+      "step": 8345
+    },
+    {
+      "epoch": 1.861699754628597,
+      "grad_norm": 0.17617610096931458,
+      "learning_rate": 6.3875187938802164e-06,
+      "loss": 0.4577,
+      "step": 8346
+    },
+    {
+      "epoch": 1.8619228195404864,
+      "grad_norm": 0.17390653491020203,
+      "learning_rate": 6.385324392443144e-06,
+      "loss": 0.4215,
+      "step": 8347
+    },
+    {
+      "epoch": 1.8621458844523757,
+      "grad_norm": 0.1813625991344452,
+      "learning_rate": 6.3831301912095465e-06,
+      "loss": 0.429,
+      "step": 8348
+    },
+    {
+      "epoch": 1.862368949364265,
+      "grad_norm": 0.17084379494190216,
+      "learning_rate": 6.380936190300948e-06,
+      "loss": 0.4402,
+      "step": 8349
+    },
+    {
+      "epoch": 1.8625920142761543,
+      "grad_norm": 0.18061068654060364,
+      "learning_rate": 6.37874238983887e-06,
+      "loss": 0.4657,
+      "step": 8350
+    },
+    {
+      "epoch": 1.8628150791880436,
+      "grad_norm": 0.18320946395397186,
+      "learning_rate": 6.376548789944817e-06,
+      "loss": 0.4526,
+      "step": 8351
+    },
+    {
+      "epoch": 1.863038144099933,
+      "grad_norm": 0.17092865705490112,
+      "learning_rate": 6.374355390740284e-06,
+      "loss": 0.4259,
+      "step": 8352
+    },
+    {
+      "epoch": 1.8632612090118226,
+      "grad_norm": 0.17330855131149292,
+      "learning_rate": 6.372162192346754e-06,
+      "loss": 0.4313,
+      "step": 8353
+    },
+    {
+      "epoch": 1.8634842739237119,
+      "grad_norm": 0.17183774709701538,
+      "learning_rate": 6.369969194885703e-06,
+      "loss": 0.4601,
+      "step": 8354
+    },
+    {
+      "epoch": 1.8637073388356011,
+      "grad_norm": 0.19768626987934113,
+      "learning_rate": 6.367776398478591e-06,
+      "loss": 0.491,
+      "step": 8355
+    },
+    {
+      "epoch": 1.8639304037474904,
+      "grad_norm": 0.1920434832572937,
+      "learning_rate": 6.365583803246872e-06,
+      "loss": 0.444,
+      "step": 8356
+    },
+    {
+      "epoch": 1.86415346865938,
+      "grad_norm": 0.18218182027339935,
+      "learning_rate": 6.3633914093119804e-06,
+      "loss": 0.5027,
+      "step": 8357
+    },
+    {
+      "epoch": 1.8643765335712692,
+      "grad_norm": 0.18595336377620697,
+      "learning_rate": 6.36119921679535e-06,
+      "loss": 0.4463,
+      "step": 8358
+    },
+    {
+      "epoch": 1.8645995984831587,
+      "grad_norm": 0.19364967942237854,
+      "learning_rate": 6.359007225818395e-06,
+      "loss": 0.4637,
+      "step": 8359
+    },
+    {
+      "epoch": 1.864822663395048,
+      "grad_norm": 0.18406571447849274,
+      "learning_rate": 6.356815436502524e-06,
+      "loss": 0.4745,
+      "step": 8360
+    },
+    {
+      "epoch": 1.8650457283069373,
+      "grad_norm": 0.18429051339626312,
+      "learning_rate": 6.354623848969128e-06,
+      "loss": 0.4379,
+      "step": 8361
+    },
+    {
+      "epoch": 1.8652687932188265,
+      "grad_norm": 0.18177488446235657,
+      "learning_rate": 6.352432463339596e-06,
+      "loss": 0.4565,
+      "step": 8362
+    },
+    {
+      "epoch": 1.865491858130716,
+      "grad_norm": 0.1876797080039978,
+      "learning_rate": 6.350241279735301e-06,
+      "loss": 0.4532,
+      "step": 8363
+    },
+    {
+      "epoch": 1.8657149230426056,
+      "grad_norm": 0.18419890105724335,
+      "learning_rate": 6.3480502982775974e-06,
+      "loss": 0.4485,
+      "step": 8364
+    },
+    {
+      "epoch": 1.8659379879544948,
+      "grad_norm": 0.1855155974626541,
+      "learning_rate": 6.3458595190878414e-06,
+      "loss": 0.455,
+      "step": 8365
+    },
+    {
+      "epoch": 1.8661610528663841,
+      "grad_norm": 0.18031683564186096,
+      "learning_rate": 6.343668942287369e-06,
+      "loss": 0.4502,
+      "step": 8366
+    },
+    {
+      "epoch": 1.8663841177782734,
+      "grad_norm": 0.18513405323028564,
+      "learning_rate": 6.341478567997512e-06,
+      "loss": 0.458,
+      "step": 8367
+    },
+    {
+      "epoch": 1.8666071826901627,
+      "grad_norm": 0.1804891675710678,
+      "learning_rate": 6.3392883963395826e-06,
+      "loss": 0.4391,
+      "step": 8368
+    },
+    {
+      "epoch": 1.8668302476020522,
+      "grad_norm": 0.17720703780651093,
+      "learning_rate": 6.337098427434889e-06,
+      "loss": 0.4451,
+      "step": 8369
+    },
+    {
+      "epoch": 1.8670533125139417,
+      "grad_norm": 0.17895428836345673,
+      "learning_rate": 6.334908661404726e-06,
+      "loss": 0.4603,
+      "step": 8370
+    },
+    {
+      "epoch": 1.867276377425831,
+      "grad_norm": 0.1729314774274826,
+      "learning_rate": 6.3327190983703745e-06,
+      "loss": 0.4544,
+      "step": 8371
+    },
+    {
+      "epoch": 1.8674994423377203,
+      "grad_norm": 0.17181377112865448,
+      "learning_rate": 6.330529738453109e-06,
+      "loss": 0.4406,
+      "step": 8372
+    },
+    {
+      "epoch": 1.8677225072496095,
+      "grad_norm": 0.17984220385551453,
+      "learning_rate": 6.328340581774187e-06,
+      "loss": 0.4796,
+      "step": 8373
+    },
+    {
+      "epoch": 1.867945572161499,
+      "grad_norm": 0.1723097413778305,
+      "learning_rate": 6.326151628454862e-06,
+      "loss": 0.4341,
+      "step": 8374
+    },
+    {
+      "epoch": 1.8681686370733883,
+      "grad_norm": 0.17812368273735046,
+      "learning_rate": 6.323962878616366e-06,
+      "loss": 0.4558,
+      "step": 8375
+    },
+    {
+      "epoch": 1.8683917019852778,
+      "grad_norm": 0.2598254680633545,
+      "learning_rate": 6.3217743323799305e-06,
+      "loss": 0.4445,
+      "step": 8376
+    },
+    {
+      "epoch": 1.868614766897167,
+      "grad_norm": 0.17721040546894073,
+      "learning_rate": 6.3195859898667675e-06,
+      "loss": 0.4308,
+      "step": 8377
+    },
+    {
+      "epoch": 1.8688378318090564,
+      "grad_norm": 0.1813073456287384,
+      "learning_rate": 6.317397851198087e-06,
+      "loss": 0.4672,
+      "step": 8378
+    },
+    {
+      "epoch": 1.8690608967209457,
+      "grad_norm": 0.17863677442073822,
+      "learning_rate": 6.315209916495076e-06,
+      "loss": 0.4606,
+      "step": 8379
+    },
+    {
+      "epoch": 1.8692839616328352,
+      "grad_norm": 0.1793542504310608,
+      "learning_rate": 6.313022185878918e-06,
+      "loss": 0.4879,
+      "step": 8380
+    },
+    {
+      "epoch": 1.8695070265447247,
+      "grad_norm": 0.18399956822395325,
+      "learning_rate": 6.310834659470781e-06,
+      "loss": 0.4591,
+      "step": 8381
+    },
+    {
+      "epoch": 1.869730091456614,
+      "grad_norm": 0.18076615035533905,
+      "learning_rate": 6.308647337391831e-06,
+      "loss": 0.4385,
+      "step": 8382
+    },
+    {
+      "epoch": 1.8699531563685032,
+      "grad_norm": 0.1910838931798935,
+      "learning_rate": 6.306460219763207e-06,
+      "loss": 0.4298,
+      "step": 8383
+    },
+    {
+      "epoch": 1.8701762212803925,
+      "grad_norm": 0.191563680768013,
+      "learning_rate": 6.304273306706052e-06,
+      "loss": 0.4685,
+      "step": 8384
+    },
+    {
+      "epoch": 1.8703992861922818,
+      "grad_norm": 0.1867865025997162,
+      "learning_rate": 6.302086598341486e-06,
+      "loss": 0.4638,
+      "step": 8385
+    },
+    {
+      "epoch": 1.8706223511041713,
+      "grad_norm": 0.18163664638996124,
+      "learning_rate": 6.2999000947906275e-06,
+      "loss": 0.4394,
+      "step": 8386
+    },
+    {
+      "epoch": 1.8708454160160608,
+      "grad_norm": 0.1798231601715088,
+      "learning_rate": 6.297713796174574e-06,
+      "loss": 0.4885,
+      "step": 8387
+    },
+    {
+      "epoch": 1.87106848092795,
+      "grad_norm": 0.19126291573047638,
+      "learning_rate": 6.295527702614422e-06,
+      "loss": 0.46,
+      "step": 8388
+    },
+    {
+      "epoch": 1.8712915458398394,
+      "grad_norm": 0.17811056971549988,
+      "learning_rate": 6.293341814231244e-06,
+      "loss": 0.4518,
+      "step": 8389
+    },
+    {
+      "epoch": 1.8715146107517286,
+      "grad_norm": 0.1750759333372116,
+      "learning_rate": 6.2911561311461126e-06,
+      "loss": 0.445,
+      "step": 8390
+    },
+    {
+      "epoch": 1.8717376756636182,
+      "grad_norm": 0.17805348336696625,
+      "learning_rate": 6.288970653480087e-06,
+      "loss": 0.4423,
+      "step": 8391
+    },
+    {
+      "epoch": 1.8719607405755074,
+      "grad_norm": 0.18071311712265015,
+      "learning_rate": 6.286785381354205e-06,
+      "loss": 0.4472,
+      "step": 8392
+    },
+    {
+      "epoch": 1.872183805487397,
+      "grad_norm": 0.1815258264541626,
+      "learning_rate": 6.284600314889511e-06,
+      "loss": 0.4749,
+      "step": 8393
+    },
+    {
+      "epoch": 1.8724068703992862,
+      "grad_norm": 0.1812569499015808,
+      "learning_rate": 6.28241545420702e-06,
+      "loss": 0.4406,
+      "step": 8394
+    },
+    {
+      "epoch": 1.8726299353111755,
+      "grad_norm": 0.1827443242073059,
+      "learning_rate": 6.280230799427747e-06,
+      "loss": 0.4698,
+      "step": 8395
+    },
+    {
+      "epoch": 1.8728530002230648,
+      "grad_norm": 0.1790318340063095,
+      "learning_rate": 6.278046350672689e-06,
+      "loss": 0.4462,
+      "step": 8396
+    },
+    {
+      "epoch": 1.8730760651349543,
+      "grad_norm": 0.18236975371837616,
+      "learning_rate": 6.275862108062837e-06,
+      "loss": 0.4655,
+      "step": 8397
+    },
+    {
+      "epoch": 1.8732991300468438,
+      "grad_norm": 0.2039221227169037,
+      "learning_rate": 6.273678071719166e-06,
+      "loss": 0.4689,
+      "step": 8398
+    },
+    {
+      "epoch": 1.873522194958733,
+      "grad_norm": 0.181179016828537,
+      "learning_rate": 6.2714942417626445e-06,
+      "loss": 0.436,
+      "step": 8399
+    },
+    {
+      "epoch": 1.8737452598706223,
+      "grad_norm": 0.1835191249847412,
+      "learning_rate": 6.269310618314223e-06,
+      "loss": 0.4576,
+      "step": 8400
+    },
+    {
+      "epoch": 1.8739683247825116,
+      "grad_norm": 0.19027259945869446,
+      "learning_rate": 6.267127201494849e-06,
+      "loss": 0.4814,
+      "step": 8401
+    },
+    {
+      "epoch": 1.874191389694401,
+      "grad_norm": 0.17249448597431183,
+      "learning_rate": 6.264943991425449e-06,
+      "loss": 0.4466,
+      "step": 8402
+    },
+    {
+      "epoch": 1.8744144546062904,
+      "grad_norm": 0.18657323718070984,
+      "learning_rate": 6.262760988226948e-06,
+      "loss": 0.4782,
+      "step": 8403
+    },
+    {
+      "epoch": 1.87463751951818,
+      "grad_norm": 0.1854049414396286,
+      "learning_rate": 6.260578192020249e-06,
+      "loss": 0.4693,
+      "step": 8404
+    },
+    {
+      "epoch": 1.8748605844300692,
+      "grad_norm": 0.1833823025226593,
+      "learning_rate": 6.2583956029262526e-06,
+      "loss": 0.4455,
+      "step": 8405
+    },
+    {
+      "epoch": 1.8750836493419585,
+      "grad_norm": 0.2055628001689911,
+      "learning_rate": 6.256213221065843e-06,
+      "loss": 0.4378,
+      "step": 8406
+    },
+    {
+      "epoch": 1.8753067142538478,
+      "grad_norm": 0.21103620529174805,
+      "learning_rate": 6.254031046559896e-06,
+      "loss": 0.4435,
+      "step": 8407
+    },
+    {
+      "epoch": 1.8755297791657373,
+      "grad_norm": 0.17623326182365417,
+      "learning_rate": 6.251849079529271e-06,
+      "loss": 0.4526,
+      "step": 8408
+    },
+    {
+      "epoch": 1.8757528440776265,
+      "grad_norm": 0.1766071766614914,
+      "learning_rate": 6.2496673200948225e-06,
+      "loss": 0.4459,
+      "step": 8409
+    },
+    {
+      "epoch": 1.875975908989516,
+      "grad_norm": 0.175866037607193,
+      "learning_rate": 6.247485768377386e-06,
+      "loss": 0.456,
+      "step": 8410
+    },
+    {
+      "epoch": 1.8761989739014053,
+      "grad_norm": 0.17706961929798126,
+      "learning_rate": 6.245304424497792e-06,
+      "loss": 0.4631,
+      "step": 8411
+    },
+    {
+      "epoch": 1.8764220388132946,
+      "grad_norm": 0.17484982311725616,
+      "learning_rate": 6.24312328857686e-06,
+      "loss": 0.4689,
+      "step": 8412
+    },
+    {
+      "epoch": 1.876645103725184,
+      "grad_norm": 0.18998976051807404,
+      "learning_rate": 6.240942360735388e-06,
+      "loss": 0.4627,
+      "step": 8413
+    },
+    {
+      "epoch": 1.8768681686370734,
+      "grad_norm": 0.1831061840057373,
+      "learning_rate": 6.238761641094178e-06,
+      "loss": 0.4503,
+      "step": 8414
+    },
+    {
+      "epoch": 1.877091233548963,
+      "grad_norm": 0.1826712042093277,
+      "learning_rate": 6.236581129774003e-06,
+      "loss": 0.4729,
+      "step": 8415
+    },
+    {
+      "epoch": 1.8773142984608522,
+      "grad_norm": 0.1810642033815384,
+      "learning_rate": 6.234400826895641e-06,
+      "loss": 0.4462,
+      "step": 8416
+    },
+    {
+      "epoch": 1.8775373633727415,
+      "grad_norm": 0.16975516080856323,
+      "learning_rate": 6.232220732579845e-06,
+      "loss": 0.4159,
+      "step": 8417
+    },
+    {
+      "epoch": 1.8777604282846307,
+      "grad_norm": 0.1825772076845169,
+      "learning_rate": 6.230040846947368e-06,
+      "loss": 0.47,
+      "step": 8418
+    },
+    {
+      "epoch": 1.87798349319652,
+      "grad_norm": 0.18529464304447174,
+      "learning_rate": 6.227861170118941e-06,
+      "loss": 0.4473,
+      "step": 8419
+    },
+    {
+      "epoch": 1.8782065581084095,
+      "grad_norm": 0.17210659384727478,
+      "learning_rate": 6.2256817022152915e-06,
+      "loss": 0.4412,
+      "step": 8420
+    },
+    {
+      "epoch": 1.878429623020299,
+      "grad_norm": 0.19293324649333954,
+      "learning_rate": 6.223502443357129e-06,
+      "loss": 0.4592,
+      "step": 8421
+    },
+    {
+      "epoch": 1.8786526879321883,
+      "grad_norm": 0.16933956742286682,
+      "learning_rate": 6.2213233936651586e-06,
+      "loss": 0.433,
+      "step": 8422
+    },
+    {
+      "epoch": 1.8788757528440776,
+      "grad_norm": 0.1740327775478363,
+      "learning_rate": 6.219144553260065e-06,
+      "loss": 0.4452,
+      "step": 8423
+    },
+    {
+      "epoch": 1.8790988177559669,
+      "grad_norm": 0.18426866829395294,
+      "learning_rate": 6.21696592226253e-06,
+      "loss": 0.4658,
+      "step": 8424
+    },
+    {
+      "epoch": 1.8793218826678564,
+      "grad_norm": 0.2005191296339035,
+      "learning_rate": 6.214787500793218e-06,
+      "loss": 0.4762,
+      "step": 8425
+    },
+    {
+      "epoch": 1.8795449475797457,
+      "grad_norm": 0.18926943838596344,
+      "learning_rate": 6.212609288972785e-06,
+      "loss": 0.4584,
+      "step": 8426
+    },
+    {
+      "epoch": 1.8797680124916352,
+      "grad_norm": 0.18263809382915497,
+      "learning_rate": 6.210431286921872e-06,
+      "loss": 0.4344,
+      "step": 8427
+    },
+    {
+      "epoch": 1.8799910774035244,
+      "grad_norm": 0.18587824702262878,
+      "learning_rate": 6.208253494761113e-06,
+      "loss": 0.457,
+      "step": 8428
+    },
+    {
+      "epoch": 1.8802141423154137,
+      "grad_norm": 0.1803131103515625,
+      "learning_rate": 6.2060759126111246e-06,
+      "loss": 0.4588,
+      "step": 8429
+    },
+    {
+      "epoch": 1.880437207227303,
+      "grad_norm": 0.17640437185764313,
+      "learning_rate": 6.203898540592517e-06,
+      "loss": 0.4593,
+      "step": 8430
+    },
+    {
+      "epoch": 1.8806602721391925,
+      "grad_norm": 0.17931349575519562,
+      "learning_rate": 6.201721378825889e-06,
+      "loss": 0.4371,
+      "step": 8431
+    },
+    {
+      "epoch": 1.880883337051082,
+      "grad_norm": 0.19341467320919037,
+      "learning_rate": 6.199544427431821e-06,
+      "loss": 0.4422,
+      "step": 8432
+    },
+    {
+      "epoch": 1.8811064019629713,
+      "grad_norm": 0.17641760408878326,
+      "learning_rate": 6.19736768653089e-06,
+      "loss": 0.4522,
+      "step": 8433
+    },
+    {
+      "epoch": 1.8813294668748606,
+      "grad_norm": 0.18412084877490997,
+      "learning_rate": 6.195191156243654e-06,
+      "loss": 0.4819,
+      "step": 8434
+    },
+    {
+      "epoch": 1.8815525317867499,
+      "grad_norm": 0.18774092197418213,
+      "learning_rate": 6.193014836690667e-06,
+      "loss": 0.4648,
+      "step": 8435
+    },
+    {
+      "epoch": 1.8817755966986391,
+      "grad_norm": 0.1818709373474121,
+      "learning_rate": 6.190838727992463e-06,
+      "loss": 0.4621,
+      "step": 8436
+    },
+    {
+      "epoch": 1.8819986616105286,
+      "grad_norm": 0.1864043027162552,
+      "learning_rate": 6.188662830269574e-06,
+      "loss": 0.4569,
+      "step": 8437
+    },
+    {
+      "epoch": 1.8822217265224181,
+      "grad_norm": 0.18635167181491852,
+      "learning_rate": 6.186487143642508e-06,
+      "loss": 0.4451,
+      "step": 8438
+    },
+    {
+      "epoch": 1.8824447914343074,
+      "grad_norm": 0.18478180468082428,
+      "learning_rate": 6.184311668231775e-06,
+      "loss": 0.4505,
+      "step": 8439
+    },
+    {
+      "epoch": 1.8826678563461967,
+      "grad_norm": 0.1775161474943161,
+      "learning_rate": 6.1821364041578615e-06,
+      "loss": 0.4432,
+      "step": 8440
+    },
+    {
+      "epoch": 1.882890921258086,
+      "grad_norm": 0.18662866950035095,
+      "learning_rate": 6.179961351541252e-06,
+      "loss": 0.4602,
+      "step": 8441
+    },
+    {
+      "epoch": 1.8831139861699755,
+      "grad_norm": 0.1838211715221405,
+      "learning_rate": 6.177786510502408e-06,
+      "loss": 0.4519,
+      "step": 8442
+    },
+    {
+      "epoch": 1.8833370510818648,
+      "grad_norm": 0.1826631873846054,
+      "learning_rate": 6.1756118811617935e-06,
+      "loss": 0.4505,
+      "step": 8443
+    },
+    {
+      "epoch": 1.8835601159937543,
+      "grad_norm": 0.18082208931446075,
+      "learning_rate": 6.173437463639846e-06,
+      "loss": 0.4717,
+      "step": 8444
+    },
+    {
+      "epoch": 1.8837831809056436,
+      "grad_norm": 0.18472546339035034,
+      "learning_rate": 6.171263258057008e-06,
+      "loss": 0.4626,
+      "step": 8445
+    },
+    {
+      "epoch": 1.8840062458175328,
+      "grad_norm": 0.1825518012046814,
+      "learning_rate": 6.16908926453369e-06,
+      "loss": 0.4531,
+      "step": 8446
+    },
+    {
+      "epoch": 1.8842293107294221,
+      "grad_norm": 0.23177875578403473,
+      "learning_rate": 6.166915483190304e-06,
+      "loss": 0.4379,
+      "step": 8447
+    },
+    {
+      "epoch": 1.8844523756413116,
+      "grad_norm": 0.17296954989433289,
+      "learning_rate": 6.164741914147252e-06,
+      "loss": 0.4359,
+      "step": 8448
+    },
+    {
+      "epoch": 1.8846754405532011,
+      "grad_norm": 0.18288472294807434,
+      "learning_rate": 6.162568557524918e-06,
+      "loss": 0.4547,
+      "step": 8449
+    },
+    {
+      "epoch": 1.8848985054650904,
+      "grad_norm": 0.17033147811889648,
+      "learning_rate": 6.160395413443674e-06,
+      "loss": 0.4164,
+      "step": 8450
+    },
+    {
+      "epoch": 1.8851215703769797,
+      "grad_norm": 0.1902470588684082,
+      "learning_rate": 6.1582224820238836e-06,
+      "loss": 0.4366,
+      "step": 8451
+    },
+    {
+      "epoch": 1.885344635288869,
+      "grad_norm": 0.17565543949604034,
+      "learning_rate": 6.156049763385901e-06,
+      "loss": 0.4693,
+      "step": 8452
+    },
+    {
+      "epoch": 1.8855677002007583,
+      "grad_norm": 0.1908363252878189,
+      "learning_rate": 6.1538772576500584e-06,
+      "loss": 0.4484,
+      "step": 8453
+    },
+    {
+      "epoch": 1.8857907651126478,
+      "grad_norm": 0.1861358880996704,
+      "learning_rate": 6.151704964936687e-06,
+      "loss": 0.4531,
+      "step": 8454
+    },
+    {
+      "epoch": 1.8860138300245373,
+      "grad_norm": 0.18762946128845215,
+      "learning_rate": 6.149532885366103e-06,
+      "loss": 0.4676,
+      "step": 8455
+    },
+    {
+      "epoch": 1.8862368949364265,
+      "grad_norm": 0.19913704693317413,
+      "learning_rate": 6.147361019058606e-06,
+      "loss": 0.4429,
+      "step": 8456
+    },
+    {
+      "epoch": 1.8864599598483158,
+      "grad_norm": 0.19710618257522583,
+      "learning_rate": 6.145189366134492e-06,
+      "loss": 0.4722,
+      "step": 8457
+    },
+    {
+      "epoch": 1.886683024760205,
+      "grad_norm": 0.1801493912935257,
+      "learning_rate": 6.143017926714036e-06,
+      "loss": 0.4598,
+      "step": 8458
+    },
+    {
+      "epoch": 1.8869060896720946,
+      "grad_norm": 0.2561694383621216,
+      "learning_rate": 6.140846700917513e-06,
+      "loss": 0.4218,
+      "step": 8459
+    },
+    {
+      "epoch": 1.887129154583984,
+      "grad_norm": 0.17985545098781586,
+      "learning_rate": 6.138675688865171e-06,
+      "loss": 0.4269,
+      "step": 8460
+    },
+    {
+      "epoch": 1.8873522194958734,
+      "grad_norm": 0.1875736266374588,
+      "learning_rate": 6.13650489067726e-06,
+      "loss": 0.4695,
+      "step": 8461
+    },
+    {
+      "epoch": 1.8875752844077627,
+      "grad_norm": 0.18572880327701569,
+      "learning_rate": 6.134334306474009e-06,
+      "loss": 0.4448,
+      "step": 8462
+    },
+    {
+      "epoch": 1.887798349319652,
+      "grad_norm": 0.17786027491092682,
+      "learning_rate": 6.132163936375641e-06,
+      "loss": 0.4505,
+      "step": 8463
+    },
+    {
+      "epoch": 1.8880214142315412,
+      "grad_norm": 0.18378950655460358,
+      "learning_rate": 6.129993780502364e-06,
+      "loss": 0.4583,
+      "step": 8464
+    },
+    {
+      "epoch": 1.8882444791434307,
+      "grad_norm": 0.1798183172941208,
+      "learning_rate": 6.127823838974375e-06,
+      "loss": 0.4491,
+      "step": 8465
+    },
+    {
+      "epoch": 1.8884675440553202,
+      "grad_norm": 0.18183192610740662,
+      "learning_rate": 6.125654111911857e-06,
+      "loss": 0.4707,
+      "step": 8466
+    },
+    {
+      "epoch": 1.8886906089672095,
+      "grad_norm": 0.17841511964797974,
+      "learning_rate": 6.1234845994349875e-06,
+      "loss": 0.4645,
+      "step": 8467
+    },
+    {
+      "epoch": 1.8889136738790988,
+      "grad_norm": 0.17050506174564362,
+      "learning_rate": 6.121315301663923e-06,
+      "loss": 0.4241,
+      "step": 8468
+    },
+    {
+      "epoch": 1.889136738790988,
+      "grad_norm": 0.18663549423217773,
+      "learning_rate": 6.119146218718817e-06,
+      "loss": 0.4736,
+      "step": 8469
+    },
+    {
+      "epoch": 1.8893598037028774,
+      "grad_norm": 0.19147509336471558,
+      "learning_rate": 6.116977350719801e-06,
+      "loss": 0.4557,
+      "step": 8470
+    },
+    {
+      "epoch": 1.8895828686147669,
+      "grad_norm": 0.18618586659431458,
+      "learning_rate": 6.114808697787009e-06,
+      "loss": 0.4469,
+      "step": 8471
+    },
+    {
+      "epoch": 1.8898059335266564,
+      "grad_norm": 0.18382149934768677,
+      "learning_rate": 6.112640260040548e-06,
+      "loss": 0.4903,
+      "step": 8472
+    },
+    {
+      "epoch": 1.8900289984385457,
+      "grad_norm": 0.20091551542282104,
+      "learning_rate": 6.110472037600523e-06,
+      "loss": 0.4578,
+      "step": 8473
+    },
+    {
+      "epoch": 1.890252063350435,
+      "grad_norm": 0.1820792704820633,
+      "learning_rate": 6.10830403058702e-06,
+      "loss": 0.4645,
+      "step": 8474
+    },
+    {
+      "epoch": 1.8904751282623242,
+      "grad_norm": 0.2219545990228653,
+      "learning_rate": 6.106136239120121e-06,
+      "loss": 0.4705,
+      "step": 8475
+    },
+    {
+      "epoch": 1.8906981931742137,
+      "grad_norm": 0.17899756133556366,
+      "learning_rate": 6.103968663319893e-06,
+      "loss": 0.4646,
+      "step": 8476
+    },
+    {
+      "epoch": 1.890921258086103,
+      "grad_norm": 0.17997188866138458,
+      "learning_rate": 6.101801303306383e-06,
+      "loss": 0.4517,
+      "step": 8477
+    },
+    {
+      "epoch": 1.8911443229979925,
+      "grad_norm": 0.16982750594615936,
+      "learning_rate": 6.099634159199641e-06,
+      "loss": 0.4286,
+      "step": 8478
+    },
+    {
+      "epoch": 1.8913673879098818,
+      "grad_norm": 0.17612092196941376,
+      "learning_rate": 6.0974672311196916e-06,
+      "loss": 0.4476,
+      "step": 8479
+    },
+    {
+      "epoch": 1.891590452821771,
+      "grad_norm": 0.18440189957618713,
+      "learning_rate": 6.095300519186555e-06,
+      "loss": 0.4456,
+      "step": 8480
+    },
+    {
+      "epoch": 1.8918135177336604,
+      "grad_norm": 0.16765691339969635,
+      "learning_rate": 6.093134023520237e-06,
+      "loss": 0.4215,
+      "step": 8481
+    },
+    {
+      "epoch": 1.8920365826455499,
+      "grad_norm": 0.1789095401763916,
+      "learning_rate": 6.090967744240734e-06,
+      "loss": 0.4711,
+      "step": 8482
+    },
+    {
+      "epoch": 1.8922596475574394,
+      "grad_norm": 0.18848133087158203,
+      "learning_rate": 6.088801681468024e-06,
+      "loss": 0.4513,
+      "step": 8483
+    },
+    {
+      "epoch": 1.8924827124693286,
+      "grad_norm": 0.17877109348773956,
+      "learning_rate": 6.086635835322081e-06,
+      "loss": 0.4506,
+      "step": 8484
+    },
+    {
+      "epoch": 1.892705777381218,
+      "grad_norm": 0.18696287274360657,
+      "learning_rate": 6.084470205922859e-06,
+      "loss": 0.4523,
+      "step": 8485
+    },
+    {
+      "epoch": 1.8929288422931072,
+      "grad_norm": 0.18571056425571442,
+      "learning_rate": 6.08230479339031e-06,
+      "loss": 0.4646,
+      "step": 8486
+    },
+    {
+      "epoch": 1.8931519072049967,
+      "grad_norm": 0.17447581887245178,
+      "learning_rate": 6.080139597844361e-06,
+      "loss": 0.4671,
+      "step": 8487
+    },
+    {
+      "epoch": 1.893374972116886,
+      "grad_norm": 0.18128472566604614,
+      "learning_rate": 6.0779746194049415e-06,
+      "loss": 0.4419,
+      "step": 8488
+    },
+    {
+      "epoch": 1.8935980370287755,
+      "grad_norm": 0.1792159378528595,
+      "learning_rate": 6.075809858191957e-06,
+      "loss": 0.4547,
+      "step": 8489
+    },
+    {
+      "epoch": 1.8938211019406648,
+      "grad_norm": 0.17733284831047058,
+      "learning_rate": 6.073645314325306e-06,
+      "loss": 0.4177,
+      "step": 8490
+    },
+    {
+      "epoch": 1.894044166852554,
+      "grad_norm": 0.1729053109884262,
+      "learning_rate": 6.071480987924875e-06,
+      "loss": 0.4555,
+      "step": 8491
+    },
+    {
+      "epoch": 1.8942672317644433,
+      "grad_norm": 0.18540537357330322,
+      "learning_rate": 6.069316879110539e-06,
+      "loss": 0.4352,
+      "step": 8492
+    },
+    {
+      "epoch": 1.8944902966763328,
+      "grad_norm": 0.17858169972896576,
+      "learning_rate": 6.0671529880021585e-06,
+      "loss": 0.4707,
+      "step": 8493
+    },
+    {
+      "epoch": 1.8947133615882221,
+      "grad_norm": 0.1803089827299118,
+      "learning_rate": 6.0649893147195845e-06,
+      "loss": 0.4438,
+      "step": 8494
+    },
+    {
+      "epoch": 1.8949364265001116,
+      "grad_norm": 0.18085269629955292,
+      "learning_rate": 6.062825859382655e-06,
+      "loss": 0.4419,
+      "step": 8495
+    },
+    {
+      "epoch": 1.895159491412001,
+      "grad_norm": 0.18501605093479156,
+      "learning_rate": 6.060662622111193e-06,
+      "loss": 0.4698,
+      "step": 8496
+    },
+    {
+      "epoch": 1.8953825563238902,
+      "grad_norm": 0.20104503631591797,
+      "learning_rate": 6.058499603025018e-06,
+      "loss": 0.459,
+      "step": 8497
+    },
+    {
+      "epoch": 1.8956056212357795,
+      "grad_norm": 0.17107710242271423,
+      "learning_rate": 6.056336802243925e-06,
+      "loss": 0.4357,
+      "step": 8498
+    },
+    {
+      "epoch": 1.895828686147669,
+      "grad_norm": 0.18555793166160583,
+      "learning_rate": 6.0541742198877085e-06,
+      "loss": 0.4562,
+      "step": 8499
+    },
+    {
+      "epoch": 1.8960517510595585,
+      "grad_norm": 0.17548789083957672,
+      "learning_rate": 6.052011856076142e-06,
+      "loss": 0.4281,
+      "step": 8500
+    },
+    {
+      "epoch": 1.8962748159714478,
+      "grad_norm": 0.19359424710273743,
+      "learning_rate": 6.049849710928995e-06,
+      "loss": 0.4761,
+      "step": 8501
+    },
+    {
+      "epoch": 1.896497880883337,
+      "grad_norm": 0.1868746131658554,
+      "learning_rate": 6.0476877845660165e-06,
+      "loss": 0.4797,
+      "step": 8502
+    },
+    {
+      "epoch": 1.8967209457952263,
+      "grad_norm": 0.18077945709228516,
+      "learning_rate": 6.045526077106951e-06,
+      "loss": 0.4245,
+      "step": 8503
+    },
+    {
+      "epoch": 1.8969440107071158,
+      "grad_norm": 0.1770927757024765,
+      "learning_rate": 6.0433645886715255e-06,
+      "loss": 0.4483,
+      "step": 8504
+    },
+    {
+      "epoch": 1.897167075619005,
+      "grad_norm": 0.18114930391311646,
+      "learning_rate": 6.041203319379457e-06,
+      "loss": 0.4253,
+      "step": 8505
+    },
+    {
+      "epoch": 1.8973901405308946,
+      "grad_norm": 0.17783895134925842,
+      "learning_rate": 6.039042269350451e-06,
+      "loss": 0.4396,
+      "step": 8506
+    },
+    {
+      "epoch": 1.897613205442784,
+      "grad_norm": 0.18055394291877747,
+      "learning_rate": 6.036881438704198e-06,
+      "loss": 0.4561,
+      "step": 8507
+    },
+    {
+      "epoch": 1.8978362703546732,
+      "grad_norm": 0.17930994927883148,
+      "learning_rate": 6.034720827560381e-06,
+      "loss": 0.4489,
+      "step": 8508
+    },
+    {
+      "epoch": 1.8980593352665625,
+      "grad_norm": 0.18059539794921875,
+      "learning_rate": 6.0325604360386684e-06,
+      "loss": 0.4519,
+      "step": 8509
+    },
+    {
+      "epoch": 1.898282400178452,
+      "grad_norm": 0.19161933660507202,
+      "learning_rate": 6.0304002642587135e-06,
+      "loss": 0.4771,
+      "step": 8510
+    },
+    {
+      "epoch": 1.8985054650903412,
+      "grad_norm": 0.1823475956916809,
+      "learning_rate": 6.028240312340163e-06,
+      "loss": 0.4446,
+      "step": 8511
+    },
+    {
+      "epoch": 1.8987285300022307,
+      "grad_norm": 0.1844417154788971,
+      "learning_rate": 6.026080580402645e-06,
+      "loss": 0.4526,
+      "step": 8512
+    },
+    {
+      "epoch": 1.89895159491412,
+      "grad_norm": 0.1865631490945816,
+      "learning_rate": 6.023921068565783e-06,
+      "loss": 0.4767,
+      "step": 8513
+    },
+    {
+      "epoch": 1.8991746598260093,
+      "grad_norm": 0.18571047484874725,
+      "learning_rate": 6.021761776949182e-06,
+      "loss": 0.4429,
+      "step": 8514
+    },
+    {
+      "epoch": 1.8993977247378986,
+      "grad_norm": 0.18585099279880524,
+      "learning_rate": 6.019602705672441e-06,
+      "loss": 0.4608,
+      "step": 8515
+    },
+    {
+      "epoch": 1.899620789649788,
+      "grad_norm": 0.17889703810214996,
+      "learning_rate": 6.017443854855136e-06,
+      "loss": 0.4121,
+      "step": 8516
+    },
+    {
+      "epoch": 1.8998438545616776,
+      "grad_norm": 0.1884198635816574,
+      "learning_rate": 6.015285224616843e-06,
+      "loss": 0.4274,
+      "step": 8517
+    },
+    {
+      "epoch": 1.9000669194735669,
+      "grad_norm": 0.18577080965042114,
+      "learning_rate": 6.0131268150771194e-06,
+      "loss": 0.4466,
+      "step": 8518
+    },
+    {
+      "epoch": 1.9002899843854562,
+      "grad_norm": 0.1771336793899536,
+      "learning_rate": 6.010968626355509e-06,
+      "loss": 0.4567,
+      "step": 8519
+    },
+    {
+      "epoch": 1.9005130492973454,
+      "grad_norm": 0.18902990221977234,
+      "learning_rate": 6.008810658571551e-06,
+      "loss": 0.4537,
+      "step": 8520
+    },
+    {
+      "epoch": 1.900736114209235,
+      "grad_norm": 0.20060394704341888,
+      "learning_rate": 6.006652911844763e-06,
+      "loss": 0.4792,
+      "step": 8521
+    },
+    {
+      "epoch": 1.9009591791211242,
+      "grad_norm": 0.1900641918182373,
+      "learning_rate": 6.004495386294657e-06,
+      "loss": 0.4775,
+      "step": 8522
+    },
+    {
+      "epoch": 1.9011822440330137,
+      "grad_norm": 0.18330518901348114,
+      "learning_rate": 6.002338082040727e-06,
+      "loss": 0.4522,
+      "step": 8523
+    },
+    {
+      "epoch": 1.901405308944903,
+      "grad_norm": 0.18432478606700897,
+      "learning_rate": 6.000180999202463e-06,
+      "loss": 0.4619,
+      "step": 8524
+    },
+    {
+      "epoch": 1.9016283738567923,
+      "grad_norm": 0.17643876373767853,
+      "learning_rate": 5.998024137899333e-06,
+      "loss": 0.4453,
+      "step": 8525
+    },
+    {
+      "epoch": 1.9018514387686816,
+      "grad_norm": 0.17010371387004852,
+      "learning_rate": 5.995867498250804e-06,
+      "loss": 0.4264,
+      "step": 8526
+    },
+    {
+      "epoch": 1.902074503680571,
+      "grad_norm": 0.17619560658931732,
+      "learning_rate": 5.993711080376317e-06,
+      "loss": 0.4522,
+      "step": 8527
+    },
+    {
+      "epoch": 1.9022975685924604,
+      "grad_norm": 0.18597012758255005,
+      "learning_rate": 5.99155488439531e-06,
+      "loss": 0.4814,
+      "step": 8528
+    },
+    {
+      "epoch": 1.9025206335043499,
+      "grad_norm": 0.20063801109790802,
+      "learning_rate": 5.989398910427209e-06,
+      "loss": 0.4966,
+      "step": 8529
+    },
+    {
+      "epoch": 1.9027436984162391,
+      "grad_norm": 0.17774170637130737,
+      "learning_rate": 5.987243158591422e-06,
+      "loss": 0.436,
+      "step": 8530
+    },
+    {
+      "epoch": 1.9029667633281284,
+      "grad_norm": 0.1837313175201416,
+      "learning_rate": 5.985087629007353e-06,
+      "loss": 0.4648,
+      "step": 8531
+    },
+    {
+      "epoch": 1.9031898282400177,
+      "grad_norm": 0.17951352894306183,
+      "learning_rate": 5.982932321794383e-06,
+      "loss": 0.4547,
+      "step": 8532
+    },
+    {
+      "epoch": 1.9034128931519072,
+      "grad_norm": 0.1875593215227127,
+      "learning_rate": 5.980777237071891e-06,
+      "loss": 0.4525,
+      "step": 8533
+    },
+    {
+      "epoch": 1.9036359580637967,
+      "grad_norm": 0.177789106965065,
+      "learning_rate": 5.978622374959235e-06,
+      "loss": 0.4369,
+      "step": 8534
+    },
+    {
+      "epoch": 1.903859022975686,
+      "grad_norm": 0.1993599534034729,
+      "learning_rate": 5.9764677355757705e-06,
+      "loss": 0.4327,
+      "step": 8535
+    },
+    {
+      "epoch": 1.9040820878875753,
+      "grad_norm": 0.18298867344856262,
+      "learning_rate": 5.97431331904083e-06,
+      "loss": 0.4487,
+      "step": 8536
+    },
+    {
+      "epoch": 1.9043051527994646,
+      "grad_norm": 0.19131208956241608,
+      "learning_rate": 5.972159125473742e-06,
+      "loss": 0.4426,
+      "step": 8537
+    },
+    {
+      "epoch": 1.904528217711354,
+      "grad_norm": 0.1916476935148239,
+      "learning_rate": 5.970005154993816e-06,
+      "loss": 0.4456,
+      "step": 8538
+    },
+    {
+      "epoch": 1.9047512826232433,
+      "grad_norm": 0.18648818135261536,
+      "learning_rate": 5.9678514077203554e-06,
+      "loss": 0.4564,
+      "step": 8539
+    },
+    {
+      "epoch": 1.9049743475351328,
+      "grad_norm": 0.18742208182811737,
+      "learning_rate": 5.965697883772648e-06,
+      "loss": 0.4665,
+      "step": 8540
+    },
+    {
+      "epoch": 1.9051974124470221,
+      "grad_norm": 0.20118087530136108,
+      "learning_rate": 5.963544583269965e-06,
+      "loss": 0.5012,
+      "step": 8541
+    },
+    {
+      "epoch": 1.9054204773589114,
+      "grad_norm": 0.181925430893898,
+      "learning_rate": 5.9613915063315775e-06,
+      "loss": 0.4185,
+      "step": 8542
+    },
+    {
+      "epoch": 1.9056435422708007,
+      "grad_norm": 0.18299917876720428,
+      "learning_rate": 5.9592386530767285e-06,
+      "loss": 0.4633,
+      "step": 8543
+    },
+    {
+      "epoch": 1.9058666071826902,
+      "grad_norm": 0.1903630495071411,
+      "learning_rate": 5.957086023624663e-06,
+      "loss": 0.452,
+      "step": 8544
+    },
+    {
+      "epoch": 1.9060896720945795,
+      "grad_norm": 0.17144551873207092,
+      "learning_rate": 5.954933618094603e-06,
+      "loss": 0.446,
+      "step": 8545
+    },
+    {
+      "epoch": 1.906312737006469,
+      "grad_norm": 0.1903725415468216,
+      "learning_rate": 5.952781436605766e-06,
+      "loss": 0.4643,
+      "step": 8546
+    },
+    {
+      "epoch": 1.9065358019183583,
+      "grad_norm": 0.17307879030704498,
+      "learning_rate": 5.9506294792773475e-06,
+      "loss": 0.4358,
+      "step": 8547
+    },
+    {
+      "epoch": 1.9067588668302475,
+      "grad_norm": 0.18062752485275269,
+      "learning_rate": 5.9484777462285425e-06,
+      "loss": 0.4517,
+      "step": 8548
+    },
+    {
+      "epoch": 1.9069819317421368,
+      "grad_norm": 0.19149060547351837,
+      "learning_rate": 5.946326237578524e-06,
+      "loss": 0.4513,
+      "step": 8549
+    },
+    {
+      "epoch": 1.9072049966540263,
+      "grad_norm": 0.18597029149532318,
+      "learning_rate": 5.944174953446457e-06,
+      "loss": 0.4324,
+      "step": 8550
+    },
+    {
+      "epoch": 1.9074280615659158,
+      "grad_norm": 0.17686405777931213,
+      "learning_rate": 5.942023893951494e-06,
+      "loss": 0.4175,
+      "step": 8551
+    },
+    {
+      "epoch": 1.907651126477805,
+      "grad_norm": 0.17996470630168915,
+      "learning_rate": 5.939873059212775e-06,
+      "loss": 0.4548,
+      "step": 8552
+    },
+    {
+      "epoch": 1.9078741913896944,
+      "grad_norm": 0.18042857944965363,
+      "learning_rate": 5.937722449349421e-06,
+      "loss": 0.438,
+      "step": 8553
+    },
+    {
+      "epoch": 1.9080972563015837,
+      "grad_norm": 0.17731131613254547,
+      "learning_rate": 5.935572064480555e-06,
+      "loss": 0.4301,
+      "step": 8554
+    },
+    {
+      "epoch": 1.9083203212134732,
+      "grad_norm": 0.18161912262439728,
+      "learning_rate": 5.933421904725271e-06,
+      "loss": 0.4805,
+      "step": 8555
+    },
+    {
+      "epoch": 1.9085433861253625,
+      "grad_norm": 0.17587031424045563,
+      "learning_rate": 5.931271970202664e-06,
+      "loss": 0.4361,
+      "step": 8556
+    },
+    {
+      "epoch": 1.908766451037252,
+      "grad_norm": 0.1959368884563446,
+      "learning_rate": 5.929122261031806e-06,
+      "loss": 0.4652,
+      "step": 8557
+    },
+    {
+      "epoch": 1.9089895159491412,
+      "grad_norm": 0.17692820727825165,
+      "learning_rate": 5.926972777331767e-06,
+      "loss": 0.4492,
+      "step": 8558
+    },
+    {
+      "epoch": 1.9092125808610305,
+      "grad_norm": 0.17724066972732544,
+      "learning_rate": 5.924823519221593e-06,
+      "loss": 0.4458,
+      "step": 8559
+    },
+    {
+      "epoch": 1.9094356457729198,
+      "grad_norm": 0.17765964567661285,
+      "learning_rate": 5.922674486820327e-06,
+      "loss": 0.4259,
+      "step": 8560
+    },
+    {
+      "epoch": 1.9096587106848093,
+      "grad_norm": 0.18779602646827698,
+      "learning_rate": 5.920525680246997e-06,
+      "loss": 0.4595,
+      "step": 8561
+    },
+    {
+      "epoch": 1.9098817755966988,
+      "grad_norm": 0.17274630069732666,
+      "learning_rate": 5.918377099620613e-06,
+      "loss": 0.4408,
+      "step": 8562
+    },
+    {
+      "epoch": 1.910104840508588,
+      "grad_norm": 0.1717062145471573,
+      "learning_rate": 5.916228745060184e-06,
+      "loss": 0.4308,
+      "step": 8563
+    },
+    {
+      "epoch": 1.9103279054204774,
+      "grad_norm": 0.18208001554012299,
+      "learning_rate": 5.914080616684691e-06,
+      "loss": 0.4582,
+      "step": 8564
+    },
+    {
+      "epoch": 1.9105509703323666,
+      "grad_norm": 0.18288151919841766,
+      "learning_rate": 5.911932714613118e-06,
+      "loss": 0.4404,
+      "step": 8565
+    },
+    {
+      "epoch": 1.910774035244256,
+      "grad_norm": 0.1811714917421341,
+      "learning_rate": 5.909785038964424e-06,
+      "loss": 0.4544,
+      "step": 8566
+    },
+    {
+      "epoch": 1.9109971001561454,
+      "grad_norm": 0.2010553777217865,
+      "learning_rate": 5.907637589857565e-06,
+      "loss": 0.4497,
+      "step": 8567
+    },
+    {
+      "epoch": 1.911220165068035,
+      "grad_norm": 0.1751619428396225,
+      "learning_rate": 5.905490367411475e-06,
+      "loss": 0.4761,
+      "step": 8568
+    },
+    {
+      "epoch": 1.9114432299799242,
+      "grad_norm": 0.1781492680311203,
+      "learning_rate": 5.903343371745087e-06,
+      "loss": 0.4393,
+      "step": 8569
+    },
+    {
+      "epoch": 1.9116662948918135,
+      "grad_norm": 0.1823146939277649,
+      "learning_rate": 5.9011966029773105e-06,
+      "loss": 0.4438,
+      "step": 8570
+    },
+    {
+      "epoch": 1.9118893598037028,
+      "grad_norm": 0.1732335388660431,
+      "learning_rate": 5.89905006122705e-06,
+      "loss": 0.4574,
+      "step": 8571
+    },
+    {
+      "epoch": 1.9121124247155923,
+      "grad_norm": 0.1782243847846985,
+      "learning_rate": 5.896903746613191e-06,
+      "loss": 0.448,
+      "step": 8572
+    },
+    {
+      "epoch": 1.9123354896274816,
+      "grad_norm": 0.16777978837490082,
+      "learning_rate": 5.894757659254614e-06,
+      "loss": 0.4136,
+      "step": 8573
+    },
+    {
+      "epoch": 1.912558554539371,
+      "grad_norm": 0.18306882679462433,
+      "learning_rate": 5.892611799270179e-06,
+      "loss": 0.472,
+      "step": 8574
+    },
+    {
+      "epoch": 1.9127816194512604,
+      "grad_norm": 0.19164350628852844,
+      "learning_rate": 5.890466166778741e-06,
+      "loss": 0.4304,
+      "step": 8575
+    },
+    {
+      "epoch": 1.9130046843631496,
+      "grad_norm": 0.18419405817985535,
+      "learning_rate": 5.888320761899136e-06,
+      "loss": 0.4458,
+      "step": 8576
+    },
+    {
+      "epoch": 1.913227749275039,
+      "grad_norm": 0.1842220425605774,
+      "learning_rate": 5.886175584750191e-06,
+      "loss": 0.4658,
+      "step": 8577
+    },
+    {
+      "epoch": 1.9134508141869284,
+      "grad_norm": 0.18262866139411926,
+      "learning_rate": 5.884030635450717e-06,
+      "loss": 0.4389,
+      "step": 8578
+    },
+    {
+      "epoch": 1.913673879098818,
+      "grad_norm": 0.17716680467128754,
+      "learning_rate": 5.88188591411952e-06,
+      "loss": 0.4465,
+      "step": 8579
+    },
+    {
+      "epoch": 1.9138969440107072,
+      "grad_norm": 0.18368251621723175,
+      "learning_rate": 5.879741420875382e-06,
+      "loss": 0.4687,
+      "step": 8580
+    },
+    {
+      "epoch": 1.9141200089225965,
+      "grad_norm": 0.18374212086200714,
+      "learning_rate": 5.877597155837082e-06,
+      "loss": 0.4255,
+      "step": 8581
+    },
+    {
+      "epoch": 1.9143430738344858,
+      "grad_norm": 0.2865939736366272,
+      "learning_rate": 5.8754531191233845e-06,
+      "loss": 0.4244,
+      "step": 8582
+    },
+    {
+      "epoch": 1.914566138746375,
+      "grad_norm": 0.16909492015838623,
+      "learning_rate": 5.873309310853037e-06,
+      "loss": 0.4432,
+      "step": 8583
+    },
+    {
+      "epoch": 1.9147892036582645,
+      "grad_norm": 0.18193134665489197,
+      "learning_rate": 5.8711657311447775e-06,
+      "loss": 0.4461,
+      "step": 8584
+    },
+    {
+      "epoch": 1.915012268570154,
+      "grad_norm": 0.1864008605480194,
+      "learning_rate": 5.8690223801173305e-06,
+      "loss": 0.4525,
+      "step": 8585
+    },
+    {
+      "epoch": 1.9152353334820433,
+      "grad_norm": 0.17558574676513672,
+      "learning_rate": 5.866879257889411e-06,
+      "loss": 0.4376,
+      "step": 8586
+    },
+    {
+      "epoch": 1.9154583983939326,
+      "grad_norm": 0.18040230870246887,
+      "learning_rate": 5.864736364579714e-06,
+      "loss": 0.4505,
+      "step": 8587
+    },
+    {
+      "epoch": 1.915681463305822,
+      "grad_norm": 0.1805705428123474,
+      "learning_rate": 5.862593700306931e-06,
+      "loss": 0.4497,
+      "step": 8588
+    },
+    {
+      "epoch": 1.9159045282177114,
+      "grad_norm": 0.17601102590560913,
+      "learning_rate": 5.860451265189733e-06,
+      "loss": 0.4287,
+      "step": 8589
+    },
+    {
+      "epoch": 1.9161275931296007,
+      "grad_norm": 0.19040507078170776,
+      "learning_rate": 5.858309059346784e-06,
+      "loss": 0.4736,
+      "step": 8590
+    },
+    {
+      "epoch": 1.9163506580414902,
+      "grad_norm": 0.18196603655815125,
+      "learning_rate": 5.85616708289673e-06,
+      "loss": 0.4527,
+      "step": 8591
+    },
+    {
+      "epoch": 1.9165737229533795,
+      "grad_norm": 0.19250567257404327,
+      "learning_rate": 5.85402533595821e-06,
+      "loss": 0.4137,
+      "step": 8592
+    },
+    {
+      "epoch": 1.9167967878652687,
+      "grad_norm": 0.1818055659532547,
+      "learning_rate": 5.851883818649845e-06,
+      "loss": 0.4603,
+      "step": 8593
+    },
+    {
+      "epoch": 1.917019852777158,
+      "grad_norm": 0.21804414689540863,
+      "learning_rate": 5.8497425310902474e-06,
+      "loss": 0.4716,
+      "step": 8594
+    },
+    {
+      "epoch": 1.9172429176890475,
+      "grad_norm": 0.19273175299167633,
+      "learning_rate": 5.847601473398014e-06,
+      "loss": 0.4345,
+      "step": 8595
+    },
+    {
+      "epoch": 1.917465982600937,
+      "grad_norm": 0.18502821028232574,
+      "learning_rate": 5.845460645691732e-06,
+      "loss": 0.4634,
+      "step": 8596
+    },
+    {
+      "epoch": 1.9176890475128263,
+      "grad_norm": 0.18303586542606354,
+      "learning_rate": 5.843320048089971e-06,
+      "loss": 0.4813,
+      "step": 8597
+    },
+    {
+      "epoch": 1.9179121124247156,
+      "grad_norm": 0.18128132820129395,
+      "learning_rate": 5.841179680711294e-06,
+      "loss": 0.4677,
+      "step": 8598
+    },
+    {
+      "epoch": 1.9181351773366049,
+      "grad_norm": 0.17710062861442566,
+      "learning_rate": 5.839039543674245e-06,
+      "loss": 0.4321,
+      "step": 8599
+    },
+    {
+      "epoch": 1.9183582422484942,
+      "grad_norm": 0.187991201877594,
+      "learning_rate": 5.836899637097358e-06,
+      "loss": 0.4803,
+      "step": 8600
+    },
+    {
+      "epoch": 1.9185813071603837,
+      "grad_norm": 0.17623665928840637,
+      "learning_rate": 5.834759961099157e-06,
+      "loss": 0.4064,
+      "step": 8601
+    },
+    {
+      "epoch": 1.9188043720722732,
+      "grad_norm": 0.17650528252124786,
+      "learning_rate": 5.832620515798154e-06,
+      "loss": 0.463,
+      "step": 8602
+    },
+    {
+      "epoch": 1.9190274369841624,
+      "grad_norm": 0.17875365912914276,
+      "learning_rate": 5.830481301312836e-06,
+      "loss": 0.4543,
+      "step": 8603
+    },
+    {
+      "epoch": 1.9192505018960517,
+      "grad_norm": 0.2137259989976883,
+      "learning_rate": 5.828342317761692e-06,
+      "loss": 0.4586,
+      "step": 8604
+    },
+    {
+      "epoch": 1.919473566807941,
+      "grad_norm": 0.18043895065784454,
+      "learning_rate": 5.8262035652631896e-06,
+      "loss": 0.4573,
+      "step": 8605
+    },
+    {
+      "epoch": 1.9196966317198305,
+      "grad_norm": 0.18645495176315308,
+      "learning_rate": 5.824065043935794e-06,
+      "loss": 0.4812,
+      "step": 8606
+    },
+    {
+      "epoch": 1.9199196966317198,
+      "grad_norm": 0.18367193639278412,
+      "learning_rate": 5.821926753897942e-06,
+      "loss": 0.4644,
+      "step": 8607
+    },
+    {
+      "epoch": 1.9201427615436093,
+      "grad_norm": 0.17733469605445862,
+      "learning_rate": 5.819788695268065e-06,
+      "loss": 0.4581,
+      "step": 8608
+    },
+    {
+      "epoch": 1.9203658264554986,
+      "grad_norm": 0.17940470576286316,
+      "learning_rate": 5.817650868164584e-06,
+      "loss": 0.4452,
+      "step": 8609
+    },
+    {
+      "epoch": 1.9205888913673879,
+      "grad_norm": 0.1958422213792801,
+      "learning_rate": 5.815513272705905e-06,
+      "loss": 0.47,
+      "step": 8610
+    },
+    {
+      "epoch": 1.9208119562792771,
+      "grad_norm": 0.21605324745178223,
+      "learning_rate": 5.813375909010427e-06,
+      "loss": 0.4444,
+      "step": 8611
+    },
+    {
+      "epoch": 1.9210350211911666,
+      "grad_norm": 0.18394535779953003,
+      "learning_rate": 5.811238777196522e-06,
+      "loss": 0.4527,
+      "step": 8612
+    },
+    {
+      "epoch": 1.9212580861030562,
+      "grad_norm": 0.17584146559238434,
+      "learning_rate": 5.809101877382562e-06,
+      "loss": 0.4181,
+      "step": 8613
+    },
+    {
+      "epoch": 1.9214811510149454,
+      "grad_norm": 0.1733761727809906,
+      "learning_rate": 5.806965209686903e-06,
+      "loss": 0.4431,
+      "step": 8614
+    },
+    {
+      "epoch": 1.9217042159268347,
+      "grad_norm": 0.17333033680915833,
+      "learning_rate": 5.8048287742278886e-06,
+      "loss": 0.4301,
+      "step": 8615
+    },
+    {
+      "epoch": 1.921927280838724,
+      "grad_norm": 0.17815978825092316,
+      "learning_rate": 5.802692571123843e-06,
+      "loss": 0.4417,
+      "step": 8616
+    },
+    {
+      "epoch": 1.9221503457506133,
+      "grad_norm": 0.17262916266918182,
+      "learning_rate": 5.8005566004930826e-06,
+      "loss": 0.424,
+      "step": 8617
+    },
+    {
+      "epoch": 1.9223734106625028,
+      "grad_norm": 0.1765177696943283,
+      "learning_rate": 5.798420862453914e-06,
+      "loss": 0.4553,
+      "step": 8618
+    },
+    {
+      "epoch": 1.9225964755743923,
+      "grad_norm": 0.18109013140201569,
+      "learning_rate": 5.796285357124632e-06,
+      "loss": 0.4362,
+      "step": 8619
+    },
+    {
+      "epoch": 1.9228195404862816,
+      "grad_norm": 0.2021632194519043,
+      "learning_rate": 5.7941500846235045e-06,
+      "loss": 0.4319,
+      "step": 8620
+    },
+    {
+      "epoch": 1.9230426053981708,
+      "grad_norm": 0.18348877131938934,
+      "learning_rate": 5.7920150450688e-06,
+      "loss": 0.4166,
+      "step": 8621
+    },
+    {
+      "epoch": 1.9232656703100601,
+      "grad_norm": 0.18288786709308624,
+      "learning_rate": 5.789880238578773e-06,
+      "loss": 0.4756,
+      "step": 8622
+    },
+    {
+      "epoch": 1.9234887352219496,
+      "grad_norm": 0.2258935421705246,
+      "learning_rate": 5.787745665271664e-06,
+      "loss": 0.4458,
+      "step": 8623
+    },
+    {
+      "epoch": 1.923711800133839,
+      "grad_norm": 0.19700488448143005,
+      "learning_rate": 5.785611325265694e-06,
+      "loss": 0.4566,
+      "step": 8624
+    },
+    {
+      "epoch": 1.9239348650457284,
+      "grad_norm": 0.177838996052742,
+      "learning_rate": 5.783477218679078e-06,
+      "loss": 0.4348,
+      "step": 8625
+    },
+    {
+      "epoch": 1.9241579299576177,
+      "grad_norm": 0.18436270952224731,
+      "learning_rate": 5.781343345630016e-06,
+      "loss": 0.4323,
+      "step": 8626
+    },
+    {
+      "epoch": 1.924380994869507,
+      "grad_norm": 0.18075759708881378,
+      "learning_rate": 5.779209706236696e-06,
+      "loss": 0.4355,
+      "step": 8627
+    },
+    {
+      "epoch": 1.9246040597813963,
+      "grad_norm": 0.18219725787639618,
+      "learning_rate": 5.7770763006172966e-06,
+      "loss": 0.4712,
+      "step": 8628
+    },
+    {
+      "epoch": 1.9248271246932858,
+      "grad_norm": 0.18282842636108398,
+      "learning_rate": 5.77494312888997e-06,
+      "loss": 0.4449,
+      "step": 8629
+    },
+    {
+      "epoch": 1.9250501896051753,
+      "grad_norm": 0.19060496985912323,
+      "learning_rate": 5.77281019117287e-06,
+      "loss": 0.4629,
+      "step": 8630
+    },
+    {
+      "epoch": 1.9252732545170645,
+      "grad_norm": 0.1835882067680359,
+      "learning_rate": 5.770677487584133e-06,
+      "loss": 0.4504,
+      "step": 8631
+    },
+    {
+      "epoch": 1.9254963194289538,
+      "grad_norm": 0.19525527954101562,
+      "learning_rate": 5.7685450182418825e-06,
+      "loss": 0.4748,
+      "step": 8632
+    },
+    {
+      "epoch": 1.925719384340843,
+      "grad_norm": 0.18883006274700165,
+      "learning_rate": 5.7664127832642225e-06,
+      "loss": 0.4342,
+      "step": 8633
+    },
+    {
+      "epoch": 1.9259424492527324,
+      "grad_norm": 0.18440766632556915,
+      "learning_rate": 5.764280782769254e-06,
+      "loss": 0.4448,
+      "step": 8634
+    },
+    {
+      "epoch": 1.926165514164622,
+      "grad_norm": 0.18678854405879974,
+      "learning_rate": 5.762149016875057e-06,
+      "loss": 0.4417,
+      "step": 8635
+    },
+    {
+      "epoch": 1.9263885790765114,
+      "grad_norm": 0.1747465282678604,
+      "learning_rate": 5.760017485699711e-06,
+      "loss": 0.442,
+      "step": 8636
+    },
+    {
+      "epoch": 1.9266116439884007,
+      "grad_norm": 0.17557400465011597,
+      "learning_rate": 5.757886189361261e-06,
+      "loss": 0.4598,
+      "step": 8637
+    },
+    {
+      "epoch": 1.92683470890029,
+      "grad_norm": 0.1763993501663208,
+      "learning_rate": 5.755755127977759e-06,
+      "loss": 0.4279,
+      "step": 8638
+    },
+    {
+      "epoch": 1.9270577738121792,
+      "grad_norm": 0.18249794840812683,
+      "learning_rate": 5.753624301667235e-06,
+      "loss": 0.4529,
+      "step": 8639
+    },
+    {
+      "epoch": 1.9272808387240687,
+      "grad_norm": 0.16952653229236603,
+      "learning_rate": 5.751493710547713e-06,
+      "loss": 0.4208,
+      "step": 8640
+    },
+    {
+      "epoch": 1.927503903635958,
+      "grad_norm": 0.1794244945049286,
+      "learning_rate": 5.749363354737188e-06,
+      "loss": 0.456,
+      "step": 8641
+    },
+    {
+      "epoch": 1.9277269685478475,
+      "grad_norm": 0.1846603900194168,
+      "learning_rate": 5.74723323435366e-06,
+      "loss": 0.4559,
+      "step": 8642
+    },
+    {
+      "epoch": 1.9279500334597368,
+      "grad_norm": 0.18191924691200256,
+      "learning_rate": 5.745103349515104e-06,
+      "loss": 0.4466,
+      "step": 8643
+    },
+    {
+      "epoch": 1.928173098371626,
+      "grad_norm": 0.1772155612707138,
+      "learning_rate": 5.742973700339494e-06,
+      "loss": 0.4005,
+      "step": 8644
+    },
+    {
+      "epoch": 1.9283961632835154,
+      "grad_norm": 0.17686043679714203,
+      "learning_rate": 5.740844286944775e-06,
+      "loss": 0.4279,
+      "step": 8645
+    },
+    {
+      "epoch": 1.9286192281954049,
+      "grad_norm": 0.19069218635559082,
+      "learning_rate": 5.738715109448889e-06,
+      "loss": 0.4725,
+      "step": 8646
+    },
+    {
+      "epoch": 1.9288422931072944,
+      "grad_norm": 0.17691883444786072,
+      "learning_rate": 5.736586167969767e-06,
+      "loss": 0.4383,
+      "step": 8647
+    },
+    {
+      "epoch": 1.9290653580191837,
+      "grad_norm": 0.1703954041004181,
+      "learning_rate": 5.734457462625318e-06,
+      "loss": 0.4225,
+      "step": 8648
+    },
+    {
+      "epoch": 1.929288422931073,
+      "grad_norm": 0.17984159290790558,
+      "learning_rate": 5.732328993533452e-06,
+      "loss": 0.4681,
+      "step": 8649
+    },
+    {
+      "epoch": 1.9295114878429622,
+      "grad_norm": 0.18141353130340576,
+      "learning_rate": 5.730200760812045e-06,
+      "loss": 0.4654,
+      "step": 8650
+    },
+    {
+      "epoch": 1.9297345527548515,
+      "grad_norm": 0.19043731689453125,
+      "learning_rate": 5.72807276457898e-06,
+      "loss": 0.4443,
+      "step": 8651
+    },
+    {
+      "epoch": 1.929957617666741,
+      "grad_norm": 0.6355077624320984,
+      "learning_rate": 5.725945004952118e-06,
+      "loss": 0.4525,
+      "step": 8652
+    },
+    {
+      "epoch": 1.9301806825786305,
+      "grad_norm": 0.17925289273262024,
+      "learning_rate": 5.723817482049307e-06,
+      "loss": 0.4438,
+      "step": 8653
+    },
+    {
+      "epoch": 1.9304037474905198,
+      "grad_norm": 0.1707497388124466,
+      "learning_rate": 5.721690195988381e-06,
+      "loss": 0.4078,
+      "step": 8654
+    },
+    {
+      "epoch": 1.930626812402409,
+      "grad_norm": 0.1779852956533432,
+      "learning_rate": 5.719563146887163e-06,
+      "loss": 0.4304,
+      "step": 8655
+    },
+    {
+      "epoch": 1.9308498773142984,
+      "grad_norm": 0.18374204635620117,
+      "learning_rate": 5.717436334863462e-06,
+      "loss": 0.4503,
+      "step": 8656
+    },
+    {
+      "epoch": 1.9310729422261879,
+      "grad_norm": 0.18949589133262634,
+      "learning_rate": 5.715309760035079e-06,
+      "loss": 0.4901,
+      "step": 8657
+    },
+    {
+      "epoch": 1.9312960071380771,
+      "grad_norm": 0.1836424320936203,
+      "learning_rate": 5.71318342251979e-06,
+      "loss": 0.462,
+      "step": 8658
+    },
+    {
+      "epoch": 1.9315190720499666,
+      "grad_norm": 0.1780422180891037,
+      "learning_rate": 5.711057322435369e-06,
+      "loss": 0.4386,
+      "step": 8659
+    },
+    {
+      "epoch": 1.931742136961856,
+      "grad_norm": 0.17757810652256012,
+      "learning_rate": 5.708931459899571e-06,
+      "loss": 0.4421,
+      "step": 8660
+    },
+    {
+      "epoch": 1.9319652018737452,
+      "grad_norm": 0.1806020736694336,
+      "learning_rate": 5.7068058350301445e-06,
+      "loss": 0.4349,
+      "step": 8661
+    },
+    {
+      "epoch": 1.9321882667856345,
+      "grad_norm": 0.1809801310300827,
+      "learning_rate": 5.704680447944812e-06,
+      "loss": 0.4557,
+      "step": 8662
+    },
+    {
+      "epoch": 1.932411331697524,
+      "grad_norm": 0.22528956830501556,
+      "learning_rate": 5.702555298761297e-06,
+      "loss": 0.4204,
+      "step": 8663
+    },
+    {
+      "epoch": 1.9326343966094135,
+      "grad_norm": 0.17288967967033386,
+      "learning_rate": 5.700430387597299e-06,
+      "loss": 0.4379,
+      "step": 8664
+    },
+    {
+      "epoch": 1.9328574615213028,
+      "grad_norm": 0.18082986772060394,
+      "learning_rate": 5.698305714570517e-06,
+      "loss": 0.4727,
+      "step": 8665
+    },
+    {
+      "epoch": 1.933080526433192,
+      "grad_norm": 0.17780068516731262,
+      "learning_rate": 5.696181279798618e-06,
+      "loss": 0.4321,
+      "step": 8666
+    },
+    {
+      "epoch": 1.9333035913450813,
+      "grad_norm": 0.17463639378547668,
+      "learning_rate": 5.694057083399272e-06,
+      "loss": 0.4647,
+      "step": 8667
+    },
+    {
+      "epoch": 1.9335266562569706,
+      "grad_norm": 0.18342427909374237,
+      "learning_rate": 5.691933125490131e-06,
+      "loss": 0.434,
+      "step": 8668
+    },
+    {
+      "epoch": 1.9337497211688601,
+      "grad_norm": 0.1869836300611496,
+      "learning_rate": 5.689809406188832e-06,
+      "loss": 0.4545,
+      "step": 8669
+    },
+    {
+      "epoch": 1.9339727860807496,
+      "grad_norm": 0.19317130744457245,
+      "learning_rate": 5.687685925613005e-06,
+      "loss": 0.4475,
+      "step": 8670
+    },
+    {
+      "epoch": 1.934195850992639,
+      "grad_norm": 0.18191368877887726,
+      "learning_rate": 5.685562683880252e-06,
+      "loss": 0.4646,
+      "step": 8671
+    },
+    {
+      "epoch": 1.9344189159045282,
+      "grad_norm": 0.1778983622789383,
+      "learning_rate": 5.683439681108176e-06,
+      "loss": 0.4122,
+      "step": 8672
+    },
+    {
+      "epoch": 1.9346419808164175,
+      "grad_norm": 0.18883691728115082,
+      "learning_rate": 5.681316917414363e-06,
+      "loss": 0.4493,
+      "step": 8673
+    },
+    {
+      "epoch": 1.934865045728307,
+      "grad_norm": 0.1791423112154007,
+      "learning_rate": 5.6791943929163895e-06,
+      "loss": 0.4312,
+      "step": 8674
+    },
+    {
+      "epoch": 1.9350881106401963,
+      "grad_norm": 0.1763908416032791,
+      "learning_rate": 5.677072107731805e-06,
+      "loss": 0.4833,
+      "step": 8675
+    },
+    {
+      "epoch": 1.9353111755520858,
+      "grad_norm": 0.17612230777740479,
+      "learning_rate": 5.6749500619781595e-06,
+      "loss": 0.4412,
+      "step": 8676
+    },
+    {
+      "epoch": 1.935534240463975,
+      "grad_norm": 0.1743626892566681,
+      "learning_rate": 5.672828255772986e-06,
+      "loss": 0.4195,
+      "step": 8677
+    },
+    {
+      "epoch": 1.9357573053758643,
+      "grad_norm": 0.18442870676517487,
+      "learning_rate": 5.6707066892338054e-06,
+      "loss": 0.4564,
+      "step": 8678
+    },
+    {
+      "epoch": 1.9359803702877536,
+      "grad_norm": 0.2595793306827545,
+      "learning_rate": 5.668585362478117e-06,
+      "loss": 0.4738,
+      "step": 8679
+    },
+    {
+      "epoch": 1.936203435199643,
+      "grad_norm": 0.19340401887893677,
+      "learning_rate": 5.666464275623418e-06,
+      "loss": 0.4409,
+      "step": 8680
+    },
+    {
+      "epoch": 1.9364265001115326,
+      "grad_norm": 0.17830665409564972,
+      "learning_rate": 5.664343428787185e-06,
+      "loss": 0.4476,
+      "step": 8681
+    },
+    {
+      "epoch": 1.936649565023422,
+      "grad_norm": 0.18472926318645477,
+      "learning_rate": 5.66222282208689e-06,
+      "loss": 0.4597,
+      "step": 8682
+    },
+    {
+      "epoch": 1.9368726299353112,
+      "grad_norm": 0.18445588648319244,
+      "learning_rate": 5.660102455639978e-06,
+      "loss": 0.4724,
+      "step": 8683
+    },
+    {
+      "epoch": 1.9370956948472005,
+      "grad_norm": 0.20936332643032074,
+      "learning_rate": 5.657982329563889e-06,
+      "loss": 0.4823,
+      "step": 8684
+    },
+    {
+      "epoch": 1.9373187597590897,
+      "grad_norm": 0.17280444502830505,
+      "learning_rate": 5.655862443976051e-06,
+      "loss": 0.4476,
+      "step": 8685
+    },
+    {
+      "epoch": 1.9375418246709792,
+      "grad_norm": 0.18051601946353912,
+      "learning_rate": 5.653742798993877e-06,
+      "loss": 0.4426,
+      "step": 8686
+    },
+    {
+      "epoch": 1.9377648895828687,
+      "grad_norm": 0.18293160200119019,
+      "learning_rate": 5.651623394734771e-06,
+      "loss": 0.4589,
+      "step": 8687
+    },
+    {
+      "epoch": 1.937987954494758,
+      "grad_norm": 0.18495197594165802,
+      "learning_rate": 5.649504231316109e-06,
+      "loss": 0.463,
+      "step": 8688
+    },
+    {
+      "epoch": 1.9382110194066473,
+      "grad_norm": 0.18280479311943054,
+      "learning_rate": 5.647385308855271e-06,
+      "loss": 0.457,
+      "step": 8689
+    },
+    {
+      "epoch": 1.9384340843185366,
+      "grad_norm": 0.1783073991537094,
+      "learning_rate": 5.645266627469612e-06,
+      "loss": 0.4672,
+      "step": 8690
+    },
+    {
+      "epoch": 1.938657149230426,
+      "grad_norm": 0.18390284478664398,
+      "learning_rate": 5.643148187276477e-06,
+      "loss": 0.47,
+      "step": 8691
+    },
+    {
+      "epoch": 1.9388802141423154,
+      "grad_norm": 0.176648810505867,
+      "learning_rate": 5.641029988393208e-06,
+      "loss": 0.4732,
+      "step": 8692
+    },
+    {
+      "epoch": 1.9391032790542049,
+      "grad_norm": 0.19437986612319946,
+      "learning_rate": 5.638912030937112e-06,
+      "loss": 0.4631,
+      "step": 8693
+    },
+    {
+      "epoch": 1.9393263439660942,
+      "grad_norm": 0.1740999072790146,
+      "learning_rate": 5.636794315025499e-06,
+      "loss": 0.4326,
+      "step": 8694
+    },
+    {
+      "epoch": 1.9395494088779834,
+      "grad_norm": 0.17980170249938965,
+      "learning_rate": 5.634676840775664e-06,
+      "loss": 0.4373,
+      "step": 8695
+    },
+    {
+      "epoch": 1.9397724737898727,
+      "grad_norm": 0.1797953099012375,
+      "learning_rate": 5.632559608304886e-06,
+      "loss": 0.4681,
+      "step": 8696
+    },
+    {
+      "epoch": 1.9399955387017622,
+      "grad_norm": 0.17781023681163788,
+      "learning_rate": 5.630442617730427e-06,
+      "loss": 0.447,
+      "step": 8697
+    },
+    {
+      "epoch": 1.9402186036136517,
+      "grad_norm": 0.19193176925182343,
+      "learning_rate": 5.628325869169542e-06,
+      "loss": 0.4399,
+      "step": 8698
+    },
+    {
+      "epoch": 1.940441668525541,
+      "grad_norm": 0.18457980453968048,
+      "learning_rate": 5.62620936273947e-06,
+      "loss": 0.4717,
+      "step": 8699
+    },
+    {
+      "epoch": 1.9406647334374303,
+      "grad_norm": 0.1887241154909134,
+      "learning_rate": 5.624093098557437e-06,
+      "loss": 0.4239,
+      "step": 8700
+    },
+    {
+      "epoch": 1.9408877983493196,
+      "grad_norm": 0.18782730400562286,
+      "learning_rate": 5.621977076740652e-06,
+      "loss": 0.4417,
+      "step": 8701
+    },
+    {
+      "epoch": 1.9411108632612089,
+      "grad_norm": 0.21468473970890045,
+      "learning_rate": 5.619861297406315e-06,
+      "loss": 0.4635,
+      "step": 8702
+    },
+    {
+      "epoch": 1.9413339281730984,
+      "grad_norm": 0.18426984548568726,
+      "learning_rate": 5.61774576067161e-06,
+      "loss": 0.435,
+      "step": 8703
+    },
+    {
+      "epoch": 1.9415569930849879,
+      "grad_norm": 0.17525528371334076,
+      "learning_rate": 5.6156304666537166e-06,
+      "loss": 0.4459,
+      "step": 8704
+    },
+    {
+      "epoch": 1.9417800579968771,
+      "grad_norm": 0.17843464016914368,
+      "learning_rate": 5.613515415469782e-06,
+      "loss": 0.4455,
+      "step": 8705
+    },
+    {
+      "epoch": 1.9420031229087664,
+      "grad_norm": 0.1760268360376358,
+      "learning_rate": 5.6114006072369544e-06,
+      "loss": 0.4394,
+      "step": 8706
+    },
+    {
+      "epoch": 1.9422261878206557,
+      "grad_norm": 0.18088506162166595,
+      "learning_rate": 5.609286042072366e-06,
+      "loss": 0.4819,
+      "step": 8707
+    },
+    {
+      "epoch": 1.9424492527325452,
+      "grad_norm": 0.17896127700805664,
+      "learning_rate": 5.607171720093141e-06,
+      "loss": 0.4474,
+      "step": 8708
+    },
+    {
+      "epoch": 1.9426723176444345,
+      "grad_norm": 0.1753784865140915,
+      "learning_rate": 5.605057641416374e-06,
+      "loss": 0.4743,
+      "step": 8709
+    },
+    {
+      "epoch": 1.942895382556324,
+      "grad_norm": 0.1875680536031723,
+      "learning_rate": 5.60294380615916e-06,
+      "loss": 0.4491,
+      "step": 8710
+    },
+    {
+      "epoch": 1.9431184474682133,
+      "grad_norm": 0.18972131609916687,
+      "learning_rate": 5.600830214438577e-06,
+      "loss": 0.4332,
+      "step": 8711
+    },
+    {
+      "epoch": 1.9433415123801026,
+      "grad_norm": 0.25385287404060364,
+      "learning_rate": 5.598716866371688e-06,
+      "loss": 0.4125,
+      "step": 8712
+    },
+    {
+      "epoch": 1.9435645772919918,
+      "grad_norm": 0.19474051892757416,
+      "learning_rate": 5.5966037620755475e-06,
+      "loss": 0.4405,
+      "step": 8713
+    },
+    {
+      "epoch": 1.9437876422038813,
+      "grad_norm": 0.18606369197368622,
+      "learning_rate": 5.594490901667187e-06,
+      "loss": 0.4364,
+      "step": 8714
+    },
+    {
+      "epoch": 1.9440107071157708,
+      "grad_norm": 0.17978952825069427,
+      "learning_rate": 5.592378285263631e-06,
+      "loss": 0.446,
+      "step": 8715
+    },
+    {
+      "epoch": 1.9442337720276601,
+      "grad_norm": 0.17850260436534882,
+      "learning_rate": 5.5902659129818896e-06,
+      "loss": 0.4761,
+      "step": 8716
+    },
+    {
+      "epoch": 1.9444568369395494,
+      "grad_norm": 0.1837419867515564,
+      "learning_rate": 5.588153784938966e-06,
+      "loss": 0.4414,
+      "step": 8717
+    },
+    {
+      "epoch": 1.9446799018514387,
+      "grad_norm": 0.18115438520908356,
+      "learning_rate": 5.586041901251832e-06,
+      "loss": 0.4248,
+      "step": 8718
+    },
+    {
+      "epoch": 1.944902966763328,
+      "grad_norm": 0.18175376951694489,
+      "learning_rate": 5.5839302620374625e-06,
+      "loss": 0.4587,
+      "step": 8719
+    },
+    {
+      "epoch": 1.9451260316752175,
+      "grad_norm": 0.18459230661392212,
+      "learning_rate": 5.581818867412814e-06,
+      "loss": 0.4576,
+      "step": 8720
+    },
+    {
+      "epoch": 1.945349096587107,
+      "grad_norm": 0.17397940158843994,
+      "learning_rate": 5.5797077174948325e-06,
+      "loss": 0.4568,
+      "step": 8721
+    },
+    {
+      "epoch": 1.9455721614989963,
+      "grad_norm": 0.18246999382972717,
+      "learning_rate": 5.577596812400438e-06,
+      "loss": 0.4431,
+      "step": 8722
+    },
+    {
+      "epoch": 1.9457952264108855,
+      "grad_norm": 0.18297289311885834,
+      "learning_rate": 5.57548615224655e-06,
+      "loss": 0.4615,
+      "step": 8723
+    },
+    {
+      "epoch": 1.9460182913227748,
+      "grad_norm": 0.17768624424934387,
+      "learning_rate": 5.57337573715007e-06,
+      "loss": 0.45,
+      "step": 8724
+    },
+    {
+      "epoch": 1.9462413562346643,
+      "grad_norm": 0.17627879977226257,
+      "learning_rate": 5.5712655672278905e-06,
+      "loss": 0.4253,
+      "step": 8725
+    },
+    {
+      "epoch": 1.9464644211465536,
+      "grad_norm": 0.1850820928812027,
+      "learning_rate": 5.569155642596878e-06,
+      "loss": 0.4427,
+      "step": 8726
+    },
+    {
+      "epoch": 1.946687486058443,
+      "grad_norm": 0.2031596601009369,
+      "learning_rate": 5.567045963373896e-06,
+      "loss": 0.5089,
+      "step": 8727
+    },
+    {
+      "epoch": 1.9469105509703324,
+      "grad_norm": 0.17893202602863312,
+      "learning_rate": 5.564936529675793e-06,
+      "loss": 0.4536,
+      "step": 8728
+    },
+    {
+      "epoch": 1.9471336158822217,
+      "grad_norm": 0.19759956002235413,
+      "learning_rate": 5.562827341619407e-06,
+      "loss": 0.458,
+      "step": 8729
+    },
+    {
+      "epoch": 1.947356680794111,
+      "grad_norm": 0.18113680183887482,
+      "learning_rate": 5.560718399321549e-06,
+      "loss": 0.4763,
+      "step": 8730
+    },
+    {
+      "epoch": 1.9475797457060005,
+      "grad_norm": 0.1818569451570511,
+      "learning_rate": 5.558609702899032e-06,
+      "loss": 0.4484,
+      "step": 8731
+    },
+    {
+      "epoch": 1.94780281061789,
+      "grad_norm": 0.1860044002532959,
+      "learning_rate": 5.556501252468647e-06,
+      "loss": 0.4312,
+      "step": 8732
+    },
+    {
+      "epoch": 1.9480258755297792,
+      "grad_norm": 0.18196377158164978,
+      "learning_rate": 5.554393048147172e-06,
+      "loss": 0.4708,
+      "step": 8733
+    },
+    {
+      "epoch": 1.9482489404416685,
+      "grad_norm": 0.18474991619586945,
+      "learning_rate": 5.552285090051379e-06,
+      "loss": 0.4675,
+      "step": 8734
+    },
+    {
+      "epoch": 1.9484720053535578,
+      "grad_norm": 0.2845540940761566,
+      "learning_rate": 5.55017737829801e-06,
+      "loss": 0.4609,
+      "step": 8735
+    },
+    {
+      "epoch": 1.948695070265447,
+      "grad_norm": 0.17991533875465393,
+      "learning_rate": 5.548069913003808e-06,
+      "loss": 0.4612,
+      "step": 8736
+    },
+    {
+      "epoch": 1.9489181351773366,
+      "grad_norm": 0.1794772446155548,
+      "learning_rate": 5.5459626942855e-06,
+      "loss": 0.429,
+      "step": 8737
+    },
+    {
+      "epoch": 1.949141200089226,
+      "grad_norm": 0.18330740928649902,
+      "learning_rate": 5.543855722259797e-06,
+      "loss": 0.4359,
+      "step": 8738
+    },
+    {
+      "epoch": 1.9493642650011154,
+      "grad_norm": 0.17791210114955902,
+      "learning_rate": 5.541748997043392e-06,
+      "loss": 0.4532,
+      "step": 8739
+    },
+    {
+      "epoch": 1.9495873299130047,
+      "grad_norm": 0.18846653401851654,
+      "learning_rate": 5.5396425187529705e-06,
+      "loss": 0.4784,
+      "step": 8740
+    },
+    {
+      "epoch": 1.949810394824894,
+      "grad_norm": 0.18275344371795654,
+      "learning_rate": 5.537536287505203e-06,
+      "loss": 0.4651,
+      "step": 8741
+    },
+    {
+      "epoch": 1.9500334597367834,
+      "grad_norm": 0.28151261806488037,
+      "learning_rate": 5.5354303034167486e-06,
+      "loss": 0.4302,
+      "step": 8742
+    },
+    {
+      "epoch": 1.9502565246486727,
+      "grad_norm": 0.1811457872390747,
+      "learning_rate": 5.533324566604246e-06,
+      "loss": 0.4572,
+      "step": 8743
+    },
+    {
+      "epoch": 1.9504795895605622,
+      "grad_norm": 0.18251028656959534,
+      "learning_rate": 5.531219077184322e-06,
+      "loss": 0.4651,
+      "step": 8744
+    },
+    {
+      "epoch": 1.9507026544724515,
+      "grad_norm": 0.18917511403560638,
+      "learning_rate": 5.529113835273599e-06,
+      "loss": 0.431,
+      "step": 8745
+    },
+    {
+      "epoch": 1.9509257193843408,
+      "grad_norm": 0.19956184923648834,
+      "learning_rate": 5.527008840988676e-06,
+      "loss": 0.4789,
+      "step": 8746
+    },
+    {
+      "epoch": 1.95114878429623,
+      "grad_norm": 0.17988136410713196,
+      "learning_rate": 5.5249040944461366e-06,
+      "loss": 0.4427,
+      "step": 8747
+    },
+    {
+      "epoch": 1.9513718492081196,
+      "grad_norm": 0.18015359342098236,
+      "learning_rate": 5.522799595762558e-06,
+      "loss": 0.4454,
+      "step": 8748
+    },
+    {
+      "epoch": 1.951594914120009,
+      "grad_norm": 0.17955633997917175,
+      "learning_rate": 5.5206953450545e-06,
+      "loss": 0.4477,
+      "step": 8749
+    },
+    {
+      "epoch": 1.9518179790318984,
+      "grad_norm": 0.18171581625938416,
+      "learning_rate": 5.518591342438513e-06,
+      "loss": 0.4489,
+      "step": 8750
+    },
+    {
+      "epoch": 1.9520410439437876,
+      "grad_norm": 0.1831309199333191,
+      "learning_rate": 5.516487588031124e-06,
+      "loss": 0.4686,
+      "step": 8751
+    },
+    {
+      "epoch": 1.952264108855677,
+      "grad_norm": 0.1874886304140091,
+      "learning_rate": 5.514384081948855e-06,
+      "loss": 0.4631,
+      "step": 8752
+    },
+    {
+      "epoch": 1.9524871737675662,
+      "grad_norm": 0.19060033559799194,
+      "learning_rate": 5.51228082430821e-06,
+      "loss": 0.4694,
+      "step": 8753
+    },
+    {
+      "epoch": 1.9527102386794557,
+      "grad_norm": 0.18917007744312286,
+      "learning_rate": 5.510177815225683e-06,
+      "loss": 0.4681,
+      "step": 8754
+    },
+    {
+      "epoch": 1.9529333035913452,
+      "grad_norm": 0.1917768120765686,
+      "learning_rate": 5.508075054817755e-06,
+      "loss": 0.4628,
+      "step": 8755
+    },
+    {
+      "epoch": 1.9531563685032345,
+      "grad_norm": 0.18906119465827942,
+      "learning_rate": 5.505972543200879e-06,
+      "loss": 0.4553,
+      "step": 8756
+    },
+    {
+      "epoch": 1.9533794334151238,
+      "grad_norm": 0.18696992099285126,
+      "learning_rate": 5.503870280491514e-06,
+      "loss": 0.446,
+      "step": 8757
+    },
+    {
+      "epoch": 1.953602498327013,
+      "grad_norm": 0.18778546154499054,
+      "learning_rate": 5.501768266806095e-06,
+      "loss": 0.4418,
+      "step": 8758
+    },
+    {
+      "epoch": 1.9538255632389026,
+      "grad_norm": 0.182692751288414,
+      "learning_rate": 5.4996665022610495e-06,
+      "loss": 0.4688,
+      "step": 8759
+    },
+    {
+      "epoch": 1.9540486281507918,
+      "grad_norm": 0.20348599553108215,
+      "learning_rate": 5.497564986972775e-06,
+      "loss": 0.4815,
+      "step": 8760
+    },
+    {
+      "epoch": 1.9542716930626813,
+      "grad_norm": 0.1796850711107254,
+      "learning_rate": 5.495463721057675e-06,
+      "loss": 0.4688,
+      "step": 8761
+    },
+    {
+      "epoch": 1.9544947579745706,
+      "grad_norm": 0.1841420829296112,
+      "learning_rate": 5.493362704632127e-06,
+      "loss": 0.4444,
+      "step": 8762
+    },
+    {
+      "epoch": 1.95471782288646,
+      "grad_norm": 0.20517012476921082,
+      "learning_rate": 5.491261937812504e-06,
+      "loss": 0.4471,
+      "step": 8763
+    },
+    {
+      "epoch": 1.9549408877983492,
+      "grad_norm": 0.1862691342830658,
+      "learning_rate": 5.489161420715154e-06,
+      "loss": 0.4669,
+      "step": 8764
+    },
+    {
+      "epoch": 1.9551639527102387,
+      "grad_norm": 0.1903495490550995,
+      "learning_rate": 5.487061153456418e-06,
+      "loss": 0.4476,
+      "step": 8765
+    },
+    {
+      "epoch": 1.9553870176221282,
+      "grad_norm": 0.17540110647678375,
+      "learning_rate": 5.484961136152622e-06,
+      "loss": 0.4416,
+      "step": 8766
+    },
+    {
+      "epoch": 1.9556100825340175,
+      "grad_norm": 0.18369555473327637,
+      "learning_rate": 5.482861368920084e-06,
+      "loss": 0.4645,
+      "step": 8767
+    },
+    {
+      "epoch": 1.9558331474459068,
+      "grad_norm": 0.17829713225364685,
+      "learning_rate": 5.4807618518750925e-06,
+      "loss": 0.4156,
+      "step": 8768
+    },
+    {
+      "epoch": 1.956056212357796,
+      "grad_norm": 0.17648525536060333,
+      "learning_rate": 5.478662585133938e-06,
+      "loss": 0.4493,
+      "step": 8769
+    },
+    {
+      "epoch": 1.9562792772696853,
+      "grad_norm": 0.18088924884796143,
+      "learning_rate": 5.476563568812893e-06,
+      "loss": 0.455,
+      "step": 8770
+    },
+    {
+      "epoch": 1.9565023421815748,
+      "grad_norm": 0.18878251314163208,
+      "learning_rate": 5.474464803028206e-06,
+      "loss": 0.4492,
+      "step": 8771
+    },
+    {
+      "epoch": 1.9567254070934643,
+      "grad_norm": 0.19534710049629211,
+      "learning_rate": 5.472366287896132e-06,
+      "loss": 0.4368,
+      "step": 8772
+    },
+    {
+      "epoch": 1.9569484720053536,
+      "grad_norm": 0.1891869604587555,
+      "learning_rate": 5.470268023532885e-06,
+      "loss": 0.4573,
+      "step": 8773
+    },
+    {
+      "epoch": 1.9571715369172429,
+      "grad_norm": 0.1803818792104721,
+      "learning_rate": 5.468170010054691e-06,
+      "loss": 0.4338,
+      "step": 8774
+    },
+    {
+      "epoch": 1.9573946018291322,
+      "grad_norm": 0.17830346524715424,
+      "learning_rate": 5.4660722475777475e-06,
+      "loss": 0.4367,
+      "step": 8775
+    },
+    {
+      "epoch": 1.9576176667410217,
+      "grad_norm": 0.1889098584651947,
+      "learning_rate": 5.463974736218243e-06,
+      "loss": 0.4494,
+      "step": 8776
+    },
+    {
+      "epoch": 1.957840731652911,
+      "grad_norm": 0.1826130598783493,
+      "learning_rate": 5.461877476092352e-06,
+      "loss": 0.4531,
+      "step": 8777
+    },
+    {
+      "epoch": 1.9580637965648005,
+      "grad_norm": 0.17219486832618713,
+      "learning_rate": 5.4597804673162315e-06,
+      "loss": 0.4272,
+      "step": 8778
+    },
+    {
+      "epoch": 1.9582868614766897,
+      "grad_norm": 0.18349269032478333,
+      "learning_rate": 5.457683710006026e-06,
+      "loss": 0.4443,
+      "step": 8779
+    },
+    {
+      "epoch": 1.958509926388579,
+      "grad_norm": 0.1871849000453949,
+      "learning_rate": 5.455587204277868e-06,
+      "loss": 0.4588,
+      "step": 8780
+    },
+    {
+      "epoch": 1.9587329913004683,
+      "grad_norm": 0.18738840520381927,
+      "learning_rate": 5.453490950247882e-06,
+      "loss": 0.478,
+      "step": 8781
+    },
+    {
+      "epoch": 1.9589560562123578,
+      "grad_norm": 0.17778688669204712,
+      "learning_rate": 5.45139494803216e-06,
+      "loss": 0.428,
+      "step": 8782
+    },
+    {
+      "epoch": 1.9591791211242473,
+      "grad_norm": 0.18447647988796234,
+      "learning_rate": 5.4492991977468e-06,
+      "loss": 0.4422,
+      "step": 8783
+    },
+    {
+      "epoch": 1.9594021860361366,
+      "grad_norm": 0.19066467881202698,
+      "learning_rate": 5.4472036995078734e-06,
+      "loss": 0.4328,
+      "step": 8784
+    },
+    {
+      "epoch": 1.9596252509480259,
+      "grad_norm": 0.171788290143013,
+      "learning_rate": 5.445108453431449e-06,
+      "loss": 0.428,
+      "step": 8785
+    },
+    {
+      "epoch": 1.9598483158599151,
+      "grad_norm": 0.1862727701663971,
+      "learning_rate": 5.443013459633566e-06,
+      "loss": 0.4497,
+      "step": 8786
+    },
+    {
+      "epoch": 1.9600713807718046,
+      "grad_norm": 0.19083817303180695,
+      "learning_rate": 5.440918718230263e-06,
+      "loss": 0.4435,
+      "step": 8787
+    },
+    {
+      "epoch": 1.960294445683694,
+      "grad_norm": 0.18043813109397888,
+      "learning_rate": 5.438824229337558e-06,
+      "loss": 0.4386,
+      "step": 8788
+    },
+    {
+      "epoch": 1.9605175105955834,
+      "grad_norm": 0.17163397371768951,
+      "learning_rate": 5.4367299930714615e-06,
+      "loss": 0.4406,
+      "step": 8789
+    },
+    {
+      "epoch": 1.9607405755074727,
+      "grad_norm": 0.17972704768180847,
+      "learning_rate": 5.43463600954796e-06,
+      "loss": 0.4606,
+      "step": 8790
+    },
+    {
+      "epoch": 1.960963640419362,
+      "grad_norm": 0.1928553283214569,
+      "learning_rate": 5.432542278883031e-06,
+      "loss": 0.4587,
+      "step": 8791
+    },
+    {
+      "epoch": 1.9611867053312513,
+      "grad_norm": 0.18378639221191406,
+      "learning_rate": 5.430448801192642e-06,
+      "loss": 0.4554,
+      "step": 8792
+    },
+    {
+      "epoch": 1.9614097702431408,
+      "grad_norm": 0.18218287825584412,
+      "learning_rate": 5.428355576592748e-06,
+      "loss": 0.4634,
+      "step": 8793
+    },
+    {
+      "epoch": 1.96163283515503,
+      "grad_norm": 0.18878161907196045,
+      "learning_rate": 5.426262605199273e-06,
+      "loss": 0.4567,
+      "step": 8794
+    },
+    {
+      "epoch": 1.9618559000669196,
+      "grad_norm": 0.17760403454303741,
+      "learning_rate": 5.424169887128145e-06,
+      "loss": 0.4414,
+      "step": 8795
+    },
+    {
+      "epoch": 1.9620789649788088,
+      "grad_norm": 0.18763794004917145,
+      "learning_rate": 5.4220774224952735e-06,
+      "loss": 0.4499,
+      "step": 8796
+    },
+    {
+      "epoch": 1.9623020298906981,
+      "grad_norm": 0.18167226016521454,
+      "learning_rate": 5.419985211416548e-06,
+      "loss": 0.4325,
+      "step": 8797
+    },
+    {
+      "epoch": 1.9625250948025874,
+      "grad_norm": 0.18740572035312653,
+      "learning_rate": 5.417893254007857e-06,
+      "loss": 0.453,
+      "step": 8798
+    },
+    {
+      "epoch": 1.962748159714477,
+      "grad_norm": 0.18634778261184692,
+      "learning_rate": 5.415801550385056e-06,
+      "loss": 0.4658,
+      "step": 8799
+    },
+    {
+      "epoch": 1.9629712246263664,
+      "grad_norm": 0.1778239905834198,
+      "learning_rate": 5.4137101006639994e-06,
+      "loss": 0.4438,
+      "step": 8800
+    },
+    {
+      "epoch": 1.9631942895382557,
+      "grad_norm": 0.17299148440361023,
+      "learning_rate": 5.411618904960527e-06,
+      "loss": 0.4038,
+      "step": 8801
+    },
+    {
+      "epoch": 1.963417354450145,
+      "grad_norm": 0.17823709547519684,
+      "learning_rate": 5.4095279633904665e-06,
+      "loss": 0.412,
+      "step": 8802
+    },
+    {
+      "epoch": 1.9636404193620343,
+      "grad_norm": 0.18503381311893463,
+      "learning_rate": 5.407437276069618e-06,
+      "loss": 0.4553,
+      "step": 8803
+    },
+    {
+      "epoch": 1.9638634842739238,
+      "grad_norm": 0.17986641824245453,
+      "learning_rate": 5.4053468431137814e-06,
+      "loss": 0.4714,
+      "step": 8804
+    },
+    {
+      "epoch": 1.964086549185813,
+      "grad_norm": 0.17366492748260498,
+      "learning_rate": 5.403256664638738e-06,
+      "loss": 0.4264,
+      "step": 8805
+    },
+    {
+      "epoch": 1.9643096140977025,
+      "grad_norm": 0.18994669616222382,
+      "learning_rate": 5.40116674076026e-06,
+      "loss": 0.447,
+      "step": 8806
+    },
+    {
+      "epoch": 1.9645326790095918,
+      "grad_norm": 0.17680832743644714,
+      "learning_rate": 5.3990770715940895e-06,
+      "loss": 0.4308,
+      "step": 8807
+    },
+    {
+      "epoch": 1.9647557439214811,
+      "grad_norm": 0.17399604618549347,
+      "learning_rate": 5.396987657255974e-06,
+      "loss": 0.4405,
+      "step": 8808
+    },
+    {
+      "epoch": 1.9649788088333704,
+      "grad_norm": 0.21814000606536865,
+      "learning_rate": 5.3948984978616355e-06,
+      "loss": 0.4643,
+      "step": 8809
+    },
+    {
+      "epoch": 1.96520187374526,
+      "grad_norm": 0.17908701300621033,
+      "learning_rate": 5.39280959352679e-06,
+      "loss": 0.4189,
+      "step": 8810
+    },
+    {
+      "epoch": 1.9654249386571492,
+      "grad_norm": 0.17751288414001465,
+      "learning_rate": 5.390720944367125e-06,
+      "loss": 0.4502,
+      "step": 8811
+    },
+    {
+      "epoch": 1.9656480035690387,
+      "grad_norm": 0.21767963469028473,
+      "learning_rate": 5.388632550498327e-06,
+      "loss": 0.4495,
+      "step": 8812
+    },
+    {
+      "epoch": 1.965871068480928,
+      "grad_norm": 0.1864757090806961,
+      "learning_rate": 5.386544412036066e-06,
+      "loss": 0.4775,
+      "step": 8813
+    },
+    {
+      "epoch": 1.9660941333928172,
+      "grad_norm": 0.17667576670646667,
+      "learning_rate": 5.384456529095999e-06,
+      "loss": 0.4515,
+      "step": 8814
+    },
+    {
+      "epoch": 1.9663171983047065,
+      "grad_norm": 0.18594609200954437,
+      "learning_rate": 5.382368901793759e-06,
+      "loss": 0.4855,
+      "step": 8815
+    },
+    {
+      "epoch": 1.966540263216596,
+      "grad_norm": 0.1851484477519989,
+      "learning_rate": 5.380281530244974e-06,
+      "loss": 0.417,
+      "step": 8816
+    },
+    {
+      "epoch": 1.9667633281284855,
+      "grad_norm": 0.18481191992759705,
+      "learning_rate": 5.378194414565258e-06,
+      "loss": 0.4504,
+      "step": 8817
+    },
+    {
+      "epoch": 1.9669863930403748,
+      "grad_norm": 0.17994412779808044,
+      "learning_rate": 5.376107554870207e-06,
+      "loss": 0.4565,
+      "step": 8818
+    },
+    {
+      "epoch": 1.967209457952264,
+      "grad_norm": 0.17830024659633636,
+      "learning_rate": 5.374020951275409e-06,
+      "loss": 0.4562,
+      "step": 8819
+    },
+    {
+      "epoch": 1.9674325228641534,
+      "grad_norm": 0.19198620319366455,
+      "learning_rate": 5.371934603896426e-06,
+      "loss": 0.4414,
+      "step": 8820
+    },
+    {
+      "epoch": 1.9676555877760429,
+      "grad_norm": 0.17319843173027039,
+      "learning_rate": 5.369848512848816e-06,
+      "loss": 0.4351,
+      "step": 8821
+    },
+    {
+      "epoch": 1.9678786526879322,
+      "grad_norm": 0.1835477352142334,
+      "learning_rate": 5.3677626782481205e-06,
+      "loss": 0.4447,
+      "step": 8822
+    },
+    {
+      "epoch": 1.9681017175998217,
+      "grad_norm": 0.17269068956375122,
+      "learning_rate": 5.365677100209868e-06,
+      "loss": 0.4313,
+      "step": 8823
+    },
+    {
+      "epoch": 1.968324782511711,
+      "grad_norm": 0.17792171239852905,
+      "learning_rate": 5.363591778849566e-06,
+      "loss": 0.4276,
+      "step": 8824
+    },
+    {
+      "epoch": 1.9685478474236002,
+      "grad_norm": 0.17253033816814423,
+      "learning_rate": 5.361506714282716e-06,
+      "loss": 0.4399,
+      "step": 8825
+    },
+    {
+      "epoch": 1.9687709123354895,
+      "grad_norm": 0.17550261318683624,
+      "learning_rate": 5.359421906624802e-06,
+      "loss": 0.4381,
+      "step": 8826
+    },
+    {
+      "epoch": 1.968993977247379,
+      "grad_norm": 0.18501095473766327,
+      "learning_rate": 5.357337355991296e-06,
+      "loss": 0.4414,
+      "step": 8827
+    },
+    {
+      "epoch": 1.9692170421592683,
+      "grad_norm": 0.17807962000370026,
+      "learning_rate": 5.3552530624976475e-06,
+      "loss": 0.4627,
+      "step": 8828
+    },
+    {
+      "epoch": 1.9694401070711578,
+      "grad_norm": 0.17999139428138733,
+      "learning_rate": 5.3531690262593e-06,
+      "loss": 0.446,
+      "step": 8829
+    },
+    {
+      "epoch": 1.969663171983047,
+      "grad_norm": 0.17978177964687347,
+      "learning_rate": 5.351085247391681e-06,
+      "loss": 0.4616,
+      "step": 8830
+    },
+    {
+      "epoch": 1.9698862368949364,
+      "grad_norm": 0.18131205439567566,
+      "learning_rate": 5.34900172601021e-06,
+      "loss": 0.4418,
+      "step": 8831
+    },
+    {
+      "epoch": 1.9701093018068256,
+      "grad_norm": 0.17810958623886108,
+      "learning_rate": 5.3469184622302725e-06,
+      "loss": 0.4492,
+      "step": 8832
+    },
+    {
+      "epoch": 1.9703323667187151,
+      "grad_norm": 0.1817857027053833,
+      "learning_rate": 5.344835456167262e-06,
+      "loss": 0.4706,
+      "step": 8833
+    },
+    {
+      "epoch": 1.9705554316306046,
+      "grad_norm": 0.18190069496631622,
+      "learning_rate": 5.342752707936546e-06,
+      "loss": 0.4614,
+      "step": 8834
+    },
+    {
+      "epoch": 1.970778496542494,
+      "grad_norm": 0.18769033253192902,
+      "learning_rate": 5.340670217653483e-06,
+      "loss": 0.4537,
+      "step": 8835
+    },
+    {
+      "epoch": 1.9710015614543832,
+      "grad_norm": 0.18518251180648804,
+      "learning_rate": 5.338587985433409e-06,
+      "loss": 0.4551,
+      "step": 8836
+    },
+    {
+      "epoch": 1.9712246263662725,
+      "grad_norm": 0.1764073669910431,
+      "learning_rate": 5.336506011391653e-06,
+      "loss": 0.4309,
+      "step": 8837
+    },
+    {
+      "epoch": 1.971447691278162,
+      "grad_norm": 0.3080836534500122,
+      "learning_rate": 5.334424295643529e-06,
+      "loss": 0.4195,
+      "step": 8838
+    },
+    {
+      "epoch": 1.9716707561900513,
+      "grad_norm": 0.18763823807239532,
+      "learning_rate": 5.332342838304335e-06,
+      "loss": 0.4559,
+      "step": 8839
+    },
+    {
+      "epoch": 1.9718938211019408,
+      "grad_norm": 0.19015417993068695,
+      "learning_rate": 5.330261639489359e-06,
+      "loss": 0.4487,
+      "step": 8840
+    },
+    {
+      "epoch": 1.97211688601383,
+      "grad_norm": 0.18165238201618195,
+      "learning_rate": 5.328180699313864e-06,
+      "loss": 0.4439,
+      "step": 8841
+    },
+    {
+      "epoch": 1.9723399509257193,
+      "grad_norm": 0.28610339760780334,
+      "learning_rate": 5.326100017893107e-06,
+      "loss": 0.4331,
+      "step": 8842
+    },
+    {
+      "epoch": 1.9725630158376086,
+      "grad_norm": 0.1734454482793808,
+      "learning_rate": 5.324019595342334e-06,
+      "loss": 0.4424,
+      "step": 8843
+    },
+    {
+      "epoch": 1.9727860807494981,
+      "grad_norm": 0.18375049531459808,
+      "learning_rate": 5.32193943177677e-06,
+      "loss": 0.4364,
+      "step": 8844
+    },
+    {
+      "epoch": 1.9730091456613874,
+      "grad_norm": 0.1872616410255432,
+      "learning_rate": 5.319859527311624e-06,
+      "loss": 0.4591,
+      "step": 8845
+    },
+    {
+      "epoch": 1.973232210573277,
+      "grad_norm": 0.18444837629795074,
+      "learning_rate": 5.3177798820620975e-06,
+      "loss": 0.4462,
+      "step": 8846
+    },
+    {
+      "epoch": 1.9734552754851662,
+      "grad_norm": 0.1809314340353012,
+      "learning_rate": 5.315700496143372e-06,
+      "loss": 0.4382,
+      "step": 8847
+    },
+    {
+      "epoch": 1.9736783403970555,
+      "grad_norm": 0.18612340092658997,
+      "learning_rate": 5.313621369670623e-06,
+      "loss": 0.4391,
+      "step": 8848
+    },
+    {
+      "epoch": 1.9739014053089448,
+      "grad_norm": 0.18118642270565033,
+      "learning_rate": 5.311542502758996e-06,
+      "loss": 0.4659,
+      "step": 8849
+    },
+    {
+      "epoch": 1.9741244702208343,
+      "grad_norm": 0.19646091759204865,
+      "learning_rate": 5.309463895523638e-06,
+      "loss": 0.4594,
+      "step": 8850
+    },
+    {
+      "epoch": 1.9743475351327238,
+      "grad_norm": 0.18082433938980103,
+      "learning_rate": 5.3073855480796735e-06,
+      "loss": 0.472,
+      "step": 8851
+    },
+    {
+      "epoch": 1.974570600044613,
+      "grad_norm": 0.17272797226905823,
+      "learning_rate": 5.305307460542219e-06,
+      "loss": 0.4375,
+      "step": 8852
+    },
+    {
+      "epoch": 1.9747936649565023,
+      "grad_norm": 0.19139288365840912,
+      "learning_rate": 5.303229633026368e-06,
+      "loss": 0.459,
+      "step": 8853
+    },
+    {
+      "epoch": 1.9750167298683916,
+      "grad_norm": 0.18878120183944702,
+      "learning_rate": 5.3011520656472e-06,
+      "loss": 0.4701,
+      "step": 8854
+    },
+    {
+      "epoch": 1.975239794780281,
+      "grad_norm": 0.1769939512014389,
+      "learning_rate": 5.299074758519788e-06,
+      "loss": 0.4322,
+      "step": 8855
+    },
+    {
+      "epoch": 1.9754628596921704,
+      "grad_norm": 0.1793774664402008,
+      "learning_rate": 5.296997711759186e-06,
+      "loss": 0.4352,
+      "step": 8856
+    },
+    {
+      "epoch": 1.97568592460406,
+      "grad_norm": 0.17794568836688995,
+      "learning_rate": 5.2949209254804366e-06,
+      "loss": 0.4493,
+      "step": 8857
+    },
+    {
+      "epoch": 1.9759089895159492,
+      "grad_norm": 0.185879185795784,
+      "learning_rate": 5.292844399798559e-06,
+      "loss": 0.4612,
+      "step": 8858
+    },
+    {
+      "epoch": 1.9761320544278385,
+      "grad_norm": 0.18604058027267456,
+      "learning_rate": 5.290768134828568e-06,
+      "loss": 0.4597,
+      "step": 8859
+    },
+    {
+      "epoch": 1.9763551193397277,
+      "grad_norm": 0.19229385256767273,
+      "learning_rate": 5.28869213068546e-06,
+      "loss": 0.4597,
+      "step": 8860
+    },
+    {
+      "epoch": 1.9765781842516172,
+      "grad_norm": 0.1805671751499176,
+      "learning_rate": 5.286616387484217e-06,
+      "loss": 0.4478,
+      "step": 8861
+    },
+    {
+      "epoch": 1.9768012491635067,
+      "grad_norm": 0.18257898092269897,
+      "learning_rate": 5.28454090533981e-06,
+      "loss": 0.4505,
+      "step": 8862
+    },
+    {
+      "epoch": 1.977024314075396,
+      "grad_norm": 0.20359763503074646,
+      "learning_rate": 5.2824656843671865e-06,
+      "loss": 0.4693,
+      "step": 8863
+    },
+    {
+      "epoch": 1.9772473789872853,
+      "grad_norm": 0.1799335926771164,
+      "learning_rate": 5.280390724681288e-06,
+      "loss": 0.414,
+      "step": 8864
+    },
+    {
+      "epoch": 1.9774704438991746,
+      "grad_norm": 0.17781312763690948,
+      "learning_rate": 5.278316026397037e-06,
+      "loss": 0.4495,
+      "step": 8865
+    },
+    {
+      "epoch": 1.9776935088110639,
+      "grad_norm": 0.18418540060520172,
+      "learning_rate": 5.276241589629349e-06,
+      "loss": 0.4328,
+      "step": 8866
+    },
+    {
+      "epoch": 1.9779165737229534,
+      "grad_norm": 0.1901324838399887,
+      "learning_rate": 5.274167414493113e-06,
+      "loss": 0.4648,
+      "step": 8867
+    },
+    {
+      "epoch": 1.9781396386348429,
+      "grad_norm": 0.17587338387966156,
+      "learning_rate": 5.272093501103213e-06,
+      "loss": 0.4451,
+      "step": 8868
+    },
+    {
+      "epoch": 1.9783627035467322,
+      "grad_norm": 0.17703662812709808,
+      "learning_rate": 5.270019849574513e-06,
+      "loss": 0.4693,
+      "step": 8869
+    },
+    {
+      "epoch": 1.9785857684586214,
+      "grad_norm": 0.2079380750656128,
+      "learning_rate": 5.26794646002187e-06,
+      "loss": 0.47,
+      "step": 8870
+    },
+    {
+      "epoch": 1.9788088333705107,
+      "grad_norm": 0.21435651183128357,
+      "learning_rate": 5.265873332560114e-06,
+      "loss": 0.4404,
+      "step": 8871
+    },
+    {
+      "epoch": 1.9790318982824002,
+      "grad_norm": 0.1856585443019867,
+      "learning_rate": 5.263800467304072e-06,
+      "loss": 0.4442,
+      "step": 8872
+    },
+    {
+      "epoch": 1.9792549631942895,
+      "grad_norm": 0.178078293800354,
+      "learning_rate": 5.261727864368554e-06,
+      "loss": 0.4459,
+      "step": 8873
+    },
+    {
+      "epoch": 1.979478028106179,
+      "grad_norm": 0.17547845840454102,
+      "learning_rate": 5.259655523868352e-06,
+      "loss": 0.4636,
+      "step": 8874
+    },
+    {
+      "epoch": 1.9797010930180683,
+      "grad_norm": 0.17804180085659027,
+      "learning_rate": 5.257583445918244e-06,
+      "loss": 0.4311,
+      "step": 8875
+    },
+    {
+      "epoch": 1.9799241579299576,
+      "grad_norm": 0.180350661277771,
+      "learning_rate": 5.2555116306329945e-06,
+      "loss": 0.4316,
+      "step": 8876
+    },
+    {
+      "epoch": 1.9801472228418469,
+      "grad_norm": 0.17848455905914307,
+      "learning_rate": 5.253440078127355e-06,
+      "loss": 0.4671,
+      "step": 8877
+    },
+    {
+      "epoch": 1.9803702877537364,
+      "grad_norm": 0.1925760954618454,
+      "learning_rate": 5.251368788516066e-06,
+      "loss": 0.462,
+      "step": 8878
+    },
+    {
+      "epoch": 1.9805933526656259,
+      "grad_norm": 0.19506682455539703,
+      "learning_rate": 5.249297761913839e-06,
+      "loss": 0.4718,
+      "step": 8879
+    },
+    {
+      "epoch": 1.9808164175775151,
+      "grad_norm": 0.18116964399814606,
+      "learning_rate": 5.2472269984353845e-06,
+      "loss": 0.4597,
+      "step": 8880
+    },
+    {
+      "epoch": 1.9810394824894044,
+      "grad_norm": 0.18594235181808472,
+      "learning_rate": 5.245156498195395e-06,
+      "loss": 0.453,
+      "step": 8881
+    },
+    {
+      "epoch": 1.9812625474012937,
+      "grad_norm": 0.1806681901216507,
+      "learning_rate": 5.243086261308548e-06,
+      "loss": 0.4442,
+      "step": 8882
+    },
+    {
+      "epoch": 1.981485612313183,
+      "grad_norm": 0.18260101974010468,
+      "learning_rate": 5.241016287889511e-06,
+      "loss": 0.4457,
+      "step": 8883
+    },
+    {
+      "epoch": 1.9817086772250725,
+      "grad_norm": 0.17734308540821075,
+      "learning_rate": 5.238946578052921e-06,
+      "loss": 0.449,
+      "step": 8884
+    },
+    {
+      "epoch": 1.981931742136962,
+      "grad_norm": 0.1897689253091812,
+      "learning_rate": 5.236877131913421e-06,
+      "loss": 0.4611,
+      "step": 8885
+    },
+    {
+      "epoch": 1.9821548070488513,
+      "grad_norm": 0.1725919246673584,
+      "learning_rate": 5.234807949585623e-06,
+      "loss": 0.4231,
+      "step": 8886
+    },
+    {
+      "epoch": 1.9823778719607406,
+      "grad_norm": 0.18596228957176208,
+      "learning_rate": 5.2327390311841404e-06,
+      "loss": 0.45,
+      "step": 8887
+    },
+    {
+      "epoch": 1.9826009368726298,
+      "grad_norm": 0.18885090947151184,
+      "learning_rate": 5.230670376823555e-06,
+      "loss": 0.4753,
+      "step": 8888
+    },
+    {
+      "epoch": 1.9828240017845193,
+      "grad_norm": 0.18967807292938232,
+      "learning_rate": 5.228601986618442e-06,
+      "loss": 0.449,
+      "step": 8889
+    },
+    {
+      "epoch": 1.9830470666964086,
+      "grad_norm": 0.18121550977230072,
+      "learning_rate": 5.226533860683366e-06,
+      "loss": 0.4322,
+      "step": 8890
+    },
+    {
+      "epoch": 1.9832701316082981,
+      "grad_norm": 0.181602343916893,
+      "learning_rate": 5.224465999132875e-06,
+      "loss": 0.4414,
+      "step": 8891
+    },
+    {
+      "epoch": 1.9834931965201874,
+      "grad_norm": 0.17742736637592316,
+      "learning_rate": 5.222398402081492e-06,
+      "loss": 0.4493,
+      "step": 8892
+    },
+    {
+      "epoch": 1.9837162614320767,
+      "grad_norm": 0.17834076285362244,
+      "learning_rate": 5.220331069643737e-06,
+      "loss": 0.4529,
+      "step": 8893
+    },
+    {
+      "epoch": 1.983939326343966,
+      "grad_norm": 0.17975963652133942,
+      "learning_rate": 5.218264001934114e-06,
+      "loss": 0.4164,
+      "step": 8894
+    },
+    {
+      "epoch": 1.9841623912558555,
+      "grad_norm": 0.18679198622703552,
+      "learning_rate": 5.216197199067112e-06,
+      "loss": 0.431,
+      "step": 8895
+    },
+    {
+      "epoch": 1.984385456167745,
+      "grad_norm": 0.1757335662841797,
+      "learning_rate": 5.2141306611571965e-06,
+      "loss": 0.4427,
+      "step": 8896
+    },
+    {
+      "epoch": 1.9846085210796343,
+      "grad_norm": 0.19077639281749725,
+      "learning_rate": 5.21206438831883e-06,
+      "loss": 0.4358,
+      "step": 8897
+    },
+    {
+      "epoch": 1.9848315859915235,
+      "grad_norm": 0.18716174364089966,
+      "learning_rate": 5.2099983806664546e-06,
+      "loss": 0.4621,
+      "step": 8898
+    },
+    {
+      "epoch": 1.9850546509034128,
+      "grad_norm": 0.1881352961063385,
+      "learning_rate": 5.207932638314502e-06,
+      "loss": 0.461,
+      "step": 8899
+    },
+    {
+      "epoch": 1.985277715815302,
+      "grad_norm": 0.18843919038772583,
+      "learning_rate": 5.2058671613773805e-06,
+      "loss": 0.4489,
+      "step": 8900
+    },
+    {
+      "epoch": 1.9855007807271916,
+      "grad_norm": 0.18741180002689362,
+      "learning_rate": 5.203801949969491e-06,
+      "loss": 0.4648,
+      "step": 8901
+    },
+    {
+      "epoch": 1.985723845639081,
+      "grad_norm": 0.17685861885547638,
+      "learning_rate": 5.2017370042052205e-06,
+      "loss": 0.4629,
+      "step": 8902
+    },
+    {
+      "epoch": 1.9859469105509704,
+      "grad_norm": 0.16930168867111206,
+      "learning_rate": 5.199672324198935e-06,
+      "loss": 0.4395,
+      "step": 8903
+    },
+    {
+      "epoch": 1.9861699754628597,
+      "grad_norm": 0.17272312939167023,
+      "learning_rate": 5.197607910064997e-06,
+      "loss": 0.4509,
+      "step": 8904
+    },
+    {
+      "epoch": 1.986393040374749,
+      "grad_norm": 0.18139375746250153,
+      "learning_rate": 5.195543761917736e-06,
+      "loss": 0.4504,
+      "step": 8905
+    },
+    {
+      "epoch": 1.9866161052866385,
+      "grad_norm": 0.18100585043430328,
+      "learning_rate": 5.193479879871483e-06,
+      "loss": 0.4445,
+      "step": 8906
+    },
+    {
+      "epoch": 1.9868391701985277,
+      "grad_norm": 0.2222553938627243,
+      "learning_rate": 5.19141626404055e-06,
+      "loss": 0.4387,
+      "step": 8907
+    },
+    {
+      "epoch": 1.9870622351104172,
+      "grad_norm": 0.17244549095630646,
+      "learning_rate": 5.189352914539233e-06,
+      "loss": 0.4727,
+      "step": 8908
+    },
+    {
+      "epoch": 1.9872853000223065,
+      "grad_norm": 0.1766839474439621,
+      "learning_rate": 5.18728983148181e-06,
+      "loss": 0.4521,
+      "step": 8909
+    },
+    {
+      "epoch": 1.9875083649341958,
+      "grad_norm": 0.1754491627216339,
+      "learning_rate": 5.185227014982548e-06,
+      "loss": 0.4548,
+      "step": 8910
+    },
+    {
+      "epoch": 1.987731429846085,
+      "grad_norm": 0.17780523002147675,
+      "learning_rate": 5.183164465155699e-06,
+      "loss": 0.4563,
+      "step": 8911
+    },
+    {
+      "epoch": 1.9879544947579746,
+      "grad_norm": 0.1709602177143097,
+      "learning_rate": 5.181102182115507e-06,
+      "loss": 0.4194,
+      "step": 8912
+    },
+    {
+      "epoch": 1.988177559669864,
+      "grad_norm": 0.1960027664899826,
+      "learning_rate": 5.179040165976183e-06,
+      "loss": 0.4324,
+      "step": 8913
+    },
+    {
+      "epoch": 1.9884006245817534,
+      "grad_norm": 0.18167687952518463,
+      "learning_rate": 5.176978416851941e-06,
+      "loss": 0.4349,
+      "step": 8914
+    },
+    {
+      "epoch": 1.9886236894936427,
+      "grad_norm": 0.1885111927986145,
+      "learning_rate": 5.17491693485697e-06,
+      "loss": 0.4979,
+      "step": 8915
+    },
+    {
+      "epoch": 1.988846754405532,
+      "grad_norm": 0.18724943697452545,
+      "learning_rate": 5.172855720105456e-06,
+      "loss": 0.4774,
+      "step": 8916
+    },
+    {
+      "epoch": 1.9890698193174212,
+      "grad_norm": 0.1897243857383728,
+      "learning_rate": 5.1707947727115515e-06,
+      "loss": 0.4415,
+      "step": 8917
+    },
+    {
+      "epoch": 1.9892928842293107,
+      "grad_norm": 0.18215550482273102,
+      "learning_rate": 5.16873409278941e-06,
+      "loss": 0.4539,
+      "step": 8918
+    },
+    {
+      "epoch": 1.9895159491412002,
+      "grad_norm": 0.17774201929569244,
+      "learning_rate": 5.1666736804531646e-06,
+      "loss": 0.4199,
+      "step": 8919
+    },
+    {
+      "epoch": 1.9897390140530895,
+      "grad_norm": 0.17743372917175293,
+      "learning_rate": 5.164613535816937e-06,
+      "loss": 0.4457,
+      "step": 8920
+    },
+    {
+      "epoch": 1.9899620789649788,
+      "grad_norm": 0.17990683019161224,
+      "learning_rate": 5.162553658994823e-06,
+      "loss": 0.4648,
+      "step": 8921
+    },
+    {
+      "epoch": 1.990185143876868,
+      "grad_norm": 0.18305742740631104,
+      "learning_rate": 5.160494050100917e-06,
+      "loss": 0.4792,
+      "step": 8922
+    },
+    {
+      "epoch": 1.9904082087887576,
+      "grad_norm": 0.18619424104690552,
+      "learning_rate": 5.158434709249291e-06,
+      "loss": 0.425,
+      "step": 8923
+    },
+    {
+      "epoch": 1.9906312737006469,
+      "grad_norm": 0.17404185235500336,
+      "learning_rate": 5.156375636554007e-06,
+      "loss": 0.4299,
+      "step": 8924
+    },
+    {
+      "epoch": 1.9908543386125364,
+      "grad_norm": 0.17407676577568054,
+      "learning_rate": 5.1543168321291115e-06,
+      "loss": 0.4386,
+      "step": 8925
+    },
+    {
+      "epoch": 1.9910774035244256,
+      "grad_norm": 0.17693208158016205,
+      "learning_rate": 5.152258296088626e-06,
+      "loss": 0.4306,
+      "step": 8926
+    },
+    {
+      "epoch": 1.991300468436315,
+      "grad_norm": 0.1785212606191635,
+      "learning_rate": 5.15020002854657e-06,
+      "loss": 0.4715,
+      "step": 8927
+    },
+    {
+      "epoch": 1.9915235333482042,
+      "grad_norm": 0.18704737722873688,
+      "learning_rate": 5.148142029616943e-06,
+      "loss": 0.4653,
+      "step": 8928
+    },
+    {
+      "epoch": 1.9917465982600937,
+      "grad_norm": 0.19586940109729767,
+      "learning_rate": 5.146084299413732e-06,
+      "loss": 0.4784,
+      "step": 8929
+    },
+    {
+      "epoch": 1.9919696631719832,
+      "grad_norm": 0.18988698720932007,
+      "learning_rate": 5.144026838050902e-06,
+      "loss": 0.4543,
+      "step": 8930
+    },
+    {
+      "epoch": 1.9921927280838725,
+      "grad_norm": 0.17881841957569122,
+      "learning_rate": 5.141969645642412e-06,
+      "loss": 0.4286,
+      "step": 8931
+    },
+    {
+      "epoch": 1.9924157929957618,
+      "grad_norm": 0.18214870989322662,
+      "learning_rate": 5.139912722302201e-06,
+      "loss": 0.435,
+      "step": 8932
+    },
+    {
+      "epoch": 1.992638857907651,
+      "grad_norm": 0.26209428906440735,
+      "learning_rate": 5.137856068144197e-06,
+      "loss": 0.4628,
+      "step": 8933
+    },
+    {
+      "epoch": 1.9928619228195403,
+      "grad_norm": 0.17895345389842987,
+      "learning_rate": 5.135799683282309e-06,
+      "loss": 0.4708,
+      "step": 8934
+    },
+    {
+      "epoch": 1.9930849877314298,
+      "grad_norm": 0.18702343106269836,
+      "learning_rate": 5.133743567830427e-06,
+      "loss": 0.4324,
+      "step": 8935
+    },
+    {
+      "epoch": 1.9933080526433193,
+      "grad_norm": 0.18158088624477386,
+      "learning_rate": 5.1316877219024375e-06,
+      "loss": 0.4554,
+      "step": 8936
+    },
+    {
+      "epoch": 1.9935311175552086,
+      "grad_norm": 0.1866513192653656,
+      "learning_rate": 5.129632145612204e-06,
+      "loss": 0.4397,
+      "step": 8937
+    },
+    {
+      "epoch": 1.993754182467098,
+      "grad_norm": 0.19149568676948547,
+      "learning_rate": 5.127576839073583e-06,
+      "loss": 0.432,
+      "step": 8938
+    },
+    {
+      "epoch": 1.9939772473789872,
+      "grad_norm": 0.18493977189064026,
+      "learning_rate": 5.1255218024004e-06,
+      "loss": 0.4445,
+      "step": 8939
+    },
+    {
+      "epoch": 1.9942003122908767,
+      "grad_norm": 0.18081733584403992,
+      "learning_rate": 5.123467035706482e-06,
+      "loss": 0.4412,
+      "step": 8940
+    },
+    {
+      "epoch": 1.994423377202766,
+      "grad_norm": 0.17544502019882202,
+      "learning_rate": 5.121412539105635e-06,
+      "loss": 0.4076,
+      "step": 8941
+    },
+    {
+      "epoch": 1.9946464421146555,
+      "grad_norm": 0.20097681879997253,
+      "learning_rate": 5.119358312711651e-06,
+      "loss": 0.4676,
+      "step": 8942
+    },
+    {
+      "epoch": 1.9948695070265448,
+      "grad_norm": 0.17874199151992798,
+      "learning_rate": 5.117304356638301e-06,
+      "loss": 0.4525,
+      "step": 8943
+    },
+    {
+      "epoch": 1.995092571938434,
+      "grad_norm": 0.18695956468582153,
+      "learning_rate": 5.11525067099935e-06,
+      "loss": 0.4498,
+      "step": 8944
+    },
+    {
+      "epoch": 1.9953156368503233,
+      "grad_norm": 0.18029382824897766,
+      "learning_rate": 5.113197255908543e-06,
+      "loss": 0.454,
+      "step": 8945
+    },
+    {
+      "epoch": 1.9955387017622128,
+      "grad_norm": 0.17514650523662567,
+      "learning_rate": 5.111144111479611e-06,
+      "loss": 0.4191,
+      "step": 8946
+    },
+    {
+      "epoch": 1.9957617666741023,
+      "grad_norm": 0.17547555267810822,
+      "learning_rate": 5.109091237826273e-06,
+      "loss": 0.4276,
+      "step": 8947
+    },
+    {
+      "epoch": 1.9959848315859916,
+      "grad_norm": 0.18239977955818176,
+      "learning_rate": 5.107038635062225e-06,
+      "loss": 0.4447,
+      "step": 8948
+    },
+    {
+      "epoch": 1.9962078964978809,
+      "grad_norm": 0.1756354421377182,
+      "learning_rate": 5.1049863033011535e-06,
+      "loss": 0.4251,
+      "step": 8949
+    },
+    {
+      "epoch": 1.9964309614097702,
+      "grad_norm": 0.18526077270507812,
+      "learning_rate": 5.1029342426567345e-06,
+      "loss": 0.4603,
+      "step": 8950
+    },
+    {
+      "epoch": 1.9966540263216594,
+      "grad_norm": 0.20123088359832764,
+      "learning_rate": 5.100882453242622e-06,
+      "loss": 0.4473,
+      "step": 8951
+    },
+    {
+      "epoch": 1.996877091233549,
+      "grad_norm": 0.17634332180023193,
+      "learning_rate": 5.098830935172453e-06,
+      "loss": 0.4318,
+      "step": 8952
+    },
+    {
+      "epoch": 1.9971001561454385,
+      "grad_norm": 0.19645899534225464,
+      "learning_rate": 5.096779688559857e-06,
+      "loss": 0.4572,
+      "step": 8953
+    },
+    {
+      "epoch": 1.9973232210573277,
+      "grad_norm": 0.17646844685077667,
+      "learning_rate": 5.094728713518442e-06,
+      "loss": 0.4428,
+      "step": 8954
+    },
+    {
+      "epoch": 1.997546285969217,
+      "grad_norm": 0.17658363282680511,
+      "learning_rate": 5.092678010161812e-06,
+      "loss": 0.4518,
+      "step": 8955
+    },
+    {
+      "epoch": 1.9977693508811063,
+      "grad_norm": 0.17553554475307465,
+      "learning_rate": 5.090627578603537e-06,
+      "loss": 0.4602,
+      "step": 8956
+    },
+    {
+      "epoch": 1.9979924157929958,
+      "grad_norm": 0.18948344886302948,
+      "learning_rate": 5.08857741895719e-06,
+      "loss": 0.4539,
+      "step": 8957
+    },
+    {
+      "epoch": 1.998215480704885,
+      "grad_norm": 0.17445118725299835,
+      "learning_rate": 5.086527531336318e-06,
+      "loss": 0.4532,
+      "step": 8958
+    },
+    {
+      "epoch": 1.9984385456167746,
+      "grad_norm": 0.1809672862291336,
+      "learning_rate": 5.084477915854462e-06,
+      "loss": 0.4203,
+      "step": 8959
+    },
+    {
+      "epoch": 1.9986616105286639,
+      "grad_norm": 0.17913390696048737,
+      "learning_rate": 5.082428572625136e-06,
+      "loss": 0.4322,
+      "step": 8960
+    },
+    {
+      "epoch": 1.9988846754405531,
+      "grad_norm": 0.18459449708461761,
+      "learning_rate": 5.080379501761848e-06,
+      "loss": 0.4865,
+      "step": 8961
+    },
+    {
+      "epoch": 1.9991077403524424,
+      "grad_norm": 0.18868421018123627,
+      "learning_rate": 5.0783307033780895e-06,
+      "loss": 0.4571,
+      "step": 8962
+    },
+    {
+      "epoch": 1.999330805264332,
+      "grad_norm": 0.17596866190433502,
+      "learning_rate": 5.076282177587339e-06,
+      "loss": 0.4382,
+      "step": 8963
+    },
+    {
+      "epoch": 1.9995538701762214,
+      "grad_norm": 0.174740731716156,
+      "learning_rate": 5.074233924503047e-06,
+      "loss": 0.4407,
+      "step": 8964
+    },
+    {
+      "epoch": 1.9997769350881107,
+      "grad_norm": 0.17470811307430267,
+      "learning_rate": 5.072185944238665e-06,
+      "loss": 0.4456,
+      "step": 8965
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.18581973016262054,
+      "learning_rate": 5.070138236907625e-06,
+      "loss": 0.4513,
+      "step": 8966
+    },
+    {
+      "epoch": 2.0,
+      "eval_loss": 0.31910037994384766,
+      "eval_runtime": 667.801,
+      "eval_samples_per_second": 94.504,
+      "eval_steps_per_second": 1.478,
+      "step": 8966
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 13449,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 6.876226391488531e+20,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}