diff --git "a/checkpoint-1784/trainer_state.json" "b/checkpoint-1784/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-1784/trainer_state.json"
@@ -0,0 +1,12521 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9981329350261388,
+  "eval_steps": 500,
+  "global_step": 1784,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011202389843166542,
+      "grad_norm": 2.8858156204223633,
+      "learning_rate": 5.0000000000000004e-08,
+      "loss": 0.9141,
+      "step": 1
+    },
+    {
+      "epoch": 0.0022404779686333084,
+      "grad_norm": 2.804832935333252,
+      "learning_rate": 1.0000000000000001e-07,
+      "loss": 0.9218,
+      "step": 2
+    },
+    {
+      "epoch": 0.0033607169529499626,
+      "grad_norm": 2.7312190532684326,
+      "learning_rate": 1.5000000000000002e-07,
+      "loss": 0.9026,
+      "step": 3
+    },
+    {
+      "epoch": 0.004480955937266617,
+      "grad_norm": 2.466888904571533,
+      "learning_rate": 2.0000000000000002e-07,
+      "loss": 0.9196,
+      "step": 4
+    },
+    {
+      "epoch": 0.005601194921583271,
+      "grad_norm": 2.629023551940918,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 0.8756,
+      "step": 5
+    },
+    {
+      "epoch": 0.006721433905899925,
+      "grad_norm": 2.747882604598999,
+      "learning_rate": 3.0000000000000004e-07,
+      "loss": 0.9368,
+      "step": 6
+    },
+    {
+      "epoch": 0.00784167289021658,
+      "grad_norm": 2.7748570442199707,
+      "learning_rate": 3.5000000000000004e-07,
+      "loss": 0.9265,
+      "step": 7
+    },
+    {
+      "epoch": 0.008961911874533234,
+      "grad_norm": 2.38838267326355,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 0.9154,
+      "step": 8
+    },
+    {
+      "epoch": 0.010082150858849889,
+      "grad_norm": 2.9623191356658936,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 0.9322,
+      "step": 9
+    },
+    {
+      "epoch": 0.011202389843166542,
+      "grad_norm": 2.4339418411254883,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 0.9117,
+      "step": 10
+    },
+    {
+      "epoch": 0.012322628827483197,
+      "grad_norm": 2.7546939849853516,
+      "learning_rate": 5.5e-07,
+      "loss": 0.9146,
+      "step": 11
+    },
+    {
+      "epoch": 0.01344286781179985,
+      "grad_norm": 2.2417454719543457,
+      "learning_rate": 6.000000000000001e-07,
+      "loss": 0.8947,
+      "step": 12
+    },
+    {
+      "epoch": 0.014563106796116505,
+      "grad_norm": 2.443024158477783,
+      "learning_rate": 6.5e-07,
+      "loss": 0.8952,
+      "step": 13
+    },
+    {
+      "epoch": 0.01568334578043316,
+      "grad_norm": 2.1837103366851807,
+      "learning_rate": 7.000000000000001e-07,
+      "loss": 0.9105,
+      "step": 14
+    },
+    {
+      "epoch": 0.016803584764749812,
+      "grad_norm": 2.1527903079986572,
+      "learning_rate": 7.5e-07,
+      "loss": 0.8899,
+      "step": 15
+    },
+    {
+      "epoch": 0.017923823749066467,
+      "grad_norm": 2.161756753921509,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 0.8655,
+      "step": 16
+    },
+    {
+      "epoch": 0.019044062733383122,
+      "grad_norm": 1.898628830909729,
+      "learning_rate": 8.500000000000001e-07,
+      "loss": 0.85,
+      "step": 17
+    },
+    {
+      "epoch": 0.020164301717699777,
+      "grad_norm": 1.7848337888717651,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 0.8654,
+      "step": 18
+    },
+    {
+      "epoch": 0.02128454070201643,
+      "grad_norm": 1.9204438924789429,
+      "learning_rate": 9.500000000000001e-07,
+      "loss": 0.8685,
+      "step": 19
+    },
+    {
+      "epoch": 0.022404779686333084,
+      "grad_norm": 1.6950091123580933,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 0.8981,
+      "step": 20
+    },
+    {
+      "epoch": 0.02352501867064974,
+      "grad_norm": 1.774417519569397,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 0.8953,
+      "step": 21
+    },
+    {
+      "epoch": 0.024645257654966394,
+      "grad_norm": 1.7463773488998413,
+      "learning_rate": 1.1e-06,
+      "loss": 0.8668,
+      "step": 22
+    },
+    {
+      "epoch": 0.025765496639283045,
+      "grad_norm": 1.5141345262527466,
+      "learning_rate": 1.1500000000000002e-06,
+      "loss": 0.8523,
+      "step": 23
+    },
+    {
+      "epoch": 0.0268857356235997,
+      "grad_norm": 1.5256919860839844,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 0.837,
+      "step": 24
+    },
+    {
+      "epoch": 0.028005974607916356,
+      "grad_norm": 1.3926639556884766,
+      "learning_rate": 1.25e-06,
+      "loss": 0.8498,
+      "step": 25
+    },
+    {
+      "epoch": 0.02912621359223301,
+      "grad_norm": 1.2967628240585327,
+      "learning_rate": 1.3e-06,
+      "loss": 0.7906,
+      "step": 26
+    },
+    {
+      "epoch": 0.030246452576549662,
+      "grad_norm": 1.2432132959365845,
+      "learning_rate": 1.3500000000000002e-06,
+      "loss": 0.8143,
+      "step": 27
+    },
+    {
+      "epoch": 0.03136669156086632,
+      "grad_norm": 1.1862553358078003,
+      "learning_rate": 1.4000000000000001e-06,
+      "loss": 0.7959,
+      "step": 28
+    },
+    {
+      "epoch": 0.03248693054518297,
+      "grad_norm": 1.18585205078125,
+      "learning_rate": 1.45e-06,
+      "loss": 0.8263,
+      "step": 29
+    },
+    {
+      "epoch": 0.033607169529499624,
+      "grad_norm": 1.118574857711792,
+      "learning_rate": 1.5e-06,
+      "loss": 0.7765,
+      "step": 30
+    },
+    {
+      "epoch": 0.03472740851381628,
+      "grad_norm": 1.1179440021514893,
+      "learning_rate": 1.5500000000000002e-06,
+      "loss": 0.8037,
+      "step": 31
+    },
+    {
+      "epoch": 0.035847647498132934,
+      "grad_norm": 1.100803017616272,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 0.8113,
+      "step": 32
+    },
+    {
+      "epoch": 0.03696788648244959,
+      "grad_norm": 1.097655177116394,
+      "learning_rate": 1.6500000000000003e-06,
+      "loss": 0.7993,
+      "step": 33
+    },
+    {
+      "epoch": 0.038088125466766244,
+      "grad_norm": 1.0662972927093506,
+      "learning_rate": 1.7000000000000002e-06,
+      "loss": 0.8029,
+      "step": 34
+    },
+    {
+      "epoch": 0.039208364451082896,
+      "grad_norm": 1.109535574913025,
+      "learning_rate": 1.75e-06,
+      "loss": 0.8218,
+      "step": 35
+    },
+    {
+      "epoch": 0.040328603435399554,
+      "grad_norm": 1.0223349332809448,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 0.7871,
+      "step": 36
+    },
+    {
+      "epoch": 0.041448842419716206,
+      "grad_norm": 1.0227162837982178,
+      "learning_rate": 1.85e-06,
+      "loss": 0.7944,
+      "step": 37
+    },
+    {
+      "epoch": 0.04256908140403286,
+      "grad_norm": 0.9250116348266602,
+      "learning_rate": 1.9000000000000002e-06,
+      "loss": 0.7522,
+      "step": 38
+    },
+    {
+      "epoch": 0.043689320388349516,
+      "grad_norm": 0.9528149366378784,
+      "learning_rate": 1.9500000000000004e-06,
+      "loss": 0.7645,
+      "step": 39
+    },
+    {
+      "epoch": 0.04480955937266617,
+      "grad_norm": 0.8865862488746643,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 0.7861,
+      "step": 40
+    },
+    {
+      "epoch": 0.045929798356982826,
+      "grad_norm": 0.9337865114212036,
+      "learning_rate": 2.05e-06,
+      "loss": 0.7427,
+      "step": 41
+    },
+    {
+      "epoch": 0.04705003734129948,
+      "grad_norm": 0.880203127861023,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 0.7411,
+      "step": 42
+    },
+    {
+      "epoch": 0.04817027632561613,
+      "grad_norm": 0.8911599516868591,
+      "learning_rate": 2.15e-06,
+      "loss": 0.7434,
+      "step": 43
+    },
+    {
+      "epoch": 0.04929051530993279,
+      "grad_norm": 0.8979325294494629,
+      "learning_rate": 2.2e-06,
+      "loss": 0.7598,
+      "step": 44
+    },
+    {
+      "epoch": 0.05041075429424944,
+      "grad_norm": 0.8859411478042603,
+      "learning_rate": 2.25e-06,
+      "loss": 0.7646,
+      "step": 45
+    },
+    {
+      "epoch": 0.05153099327856609,
+      "grad_norm": 0.8641071915626526,
+      "learning_rate": 2.3000000000000004e-06,
+      "loss": 0.7559,
+      "step": 46
+    },
+    {
+      "epoch": 0.05265123226288275,
+      "grad_norm": 2.738354444503784,
+      "learning_rate": 2.35e-06,
+      "loss": 0.7826,
+      "step": 47
+    },
+    {
+      "epoch": 0.0537714712471994,
+      "grad_norm": 0.847560465335846,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 0.7469,
+      "step": 48
+    },
+    {
+      "epoch": 0.05489171023151606,
+      "grad_norm": 0.8350309729576111,
+      "learning_rate": 2.4500000000000003e-06,
+      "loss": 0.7878,
+      "step": 49
+    },
+    {
+      "epoch": 0.05601194921583271,
+      "grad_norm": 0.8858978748321533,
+      "learning_rate": 2.5e-06,
+      "loss": 0.7154,
+      "step": 50
+    },
+    {
+      "epoch": 0.05713218820014936,
+      "grad_norm": 0.8486636877059937,
+      "learning_rate": 2.55e-06,
+      "loss": 0.7521,
+      "step": 51
+    },
+    {
+      "epoch": 0.05825242718446602,
+      "grad_norm": 0.8095760345458984,
+      "learning_rate": 2.6e-06,
+      "loss": 0.7355,
+      "step": 52
+    },
+    {
+      "epoch": 0.05937266616878267,
+      "grad_norm": 1.1256295442581177,
+      "learning_rate": 2.6500000000000005e-06,
+      "loss": 0.7312,
+      "step": 53
+    },
+    {
+      "epoch": 0.060492905153099324,
+      "grad_norm": 0.8093723058700562,
+      "learning_rate": 2.7000000000000004e-06,
+      "loss": 0.738,
+      "step": 54
+    },
+    {
+      "epoch": 0.06161314413741598,
+      "grad_norm": 0.807765543460846,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 0.7055,
+      "step": 55
+    },
+    {
+      "epoch": 0.06273338312173264,
+      "grad_norm": 0.8036230206489563,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 0.7354,
+      "step": 56
+    },
+    {
+      "epoch": 0.0638536221060493,
+      "grad_norm": 0.809104859828949,
+      "learning_rate": 2.85e-06,
+      "loss": 0.7469,
+      "step": 57
+    },
+    {
+      "epoch": 0.06497386109036594,
+      "grad_norm": 0.8124133944511414,
+      "learning_rate": 2.9e-06,
+      "loss": 0.7039,
+      "step": 58
+    },
+    {
+      "epoch": 0.0660941000746826,
+      "grad_norm": 0.8039435744285583,
+      "learning_rate": 2.95e-06,
+      "loss": 0.7209,
+      "step": 59
+    },
+    {
+      "epoch": 0.06721433905899925,
+      "grad_norm": 0.8367541432380676,
+      "learning_rate": 3e-06,
+      "loss": 0.7759,
+      "step": 60
+    },
+    {
+      "epoch": 0.06833457804331591,
+      "grad_norm": 0.7983250021934509,
+      "learning_rate": 3.05e-06,
+      "loss": 0.7251,
+      "step": 61
+    },
+    {
+      "epoch": 0.06945481702763256,
+      "grad_norm": 0.7989967465400696,
+      "learning_rate": 3.1000000000000004e-06,
+      "loss": 0.7164,
+      "step": 62
+    },
+    {
+      "epoch": 0.07057505601194922,
+      "grad_norm": 0.7835646867752075,
+      "learning_rate": 3.1500000000000003e-06,
+      "loss": 0.7124,
+      "step": 63
+    },
+    {
+      "epoch": 0.07169529499626587,
+      "grad_norm": 0.7820047736167908,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 0.6997,
+      "step": 64
+    },
+    {
+      "epoch": 0.07281553398058252,
+      "grad_norm": 0.8633654713630676,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 0.7248,
+      "step": 65
+    },
+    {
+      "epoch": 0.07393577296489919,
+      "grad_norm": 0.7743967175483704,
+      "learning_rate": 3.3000000000000006e-06,
+      "loss": 0.7237,
+      "step": 66
+    },
+    {
+      "epoch": 0.07505601194921584,
+      "grad_norm": 0.7750105857849121,
+      "learning_rate": 3.3500000000000005e-06,
+      "loss": 0.6874,
+      "step": 67
+    },
+    {
+      "epoch": 0.07617625093353249,
+      "grad_norm": 0.7676857113838196,
+      "learning_rate": 3.4000000000000005e-06,
+      "loss": 0.6771,
+      "step": 68
+    },
+    {
+      "epoch": 0.07729648991784914,
+      "grad_norm": 0.7877640724182129,
+      "learning_rate": 3.45e-06,
+      "loss": 0.7063,
+      "step": 69
+    },
+    {
+      "epoch": 0.07841672890216579,
+      "grad_norm": 0.7854837775230408,
+      "learning_rate": 3.5e-06,
+      "loss": 0.6957,
+      "step": 70
+    },
+    {
+      "epoch": 0.07953696788648244,
+      "grad_norm": 0.8387333750724792,
+      "learning_rate": 3.5500000000000003e-06,
+      "loss": 0.6923,
+      "step": 71
+    },
+    {
+      "epoch": 0.08065720687079911,
+      "grad_norm": 0.7769498229026794,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 0.7229,
+      "step": 72
+    },
+    {
+      "epoch": 0.08177744585511576,
+      "grad_norm": 0.8273895978927612,
+      "learning_rate": 3.65e-06,
+      "loss": 0.7011,
+      "step": 73
+    },
+    {
+      "epoch": 0.08289768483943241,
+      "grad_norm": 0.7759560346603394,
+      "learning_rate": 3.7e-06,
+      "loss": 0.7246,
+      "step": 74
+    },
+    {
+      "epoch": 0.08401792382374906,
+      "grad_norm": 0.7503573298454285,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.6711,
+      "step": 75
+    },
+    {
+      "epoch": 0.08513816280806571,
+      "grad_norm": 0.7817795276641846,
+      "learning_rate": 3.8000000000000005e-06,
+      "loss": 0.6771,
+      "step": 76
+    },
+    {
+      "epoch": 0.08625840179238238,
+      "grad_norm": 0.7503968477249146,
+      "learning_rate": 3.85e-06,
+      "loss": 0.6903,
+      "step": 77
+    },
+    {
+      "epoch": 0.08737864077669903,
+      "grad_norm": 0.7876687049865723,
+      "learning_rate": 3.900000000000001e-06,
+      "loss": 0.6963,
+      "step": 78
+    },
+    {
+      "epoch": 0.08849887976101568,
+      "grad_norm": 0.7441173791885376,
+      "learning_rate": 3.95e-06,
+      "loss": 0.6872,
+      "step": 79
+    },
+    {
+      "epoch": 0.08961911874533234,
+      "grad_norm": 0.7741108536720276,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 0.6563,
+      "step": 80
+    },
+    {
+      "epoch": 0.09073935772964899,
+      "grad_norm": 0.7747477889060974,
+      "learning_rate": 4.05e-06,
+      "loss": 0.7077,
+      "step": 81
+    },
+    {
+      "epoch": 0.09185959671396565,
+      "grad_norm": 0.8173837065696716,
+      "learning_rate": 4.1e-06,
+      "loss": 0.708,
+      "step": 82
+    },
+    {
+      "epoch": 0.0929798356982823,
+      "grad_norm": 0.7583895325660706,
+      "learning_rate": 4.15e-06,
+      "loss": 0.6696,
+      "step": 83
+    },
+    {
+      "epoch": 0.09410007468259896,
+      "grad_norm": 0.7949450612068176,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 0.7041,
+      "step": 84
+    },
+    {
+      "epoch": 0.09522031366691561,
+      "grad_norm": 0.7842839360237122,
+      "learning_rate": 4.25e-06,
+      "loss": 0.6552,
+      "step": 85
+    },
+    {
+      "epoch": 0.09634055265123226,
+      "grad_norm": 0.769949734210968,
+      "learning_rate": 4.3e-06,
+      "loss": 0.7012,
+      "step": 86
+    },
+    {
+      "epoch": 0.09746079163554892,
+      "grad_norm": 0.7680435180664062,
+      "learning_rate": 4.350000000000001e-06,
+      "loss": 0.7026,
+      "step": 87
+    },
+    {
+      "epoch": 0.09858103061986558,
+      "grad_norm": 0.7704882621765137,
+      "learning_rate": 4.4e-06,
+      "loss": 0.676,
+      "step": 88
+    },
+    {
+      "epoch": 0.09970126960418223,
+      "grad_norm": 0.7687551379203796,
+      "learning_rate": 4.450000000000001e-06,
+      "loss": 0.662,
+      "step": 89
+    },
+    {
+      "epoch": 0.10082150858849888,
+      "grad_norm": 0.8296576738357544,
+      "learning_rate": 4.5e-06,
+      "loss": 0.6815,
+      "step": 90
+    },
+    {
+      "epoch": 0.10194174757281553,
+      "grad_norm": 0.807574987411499,
+      "learning_rate": 4.5500000000000005e-06,
+      "loss": 0.6959,
+      "step": 91
+    },
+    {
+      "epoch": 0.10306198655713218,
+      "grad_norm": 0.8537940979003906,
+      "learning_rate": 4.600000000000001e-06,
+      "loss": 0.6568,
+      "step": 92
+    },
+    {
+      "epoch": 0.10418222554144885,
+      "grad_norm": 0.776905357837677,
+      "learning_rate": 4.65e-06,
+      "loss": 0.7027,
+      "step": 93
+    },
+    {
+      "epoch": 0.1053024645257655,
+      "grad_norm": 0.7668991684913635,
+      "learning_rate": 4.7e-06,
+      "loss": 0.6537,
+      "step": 94
+    },
+    {
+      "epoch": 0.10642270351008215,
+      "grad_norm": 0.7789910435676575,
+      "learning_rate": 4.75e-06,
+      "loss": 0.6637,
+      "step": 95
+    },
+    {
+      "epoch": 0.1075429424943988,
+      "grad_norm": 0.7807944416999817,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 0.7103,
+      "step": 96
+    },
+    {
+      "epoch": 0.10866318147871545,
+      "grad_norm": 0.7590618133544922,
+      "learning_rate": 4.85e-06,
+      "loss": 0.6765,
+      "step": 97
+    },
+    {
+      "epoch": 0.10978342046303212,
+      "grad_norm": 0.7463409900665283,
+      "learning_rate": 4.9000000000000005e-06,
+      "loss": 0.6795,
+      "step": 98
+    },
+    {
+      "epoch": 0.11090365944734877,
+      "grad_norm": 0.8291321992874146,
+      "learning_rate": 4.95e-06,
+      "loss": 0.7311,
+      "step": 99
+    },
+    {
+      "epoch": 0.11202389843166542,
+      "grad_norm": 0.7995114326477051,
+      "learning_rate": 5e-06,
+      "loss": 0.684,
+      "step": 100
+    },
+    {
+      "epoch": 0.11314413741598207,
+      "grad_norm": 0.7649957537651062,
+      "learning_rate": 4.9999995527397415e-06,
+      "loss": 0.6645,
+      "step": 101
+    },
+    {
+      "epoch": 0.11426437640029873,
+      "grad_norm": 0.8313412070274353,
+      "learning_rate": 4.999998210959126e-06,
+      "loss": 0.6656,
+      "step": 102
+    },
+    {
+      "epoch": 0.11538461538461539,
+      "grad_norm": 0.7994295358657837,
+      "learning_rate": 4.999995974658632e-06,
+      "loss": 0.6562,
+      "step": 103
+    },
+    {
+      "epoch": 0.11650485436893204,
+      "grad_norm": 0.8024242520332336,
+      "learning_rate": 4.999992843839062e-06,
+      "loss": 0.6946,
+      "step": 104
+    },
+    {
+      "epoch": 0.1176250933532487,
+      "grad_norm": 0.7950218319892883,
+      "learning_rate": 4.999988818501535e-06,
+      "loss": 0.6833,
+      "step": 105
+    },
+    {
+      "epoch": 0.11874533233756535,
+      "grad_norm": 0.8187536597251892,
+      "learning_rate": 4.9999838986474925e-06,
+      "loss": 0.6869,
+      "step": 106
+    },
+    {
+      "epoch": 0.119865571321882,
+      "grad_norm": 0.7608385682106018,
+      "learning_rate": 4.999978084278693e-06,
+      "loss": 0.6536,
+      "step": 107
+    },
+    {
+      "epoch": 0.12098581030619865,
+      "grad_norm": 0.9910794496536255,
+      "learning_rate": 4.999971375397218e-06,
+      "loss": 0.6444,
+      "step": 108
+    },
+    {
+      "epoch": 0.12210604929051531,
+      "grad_norm": 0.8221011757850647,
+      "learning_rate": 4.9999637720054675e-06,
+      "loss": 0.6986,
+      "step": 109
+    },
+    {
+      "epoch": 0.12322628827483197,
+      "grad_norm": 0.8212369680404663,
+      "learning_rate": 4.999955274106163e-06,
+      "loss": 0.6559,
+      "step": 110
+    },
+    {
+      "epoch": 0.12434652725914862,
+      "grad_norm": 0.7603834867477417,
+      "learning_rate": 4.999945881702344e-06,
+      "loss": 0.6644,
+      "step": 111
+    },
+    {
+      "epoch": 0.12546676624346528,
+      "grad_norm": 0.864615261554718,
+      "learning_rate": 4.999935594797372e-06,
+      "loss": 0.7045,
+      "step": 112
+    },
+    {
+      "epoch": 0.12658700522778193,
+      "grad_norm": 0.7939414381980896,
+      "learning_rate": 4.999924413394926e-06,
+      "loss": 0.6506,
+      "step": 113
+    },
+    {
+      "epoch": 0.1277072442120986,
+      "grad_norm": 0.7965752482414246,
+      "learning_rate": 4.99991233749901e-06,
+      "loss": 0.6634,
+      "step": 114
+    },
+    {
+      "epoch": 0.12882748319641524,
+      "grad_norm": 0.809953510761261,
+      "learning_rate": 4.999899367113943e-06,
+      "loss": 0.6613,
+      "step": 115
+    },
+    {
+      "epoch": 0.1299477221807319,
+      "grad_norm": 0.7917526960372925,
+      "learning_rate": 4.999885502244365e-06,
+      "loss": 0.6503,
+      "step": 116
+    },
+    {
+      "epoch": 0.13106796116504854,
+      "grad_norm": 0.7986327409744263,
+      "learning_rate": 4.999870742895239e-06,
+      "loss": 0.6607,
+      "step": 117
+    },
+    {
+      "epoch": 0.1321882001493652,
+      "grad_norm": 1.0296788215637207,
+      "learning_rate": 4.999855089071844e-06,
+      "loss": 0.6705,
+      "step": 118
+    },
+    {
+      "epoch": 0.13330843913368184,
+      "grad_norm": 0.7908150553703308,
+      "learning_rate": 4.999838540779782e-06,
+      "loss": 0.6622,
+      "step": 119
+    },
+    {
+      "epoch": 0.1344286781179985,
+      "grad_norm": 0.8663723468780518,
+      "learning_rate": 4.999821098024975e-06,
+      "loss": 0.6592,
+      "step": 120
+    },
+    {
+      "epoch": 0.13554891710231515,
+      "grad_norm": 0.8147454857826233,
+      "learning_rate": 4.999802760813662e-06,
+      "loss": 0.6478,
+      "step": 121
+    },
+    {
+      "epoch": 0.13666915608663183,
+      "grad_norm": 0.810529351234436,
+      "learning_rate": 4.9997835291524065e-06,
+      "loss": 0.6438,
+      "step": 122
+    },
+    {
+      "epoch": 0.13778939507094848,
+      "grad_norm": 0.7966436743736267,
+      "learning_rate": 4.999763403048089e-06,
+      "loss": 0.6425,
+      "step": 123
+    },
+    {
+      "epoch": 0.13890963405526513,
+      "grad_norm": 0.789413571357727,
+      "learning_rate": 4.999742382507909e-06,
+      "loss": 0.6227,
+      "step": 124
+    },
+    {
+      "epoch": 0.14002987303958178,
+      "grad_norm": 0.7998204827308655,
+      "learning_rate": 4.999720467539391e-06,
+      "loss": 0.6635,
+      "step": 125
+    },
+    {
+      "epoch": 0.14115011202389843,
+      "grad_norm": 0.8009337782859802,
+      "learning_rate": 4.999697658150374e-06,
+      "loss": 0.6673,
+      "step": 126
+    },
+    {
+      "epoch": 0.14227035100821508,
+      "grad_norm": 0.7708600163459778,
+      "learning_rate": 4.99967395434902e-06,
+      "loss": 0.6336,
+      "step": 127
+    },
+    {
+      "epoch": 0.14339058999253174,
+      "grad_norm": 0.7975388765335083,
+      "learning_rate": 4.999649356143811e-06,
+      "loss": 0.6673,
+      "step": 128
+    },
+    {
+      "epoch": 0.1445108289768484,
+      "grad_norm": 0.83116215467453,
+      "learning_rate": 4.999623863543548e-06,
+      "loss": 0.6902,
+      "step": 129
+    },
+    {
+      "epoch": 0.14563106796116504,
+      "grad_norm": 0.8246234655380249,
+      "learning_rate": 4.999597476557352e-06,
+      "loss": 0.6586,
+      "step": 130
+    },
+    {
+      "epoch": 0.1467513069454817,
+      "grad_norm": 0.7604647278785706,
+      "learning_rate": 4.999570195194665e-06,
+      "loss": 0.6455,
+      "step": 131
+    },
+    {
+      "epoch": 0.14787154592979837,
+      "grad_norm": 0.7587922215461731,
+      "learning_rate": 4.9995420194652475e-06,
+      "loss": 0.6107,
+      "step": 132
+    },
+    {
+      "epoch": 0.14899178491411502,
+      "grad_norm": 0.909156858921051,
+      "learning_rate": 4.999512949379183e-06,
+      "loss": 0.679,
+      "step": 133
+    },
+    {
+      "epoch": 0.15011202389843167,
+      "grad_norm": 0.8901562690734863,
+      "learning_rate": 4.999482984946872e-06,
+      "loss": 0.6887,
+      "step": 134
+    },
+    {
+      "epoch": 0.15123226288274833,
+      "grad_norm": 0.7636528611183167,
+      "learning_rate": 4.999452126179035e-06,
+      "loss": 0.6514,
+      "step": 135
+    },
+    {
+      "epoch": 0.15235250186706498,
+      "grad_norm": 0.8096560835838318,
+      "learning_rate": 4.999420373086715e-06,
+      "loss": 0.6937,
+      "step": 136
+    },
+    {
+      "epoch": 0.15347274085138163,
+      "grad_norm": 0.8607243299484253,
+      "learning_rate": 4.999387725681274e-06,
+      "loss": 0.6812,
+      "step": 137
+    },
+    {
+      "epoch": 0.15459297983569828,
+      "grad_norm": 0.8863621354103088,
+      "learning_rate": 4.9993541839743906e-06,
+      "loss": 0.6829,
+      "step": 138
+    },
+    {
+      "epoch": 0.15571321882001493,
+      "grad_norm": 0.8292368054389954,
+      "learning_rate": 4.99931974797807e-06,
+      "loss": 0.6714,
+      "step": 139
+    },
+    {
+      "epoch": 0.15683345780433158,
+      "grad_norm": 0.8053487539291382,
+      "learning_rate": 4.999284417704631e-06,
+      "loss": 0.656,
+      "step": 140
+    },
+    {
+      "epoch": 0.15795369678864823,
+      "grad_norm": 0.8153090476989746,
+      "learning_rate": 4.999248193166715e-06,
+      "loss": 0.6529,
+      "step": 141
+    },
+    {
+      "epoch": 0.1590739357729649,
+      "grad_norm": 0.8820142149925232,
+      "learning_rate": 4.999211074377285e-06,
+      "loss": 0.6507,
+      "step": 142
+    },
+    {
+      "epoch": 0.16019417475728157,
+      "grad_norm": 0.8246594071388245,
+      "learning_rate": 4.999173061349623e-06,
+      "loss": 0.6767,
+      "step": 143
+    },
+    {
+      "epoch": 0.16131441374159822,
+      "grad_norm": 0.7902995944023132,
+      "learning_rate": 4.999134154097328e-06,
+      "loss": 0.6596,
+      "step": 144
+    },
+    {
+      "epoch": 0.16243465272591487,
+      "grad_norm": 0.8101757764816284,
+      "learning_rate": 4.999094352634322e-06,
+      "loss": 0.6462,
+      "step": 145
+    },
+    {
+      "epoch": 0.16355489171023152,
+      "grad_norm": 0.8051759600639343,
+      "learning_rate": 4.999053656974847e-06,
+      "loss": 0.6659,
+      "step": 146
+    },
+    {
+      "epoch": 0.16467513069454817,
+      "grad_norm": 0.8006038665771484,
+      "learning_rate": 4.999012067133465e-06,
+      "loss": 0.6865,
+      "step": 147
+    },
+    {
+      "epoch": 0.16579536967886482,
+      "grad_norm": 0.8025104999542236,
+      "learning_rate": 4.998969583125055e-06,
+      "loss": 0.6558,
+      "step": 148
+    },
+    {
+      "epoch": 0.16691560866318148,
+      "grad_norm": 0.7755445241928101,
+      "learning_rate": 4.99892620496482e-06,
+      "loss": 0.6373,
+      "step": 149
+    },
+    {
+      "epoch": 0.16803584764749813,
+      "grad_norm": 0.8465038537979126,
+      "learning_rate": 4.99888193266828e-06,
+      "loss": 0.6476,
+      "step": 150
+    },
+    {
+      "epoch": 0.16915608663181478,
+      "grad_norm": 0.8304318189620972,
+      "learning_rate": 4.9988367662512775e-06,
+      "loss": 0.6636,
+      "step": 151
+    },
+    {
+      "epoch": 0.17027632561613143,
+      "grad_norm": 1.0744889974594116,
+      "learning_rate": 4.99879070572997e-06,
+      "loss": 0.6329,
+      "step": 152
+    },
+    {
+      "epoch": 0.1713965646004481,
+      "grad_norm": 0.8096840381622314,
+      "learning_rate": 4.998743751120843e-06,
+      "loss": 0.6515,
+      "step": 153
+    },
+    {
+      "epoch": 0.17251680358476476,
+      "grad_norm": 0.8647177219390869,
+      "learning_rate": 4.998695902440693e-06,
+      "loss": 0.669,
+      "step": 154
+    },
+    {
+      "epoch": 0.1736370425690814,
+      "grad_norm": 0.8000180125236511,
+      "learning_rate": 4.9986471597066435e-06,
+      "loss": 0.6353,
+      "step": 155
+    },
+    {
+      "epoch": 0.17475728155339806,
+      "grad_norm": 0.7994810938835144,
+      "learning_rate": 4.998597522936134e-06,
+      "loss": 0.6312,
+      "step": 156
+    },
+    {
+      "epoch": 0.17587752053771472,
+      "grad_norm": 0.8086467385292053,
+      "learning_rate": 4.9985469921469245e-06,
+      "loss": 0.6426,
+      "step": 157
+    },
+    {
+      "epoch": 0.17699775952203137,
+      "grad_norm": 0.8089617490768433,
+      "learning_rate": 4.998495567357097e-06,
+      "loss": 0.6671,
+      "step": 158
+    },
+    {
+      "epoch": 0.17811799850634802,
+      "grad_norm": 0.7937355041503906,
+      "learning_rate": 4.9984432485850495e-06,
+      "loss": 0.6433,
+      "step": 159
+    },
+    {
+      "epoch": 0.17923823749066467,
+      "grad_norm": 0.7971526980400085,
+      "learning_rate": 4.998390035849503e-06,
+      "loss": 0.6371,
+      "step": 160
+    },
+    {
+      "epoch": 0.18035847647498132,
+      "grad_norm": 0.8417288064956665,
+      "learning_rate": 4.998335929169498e-06,
+      "loss": 0.6607,
+      "step": 161
+    },
+    {
+      "epoch": 0.18147871545929797,
+      "grad_norm": 2.7477307319641113,
+      "learning_rate": 4.998280928564394e-06,
+      "loss": 0.6445,
+      "step": 162
+    },
+    {
+      "epoch": 0.18259895444361463,
+      "grad_norm": 0.8154010772705078,
+      "learning_rate": 4.998225034053871e-06,
+      "loss": 0.6518,
+      "step": 163
+    },
+    {
+      "epoch": 0.1837191934279313,
+      "grad_norm": 0.8180714249610901,
+      "learning_rate": 4.9981682456579275e-06,
+      "loss": 0.6524,
+      "step": 164
+    },
+    {
+      "epoch": 0.18483943241224796,
+      "grad_norm": 0.8008469939231873,
+      "learning_rate": 4.998110563396883e-06,
+      "loss": 0.6508,
+      "step": 165
+    },
+    {
+      "epoch": 0.1859596713965646,
+      "grad_norm": 0.7731476426124573,
+      "learning_rate": 4.998051987291378e-06,
+      "loss": 0.6262,
+      "step": 166
+    },
+    {
+      "epoch": 0.18707991038088126,
+      "grad_norm": 0.7730596661567688,
+      "learning_rate": 4.99799251736237e-06,
+      "loss": 0.6356,
+      "step": 167
+    },
+    {
+      "epoch": 0.1882001493651979,
+      "grad_norm": 0.8305966854095459,
+      "learning_rate": 4.997932153631139e-06,
+      "loss": 0.6584,
+      "step": 168
+    },
+    {
+      "epoch": 0.18932038834951456,
+      "grad_norm": 0.8761897683143616,
+      "learning_rate": 4.997870896119282e-06,
+      "loss": 0.6593,
+      "step": 169
+    },
+    {
+      "epoch": 0.19044062733383121,
+      "grad_norm": 0.8341963887214661,
+      "learning_rate": 4.9978087448487204e-06,
+      "loss": 0.6293,
+      "step": 170
+    },
+    {
+      "epoch": 0.19156086631814787,
+      "grad_norm": 0.8244994282722473,
+      "learning_rate": 4.997745699841689e-06,
+      "loss": 0.6512,
+      "step": 171
+    },
+    {
+      "epoch": 0.19268110530246452,
+      "grad_norm": 0.8051140904426575,
+      "learning_rate": 4.997681761120749e-06,
+      "loss": 0.6535,
+      "step": 172
+    },
+    {
+      "epoch": 0.19380134428678117,
+      "grad_norm": 0.7865309119224548,
+      "learning_rate": 4.997616928708774e-06,
+      "loss": 0.6457,
+      "step": 173
+    },
+    {
+      "epoch": 0.19492158327109785,
+      "grad_norm": 0.7985602021217346,
+      "learning_rate": 4.997551202628966e-06,
+      "loss": 0.6562,
+      "step": 174
+    },
+    {
+      "epoch": 0.1960418222554145,
+      "grad_norm": 0.7812699675559998,
+      "learning_rate": 4.99748458290484e-06,
+      "loss": 0.6505,
+      "step": 175
+    },
+    {
+      "epoch": 0.19716206123973115,
+      "grad_norm": 0.7868354916572571,
+      "learning_rate": 4.997417069560234e-06,
+      "loss": 0.6295,
+      "step": 176
+    },
+    {
+      "epoch": 0.1982823002240478,
+      "grad_norm": 0.8139058947563171,
+      "learning_rate": 4.997348662619305e-06,
+      "loss": 0.6637,
+      "step": 177
+    },
+    {
+      "epoch": 0.19940253920836445,
+      "grad_norm": 0.7864823937416077,
+      "learning_rate": 4.997279362106528e-06,
+      "loss": 0.6478,
+      "step": 178
+    },
+    {
+      "epoch": 0.2005227781926811,
+      "grad_norm": 0.8095872402191162,
+      "learning_rate": 4.9972091680467e-06,
+      "loss": 0.668,
+      "step": 179
+    },
+    {
+      "epoch": 0.20164301717699776,
+      "grad_norm": 0.8363040089607239,
+      "learning_rate": 4.997138080464938e-06,
+      "loss": 0.6757,
+      "step": 180
+    },
+    {
+      "epoch": 0.2027632561613144,
+      "grad_norm": 0.8066425323486328,
+      "learning_rate": 4.9970660993866756e-06,
+      "loss": 0.6617,
+      "step": 181
+    },
+    {
+      "epoch": 0.20388349514563106,
+      "grad_norm": 0.8129490613937378,
+      "learning_rate": 4.996993224837672e-06,
+      "loss": 0.6463,
+      "step": 182
+    },
+    {
+      "epoch": 0.2050037341299477,
+      "grad_norm": 0.8092365264892578,
+      "learning_rate": 4.996919456843998e-06,
+      "loss": 0.6441,
+      "step": 183
+    },
+    {
+      "epoch": 0.20612397311426436,
+      "grad_norm": 0.7847099900245667,
+      "learning_rate": 4.996844795432051e-06,
+      "loss": 0.6608,
+      "step": 184
+    },
+    {
+      "epoch": 0.20724421209858104,
+      "grad_norm": 0.7771169543266296,
+      "learning_rate": 4.9967692406285454e-06,
+      "loss": 0.6198,
+      "step": 185
+    },
+    {
+      "epoch": 0.2083644510828977,
+      "grad_norm": 0.8057869076728821,
+      "learning_rate": 4.996692792460513e-06,
+      "loss": 0.6312,
+      "step": 186
+    },
+    {
+      "epoch": 0.20948469006721435,
+      "grad_norm": 0.8194419145584106,
+      "learning_rate": 4.996615450955311e-06,
+      "loss": 0.66,
+      "step": 187
+    },
+    {
+      "epoch": 0.210604929051531,
+      "grad_norm": 0.8123747110366821,
+      "learning_rate": 4.99653721614061e-06,
+      "loss": 0.6527,
+      "step": 188
+    },
+    {
+      "epoch": 0.21172516803584765,
+      "grad_norm": 0.8024148344993591,
+      "learning_rate": 4.996458088044405e-06,
+      "loss": 0.6426,
+      "step": 189
+    },
+    {
+      "epoch": 0.2128454070201643,
+      "grad_norm": 0.8563250303268433,
+      "learning_rate": 4.996378066695008e-06,
+      "loss": 0.6444,
+      "step": 190
+    },
+    {
+      "epoch": 0.21396564600448095,
+      "grad_norm": 0.8127154111862183,
+      "learning_rate": 4.99629715212105e-06,
+      "loss": 0.6478,
+      "step": 191
+    },
+    {
+      "epoch": 0.2150858849887976,
+      "grad_norm": 0.816790759563446,
+      "learning_rate": 4.996215344351485e-06,
+      "loss": 0.6304,
+      "step": 192
+    },
+    {
+      "epoch": 0.21620612397311426,
+      "grad_norm": 0.8222427368164062,
+      "learning_rate": 4.9961326434155834e-06,
+      "loss": 0.6418,
+      "step": 193
+    },
+    {
+      "epoch": 0.2173263629574309,
+      "grad_norm": 0.8139140605926514,
+      "learning_rate": 4.996049049342936e-06,
+      "loss": 0.6305,
+      "step": 194
+    },
+    {
+      "epoch": 0.21844660194174756,
+      "grad_norm": 0.8003841042518616,
+      "learning_rate": 4.995964562163455e-06,
+      "loss": 0.6474,
+      "step": 195
+    },
+    {
+      "epoch": 0.21956684092606424,
+      "grad_norm": 0.8458191156387329,
+      "learning_rate": 4.9958791819073684e-06,
+      "loss": 0.6637,
+      "step": 196
+    },
+    {
+      "epoch": 0.2206870799103809,
+      "grad_norm": 0.801009476184845,
+      "learning_rate": 4.995792908605227e-06,
+      "loss": 0.6388,
+      "step": 197
+    },
+    {
+      "epoch": 0.22180731889469754,
+      "grad_norm": 0.8174487948417664,
+      "learning_rate": 4.995705742287901e-06,
+      "loss": 0.653,
+      "step": 198
+    },
+    {
+      "epoch": 0.2229275578790142,
+      "grad_norm": 0.814853310585022,
+      "learning_rate": 4.995617682986578e-06,
+      "loss": 0.6217,
+      "step": 199
+    },
+    {
+      "epoch": 0.22404779686333084,
+      "grad_norm": 0.8986819982528687,
+      "learning_rate": 4.995528730732766e-06,
+      "loss": 0.6547,
+      "step": 200
+    },
+    {
+      "epoch": 0.2251680358476475,
+      "grad_norm": 0.8147602081298828,
+      "learning_rate": 4.995438885558294e-06,
+      "loss": 0.6475,
+      "step": 201
+    },
+    {
+      "epoch": 0.22628827483196415,
+      "grad_norm": 0.8108645081520081,
+      "learning_rate": 4.995348147495309e-06,
+      "loss": 0.6434,
+      "step": 202
+    },
+    {
+      "epoch": 0.2274085138162808,
+      "grad_norm": 0.8559293746948242,
+      "learning_rate": 4.995256516576278e-06,
+      "loss": 0.6559,
+      "step": 203
+    },
+    {
+      "epoch": 0.22852875280059745,
+      "grad_norm": 0.810727596282959,
+      "learning_rate": 4.995163992833986e-06,
+      "loss": 0.6365,
+      "step": 204
+    },
+    {
+      "epoch": 0.2296489917849141,
+      "grad_norm": 0.8541222214698792,
+      "learning_rate": 4.99507057630154e-06,
+      "loss": 0.6624,
+      "step": 205
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.8096019625663757,
+      "learning_rate": 4.994976267012366e-06,
+      "loss": 0.6213,
+      "step": 206
+    },
+    {
+      "epoch": 0.23188946975354743,
+      "grad_norm": 0.868386447429657,
+      "learning_rate": 4.994881065000206e-06,
+      "loss": 0.6397,
+      "step": 207
+    },
+    {
+      "epoch": 0.23300970873786409,
+      "grad_norm": 0.8383010029792786,
+      "learning_rate": 4.9947849702991275e-06,
+      "loss": 0.6477,
+      "step": 208
+    },
+    {
+      "epoch": 0.23412994772218074,
+      "grad_norm": 0.8286524415016174,
+      "learning_rate": 4.994687982943511e-06,
+      "loss": 0.6321,
+      "step": 209
+    },
+    {
+      "epoch": 0.2352501867064974,
+      "grad_norm": 0.7969548106193542,
+      "learning_rate": 4.994590102968061e-06,
+      "loss": 0.6284,
+      "step": 210
+    },
+    {
+      "epoch": 0.23637042569081404,
+      "grad_norm": 0.8124633431434631,
+      "learning_rate": 4.994491330407799e-06,
+      "loss": 0.6508,
+      "step": 211
+    },
+    {
+      "epoch": 0.2374906646751307,
+      "grad_norm": 0.8430159687995911,
+      "learning_rate": 4.994391665298066e-06,
+      "loss": 0.6392,
+      "step": 212
+    },
+    {
+      "epoch": 0.23861090365944734,
+      "grad_norm": 1.1165465116500854,
+      "learning_rate": 4.994291107674525e-06,
+      "loss": 0.6609,
+      "step": 213
+    },
+    {
+      "epoch": 0.239731142643764,
+      "grad_norm": 0.7861813902854919,
+      "learning_rate": 4.9941896575731555e-06,
+      "loss": 0.627,
+      "step": 214
+    },
+    {
+      "epoch": 0.24085138162808065,
+      "grad_norm": 0.8439143300056458,
+      "learning_rate": 4.994087315030256e-06,
+      "loss": 0.6511,
+      "step": 215
+    },
+    {
+      "epoch": 0.2419716206123973,
+      "grad_norm": 0.9279224276542664,
+      "learning_rate": 4.993984080082447e-06,
+      "loss": 0.6276,
+      "step": 216
+    },
+    {
+      "epoch": 0.24309185959671398,
+      "grad_norm": 0.8302883505821228,
+      "learning_rate": 4.993879952766666e-06,
+      "loss": 0.6399,
+      "step": 217
+    },
+    {
+      "epoch": 0.24421209858103063,
+      "grad_norm": 0.772523820400238,
+      "learning_rate": 4.993774933120171e-06,
+      "loss": 0.6284,
+      "step": 218
+    },
+    {
+      "epoch": 0.24533233756534728,
+      "grad_norm": 0.8319805264472961,
+      "learning_rate": 4.993669021180539e-06,
+      "loss": 0.6407,
+      "step": 219
+    },
+    {
+      "epoch": 0.24645257654966393,
+      "grad_norm": 0.8110228776931763,
+      "learning_rate": 4.993562216985665e-06,
+      "loss": 0.6598,
+      "step": 220
+    },
+    {
+      "epoch": 0.24757281553398058,
+      "grad_norm": 0.8476558327674866,
+      "learning_rate": 4.9934545205737665e-06,
+      "loss": 0.6105,
+      "step": 221
+    },
+    {
+      "epoch": 0.24869305451829724,
+      "grad_norm": 0.8189626932144165,
+      "learning_rate": 4.993345931983376e-06,
+      "loss": 0.6224,
+      "step": 222
+    },
+    {
+      "epoch": 0.2498132935026139,
+      "grad_norm": 0.8342490792274475,
+      "learning_rate": 4.9932364512533485e-06,
+      "loss": 0.6217,
+      "step": 223
+    },
+    {
+      "epoch": 0.25093353248693057,
+      "grad_norm": 0.838699460029602,
+      "learning_rate": 4.993126078422856e-06,
+      "loss": 0.6432,
+      "step": 224
+    },
+    {
+      "epoch": 0.2520537714712472,
+      "grad_norm": 0.801331102848053,
+      "learning_rate": 4.9930148135313935e-06,
+      "loss": 0.6219,
+      "step": 225
+    },
+    {
+      "epoch": 0.25317401045556387,
+      "grad_norm": 0.8075932860374451,
+      "learning_rate": 4.99290265661877e-06,
+      "loss": 0.649,
+      "step": 226
+    },
+    {
+      "epoch": 0.2542942494398805,
+      "grad_norm": 0.8010944724082947,
+      "learning_rate": 4.992789607725116e-06,
+      "loss": 0.6159,
+      "step": 227
+    },
+    {
+      "epoch": 0.2554144884241972,
+      "grad_norm": 0.8093600273132324,
+      "learning_rate": 4.992675666890884e-06,
+      "loss": 0.6162,
+      "step": 228
+    },
+    {
+      "epoch": 0.2565347274085138,
+      "grad_norm": 0.8651375770568848,
+      "learning_rate": 4.99256083415684e-06,
+      "loss": 0.6391,
+      "step": 229
+    },
+    {
+      "epoch": 0.2576549663928305,
+      "grad_norm": 0.8180515766143799,
+      "learning_rate": 4.992445109564074e-06,
+      "loss": 0.6627,
+      "step": 230
+    },
+    {
+      "epoch": 0.2587752053771471,
+      "grad_norm": 0.8492463231086731,
+      "learning_rate": 4.992328493153992e-06,
+      "loss": 0.6533,
+      "step": 231
+    },
+    {
+      "epoch": 0.2598954443614638,
+      "grad_norm": 0.7910256385803223,
+      "learning_rate": 4.992210984968321e-06,
+      "loss": 0.6212,
+      "step": 232
+    },
+    {
+      "epoch": 0.26101568334578046,
+      "grad_norm": 0.8081009984016418,
+      "learning_rate": 4.992092585049106e-06,
+      "loss": 0.6464,
+      "step": 233
+    },
+    {
+      "epoch": 0.2621359223300971,
+      "grad_norm": 0.8275237083435059,
+      "learning_rate": 4.991973293438712e-06,
+      "loss": 0.6447,
+      "step": 234
+    },
+    {
+      "epoch": 0.26325616131441376,
+      "grad_norm": 0.7819504141807556,
+      "learning_rate": 4.991853110179822e-06,
+      "loss": 0.6278,
+      "step": 235
+    },
+    {
+      "epoch": 0.2643764002987304,
+      "grad_norm": 0.7954742908477783,
+      "learning_rate": 4.991732035315439e-06,
+      "loss": 0.6259,
+      "step": 236
+    },
+    {
+      "epoch": 0.26549663928304706,
+      "grad_norm": 0.7841144800186157,
+      "learning_rate": 4.9916100688888835e-06,
+      "loss": 0.6209,
+      "step": 237
+    },
+    {
+      "epoch": 0.2666168782673637,
+      "grad_norm": 0.783143937587738,
+      "learning_rate": 4.9914872109437975e-06,
+      "loss": 0.6015,
+      "step": 238
+    },
+    {
+      "epoch": 0.26773711725168037,
+      "grad_norm": 0.8399513363838196,
+      "learning_rate": 4.9913634615241394e-06,
+      "loss": 0.6624,
+      "step": 239
+    },
+    {
+      "epoch": 0.268857356235997,
+      "grad_norm": 0.8158865571022034,
+      "learning_rate": 4.9912388206741895e-06,
+      "loss": 0.6607,
+      "step": 240
+    },
+    {
+      "epoch": 0.26997759522031367,
+      "grad_norm": 0.8276694416999817,
+      "learning_rate": 4.991113288438543e-06,
+      "loss": 0.6474,
+      "step": 241
+    },
+    {
+      "epoch": 0.2710978342046303,
+      "grad_norm": 0.8164044618606567,
+      "learning_rate": 4.9909868648621185e-06,
+      "loss": 0.6352,
+      "step": 242
+    },
+    {
+      "epoch": 0.272218073188947,
+      "grad_norm": 0.8372899889945984,
+      "learning_rate": 4.99085954999015e-06,
+      "loss": 0.638,
+      "step": 243
+    },
+    {
+      "epoch": 0.27333831217326365,
+      "grad_norm": 0.8520254492759705,
+      "learning_rate": 4.990731343868192e-06,
+      "loss": 0.6591,
+      "step": 244
+    },
+    {
+      "epoch": 0.2744585511575803,
+      "grad_norm": 0.8587496876716614,
+      "learning_rate": 4.990602246542118e-06,
+      "loss": 0.6442,
+      "step": 245
+    },
+    {
+      "epoch": 0.27557879014189696,
+      "grad_norm": 0.8117641806602478,
+      "learning_rate": 4.99047225805812e-06,
+      "loss": 0.6169,
+      "step": 246
+    },
+    {
+      "epoch": 0.2766990291262136,
+      "grad_norm": 0.7888590693473816,
+      "learning_rate": 4.99034137846271e-06,
+      "loss": 0.6345,
+      "step": 247
+    },
+    {
+      "epoch": 0.27781926811053026,
+      "grad_norm": 0.8112688064575195,
+      "learning_rate": 4.990209607802715e-06,
+      "loss": 0.6246,
+      "step": 248
+    },
+    {
+      "epoch": 0.2789395070948469,
+      "grad_norm": 0.8172190189361572,
+      "learning_rate": 4.990076946125286e-06,
+      "loss": 0.6323,
+      "step": 249
+    },
+    {
+      "epoch": 0.28005974607916356,
+      "grad_norm": 0.7784557938575745,
+      "learning_rate": 4.9899433934778905e-06,
+      "loss": 0.6186,
+      "step": 250
+    },
+    {
+      "epoch": 0.2811799850634802,
+      "grad_norm": 0.8571611642837524,
+      "learning_rate": 4.9898089499083135e-06,
+      "loss": 0.6157,
+      "step": 251
+    },
+    {
+      "epoch": 0.28230022404779687,
+      "grad_norm": 0.798240065574646,
+      "learning_rate": 4.989673615464661e-06,
+      "loss": 0.6201,
+      "step": 252
+    },
+    {
+      "epoch": 0.2834204630321135,
+      "grad_norm": 0.8300586342811584,
+      "learning_rate": 4.989537390195356e-06,
+      "loss": 0.6481,
+      "step": 253
+    },
+    {
+      "epoch": 0.28454070201643017,
+      "grad_norm": 0.9545755982398987,
+      "learning_rate": 4.989400274149141e-06,
+      "loss": 0.6315,
+      "step": 254
+    },
+    {
+      "epoch": 0.28566094100074685,
+      "grad_norm": 0.80037522315979,
+      "learning_rate": 4.989262267375078e-06,
+      "loss": 0.6206,
+      "step": 255
+    },
+    {
+      "epoch": 0.2867811799850635,
+      "grad_norm": 0.8587538003921509,
+      "learning_rate": 4.989123369922547e-06,
+      "loss": 0.622,
+      "step": 256
+    },
+    {
+      "epoch": 0.28790141896938015,
+      "grad_norm": 0.7835617661476135,
+      "learning_rate": 4.988983581841246e-06,
+      "loss": 0.6255,
+      "step": 257
+    },
+    {
+      "epoch": 0.2890216579536968,
+      "grad_norm": 0.8115828037261963,
+      "learning_rate": 4.988842903181192e-06,
+      "loss": 0.6122,
+      "step": 258
+    },
+    {
+      "epoch": 0.29014189693801345,
+      "grad_norm": 0.7731125950813293,
+      "learning_rate": 4.988701333992722e-06,
+      "loss": 0.5881,
+      "step": 259
+    },
+    {
+      "epoch": 0.2912621359223301,
+      "grad_norm": 0.8486718535423279,
+      "learning_rate": 4.988558874326489e-06,
+      "loss": 0.676,
+      "step": 260
+    },
+    {
+      "epoch": 0.29238237490664676,
+      "grad_norm": 0.7868174910545349,
+      "learning_rate": 4.988415524233469e-06,
+      "loss": 0.6267,
+      "step": 261
+    },
+    {
+      "epoch": 0.2935026138909634,
+      "grad_norm": 0.7890263795852661,
+      "learning_rate": 4.9882712837649515e-06,
+      "loss": 0.6171,
+      "step": 262
+    },
+    {
+      "epoch": 0.29462285287528006,
+      "grad_norm": 0.799913763999939,
+      "learning_rate": 4.988126152972548e-06,
+      "loss": 0.6353,
+      "step": 263
+    },
+    {
+      "epoch": 0.29574309185959674,
+      "grad_norm": 0.8258731961250305,
+      "learning_rate": 4.987980131908186e-06,
+      "loss": 0.6371,
+      "step": 264
+    },
+    {
+      "epoch": 0.29686333084391336,
+      "grad_norm": 0.8256173729896545,
+      "learning_rate": 4.987833220624116e-06,
+      "loss": 0.609,
+      "step": 265
+    },
+    {
+      "epoch": 0.29798356982823004,
+      "grad_norm": 0.7728785872459412,
+      "learning_rate": 4.9876854191729e-06,
+      "loss": 0.6021,
+      "step": 266
+    },
+    {
+      "epoch": 0.29910380881254667,
+      "grad_norm": 0.8612363338470459,
+      "learning_rate": 4.987536727607426e-06,
+      "loss": 0.6031,
+      "step": 267
+    },
+    {
+      "epoch": 0.30022404779686335,
+      "grad_norm": 0.8287588357925415,
+      "learning_rate": 4.987387145980896e-06,
+      "loss": 0.6365,
+      "step": 268
+    },
+    {
+      "epoch": 0.30134428678117997,
+      "grad_norm": 0.7820507884025574,
+      "learning_rate": 4.9872366743468305e-06,
+      "loss": 0.6257,
+      "step": 269
+    },
+    {
+      "epoch": 0.30246452576549665,
+      "grad_norm": 0.8158690929412842,
+      "learning_rate": 4.987085312759071e-06,
+      "loss": 0.6465,
+      "step": 270
+    },
+    {
+      "epoch": 0.3035847647498133,
+      "grad_norm": 0.804691731929779,
+      "learning_rate": 4.986933061271776e-06,
+      "loss": 0.6384,
+      "step": 271
+    },
+    {
+      "epoch": 0.30470500373412995,
+      "grad_norm": 0.8142663240432739,
+      "learning_rate": 4.9867799199394205e-06,
+      "loss": 0.6061,
+      "step": 272
+    },
+    {
+      "epoch": 0.3058252427184466,
+      "grad_norm": 0.8177836537361145,
+      "learning_rate": 4.9866258888168e-06,
+      "loss": 0.6186,
+      "step": 273
+    },
+    {
+      "epoch": 0.30694548170276326,
+      "grad_norm": 0.8021143078804016,
+      "learning_rate": 4.986470967959031e-06,
+      "loss": 0.6262,
+      "step": 274
+    },
+    {
+      "epoch": 0.30806572068707994,
+      "grad_norm": 0.7913358807563782,
+      "learning_rate": 4.986315157421543e-06,
+      "loss": 0.6011,
+      "step": 275
+    },
+    {
+      "epoch": 0.30918595967139656,
+      "grad_norm": 0.8460528254508972,
+      "learning_rate": 4.986158457260086e-06,
+      "loss": 0.5983,
+      "step": 276
+    },
+    {
+      "epoch": 0.31030619865571324,
+      "grad_norm": 0.8357203006744385,
+      "learning_rate": 4.986000867530729e-06,
+      "loss": 0.6081,
+      "step": 277
+    },
+    {
+      "epoch": 0.31142643764002986,
+      "grad_norm": 0.8164505362510681,
+      "learning_rate": 4.9858423882898595e-06,
+      "loss": 0.6408,
+      "step": 278
+    },
+    {
+      "epoch": 0.31254667662434654,
+      "grad_norm": 0.8584355115890503,
+      "learning_rate": 4.985683019594183e-06,
+      "loss": 0.6206,
+      "step": 279
+    },
+    {
+      "epoch": 0.31366691560866317,
+      "grad_norm": 0.8428062796592712,
+      "learning_rate": 4.985522761500722e-06,
+      "loss": 0.6622,
+      "step": 280
+    },
+    {
+      "epoch": 0.31478715459297985,
+      "grad_norm": 0.9295979142189026,
+      "learning_rate": 4.985361614066818e-06,
+      "loss": 0.6374,
+      "step": 281
+    },
+    {
+      "epoch": 0.31590739357729647,
+      "grad_norm": 0.7724334597587585,
+      "learning_rate": 4.985199577350132e-06,
+      "loss": 0.6148,
+      "step": 282
+    },
+    {
+      "epoch": 0.31702763256161315,
+      "grad_norm": 0.785025954246521,
+      "learning_rate": 4.9850366514086415e-06,
+      "loss": 0.6299,
+      "step": 283
+    },
+    {
+      "epoch": 0.3181478715459298,
+      "grad_norm": 0.8610864877700806,
+      "learning_rate": 4.984872836300641e-06,
+      "loss": 0.6464,
+      "step": 284
+    },
+    {
+      "epoch": 0.31926811053024645,
+      "grad_norm": 0.8281784653663635,
+      "learning_rate": 4.9847081320847475e-06,
+      "loss": 0.6297,
+      "step": 285
+    },
+    {
+      "epoch": 0.32038834951456313,
+      "grad_norm": 0.8076677322387695,
+      "learning_rate": 4.984542538819893e-06,
+      "loss": 0.6144,
+      "step": 286
+    },
+    {
+      "epoch": 0.32150858849887975,
+      "grad_norm": 1.0354986190795898,
+      "learning_rate": 4.984376056565328e-06,
+      "loss": 0.6246,
+      "step": 287
+    },
+    {
+      "epoch": 0.32262882748319643,
+      "grad_norm": 0.7936146259307861,
+      "learning_rate": 4.98420868538062e-06,
+      "loss": 0.6276,
+      "step": 288
+    },
+    {
+      "epoch": 0.32374906646751306,
+      "grad_norm": 0.8247389793395996,
+      "learning_rate": 4.984040425325657e-06,
+      "loss": 0.6326,
+      "step": 289
+    },
+    {
+      "epoch": 0.32486930545182974,
+      "grad_norm": 0.8053764700889587,
+      "learning_rate": 4.983871276460643e-06,
+      "loss": 0.6298,
+      "step": 290
+    },
+    {
+      "epoch": 0.32598954443614636,
+      "grad_norm": 0.7672686576843262,
+      "learning_rate": 4.983701238846102e-06,
+      "loss": 0.6192,
+      "step": 291
+    },
+    {
+      "epoch": 0.32710978342046304,
+      "grad_norm": 0.8261781930923462,
+      "learning_rate": 4.983530312542874e-06,
+      "loss": 0.6571,
+      "step": 292
+    },
+    {
+      "epoch": 0.32823002240477966,
+      "grad_norm": 0.8620947599411011,
+      "learning_rate": 4.983358497612118e-06,
+      "loss": 0.6195,
+      "step": 293
+    },
+    {
+      "epoch": 0.32935026138909634,
+      "grad_norm": 0.8011974692344666,
+      "learning_rate": 4.983185794115311e-06,
+      "loss": 0.6157,
+      "step": 294
+    },
+    {
+      "epoch": 0.33047050037341297,
+      "grad_norm": 0.8179512619972229,
+      "learning_rate": 4.983012202114247e-06,
+      "loss": 0.6086,
+      "step": 295
+    },
+    {
+      "epoch": 0.33159073935772965,
+      "grad_norm": 0.82745760679245,
+      "learning_rate": 4.98283772167104e-06,
+      "loss": 0.651,
+      "step": 296
+    },
+    {
+      "epoch": 0.3327109783420463,
+      "grad_norm": 0.8494473695755005,
+      "learning_rate": 4.982662352848119e-06,
+      "loss": 0.6114,
+      "step": 297
+    },
+    {
+      "epoch": 0.33383121732636295,
+      "grad_norm": 0.80657559633255,
+      "learning_rate": 4.982486095708234e-06,
+      "loss": 0.6105,
+      "step": 298
+    },
+    {
+      "epoch": 0.33495145631067963,
+      "grad_norm": 0.8035582900047302,
+      "learning_rate": 4.982308950314449e-06,
+      "loss": 0.6373,
+      "step": 299
+    },
+    {
+      "epoch": 0.33607169529499625,
+      "grad_norm": 0.7871240377426147,
+      "learning_rate": 4.982130916730151e-06,
+      "loss": 0.6004,
+      "step": 300
+    },
+    {
+      "epoch": 0.33719193427931293,
+      "grad_norm": 0.8356077075004578,
+      "learning_rate": 4.98195199501904e-06,
+      "loss": 0.6164,
+      "step": 301
+    },
+    {
+      "epoch": 0.33831217326362956,
+      "grad_norm": 0.7815148234367371,
+      "learning_rate": 4.981772185245135e-06,
+      "loss": 0.6173,
+      "step": 302
+    },
+    {
+      "epoch": 0.33943241224794624,
+      "grad_norm": 0.8669472336769104,
+      "learning_rate": 4.981591487472776e-06,
+      "loss": 0.6209,
+      "step": 303
+    },
+    {
+      "epoch": 0.34055265123226286,
+      "grad_norm": 0.7833518385887146,
+      "learning_rate": 4.981409901766615e-06,
+      "loss": 0.6178,
+      "step": 304
+    },
+    {
+      "epoch": 0.34167289021657954,
+      "grad_norm": 0.79278165102005,
+      "learning_rate": 4.9812274281916275e-06,
+      "loss": 0.6268,
+      "step": 305
+    },
+    {
+      "epoch": 0.3427931292008962,
+      "grad_norm": 0.802302360534668,
+      "learning_rate": 4.981044066813102e-06,
+      "loss": 0.5878,
+      "step": 306
+    },
+    {
+      "epoch": 0.34391336818521284,
+      "grad_norm": 0.8093838691711426,
+      "learning_rate": 4.9808598176966485e-06,
+      "loss": 0.6419,
+      "step": 307
+    },
+    {
+      "epoch": 0.3450336071695295,
+      "grad_norm": 0.7981029152870178,
+      "learning_rate": 4.980674680908192e-06,
+      "loss": 0.6203,
+      "step": 308
+    },
+    {
+      "epoch": 0.34615384615384615,
+      "grad_norm": 0.8186250329017639,
+      "learning_rate": 4.980488656513976e-06,
+      "loss": 0.6346,
+      "step": 309
+    },
+    {
+      "epoch": 0.3472740851381628,
+      "grad_norm": 0.7659736275672913,
+      "learning_rate": 4.980301744580561e-06,
+      "loss": 0.5925,
+      "step": 310
+    },
+    {
+      "epoch": 0.34839432412247945,
+      "grad_norm": 0.8284393548965454,
+      "learning_rate": 4.980113945174826e-06,
+      "loss": 0.6471,
+      "step": 311
+    },
+    {
+      "epoch": 0.34951456310679613,
+      "grad_norm": 0.7763761878013611,
+      "learning_rate": 4.979925258363968e-06,
+      "loss": 0.6095,
+      "step": 312
+    },
+    {
+      "epoch": 0.35063480209111275,
+      "grad_norm": 0.820344090461731,
+      "learning_rate": 4.979735684215499e-06,
+      "loss": 0.6044,
+      "step": 313
+    },
+    {
+      "epoch": 0.35175504107542943,
+      "grad_norm": 0.8457436561584473,
+      "learning_rate": 4.979545222797252e-06,
+      "loss": 0.6176,
+      "step": 314
+    },
+    {
+      "epoch": 0.35287528005974605,
+      "grad_norm": 0.8047285079956055,
+      "learning_rate": 4.979353874177375e-06,
+      "loss": 0.5912,
+      "step": 315
+    },
+    {
+      "epoch": 0.35399551904406273,
+      "grad_norm": 0.8075456619262695,
+      "learning_rate": 4.9791616384243335e-06,
+      "loss": 0.5807,
+      "step": 316
+    },
+    {
+      "epoch": 0.3551157580283794,
+      "grad_norm": 0.7939711809158325,
+      "learning_rate": 4.978968515606912e-06,
+      "loss": 0.5862,
+      "step": 317
+    },
+    {
+      "epoch": 0.35623599701269604,
+      "grad_norm": 0.8410739302635193,
+      "learning_rate": 4.97877450579421e-06,
+      "loss": 0.6144,
+      "step": 318
+    },
+    {
+      "epoch": 0.3573562359970127,
+      "grad_norm": 0.8608905076980591,
+      "learning_rate": 4.9785796090556485e-06,
+      "loss": 0.6428,
+      "step": 319
+    },
+    {
+      "epoch": 0.35847647498132934,
+      "grad_norm": 0.8229684233665466,
+      "learning_rate": 4.978383825460959e-06,
+      "loss": 0.6124,
+      "step": 320
+    },
+    {
+      "epoch": 0.359596713965646,
+      "grad_norm": 0.790482223033905,
+      "learning_rate": 4.978187155080199e-06,
+      "loss": 0.6127,
+      "step": 321
+    },
+    {
+      "epoch": 0.36071695294996264,
+      "grad_norm": 0.7994086146354675,
+      "learning_rate": 4.9779895979837355e-06,
+      "loss": 0.6337,
+      "step": 322
+    },
+    {
+      "epoch": 0.3618371919342793,
+      "grad_norm": 0.8186279535293579,
+      "learning_rate": 4.977791154242257e-06,
+      "loss": 0.6257,
+      "step": 323
+    },
+    {
+      "epoch": 0.36295743091859595,
+      "grad_norm": 0.7845602631568909,
+      "learning_rate": 4.977591823926769e-06,
+      "loss": 0.5702,
+      "step": 324
+    },
+    {
+      "epoch": 0.3640776699029126,
+      "grad_norm": 0.7722570300102234,
+      "learning_rate": 4.977391607108593e-06,
+      "loss": 0.5897,
+      "step": 325
+    },
+    {
+      "epoch": 0.36519790888722925,
+      "grad_norm": 0.8029158115386963,
+      "learning_rate": 4.977190503859368e-06,
+      "loss": 0.6142,
+      "step": 326
+    },
+    {
+      "epoch": 0.36631814787154593,
+      "grad_norm": 0.801633894443512,
+      "learning_rate": 4.976988514251051e-06,
+      "loss": 0.6298,
+      "step": 327
+    },
+    {
+      "epoch": 0.3674383868558626,
+      "grad_norm": 0.8345485925674438,
+      "learning_rate": 4.976785638355914e-06,
+      "loss": 0.6351,
+      "step": 328
+    },
+    {
+      "epoch": 0.36855862584017923,
+      "grad_norm": 0.793038547039032,
+      "learning_rate": 4.97658187624655e-06,
+      "loss": 0.6463,
+      "step": 329
+    },
+    {
+      "epoch": 0.3696788648244959,
+      "grad_norm": 0.8270999789237976,
+      "learning_rate": 4.976377227995866e-06,
+      "loss": 0.6193,
+      "step": 330
+    },
+    {
+      "epoch": 0.37079910380881254,
+      "grad_norm": 0.8013745546340942,
+      "learning_rate": 4.9761716936770855e-06,
+      "loss": 0.6047,
+      "step": 331
+    },
+    {
+      "epoch": 0.3719193427931292,
+      "grad_norm": 0.8266115784645081,
+      "learning_rate": 4.975965273363751e-06,
+      "loss": 0.6226,
+      "step": 332
+    },
+    {
+      "epoch": 0.37303958177744584,
+      "grad_norm": 0.8072906732559204,
+      "learning_rate": 4.975757967129722e-06,
+      "loss": 0.615,
+      "step": 333
+    },
+    {
+      "epoch": 0.3741598207617625,
+      "grad_norm": 0.7997392416000366,
+      "learning_rate": 4.975549775049174e-06,
+      "loss": 0.6271,
+      "step": 334
+    },
+    {
+      "epoch": 0.37528005974607914,
+      "grad_norm": 0.8856087923049927,
+      "learning_rate": 4.9753406971966e-06,
+      "loss": 0.6066,
+      "step": 335
+    },
+    {
+      "epoch": 0.3764002987303958,
+      "grad_norm": 0.7950431704521179,
+      "learning_rate": 4.975130733646809e-06,
+      "loss": 0.6001,
+      "step": 336
+    },
+    {
+      "epoch": 0.37752053771471245,
+      "grad_norm": 0.8051393032073975,
+      "learning_rate": 4.974919884474929e-06,
+      "loss": 0.6316,
+      "step": 337
+    },
+    {
+      "epoch": 0.3786407766990291,
+      "grad_norm": 0.8180161714553833,
+      "learning_rate": 4.974708149756403e-06,
+      "loss": 0.6362,
+      "step": 338
+    },
+    {
+      "epoch": 0.3797610156833458,
+      "grad_norm": 0.8257321119308472,
+      "learning_rate": 4.974495529566991e-06,
+      "loss": 0.626,
+      "step": 339
+    },
+    {
+      "epoch": 0.38088125466766243,
+      "grad_norm": 0.8660753965377808,
+      "learning_rate": 4.9742820239827705e-06,
+      "loss": 0.6191,
+      "step": 340
+    },
+    {
+      "epoch": 0.3820014936519791,
+      "grad_norm": 0.7817429900169373,
+      "learning_rate": 4.974067633080135e-06,
+      "loss": 0.6297,
+      "step": 341
+    },
+    {
+      "epoch": 0.38312173263629573,
+      "grad_norm": 0.807461678981781,
+      "learning_rate": 4.973852356935797e-06,
+      "loss": 0.6356,
+      "step": 342
+    },
+    {
+      "epoch": 0.3842419716206124,
+      "grad_norm": 0.8301308155059814,
+      "learning_rate": 4.973636195626783e-06,
+      "loss": 0.6205,
+      "step": 343
+    },
+    {
+      "epoch": 0.38536221060492903,
+      "grad_norm": 0.897199809551239,
+      "learning_rate": 4.973419149230437e-06,
+      "loss": 0.6455,
+      "step": 344
+    },
+    {
+      "epoch": 0.3864824495892457,
+      "grad_norm": 0.8273870944976807,
+      "learning_rate": 4.9732012178244194e-06,
+      "loss": 0.6458,
+      "step": 345
+    },
+    {
+      "epoch": 0.38760268857356234,
+      "grad_norm": 0.9393578767776489,
+      "learning_rate": 4.972982401486709e-06,
+      "loss": 0.61,
+      "step": 346
+    },
+    {
+      "epoch": 0.388722927557879,
+      "grad_norm": 0.8364170789718628,
+      "learning_rate": 4.9727627002956e-06,
+      "loss": 0.6195,
+      "step": 347
+    },
+    {
+      "epoch": 0.3898431665421957,
+      "grad_norm": 0.8208838701248169,
+      "learning_rate": 4.972542114329704e-06,
+      "loss": 0.6246,
+      "step": 348
+    },
+    {
+      "epoch": 0.3909634055265123,
+      "grad_norm": 0.8595433831214905,
+      "learning_rate": 4.972320643667946e-06,
+      "loss": 0.6251,
+      "step": 349
+    },
+    {
+      "epoch": 0.392083644510829,
+      "grad_norm": 0.7956632375717163,
+      "learning_rate": 4.972098288389572e-06,
+      "loss": 0.608,
+      "step": 350
+    },
+    {
+      "epoch": 0.3932038834951456,
+      "grad_norm": 0.8434047698974609,
+      "learning_rate": 4.9718750485741425e-06,
+      "loss": 0.6137,
+      "step": 351
+    },
+    {
+      "epoch": 0.3943241224794623,
+      "grad_norm": 0.8272572755813599,
+      "learning_rate": 4.971650924301534e-06,
+      "loss": 0.623,
+      "step": 352
+    },
+    {
+      "epoch": 0.3954443614637789,
+      "grad_norm": 0.8429244756698608,
+      "learning_rate": 4.9714259156519406e-06,
+      "loss": 0.6325,
+      "step": 353
+    },
+    {
+      "epoch": 0.3965646004480956,
+      "grad_norm": 0.8181564807891846,
+      "learning_rate": 4.971200022705871e-06,
+      "loss": 0.6421,
+      "step": 354
+    },
+    {
+      "epoch": 0.39768483943241223,
+      "grad_norm": 0.7761826515197754,
+      "learning_rate": 4.970973245544153e-06,
+      "loss": 0.6063,
+      "step": 355
+    },
+    {
+      "epoch": 0.3988050784167289,
+      "grad_norm": 0.8283432126045227,
+      "learning_rate": 4.970745584247928e-06,
+      "loss": 0.6125,
+      "step": 356
+    },
+    {
+      "epoch": 0.39992531740104553,
+      "grad_norm": 0.8041115999221802,
+      "learning_rate": 4.9705170388986565e-06,
+      "loss": 0.6227,
+      "step": 357
+    },
+    {
+      "epoch": 0.4010455563853622,
+      "grad_norm": 0.792184054851532,
+      "learning_rate": 4.970287609578112e-06,
+      "loss": 0.624,
+      "step": 358
+    },
+    {
+      "epoch": 0.4021657953696789,
+      "grad_norm": 0.8069051504135132,
+      "learning_rate": 4.970057296368388e-06,
+      "loss": 0.6162,
+      "step": 359
+    },
+    {
+      "epoch": 0.4032860343539955,
+      "grad_norm": 0.7888614535331726,
+      "learning_rate": 4.969826099351892e-06,
+      "loss": 0.6088,
+      "step": 360
+    },
+    {
+      "epoch": 0.4044062733383122,
+      "grad_norm": 0.8092378377914429,
+      "learning_rate": 4.969594018611348e-06,
+      "loss": 0.6062,
+      "step": 361
+    },
+    {
+      "epoch": 0.4055265123226288,
+      "grad_norm": 0.8325386047363281,
+      "learning_rate": 4.969361054229796e-06,
+      "loss": 0.6277,
+      "step": 362
+    },
+    {
+      "epoch": 0.4066467513069455,
+      "grad_norm": 0.860455334186554,
+      "learning_rate": 4.9691272062905935e-06,
+      "loss": 0.6202,
+      "step": 363
+    },
+    {
+      "epoch": 0.4077669902912621,
+      "grad_norm": 0.8035297989845276,
+      "learning_rate": 4.968892474877412e-06,
+      "loss": 0.6087,
+      "step": 364
+    },
+    {
+      "epoch": 0.4088872292755788,
+      "grad_norm": 0.792544424533844,
+      "learning_rate": 4.968656860074241e-06,
+      "loss": 0.5919,
+      "step": 365
+    },
+    {
+      "epoch": 0.4100074682598954,
+      "grad_norm": 0.8256428837776184,
+      "learning_rate": 4.968420361965386e-06,
+      "loss": 0.5922,
+      "step": 366
+    },
+    {
+      "epoch": 0.4111277072442121,
+      "grad_norm": 0.8024367094039917,
+      "learning_rate": 4.968182980635468e-06,
+      "loss": 0.6332,
+      "step": 367
+    },
+    {
+      "epoch": 0.4122479462285287,
+      "grad_norm": 0.7686941027641296,
+      "learning_rate": 4.967944716169421e-06,
+      "loss": 0.5954,
+      "step": 368
+    },
+    {
+      "epoch": 0.4133681852128454,
+      "grad_norm": 0.814935028553009,
+      "learning_rate": 4.967705568652502e-06,
+      "loss": 0.627,
+      "step": 369
+    },
+    {
+      "epoch": 0.4144884241971621,
+      "grad_norm": 0.7745496034622192,
+      "learning_rate": 4.967465538170279e-06,
+      "loss": 0.6051,
+      "step": 370
+    },
+    {
+      "epoch": 0.4156086631814787,
+      "grad_norm": 0.8013162612915039,
+      "learning_rate": 4.9672246248086344e-06,
+      "loss": 0.6005,
+      "step": 371
+    },
+    {
+      "epoch": 0.4167289021657954,
+      "grad_norm": 0.8392379879951477,
+      "learning_rate": 4.966982828653772e-06,
+      "loss": 0.624,
+      "step": 372
+    },
+    {
+      "epoch": 0.417849141150112,
+      "grad_norm": 0.781700849533081,
+      "learning_rate": 4.966740149792206e-06,
+      "loss": 0.6064,
+      "step": 373
+    },
+    {
+      "epoch": 0.4189693801344287,
+      "grad_norm": 0.8092135190963745,
+      "learning_rate": 4.966496588310771e-06,
+      "loss": 0.6355,
+      "step": 374
+    },
+    {
+      "epoch": 0.4200896191187453,
+      "grad_norm": 0.7815409898757935,
+      "learning_rate": 4.966252144296614e-06,
+      "loss": 0.6039,
+      "step": 375
+    },
+    {
+      "epoch": 0.421209858103062,
+      "grad_norm": 0.7987800240516663,
+      "learning_rate": 4.9660068178372e-06,
+      "loss": 0.6341,
+      "step": 376
+    },
+    {
+      "epoch": 0.4223300970873786,
+      "grad_norm": 0.8279643058776855,
+      "learning_rate": 4.965760609020308e-06,
+      "loss": 0.6119,
+      "step": 377
+    },
+    {
+      "epoch": 0.4234503360716953,
+      "grad_norm": 0.79878830909729,
+      "learning_rate": 4.965513517934033e-06,
+      "loss": 0.6334,
+      "step": 378
+    },
+    {
+      "epoch": 0.4245705750560119,
+      "grad_norm": 0.8361204862594604,
+      "learning_rate": 4.965265544666788e-06,
+      "loss": 0.6078,
+      "step": 379
+    },
+    {
+      "epoch": 0.4256908140403286,
+      "grad_norm": 0.8225858211517334,
+      "learning_rate": 4.965016689307298e-06,
+      "loss": 0.6264,
+      "step": 380
+    },
+    {
+      "epoch": 0.4268110530246453,
+      "grad_norm": 0.8204994201660156,
+      "learning_rate": 4.964766951944607e-06,
+      "loss": 0.606,
+      "step": 381
+    },
+    {
+      "epoch": 0.4279312920089619,
+      "grad_norm": 0.8047780990600586,
+      "learning_rate": 4.964516332668073e-06,
+      "loss": 0.6053,
+      "step": 382
+    },
+    {
+      "epoch": 0.4290515309932786,
+      "grad_norm": 0.8498148322105408,
+      "learning_rate": 4.964264831567368e-06,
+      "loss": 0.6211,
+      "step": 383
+    },
+    {
+      "epoch": 0.4301717699775952,
+      "grad_norm": 0.8710355758666992,
+      "learning_rate": 4.964012448732482e-06,
+      "loss": 0.6134,
+      "step": 384
+    },
+    {
+      "epoch": 0.4312920089619119,
+      "grad_norm": 0.8357451558113098,
+      "learning_rate": 4.963759184253721e-06,
+      "loss": 0.6339,
+      "step": 385
+    },
+    {
+      "epoch": 0.4324122479462285,
+      "grad_norm": 0.8333609700202942,
+      "learning_rate": 4.963505038221704e-06,
+      "loss": 0.623,
+      "step": 386
+    },
+    {
+      "epoch": 0.4335324869305452,
+      "grad_norm": 0.8006998300552368,
+      "learning_rate": 4.963250010727366e-06,
+      "loss": 0.621,
+      "step": 387
+    },
+    {
+      "epoch": 0.4346527259148618,
+      "grad_norm": 0.8830387592315674,
+      "learning_rate": 4.962994101861959e-06,
+      "loss": 0.5945,
+      "step": 388
+    },
+    {
+      "epoch": 0.4357729648991785,
+      "grad_norm": 0.8361951112747192,
+      "learning_rate": 4.9627373117170476e-06,
+      "loss": 0.6042,
+      "step": 389
+    },
+    {
+      "epoch": 0.4368932038834951,
+      "grad_norm": 0.8356959819793701,
+      "learning_rate": 4.962479640384516e-06,
+      "loss": 0.6423,
+      "step": 390
+    },
+    {
+      "epoch": 0.4380134428678118,
+      "grad_norm": 0.8058585524559021,
+      "learning_rate": 4.962221087956558e-06,
+      "loss": 0.622,
+      "step": 391
+    },
+    {
+      "epoch": 0.4391336818521285,
+      "grad_norm": 0.8438712358474731,
+      "learning_rate": 4.961961654525689e-06,
+      "loss": 0.619,
+      "step": 392
+    },
+    {
+      "epoch": 0.4402539208364451,
+      "grad_norm": 0.8207183480262756,
+      "learning_rate": 4.961701340184734e-06,
+      "loss": 0.6095,
+      "step": 393
+    },
+    {
+      "epoch": 0.4413741598207618,
+      "grad_norm": 0.8222430348396301,
+      "learning_rate": 4.9614401450268364e-06,
+      "loss": 0.6266,
+      "step": 394
+    },
+    {
+      "epoch": 0.4424943988050784,
+      "grad_norm": 0.8192475438117981,
+      "learning_rate": 4.961178069145455e-06,
+      "loss": 0.6146,
+      "step": 395
+    },
+    {
+      "epoch": 0.4436146377893951,
+      "grad_norm": 0.7848315238952637,
+      "learning_rate": 4.960915112634362e-06,
+      "loss": 0.6046,
+      "step": 396
+    },
+    {
+      "epoch": 0.4447348767737117,
+      "grad_norm": 0.8019006848335266,
+      "learning_rate": 4.960651275587644e-06,
+      "loss": 0.6253,
+      "step": 397
+    },
+    {
+      "epoch": 0.4458551157580284,
+      "grad_norm": 0.8097095489501953,
+      "learning_rate": 4.960386558099706e-06,
+      "loss": 0.6284,
+      "step": 398
+    },
+    {
+      "epoch": 0.446975354742345,
+      "grad_norm": 0.796343207359314,
+      "learning_rate": 4.9601209602652644e-06,
+      "loss": 0.6051,
+      "step": 399
+    },
+    {
+      "epoch": 0.4480955937266617,
+      "grad_norm": 0.8076500296592712,
+      "learning_rate": 4.959854482179354e-06,
+      "loss": 0.6215,
+      "step": 400
+    },
+    {
+      "epoch": 0.44921583271097837,
+      "grad_norm": 0.8721750974655151,
+      "learning_rate": 4.9595871239373215e-06,
+      "loss": 0.6534,
+      "step": 401
+    },
+    {
+      "epoch": 0.450336071695295,
+      "grad_norm": 0.8260684013366699,
+      "learning_rate": 4.959318885634831e-06,
+      "loss": 0.6073,
+      "step": 402
+    },
+    {
+      "epoch": 0.45145631067961167,
+      "grad_norm": 0.8113908767700195,
+      "learning_rate": 4.95904976736786e-06,
+      "loss": 0.6004,
+      "step": 403
+    },
+    {
+      "epoch": 0.4525765496639283,
+      "grad_norm": 0.8291077017784119,
+      "learning_rate": 4.9587797692327e-06,
+      "loss": 0.622,
+      "step": 404
+    },
+    {
+      "epoch": 0.453696788648245,
+      "grad_norm": 0.8331876993179321,
+      "learning_rate": 4.95850889132596e-06,
+      "loss": 0.6367,
+      "step": 405
+    },
+    {
+      "epoch": 0.4548170276325616,
+      "grad_norm": 0.8167732357978821,
+      "learning_rate": 4.9582371337445624e-06,
+      "loss": 0.6191,
+      "step": 406
+    },
+    {
+      "epoch": 0.4559372666168783,
+      "grad_norm": 0.8163003921508789,
+      "learning_rate": 4.957964496585743e-06,
+      "loss": 0.6197,
+      "step": 407
+    },
+    {
+      "epoch": 0.4570575056011949,
+      "grad_norm": 0.8027840852737427,
+      "learning_rate": 4.957690979947055e-06,
+      "loss": 0.6119,
+      "step": 408
+    },
+    {
+      "epoch": 0.4581777445855116,
+      "grad_norm": 0.8352331519126892,
+      "learning_rate": 4.957416583926364e-06,
+      "loss": 0.6176,
+      "step": 409
+    },
+    {
+      "epoch": 0.4592979835698282,
+      "grad_norm": 0.846203625202179,
+      "learning_rate": 4.957141308621851e-06,
+      "loss": 0.608,
+      "step": 410
+    },
+    {
+      "epoch": 0.4604182225541449,
+      "grad_norm": 0.8850703239440918,
+      "learning_rate": 4.956865154132013e-06,
+      "loss": 0.6359,
+      "step": 411
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.7965673804283142,
+      "learning_rate": 4.95658812055566e-06,
+      "loss": 0.5849,
+      "step": 412
+    },
+    {
+      "epoch": 0.4626587005227782,
+      "grad_norm": 0.862568199634552,
+      "learning_rate": 4.956310207991916e-06,
+      "loss": 0.6176,
+      "step": 413
+    },
+    {
+      "epoch": 0.46377893950709487,
+      "grad_norm": 0.8031529784202576,
+      "learning_rate": 4.95603141654022e-06,
+      "loss": 0.5987,
+      "step": 414
+    },
+    {
+      "epoch": 0.4648991784914115,
+      "grad_norm": 0.8270638585090637,
+      "learning_rate": 4.9557517463003275e-06,
+      "loss": 0.6611,
+      "step": 415
+    },
+    {
+      "epoch": 0.46601941747572817,
+      "grad_norm": 0.7914713025093079,
+      "learning_rate": 4.9554711973723055e-06,
+      "loss": 0.5944,
+      "step": 416
+    },
+    {
+      "epoch": 0.4671396564600448,
+      "grad_norm": 0.8575369119644165,
+      "learning_rate": 4.9551897698565376e-06,
+      "loss": 0.6282,
+      "step": 417
+    },
+    {
+      "epoch": 0.4682598954443615,
+      "grad_norm": 0.766158401966095,
+      "learning_rate": 4.9549074638537205e-06,
+      "loss": 0.5798,
+      "step": 418
+    },
+    {
+      "epoch": 0.4693801344286781,
+      "grad_norm": 0.7910577654838562,
+      "learning_rate": 4.954624279464866e-06,
+      "loss": 0.5881,
+      "step": 419
+    },
+    {
+      "epoch": 0.4705003734129948,
+      "grad_norm": 0.8028966188430786,
+      "learning_rate": 4.9543402167912985e-06,
+      "loss": 0.6126,
+      "step": 420
+    },
+    {
+      "epoch": 0.4716206123973114,
+      "grad_norm": 0.8077728748321533,
+      "learning_rate": 4.9540552759346596e-06,
+      "loss": 0.6072,
+      "step": 421
+    },
+    {
+      "epoch": 0.4727408513816281,
+      "grad_norm": 0.8463570475578308,
+      "learning_rate": 4.953769456996902e-06,
+      "loss": 0.5721,
+      "step": 422
+    },
+    {
+      "epoch": 0.47386109036594476,
+      "grad_norm": 0.7810856699943542,
+      "learning_rate": 4.9534827600802956e-06,
+      "loss": 0.5852,
+      "step": 423
+    },
+    {
+      "epoch": 0.4749813293502614,
+      "grad_norm": 0.7997978329658508,
+      "learning_rate": 4.953195185287421e-06,
+      "loss": 0.5766,
+      "step": 424
+    },
+    {
+      "epoch": 0.47610156833457806,
+      "grad_norm": 0.7943705320358276,
+      "learning_rate": 4.952906732721177e-06,
+      "loss": 0.5915,
+      "step": 425
+    },
+    {
+      "epoch": 0.4772218073188947,
+      "grad_norm": 0.7930023670196533,
+      "learning_rate": 4.952617402484773e-06,
+      "loss": 0.6025,
+      "step": 426
+    },
+    {
+      "epoch": 0.47834204630321137,
+      "grad_norm": 0.7801458835601807,
+      "learning_rate": 4.952327194681734e-06,
+      "loss": 0.5947,
+      "step": 427
+    },
+    {
+      "epoch": 0.479462285287528,
+      "grad_norm": 0.8124021291732788,
+      "learning_rate": 4.952036109415899e-06,
+      "loss": 0.6175,
+      "step": 428
+    },
+    {
+      "epoch": 0.48058252427184467,
+      "grad_norm": 0.8290462493896484,
+      "learning_rate": 4.9517441467914195e-06,
+      "loss": 0.6229,
+      "step": 429
+    },
+    {
+      "epoch": 0.4817027632561613,
+      "grad_norm": 0.7960165739059448,
+      "learning_rate": 4.951451306912764e-06,
+      "loss": 0.6078,
+      "step": 430
+    },
+    {
+      "epoch": 0.48282300224047797,
+      "grad_norm": 0.8182939887046814,
+      "learning_rate": 4.951157589884711e-06,
+      "loss": 0.6145,
+      "step": 431
+    },
+    {
+      "epoch": 0.4839432412247946,
+      "grad_norm": 0.7847172021865845,
+      "learning_rate": 4.950862995812356e-06,
+      "loss": 0.592,
+      "step": 432
+    },
+    {
+      "epoch": 0.4850634802091113,
+      "grad_norm": 0.8396735191345215,
+      "learning_rate": 4.9505675248011085e-06,
+      "loss": 0.6323,
+      "step": 433
+    },
+    {
+      "epoch": 0.48618371919342795,
+      "grad_norm": 0.8656349778175354,
+      "learning_rate": 4.950271176956688e-06,
+      "loss": 0.6237,
+      "step": 434
+    },
+    {
+      "epoch": 0.4873039581777446,
+      "grad_norm": 0.8108944296836853,
+      "learning_rate": 4.949973952385131e-06,
+      "loss": 0.6148,
+      "step": 435
+    },
+    {
+      "epoch": 0.48842419716206126,
+      "grad_norm": 0.839131236076355,
+      "learning_rate": 4.949675851192786e-06,
+      "loss": 0.6475,
+      "step": 436
+    },
+    {
+      "epoch": 0.4895444361463779,
+      "grad_norm": 0.8033477067947388,
+      "learning_rate": 4.949376873486319e-06,
+      "loss": 0.6095,
+      "step": 437
+    },
+    {
+      "epoch": 0.49066467513069456,
+      "grad_norm": 0.8268027901649475,
+      "learning_rate": 4.949077019372704e-06,
+      "loss": 0.6096,
+      "step": 438
+    },
+    {
+      "epoch": 0.4917849141150112,
+      "grad_norm": 0.8278724551200867,
+      "learning_rate": 4.948776288959233e-06,
+      "loss": 0.6225,
+      "step": 439
+    },
+    {
+      "epoch": 0.49290515309932786,
+      "grad_norm": 0.8136724233627319,
+      "learning_rate": 4.948474682353508e-06,
+      "loss": 0.6063,
+      "step": 440
+    },
+    {
+      "epoch": 0.4940253920836445,
+      "grad_norm": 0.821102499961853,
+      "learning_rate": 4.948172199663448e-06,
+      "loss": 0.6214,
+      "step": 441
+    },
+    {
+      "epoch": 0.49514563106796117,
+      "grad_norm": 0.8143325448036194,
+      "learning_rate": 4.947868840997282e-06,
+      "loss": 0.6021,
+      "step": 442
+    },
+    {
+      "epoch": 0.49626587005227785,
+      "grad_norm": 0.8387938141822815,
+      "learning_rate": 4.947564606463556e-06,
+      "loss": 0.6299,
+      "step": 443
+    },
+    {
+      "epoch": 0.49738610903659447,
+      "grad_norm": 0.8075438737869263,
+      "learning_rate": 4.9472594961711266e-06,
+      "loss": 0.603,
+      "step": 444
+    },
+    {
+      "epoch": 0.49850634802091115,
+      "grad_norm": 0.8184303641319275,
+      "learning_rate": 4.946953510229164e-06,
+      "loss": 0.6126,
+      "step": 445
+    },
+    {
+      "epoch": 0.4996265870052278,
+      "grad_norm": 0.8168706893920898,
+      "learning_rate": 4.946646648747155e-06,
+      "loss": 0.5987,
+      "step": 446
+    },
+    {
+      "epoch": 0.5007468259895445,
+      "grad_norm": 0.8080923557281494,
+      "learning_rate": 4.946338911834895e-06,
+      "loss": 0.5998,
+      "step": 447
+    },
+    {
+      "epoch": 0.5018670649738611,
+      "grad_norm": 0.8014161586761475,
+      "learning_rate": 4.946030299602496e-06,
+      "loss": 0.6279,
+      "step": 448
+    },
+    {
+      "epoch": 0.5029873039581777,
+      "grad_norm": 0.7952543497085571,
+      "learning_rate": 4.94572081216038e-06,
+      "loss": 0.5883,
+      "step": 449
+    },
+    {
+      "epoch": 0.5041075429424944,
+      "grad_norm": 0.8177149295806885,
+      "learning_rate": 4.945410449619286e-06,
+      "loss": 0.5988,
+      "step": 450
+    },
+    {
+      "epoch": 0.5052277819268111,
+      "grad_norm": 0.8595210909843445,
+      "learning_rate": 4.945099212090265e-06,
+      "loss": 0.6387,
+      "step": 451
+    },
+    {
+      "epoch": 0.5063480209111277,
+      "grad_norm": 0.8385729193687439,
+      "learning_rate": 4.944787099684678e-06,
+      "loss": 0.6276,
+      "step": 452
+    },
+    {
+      "epoch": 0.5074682598954443,
+      "grad_norm": 0.7771918177604675,
+      "learning_rate": 4.944474112514204e-06,
+      "loss": 0.5967,
+      "step": 453
+    },
+    {
+      "epoch": 0.508588498879761,
+      "grad_norm": 0.7973781228065491,
+      "learning_rate": 4.9441602506908295e-06,
+      "loss": 0.6055,
+      "step": 454
+    },
+    {
+      "epoch": 0.5097087378640777,
+      "grad_norm": 0.8014892935752869,
+      "learning_rate": 4.9438455143268596e-06,
+      "loss": 0.6018,
+      "step": 455
+    },
+    {
+      "epoch": 0.5108289768483943,
+      "grad_norm": 0.8037407994270325,
+      "learning_rate": 4.943529903534907e-06,
+      "loss": 0.6117,
+      "step": 456
+    },
+    {
+      "epoch": 0.511949215832711,
+      "grad_norm": 0.8133897185325623,
+      "learning_rate": 4.943213418427901e-06,
+      "loss": 0.5902,
+      "step": 457
+    },
+    {
+      "epoch": 0.5130694548170276,
+      "grad_norm": 0.81773841381073,
+      "learning_rate": 4.9428960591190834e-06,
+      "loss": 0.6017,
+      "step": 458
+    },
+    {
+      "epoch": 0.5141896938013443,
+      "grad_norm": 0.8633530735969543,
+      "learning_rate": 4.942577825722007e-06,
+      "loss": 0.597,
+      "step": 459
+    },
+    {
+      "epoch": 0.515309932785661,
+      "grad_norm": 0.7943804264068604,
+      "learning_rate": 4.942258718350539e-06,
+      "loss": 0.605,
+      "step": 460
+    },
+    {
+      "epoch": 0.5164301717699776,
+      "grad_norm": 0.8280025124549866,
+      "learning_rate": 4.941938737118857e-06,
+      "loss": 0.5855,
+      "step": 461
+    },
+    {
+      "epoch": 0.5175504107542942,
+      "grad_norm": 0.7869625091552734,
+      "learning_rate": 4.941617882141455e-06,
+      "loss": 0.6035,
+      "step": 462
+    },
+    {
+      "epoch": 0.5186706497386109,
+      "grad_norm": 0.8214724659919739,
+      "learning_rate": 4.9412961535331365e-06,
+      "loss": 0.6402,
+      "step": 463
+    },
+    {
+      "epoch": 0.5197908887229276,
+      "grad_norm": 0.8125516772270203,
+      "learning_rate": 4.940973551409018e-06,
+      "loss": 0.6135,
+      "step": 464
+    },
+    {
+      "epoch": 0.5209111277072442,
+      "grad_norm": 0.8744713068008423,
+      "learning_rate": 4.9406500758845305e-06,
+      "loss": 0.5904,
+      "step": 465
+    },
+    {
+      "epoch": 0.5220313666915609,
+      "grad_norm": 0.8251977562904358,
+      "learning_rate": 4.940325727075415e-06,
+      "loss": 0.6185,
+      "step": 466
+    },
+    {
+      "epoch": 0.5231516056758775,
+      "grad_norm": 0.7986443638801575,
+      "learning_rate": 4.940000505097728e-06,
+      "loss": 0.6177,
+      "step": 467
+    },
+    {
+      "epoch": 0.5242718446601942,
+      "grad_norm": 0.8290914297103882,
+      "learning_rate": 4.939674410067834e-06,
+      "loss": 0.6531,
+      "step": 468
+    },
+    {
+      "epoch": 0.5253920836445108,
+      "grad_norm": 0.7911964058876038,
+      "learning_rate": 4.939347442102415e-06,
+      "loss": 0.6278,
+      "step": 469
+    },
+    {
+      "epoch": 0.5265123226288275,
+      "grad_norm": 0.7842428088188171,
+      "learning_rate": 4.9390196013184614e-06,
+      "loss": 0.6014,
+      "step": 470
+    },
+    {
+      "epoch": 0.5276325616131441,
+      "grad_norm": 0.8033813834190369,
+      "learning_rate": 4.9386908878332775e-06,
+      "loss": 0.5975,
+      "step": 471
+    },
+    {
+      "epoch": 0.5287528005974608,
+      "grad_norm": 0.7822708487510681,
+      "learning_rate": 4.9383613017644796e-06,
+      "loss": 0.5812,
+      "step": 472
+    },
+    {
+      "epoch": 0.5298730395817774,
+      "grad_norm": 0.8360676169395447,
+      "learning_rate": 4.9380308432299976e-06,
+      "loss": 0.6112,
+      "step": 473
+    },
+    {
+      "epoch": 0.5309932785660941,
+      "grad_norm": 0.8052594065666199,
+      "learning_rate": 4.93769951234807e-06,
+      "loss": 0.6044,
+      "step": 474
+    },
+    {
+      "epoch": 0.5321135175504108,
+      "grad_norm": 0.7881698608398438,
+      "learning_rate": 4.937367309237251e-06,
+      "loss": 0.5778,
+      "step": 475
+    },
+    {
+      "epoch": 0.5332337565347274,
+      "grad_norm": 0.8374946117401123,
+      "learning_rate": 4.937034234016406e-06,
+      "loss": 0.6213,
+      "step": 476
+    },
+    {
+      "epoch": 0.5343539955190441,
+      "grad_norm": 0.8044434189796448,
+      "learning_rate": 4.936700286804711e-06,
+      "loss": 0.5984,
+      "step": 477
+    },
+    {
+      "epoch": 0.5354742345033607,
+      "grad_norm": 0.7946317195892334,
+      "learning_rate": 4.936365467721655e-06,
+      "loss": 0.6027,
+      "step": 478
+    },
+    {
+      "epoch": 0.5365944734876774,
+      "grad_norm": 0.8196859955787659,
+      "learning_rate": 4.93602977688704e-06,
+      "loss": 0.6009,
+      "step": 479
+    },
+    {
+      "epoch": 0.537714712471994,
+      "grad_norm": 0.8330888748168945,
+      "learning_rate": 4.935693214420979e-06,
+      "loss": 0.6396,
+      "step": 480
+    },
+    {
+      "epoch": 0.5388349514563107,
+      "grad_norm": 0.7970585823059082,
+      "learning_rate": 4.935355780443896e-06,
+      "loss": 0.6134,
+      "step": 481
+    },
+    {
+      "epoch": 0.5399551904406273,
+      "grad_norm": 0.8149996399879456,
+      "learning_rate": 4.935017475076529e-06,
+      "loss": 0.6214,
+      "step": 482
+    },
+    {
+      "epoch": 0.541075429424944,
+      "grad_norm": 0.7842663526535034,
+      "learning_rate": 4.9346782984399246e-06,
+      "loss": 0.6084,
+      "step": 483
+    },
+    {
+      "epoch": 0.5421956684092606,
+      "grad_norm": 0.8166002035140991,
+      "learning_rate": 4.934338250655444e-06,
+      "loss": 0.5939,
+      "step": 484
+    },
+    {
+      "epoch": 0.5433159073935773,
+      "grad_norm": 0.8198996186256409,
+      "learning_rate": 4.933997331844758e-06,
+      "loss": 0.5925,
+      "step": 485
+    },
+    {
+      "epoch": 0.544436146377894,
+      "grad_norm": 0.8149311542510986,
+      "learning_rate": 4.933655542129852e-06,
+      "loss": 0.618,
+      "step": 486
+    },
+    {
+      "epoch": 0.5455563853622106,
+      "grad_norm": 0.8132028579711914,
+      "learning_rate": 4.93331288163302e-06,
+      "loss": 0.5923,
+      "step": 487
+    },
+    {
+      "epoch": 0.5466766243465273,
+      "grad_norm": 0.8155807852745056,
+      "learning_rate": 4.9329693504768686e-06,
+      "loss": 0.5966,
+      "step": 488
+    },
+    {
+      "epoch": 0.5477968633308439,
+      "grad_norm": 0.8111668229103088,
+      "learning_rate": 4.932624948784317e-06,
+      "loss": 0.613,
+      "step": 489
+    },
+    {
+      "epoch": 0.5489171023151606,
+      "grad_norm": 0.8030702471733093,
+      "learning_rate": 4.932279676678594e-06,
+      "loss": 0.6077,
+      "step": 490
+    },
+    {
+      "epoch": 0.5500373412994772,
+      "grad_norm": 0.8306952118873596,
+      "learning_rate": 4.931933534283241e-06,
+      "loss": 0.6182,
+      "step": 491
+    },
+    {
+      "epoch": 0.5511575802837939,
+      "grad_norm": 0.8086608052253723,
+      "learning_rate": 4.931586521722112e-06,
+      "loss": 0.5989,
+      "step": 492
+    },
+    {
+      "epoch": 0.5522778192681105,
+      "grad_norm": 0.8051784634590149,
+      "learning_rate": 4.931238639119368e-06,
+      "loss": 0.595,
+      "step": 493
+    },
+    {
+      "epoch": 0.5533980582524272,
+      "grad_norm": 0.7761525511741638,
+      "learning_rate": 4.930889886599486e-06,
+      "loss": 0.5773,
+      "step": 494
+    },
+    {
+      "epoch": 0.5545182972367438,
+      "grad_norm": 0.7963784337043762,
+      "learning_rate": 4.9305402642872534e-06,
+      "loss": 0.578,
+      "step": 495
+    },
+    {
+      "epoch": 0.5556385362210605,
+      "grad_norm": 0.7898165583610535,
+      "learning_rate": 4.9301897723077664e-06,
+      "loss": 0.5954,
+      "step": 496
+    },
+    {
+      "epoch": 0.5567587752053772,
+      "grad_norm": 0.7823036909103394,
+      "learning_rate": 4.929838410786435e-06,
+      "loss": 0.5871,
+      "step": 497
+    },
+    {
+      "epoch": 0.5578790141896938,
+      "grad_norm": 0.8508079648017883,
+      "learning_rate": 4.929486179848979e-06,
+      "loss": 0.6044,
+      "step": 498
+    },
+    {
+      "epoch": 0.5589992531740104,
+      "grad_norm": 0.8375908136367798,
+      "learning_rate": 4.9291330796214285e-06,
+      "loss": 0.6175,
+      "step": 499
+    },
+    {
+      "epoch": 0.5601194921583271,
+      "grad_norm": 0.8124059438705444,
+      "learning_rate": 4.928779110230126e-06,
+      "loss": 0.5857,
+      "step": 500
+    },
+    {
+      "epoch": 0.5612397311426438,
+      "grad_norm": 0.8159031271934509,
+      "learning_rate": 4.928424271801726e-06,
+      "loss": 0.6013,
+      "step": 501
+    },
+    {
+      "epoch": 0.5623599701269604,
+      "grad_norm": 0.7889281511306763,
+      "learning_rate": 4.928068564463192e-06,
+      "loss": 0.5959,
+      "step": 502
+    },
+    {
+      "epoch": 0.563480209111277,
+      "grad_norm": 0.8010871410369873,
+      "learning_rate": 4.927711988341798e-06,
+      "loss": 0.591,
+      "step": 503
+    },
+    {
+      "epoch": 0.5646004480955937,
+      "grad_norm": 0.8424665927886963,
+      "learning_rate": 4.927354543565131e-06,
+      "loss": 0.5844,
+      "step": 504
+    },
+    {
+      "epoch": 0.5657206870799104,
+      "grad_norm": 0.7999626398086548,
+      "learning_rate": 4.9269962302610866e-06,
+      "loss": 0.5836,
+      "step": 505
+    },
+    {
+      "epoch": 0.566840926064227,
+      "grad_norm": 0.8242966532707214,
+      "learning_rate": 4.926637048557873e-06,
+      "loss": 0.5929,
+      "step": 506
+    },
+    {
+      "epoch": 0.5679611650485437,
+      "grad_norm": 0.8201645016670227,
+      "learning_rate": 4.926276998584008e-06,
+      "loss": 0.5937,
+      "step": 507
+    },
+    {
+      "epoch": 0.5690814040328603,
+      "grad_norm": 0.8664683699607849,
+      "learning_rate": 4.925916080468321e-06,
+      "loss": 0.6258,
+      "step": 508
+    },
+    {
+      "epoch": 0.570201643017177,
+      "grad_norm": 0.8183404207229614,
+      "learning_rate": 4.925554294339951e-06,
+      "loss": 0.6088,
+      "step": 509
+    },
+    {
+      "epoch": 0.5713218820014937,
+      "grad_norm": 0.8237192034721375,
+      "learning_rate": 4.925191640328348e-06,
+      "loss": 0.6054,
+      "step": 510
+    },
+    {
+      "epoch": 0.5724421209858103,
+      "grad_norm": 0.8951362371444702,
+      "learning_rate": 4.924828118563273e-06,
+      "loss": 0.5847,
+      "step": 511
+    },
+    {
+      "epoch": 0.573562359970127,
+      "grad_norm": 0.8432679176330566,
+      "learning_rate": 4.924463729174796e-06,
+      "loss": 0.6185,
+      "step": 512
+    },
+    {
+      "epoch": 0.5746825989544436,
+      "grad_norm": 0.8540483117103577,
+      "learning_rate": 4.924098472293301e-06,
+      "loss": 0.6119,
+      "step": 513
+    },
+    {
+      "epoch": 0.5758028379387603,
+      "grad_norm": 0.8214981555938721,
+      "learning_rate": 4.923732348049477e-06,
+      "loss": 0.5815,
+      "step": 514
+    },
+    {
+      "epoch": 0.5769230769230769,
+      "grad_norm": 0.8091721534729004,
+      "learning_rate": 4.9233653565743276e-06,
+      "loss": 0.5912,
+      "step": 515
+    },
+    {
+      "epoch": 0.5780433159073936,
+      "grad_norm": 0.8201042413711548,
+      "learning_rate": 4.922997497999166e-06,
+      "loss": 0.5754,
+      "step": 516
+    },
+    {
+      "epoch": 0.5791635548917102,
+      "grad_norm": 0.8555132746696472,
+      "learning_rate": 4.922628772455613e-06,
+      "loss": 0.6227,
+      "step": 517
+    },
+    {
+      "epoch": 0.5802837938760269,
+      "grad_norm": 0.805891752243042,
+      "learning_rate": 4.922259180075604e-06,
+      "loss": 0.5782,
+      "step": 518
+    },
+    {
+      "epoch": 0.5814040328603436,
+      "grad_norm": 0.8111713528633118,
+      "learning_rate": 4.921888720991381e-06,
+      "loss": 0.5918,
+      "step": 519
+    },
+    {
+      "epoch": 0.5825242718446602,
+      "grad_norm": 0.8208431601524353,
+      "learning_rate": 4.921517395335497e-06,
+      "loss": 0.6038,
+      "step": 520
+    },
+    {
+      "epoch": 0.5836445108289768,
+      "grad_norm": 0.826924204826355,
+      "learning_rate": 4.921145203240816e-06,
+      "loss": 0.5922,
+      "step": 521
+    },
+    {
+      "epoch": 0.5847647498132935,
+      "grad_norm": 0.8484541177749634,
+      "learning_rate": 4.92077214484051e-06,
+      "loss": 0.596,
+      "step": 522
+    },
+    {
+      "epoch": 0.5858849887976102,
+      "grad_norm": 0.8284783363342285,
+      "learning_rate": 4.920398220268064e-06,
+      "loss": 0.599,
+      "step": 523
+    },
+    {
+      "epoch": 0.5870052277819268,
+      "grad_norm": 0.8378069400787354,
+      "learning_rate": 4.920023429657273e-06,
+      "loss": 0.6259,
+      "step": 524
+    },
+    {
+      "epoch": 0.5881254667662434,
+      "grad_norm": 0.8292249441146851,
+      "learning_rate": 4.919647773142235e-06,
+      "loss": 0.5925,
+      "step": 525
+    },
+    {
+      "epoch": 0.5892457057505601,
+      "grad_norm": 0.8202025890350342,
+      "learning_rate": 4.919271250857367e-06,
+      "loss": 0.5932,
+      "step": 526
+    },
+    {
+      "epoch": 0.5903659447348768,
+      "grad_norm": 0.8431236147880554,
+      "learning_rate": 4.9188938629373915e-06,
+      "loss": 0.5958,
+      "step": 527
+    },
+    {
+      "epoch": 0.5914861837191935,
+      "grad_norm": 0.8396396040916443,
+      "learning_rate": 4.9185156095173395e-06,
+      "loss": 0.6148,
+      "step": 528
+    },
+    {
+      "epoch": 0.59260642270351,
+      "grad_norm": 0.8345813155174255,
+      "learning_rate": 4.9181364907325535e-06,
+      "loss": 0.5651,
+      "step": 529
+    },
+    {
+      "epoch": 0.5937266616878267,
+      "grad_norm": 0.8040415644645691,
+      "learning_rate": 4.917756506718686e-06,
+      "loss": 0.6102,
+      "step": 530
+    },
+    {
+      "epoch": 0.5948469006721434,
+      "grad_norm": 0.7991846799850464,
+      "learning_rate": 4.917375657611697e-06,
+      "loss": 0.6,
+      "step": 531
+    },
+    {
+      "epoch": 0.5959671396564601,
+      "grad_norm": 0.8031482696533203,
+      "learning_rate": 4.9169939435478595e-06,
+      "loss": 0.5883,
+      "step": 532
+    },
+    {
+      "epoch": 0.5970873786407767,
+      "grad_norm": 0.8042899370193481,
+      "learning_rate": 4.9166113646637534e-06,
+      "loss": 0.561,
+      "step": 533
+    },
+    {
+      "epoch": 0.5982076176250933,
+      "grad_norm": 0.8222240209579468,
+      "learning_rate": 4.916227921096267e-06,
+      "loss": 0.5649,
+      "step": 534
+    },
+    {
+      "epoch": 0.59932785660941,
+      "grad_norm": 0.7763756513595581,
+      "learning_rate": 4.915843612982601e-06,
+      "loss": 0.577,
+      "step": 535
+    },
+    {
+      "epoch": 0.6004480955937267,
+      "grad_norm": 0.8542895913124084,
+      "learning_rate": 4.915458440460265e-06,
+      "loss": 0.6099,
+      "step": 536
+    },
+    {
+      "epoch": 0.6015683345780433,
+      "grad_norm": 0.8213333487510681,
+      "learning_rate": 4.915072403667074e-06,
+      "loss": 0.6131,
+      "step": 537
+    },
+    {
+      "epoch": 0.6026885735623599,
+      "grad_norm": 0.8075094819068909,
+      "learning_rate": 4.914685502741157e-06,
+      "loss": 0.5938,
+      "step": 538
+    },
+    {
+      "epoch": 0.6038088125466766,
+      "grad_norm": 0.843874454498291,
+      "learning_rate": 4.91429773782095e-06,
+      "loss": 0.5921,
+      "step": 539
+    },
+    {
+      "epoch": 0.6049290515309933,
+      "grad_norm": 0.8286192417144775,
+      "learning_rate": 4.913909109045198e-06,
+      "loss": 0.6035,
+      "step": 540
+    },
+    {
+      "epoch": 0.60604929051531,
+      "grad_norm": 0.7961990237236023,
+      "learning_rate": 4.913519616552957e-06,
+      "loss": 0.6176,
+      "step": 541
+    },
+    {
+      "epoch": 0.6071695294996265,
+      "grad_norm": 0.7807387709617615,
+      "learning_rate": 4.913129260483589e-06,
+      "loss": 0.5905,
+      "step": 542
+    },
+    {
+      "epoch": 0.6082897684839432,
+      "grad_norm": 0.8079019784927368,
+      "learning_rate": 4.912738040976767e-06,
+      "loss": 0.614,
+      "step": 543
+    },
+    {
+      "epoch": 0.6094100074682599,
+      "grad_norm": 0.7950747609138489,
+      "learning_rate": 4.912345958172474e-06,
+      "loss": 0.6129,
+      "step": 544
+    },
+    {
+      "epoch": 0.6105302464525766,
+      "grad_norm": 0.7880637645721436,
+      "learning_rate": 4.911953012210998e-06,
+      "loss": 0.5745,
+      "step": 545
+    },
+    {
+      "epoch": 0.6116504854368932,
+      "grad_norm": 0.797683596611023,
+      "learning_rate": 4.911559203232941e-06,
+      "loss": 0.5704,
+      "step": 546
+    },
+    {
+      "epoch": 0.6127707244212098,
+      "grad_norm": 0.7976189851760864,
+      "learning_rate": 4.911164531379208e-06,
+      "loss": 0.5753,
+      "step": 547
+    },
+    {
+      "epoch": 0.6138909634055265,
+      "grad_norm": 0.797524631023407,
+      "learning_rate": 4.910768996791018e-06,
+      "loss": 0.5965,
+      "step": 548
+    },
+    {
+      "epoch": 0.6150112023898432,
+      "grad_norm": 0.7966054677963257,
+      "learning_rate": 4.910372599609896e-06,
+      "loss": 0.5926,
+      "step": 549
+    },
+    {
+      "epoch": 0.6161314413741599,
+      "grad_norm": 0.8061427474021912,
+      "learning_rate": 4.909975339977676e-06,
+      "loss": 0.6076,
+      "step": 550
+    },
+    {
+      "epoch": 0.6172516803584764,
+      "grad_norm": 0.8192665576934814,
+      "learning_rate": 4.909577218036501e-06,
+      "loss": 0.5846,
+      "step": 551
+    },
+    {
+      "epoch": 0.6183719193427931,
+      "grad_norm": 0.8142986297607422,
+      "learning_rate": 4.909178233928822e-06,
+      "loss": 0.6269,
+      "step": 552
+    },
+    {
+      "epoch": 0.6194921583271098,
+      "grad_norm": 0.8129419684410095,
+      "learning_rate": 4.908778387797399e-06,
+      "loss": 0.6161,
+      "step": 553
+    },
+    {
+      "epoch": 0.6206123973114265,
+      "grad_norm": 0.7790163159370422,
+      "learning_rate": 4.9083776797853015e-06,
+      "loss": 0.6039,
+      "step": 554
+    },
+    {
+      "epoch": 0.621732636295743,
+      "grad_norm": 0.8370928168296814,
+      "learning_rate": 4.907976110035903e-06,
+      "loss": 0.5813,
+      "step": 555
+    },
+    {
+      "epoch": 0.6228528752800597,
+      "grad_norm": 0.8461244702339172,
+      "learning_rate": 4.907573678692891e-06,
+      "loss": 0.5879,
+      "step": 556
+    },
+    {
+      "epoch": 0.6239731142643764,
+      "grad_norm": 0.7943955659866333,
+      "learning_rate": 4.907170385900258e-06,
+      "loss": 0.579,
+      "step": 557
+    },
+    {
+      "epoch": 0.6250933532486931,
+      "grad_norm": 0.8056941628456116,
+      "learning_rate": 4.9067662318023056e-06,
+      "loss": 0.6096,
+      "step": 558
+    },
+    {
+      "epoch": 0.6262135922330098,
+      "grad_norm": 0.8091793060302734,
+      "learning_rate": 4.906361216543643e-06,
+      "loss": 0.5824,
+      "step": 559
+    },
+    {
+      "epoch": 0.6273338312173263,
+      "grad_norm": 0.8035262227058411,
+      "learning_rate": 4.905955340269188e-06,
+      "loss": 0.6232,
+      "step": 560
+    },
+    {
+      "epoch": 0.628454070201643,
+      "grad_norm": 0.7793251276016235,
+      "learning_rate": 4.9055486031241675e-06,
+      "loss": 0.5901,
+      "step": 561
+    },
+    {
+      "epoch": 0.6295743091859597,
+      "grad_norm": 0.7671579122543335,
+      "learning_rate": 4.9051410052541135e-06,
+      "loss": 0.5843,
+      "step": 562
+    },
+    {
+      "epoch": 0.6306945481702764,
+      "grad_norm": 0.8493455052375793,
+      "learning_rate": 4.90473254680487e-06,
+      "loss": 0.6254,
+      "step": 563
+    },
+    {
+      "epoch": 0.6318147871545929,
+      "grad_norm": 0.8092489838600159,
+      "learning_rate": 4.904323227922586e-06,
+      "loss": 0.5906,
+      "step": 564
+    },
+    {
+      "epoch": 0.6329350261389096,
+      "grad_norm": 0.8145341873168945,
+      "learning_rate": 4.903913048753719e-06,
+      "loss": 0.6129,
+      "step": 565
+    },
+    {
+      "epoch": 0.6340552651232263,
+      "grad_norm": 0.8953230381011963,
+      "learning_rate": 4.903502009445033e-06,
+      "loss": 0.5975,
+      "step": 566
+    },
+    {
+      "epoch": 0.635175504107543,
+      "grad_norm": 0.8133732676506042,
+      "learning_rate": 4.903090110143605e-06,
+      "loss": 0.558,
+      "step": 567
+    },
+    {
+      "epoch": 0.6362957430918595,
+      "grad_norm": 0.8249391317367554,
+      "learning_rate": 4.902677350996812e-06,
+      "loss": 0.5893,
+      "step": 568
+    },
+    {
+      "epoch": 0.6374159820761762,
+      "grad_norm": 0.8067794442176819,
+      "learning_rate": 4.902263732152345e-06,
+      "loss": 0.5823,
+      "step": 569
+    },
+    {
+      "epoch": 0.6385362210604929,
+      "grad_norm": 0.8032618761062622,
+      "learning_rate": 4.901849253758199e-06,
+      "loss": 0.6003,
+      "step": 570
+    },
+    {
+      "epoch": 0.6396564600448096,
+      "grad_norm": 0.8413413166999817,
+      "learning_rate": 4.901433915962678e-06,
+      "loss": 0.5742,
+      "step": 571
+    },
+    {
+      "epoch": 0.6407766990291263,
+      "grad_norm": 0.8067041039466858,
+      "learning_rate": 4.9010177189143935e-06,
+      "loss": 0.5902,
+      "step": 572
+    },
+    {
+      "epoch": 0.6418969380134428,
+      "grad_norm": 0.8752678632736206,
+      "learning_rate": 4.900600662762265e-06,
+      "loss": 0.6143,
+      "step": 573
+    },
+    {
+      "epoch": 0.6430171769977595,
+      "grad_norm": 0.8311488628387451,
+      "learning_rate": 4.9001827476555166e-06,
+      "loss": 0.5901,
+      "step": 574
+    },
+    {
+      "epoch": 0.6441374159820762,
+      "grad_norm": 0.8206403255462646,
+      "learning_rate": 4.899763973743684e-06,
+      "loss": 0.6056,
+      "step": 575
+    },
+    {
+      "epoch": 0.6452576549663929,
+      "grad_norm": 0.7974233627319336,
+      "learning_rate": 4.899344341176606e-06,
+      "loss": 0.5889,
+      "step": 576
+    },
+    {
+      "epoch": 0.6463778939507094,
+      "grad_norm": 0.8065476417541504,
+      "learning_rate": 4.898923850104432e-06,
+      "loss": 0.6,
+      "step": 577
+    },
+    {
+      "epoch": 0.6474981329350261,
+      "grad_norm": 0.8313128352165222,
+      "learning_rate": 4.898502500677617e-06,
+      "loss": 0.5973,
+      "step": 578
+    },
+    {
+      "epoch": 0.6486183719193428,
+      "grad_norm": 0.8816797733306885,
+      "learning_rate": 4.898080293046923e-06,
+      "loss": 0.5843,
+      "step": 579
+    },
+    {
+      "epoch": 0.6497386109036595,
+      "grad_norm": 0.816006600856781,
+      "learning_rate": 4.897657227363419e-06,
+      "loss": 0.6189,
+      "step": 580
+    },
+    {
+      "epoch": 0.6508588498879762,
+      "grad_norm": 0.8070967197418213,
+      "learning_rate": 4.897233303778481e-06,
+      "loss": 0.5872,
+      "step": 581
+    },
+    {
+      "epoch": 0.6519790888722927,
+      "grad_norm": 0.8039142489433289,
+      "learning_rate": 4.896808522443794e-06,
+      "loss": 0.5747,
+      "step": 582
+    },
+    {
+      "epoch": 0.6530993278566094,
+      "grad_norm": 0.8226941227912903,
+      "learning_rate": 4.896382883511347e-06,
+      "loss": 0.5845,
+      "step": 583
+    },
+    {
+      "epoch": 0.6542195668409261,
+      "grad_norm": 0.8548035025596619,
+      "learning_rate": 4.895956387133437e-06,
+      "loss": 0.5959,
+      "step": 584
+    },
+    {
+      "epoch": 0.6553398058252428,
+      "grad_norm": 0.8523133993148804,
+      "learning_rate": 4.895529033462669e-06,
+      "loss": 0.6107,
+      "step": 585
+    },
+    {
+      "epoch": 0.6564600448095593,
+      "grad_norm": 0.8199976086616516,
+      "learning_rate": 4.8951008226519525e-06,
+      "loss": 0.5981,
+      "step": 586
+    },
+    {
+      "epoch": 0.657580283793876,
+      "grad_norm": 0.8249738812446594,
+      "learning_rate": 4.894671754854505e-06,
+      "loss": 0.5888,
+      "step": 587
+    },
+    {
+      "epoch": 0.6587005227781927,
+      "grad_norm": 0.8519604206085205,
+      "learning_rate": 4.894241830223851e-06,
+      "loss": 0.6223,
+      "step": 588
+    },
+    {
+      "epoch": 0.6598207617625094,
+      "grad_norm": 0.8221259713172913,
+      "learning_rate": 4.8938110489138205e-06,
+      "loss": 0.5774,
+      "step": 589
+    },
+    {
+      "epoch": 0.6609410007468259,
+      "grad_norm": 0.8143469095230103,
+      "learning_rate": 4.893379411078552e-06,
+      "loss": 0.6049,
+      "step": 590
+    },
+    {
+      "epoch": 0.6620612397311426,
+      "grad_norm": 0.7872767448425293,
+      "learning_rate": 4.892946916872486e-06,
+      "loss": 0.5796,
+      "step": 591
+    },
+    {
+      "epoch": 0.6631814787154593,
+      "grad_norm": 0.8111532926559448,
+      "learning_rate": 4.8925135664503755e-06,
+      "loss": 0.6041,
+      "step": 592
+    },
+    {
+      "epoch": 0.664301717699776,
+      "grad_norm": 0.8038745522499084,
+      "learning_rate": 4.892079359967276e-06,
+      "loss": 0.5929,
+      "step": 593
+    },
+    {
+      "epoch": 0.6654219566840927,
+      "grad_norm": 0.8693163990974426,
+      "learning_rate": 4.89164429757855e-06,
+      "loss": 0.5892,
+      "step": 594
+    },
+    {
+      "epoch": 0.6665421956684092,
+      "grad_norm": 0.7890527248382568,
+      "learning_rate": 4.891208379439866e-06,
+      "loss": 0.5973,
+      "step": 595
+    },
+    {
+      "epoch": 0.6676624346527259,
+      "grad_norm": 0.8141840100288391,
+      "learning_rate": 4.890771605707199e-06,
+      "loss": 0.6031,
+      "step": 596
+    },
+    {
+      "epoch": 0.6687826736370426,
+      "grad_norm": 0.8265807032585144,
+      "learning_rate": 4.890333976536831e-06,
+      "loss": 0.5893,
+      "step": 597
+    },
+    {
+      "epoch": 0.6699029126213593,
+      "grad_norm": 0.8645585179328918,
+      "learning_rate": 4.8898954920853484e-06,
+      "loss": 0.6189,
+      "step": 598
+    },
+    {
+      "epoch": 0.6710231516056758,
+      "grad_norm": 0.8112260699272156,
+      "learning_rate": 4.889456152509646e-06,
+      "loss": 0.589,
+      "step": 599
+    },
+    {
+      "epoch": 0.6721433905899925,
+      "grad_norm": 0.8134514093399048,
+      "learning_rate": 4.8890159579669214e-06,
+      "loss": 0.5703,
+      "step": 600
+    },
+    {
+      "epoch": 0.6732636295743092,
+      "grad_norm": 0.8513573408126831,
+      "learning_rate": 4.888574908614681e-06,
+      "loss": 0.5971,
+      "step": 601
+    },
+    {
+      "epoch": 0.6743838685586259,
+      "grad_norm": 0.8489837646484375,
+      "learning_rate": 4.888133004610735e-06,
+      "loss": 0.6084,
+      "step": 602
+    },
+    {
+      "epoch": 0.6755041075429425,
+      "grad_norm": 0.8141038417816162,
+      "learning_rate": 4.887690246113201e-06,
+      "loss": 0.596,
+      "step": 603
+    },
+    {
+      "epoch": 0.6766243465272591,
+      "grad_norm": 0.8250095844268799,
+      "learning_rate": 4.887246633280501e-06,
+      "loss": 0.5535,
+      "step": 604
+    },
+    {
+      "epoch": 0.6777445855115758,
+      "grad_norm": 0.8543534874916077,
+      "learning_rate": 4.886802166271365e-06,
+      "loss": 0.6056,
+      "step": 605
+    },
+    {
+      "epoch": 0.6788648244958925,
+      "grad_norm": 0.8766425848007202,
+      "learning_rate": 4.886356845244824e-06,
+      "loss": 0.5984,
+      "step": 606
+    },
+    {
+      "epoch": 0.6799850634802092,
+      "grad_norm": 0.8217388391494751,
+      "learning_rate": 4.88591067036022e-06,
+      "loss": 0.6037,
+      "step": 607
+    },
+    {
+      "epoch": 0.6811053024645257,
+      "grad_norm": 0.7816200852394104,
+      "learning_rate": 4.885463641777197e-06,
+      "loss": 0.5451,
+      "step": 608
+    },
+    {
+      "epoch": 0.6822255414488424,
+      "grad_norm": 0.8485130071640015,
+      "learning_rate": 4.885015759655706e-06,
+      "loss": 0.5917,
+      "step": 609
+    },
+    {
+      "epoch": 0.6833457804331591,
+      "grad_norm": 0.8265321850776672,
+      "learning_rate": 4.884567024156002e-06,
+      "loss": 0.5676,
+      "step": 610
+    },
+    {
+      "epoch": 0.6844660194174758,
+      "grad_norm": 0.8090165853500366,
+      "learning_rate": 4.884117435438646e-06,
+      "loss": 0.616,
+      "step": 611
+    },
+    {
+      "epoch": 0.6855862584017924,
+      "grad_norm": 0.8049173951148987,
+      "learning_rate": 4.883666993664507e-06,
+      "loss": 0.5814,
+      "step": 612
+    },
+    {
+      "epoch": 0.686706497386109,
+      "grad_norm": 0.8055282831192017,
+      "learning_rate": 4.883215698994755e-06,
+      "loss": 0.5947,
+      "step": 613
+    },
+    {
+      "epoch": 0.6878267363704257,
+      "grad_norm": 0.8200820088386536,
+      "learning_rate": 4.882763551590865e-06,
+      "loss": 0.6262,
+      "step": 614
+    },
+    {
+      "epoch": 0.6889469753547424,
+      "grad_norm": 0.808571994304657,
+      "learning_rate": 4.882310551614623e-06,
+      "loss": 0.6029,
+      "step": 615
+    },
+    {
+      "epoch": 0.690067214339059,
+      "grad_norm": 0.8104593753814697,
+      "learning_rate": 4.881856699228114e-06,
+      "loss": 0.5979,
+      "step": 616
+    },
+    {
+      "epoch": 0.6911874533233756,
+      "grad_norm": 0.7940874695777893,
+      "learning_rate": 4.88140199459373e-06,
+      "loss": 0.5818,
+      "step": 617
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.7887553572654724,
+      "learning_rate": 4.880946437874169e-06,
+      "loss": 0.573,
+      "step": 618
+    },
+    {
+      "epoch": 0.693427931292009,
+      "grad_norm": 0.8413677215576172,
+      "learning_rate": 4.880490029232431e-06,
+      "loss": 0.6158,
+      "step": 619
+    },
+    {
+      "epoch": 0.6945481702763256,
+      "grad_norm": 0.8174417614936829,
+      "learning_rate": 4.880032768831825e-06,
+      "loss": 0.57,
+      "step": 620
+    },
+    {
+      "epoch": 0.6956684092606422,
+      "grad_norm": 0.8150051236152649,
+      "learning_rate": 4.879574656835962e-06,
+      "loss": 0.6071,
+      "step": 621
+    },
+    {
+      "epoch": 0.6967886482449589,
+      "grad_norm": 0.8536598682403564,
+      "learning_rate": 4.879115693408757e-06,
+      "loss": 0.6294,
+      "step": 622
+    },
+    {
+      "epoch": 0.6979088872292756,
+      "grad_norm": 0.8469665050506592,
+      "learning_rate": 4.8786558787144314e-06,
+      "loss": 0.6146,
+      "step": 623
+    },
+    {
+      "epoch": 0.6990291262135923,
+      "grad_norm": 0.7985464930534363,
+      "learning_rate": 4.878195212917511e-06,
+      "loss": 0.5528,
+      "step": 624
+    },
+    {
+      "epoch": 0.7001493651979089,
+      "grad_norm": 0.8789647817611694,
+      "learning_rate": 4.877733696182827e-06,
+      "loss": 0.5926,
+      "step": 625
+    },
+    {
+      "epoch": 0.7012696041822255,
+      "grad_norm": 0.7992759943008423,
+      "learning_rate": 4.877271328675511e-06,
+      "loss": 0.5912,
+      "step": 626
+    },
+    {
+      "epoch": 0.7023898431665422,
+      "grad_norm": 0.7898268699645996,
+      "learning_rate": 4.876808110561004e-06,
+      "loss": 0.5722,
+      "step": 627
+    },
+    {
+      "epoch": 0.7035100821508589,
+      "grad_norm": 0.8306590914726257,
+      "learning_rate": 4.87634404200505e-06,
+      "loss": 0.627,
+      "step": 628
+    },
+    {
+      "epoch": 0.7046303211351755,
+      "grad_norm": 0.8404473662376404,
+      "learning_rate": 4.875879123173694e-06,
+      "loss": 0.5841,
+      "step": 629
+    },
+    {
+      "epoch": 0.7057505601194921,
+      "grad_norm": 0.8321950435638428,
+      "learning_rate": 4.87541335423329e-06,
+      "loss": 0.6141,
+      "step": 630
+    },
+    {
+      "epoch": 0.7068707991038088,
+      "grad_norm": 0.7946979403495789,
+      "learning_rate": 4.874946735350492e-06,
+      "loss": 0.5849,
+      "step": 631
+    },
+    {
+      "epoch": 0.7079910380881255,
+      "grad_norm": 0.8189645409584045,
+      "learning_rate": 4.874479266692261e-06,
+      "loss": 0.6146,
+      "step": 632
+    },
+    {
+      "epoch": 0.7091112770724421,
+      "grad_norm": 0.8169258832931519,
+      "learning_rate": 4.8740109484258625e-06,
+      "loss": 0.5926,
+      "step": 633
+    },
+    {
+      "epoch": 0.7102315160567588,
+      "grad_norm": 0.7945389151573181,
+      "learning_rate": 4.873541780718863e-06,
+      "loss": 0.5929,
+      "step": 634
+    },
+    {
+      "epoch": 0.7113517550410754,
+      "grad_norm": 0.7935434579849243,
+      "learning_rate": 4.873071763739135e-06,
+      "loss": 0.5942,
+      "step": 635
+    },
+    {
+      "epoch": 0.7124719940253921,
+      "grad_norm": 0.8001875877380371,
+      "learning_rate": 4.872600897654853e-06,
+      "loss": 0.5957,
+      "step": 636
+    },
+    {
+      "epoch": 0.7135922330097088,
+      "grad_norm": 0.8338243961334229,
+      "learning_rate": 4.872129182634499e-06,
+      "loss": 0.5908,
+      "step": 637
+    },
+    {
+      "epoch": 0.7147124719940254,
+      "grad_norm": 0.8988234400749207,
+      "learning_rate": 4.871656618846855e-06,
+      "loss": 0.5955,
+      "step": 638
+    },
+    {
+      "epoch": 0.715832710978342,
+      "grad_norm": 0.7992603778839111,
+      "learning_rate": 4.87118320646101e-06,
+      "loss": 0.5852,
+      "step": 639
+    },
+    {
+      "epoch": 0.7169529499626587,
+      "grad_norm": 0.8259397745132446,
+      "learning_rate": 4.870708945646353e-06,
+      "loss": 0.6156,
+      "step": 640
+    },
+    {
+      "epoch": 0.7180731889469754,
+      "grad_norm": 0.893584132194519,
+      "learning_rate": 4.870233836572579e-06,
+      "loss": 0.6117,
+      "step": 641
+    },
+    {
+      "epoch": 0.719193427931292,
+      "grad_norm": 0.8482956290245056,
+      "learning_rate": 4.869757879409686e-06,
+      "loss": 0.6018,
+      "step": 642
+    },
+    {
+      "epoch": 0.7203136669156086,
+      "grad_norm": 0.7782022356987,
+      "learning_rate": 4.869281074327975e-06,
+      "loss": 0.5761,
+      "step": 643
+    },
+    {
+      "epoch": 0.7214339058999253,
+      "grad_norm": 0.8154391050338745,
+      "learning_rate": 4.8688034214980515e-06,
+      "loss": 0.5846,
+      "step": 644
+    },
+    {
+      "epoch": 0.722554144884242,
+      "grad_norm": 0.8113802671432495,
+      "learning_rate": 4.868324921090822e-06,
+      "loss": 0.5616,
+      "step": 645
+    },
+    {
+      "epoch": 0.7236743838685586,
+      "grad_norm": 0.830033540725708,
+      "learning_rate": 4.867845573277501e-06,
+      "loss": 0.63,
+      "step": 646
+    },
+    {
+      "epoch": 0.7247946228528753,
+      "grad_norm": 0.8050087690353394,
+      "learning_rate": 4.867365378229601e-06,
+      "loss": 0.5894,
+      "step": 647
+    },
+    {
+      "epoch": 0.7259148618371919,
+      "grad_norm": 0.8076449036598206,
+      "learning_rate": 4.8668843361189385e-06,
+      "loss": 0.6145,
+      "step": 648
+    },
+    {
+      "epoch": 0.7270351008215086,
+      "grad_norm": 0.8889169692993164,
+      "learning_rate": 4.8664024471176366e-06,
+      "loss": 0.6154,
+      "step": 649
+    },
+    {
+      "epoch": 0.7281553398058253,
+      "grad_norm": 0.8604964017868042,
+      "learning_rate": 4.865919711398118e-06,
+      "loss": 0.5984,
+      "step": 650
+    },
+    {
+      "epoch": 0.7292755787901419,
+      "grad_norm": 0.828212320804596,
+      "learning_rate": 4.86543612913311e-06,
+      "loss": 0.5785,
+      "step": 651
+    },
+    {
+      "epoch": 0.7303958177744585,
+      "grad_norm": 0.786229133605957,
+      "learning_rate": 4.864951700495642e-06,
+      "loss": 0.5983,
+      "step": 652
+    },
+    {
+      "epoch": 0.7315160567587752,
+      "grad_norm": 0.8467420935630798,
+      "learning_rate": 4.864466425659046e-06,
+      "loss": 0.6037,
+      "step": 653
+    },
+    {
+      "epoch": 0.7326362957430919,
+      "grad_norm": 0.7861534953117371,
+      "learning_rate": 4.863980304796958e-06,
+      "loss": 0.564,
+      "step": 654
+    },
+    {
+      "epoch": 0.7337565347274085,
+      "grad_norm": 0.7819376587867737,
+      "learning_rate": 4.863493338083316e-06,
+      "loss": 0.5783,
+      "step": 655
+    },
+    {
+      "epoch": 0.7348767737117252,
+      "grad_norm": 0.7864521741867065,
+      "learning_rate": 4.863005525692362e-06,
+      "loss": 0.594,
+      "step": 656
+    },
+    {
+      "epoch": 0.7359970126960418,
+      "grad_norm": 0.7910285592079163,
+      "learning_rate": 4.862516867798636e-06,
+      "loss": 0.5888,
+      "step": 657
+    },
+    {
+      "epoch": 0.7371172516803585,
+      "grad_norm": 0.8056003451347351,
+      "learning_rate": 4.862027364576987e-06,
+      "loss": 0.5895,
+      "step": 658
+    },
+    {
+      "epoch": 0.7382374906646751,
+      "grad_norm": 0.8140441179275513,
+      "learning_rate": 4.861537016202561e-06,
+      "loss": 0.6015,
+      "step": 659
+    },
+    {
+      "epoch": 0.7393577296489918,
+      "grad_norm": 0.8237465620040894,
+      "learning_rate": 4.861045822850811e-06,
+      "loss": 0.6072,
+      "step": 660
+    },
+    {
+      "epoch": 0.7404779686333084,
+      "grad_norm": 0.832146942615509,
+      "learning_rate": 4.860553784697487e-06,
+      "loss": 0.6171,
+      "step": 661
+    },
+    {
+      "epoch": 0.7415982076176251,
+      "grad_norm": 0.8073914051055908,
+      "learning_rate": 4.860060901918648e-06,
+      "loss": 0.5931,
+      "step": 662
+    },
+    {
+      "epoch": 0.7427184466019418,
+      "grad_norm": 0.8193264007568359,
+      "learning_rate": 4.859567174690647e-06,
+      "loss": 0.5948,
+      "step": 663
+    },
+    {
+      "epoch": 0.7438386855862584,
+      "grad_norm": 0.8580235838890076,
+      "learning_rate": 4.859072603190148e-06,
+      "loss": 0.6157,
+      "step": 664
+    },
+    {
+      "epoch": 0.7449589245705751,
+      "grad_norm": 0.7729256749153137,
+      "learning_rate": 4.8585771875941105e-06,
+      "loss": 0.5876,
+      "step": 665
+    },
+    {
+      "epoch": 0.7460791635548917,
+      "grad_norm": 0.844573438167572,
+      "learning_rate": 4.858080928079798e-06,
+      "loss": 0.6206,
+      "step": 666
+    },
+    {
+      "epoch": 0.7471994025392084,
+      "grad_norm": 0.816228449344635,
+      "learning_rate": 4.857583824824777e-06,
+      "loss": 0.6247,
+      "step": 667
+    },
+    {
+      "epoch": 0.748319641523525,
+      "grad_norm": 0.7984215021133423,
+      "learning_rate": 4.8570858780069155e-06,
+      "loss": 0.6038,
+      "step": 668
+    },
+    {
+      "epoch": 0.7494398805078417,
+      "grad_norm": 0.850842297077179,
+      "learning_rate": 4.856587087804382e-06,
+      "loss": 0.6283,
+      "step": 669
+    },
+    {
+      "epoch": 0.7505601194921583,
+      "grad_norm": 0.7965848445892334,
+      "learning_rate": 4.856087454395648e-06,
+      "loss": 0.6018,
+      "step": 670
+    },
+    {
+      "epoch": 0.751680358476475,
+      "grad_norm": 0.7891208529472351,
+      "learning_rate": 4.855586977959487e-06,
+      "loss": 0.5965,
+      "step": 671
+    },
+    {
+      "epoch": 0.7528005974607916,
+      "grad_norm": 0.7875372767448425,
+      "learning_rate": 4.855085658674973e-06,
+      "loss": 0.6004,
+      "step": 672
+    },
+    {
+      "epoch": 0.7539208364451083,
+      "grad_norm": 0.7898383736610413,
+      "learning_rate": 4.854583496721482e-06,
+      "loss": 0.5566,
+      "step": 673
+    },
+    {
+      "epoch": 0.7550410754294249,
+      "grad_norm": 0.8215773701667786,
+      "learning_rate": 4.854080492278693e-06,
+      "loss": 0.5837,
+      "step": 674
+    },
+    {
+      "epoch": 0.7561613144137416,
+      "grad_norm": 0.810920000076294,
+      "learning_rate": 4.853576645526584e-06,
+      "loss": 0.5893,
+      "step": 675
+    },
+    {
+      "epoch": 0.7572815533980582,
+      "grad_norm": 0.7786296606063843,
+      "learning_rate": 4.8530719566454345e-06,
+      "loss": 0.5604,
+      "step": 676
+    },
+    {
+      "epoch": 0.7584017923823749,
+      "grad_norm": 0.7917574048042297,
+      "learning_rate": 4.852566425815829e-06,
+      "loss": 0.582,
+      "step": 677
+    },
+    {
+      "epoch": 0.7595220313666916,
+      "grad_norm": 0.8070117235183716,
+      "learning_rate": 4.8520600532186475e-06,
+      "loss": 0.604,
+      "step": 678
+    },
+    {
+      "epoch": 0.7606422703510082,
+      "grad_norm": 0.8068091869354248,
+      "learning_rate": 4.8515528390350765e-06,
+      "loss": 0.5896,
+      "step": 679
+    },
+    {
+      "epoch": 0.7617625093353249,
+      "grad_norm": 0.8220896124839783,
+      "learning_rate": 4.8510447834466e-06,
+      "loss": 0.593,
+      "step": 680
+    },
+    {
+      "epoch": 0.7628827483196415,
+      "grad_norm": 0.7868808507919312,
+      "learning_rate": 4.850535886635006e-06,
+      "loss": 0.6,
+      "step": 681
+    },
+    {
+      "epoch": 0.7640029873039582,
+      "grad_norm": 0.8102167248725891,
+      "learning_rate": 4.850026148782381e-06,
+      "loss": 0.5813,
+      "step": 682
+    },
+    {
+      "epoch": 0.7651232262882748,
+      "grad_norm": 0.8151317834854126,
+      "learning_rate": 4.849515570071113e-06,
+      "loss": 0.5747,
+      "step": 683
+    },
+    {
+      "epoch": 0.7662434652725915,
+      "grad_norm": 0.8038537502288818,
+      "learning_rate": 4.849004150683892e-06,
+      "loss": 0.6192,
+      "step": 684
+    },
+    {
+      "epoch": 0.7673637042569081,
+      "grad_norm": 0.8006526827812195,
+      "learning_rate": 4.848491890803708e-06,
+      "loss": 0.5866,
+      "step": 685
+    },
+    {
+      "epoch": 0.7684839432412248,
+      "grad_norm": 0.7970767617225647,
+      "learning_rate": 4.847978790613851e-06,
+      "loss": 0.6098,
+      "step": 686
+    },
+    {
+      "epoch": 0.7696041822255415,
+      "grad_norm": 0.7953868508338928,
+      "learning_rate": 4.847464850297914e-06,
+      "loss": 0.5758,
+      "step": 687
+    },
+    {
+      "epoch": 0.7707244212098581,
+      "grad_norm": 0.8159863352775574,
+      "learning_rate": 4.8469500700397875e-06,
+      "loss": 0.5949,
+      "step": 688
+    },
+    {
+      "epoch": 0.7718446601941747,
+      "grad_norm": 0.7919224500656128,
+      "learning_rate": 4.8464344500236645e-06,
+      "loss": 0.5769,
+      "step": 689
+    },
+    {
+      "epoch": 0.7729648991784914,
+      "grad_norm": 0.7827619314193726,
+      "learning_rate": 4.845917990434039e-06,
+      "loss": 0.5923,
+      "step": 690
+    },
+    {
+      "epoch": 0.7740851381628081,
+      "grad_norm": 0.7937460541725159,
+      "learning_rate": 4.845400691455703e-06,
+      "loss": 0.5759,
+      "step": 691
+    },
+    {
+      "epoch": 0.7752053771471247,
+      "grad_norm": 0.7901214361190796,
+      "learning_rate": 4.844882553273752e-06,
+      "loss": 0.5657,
+      "step": 692
+    },
+    {
+      "epoch": 0.7763256161314414,
+      "grad_norm": 0.8160180449485779,
+      "learning_rate": 4.844363576073579e-06,
+      "loss": 0.5971,
+      "step": 693
+    },
+    {
+      "epoch": 0.777445855115758,
+      "grad_norm": 0.8019487261772156,
+      "learning_rate": 4.843843760040877e-06,
+      "loss": 0.5804,
+      "step": 694
+    },
+    {
+      "epoch": 0.7785660941000747,
+      "grad_norm": 0.7898805141448975,
+      "learning_rate": 4.843323105361643e-06,
+      "loss": 0.5854,
+      "step": 695
+    },
+    {
+      "epoch": 0.7796863330843914,
+      "grad_norm": 0.829105794429779,
+      "learning_rate": 4.842801612222171e-06,
+      "loss": 0.5914,
+      "step": 696
+    },
+    {
+      "epoch": 0.780806572068708,
+      "grad_norm": 0.7974743843078613,
+      "learning_rate": 4.842279280809054e-06,
+      "loss": 0.5856,
+      "step": 697
+    },
+    {
+      "epoch": 0.7819268110530246,
+      "grad_norm": 0.7757387161254883,
+      "learning_rate": 4.841756111309188e-06,
+      "loss": 0.5679,
+      "step": 698
+    },
+    {
+      "epoch": 0.7830470500373413,
+      "grad_norm": 0.827438473701477,
+      "learning_rate": 4.841232103909766e-06,
+      "loss": 0.5657,
+      "step": 699
+    },
+    {
+      "epoch": 0.784167289021658,
+      "grad_norm": 0.8150497078895569,
+      "learning_rate": 4.840707258798284e-06,
+      "loss": 0.6065,
+      "step": 700
+    },
+    {
+      "epoch": 0.7852875280059746,
+      "grad_norm": 0.7733539342880249,
+      "learning_rate": 4.8401815761625345e-06,
+      "loss": 0.5806,
+      "step": 701
+    },
+    {
+      "epoch": 0.7864077669902912,
+      "grad_norm": 0.8224393129348755,
+      "learning_rate": 4.839655056190611e-06,
+      "loss": 0.5723,
+      "step": 702
+    },
+    {
+      "epoch": 0.7875280059746079,
+      "grad_norm": 0.8295096755027771,
+      "learning_rate": 4.839127699070907e-06,
+      "loss": 0.609,
+      "step": 703
+    },
+    {
+      "epoch": 0.7886482449589246,
+      "grad_norm": 0.7875826358795166,
+      "learning_rate": 4.838599504992116e-06,
+      "loss": 0.5625,
+      "step": 704
+    },
+    {
+      "epoch": 0.7897684839432412,
+      "grad_norm": 0.8078942894935608,
+      "learning_rate": 4.838070474143228e-06,
+      "loss": 0.6059,
+      "step": 705
+    },
+    {
+      "epoch": 0.7908887229275579,
+      "grad_norm": 0.8128692507743835,
+      "learning_rate": 4.837540606713538e-06,
+      "loss": 0.5878,
+      "step": 706
+    },
+    {
+      "epoch": 0.7920089619118745,
+      "grad_norm": 0.7869188785552979,
+      "learning_rate": 4.837009902892633e-06,
+      "loss": 0.5741,
+      "step": 707
+    },
+    {
+      "epoch": 0.7931292008961912,
+      "grad_norm": 0.7958232164382935,
+      "learning_rate": 4.836478362870406e-06,
+      "loss": 0.5879,
+      "step": 708
+    },
+    {
+      "epoch": 0.7942494398805079,
+      "grad_norm": 0.7908474802970886,
+      "learning_rate": 4.835945986837045e-06,
+      "loss": 0.5766,
+      "step": 709
+    },
+    {
+      "epoch": 0.7953696788648245,
+      "grad_norm": 0.8136919736862183,
+      "learning_rate": 4.83541277498304e-06,
+      "loss": 0.587,
+      "step": 710
+    },
+    {
+      "epoch": 0.7964899178491411,
+      "grad_norm": 0.8330440521240234,
+      "learning_rate": 4.8348787274991765e-06,
+      "loss": 0.5935,
+      "step": 711
+    },
+    {
+      "epoch": 0.7976101568334578,
+      "grad_norm": 0.8112781643867493,
+      "learning_rate": 4.834343844576543e-06,
+      "loss": 0.5952,
+      "step": 712
+    },
+    {
+      "epoch": 0.7987303958177745,
+      "grad_norm": 0.8082567453384399,
+      "learning_rate": 4.8338081264065235e-06,
+      "loss": 0.5717,
+      "step": 713
+    },
+    {
+      "epoch": 0.7998506348020911,
+      "grad_norm": 0.9035626649856567,
+      "learning_rate": 4.8332715731808035e-06,
+      "loss": 0.583,
+      "step": 714
+    },
+    {
+      "epoch": 0.8009708737864077,
+      "grad_norm": 0.8441537022590637,
+      "learning_rate": 4.8327341850913645e-06,
+      "loss": 0.6096,
+      "step": 715
+    },
+    {
+      "epoch": 0.8020911127707244,
+      "grad_norm": 0.8210310339927673,
+      "learning_rate": 4.832195962330491e-06,
+      "loss": 0.6072,
+      "step": 716
+    },
+    {
+      "epoch": 0.8032113517550411,
+      "grad_norm": 0.8119834065437317,
+      "learning_rate": 4.831656905090761e-06,
+      "loss": 0.5864,
+      "step": 717
+    },
+    {
+      "epoch": 0.8043315907393578,
+      "grad_norm": 0.8333593606948853,
+      "learning_rate": 4.831117013565056e-06,
+      "loss": 0.6003,
+      "step": 718
+    },
+    {
+      "epoch": 0.8054518297236744,
+      "grad_norm": 0.8569757342338562,
+      "learning_rate": 4.83057628794655e-06,
+      "loss": 0.5936,
+      "step": 719
+    },
+    {
+      "epoch": 0.806572068707991,
+      "grad_norm": 0.7970093488693237,
+      "learning_rate": 4.830034728428723e-06,
+      "loss": 0.5776,
+      "step": 720
+    },
+    {
+      "epoch": 0.8076923076923077,
+      "grad_norm": 0.8280375003814697,
+      "learning_rate": 4.829492335205347e-06,
+      "loss": 0.5952,
+      "step": 721
+    },
+    {
+      "epoch": 0.8088125466766244,
+      "grad_norm": 0.8139337301254272,
+      "learning_rate": 4.828949108470496e-06,
+      "loss": 0.6314,
+      "step": 722
+    },
+    {
+      "epoch": 0.809932785660941,
+      "grad_norm": 0.8315449357032776,
+      "learning_rate": 4.82840504841854e-06,
+      "loss": 0.5978,
+      "step": 723
+    },
+    {
+      "epoch": 0.8110530246452576,
+      "grad_norm": 0.8375460505485535,
+      "learning_rate": 4.827860155244149e-06,
+      "loss": 0.588,
+      "step": 724
+    },
+    {
+      "epoch": 0.8121732636295743,
+      "grad_norm": 0.7952830791473389,
+      "learning_rate": 4.827314429142291e-06,
+      "loss": 0.582,
+      "step": 725
+    },
+    {
+      "epoch": 0.813293502613891,
+      "grad_norm": 0.8073609471321106,
+      "learning_rate": 4.826767870308228e-06,
+      "loss": 0.5678,
+      "step": 726
+    },
+    {
+      "epoch": 0.8144137415982076,
+      "grad_norm": 0.8218507766723633,
+      "learning_rate": 4.826220478937527e-06,
+      "loss": 0.6072,
+      "step": 727
+    },
+    {
+      "epoch": 0.8155339805825242,
+      "grad_norm": 0.7980933785438538,
+      "learning_rate": 4.825672255226047e-06,
+      "loss": 0.5524,
+      "step": 728
+    },
+    {
+      "epoch": 0.8166542195668409,
+      "grad_norm": 0.8054032921791077,
+      "learning_rate": 4.825123199369947e-06,
+      "loss": 0.5913,
+      "step": 729
+    },
+    {
+      "epoch": 0.8177744585511576,
+      "grad_norm": 0.8572160601615906,
+      "learning_rate": 4.824573311565685e-06,
+      "loss": 0.5954,
+      "step": 730
+    },
+    {
+      "epoch": 0.8188946975354743,
+      "grad_norm": 0.8065678477287292,
+      "learning_rate": 4.824022592010013e-06,
+      "loss": 0.5966,
+      "step": 731
+    },
+    {
+      "epoch": 0.8200149365197908,
+      "grad_norm": 0.812080442905426,
+      "learning_rate": 4.8234710408999865e-06,
+      "loss": 0.5666,
+      "step": 732
+    },
+    {
+      "epoch": 0.8211351755041075,
+      "grad_norm": 0.8508359789848328,
+      "learning_rate": 4.822918658432952e-06,
+      "loss": 0.6137,
+      "step": 733
+    },
+    {
+      "epoch": 0.8222554144884242,
+      "grad_norm": 0.7996507287025452,
+      "learning_rate": 4.8223654448065585e-06,
+      "loss": 0.6038,
+      "step": 734
+    },
+    {
+      "epoch": 0.8233756534727409,
+      "grad_norm": 0.8098906874656677,
+      "learning_rate": 4.821811400218748e-06,
+      "loss": 0.5759,
+      "step": 735
+    },
+    {
+      "epoch": 0.8244958924570575,
+      "grad_norm": 0.8356794118881226,
+      "learning_rate": 4.821256524867764e-06,
+      "loss": 0.6151,
+      "step": 736
+    },
+    {
+      "epoch": 0.8256161314413741,
+      "grad_norm": 0.8081963658332825,
+      "learning_rate": 4.820700818952145e-06,
+      "loss": 0.5943,
+      "step": 737
+    },
+    {
+      "epoch": 0.8267363704256908,
+      "grad_norm": 0.825581967830658,
+      "learning_rate": 4.820144282670728e-06,
+      "loss": 0.5769,
+      "step": 738
+    },
+    {
+      "epoch": 0.8278566094100075,
+      "grad_norm": 0.8103387355804443,
+      "learning_rate": 4.819586916222644e-06,
+      "loss": 0.5789,
+      "step": 739
+    },
+    {
+      "epoch": 0.8289768483943242,
+      "grad_norm": 0.8200883865356445,
+      "learning_rate": 4.819028719807326e-06,
+      "loss": 0.602,
+      "step": 740
+    },
+    {
+      "epoch": 0.8300970873786407,
+      "grad_norm": 0.8113400340080261,
+      "learning_rate": 4.818469693624499e-06,
+      "loss": 0.5951,
+      "step": 741
+    },
+    {
+      "epoch": 0.8312173263629574,
+      "grad_norm": 0.8407019972801208,
+      "learning_rate": 4.817909837874189e-06,
+      "loss": 0.5909,
+      "step": 742
+    },
+    {
+      "epoch": 0.8323375653472741,
+      "grad_norm": 0.8122291564941406,
+      "learning_rate": 4.8173491527567155e-06,
+      "loss": 0.5964,
+      "step": 743
+    },
+    {
+      "epoch": 0.8334578043315908,
+      "grad_norm": 0.8028641939163208,
+      "learning_rate": 4.816787638472697e-06,
+      "loss": 0.5926,
+      "step": 744
+    },
+    {
+      "epoch": 0.8345780433159073,
+      "grad_norm": 0.8061012625694275,
+      "learning_rate": 4.816225295223048e-06,
+      "loss": 0.5835,
+      "step": 745
+    },
+    {
+      "epoch": 0.835698282300224,
+      "grad_norm": 0.8372674584388733,
+      "learning_rate": 4.815662123208979e-06,
+      "loss": 0.6048,
+      "step": 746
+    },
+    {
+      "epoch": 0.8368185212845407,
+      "grad_norm": 0.7943447828292847,
+      "learning_rate": 4.815098122631998e-06,
+      "loss": 0.5999,
+      "step": 747
+    },
+    {
+      "epoch": 0.8379387602688574,
+      "grad_norm": 0.7981782555580139,
+      "learning_rate": 4.814533293693909e-06,
+      "loss": 0.5838,
+      "step": 748
+    },
+    {
+      "epoch": 0.8390589992531741,
+      "grad_norm": 0.8036529421806335,
+      "learning_rate": 4.8139676365968125e-06,
+      "loss": 0.59,
+      "step": 749
+    },
+    {
+      "epoch": 0.8401792382374906,
+      "grad_norm": 0.8037401437759399,
+      "learning_rate": 4.813401151543105e-06,
+      "loss": 0.5603,
+      "step": 750
+    },
+    {
+      "epoch": 0.8412994772218073,
+      "grad_norm": 0.8377910852432251,
+      "learning_rate": 4.812833838735479e-06,
+      "loss": 0.5984,
+      "step": 751
+    },
+    {
+      "epoch": 0.842419716206124,
+      "grad_norm": 0.7984040975570679,
+      "learning_rate": 4.812265698376924e-06,
+      "loss": 0.5719,
+      "step": 752
+    },
+    {
+      "epoch": 0.8435399551904407,
+      "grad_norm": 0.8077371716499329,
+      "learning_rate": 4.811696730670726e-06,
+      "loss": 0.5824,
+      "step": 753
+    },
+    {
+      "epoch": 0.8446601941747572,
+      "grad_norm": 0.7634063959121704,
+      "learning_rate": 4.811126935820466e-06,
+      "loss": 0.5584,
+      "step": 754
+    },
+    {
+      "epoch": 0.8457804331590739,
+      "grad_norm": 0.8120932579040527,
+      "learning_rate": 4.810556314030021e-06,
+      "loss": 0.5729,
+      "step": 755
+    },
+    {
+      "epoch": 0.8469006721433906,
+      "grad_norm": 0.7915536165237427,
+      "learning_rate": 4.8099848655035655e-06,
+      "loss": 0.6039,
+      "step": 756
+    },
+    {
+      "epoch": 0.8480209111277073,
+      "grad_norm": 0.8074361085891724,
+      "learning_rate": 4.809412590445566e-06,
+      "loss": 0.5888,
+      "step": 757
+    },
+    {
+      "epoch": 0.8491411501120238,
+      "grad_norm": 0.8198902010917664,
+      "learning_rate": 4.808839489060788e-06,
+      "loss": 0.5895,
+      "step": 758
+    },
+    {
+      "epoch": 0.8502613890963405,
+      "grad_norm": 0.8113036751747131,
+      "learning_rate": 4.808265561554293e-06,
+      "loss": 0.5938,
+      "step": 759
+    },
+    {
+      "epoch": 0.8513816280806572,
+      "grad_norm": 0.7734106183052063,
+      "learning_rate": 4.807690808131437e-06,
+      "loss": 0.5981,
+      "step": 760
+    },
+    {
+      "epoch": 0.8525018670649739,
+      "grad_norm": 0.7905157208442688,
+      "learning_rate": 4.80711522899787e-06,
+      "loss": 0.5862,
+      "step": 761
+    },
+    {
+      "epoch": 0.8536221060492906,
+      "grad_norm": 0.7930572628974915,
+      "learning_rate": 4.80653882435954e-06,
+      "loss": 0.6078,
+      "step": 762
+    },
+    {
+      "epoch": 0.8547423450336071,
+      "grad_norm": 0.7793716788291931,
+      "learning_rate": 4.8059615944226885e-06,
+      "loss": 0.6005,
+      "step": 763
+    },
+    {
+      "epoch": 0.8558625840179238,
+      "grad_norm": 0.7978621125221252,
+      "learning_rate": 4.8053835393938545e-06,
+      "loss": 0.5989,
+      "step": 764
+    },
+    {
+      "epoch": 0.8569828230022405,
+      "grad_norm": 0.8089454770088196,
+      "learning_rate": 4.804804659479869e-06,
+      "loss": 0.5576,
+      "step": 765
+    },
+    {
+      "epoch": 0.8581030619865572,
+      "grad_norm": 0.7966460585594177,
+      "learning_rate": 4.804224954887862e-06,
+      "loss": 0.5777,
+      "step": 766
+    },
+    {
+      "epoch": 0.8592233009708737,
+      "grad_norm": 0.8103785514831543,
+      "learning_rate": 4.8036444258252555e-06,
+      "loss": 0.5965,
+      "step": 767
+    },
+    {
+      "epoch": 0.8603435399551904,
+      "grad_norm": 0.8107985854148865,
+      "learning_rate": 4.803063072499768e-06,
+      "loss": 0.5909,
+      "step": 768
+    },
+    {
+      "epoch": 0.8614637789395071,
+      "grad_norm": 0.8095008730888367,
+      "learning_rate": 4.8024808951194105e-06,
+      "loss": 0.565,
+      "step": 769
+    },
+    {
+      "epoch": 0.8625840179238238,
+      "grad_norm": 0.8257828950881958,
+      "learning_rate": 4.801897893892493e-06,
+      "loss": 0.5825,
+      "step": 770
+    },
+    {
+      "epoch": 0.8637042569081405,
+      "grad_norm": 0.8019899725914001,
+      "learning_rate": 4.8013140690276186e-06,
+      "loss": 0.5806,
+      "step": 771
+    },
+    {
+      "epoch": 0.864824495892457,
+      "grad_norm": 0.809320330619812,
+      "learning_rate": 4.800729420733683e-06,
+      "loss": 0.5778,
+      "step": 772
+    },
+    {
+      "epoch": 0.8659447348767737,
+      "grad_norm": 0.8071535229682922,
+      "learning_rate": 4.800143949219879e-06,
+      "loss": 0.5702,
+      "step": 773
+    },
+    {
+      "epoch": 0.8670649738610904,
+      "grad_norm": 0.8018308281898499,
+      "learning_rate": 4.799557654695692e-06,
+      "loss": 0.6004,
+      "step": 774
+    },
+    {
+      "epoch": 0.8681852128454071,
+      "grad_norm": 0.8141055107116699,
+      "learning_rate": 4.798970537370904e-06,
+      "loss": 0.57,
+      "step": 775
+    },
+    {
+      "epoch": 0.8693054518297236,
+      "grad_norm": 0.7862063050270081,
+      "learning_rate": 4.798382597455591e-06,
+      "loss": 0.5528,
+      "step": 776
+    },
+    {
+      "epoch": 0.8704256908140403,
+      "grad_norm": 0.8331758975982666,
+      "learning_rate": 4.797793835160122e-06,
+      "loss": 0.6008,
+      "step": 777
+    },
+    {
+      "epoch": 0.871545929798357,
+      "grad_norm": 0.823685884475708,
+      "learning_rate": 4.797204250695161e-06,
+      "loss": 0.6154,
+      "step": 778
+    },
+    {
+      "epoch": 0.8726661687826737,
+      "grad_norm": 0.7857784032821655,
+      "learning_rate": 4.796613844271665e-06,
+      "loss": 0.5994,
+      "step": 779
+    },
+    {
+      "epoch": 0.8737864077669902,
+      "grad_norm": 0.8060506582260132,
+      "learning_rate": 4.796022616100889e-06,
+      "loss": 0.5633,
+      "step": 780
+    },
+    {
+      "epoch": 0.8749066467513069,
+      "grad_norm": 0.8090853095054626,
+      "learning_rate": 4.795430566394377e-06,
+      "loss": 0.5573,
+      "step": 781
+    },
+    {
+      "epoch": 0.8760268857356236,
+      "grad_norm": 0.8793733716011047,
+      "learning_rate": 4.7948376953639695e-06,
+      "loss": 0.619,
+      "step": 782
+    },
+    {
+      "epoch": 0.8771471247199403,
+      "grad_norm": 0.8061932921409607,
+      "learning_rate": 4.794244003221801e-06,
+      "loss": 0.6047,
+      "step": 783
+    },
+    {
+      "epoch": 0.878267363704257,
+      "grad_norm": 0.8307778239250183,
+      "learning_rate": 4.7936494901803e-06,
+      "loss": 0.5825,
+      "step": 784
+    },
+    {
+      "epoch": 0.8793876026885735,
+      "grad_norm": 0.8569943308830261,
+      "learning_rate": 4.793054156452187e-06,
+      "loss": 0.5754,
+      "step": 785
+    },
+    {
+      "epoch": 0.8805078416728902,
+      "grad_norm": 0.793869137763977,
+      "learning_rate": 4.792458002250479e-06,
+      "loss": 0.5864,
+      "step": 786
+    },
+    {
+      "epoch": 0.8816280806572069,
+      "grad_norm": 0.796796977519989,
+      "learning_rate": 4.791861027788483e-06,
+      "loss": 0.5988,
+      "step": 787
+    },
+    {
+      "epoch": 0.8827483196415236,
+      "grad_norm": 0.8137564063072205,
+      "learning_rate": 4.791263233279802e-06,
+      "loss": 0.5768,
+      "step": 788
+    },
+    {
+      "epoch": 0.8838685586258401,
+      "grad_norm": 0.7946325540542603,
+      "learning_rate": 4.790664618938332e-06,
+      "loss": 0.5674,
+      "step": 789
+    },
+    {
+      "epoch": 0.8849887976101568,
+      "grad_norm": 0.8001543283462524,
+      "learning_rate": 4.790065184978263e-06,
+      "loss": 0.5705,
+      "step": 790
+    },
+    {
+      "epoch": 0.8861090365944735,
+      "grad_norm": 0.8289190530776978,
+      "learning_rate": 4.789464931614075e-06,
+      "loss": 0.6023,
+      "step": 791
+    },
+    {
+      "epoch": 0.8872292755787902,
+      "grad_norm": 0.8041342496871948,
+      "learning_rate": 4.788863859060546e-06,
+      "loss": 0.5963,
+      "step": 792
+    },
+    {
+      "epoch": 0.8883495145631068,
+      "grad_norm": 0.8325103521347046,
+      "learning_rate": 4.788261967532743e-06,
+      "loss": 0.6005,
+      "step": 793
+    },
+    {
+      "epoch": 0.8894697535474234,
+      "grad_norm": 0.7777535319328308,
+      "learning_rate": 4.787659257246029e-06,
+      "loss": 0.5701,
+      "step": 794
+    },
+    {
+      "epoch": 0.8905899925317401,
+      "grad_norm": 0.8024612069129944,
+      "learning_rate": 4.787055728416058e-06,
+      "loss": 0.593,
+      "step": 795
+    },
+    {
+      "epoch": 0.8917102315160568,
+      "grad_norm": 0.7804939150810242,
+      "learning_rate": 4.786451381258777e-06,
+      "loss": 0.5547,
+      "step": 796
+    },
+    {
+      "epoch": 0.8928304705003735,
+      "grad_norm": 0.8120240569114685,
+      "learning_rate": 4.7858462159904284e-06,
+      "loss": 0.5955,
+      "step": 797
+    },
+    {
+      "epoch": 0.89395070948469,
+      "grad_norm": 0.7861335873603821,
+      "learning_rate": 4.785240232827543e-06,
+      "loss": 0.5769,
+      "step": 798
+    },
+    {
+      "epoch": 0.8950709484690067,
+      "grad_norm": 0.7677375674247742,
+      "learning_rate": 4.784633431986948e-06,
+      "loss": 0.572,
+      "step": 799
+    },
+    {
+      "epoch": 0.8961911874533234,
+      "grad_norm": 0.8227370381355286,
+      "learning_rate": 4.784025813685761e-06,
+      "loss": 0.5968,
+      "step": 800
+    },
+    {
+      "epoch": 0.8973114264376401,
+      "grad_norm": 0.8148797750473022,
+      "learning_rate": 4.783417378141393e-06,
+      "loss": 0.587,
+      "step": 801
+    },
+    {
+      "epoch": 0.8984316654219567,
+      "grad_norm": 0.8100507855415344,
+      "learning_rate": 4.782808125571547e-06,
+      "loss": 0.581,
+      "step": 802
+    },
+    {
+      "epoch": 0.8995519044062733,
+      "grad_norm": 0.8154464960098267,
+      "learning_rate": 4.782198056194219e-06,
+      "loss": 0.5932,
+      "step": 803
+    },
+    {
+      "epoch": 0.90067214339059,
+      "grad_norm": 0.7960717678070068,
+      "learning_rate": 4.781587170227698e-06,
+      "loss": 0.568,
+      "step": 804
+    },
+    {
+      "epoch": 0.9017923823749067,
+      "grad_norm": 0.7840545177459717,
+      "learning_rate": 4.780975467890561e-06,
+      "loss": 0.5709,
+      "step": 805
+    },
+    {
+      "epoch": 0.9029126213592233,
+      "grad_norm": 0.7957537174224854,
+      "learning_rate": 4.780362949401683e-06,
+      "loss": 0.5946,
+      "step": 806
+    },
+    {
+      "epoch": 0.9040328603435399,
+      "grad_norm": 0.8008480668067932,
+      "learning_rate": 4.779749614980225e-06,
+      "loss": 0.6082,
+      "step": 807
+    },
+    {
+      "epoch": 0.9051530993278566,
+      "grad_norm": 0.8121342062950134,
+      "learning_rate": 4.779135464845647e-06,
+      "loss": 0.5799,
+      "step": 808
+    },
+    {
+      "epoch": 0.9062733383121733,
+      "grad_norm": 0.7985235452651978,
+      "learning_rate": 4.778520499217694e-06,
+      "loss": 0.6025,
+      "step": 809
+    },
+    {
+      "epoch": 0.90739357729649,
+      "grad_norm": 0.7896565198898315,
+      "learning_rate": 4.7779047183164075e-06,
+      "loss": 0.5954,
+      "step": 810
+    },
+    {
+      "epoch": 0.9085138162808065,
+      "grad_norm": 0.807032585144043,
+      "learning_rate": 4.7772881223621175e-06,
+      "loss": 0.6022,
+      "step": 811
+    },
+    {
+      "epoch": 0.9096340552651232,
+      "grad_norm": 0.8102108836174011,
+      "learning_rate": 4.7766707115754485e-06,
+      "loss": 0.5775,
+      "step": 812
+    },
+    {
+      "epoch": 0.9107542942494399,
+      "grad_norm": 0.8097600340843201,
+      "learning_rate": 4.776052486177314e-06,
+      "loss": 0.5877,
+      "step": 813
+    },
+    {
+      "epoch": 0.9118745332337566,
+      "grad_norm": 0.7986764907836914,
+      "learning_rate": 4.775433446388921e-06,
+      "loss": 0.5625,
+      "step": 814
+    },
+    {
+      "epoch": 0.9129947722180732,
+      "grad_norm": 0.8128889203071594,
+      "learning_rate": 4.7748135924317655e-06,
+      "loss": 0.6031,
+      "step": 815
+    },
+    {
+      "epoch": 0.9141150112023898,
+      "grad_norm": 0.7872529029846191,
+      "learning_rate": 4.774192924527638e-06,
+      "loss": 0.5624,
+      "step": 816
+    },
+    {
+      "epoch": 0.9152352501867065,
+      "grad_norm": 0.8138707280158997,
+      "learning_rate": 4.773571442898618e-06,
+      "loss": 0.5848,
+      "step": 817
+    },
+    {
+      "epoch": 0.9163554891710232,
+      "grad_norm": 0.8301820158958435,
+      "learning_rate": 4.772949147767077e-06,
+      "loss": 0.6031,
+      "step": 818
+    },
+    {
+      "epoch": 0.9174757281553398,
+      "grad_norm": 0.8014585971832275,
+      "learning_rate": 4.772326039355677e-06,
+      "loss": 0.5575,
+      "step": 819
+    },
+    {
+      "epoch": 0.9185959671396564,
+      "grad_norm": 0.8259865641593933,
+      "learning_rate": 4.771702117887371e-06,
+      "loss": 0.596,
+      "step": 820
+    },
+    {
+      "epoch": 0.9197162061239731,
+      "grad_norm": 0.8028122782707214,
+      "learning_rate": 4.771077383585404e-06,
+      "loss": 0.5851,
+      "step": 821
+    },
+    {
+      "epoch": 0.9208364451082898,
+      "grad_norm": 0.7960208058357239,
+      "learning_rate": 4.7704518366733095e-06,
+      "loss": 0.5999,
+      "step": 822
+    },
+    {
+      "epoch": 0.9219566840926064,
+      "grad_norm": 0.797467827796936,
+      "learning_rate": 4.769825477374915e-06,
+      "loss": 0.5809,
+      "step": 823
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.8268144130706787,
+      "learning_rate": 4.7691983059143374e-06,
+      "loss": 0.5821,
+      "step": 824
+    },
+    {
+      "epoch": 0.9241971620612397,
+      "grad_norm": 0.803213357925415,
+      "learning_rate": 4.768570322515982e-06,
+      "loss": 0.5606,
+      "step": 825
+    },
+    {
+      "epoch": 0.9253174010455564,
+      "grad_norm": 0.8422443866729736,
+      "learning_rate": 4.767941527404547e-06,
+      "loss": 0.5813,
+      "step": 826
+    },
+    {
+      "epoch": 0.926437640029873,
+      "grad_norm": 0.7983731031417847,
+      "learning_rate": 4.767311920805021e-06,
+      "loss": 0.5747,
+      "step": 827
+    },
+    {
+      "epoch": 0.9275578790141897,
+      "grad_norm": 0.827704906463623,
+      "learning_rate": 4.766681502942681e-06,
+      "loss": 0.5587,
+      "step": 828
+    },
+    {
+      "epoch": 0.9286781179985063,
+      "grad_norm": 0.832988977432251,
+      "learning_rate": 4.7660502740430985e-06,
+      "loss": 0.572,
+      "step": 829
+    },
+    {
+      "epoch": 0.929798356982823,
+      "grad_norm": 0.8249674439430237,
+      "learning_rate": 4.76541823433213e-06,
+      "loss": 0.5801,
+      "step": 830
+    },
+    {
+      "epoch": 0.9309185959671397,
+      "grad_norm": 0.8318766355514526,
+      "learning_rate": 4.764785384035925e-06,
+      "loss": 0.5793,
+      "step": 831
+    },
+    {
+      "epoch": 0.9320388349514563,
+      "grad_norm": 0.8395798206329346,
+      "learning_rate": 4.764151723380923e-06,
+      "loss": 0.587,
+      "step": 832
+    },
+    {
+      "epoch": 0.933159073935773,
+      "grad_norm": 0.8279564380645752,
+      "learning_rate": 4.763517252593852e-06,
+      "loss": 0.597,
+      "step": 833
+    },
+    {
+      "epoch": 0.9342793129200896,
+      "grad_norm": 0.856611967086792,
+      "learning_rate": 4.762881971901732e-06,
+      "loss": 0.5488,
+      "step": 834
+    },
+    {
+      "epoch": 0.9353995519044063,
+      "grad_norm": 0.8096729516983032,
+      "learning_rate": 4.762245881531872e-06,
+      "loss": 0.577,
+      "step": 835
+    },
+    {
+      "epoch": 0.936519790888723,
+      "grad_norm": 0.8170201778411865,
+      "learning_rate": 4.761608981711868e-06,
+      "loss": 0.5602,
+      "step": 836
+    },
+    {
+      "epoch": 0.9376400298730396,
+      "grad_norm": 0.8169147968292236,
+      "learning_rate": 4.76097127266961e-06,
+      "loss": 0.5974,
+      "step": 837
+    },
+    {
+      "epoch": 0.9387602688573562,
+      "grad_norm": 0.806937038898468,
+      "learning_rate": 4.760332754633276e-06,
+      "loss": 0.5795,
+      "step": 838
+    },
+    {
+      "epoch": 0.9398805078416729,
+      "grad_norm": 0.8193455934524536,
+      "learning_rate": 4.759693427831331e-06,
+      "loss": 0.5886,
+      "step": 839
+    },
+    {
+      "epoch": 0.9410007468259896,
+      "grad_norm": 0.8178059458732605,
+      "learning_rate": 4.759053292492532e-06,
+      "loss": 0.5951,
+      "step": 840
+    },
+    {
+      "epoch": 0.9421209858103062,
+      "grad_norm": 0.8102588057518005,
+      "learning_rate": 4.758412348845925e-06,
+      "loss": 0.5724,
+      "step": 841
+    },
+    {
+      "epoch": 0.9432412247946228,
+      "grad_norm": 0.8178951144218445,
+      "learning_rate": 4.7577705971208466e-06,
+      "loss": 0.5849,
+      "step": 842
+    },
+    {
+      "epoch": 0.9443614637789395,
+      "grad_norm": 0.8069207668304443,
+      "learning_rate": 4.757128037546918e-06,
+      "loss": 0.5978,
+      "step": 843
+    },
+    {
+      "epoch": 0.9454817027632562,
+      "grad_norm": 0.8033669590950012,
+      "learning_rate": 4.756484670354054e-06,
+      "loss": 0.5794,
+      "step": 844
+    },
+    {
+      "epoch": 0.9466019417475728,
+      "grad_norm": 0.8094858527183533,
+      "learning_rate": 4.755840495772455e-06,
+      "loss": 0.6082,
+      "step": 845
+    },
+    {
+      "epoch": 0.9477221807318895,
+      "grad_norm": 0.8048908114433289,
+      "learning_rate": 4.755195514032614e-06,
+      "loss": 0.5644,
+      "step": 846
+    },
+    {
+      "epoch": 0.9488424197162061,
+      "grad_norm": 0.8288162350654602,
+      "learning_rate": 4.754549725365311e-06,
+      "loss": 0.5665,
+      "step": 847
+    },
+    {
+      "epoch": 0.9499626587005228,
+      "grad_norm": 0.8059208989143372,
+      "learning_rate": 4.753903130001612e-06,
+      "loss": 0.6111,
+      "step": 848
+    },
+    {
+      "epoch": 0.9510828976848394,
+      "grad_norm": 0.7926841974258423,
+      "learning_rate": 4.753255728172875e-06,
+      "loss": 0.5632,
+      "step": 849
+    },
+    {
+      "epoch": 0.9522031366691561,
+      "grad_norm": 0.8072546720504761,
+      "learning_rate": 4.752607520110748e-06,
+      "loss": 0.5953,
+      "step": 850
+    },
+    {
+      "epoch": 0.9533233756534727,
+      "grad_norm": 0.8185967803001404,
+      "learning_rate": 4.7519585060471616e-06,
+      "loss": 0.6158,
+      "step": 851
+    },
+    {
+      "epoch": 0.9544436146377894,
+      "grad_norm": 0.8182018399238586,
+      "learning_rate": 4.751308686214341e-06,
+      "loss": 0.586,
+      "step": 852
+    },
+    {
+      "epoch": 0.955563853622106,
+      "grad_norm": 0.8257445693016052,
+      "learning_rate": 4.750658060844796e-06,
+      "loss": 0.5818,
+      "step": 853
+    },
+    {
+      "epoch": 0.9566840926064227,
+      "grad_norm": 0.7723025679588318,
+      "learning_rate": 4.750006630171325e-06,
+      "loss": 0.5754,
+      "step": 854
+    },
+    {
+      "epoch": 0.9578043315907394,
+      "grad_norm": 0.7873339653015137,
+      "learning_rate": 4.7493543944270174e-06,
+      "loss": 0.5641,
+      "step": 855
+    },
+    {
+      "epoch": 0.958924570575056,
+      "grad_norm": 0.8600917458534241,
+      "learning_rate": 4.748701353845247e-06,
+      "loss": 0.5993,
+      "step": 856
+    },
+    {
+      "epoch": 0.9600448095593727,
+      "grad_norm": 0.7842808365821838,
+      "learning_rate": 4.748047508659677e-06,
+      "loss": 0.5928,
+      "step": 857
+    },
+    {
+      "epoch": 0.9611650485436893,
+      "grad_norm": 0.7855350375175476,
+      "learning_rate": 4.747392859104258e-06,
+      "loss": 0.5814,
+      "step": 858
+    },
+    {
+      "epoch": 0.962285287528006,
+      "grad_norm": 0.8112151026725769,
+      "learning_rate": 4.746737405413231e-06,
+      "loss": 0.5753,
+      "step": 859
+    },
+    {
+      "epoch": 0.9634055265123226,
+      "grad_norm": 0.8161853551864624,
+      "learning_rate": 4.746081147821121e-06,
+      "loss": 0.5853,
+      "step": 860
+    },
+    {
+      "epoch": 0.9645257654966393,
+      "grad_norm": 0.7908380627632141,
+      "learning_rate": 4.745424086562743e-06,
+      "loss": 0.5854,
+      "step": 861
+    },
+    {
+      "epoch": 0.9656460044809559,
+      "grad_norm": 0.8022582530975342,
+      "learning_rate": 4.7447662218731995e-06,
+      "loss": 0.5602,
+      "step": 862
+    },
+    {
+      "epoch": 0.9667662434652726,
+      "grad_norm": 0.8379324674606323,
+      "learning_rate": 4.744107553987878e-06,
+      "loss": 0.5921,
+      "step": 863
+    },
+    {
+      "epoch": 0.9678864824495892,
+      "grad_norm": 0.8459264636039734,
+      "learning_rate": 4.743448083142458e-06,
+      "loss": 0.5984,
+      "step": 864
+    },
+    {
+      "epoch": 0.9690067214339059,
+      "grad_norm": 0.8297575116157532,
+      "learning_rate": 4.742787809572901e-06,
+      "loss": 0.5897,
+      "step": 865
+    },
+    {
+      "epoch": 0.9701269604182226,
+      "grad_norm": 0.8278453946113586,
+      "learning_rate": 4.742126733515459e-06,
+      "loss": 0.597,
+      "step": 866
+    },
+    {
+      "epoch": 0.9712471994025392,
+      "grad_norm": 0.8626736402511597,
+      "learning_rate": 4.741464855206671e-06,
+      "loss": 0.6084,
+      "step": 867
+    },
+    {
+      "epoch": 0.9723674383868559,
+      "grad_norm": 0.8282763957977295,
+      "learning_rate": 4.740802174883363e-06,
+      "loss": 0.5668,
+      "step": 868
+    },
+    {
+      "epoch": 0.9734876773711725,
+      "grad_norm": 0.8218764066696167,
+      "learning_rate": 4.740138692782646e-06,
+      "loss": 0.5937,
+      "step": 869
+    },
+    {
+      "epoch": 0.9746079163554892,
+      "grad_norm": 0.7809569835662842,
+      "learning_rate": 4.73947440914192e-06,
+      "loss": 0.5895,
+      "step": 870
+    },
+    {
+      "epoch": 0.9757281553398058,
+      "grad_norm": 0.8076401948928833,
+      "learning_rate": 4.738809324198872e-06,
+      "loss": 0.5757,
+      "step": 871
+    },
+    {
+      "epoch": 0.9768483943241225,
+      "grad_norm": 0.8134318590164185,
+      "learning_rate": 4.738143438191473e-06,
+      "loss": 0.5326,
+      "step": 872
+    },
+    {
+      "epoch": 0.9779686333084391,
+      "grad_norm": 0.7635218501091003,
+      "learning_rate": 4.737476751357985e-06,
+      "loss": 0.5476,
+      "step": 873
+    },
+    {
+      "epoch": 0.9790888722927558,
+      "grad_norm": 0.8077232241630554,
+      "learning_rate": 4.736809263936952e-06,
+      "loss": 0.5722,
+      "step": 874
+    },
+    {
+      "epoch": 0.9802091112770724,
+      "grad_norm": 0.8017358183860779,
+      "learning_rate": 4.736140976167206e-06,
+      "loss": 0.5907,
+      "step": 875
+    },
+    {
+      "epoch": 0.9813293502613891,
+      "grad_norm": 0.8405149579048157,
+      "learning_rate": 4.735471888287868e-06,
+      "loss": 0.6333,
+      "step": 876
+    },
+    {
+      "epoch": 0.9824495892457058,
+      "grad_norm": 0.7879468202590942,
+      "learning_rate": 4.734802000538342e-06,
+      "loss": 0.5496,
+      "step": 877
+    },
+    {
+      "epoch": 0.9835698282300224,
+      "grad_norm": 0.8601053357124329,
+      "learning_rate": 4.734131313158319e-06,
+      "loss": 0.5751,
+      "step": 878
+    },
+    {
+      "epoch": 0.984690067214339,
+      "grad_norm": 0.8143794536590576,
+      "learning_rate": 4.733459826387777e-06,
+      "loss": 0.583,
+      "step": 879
+    },
+    {
+      "epoch": 0.9858103061986557,
+      "grad_norm": 0.8531054258346558,
+      "learning_rate": 4.732787540466979e-06,
+      "loss": 0.6068,
+      "step": 880
+    },
+    {
+      "epoch": 0.9869305451829724,
+      "grad_norm": 0.8039345145225525,
+      "learning_rate": 4.732114455636475e-06,
+      "loss": 0.5873,
+      "step": 881
+    },
+    {
+      "epoch": 0.988050784167289,
+      "grad_norm": 0.8260403275489807,
+      "learning_rate": 4.7314405721371e-06,
+      "loss": 0.5922,
+      "step": 882
+    },
+    {
+      "epoch": 0.9891710231516057,
+      "grad_norm": 0.8372035622596741,
+      "learning_rate": 4.7307658902099755e-06,
+      "loss": 0.5862,
+      "step": 883
+    },
+    {
+      "epoch": 0.9902912621359223,
+      "grad_norm": 0.8514466881752014,
+      "learning_rate": 4.730090410096507e-06,
+      "loss": 0.5981,
+      "step": 884
+    },
+    {
+      "epoch": 0.991411501120239,
+      "grad_norm": 0.7753927111625671,
+      "learning_rate": 4.7294141320383875e-06,
+      "loss": 0.6009,
+      "step": 885
+    },
+    {
+      "epoch": 0.9925317401045557,
+      "grad_norm": 0.8559523820877075,
+      "learning_rate": 4.7287370562775955e-06,
+      "loss": 0.5903,
+      "step": 886
+    },
+    {
+      "epoch": 0.9936519790888723,
+      "grad_norm": 0.7929627895355225,
+      "learning_rate": 4.728059183056394e-06,
+      "loss": 0.5727,
+      "step": 887
+    },
+    {
+      "epoch": 0.9947722180731889,
+      "grad_norm": 0.8660935759544373,
+      "learning_rate": 4.72738051261733e-06,
+      "loss": 0.5701,
+      "step": 888
+    },
+    {
+      "epoch": 0.9958924570575056,
+      "grad_norm": 0.8257469534873962,
+      "learning_rate": 4.72670104520324e-06,
+      "loss": 0.5592,
+      "step": 889
+    },
+    {
+      "epoch": 0.9970126960418223,
+      "grad_norm": 0.8579026460647583,
+      "learning_rate": 4.726020781057241e-06,
+      "loss": 0.5888,
+      "step": 890
+    },
+    {
+      "epoch": 0.9981329350261389,
+      "grad_norm": 0.791415810585022,
+      "learning_rate": 4.7253397204227385e-06,
+      "loss": 0.5572,
+      "step": 891
+    },
+    {
+      "epoch": 0.9992531740104555,
+      "grad_norm": 0.8314563035964966,
+      "learning_rate": 4.724657863543419e-06,
+      "loss": 0.5832,
+      "step": 892
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8314563035964966,
+      "learning_rate": 4.72397521066326e-06,
+      "loss": 0.6103,
+      "step": 893
+    },
+    {
+      "epoch": 1.0011202389843166,
+      "grad_norm": 1.2335702180862427,
+      "learning_rate": 4.723291762026519e-06,
+      "loss": 0.5434,
+      "step": 894
+    },
+    {
+      "epoch": 1.0022404779686334,
+      "grad_norm": 0.8101676106452942,
+      "learning_rate": 4.722607517877737e-06,
+      "loss": 0.5393,
+      "step": 895
+    },
+    {
+      "epoch": 1.00336071695295,
+      "grad_norm": 0.7916525602340698,
+      "learning_rate": 4.7219224784617465e-06,
+      "loss": 0.5638,
+      "step": 896
+    },
+    {
+      "epoch": 1.0044809559372667,
+      "grad_norm": 0.8107758164405823,
+      "learning_rate": 4.721236644023657e-06,
+      "loss": 0.5773,
+      "step": 897
+    },
+    {
+      "epoch": 1.0056011949215833,
+      "grad_norm": 0.8929505348205566,
+      "learning_rate": 4.720550014808867e-06,
+      "loss": 0.5105,
+      "step": 898
+    },
+    {
+      "epoch": 1.0067214339058999,
+      "grad_norm": 0.8275009989738464,
+      "learning_rate": 4.719862591063058e-06,
+      "loss": 0.5629,
+      "step": 899
+    },
+    {
+      "epoch": 1.0078416728902166,
+      "grad_norm": 0.8205452561378479,
+      "learning_rate": 4.7191743730321955e-06,
+      "loss": 0.5458,
+      "step": 900
+    },
+    {
+      "epoch": 1.0089619118745332,
+      "grad_norm": 0.8630054593086243,
+      "learning_rate": 4.718485360962529e-06,
+      "loss": 0.562,
+      "step": 901
+    },
+    {
+      "epoch": 1.0100821508588498,
+      "grad_norm": 0.839333176612854,
+      "learning_rate": 4.717795555100595e-06,
+      "loss": 0.5631,
+      "step": 902
+    },
+    {
+      "epoch": 1.0112023898431666,
+      "grad_norm": 0.7813014388084412,
+      "learning_rate": 4.717104955693209e-06,
+      "loss": 0.5369,
+      "step": 903
+    },
+    {
+      "epoch": 1.0123226288274831,
+      "grad_norm": 0.7793049216270447,
+      "learning_rate": 4.716413562987475e-06,
+      "loss": 0.5637,
+      "step": 904
+    },
+    {
+      "epoch": 1.0134428678118,
+      "grad_norm": 0.8571561574935913,
+      "learning_rate": 4.715721377230778e-06,
+      "loss": 0.5722,
+      "step": 905
+    },
+    {
+      "epoch": 1.0145631067961165,
+      "grad_norm": 0.8131767511367798,
+      "learning_rate": 4.715028398670787e-06,
+      "loss": 0.5608,
+      "step": 906
+    },
+    {
+      "epoch": 1.015683345780433,
+      "grad_norm": 0.8168996572494507,
+      "learning_rate": 4.714334627555458e-06,
+      "loss": 0.5529,
+      "step": 907
+    },
+    {
+      "epoch": 1.0168035847647499,
+      "grad_norm": 0.823046088218689,
+      "learning_rate": 4.7136400641330245e-06,
+      "loss": 0.5682,
+      "step": 908
+    },
+    {
+      "epoch": 1.0179238237490664,
+      "grad_norm": 0.8340138792991638,
+      "learning_rate": 4.712944708652011e-06,
+      "loss": 0.5731,
+      "step": 909
+    },
+    {
+      "epoch": 1.0190440627333832,
+      "grad_norm": 0.8534042835235596,
+      "learning_rate": 4.712248561361218e-06,
+      "loss": 0.5679,
+      "step": 910
+    },
+    {
+      "epoch": 1.0201643017176998,
+      "grad_norm": 0.8459539413452148,
+      "learning_rate": 4.711551622509734e-06,
+      "loss": 0.5607,
+      "step": 911
+    },
+    {
+      "epoch": 1.0212845407020164,
+      "grad_norm": 0.8298152685165405,
+      "learning_rate": 4.7108538923469296e-06,
+      "loss": 0.5806,
+      "step": 912
+    },
+    {
+      "epoch": 1.0224047796863331,
+      "grad_norm": 0.8335070610046387,
+      "learning_rate": 4.710155371122458e-06,
+      "loss": 0.5285,
+      "step": 913
+    },
+    {
+      "epoch": 1.0235250186706497,
+      "grad_norm": 0.7781080603599548,
+      "learning_rate": 4.709456059086257e-06,
+      "loss": 0.5824,
+      "step": 914
+    },
+    {
+      "epoch": 1.0246452576549665,
+      "grad_norm": 0.8258634805679321,
+      "learning_rate": 4.7087559564885445e-06,
+      "loss": 0.5431,
+      "step": 915
+    },
+    {
+      "epoch": 1.025765496639283,
+      "grad_norm": 0.8281342387199402,
+      "learning_rate": 4.708055063579824e-06,
+      "loss": 0.5583,
+      "step": 916
+    },
+    {
+      "epoch": 1.0268857356235996,
+      "grad_norm": 0.7836266160011292,
+      "learning_rate": 4.707353380610879e-06,
+      "loss": 0.5636,
+      "step": 917
+    },
+    {
+      "epoch": 1.0280059746079164,
+      "grad_norm": 0.834606409072876,
+      "learning_rate": 4.706650907832781e-06,
+      "loss": 0.5432,
+      "step": 918
+    },
+    {
+      "epoch": 1.029126213592233,
+      "grad_norm": 0.8574039340019226,
+      "learning_rate": 4.705947645496877e-06,
+      "loss": 0.5452,
+      "step": 919
+    },
+    {
+      "epoch": 1.0302464525765496,
+      "grad_norm": 0.7969507575035095,
+      "learning_rate": 4.705243593854802e-06,
+      "loss": 0.5357,
+      "step": 920
+    },
+    {
+      "epoch": 1.0313666915608664,
+      "grad_norm": 0.8153135180473328,
+      "learning_rate": 4.7045387531584704e-06,
+      "loss": 0.5316,
+      "step": 921
+    },
+    {
+      "epoch": 1.032486930545183,
+      "grad_norm": 0.7883548736572266,
+      "learning_rate": 4.703833123660081e-06,
+      "loss": 0.5232,
+      "step": 922
+    },
+    {
+      "epoch": 1.0336071695294997,
+      "grad_norm": 0.8280951380729675,
+      "learning_rate": 4.703126705612113e-06,
+      "loss": 0.5693,
+      "step": 923
+    },
+    {
+      "epoch": 1.0347274085138163,
+      "grad_norm": 0.8673230409622192,
+      "learning_rate": 4.702419499267329e-06,
+      "loss": 0.5508,
+      "step": 924
+    },
+    {
+      "epoch": 1.0358476474981329,
+      "grad_norm": 0.8065434694290161,
+      "learning_rate": 4.701711504878773e-06,
+      "loss": 0.5389,
+      "step": 925
+    },
+    {
+      "epoch": 1.0369678864824496,
+      "grad_norm": 0.797985851764679,
+      "learning_rate": 4.701002722699771e-06,
+      "loss": 0.5393,
+      "step": 926
+    },
+    {
+      "epoch": 1.0380881254667662,
+      "grad_norm": 0.8205111026763916,
+      "learning_rate": 4.7002931529839315e-06,
+      "loss": 0.5602,
+      "step": 927
+    },
+    {
+      "epoch": 1.039208364451083,
+      "grad_norm": 0.8055931329727173,
+      "learning_rate": 4.699582795985144e-06,
+      "loss": 0.5447,
+      "step": 928
+    },
+    {
+      "epoch": 1.0403286034353996,
+      "grad_norm": 0.8045268058776855,
+      "learning_rate": 4.698871651957581e-06,
+      "loss": 0.5299,
+      "step": 929
+    },
+    {
+      "epoch": 1.0414488424197161,
+      "grad_norm": 0.7948052883148193,
+      "learning_rate": 4.698159721155694e-06,
+      "loss": 0.5388,
+      "step": 930
+    },
+    {
+      "epoch": 1.042569081404033,
+      "grad_norm": 0.8069709539413452,
+      "learning_rate": 4.697447003834219e-06,
+      "loss": 0.5443,
+      "step": 931
+    },
+    {
+      "epoch": 1.0436893203883495,
+      "grad_norm": 0.8315683007240295,
+      "learning_rate": 4.6967335002481724e-06,
+      "loss": 0.563,
+      "step": 932
+    },
+    {
+      "epoch": 1.044809559372666,
+      "grad_norm": 0.8262251615524292,
+      "learning_rate": 4.69601921065285e-06,
+      "loss": 0.5802,
+      "step": 933
+    },
+    {
+      "epoch": 1.0459297983569829,
+      "grad_norm": 0.8720364570617676,
+      "learning_rate": 4.6953041353038324e-06,
+      "loss": 0.5285,
+      "step": 934
+    },
+    {
+      "epoch": 1.0470500373412994,
+      "grad_norm": 0.7995201349258423,
+      "learning_rate": 4.694588274456978e-06,
+      "loss": 0.5551,
+      "step": 935
+    },
+    {
+      "epoch": 1.0481702763256162,
+      "grad_norm": 0.8372474312782288,
+      "learning_rate": 4.693871628368428e-06,
+      "loss": 0.5293,
+      "step": 936
+    },
+    {
+      "epoch": 1.0492905153099328,
+      "grad_norm": 0.854775071144104,
+      "learning_rate": 4.693154197294606e-06,
+      "loss": 0.5472,
+      "step": 937
+    },
+    {
+      "epoch": 1.0504107542942493,
+      "grad_norm": 0.8773542046546936,
+      "learning_rate": 4.692435981492212e-06,
+      "loss": 0.5347,
+      "step": 938
+    },
+    {
+      "epoch": 1.0515309932785661,
+      "grad_norm": 0.7934802174568176,
+      "learning_rate": 4.69171698121823e-06,
+      "loss": 0.5594,
+      "step": 939
+    },
+    {
+      "epoch": 1.0526512322628827,
+      "grad_norm": 0.840215265750885,
+      "learning_rate": 4.690997196729925e-06,
+      "loss": 0.5637,
+      "step": 940
+    },
+    {
+      "epoch": 1.0537714712471995,
+      "grad_norm": 0.813681423664093,
+      "learning_rate": 4.690276628284842e-06,
+      "loss": 0.5453,
+      "step": 941
+    },
+    {
+      "epoch": 1.054891710231516,
+      "grad_norm": 0.833226203918457,
+      "learning_rate": 4.689555276140807e-06,
+      "loss": 0.5625,
+      "step": 942
+    },
+    {
+      "epoch": 1.0560119492158326,
+      "grad_norm": 0.8451018929481506,
+      "learning_rate": 4.688833140555923e-06,
+      "loss": 0.5377,
+      "step": 943
+    },
+    {
+      "epoch": 1.0571321882001494,
+      "grad_norm": 0.8202646970748901,
+      "learning_rate": 4.688110221788579e-06,
+      "loss": 0.5439,
+      "step": 944
+    },
+    {
+      "epoch": 1.058252427184466,
+      "grad_norm": 0.8503218293190002,
+      "learning_rate": 4.68738652009744e-06,
+      "loss": 0.5693,
+      "step": 945
+    },
+    {
+      "epoch": 1.0593726661687826,
+      "grad_norm": 0.8435008525848389,
+      "learning_rate": 4.686662035741452e-06,
+      "loss": 0.5549,
+      "step": 946
+    },
+    {
+      "epoch": 1.0604929051530994,
+      "grad_norm": 0.8627398014068604,
+      "learning_rate": 4.685936768979842e-06,
+      "loss": 0.5509,
+      "step": 947
+    },
+    {
+      "epoch": 1.061613144137416,
+      "grad_norm": 0.7947062253952026,
+      "learning_rate": 4.685210720072115e-06,
+      "loss": 0.5359,
+      "step": 948
+    },
+    {
+      "epoch": 1.0627333831217327,
+      "grad_norm": 0.7926372289657593,
+      "learning_rate": 4.68448388927806e-06,
+      "loss": 0.5033,
+      "step": 949
+    },
+    {
+      "epoch": 1.0638536221060493,
+      "grad_norm": 0.800025224685669,
+      "learning_rate": 4.683756276857741e-06,
+      "loss": 0.5379,
+      "step": 950
+    },
+    {
+      "epoch": 1.0649738610903658,
+      "grad_norm": 0.8090462684631348,
+      "learning_rate": 4.683027883071505e-06,
+      "loss": 0.5612,
+      "step": 951
+    },
+    {
+      "epoch": 1.0660941000746826,
+      "grad_norm": 0.8113719820976257,
+      "learning_rate": 4.682298708179975e-06,
+      "loss": 0.5743,
+      "step": 952
+    },
+    {
+      "epoch": 1.0672143390589992,
+      "grad_norm": 0.8232184648513794,
+      "learning_rate": 4.681568752444058e-06,
+      "loss": 0.544,
+      "step": 953
+    },
+    {
+      "epoch": 1.068334578043316,
+      "grad_norm": 0.7987185716629028,
+      "learning_rate": 4.680838016124938e-06,
+      "loss": 0.5599,
+      "step": 954
+    },
+    {
+      "epoch": 1.0694548170276326,
+      "grad_norm": 0.8448579907417297,
+      "learning_rate": 4.680106499484077e-06,
+      "loss": 0.5707,
+      "step": 955
+    },
+    {
+      "epoch": 1.0705750560119491,
+      "grad_norm": 0.8265500068664551,
+      "learning_rate": 4.679374202783219e-06,
+      "loss": 0.5315,
+      "step": 956
+    },
+    {
+      "epoch": 1.071695294996266,
+      "grad_norm": 0.8001286387443542,
+      "learning_rate": 4.678641126284385e-06,
+      "loss": 0.5567,
+      "step": 957
+    },
+    {
+      "epoch": 1.0728155339805825,
+      "grad_norm": 0.8409720063209534,
+      "learning_rate": 4.6779072702498765e-06,
+      "loss": 0.5274,
+      "step": 958
+    },
+    {
+      "epoch": 1.0739357729648993,
+      "grad_norm": 0.8332072496414185,
+      "learning_rate": 4.677172634942273e-06,
+      "loss": 0.5326,
+      "step": 959
+    },
+    {
+      "epoch": 1.0750560119492159,
+      "grad_norm": 0.8143296837806702,
+      "learning_rate": 4.676437220624432e-06,
+      "loss": 0.5283,
+      "step": 960
+    },
+    {
+      "epoch": 1.0761762509335324,
+      "grad_norm": 0.8352565765380859,
+      "learning_rate": 4.675701027559492e-06,
+      "loss": 0.5528,
+      "step": 961
+    },
+    {
+      "epoch": 1.0772964899178492,
+      "grad_norm": 0.8069751262664795,
+      "learning_rate": 4.6749640560108685e-06,
+      "loss": 0.5616,
+      "step": 962
+    },
+    {
+      "epoch": 1.0784167289021658,
+      "grad_norm": 0.8762385845184326,
+      "learning_rate": 4.674226306242256e-06,
+      "loss": 0.5461,
+      "step": 963
+    },
+    {
+      "epoch": 1.0795369678864823,
+      "grad_norm": 0.8191564679145813,
+      "learning_rate": 4.673487778517628e-06,
+      "loss": 0.5563,
+      "step": 964
+    },
+    {
+      "epoch": 1.0806572068707991,
+      "grad_norm": 0.8050143122673035,
+      "learning_rate": 4.672748473101234e-06,
+      "loss": 0.5373,
+      "step": 965
+    },
+    {
+      "epoch": 1.0817774458551157,
+      "grad_norm": 0.8051874041557312,
+      "learning_rate": 4.672008390257606e-06,
+      "loss": 0.5327,
+      "step": 966
+    },
+    {
+      "epoch": 1.0828976848394325,
+      "grad_norm": 0.8265339136123657,
+      "learning_rate": 4.6712675302515495e-06,
+      "loss": 0.5418,
+      "step": 967
+    },
+    {
+      "epoch": 1.084017923823749,
+      "grad_norm": 0.8328646421432495,
+      "learning_rate": 4.670525893348152e-06,
+      "loss": 0.5555,
+      "step": 968
+    },
+    {
+      "epoch": 1.0851381628080656,
+      "grad_norm": 0.8175073266029358,
+      "learning_rate": 4.669783479812776e-06,
+      "loss": 0.5664,
+      "step": 969
+    },
+    {
+      "epoch": 1.0862584017923824,
+      "grad_norm": 0.8268328309059143,
+      "learning_rate": 4.669040289911064e-06,
+      "loss": 0.5412,
+      "step": 970
+    },
+    {
+      "epoch": 1.087378640776699,
+      "grad_norm": 0.8998653888702393,
+      "learning_rate": 4.668296323908935e-06,
+      "loss": 0.5581,
+      "step": 971
+    },
+    {
+      "epoch": 1.0884988797610158,
+      "grad_norm": 0.8150838017463684,
+      "learning_rate": 4.6675515820725856e-06,
+      "loss": 0.5404,
+      "step": 972
+    },
+    {
+      "epoch": 1.0896191187453323,
+      "grad_norm": 0.8107709288597107,
+      "learning_rate": 4.666806064668493e-06,
+      "loss": 0.5472,
+      "step": 973
+    },
+    {
+      "epoch": 1.090739357729649,
+      "grad_norm": 0.8273866772651672,
+      "learning_rate": 4.666059771963406e-06,
+      "loss": 0.5672,
+      "step": 974
+    },
+    {
+      "epoch": 1.0918595967139657,
+      "grad_norm": 0.8533821702003479,
+      "learning_rate": 4.6653127042243565e-06,
+      "loss": 0.5573,
+      "step": 975
+    },
+    {
+      "epoch": 1.0929798356982823,
+      "grad_norm": 0.809196412563324,
+      "learning_rate": 4.664564861718651e-06,
+      "loss": 0.5339,
+      "step": 976
+    },
+    {
+      "epoch": 1.0941000746825988,
+      "grad_norm": 0.8037805557250977,
+      "learning_rate": 4.663816244713874e-06,
+      "loss": 0.5553,
+      "step": 977
+    },
+    {
+      "epoch": 1.0952203136669156,
+      "grad_norm": 0.8320299386978149,
+      "learning_rate": 4.663066853477885e-06,
+      "loss": 0.5362,
+      "step": 978
+    },
+    {
+      "epoch": 1.0963405526512322,
+      "grad_norm": 0.8118987679481506,
+      "learning_rate": 4.662316688278824e-06,
+      "loss": 0.5683,
+      "step": 979
+    },
+    {
+      "epoch": 1.097460791635549,
+      "grad_norm": 0.8285667896270752,
+      "learning_rate": 4.6615657493851075e-06,
+      "loss": 0.5496,
+      "step": 980
+    },
+    {
+      "epoch": 1.0985810306198656,
+      "grad_norm": 0.8147162795066833,
+      "learning_rate": 4.660814037065424e-06,
+      "loss": 0.5558,
+      "step": 981
+    },
+    {
+      "epoch": 1.0997012696041821,
+      "grad_norm": 0.8943469524383545,
+      "learning_rate": 4.660061551588746e-06,
+      "loss": 0.5359,
+      "step": 982
+    },
+    {
+      "epoch": 1.100821508588499,
+      "grad_norm": 0.8224088549613953,
+      "learning_rate": 4.659308293224317e-06,
+      "loss": 0.5215,
+      "step": 983
+    },
+    {
+      "epoch": 1.1019417475728155,
+      "grad_norm": 0.8107877969741821,
+      "learning_rate": 4.658554262241659e-06,
+      "loss": 0.5536,
+      "step": 984
+    },
+    {
+      "epoch": 1.1030619865571323,
+      "grad_norm": 0.8498128652572632,
+      "learning_rate": 4.6577994589105715e-06,
+      "loss": 0.5537,
+      "step": 985
+    },
+    {
+      "epoch": 1.1041822255414488,
+      "grad_norm": 0.8373145461082458,
+      "learning_rate": 4.657043883501128e-06,
+      "loss": 0.5663,
+      "step": 986
+    },
+    {
+      "epoch": 1.1053024645257654,
+      "grad_norm": 0.8539902567863464,
+      "learning_rate": 4.656287536283681e-06,
+      "loss": 0.5684,
+      "step": 987
+    },
+    {
+      "epoch": 1.1064227035100822,
+      "grad_norm": 0.8307133316993713,
+      "learning_rate": 4.655530417528856e-06,
+      "loss": 0.552,
+      "step": 988
+    },
+    {
+      "epoch": 1.1075429424943988,
+      "grad_norm": 0.8355655670166016,
+      "learning_rate": 4.654772527507558e-06,
+      "loss": 0.5439,
+      "step": 989
+    },
+    {
+      "epoch": 1.1086631814787156,
+      "grad_norm": 0.8081607222557068,
+      "learning_rate": 4.6540138664909665e-06,
+      "loss": 0.5485,
+      "step": 990
+    },
+    {
+      "epoch": 1.1097834204630321,
+      "grad_norm": 0.8134661316871643,
+      "learning_rate": 4.653254434750535e-06,
+      "loss": 0.5661,
+      "step": 991
+    },
+    {
+      "epoch": 1.1109036594473487,
+      "grad_norm": 0.8444826602935791,
+      "learning_rate": 4.652494232557995e-06,
+      "loss": 0.5517,
+      "step": 992
+    },
+    {
+      "epoch": 1.1120238984316655,
+      "grad_norm": 0.8160927891731262,
+      "learning_rate": 4.651733260185353e-06,
+      "loss": 0.5181,
+      "step": 993
+    },
+    {
+      "epoch": 1.113144137415982,
+      "grad_norm": 0.8135287761688232,
+      "learning_rate": 4.650971517904892e-06,
+      "loss": 0.5747,
+      "step": 994
+    },
+    {
+      "epoch": 1.1142643764002986,
+      "grad_norm": 0.8516098260879517,
+      "learning_rate": 4.650209005989169e-06,
+      "loss": 0.5464,
+      "step": 995
+    },
+    {
+      "epoch": 1.1153846153846154,
+      "grad_norm": 0.814786434173584,
+      "learning_rate": 4.6494457247110166e-06,
+      "loss": 0.5315,
+      "step": 996
+    },
+    {
+      "epoch": 1.116504854368932,
+      "grad_norm": 0.8316855430603027,
+      "learning_rate": 4.648681674343543e-06,
+      "loss": 0.5427,
+      "step": 997
+    },
+    {
+      "epoch": 1.1176250933532488,
+      "grad_norm": 0.8516353964805603,
+      "learning_rate": 4.647916855160134e-06,
+      "loss": 0.5578,
+      "step": 998
+    },
+    {
+      "epoch": 1.1187453323375653,
+      "grad_norm": 0.8202309608459473,
+      "learning_rate": 4.647151267434445e-06,
+      "loss": 0.5515,
+      "step": 999
+    },
+    {
+      "epoch": 1.119865571321882,
+      "grad_norm": 0.8232266306877136,
+      "learning_rate": 4.6463849114404115e-06,
+      "loss": 0.564,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1209858103061987,
+      "grad_norm": 0.8440135717391968,
+      "learning_rate": 4.645617787452241e-06,
+      "loss": 0.5571,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1221060492905153,
+      "grad_norm": 0.8753403425216675,
+      "learning_rate": 4.6448498957444185e-06,
+      "loss": 0.5479,
+      "step": 1002
+    },
+    {
+      "epoch": 1.123226288274832,
+      "grad_norm": 0.8252137303352356,
+      "learning_rate": 4.6440812365917e-06,
+      "loss": 0.534,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1243465272591486,
+      "grad_norm": 0.8255431652069092,
+      "learning_rate": 4.643311810269118e-06,
+      "loss": 0.5505,
+      "step": 1004
+    },
+    {
+      "epoch": 1.1254667662434652,
+      "grad_norm": 0.820588231086731,
+      "learning_rate": 4.642541617051981e-06,
+      "loss": 0.576,
+      "step": 1005
+    },
+    {
+      "epoch": 1.126587005227782,
+      "grad_norm": 0.845767617225647,
+      "learning_rate": 4.6417706572158705e-06,
+      "loss": 0.5095,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1277072442120986,
+      "grad_norm": 0.8620426654815674,
+      "learning_rate": 4.64099893103664e-06,
+      "loss": 0.5443,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1288274831964151,
+      "grad_norm": 0.8254401683807373,
+      "learning_rate": 4.640226438790422e-06,
+      "loss": 0.5401,
+      "step": 1008
+    },
+    {
+      "epoch": 1.129947722180732,
+      "grad_norm": 0.8236058354377747,
+      "learning_rate": 4.639453180753619e-06,
+      "loss": 0.5579,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1310679611650485,
+      "grad_norm": 0.841072678565979,
+      "learning_rate": 4.63867915720291e-06,
+      "loss": 0.5714,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1321882001493653,
+      "grad_norm": 0.8392432928085327,
+      "learning_rate": 4.637904368415247e-06,
+      "loss": 0.5301,
+      "step": 1011
+    },
+    {
+      "epoch": 1.1333084391336818,
+      "grad_norm": 0.84500652551651,
+      "learning_rate": 4.637128814667854e-06,
+      "loss": 0.5214,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1344286781179984,
+      "grad_norm": 0.7862651944160461,
+      "learning_rate": 4.636352496238234e-06,
+      "loss": 0.5186,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1355489171023152,
+      "grad_norm": 0.8130670785903931,
+      "learning_rate": 4.635575413404156e-06,
+      "loss": 0.5428,
+      "step": 1014
+    },
+    {
+      "epoch": 1.1366691560866318,
+      "grad_norm": 0.8158490061759949,
+      "learning_rate": 4.634797566443668e-06,
+      "loss": 0.5263,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1377893950709486,
+      "grad_norm": 0.8362638354301453,
+      "learning_rate": 4.634018955635092e-06,
+      "loss": 0.5503,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1389096340552651,
+      "grad_norm": 0.828471839427948,
+      "learning_rate": 4.6332395812570206e-06,
+      "loss": 0.5525,
+      "step": 1017
+    },
+    {
+      "epoch": 1.1400298730395817,
+      "grad_norm": 0.8640450239181519,
+      "learning_rate": 4.632459443588318e-06,
+      "loss": 0.5819,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1411501120238985,
+      "grad_norm": 0.8726822733879089,
+      "learning_rate": 4.631678542908126e-06,
+      "loss": 0.5516,
+      "step": 1019
+    },
+    {
+      "epoch": 1.142270351008215,
+      "grad_norm": 0.8317193388938904,
+      "learning_rate": 4.630896879495858e-06,
+      "loss": 0.5643,
+      "step": 1020
+    },
+    {
+      "epoch": 1.1433905899925318,
+      "grad_norm": 0.8218059539794922,
+      "learning_rate": 4.6301144536311985e-06,
+      "loss": 0.566,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1445108289768484,
+      "grad_norm": 0.8336435556411743,
+      "learning_rate": 4.629331265594105e-06,
+      "loss": 0.5294,
+      "step": 1022
+    },
+    {
+      "epoch": 1.145631067961165,
+      "grad_norm": 0.8649945855140686,
+      "learning_rate": 4.628547315664809e-06,
+      "loss": 0.5594,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1467513069454818,
+      "grad_norm": 0.8594834208488464,
+      "learning_rate": 4.627762604123815e-06,
+      "loss": 0.5429,
+      "step": 1024
+    },
+    {
+      "epoch": 1.1478715459297983,
+      "grad_norm": 0.8120294213294983,
+      "learning_rate": 4.626977131251898e-06,
+      "loss": 0.5396,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1489917849141151,
+      "grad_norm": 0.8123870491981506,
+      "learning_rate": 4.626190897330109e-06,
+      "loss": 0.5408,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1501120238984317,
+      "grad_norm": 0.817854642868042,
+      "learning_rate": 4.625403902639767e-06,
+      "loss": 0.5469,
+      "step": 1027
+    },
+    {
+      "epoch": 1.1512322628827483,
+      "grad_norm": 0.8520634174346924,
+      "learning_rate": 4.6246161474624656e-06,
+      "loss": 0.5545,
+      "step": 1028
+    },
+    {
+      "epoch": 1.152352501867065,
+      "grad_norm": 0.8475438952445984,
+      "learning_rate": 4.62382763208007e-06,
+      "loss": 0.5228,
+      "step": 1029
+    },
+    {
+      "epoch": 1.1534727408513816,
+      "grad_norm": 0.8179640769958496,
+      "learning_rate": 4.623038356774717e-06,
+      "loss": 0.5539,
+      "step": 1030
+    },
+    {
+      "epoch": 1.1545929798356982,
+      "grad_norm": 0.8401399850845337,
+      "learning_rate": 4.622248321828817e-06,
+      "loss": 0.543,
+      "step": 1031
+    },
+    {
+      "epoch": 1.155713218820015,
+      "grad_norm": 0.8284974098205566,
+      "learning_rate": 4.621457527525051e-06,
+      "loss": 0.5396,
+      "step": 1032
+    },
+    {
+      "epoch": 1.1568334578043316,
+      "grad_norm": 0.8201552033424377,
+      "learning_rate": 4.62066597414637e-06,
+      "loss": 0.5641,
+      "step": 1033
+    },
+    {
+      "epoch": 1.1579536967886481,
+      "grad_norm": 0.819282591342926,
+      "learning_rate": 4.619873661976e-06,
+      "loss": 0.5276,
+      "step": 1034
+    },
+    {
+      "epoch": 1.159073935772965,
+      "grad_norm": 0.8161391615867615,
+      "learning_rate": 4.619080591297435e-06,
+      "loss": 0.5521,
+      "step": 1035
+    },
+    {
+      "epoch": 1.1601941747572815,
+      "grad_norm": 0.8103516697883606,
+      "learning_rate": 4.618286762394444e-06,
+      "loss": 0.55,
+      "step": 1036
+    },
+    {
+      "epoch": 1.1613144137415983,
+      "grad_norm": 0.8195378184318542,
+      "learning_rate": 4.617492175551065e-06,
+      "loss": 0.5367,
+      "step": 1037
+    },
+    {
+      "epoch": 1.1624346527259148,
+      "grad_norm": 0.8827869892120361,
+      "learning_rate": 4.6166968310516066e-06,
+      "loss": 0.5464,
+      "step": 1038
+    },
+    {
+      "epoch": 1.1635548917102314,
+      "grad_norm": 0.8650336861610413,
+      "learning_rate": 4.6159007291806515e-06,
+      "loss": 0.5257,
+      "step": 1039
+    },
+    {
+      "epoch": 1.1646751306945482,
+      "grad_norm": 0.852587878704071,
+      "learning_rate": 4.615103870223051e-06,
+      "loss": 0.5225,
+      "step": 1040
+    },
+    {
+      "epoch": 1.1657953696788648,
+      "grad_norm": 0.8107483983039856,
+      "learning_rate": 4.614306254463926e-06,
+      "loss": 0.56,
+      "step": 1041
+    },
+    {
+      "epoch": 1.1669156086631816,
+      "grad_norm": 0.8535749912261963,
+      "learning_rate": 4.613507882188671e-06,
+      "loss": 0.5365,
+      "step": 1042
+    },
+    {
+      "epoch": 1.1680358476474981,
+      "grad_norm": 0.8128278255462646,
+      "learning_rate": 4.6127087536829514e-06,
+      "loss": 0.5369,
+      "step": 1043
+    },
+    {
+      "epoch": 1.1691560866318147,
+      "grad_norm": 0.8134152889251709,
+      "learning_rate": 4.611908869232701e-06,
+      "loss": 0.5621,
+      "step": 1044
+    },
+    {
+      "epoch": 1.1702763256161315,
+      "grad_norm": 0.8495088815689087,
+      "learning_rate": 4.611108229124124e-06,
+      "loss": 0.581,
+      "step": 1045
+    },
+    {
+      "epoch": 1.171396564600448,
+      "grad_norm": 0.8236438035964966,
+      "learning_rate": 4.610306833643697e-06,
+      "loss": 0.518,
+      "step": 1046
+    },
+    {
+      "epoch": 1.1725168035847648,
+      "grad_norm": 0.793741762638092,
+      "learning_rate": 4.6095046830781666e-06,
+      "loss": 0.5389,
+      "step": 1047
+    },
+    {
+      "epoch": 1.1736370425690814,
+      "grad_norm": 0.8526056408882141,
+      "learning_rate": 4.608701777714547e-06,
+      "loss": 0.5665,
+      "step": 1048
+    },
+    {
+      "epoch": 1.174757281553398,
+      "grad_norm": 0.8603430986404419,
+      "learning_rate": 4.6078981178401265e-06,
+      "loss": 0.5524,
+      "step": 1049
+    },
+    {
+      "epoch": 1.1758775205377148,
+      "grad_norm": 0.8882256150245667,
+      "learning_rate": 4.607093703742459e-06,
+      "loss": 0.5554,
+      "step": 1050
+    },
+    {
+      "epoch": 1.1769977595220313,
+      "grad_norm": 0.8279406428337097,
+      "learning_rate": 4.606288535709372e-06,
+      "loss": 0.5506,
+      "step": 1051
+    },
+    {
+      "epoch": 1.1781179985063481,
+      "grad_norm": 0.8276134133338928,
+      "learning_rate": 4.605482614028961e-06,
+      "loss": 0.538,
+      "step": 1052
+    },
+    {
+      "epoch": 1.1792382374906647,
+      "grad_norm": 0.8543738126754761,
+      "learning_rate": 4.604675938989591e-06,
+      "loss": 0.5354,
+      "step": 1053
+    },
+    {
+      "epoch": 1.1803584764749813,
+      "grad_norm": 0.8727709054946899,
+      "learning_rate": 4.603868510879897e-06,
+      "loss": 0.5543,
+      "step": 1054
+    },
+    {
+      "epoch": 1.181478715459298,
+      "grad_norm": 0.8450466990470886,
+      "learning_rate": 4.603060329988784e-06,
+      "loss": 0.5412,
+      "step": 1055
+    },
+    {
+      "epoch": 1.1825989544436146,
+      "grad_norm": 0.8641091585159302,
+      "learning_rate": 4.602251396605425e-06,
+      "loss": 0.5771,
+      "step": 1056
+    },
+    {
+      "epoch": 1.1837191934279314,
+      "grad_norm": 0.8307204246520996,
+      "learning_rate": 4.601441711019263e-06,
+      "loss": 0.5457,
+      "step": 1057
+    },
+    {
+      "epoch": 1.184839432412248,
+      "grad_norm": 0.8530333638191223,
+      "learning_rate": 4.600631273520011e-06,
+      "loss": 0.5494,
+      "step": 1058
+    },
+    {
+      "epoch": 1.1859596713965646,
+      "grad_norm": 0.8765411376953125,
+      "learning_rate": 4.599820084397649e-06,
+      "loss": 0.5651,
+      "step": 1059
+    },
+    {
+      "epoch": 1.1870799103808813,
+      "grad_norm": 0.8434428572654724,
+      "learning_rate": 4.599008143942428e-06,
+      "loss": 0.5585,
+      "step": 1060
+    },
+    {
+      "epoch": 1.188200149365198,
+      "grad_norm": 0.8094550967216492,
+      "learning_rate": 4.598195452444867e-06,
+      "loss": 0.5651,
+      "step": 1061
+    },
+    {
+      "epoch": 1.1893203883495145,
+      "grad_norm": 0.825619101524353,
+      "learning_rate": 4.5973820101957535e-06,
+      "loss": 0.5358,
+      "step": 1062
+    },
+    {
+      "epoch": 1.1904406273338313,
+      "grad_norm": 0.7984577417373657,
+      "learning_rate": 4.596567817486143e-06,
+      "loss": 0.5305,
+      "step": 1063
+    },
+    {
+      "epoch": 1.1915608663181478,
+      "grad_norm": 0.8166099786758423,
+      "learning_rate": 4.595752874607362e-06,
+      "loss": 0.5268,
+      "step": 1064
+    },
+    {
+      "epoch": 1.1926811053024644,
+      "grad_norm": 0.8126864433288574,
+      "learning_rate": 4.594937181851002e-06,
+      "loss": 0.5416,
+      "step": 1065
+    },
+    {
+      "epoch": 1.1938013442867812,
+      "grad_norm": 0.8082033395767212,
+      "learning_rate": 4.594120739508925e-06,
+      "loss": 0.556,
+      "step": 1066
+    },
+    {
+      "epoch": 1.1949215832710978,
+      "grad_norm": 0.822907567024231,
+      "learning_rate": 4.593303547873262e-06,
+      "loss": 0.5374,
+      "step": 1067
+    },
+    {
+      "epoch": 1.1960418222554146,
+      "grad_norm": 0.8471306562423706,
+      "learning_rate": 4.59248560723641e-06,
+      "loss": 0.5619,
+      "step": 1068
+    },
+    {
+      "epoch": 1.1971620612397311,
+      "grad_norm": 0.849753201007843,
+      "learning_rate": 4.591666917891034e-06,
+      "loss": 0.5517,
+      "step": 1069
+    },
+    {
+      "epoch": 1.1982823002240477,
+      "grad_norm": 0.8199844360351562,
+      "learning_rate": 4.590847480130069e-06,
+      "loss": 0.5431,
+      "step": 1070
+    },
+    {
+      "epoch": 1.1994025392083645,
+      "grad_norm": 0.846013069152832,
+      "learning_rate": 4.590027294246717e-06,
+      "loss": 0.5463,
+      "step": 1071
+    },
+    {
+      "epoch": 1.200522778192681,
+      "grad_norm": 0.8319349884986877,
+      "learning_rate": 4.589206360534446e-06,
+      "loss": 0.5374,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2016430171769978,
+      "grad_norm": 0.8070456981658936,
+      "learning_rate": 4.5883846792869925e-06,
+      "loss": 0.5283,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2027632561613144,
+      "grad_norm": 0.7885628342628479,
+      "learning_rate": 4.5875622507983616e-06,
+      "loss": 0.5542,
+      "step": 1074
+    },
+    {
+      "epoch": 1.203883495145631,
+      "grad_norm": 0.8185691833496094,
+      "learning_rate": 4.586739075362826e-06,
+      "loss": 0.5316,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2050037341299478,
+      "grad_norm": 0.8217731714248657,
+      "learning_rate": 4.585915153274924e-06,
+      "loss": 0.5436,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2061239731142643,
+      "grad_norm": 0.8212025761604309,
+      "learning_rate": 4.585090484829459e-06,
+      "loss": 0.5589,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2072442120985811,
+      "grad_norm": 0.8024436831474304,
+      "learning_rate": 4.5842650703215075e-06,
+      "loss": 0.5324,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2083644510828977,
+      "grad_norm": 0.8126838803291321,
+      "learning_rate": 4.583438910046409e-06,
+      "loss": 0.5519,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2094846900672143,
+      "grad_norm": 0.819503128528595,
+      "learning_rate": 4.582612004299769e-06,
+      "loss": 0.5508,
+      "step": 1080
+    },
+    {
+      "epoch": 1.210604929051531,
+      "grad_norm": 0.8436784148216248,
+      "learning_rate": 4.581784353377463e-06,
+      "loss": 0.5544,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2117251680358476,
+      "grad_norm": 0.8434869647026062,
+      "learning_rate": 4.580955957575629e-06,
+      "loss": 0.5531,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2128454070201644,
+      "grad_norm": 0.8168767094612122,
+      "learning_rate": 4.580126817190676e-06,
+      "loss": 0.5044,
+      "step": 1083
+    },
+    {
+      "epoch": 1.213965646004481,
+      "grad_norm": 0.7980238199234009,
+      "learning_rate": 4.579296932519276e-06,
+      "loss": 0.5471,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2150858849887975,
+      "grad_norm": 0.8311899304389954,
+      "learning_rate": 4.578466303858369e-06,
+      "loss": 0.5315,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2162061239731143,
+      "grad_norm": 0.8057358860969543,
+      "learning_rate": 4.577634931505162e-06,
+      "loss": 0.5133,
+      "step": 1086
+    },
+    {
+      "epoch": 1.217326362957431,
+      "grad_norm": 0.8055294156074524,
+      "learning_rate": 4.5768028157571235e-06,
+      "loss": 0.5525,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2184466019417475,
+      "grad_norm": 0.8826123476028442,
+      "learning_rate": 4.575969956911994e-06,
+      "loss": 0.5382,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2195668409260643,
+      "grad_norm": 0.8325172662734985,
+      "learning_rate": 4.575136355267777e-06,
+      "loss": 0.56,
+      "step": 1089
+    },
+    {
+      "epoch": 1.2206870799103808,
+      "grad_norm": 0.843222439289093,
+      "learning_rate": 4.574302011122741e-06,
+      "loss": 0.5589,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2218073188946976,
+      "grad_norm": 0.8231297135353088,
+      "learning_rate": 4.5734669247754225e-06,
+      "loss": 0.5376,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2229275578790142,
+      "grad_norm": 0.8703131079673767,
+      "learning_rate": 4.572631096524621e-06,
+      "loss": 0.5415,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2240477968633308,
+      "grad_norm": 0.8341230154037476,
+      "learning_rate": 4.571794526669404e-06,
+      "loss": 0.5755,
+      "step": 1093
+    },
+    {
+      "epoch": 1.2251680358476476,
+      "grad_norm": 0.8507810831069946,
+      "learning_rate": 4.570957215509102e-06,
+      "loss": 0.5324,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2262882748319641,
+      "grad_norm": 0.8084942698478699,
+      "learning_rate": 4.570119163343312e-06,
+      "loss": 0.56,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2274085138162807,
+      "grad_norm": 0.8282336592674255,
+      "learning_rate": 4.569280370471896e-06,
+      "loss": 0.5544,
+      "step": 1096
+    },
+    {
+      "epoch": 1.2285287528005975,
+      "grad_norm": 0.789263904094696,
+      "learning_rate": 4.568440837194981e-06,
+      "loss": 0.5526,
+      "step": 1097
+    },
+    {
+      "epoch": 1.229648991784914,
+      "grad_norm": 0.8346118330955505,
+      "learning_rate": 4.567600563812958e-06,
+      "loss": 0.5409,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.8182159662246704,
+      "learning_rate": 4.566759550626486e-06,
+      "loss": 0.5391,
+      "step": 1099
+    },
+    {
+      "epoch": 1.2318894697535474,
+      "grad_norm": 0.8271141648292542,
+      "learning_rate": 4.5659177979364846e-06,
+      "loss": 0.5368,
+      "step": 1100
+    },
+    {
+      "epoch": 1.233009708737864,
+      "grad_norm": 0.804466187953949,
+      "learning_rate": 4.56507530604414e-06,
+      "loss": 0.5426,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2341299477221808,
+      "grad_norm": 0.836618959903717,
+      "learning_rate": 4.564232075250904e-06,
+      "loss": 0.5828,
+      "step": 1102
+    },
+    {
+      "epoch": 1.2352501867064973,
+      "grad_norm": 0.8297340869903564,
+      "learning_rate": 4.56338810585849e-06,
+      "loss": 0.5259,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2363704256908141,
+      "grad_norm": 0.8460643887519836,
+      "learning_rate": 4.562543398168876e-06,
+      "loss": 0.5463,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2374906646751307,
+      "grad_norm": 0.8542268872261047,
+      "learning_rate": 4.561697952484309e-06,
+      "loss": 0.5487,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2386109036594473,
+      "grad_norm": 0.8352535963058472,
+      "learning_rate": 4.560851769107294e-06,
+      "loss": 0.5565,
+      "step": 1106
+    },
+    {
+      "epoch": 1.239731142643764,
+      "grad_norm": 0.8429189324378967,
+      "learning_rate": 4.560004848340602e-06,
+      "loss": 0.5627,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2408513816280806,
+      "grad_norm": 0.8104339241981506,
+      "learning_rate": 4.55915719048727e-06,
+      "loss": 0.5535,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2419716206123974,
+      "grad_norm": 0.8268340229988098,
+      "learning_rate": 4.5583087958505954e-06,
+      "loss": 0.5557,
+      "step": 1109
+    },
+    {
+      "epoch": 1.243091859596714,
+      "grad_norm": 0.8117450475692749,
+      "learning_rate": 4.5574596647341414e-06,
+      "loss": 0.5477,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2442120985810305,
+      "grad_norm": 0.8114601969718933,
+      "learning_rate": 4.556609797441734e-06,
+      "loss": 0.5413,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2453323375653473,
+      "grad_norm": 0.835013747215271,
+      "learning_rate": 4.555759194277462e-06,
+      "loss": 0.5319,
+      "step": 1112
+    },
+    {
+      "epoch": 1.246452576549664,
+      "grad_norm": 0.7998044490814209,
+      "learning_rate": 4.554907855545679e-06,
+      "loss": 0.5397,
+      "step": 1113
+    },
+    {
+      "epoch": 1.2475728155339807,
+      "grad_norm": 0.8180138468742371,
+      "learning_rate": 4.554055781551002e-06,
+      "loss": 0.5456,
+      "step": 1114
+    },
+    {
+      "epoch": 1.2486930545182973,
+      "grad_norm": 0.8064574003219604,
+      "learning_rate": 4.5532029725983086e-06,
+      "loss": 0.568,
+      "step": 1115
+    },
+    {
+      "epoch": 1.2498132935026138,
+      "grad_norm": 0.802710771560669,
+      "learning_rate": 4.55234942899274e-06,
+      "loss": 0.5389,
+      "step": 1116
+    },
+    {
+      "epoch": 1.2509335324869306,
+      "grad_norm": 0.8466383814811707,
+      "learning_rate": 4.551495151039703e-06,
+      "loss": 0.5397,
+      "step": 1117
+    },
+    {
+      "epoch": 1.2520537714712472,
+      "grad_norm": 0.8483602404594421,
+      "learning_rate": 4.550640139044864e-06,
+      "loss": 0.5491,
+      "step": 1118
+    },
+    {
+      "epoch": 1.253174010455564,
+      "grad_norm": 0.8475786447525024,
+      "learning_rate": 4.549784393314155e-06,
+      "loss": 0.5303,
+      "step": 1119
+    },
+    {
+      "epoch": 1.2542942494398805,
+      "grad_norm": 0.8381414413452148,
+      "learning_rate": 4.548927914153766e-06,
+      "loss": 0.5235,
+      "step": 1120
+    },
+    {
+      "epoch": 1.2554144884241971,
+      "grad_norm": 0.8032870888710022,
+      "learning_rate": 4.548070701870155e-06,
+      "loss": 0.5321,
+      "step": 1121
+    },
+    {
+      "epoch": 1.2565347274085137,
+      "grad_norm": 0.8096051812171936,
+      "learning_rate": 4.547212756770039e-06,
+      "loss": 0.5435,
+      "step": 1122
+    },
+    {
+      "epoch": 1.2576549663928305,
+      "grad_norm": 0.8399277925491333,
+      "learning_rate": 4.546354079160395e-06,
+      "loss": 0.5434,
+      "step": 1123
+    },
+    {
+      "epoch": 1.258775205377147,
+      "grad_norm": 0.8209378719329834,
+      "learning_rate": 4.545494669348469e-06,
+      "loss": 0.5584,
+      "step": 1124
+    },
+    {
+      "epoch": 1.2598954443614638,
+      "grad_norm": 0.8418322801589966,
+      "learning_rate": 4.544634527641764e-06,
+      "loss": 0.5545,
+      "step": 1125
+    },
+    {
+      "epoch": 1.2610156833457804,
+      "grad_norm": 0.8334120512008667,
+      "learning_rate": 4.543773654348044e-06,
+      "loss": 0.5252,
+      "step": 1126
+    },
+    {
+      "epoch": 1.262135922330097,
+      "grad_norm": 0.8128974437713623,
+      "learning_rate": 4.542912049775337e-06,
+      "loss": 0.5421,
+      "step": 1127
+    },
+    {
+      "epoch": 1.2632561613144138,
+      "grad_norm": 0.8154008984565735,
+      "learning_rate": 4.542049714231932e-06,
+      "loss": 0.5275,
+      "step": 1128
+    },
+    {
+      "epoch": 1.2643764002987303,
+      "grad_norm": 0.831755518913269,
+      "learning_rate": 4.541186648026381e-06,
+      "loss": 0.523,
+      "step": 1129
+    },
+    {
+      "epoch": 1.2654966392830471,
+      "grad_norm": 0.7766554355621338,
+      "learning_rate": 4.540322851467496e-06,
+      "loss": 0.5502,
+      "step": 1130
+    },
+    {
+      "epoch": 1.2666168782673637,
+      "grad_norm": 0.8152607083320618,
+      "learning_rate": 4.53945832486435e-06,
+      "loss": 0.5464,
+      "step": 1131
+    },
+    {
+      "epoch": 1.2677371172516803,
+      "grad_norm": 0.8785687685012817,
+      "learning_rate": 4.538593068526277e-06,
+      "loss": 0.5537,
+      "step": 1132
+    },
+    {
+      "epoch": 1.268857356235997,
+      "grad_norm": 0.8451907634735107,
+      "learning_rate": 4.537727082762873e-06,
+      "loss": 0.5665,
+      "step": 1133
+    },
+    {
+      "epoch": 1.2699775952203136,
+      "grad_norm": 0.8168203234672546,
+      "learning_rate": 4.536860367883996e-06,
+      "loss": 0.5637,
+      "step": 1134
+    },
+    {
+      "epoch": 1.2710978342046304,
+      "grad_norm": 0.8470443487167358,
+      "learning_rate": 4.535992924199764e-06,
+      "loss": 0.5134,
+      "step": 1135
+    },
+    {
+      "epoch": 1.272218073188947,
+      "grad_norm": 0.8196655511856079,
+      "learning_rate": 4.535124752020552e-06,
+      "loss": 0.5358,
+      "step": 1136
+    },
+    {
+      "epoch": 1.2733383121732635,
+      "grad_norm": 0.8088006973266602,
+      "learning_rate": 4.534255851657004e-06,
+      "loss": 0.5317,
+      "step": 1137
+    },
+    {
+      "epoch": 1.2744585511575803,
+      "grad_norm": 0.8026657700538635,
+      "learning_rate": 4.533386223420017e-06,
+      "loss": 0.5255,
+      "step": 1138
+    },
+    {
+      "epoch": 1.275578790141897,
+      "grad_norm": 0.8050847053527832,
+      "learning_rate": 4.53251586762075e-06,
+      "loss": 0.5487,
+      "step": 1139
+    },
+    {
+      "epoch": 1.2766990291262137,
+      "grad_norm": 0.828590989112854,
+      "learning_rate": 4.5316447845706255e-06,
+      "loss": 0.565,
+      "step": 1140
+    },
+    {
+      "epoch": 1.2778192681105303,
+      "grad_norm": 0.8065359592437744,
+      "learning_rate": 4.530772974581324e-06,
+      "loss": 0.529,
+      "step": 1141
+    },
+    {
+      "epoch": 1.2789395070948468,
+      "grad_norm": 0.8265255689620972,
+      "learning_rate": 4.529900437964786e-06,
+      "loss": 0.5246,
+      "step": 1142
+    },
+    {
+      "epoch": 1.2800597460791636,
+      "grad_norm": 0.8076423406600952,
+      "learning_rate": 4.5290271750332115e-06,
+      "loss": 0.5584,
+      "step": 1143
+    },
+    {
+      "epoch": 1.2811799850634802,
+      "grad_norm": 0.8156341314315796,
+      "learning_rate": 4.528153186099062e-06,
+      "loss": 0.5422,
+      "step": 1144
+    },
+    {
+      "epoch": 1.282300224047797,
+      "grad_norm": 0.8484329581260681,
+      "learning_rate": 4.527278471475057e-06,
+      "loss": 0.5154,
+      "step": 1145
+    },
+    {
+      "epoch": 1.2834204630321135,
+      "grad_norm": 0.7974900603294373,
+      "learning_rate": 4.5264030314741784e-06,
+      "loss": 0.5114,
+      "step": 1146
+    },
+    {
+      "epoch": 1.2845407020164301,
+      "grad_norm": 0.8154396414756775,
+      "learning_rate": 4.525526866409663e-06,
+      "loss": 0.5435,
+      "step": 1147
+    },
+    {
+      "epoch": 1.285660941000747,
+      "grad_norm": 0.8144055008888245,
+      "learning_rate": 4.524649976595013e-06,
+      "loss": 0.5548,
+      "step": 1148
+    },
+    {
+      "epoch": 1.2867811799850635,
+      "grad_norm": 0.8400017619132996,
+      "learning_rate": 4.5237723623439844e-06,
+      "loss": 0.565,
+      "step": 1149
+    },
+    {
+      "epoch": 1.2879014189693803,
+      "grad_norm": 0.8405039310455322,
+      "learning_rate": 4.522894023970596e-06,
+      "loss": 0.5222,
+      "step": 1150
+    },
+    {
+      "epoch": 1.2890216579536968,
+      "grad_norm": 0.8041941523551941,
+      "learning_rate": 4.522014961789124e-06,
+      "loss": 0.5437,
+      "step": 1151
+    },
+    {
+      "epoch": 1.2901418969380134,
+      "grad_norm": 0.9180984497070312,
+      "learning_rate": 4.521135176114104e-06,
+      "loss": 0.5371,
+      "step": 1152
+    },
+    {
+      "epoch": 1.29126213592233,
+      "grad_norm": 0.839199423789978,
+      "learning_rate": 4.520254667260331e-06,
+      "loss": 0.5252,
+      "step": 1153
+    },
+    {
+      "epoch": 1.2923823749066468,
+      "grad_norm": 0.8278453350067139,
+      "learning_rate": 4.519373435542857e-06,
+      "loss": 0.5851,
+      "step": 1154
+    },
+    {
+      "epoch": 1.2935026138909633,
+      "grad_norm": 0.8621891736984253,
+      "learning_rate": 4.518491481276997e-06,
+      "loss": 0.581,
+      "step": 1155
+    },
+    {
+      "epoch": 1.2946228528752801,
+      "grad_norm": 0.8634688258171082,
+      "learning_rate": 4.517608804778317e-06,
+      "loss": 0.5502,
+      "step": 1156
+    },
+    {
+      "epoch": 1.2957430918595967,
+      "grad_norm": 0.8093946576118469,
+      "learning_rate": 4.51672540636265e-06,
+      "loss": 0.5402,
+      "step": 1157
+    },
+    {
+      "epoch": 1.2968633308439133,
+      "grad_norm": 0.8530567288398743,
+      "learning_rate": 4.515841286346081e-06,
+      "loss": 0.5314,
+      "step": 1158
+    },
+    {
+      "epoch": 1.29798356982823,
+      "grad_norm": 0.8433166146278381,
+      "learning_rate": 4.514956445044955e-06,
+      "loss": 0.5439,
+      "step": 1159
+    },
+    {
+      "epoch": 1.2991038088125466,
+      "grad_norm": 0.8189598321914673,
+      "learning_rate": 4.514070882775877e-06,
+      "loss": 0.5388,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3002240477968634,
+      "grad_norm": 0.8377291560173035,
+      "learning_rate": 4.513184599855708e-06,
+      "loss": 0.5385,
+      "step": 1161
+    },
+    {
+      "epoch": 1.30134428678118,
+      "grad_norm": 0.8701413869857788,
+      "learning_rate": 4.512297596601567e-06,
+      "loss": 0.5326,
+      "step": 1162
+    },
+    {
+      "epoch": 1.3024645257654965,
+      "grad_norm": 0.8627327084541321,
+      "learning_rate": 4.511409873330832e-06,
+      "loss": 0.5465,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3035847647498133,
+      "grad_norm": 0.8221105933189392,
+      "learning_rate": 4.510521430361135e-06,
+      "loss": 0.5413,
+      "step": 1164
+    },
+    {
+      "epoch": 1.30470500373413,
+      "grad_norm": 0.8241137266159058,
+      "learning_rate": 4.5096322680103725e-06,
+      "loss": 0.5246,
+      "step": 1165
+    },
+    {
+      "epoch": 1.3058252427184467,
+      "grad_norm": 0.8309207558631897,
+      "learning_rate": 4.50874238659669e-06,
+      "loss": 0.5335,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3069454817027633,
+      "grad_norm": 0.848537027835846,
+      "learning_rate": 4.507851786438496e-06,
+      "loss": 0.561,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3080657206870798,
+      "grad_norm": 0.8590072393417358,
+      "learning_rate": 4.506960467854454e-06,
+      "loss": 0.5493,
+      "step": 1168
+    },
+    {
+      "epoch": 1.3091859596713966,
+      "grad_norm": 0.8492612838745117,
+      "learning_rate": 4.506068431163487e-06,
+      "loss": 0.5442,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3103061986557132,
+      "grad_norm": 0.8207250833511353,
+      "learning_rate": 4.505175676684771e-06,
+      "loss": 0.5088,
+      "step": 1170
+    },
+    {
+      "epoch": 1.31142643764003,
+      "grad_norm": 0.8155739307403564,
+      "learning_rate": 4.5042822047377424e-06,
+      "loss": 0.5686,
+      "step": 1171
+    },
+    {
+      "epoch": 1.3125466766243465,
+      "grad_norm": 0.8703952431678772,
+      "learning_rate": 4.503388015642091e-06,
+      "loss": 0.5401,
+      "step": 1172
+    },
+    {
+      "epoch": 1.313666915608663,
+      "grad_norm": 0.8210623860359192,
+      "learning_rate": 4.502493109717766e-06,
+      "loss": 0.5272,
+      "step": 1173
+    },
+    {
+      "epoch": 1.31478715459298,
+      "grad_norm": 0.8164148926734924,
+      "learning_rate": 4.5015974872849725e-06,
+      "loss": 0.5416,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3159073935772965,
+      "grad_norm": 0.8628402352333069,
+      "learning_rate": 4.500701148664171e-06,
+      "loss": 0.5546,
+      "step": 1175
+    },
+    {
+      "epoch": 1.3170276325616133,
+      "grad_norm": 0.8327616453170776,
+      "learning_rate": 4.499804094176078e-06,
+      "loss": 0.5482,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3181478715459298,
+      "grad_norm": 0.8208657503128052,
+      "learning_rate": 4.498906324141669e-06,
+      "loss": 0.5443,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3192681105302464,
+      "grad_norm": 0.807716429233551,
+      "learning_rate": 4.498007838882172e-06,
+      "loss": 0.526,
+      "step": 1178
+    },
+    {
+      "epoch": 1.3203883495145632,
+      "grad_norm": 0.829759418964386,
+      "learning_rate": 4.497108638719072e-06,
+      "loss": 0.5528,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3215085884988798,
+      "grad_norm": 0.799696147441864,
+      "learning_rate": 4.496208723974112e-06,
+      "loss": 0.539,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3226288274831965,
+      "grad_norm": 0.8294451236724854,
+      "learning_rate": 4.495308094969288e-06,
+      "loss": 0.5443,
+      "step": 1181
+    },
+    {
+      "epoch": 1.3237490664675131,
+      "grad_norm": 0.8449733257293701,
+      "learning_rate": 4.494406752026851e-06,
+      "loss": 0.5356,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3248693054518297,
+      "grad_norm": 0.8247659206390381,
+      "learning_rate": 4.493504695469311e-06,
+      "loss": 0.5608,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3259895444361463,
+      "grad_norm": 0.835203230381012,
+      "learning_rate": 4.492601925619431e-06,
+      "loss": 0.5756,
+      "step": 1184
+    },
+    {
+      "epoch": 1.327109783420463,
+      "grad_norm": 0.8409526348114014,
+      "learning_rate": 4.491698442800228e-06,
+      "loss": 0.5582,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3282300224047796,
+      "grad_norm": 0.8321925401687622,
+      "learning_rate": 4.490794247334977e-06,
+      "loss": 0.5195,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3293502613890964,
+      "grad_norm": 0.8160293102264404,
+      "learning_rate": 4.489889339547206e-06,
+      "loss": 0.5383,
+      "step": 1187
+    },
+    {
+      "epoch": 1.330470500373413,
+      "grad_norm": 0.8163958191871643,
+      "learning_rate": 4.488983719760698e-06,
+      "loss": 0.5617,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3315907393577295,
+      "grad_norm": 0.8268260359764099,
+      "learning_rate": 4.488077388299492e-06,
+      "loss": 0.5437,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3327109783420463,
+      "grad_norm": 0.838289201259613,
+      "learning_rate": 4.48717034548788e-06,
+      "loss": 0.5304,
+      "step": 1190
+    },
+    {
+      "epoch": 1.333831217326363,
+      "grad_norm": 0.8222647309303284,
+      "learning_rate": 4.48626259165041e-06,
+      "loss": 0.5306,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3349514563106797,
+      "grad_norm": 0.7920235991477966,
+      "learning_rate": 4.485354127111884e-06,
+      "loss": 0.5501,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3360716952949963,
+      "grad_norm": 0.8315522074699402,
+      "learning_rate": 4.484444952197358e-06,
+      "loss": 0.5149,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3371919342793128,
+      "grad_norm": 0.813721776008606,
+      "learning_rate": 4.483535067232141e-06,
+      "loss": 0.58,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3383121732636296,
+      "grad_norm": 0.8278853893280029,
+      "learning_rate": 4.482624472541799e-06,
+      "loss": 0.5347,
+      "step": 1195
+    },
+    {
+      "epoch": 1.3394324122479462,
+      "grad_norm": 0.8072376847267151,
+      "learning_rate": 4.481713168452149e-06,
+      "loss": 0.5417,
+      "step": 1196
+    },
+    {
+      "epoch": 1.340552651232263,
+      "grad_norm": 0.8479835987091064,
+      "learning_rate": 4.480801155289263e-06,
+      "loss": 0.5545,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3416728902165795,
+      "grad_norm": 0.857487142086029,
+      "learning_rate": 4.479888433379468e-06,
+      "loss": 0.5542,
+      "step": 1198
+    },
+    {
+      "epoch": 1.342793129200896,
+      "grad_norm": 0.8252971172332764,
+      "learning_rate": 4.478975003049343e-06,
+      "loss": 0.5599,
+      "step": 1199
+    },
+    {
+      "epoch": 1.343913368185213,
+      "grad_norm": 0.8503106236457825,
+      "learning_rate": 4.478060864625721e-06,
+      "loss": 0.5351,
+      "step": 1200
+    },
+    {
+      "epoch": 1.3450336071695295,
+      "grad_norm": 0.7950337529182434,
+      "learning_rate": 4.477146018435687e-06,
+      "loss": 0.5313,
+      "step": 1201
+    },
+    {
+      "epoch": 1.3461538461538463,
+      "grad_norm": 0.8159598112106323,
+      "learning_rate": 4.476230464806581e-06,
+      "loss": 0.5472,
+      "step": 1202
+    },
+    {
+      "epoch": 1.3472740851381628,
+      "grad_norm": 0.833303689956665,
+      "learning_rate": 4.475314204065997e-06,
+      "loss": 0.5386,
+      "step": 1203
+    },
+    {
+      "epoch": 1.3483943241224794,
+      "grad_norm": 0.8313762545585632,
+      "learning_rate": 4.474397236541779e-06,
+      "loss": 0.562,
+      "step": 1204
+    },
+    {
+      "epoch": 1.3495145631067962,
+      "grad_norm": 0.8206455707550049,
+      "learning_rate": 4.4734795625620265e-06,
+      "loss": 0.5418,
+      "step": 1205
+    },
+    {
+      "epoch": 1.3506348020911128,
+      "grad_norm": 0.7903845906257629,
+      "learning_rate": 4.47256118245509e-06,
+      "loss": 0.5413,
+      "step": 1206
+    },
+    {
+      "epoch": 1.3517550410754295,
+      "grad_norm": 0.8083523511886597,
+      "learning_rate": 4.4716420965495735e-06,
+      "loss": 0.5385,
+      "step": 1207
+    },
+    {
+      "epoch": 1.352875280059746,
+      "grad_norm": 0.843407928943634,
+      "learning_rate": 4.470722305174334e-06,
+      "loss": 0.5568,
+      "step": 1208
+    },
+    {
+      "epoch": 1.3539955190440627,
+      "grad_norm": 0.8617134094238281,
+      "learning_rate": 4.46980180865848e-06,
+      "loss": 0.55,
+      "step": 1209
+    },
+    {
+      "epoch": 1.3551157580283795,
+      "grad_norm": 0.8540908098220825,
+      "learning_rate": 4.468880607331373e-06,
+      "loss": 0.5242,
+      "step": 1210
+    },
+    {
+      "epoch": 1.356235997012696,
+      "grad_norm": 0.779636561870575,
+      "learning_rate": 4.4679587015226255e-06,
+      "loss": 0.5295,
+      "step": 1211
+    },
+    {
+      "epoch": 1.3573562359970128,
+      "grad_norm": 0.826751172542572,
+      "learning_rate": 4.467036091562105e-06,
+      "loss": 0.5158,
+      "step": 1212
+    },
+    {
+      "epoch": 1.3584764749813294,
+      "grad_norm": 0.854387104511261,
+      "learning_rate": 4.466112777779927e-06,
+      "loss": 0.5386,
+      "step": 1213
+    },
+    {
+      "epoch": 1.359596713965646,
+      "grad_norm": 0.7959662675857544,
+      "learning_rate": 4.465188760506461e-06,
+      "loss": 0.5212,
+      "step": 1214
+    },
+    {
+      "epoch": 1.3607169529499625,
+      "grad_norm": 0.8715735077857971,
+      "learning_rate": 4.464264040072329e-06,
+      "loss": 0.553,
+      "step": 1215
+    },
+    {
+      "epoch": 1.3618371919342793,
+      "grad_norm": 0.8566111922264099,
+      "learning_rate": 4.463338616808403e-06,
+      "loss": 0.5478,
+      "step": 1216
+    },
+    {
+      "epoch": 1.362957430918596,
+      "grad_norm": 0.8157747983932495,
+      "learning_rate": 4.462412491045806e-06,
+      "loss": 0.5432,
+      "step": 1217
+    },
+    {
+      "epoch": 1.3640776699029127,
+      "grad_norm": 0.8130636215209961,
+      "learning_rate": 4.461485663115915e-06,
+      "loss": 0.5251,
+      "step": 1218
+    },
+    {
+      "epoch": 1.3651979088872293,
+      "grad_norm": 0.8303166031837463,
+      "learning_rate": 4.460558133350356e-06,
+      "loss": 0.5606,
+      "step": 1219
+    },
+    {
+      "epoch": 1.3663181478715458,
+      "grad_norm": 0.8391600847244263,
+      "learning_rate": 4.459629902081007e-06,
+      "loss": 0.5662,
+      "step": 1220
+    },
+    {
+      "epoch": 1.3674383868558626,
+      "grad_norm": 0.8392864465713501,
+      "learning_rate": 4.458700969639997e-06,
+      "loss": 0.5547,
+      "step": 1221
+    },
+    {
+      "epoch": 1.3685586258401792,
+      "grad_norm": 0.8507291674613953,
+      "learning_rate": 4.457771336359703e-06,
+      "loss": 0.5587,
+      "step": 1222
+    },
+    {
+      "epoch": 1.369678864824496,
+      "grad_norm": 0.8064186573028564,
+      "learning_rate": 4.456841002572759e-06,
+      "loss": 0.5175,
+      "step": 1223
+    },
+    {
+      "epoch": 1.3707991038088125,
+      "grad_norm": 0.8098925948143005,
+      "learning_rate": 4.4559099686120445e-06,
+      "loss": 0.5287,
+      "step": 1224
+    },
+    {
+      "epoch": 1.371919342793129,
+      "grad_norm": 0.8226019144058228,
+      "learning_rate": 4.454978234810692e-06,
+      "loss": 0.5388,
+      "step": 1225
+    },
+    {
+      "epoch": 1.373039581777446,
+      "grad_norm": 0.8429880142211914,
+      "learning_rate": 4.454045801502082e-06,
+      "loss": 0.5407,
+      "step": 1226
+    },
+    {
+      "epoch": 1.3741598207617625,
+      "grad_norm": 0.8022751212120056,
+      "learning_rate": 4.4531126690198476e-06,
+      "loss": 0.5264,
+      "step": 1227
+    },
+    {
+      "epoch": 1.3752800597460793,
+      "grad_norm": 0.7995109558105469,
+      "learning_rate": 4.452178837697871e-06,
+      "loss": 0.5557,
+      "step": 1228
+    },
+    {
+      "epoch": 1.3764002987303958,
+      "grad_norm": 0.8224391937255859,
+      "learning_rate": 4.451244307870285e-06,
+      "loss": 0.5414,
+      "step": 1229
+    },
+    {
+      "epoch": 1.3775205377147124,
+      "grad_norm": 0.8280567526817322,
+      "learning_rate": 4.450309079871472e-06,
+      "loss": 0.5489,
+      "step": 1230
+    },
+    {
+      "epoch": 1.3786407766990292,
+      "grad_norm": 0.8450224995613098,
+      "learning_rate": 4.449373154036064e-06,
+      "loss": 0.5364,
+      "step": 1231
+    },
+    {
+      "epoch": 1.3797610156833457,
+      "grad_norm": 0.7838452458381653,
+      "learning_rate": 4.4484365306989434e-06,
+      "loss": 0.5183,
+      "step": 1232
+    },
+    {
+      "epoch": 1.3808812546676625,
+      "grad_norm": 0.8605710864067078,
+      "learning_rate": 4.447499210195242e-06,
+      "loss": 0.542,
+      "step": 1233
+    },
+    {
+      "epoch": 1.382001493651979,
+      "grad_norm": 0.8326447606086731,
+      "learning_rate": 4.44656119286034e-06,
+      "loss": 0.5341,
+      "step": 1234
+    },
+    {
+      "epoch": 1.3831217326362957,
+      "grad_norm": 0.8388296365737915,
+      "learning_rate": 4.4456224790298666e-06,
+      "loss": 0.5493,
+      "step": 1235
+    },
+    {
+      "epoch": 1.3842419716206125,
+      "grad_norm": 0.8321563005447388,
+      "learning_rate": 4.444683069039703e-06,
+      "loss": 0.5719,
+      "step": 1236
+    },
+    {
+      "epoch": 1.385362210604929,
+      "grad_norm": 0.8426681756973267,
+      "learning_rate": 4.443742963225979e-06,
+      "loss": 0.5329,
+      "step": 1237
+    },
+    {
+      "epoch": 1.3864824495892458,
+      "grad_norm": 0.8210824728012085,
+      "learning_rate": 4.44280216192507e-06,
+      "loss": 0.5537,
+      "step": 1238
+    },
+    {
+      "epoch": 1.3876026885735624,
+      "grad_norm": 0.8024011850357056,
+      "learning_rate": 4.441860665473602e-06,
+      "loss": 0.5084,
+      "step": 1239
+    },
+    {
+      "epoch": 1.388722927557879,
+      "grad_norm": 0.8122202157974243,
+      "learning_rate": 4.4409184742084514e-06,
+      "loss": 0.5225,
+      "step": 1240
+    },
+    {
+      "epoch": 1.3898431665421958,
+      "grad_norm": 0.8381528258323669,
+      "learning_rate": 4.439975588466742e-06,
+      "loss": 0.5325,
+      "step": 1241
+    },
+    {
+      "epoch": 1.3909634055265123,
+      "grad_norm": 0.8207586407661438,
+      "learning_rate": 4.439032008585846e-06,
+      "loss": 0.5557,
+      "step": 1242
+    },
+    {
+      "epoch": 1.392083644510829,
+      "grad_norm": 0.8223419189453125,
+      "learning_rate": 4.438087734903385e-06,
+      "loss": 0.544,
+      "step": 1243
+    },
+    {
+      "epoch": 1.3932038834951457,
+      "grad_norm": 0.8231616020202637,
+      "learning_rate": 4.437142767757225e-06,
+      "loss": 0.5503,
+      "step": 1244
+    },
+    {
+      "epoch": 1.3943241224794622,
+      "grad_norm": 0.8446402549743652,
+      "learning_rate": 4.436197107485485e-06,
+      "loss": 0.5196,
+      "step": 1245
+    },
+    {
+      "epoch": 1.3954443614637788,
+      "grad_norm": 0.7973191142082214,
+      "learning_rate": 4.43525075442653e-06,
+      "loss": 0.5235,
+      "step": 1246
+    },
+    {
+      "epoch": 1.3965646004480956,
+      "grad_norm": 0.812990128993988,
+      "learning_rate": 4.4343037089189735e-06,
+      "loss": 0.5423,
+      "step": 1247
+    },
+    {
+      "epoch": 1.3976848394324122,
+      "grad_norm": 0.8209276795387268,
+      "learning_rate": 4.4333559713016736e-06,
+      "loss": 0.5169,
+      "step": 1248
+    },
+    {
+      "epoch": 1.398805078416729,
+      "grad_norm": 0.8271411657333374,
+      "learning_rate": 4.432407541913741e-06,
+      "loss": 0.5337,
+      "step": 1249
+    },
+    {
+      "epoch": 1.3999253174010455,
+      "grad_norm": 0.8020209670066833,
+      "learning_rate": 4.431458421094531e-06,
+      "loss": 0.5179,
+      "step": 1250
+    },
+    {
+      "epoch": 1.401045556385362,
+      "grad_norm": 0.8035420775413513,
+      "learning_rate": 4.430508609183647e-06,
+      "loss": 0.5416,
+      "step": 1251
+    },
+    {
+      "epoch": 1.402165795369679,
+      "grad_norm": 0.819017767906189,
+      "learning_rate": 4.429558106520939e-06,
+      "loss": 0.5261,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4032860343539955,
+      "grad_norm": 0.81986004114151,
+      "learning_rate": 4.4286069134465035e-06,
+      "loss": 0.5122,
+      "step": 1253
+    },
+    {
+      "epoch": 1.4044062733383122,
+      "grad_norm": 0.8470500111579895,
+      "learning_rate": 4.427655030300688e-06,
+      "loss": 0.5393,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4055265123226288,
+      "grad_norm": 0.8163276314735413,
+      "learning_rate": 4.426702457424082e-06,
+      "loss": 0.5475,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4066467513069454,
+      "grad_norm": 0.8219355344772339,
+      "learning_rate": 4.425749195157524e-06,
+      "loss": 0.5343,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4077669902912622,
+      "grad_norm": 0.8590656518936157,
+      "learning_rate": 4.424795243842099e-06,
+      "loss": 0.5649,
+      "step": 1257
+    },
+    {
+      "epoch": 1.4088872292755787,
+      "grad_norm": 0.837833046913147,
+      "learning_rate": 4.423840603819141e-06,
+      "loss": 0.5249,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4100074682598955,
+      "grad_norm": 0.8217041492462158,
+      "learning_rate": 4.422885275430225e-06,
+      "loss": 0.55,
+      "step": 1259
+    },
+    {
+      "epoch": 1.411127707244212,
+      "grad_norm": 0.8252221941947937,
+      "learning_rate": 4.421929259017176e-06,
+      "loss": 0.5684,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4122479462285287,
+      "grad_norm": 0.8646223545074463,
+      "learning_rate": 4.420972554922066e-06,
+      "loss": 0.5623,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4133681852128455,
+      "grad_norm": 0.8612502217292786,
+      "learning_rate": 4.42001516348721e-06,
+      "loss": 0.5597,
+      "step": 1262
+    },
+    {
+      "epoch": 1.414488424197162,
+      "grad_norm": 0.8005207180976868,
+      "learning_rate": 4.419057085055171e-06,
+      "loss": 0.5252,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4156086631814788,
+      "grad_norm": 0.813919723033905,
+      "learning_rate": 4.418098319968758e-06,
+      "loss": 0.5502,
+      "step": 1264
+    },
+    {
+      "epoch": 1.4167289021657954,
+      "grad_norm": 0.8239195942878723,
+      "learning_rate": 4.4171388685710234e-06,
+      "loss": 0.5229,
+      "step": 1265
+    },
+    {
+      "epoch": 1.417849141150112,
+      "grad_norm": 0.8362986445426941,
+      "learning_rate": 4.416178731205269e-06,
+      "loss": 0.5311,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4189693801344287,
+      "grad_norm": 0.8120007514953613,
+      "learning_rate": 4.415217908215038e-06,
+      "loss": 0.5479,
+      "step": 1267
+    },
+    {
+      "epoch": 1.4200896191187453,
+      "grad_norm": 0.811576783657074,
+      "learning_rate": 4.414256399944122e-06,
+      "loss": 0.5542,
+      "step": 1268
+    },
+    {
+      "epoch": 1.421209858103062,
+      "grad_norm": 0.8243828415870667,
+      "learning_rate": 4.413294206736555e-06,
+      "loss": 0.5555,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4223300970873787,
+      "grad_norm": 0.8122988343238831,
+      "learning_rate": 4.41233132893662e-06,
+      "loss": 0.5551,
+      "step": 1270
+    },
+    {
+      "epoch": 1.4234503360716952,
+      "grad_norm": 0.8304051756858826,
+      "learning_rate": 4.41136776688884e-06,
+      "loss": 0.5734,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4245705750560118,
+      "grad_norm": 0.8483263850212097,
+      "learning_rate": 4.410403520937987e-06,
+      "loss": 0.5328,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4256908140403286,
+      "grad_norm": 0.837712287902832,
+      "learning_rate": 4.409438591429075e-06,
+      "loss": 0.542,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4268110530246454,
+      "grad_norm": 0.8439350128173828,
+      "learning_rate": 4.408472978707365e-06,
+      "loss": 0.5523,
+      "step": 1274
+    },
+    {
+      "epoch": 1.427931292008962,
+      "grad_norm": 0.8609887957572937,
+      "learning_rate": 4.4075066831183605e-06,
+      "loss": 0.5462,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4290515309932785,
+      "grad_norm": 0.8378350734710693,
+      "learning_rate": 4.406539705007809e-06,
+      "loss": 0.5579,
+      "step": 1276
+    },
+    {
+      "epoch": 1.430171769977595,
+      "grad_norm": 0.8463708758354187,
+      "learning_rate": 4.4055720447217056e-06,
+      "loss": 0.5634,
+      "step": 1277
+    },
+    {
+      "epoch": 1.4312920089619119,
+      "grad_norm": 0.8240607976913452,
+      "learning_rate": 4.404603702606286e-06,
+      "loss": 0.5471,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4324122479462285,
+      "grad_norm": 0.8117276430130005,
+      "learning_rate": 4.40363467900803e-06,
+      "loss": 0.5614,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4335324869305452,
+      "grad_norm": 0.8322970271110535,
+      "learning_rate": 4.402664974273662e-06,
+      "loss": 0.5534,
+      "step": 1280
+    },
+    {
+      "epoch": 1.4346527259148618,
+      "grad_norm": 0.8382965326309204,
+      "learning_rate": 4.401694588750153e-06,
+      "loss": 0.5546,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4357729648991784,
+      "grad_norm": 0.8176775574684143,
+      "learning_rate": 4.400723522784712e-06,
+      "loss": 0.5236,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4368932038834952,
+      "grad_norm": 0.8426209688186646,
+      "learning_rate": 4.399751776724795e-06,
+      "loss": 0.5527,
+      "step": 1283
+    },
+    {
+      "epoch": 1.4380134428678117,
+      "grad_norm": 0.823796808719635,
+      "learning_rate": 4.398779350918102e-06,
+      "loss": 0.5662,
+      "step": 1284
+    },
+    {
+      "epoch": 1.4391336818521285,
+      "grad_norm": 0.8374935388565063,
+      "learning_rate": 4.397806245712573e-06,
+      "loss": 0.5726,
+      "step": 1285
+    },
+    {
+      "epoch": 1.440253920836445,
+      "grad_norm": 0.8143332600593567,
+      "learning_rate": 4.3968324614563965e-06,
+      "loss": 0.5077,
+      "step": 1286
+    },
+    {
+      "epoch": 1.4413741598207617,
+      "grad_norm": 0.8215734958648682,
+      "learning_rate": 4.395857998497996e-06,
+      "loss": 0.5138,
+      "step": 1287
+    },
+    {
+      "epoch": 1.4424943988050785,
+      "grad_norm": 0.7970778346061707,
+      "learning_rate": 4.394882857186045e-06,
+      "loss": 0.5088,
+      "step": 1288
+    },
+    {
+      "epoch": 1.443614637789395,
+      "grad_norm": 0.8270483613014221,
+      "learning_rate": 4.393907037869457e-06,
+      "loss": 0.5374,
+      "step": 1289
+    },
+    {
+      "epoch": 1.4447348767737118,
+      "grad_norm": 0.8275690078735352,
+      "learning_rate": 4.3929305408973875e-06,
+      "loss": 0.5122,
+      "step": 1290
+    },
+    {
+      "epoch": 1.4458551157580284,
+      "grad_norm": 0.8330255746841431,
+      "learning_rate": 4.3919533666192345e-06,
+      "loss": 0.5562,
+      "step": 1291
+    },
+    {
+      "epoch": 1.446975354742345,
+      "grad_norm": 0.8581027984619141,
+      "learning_rate": 4.390975515384641e-06,
+      "loss": 0.5542,
+      "step": 1292
+    },
+    {
+      "epoch": 1.4480955937266617,
+      "grad_norm": 0.844085156917572,
+      "learning_rate": 4.389996987543488e-06,
+      "loss": 0.5506,
+      "step": 1293
+    },
+    {
+      "epoch": 1.4492158327109783,
+      "grad_norm": 0.8119167685508728,
+      "learning_rate": 4.389017783445903e-06,
+      "loss": 0.5309,
+      "step": 1294
+    },
+    {
+      "epoch": 1.450336071695295,
+      "grad_norm": 0.8669208288192749,
+      "learning_rate": 4.388037903442251e-06,
+      "loss": 0.5691,
+      "step": 1295
+    },
+    {
+      "epoch": 1.4514563106796117,
+      "grad_norm": 0.8349128365516663,
+      "learning_rate": 4.387057347883143e-06,
+      "loss": 0.5426,
+      "step": 1296
+    },
+    {
+      "epoch": 1.4525765496639282,
+      "grad_norm": 0.8149973750114441,
+      "learning_rate": 4.386076117119429e-06,
+      "loss": 0.544,
+      "step": 1297
+    },
+    {
+      "epoch": 1.453696788648245,
+      "grad_norm": 0.8381718397140503,
+      "learning_rate": 4.385094211502201e-06,
+      "loss": 0.541,
+      "step": 1298
+    },
+    {
+      "epoch": 1.4548170276325616,
+      "grad_norm": 0.8309394717216492,
+      "learning_rate": 4.3841116313827924e-06,
+      "loss": 0.555,
+      "step": 1299
+    },
+    {
+      "epoch": 1.4559372666168784,
+      "grad_norm": 0.8401052951812744,
+      "learning_rate": 4.383128377112781e-06,
+      "loss": 0.5455,
+      "step": 1300
+    },
+    {
+      "epoch": 1.457057505601195,
+      "grad_norm": 0.8229027390480042,
+      "learning_rate": 4.38214444904398e-06,
+      "loss": 0.5317,
+      "step": 1301
+    },
+    {
+      "epoch": 1.4581777445855115,
+      "grad_norm": 0.822666347026825,
+      "learning_rate": 4.38115984752845e-06,
+      "loss": 0.5143,
+      "step": 1302
+    },
+    {
+      "epoch": 1.459297983569828,
+      "grad_norm": 0.8157971501350403,
+      "learning_rate": 4.3801745729184875e-06,
+      "loss": 0.5317,
+      "step": 1303
+    },
+    {
+      "epoch": 1.4604182225541449,
+      "grad_norm": 0.8346320390701294,
+      "learning_rate": 4.379188625566632e-06,
+      "loss": 0.5533,
+      "step": 1304
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.8360608816146851,
+      "learning_rate": 4.378202005825665e-06,
+      "loss": 0.5403,
+      "step": 1305
+    },
+    {
+      "epoch": 1.4626587005227782,
+      "grad_norm": 0.8309468030929565,
+      "learning_rate": 4.3772147140486045e-06,
+      "loss": 0.5028,
+      "step": 1306
+    },
+    {
+      "epoch": 1.4637789395070948,
+      "grad_norm": 0.8045680522918701,
+      "learning_rate": 4.376226750588714e-06,
+      "loss": 0.5153,
+      "step": 1307
+    },
+    {
+      "epoch": 1.4648991784914114,
+      "grad_norm": 0.826449990272522,
+      "learning_rate": 4.375238115799494e-06,
+      "loss": 0.5193,
+      "step": 1308
+    },
+    {
+      "epoch": 1.4660194174757282,
+      "grad_norm": 0.8211467862129211,
+      "learning_rate": 4.374248810034686e-06,
+      "loss": 0.5236,
+      "step": 1309
+    },
+    {
+      "epoch": 1.4671396564600447,
+      "grad_norm": 0.8300038576126099,
+      "learning_rate": 4.373258833648273e-06,
+      "loss": 0.554,
+      "step": 1310
+    },
+    {
+      "epoch": 1.4682598954443615,
+      "grad_norm": 0.852250874042511,
+      "learning_rate": 4.372268186994474e-06,
+      "loss": 0.5673,
+      "step": 1311
+    },
+    {
+      "epoch": 1.469380134428678,
+      "grad_norm": 0.8562779426574707,
+      "learning_rate": 4.3712768704277535e-06,
+      "loss": 0.5312,
+      "step": 1312
+    },
+    {
+      "epoch": 1.4705003734129947,
+      "grad_norm": 0.8487451672554016,
+      "learning_rate": 4.37028488430281e-06,
+      "loss": 0.542,
+      "step": 1313
+    },
+    {
+      "epoch": 1.4716206123973115,
+      "grad_norm": 0.8320393562316895,
+      "learning_rate": 4.369292228974587e-06,
+      "loss": 0.5256,
+      "step": 1314
+    },
+    {
+      "epoch": 1.472740851381628,
+      "grad_norm": 0.821002185344696,
+      "learning_rate": 4.368298904798263e-06,
+      "loss": 0.5286,
+      "step": 1315
+    },
+    {
+      "epoch": 1.4738610903659448,
+      "grad_norm": 0.8211376070976257,
+      "learning_rate": 4.367304912129258e-06,
+      "loss": 0.5692,
+      "step": 1316
+    },
+    {
+      "epoch": 1.4749813293502614,
+      "grad_norm": 0.8837286233901978,
+      "learning_rate": 4.3663102513232295e-06,
+      "loss": 0.5346,
+      "step": 1317
+    },
+    {
+      "epoch": 1.476101568334578,
+      "grad_norm": 0.8166071772575378,
+      "learning_rate": 4.365314922736077e-06,
+      "loss": 0.5528,
+      "step": 1318
+    },
+    {
+      "epoch": 1.4772218073188947,
+      "grad_norm": 0.8274430632591248,
+      "learning_rate": 4.364318926723937e-06,
+      "loss": 0.5464,
+      "step": 1319
+    },
+    {
+      "epoch": 1.4783420463032113,
+      "grad_norm": 0.8482112884521484,
+      "learning_rate": 4.3633222636431845e-06,
+      "loss": 0.5459,
+      "step": 1320
+    },
+    {
+      "epoch": 1.479462285287528,
+      "grad_norm": 0.8600017428398132,
+      "learning_rate": 4.362324933850435e-06,
+      "loss": 0.5567,
+      "step": 1321
+    },
+    {
+      "epoch": 1.4805825242718447,
+      "grad_norm": 0.8248361945152283,
+      "learning_rate": 4.361326937702538e-06,
+      "loss": 0.5479,
+      "step": 1322
+    },
+    {
+      "epoch": 1.4817027632561612,
+      "grad_norm": 0.8052260279655457,
+      "learning_rate": 4.3603282755565885e-06,
+      "loss": 0.5388,
+      "step": 1323
+    },
+    {
+      "epoch": 1.482823002240478,
+      "grad_norm": 0.8378350734710693,
+      "learning_rate": 4.359328947769914e-06,
+      "loss": 0.5475,
+      "step": 1324
+    },
+    {
+      "epoch": 1.4839432412247946,
+      "grad_norm": 0.845522403717041,
+      "learning_rate": 4.358328954700083e-06,
+      "loss": 0.5133,
+      "step": 1325
+    },
+    {
+      "epoch": 1.4850634802091114,
+      "grad_norm": 0.9849204421043396,
+      "learning_rate": 4.3573282967049e-06,
+      "loss": 0.5107,
+      "step": 1326
+    },
+    {
+      "epoch": 1.486183719193428,
+      "grad_norm": 0.8313608765602112,
+      "learning_rate": 4.356326974142411e-06,
+      "loss": 0.5737,
+      "step": 1327
+    },
+    {
+      "epoch": 1.4873039581777445,
+      "grad_norm": 0.8800959587097168,
+      "learning_rate": 4.355324987370895e-06,
+      "loss": 0.5312,
+      "step": 1328
+    },
+    {
+      "epoch": 1.4884241971620613,
+      "grad_norm": 0.8424291014671326,
+      "learning_rate": 4.354322336748872e-06,
+      "loss": 0.5197,
+      "step": 1329
+    },
+    {
+      "epoch": 1.4895444361463779,
+      "grad_norm": 0.820409893989563,
+      "learning_rate": 4.353319022635098e-06,
+      "loss": 0.5734,
+      "step": 1330
+    },
+    {
+      "epoch": 1.4906646751306947,
+      "grad_norm": 0.8500661849975586,
+      "learning_rate": 4.352315045388569e-06,
+      "loss": 0.5342,
+      "step": 1331
+    },
+    {
+      "epoch": 1.4917849141150112,
+      "grad_norm": 0.806735098361969,
+      "learning_rate": 4.351310405368513e-06,
+      "loss": 0.5213,
+      "step": 1332
+    },
+    {
+      "epoch": 1.4929051530993278,
+      "grad_norm": 0.838384211063385,
+      "learning_rate": 4.3503051029344025e-06,
+      "loss": 0.5479,
+      "step": 1333
+    },
+    {
+      "epoch": 1.4940253920836444,
+      "grad_norm": 0.8513844013214111,
+      "learning_rate": 4.34929913844594e-06,
+      "loss": 0.5406,
+      "step": 1334
+    },
+    {
+      "epoch": 1.4951456310679612,
+      "grad_norm": 0.8119661211967468,
+      "learning_rate": 4.348292512263068e-06,
+      "loss": 0.5481,
+      "step": 1335
+    },
+    {
+      "epoch": 1.496265870052278,
+      "grad_norm": 0.8307415843009949,
+      "learning_rate": 4.347285224745967e-06,
+      "loss": 0.5565,
+      "step": 1336
+    },
+    {
+      "epoch": 1.4973861090365945,
+      "grad_norm": 0.8296387195587158,
+      "learning_rate": 4.346277276255053e-06,
+      "loss": 0.5067,
+      "step": 1337
+    },
+    {
+      "epoch": 1.498506348020911,
+      "grad_norm": 0.8077414035797119,
+      "learning_rate": 4.345268667150976e-06,
+      "loss": 0.5209,
+      "step": 1338
+    },
+    {
+      "epoch": 1.4996265870052277,
+      "grad_norm": 0.7943417429924011,
+      "learning_rate": 4.344259397794626e-06,
+      "loss": 0.5048,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5007468259895445,
+      "grad_norm": 0.7765065431594849,
+      "learning_rate": 4.343249468547128e-06,
+      "loss": 0.5567,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5018670649738612,
+      "grad_norm": 0.868191123008728,
+      "learning_rate": 4.3422388797698435e-06,
+      "loss": 0.5479,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5029873039581778,
+      "grad_norm": 0.830331027507782,
+      "learning_rate": 4.341227631824368e-06,
+      "loss": 0.5492,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5041075429424944,
+      "grad_norm": 0.8496562242507935,
+      "learning_rate": 4.340215725072535e-06,
+      "loss": 0.5541,
+      "step": 1343
+    },
+    {
+      "epoch": 1.505227781926811,
+      "grad_norm": 0.8115208148956299,
+      "learning_rate": 4.339203159876413e-06,
+      "loss": 0.5429,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5063480209111277,
+      "grad_norm": 0.8215025067329407,
+      "learning_rate": 4.338189936598307e-06,
+      "loss": 0.5522,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5074682598954443,
+      "grad_norm": 0.8184052109718323,
+      "learning_rate": 4.3371760556007545e-06,
+      "loss": 0.527,
+      "step": 1346
+    },
+    {
+      "epoch": 1.508588498879761,
+      "grad_norm": 0.7937412858009338,
+      "learning_rate": 4.3361615172465325e-06,
+      "loss": 0.5325,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5097087378640777,
+      "grad_norm": 0.8070276379585266,
+      "learning_rate": 4.335146321898651e-06,
+      "loss": 0.5279,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5108289768483942,
+      "grad_norm": 0.8029081225395203,
+      "learning_rate": 4.334130469920354e-06,
+      "loss": 0.5491,
+      "step": 1349
+    },
+    {
+      "epoch": 1.511949215832711,
+      "grad_norm": 0.8653881549835205,
+      "learning_rate": 4.333113961675122e-06,
+      "loss": 0.5623,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5130694548170276,
+      "grad_norm": 0.8287846446037292,
+      "learning_rate": 4.33209679752667e-06,
+      "loss": 0.5579,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5141896938013444,
+      "grad_norm": 0.8319088220596313,
+      "learning_rate": 4.331078977838949e-06,
+      "loss": 0.5572,
+      "step": 1352
+    },
+    {
+      "epoch": 1.515309932785661,
+      "grad_norm": 0.8489014506340027,
+      "learning_rate": 4.3300605029761414e-06,
+      "loss": 0.5302,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5164301717699775,
+      "grad_norm": 0.8003896474838257,
+      "learning_rate": 4.329041373302667e-06,
+      "loss": 0.5601,
+      "step": 1354
+    },
+    {
+      "epoch": 1.517550410754294,
+      "grad_norm": 0.8308793306350708,
+      "learning_rate": 4.328021589183179e-06,
+      "loss": 0.5078,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5186706497386109,
+      "grad_norm": 0.8175726532936096,
+      "learning_rate": 4.327001150982563e-06,
+      "loss": 0.5434,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5197908887229277,
+      "grad_norm": 0.8181414008140564,
+      "learning_rate": 4.325980059065942e-06,
+      "loss": 0.545,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5209111277072442,
+      "grad_norm": 0.8243163228034973,
+      "learning_rate": 4.32495831379867e-06,
+      "loss": 0.5339,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5220313666915608,
+      "grad_norm": 0.8128981590270996,
+      "learning_rate": 4.3239359155463356e-06,
+      "loss": 0.5293,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5231516056758774,
+      "grad_norm": 0.8324003219604492,
+      "learning_rate": 4.3229128646747615e-06,
+      "loss": 0.5426,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5242718446601942,
+      "grad_norm": 0.8296977281570435,
+      "learning_rate": 4.321889161550005e-06,
+      "loss": 0.5626,
+      "step": 1361
+    },
+    {
+      "epoch": 1.525392083644511,
+      "grad_norm": 0.8470244407653809,
+      "learning_rate": 4.320864806538353e-06,
+      "loss": 0.5524,
+      "step": 1362
+    },
+    {
+      "epoch": 1.5265123226288275,
+      "grad_norm": 0.8672470450401306,
+      "learning_rate": 4.31983980000633e-06,
+      "loss": 0.5557,
+      "step": 1363
+    },
+    {
+      "epoch": 1.527632561613144,
+      "grad_norm": 0.833143413066864,
+      "learning_rate": 4.318814142320692e-06,
+      "loss": 0.5337,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5287528005974607,
+      "grad_norm": 0.8296066522598267,
+      "learning_rate": 4.317787833848426e-06,
+      "loss": 0.5637,
+      "step": 1365
+    },
+    {
+      "epoch": 1.5298730395817774,
+      "grad_norm": 0.8583776950836182,
+      "learning_rate": 4.316760874956756e-06,
+      "loss": 0.5429,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5309932785660942,
+      "grad_norm": 0.8318846821784973,
+      "learning_rate": 4.3157332660131335e-06,
+      "loss": 0.5584,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5321135175504108,
+      "grad_norm": 0.8015156984329224,
+      "learning_rate": 4.314705007385248e-06,
+      "loss": 0.5332,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5332337565347274,
+      "grad_norm": 0.8311407566070557,
+      "learning_rate": 4.313676099441017e-06,
+      "loss": 0.546,
+      "step": 1369
+    },
+    {
+      "epoch": 1.534353995519044,
+      "grad_norm": 0.8014186024665833,
+      "learning_rate": 4.312646542548593e-06,
+      "loss": 0.5109,
+      "step": 1370
+    },
+    {
+      "epoch": 1.5354742345033607,
+      "grad_norm": 0.8057787418365479,
+      "learning_rate": 4.31161633707636e-06,
+      "loss": 0.5369,
+      "step": 1371
+    },
+    {
+      "epoch": 1.5365944734876775,
+      "grad_norm": 0.8253255486488342,
+      "learning_rate": 4.310585483392933e-06,
+      "loss": 0.5478,
+      "step": 1372
+    },
+    {
+      "epoch": 1.537714712471994,
+      "grad_norm": 0.8418827056884766,
+      "learning_rate": 4.309553981867162e-06,
+      "loss": 0.5471,
+      "step": 1373
+    },
+    {
+      "epoch": 1.5388349514563107,
+      "grad_norm": 0.8555072546005249,
+      "learning_rate": 4.308521832868125e-06,
+      "loss": 0.5381,
+      "step": 1374
+    },
+    {
+      "epoch": 1.5399551904406272,
+      "grad_norm": 0.8227778077125549,
+      "learning_rate": 4.307489036765134e-06,
+      "loss": 0.55,
+      "step": 1375
+    },
+    {
+      "epoch": 1.541075429424944,
+      "grad_norm": 0.8473497033119202,
+      "learning_rate": 4.306455593927731e-06,
+      "loss": 0.5355,
+      "step": 1376
+    },
+    {
+      "epoch": 1.5421956684092606,
+      "grad_norm": 0.8366703987121582,
+      "learning_rate": 4.305421504725691e-06,
+      "loss": 0.5569,
+      "step": 1377
+    },
+    {
+      "epoch": 1.5433159073935774,
+      "grad_norm": 0.815157413482666,
+      "learning_rate": 4.30438676952902e-06,
+      "loss": 0.5252,
+      "step": 1378
+    },
+    {
+      "epoch": 1.544436146377894,
+      "grad_norm": 0.877655029296875,
+      "learning_rate": 4.303351388707955e-06,
+      "loss": 0.5393,
+      "step": 1379
+    },
+    {
+      "epoch": 1.5455563853622105,
+      "grad_norm": 0.8467420935630798,
+      "learning_rate": 4.302315362632963e-06,
+      "loss": 0.5891,
+      "step": 1380
+    },
+    {
+      "epoch": 1.5466766243465273,
+      "grad_norm": 0.8801563382148743,
+      "learning_rate": 4.301278691674742e-06,
+      "loss": 0.5599,
+      "step": 1381
+    },
+    {
+      "epoch": 1.5477968633308439,
+      "grad_norm": 0.8227587342262268,
+      "learning_rate": 4.300241376204223e-06,
+      "loss": 0.5304,
+      "step": 1382
+    },
+    {
+      "epoch": 1.5489171023151607,
+      "grad_norm": 0.852873682975769,
+      "learning_rate": 4.299203416592565e-06,
+      "loss": 0.5279,
+      "step": 1383
+    },
+    {
+      "epoch": 1.5500373412994772,
+      "grad_norm": 0.8275724649429321,
+      "learning_rate": 4.298164813211158e-06,
+      "loss": 0.5514,
+      "step": 1384
+    },
+    {
+      "epoch": 1.5511575802837938,
+      "grad_norm": 0.8230090737342834,
+      "learning_rate": 4.297125566431624e-06,
+      "loss": 0.5399,
+      "step": 1385
+    },
+    {
+      "epoch": 1.5522778192681104,
+      "grad_norm": 0.8874813914299011,
+      "learning_rate": 4.296085676625813e-06,
+      "loss": 0.5386,
+      "step": 1386
+    },
+    {
+      "epoch": 1.5533980582524272,
+      "grad_norm": 0.8467227816581726,
+      "learning_rate": 4.295045144165807e-06,
+      "loss": 0.5467,
+      "step": 1387
+    },
+    {
+      "epoch": 1.554518297236744,
+      "grad_norm": 0.8453990817070007,
+      "learning_rate": 4.2940039694239165e-06,
+      "loss": 0.5677,
+      "step": 1388
+    },
+    {
+      "epoch": 1.5556385362210605,
+      "grad_norm": 0.8536325097084045,
+      "learning_rate": 4.292962152772682e-06,
+      "loss": 0.5185,
+      "step": 1389
+    },
+    {
+      "epoch": 1.556758775205377,
+      "grad_norm": 0.8449810743331909,
+      "learning_rate": 4.291919694584875e-06,
+      "loss": 0.546,
+      "step": 1390
+    },
+    {
+      "epoch": 1.5578790141896937,
+      "grad_norm": 1.027535080909729,
+      "learning_rate": 4.290876595233495e-06,
+      "loss": 0.5516,
+      "step": 1391
+    },
+    {
+      "epoch": 1.5589992531740104,
+      "grad_norm": 0.8208884000778198,
+      "learning_rate": 4.289832855091772e-06,
+      "loss": 0.5452,
+      "step": 1392
+    },
+    {
+      "epoch": 1.5601194921583272,
+      "grad_norm": 0.8302054405212402,
+      "learning_rate": 4.288788474533164e-06,
+      "loss": 0.5374,
+      "step": 1393
+    },
+    {
+      "epoch": 1.5612397311426438,
+      "grad_norm": 0.8338841199874878,
+      "learning_rate": 4.28774345393136e-06,
+      "loss": 0.5286,
+      "step": 1394
+    },
+    {
+      "epoch": 1.5623599701269604,
+      "grad_norm": 0.8848025798797607,
+      "learning_rate": 4.286697793660276e-06,
+      "loss": 0.5503,
+      "step": 1395
+    },
+    {
+      "epoch": 1.563480209111277,
+      "grad_norm": 0.8843843936920166,
+      "learning_rate": 4.285651494094058e-06,
+      "loss": 0.5394,
+      "step": 1396
+    },
+    {
+      "epoch": 1.5646004480955937,
+      "grad_norm": 0.860697329044342,
+      "learning_rate": 4.2846045556070805e-06,
+      "loss": 0.5486,
+      "step": 1397
+    },
+    {
+      "epoch": 1.5657206870799105,
+      "grad_norm": 0.8341761231422424,
+      "learning_rate": 4.283556978573948e-06,
+      "loss": 0.5845,
+      "step": 1398
+    },
+    {
+      "epoch": 1.566840926064227,
+      "grad_norm": 0.8415078520774841,
+      "learning_rate": 4.282508763369489e-06,
+      "loss": 0.5396,
+      "step": 1399
+    },
+    {
+      "epoch": 1.5679611650485437,
+      "grad_norm": 0.8659593462944031,
+      "learning_rate": 4.2814599103687686e-06,
+      "loss": 0.5341,
+      "step": 1400
+    },
+    {
+      "epoch": 1.5690814040328602,
+      "grad_norm": 0.8313730955123901,
+      "learning_rate": 4.2804104199470696e-06,
+      "loss": 0.5388,
+      "step": 1401
+    },
+    {
+      "epoch": 1.570201643017177,
+      "grad_norm": 0.8137566447257996,
+      "learning_rate": 4.279360292479911e-06,
+      "loss": 0.5298,
+      "step": 1402
+    },
+    {
+      "epoch": 1.5713218820014938,
+      "grad_norm": 0.829147219657898,
+      "learning_rate": 4.2783095283430365e-06,
+      "loss": 0.5478,
+      "step": 1403
+    },
+    {
+      "epoch": 1.5724421209858104,
+      "grad_norm": 0.8064625859260559,
+      "learning_rate": 4.27725812791242e-06,
+      "loss": 0.5332,
+      "step": 1404
+    },
+    {
+      "epoch": 1.573562359970127,
+      "grad_norm": 0.828982949256897,
+      "learning_rate": 4.276206091564257e-06,
+      "loss": 0.5317,
+      "step": 1405
+    },
+    {
+      "epoch": 1.5746825989544435,
+      "grad_norm": 0.8150440454483032,
+      "learning_rate": 4.275153419674979e-06,
+      "loss": 0.5502,
+      "step": 1406
+    },
+    {
+      "epoch": 1.5758028379387603,
+      "grad_norm": 0.8187349438667297,
+      "learning_rate": 4.274100112621238e-06,
+      "loss": 0.5377,
+      "step": 1407
+    },
+    {
+      "epoch": 1.5769230769230769,
+      "grad_norm": 0.8216383457183838,
+      "learning_rate": 4.2730461707799174e-06,
+      "loss": 0.5413,
+      "step": 1408
+    },
+    {
+      "epoch": 1.5780433159073937,
+      "grad_norm": 0.8329434394836426,
+      "learning_rate": 4.271991594528126e-06,
+      "loss": 0.5372,
+      "step": 1409
+    },
+    {
+      "epoch": 1.5791635548917102,
+      "grad_norm": 0.8837087750434875,
+      "learning_rate": 4.270936384243198e-06,
+      "loss": 0.5187,
+      "step": 1410
+    },
+    {
+      "epoch": 1.5802837938760268,
+      "grad_norm": 0.851102888584137,
+      "learning_rate": 4.2698805403027e-06,
+      "loss": 0.5528,
+      "step": 1411
+    },
+    {
+      "epoch": 1.5814040328603436,
+      "grad_norm": 0.8246365785598755,
+      "learning_rate": 4.268824063084417e-06,
+      "loss": 0.5276,
+      "step": 1412
+    },
+    {
+      "epoch": 1.5825242718446602,
+      "grad_norm": 0.8388403654098511,
+      "learning_rate": 4.267766952966369e-06,
+      "loss": 0.5545,
+      "step": 1413
+    },
+    {
+      "epoch": 1.583644510828977,
+      "grad_norm": 0.8562852740287781,
+      "learning_rate": 4.2667092103267975e-06,
+      "loss": 0.5512,
+      "step": 1414
+    },
+    {
+      "epoch": 1.5847647498132935,
+      "grad_norm": 0.8856998682022095,
+      "learning_rate": 4.2656508355441705e-06,
+      "loss": 0.5534,
+      "step": 1415
+    },
+    {
+      "epoch": 1.58588498879761,
+      "grad_norm": 0.9003668427467346,
+      "learning_rate": 4.264591828997184e-06,
+      "loss": 0.5061,
+      "step": 1416
+    },
+    {
+      "epoch": 1.5870052277819267,
+      "grad_norm": 0.8696342706680298,
+      "learning_rate": 4.263532191064759e-06,
+      "loss": 0.5448,
+      "step": 1417
+    },
+    {
+      "epoch": 1.5881254667662434,
+      "grad_norm": 0.8576314449310303,
+      "learning_rate": 4.262471922126044e-06,
+      "loss": 0.5117,
+      "step": 1418
+    },
+    {
+      "epoch": 1.5892457057505602,
+      "grad_norm": 0.9128873348236084,
+      "learning_rate": 4.261411022560411e-06,
+      "loss": 0.5376,
+      "step": 1419
+    },
+    {
+      "epoch": 1.5903659447348768,
+      "grad_norm": 0.8280308246612549,
+      "learning_rate": 4.2603494927474575e-06,
+      "loss": 0.5463,
+      "step": 1420
+    },
+    {
+      "epoch": 1.5914861837191934,
+      "grad_norm": 0.8322044610977173,
+      "learning_rate": 4.2592873330670085e-06,
+      "loss": 0.5596,
+      "step": 1421
+    },
+    {
+      "epoch": 1.59260642270351,
+      "grad_norm": 0.8214216232299805,
+      "learning_rate": 4.2582245438991135e-06,
+      "loss": 0.5135,
+      "step": 1422
+    },
+    {
+      "epoch": 1.5937266616878267,
+      "grad_norm": 0.8392521142959595,
+      "learning_rate": 4.257161125624047e-06,
+      "loss": 0.5411,
+      "step": 1423
+    },
+    {
+      "epoch": 1.5948469006721435,
+      "grad_norm": 0.8548365831375122,
+      "learning_rate": 4.256097078622309e-06,
+      "loss": 0.5287,
+      "step": 1424
+    },
+    {
+      "epoch": 1.59596713965646,
+      "grad_norm": 0.8191155791282654,
+      "learning_rate": 4.255032403274624e-06,
+      "loss": 0.5239,
+      "step": 1425
+    },
+    {
+      "epoch": 1.5970873786407767,
+      "grad_norm": 0.8004139065742493,
+      "learning_rate": 4.253967099961942e-06,
+      "loss": 0.5355,
+      "step": 1426
+    },
+    {
+      "epoch": 1.5982076176250932,
+      "grad_norm": 0.8321914672851562,
+      "learning_rate": 4.2529011690654366e-06,
+      "loss": 0.5495,
+      "step": 1427
+    },
+    {
+      "epoch": 1.59932785660941,
+      "grad_norm": 0.8587489724159241,
+      "learning_rate": 4.2518346109665074e-06,
+      "loss": 0.5502,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6004480955937268,
+      "grad_norm": 0.8176190853118896,
+      "learning_rate": 4.250767426046777e-06,
+      "loss": 0.5353,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6015683345780434,
+      "grad_norm": 0.8270806074142456,
+      "learning_rate": 4.249699614688093e-06,
+      "loss": 0.5449,
+      "step": 1430
+    },
+    {
+      "epoch": 1.60268857356236,
+      "grad_norm": 0.8037596344947815,
+      "learning_rate": 4.248631177272527e-06,
+      "loss": 0.5775,
+      "step": 1431
+    },
+    {
+      "epoch": 1.6038088125466765,
+      "grad_norm": 0.8505509495735168,
+      "learning_rate": 4.247562114182375e-06,
+      "loss": 0.5231,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6049290515309933,
+      "grad_norm": 0.8228225708007812,
+      "learning_rate": 4.246492425800157e-06,
+      "loss": 0.5336,
+      "step": 1433
+    },
+    {
+      "epoch": 1.60604929051531,
+      "grad_norm": 0.8823798298835754,
+      "learning_rate": 4.245422112508615e-06,
+      "loss": 0.5549,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6071695294996267,
+      "grad_norm": 0.8225086331367493,
+      "learning_rate": 4.244351174690717e-06,
+      "loss": 0.5347,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6082897684839432,
+      "grad_norm": 0.8210791945457458,
+      "learning_rate": 4.2432796127296535e-06,
+      "loss": 0.5349,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6094100074682598,
+      "grad_norm": 0.8335378766059875,
+      "learning_rate": 4.242207427008837e-06,
+      "loss": 0.5575,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6105302464525766,
+      "grad_norm": 0.8326417207717896,
+      "learning_rate": 4.241134617911905e-06,
+      "loss": 0.5422,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6116504854368932,
+      "grad_norm": 0.8305699825286865,
+      "learning_rate": 4.240061185822717e-06,
+      "loss": 0.553,
+      "step": 1439
+    },
+    {
+      "epoch": 1.61277072442121,
+      "grad_norm": 0.8145079016685486,
+      "learning_rate": 4.238987131125357e-06,
+      "loss": 0.5172,
+      "step": 1440
+    },
+    {
+      "epoch": 1.6138909634055265,
+      "grad_norm": 0.8260290622711182,
+      "learning_rate": 4.237912454204129e-06,
+      "loss": 0.5287,
+      "step": 1441
+    },
+    {
+      "epoch": 1.615011202389843,
+      "grad_norm": 0.812061071395874,
+      "learning_rate": 4.236837155443562e-06,
+      "loss": 0.5332,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6161314413741599,
+      "grad_norm": 0.7981239557266235,
+      "learning_rate": 4.235761235228407e-06,
+      "loss": 0.5563,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6172516803584764,
+      "grad_norm": 0.8434090614318848,
+      "learning_rate": 4.234684693943637e-06,
+      "loss": 0.5575,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6183719193427932,
+      "grad_norm": 0.8346796035766602,
+      "learning_rate": 4.233607531974447e-06,
+      "loss": 0.5328,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6194921583271098,
+      "grad_norm": 0.8098893761634827,
+      "learning_rate": 4.232529749706254e-06,
+      "loss": 0.5467,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6206123973114264,
+      "grad_norm": 0.8355947732925415,
+      "learning_rate": 4.231451347524699e-06,
+      "loss": 0.5432,
+      "step": 1447
+    },
+    {
+      "epoch": 1.621732636295743,
+      "grad_norm": 0.8329986333847046,
+      "learning_rate": 4.230372325815641e-06,
+      "loss": 0.5622,
+      "step": 1448
+    },
+    {
+      "epoch": 1.6228528752800597,
+      "grad_norm": 0.8184260725975037,
+      "learning_rate": 4.229292684965163e-06,
+      "loss": 0.5514,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6239731142643765,
+      "grad_norm": 0.8356477618217468,
+      "learning_rate": 4.2282124253595715e-06,
+      "loss": 0.546,
+      "step": 1450
+    },
+    {
+      "epoch": 1.625093353248693,
+      "grad_norm": 0.8311543464660645,
+      "learning_rate": 4.227131547385391e-06,
+      "loss": 0.5234,
+      "step": 1451
+    },
+    {
+      "epoch": 1.6262135922330097,
+      "grad_norm": 0.8295907974243164,
+      "learning_rate": 4.226050051429367e-06,
+      "loss": 0.4932,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6273338312173262,
+      "grad_norm": 0.7894620895385742,
+      "learning_rate": 4.22496793787847e-06,
+      "loss": 0.5641,
+      "step": 1453
+    },
+    {
+      "epoch": 1.628454070201643,
+      "grad_norm": 0.8411605954170227,
+      "learning_rate": 4.223885207119889e-06,
+      "loss": 0.5417,
+      "step": 1454
+    },
+    {
+      "epoch": 1.6295743091859598,
+      "grad_norm": 0.8090710639953613,
+      "learning_rate": 4.222801859541031e-06,
+      "loss": 0.5445,
+      "step": 1455
+    },
+    {
+      "epoch": 1.6306945481702764,
+      "grad_norm": 0.8332550525665283,
+      "learning_rate": 4.221717895529531e-06,
+      "loss": 0.5471,
+      "step": 1456
+    },
+    {
+      "epoch": 1.631814787154593,
+      "grad_norm": 0.9103831648826599,
+      "learning_rate": 4.220633315473236e-06,
+      "loss": 0.5384,
+      "step": 1457
+    },
+    {
+      "epoch": 1.6329350261389095,
+      "grad_norm": 0.8271507620811462,
+      "learning_rate": 4.21954811976022e-06,
+      "loss": 0.5408,
+      "step": 1458
+    },
+    {
+      "epoch": 1.6340552651232263,
+      "grad_norm": 0.8201518058776855,
+      "learning_rate": 4.218462308778776e-06,
+      "loss": 0.5633,
+      "step": 1459
+    },
+    {
+      "epoch": 1.635175504107543,
+      "grad_norm": 0.8605701327323914,
+      "learning_rate": 4.217375882917414e-06,
+      "loss": 0.5498,
+      "step": 1460
+    },
+    {
+      "epoch": 1.6362957430918597,
+      "grad_norm": 0.8359277844429016,
+      "learning_rate": 4.2162888425648665e-06,
+      "loss": 0.5467,
+      "step": 1461
+    },
+    {
+      "epoch": 1.6374159820761762,
+      "grad_norm": 0.8181623816490173,
+      "learning_rate": 4.215201188110086e-06,
+      "loss": 0.5219,
+      "step": 1462
+    },
+    {
+      "epoch": 1.6385362210604928,
+      "grad_norm": 0.8375784754753113,
+      "learning_rate": 4.214112919942243e-06,
+      "loss": 0.526,
+      "step": 1463
+    },
+    {
+      "epoch": 1.6396564600448096,
+      "grad_norm": 0.7823356986045837,
+      "learning_rate": 4.21302403845073e-06,
+      "loss": 0.5204,
+      "step": 1464
+    },
+    {
+      "epoch": 1.6407766990291264,
+      "grad_norm": 2.3768980503082275,
+      "learning_rate": 4.211934544025158e-06,
+      "loss": 0.5606,
+      "step": 1465
+    },
+    {
+      "epoch": 1.641896938013443,
+      "grad_norm": 0.84914231300354,
+      "learning_rate": 4.210844437055356e-06,
+      "loss": 0.565,
+      "step": 1466
+    },
+    {
+      "epoch": 1.6430171769977595,
+      "grad_norm": 0.8609841465950012,
+      "learning_rate": 4.209753717931373e-06,
+      "loss": 0.5757,
+      "step": 1467
+    },
+    {
+      "epoch": 1.644137415982076,
+      "grad_norm": 0.8694177269935608,
+      "learning_rate": 4.208662387043479e-06,
+      "loss": 0.5502,
+      "step": 1468
+    },
+    {
+      "epoch": 1.6452576549663929,
+      "grad_norm": 0.9158186316490173,
+      "learning_rate": 4.20757044478216e-06,
+      "loss": 0.5209,
+      "step": 1469
+    },
+    {
+      "epoch": 1.6463778939507094,
+      "grad_norm": 0.8625397682189941,
+      "learning_rate": 4.206477891538121e-06,
+      "loss": 0.5609,
+      "step": 1470
+    },
+    {
+      "epoch": 1.6474981329350262,
+      "grad_norm": 0.8475672006607056,
+      "learning_rate": 4.205384727702288e-06,
+      "loss": 0.5441,
+      "step": 1471
+    },
+    {
+      "epoch": 1.6486183719193428,
+      "grad_norm": 0.818719744682312,
+      "learning_rate": 4.204290953665804e-06,
+      "loss": 0.5266,
+      "step": 1472
+    },
+    {
+      "epoch": 1.6497386109036594,
+      "grad_norm": 0.8269298672676086,
+      "learning_rate": 4.203196569820029e-06,
+      "loss": 0.5263,
+      "step": 1473
+    },
+    {
+      "epoch": 1.6508588498879762,
+      "grad_norm": 0.7849137187004089,
+      "learning_rate": 4.2021015765565444e-06,
+      "loss": 0.5495,
+      "step": 1474
+    },
+    {
+      "epoch": 1.6519790888722927,
+      "grad_norm": 0.8485032320022583,
+      "learning_rate": 4.201005974267146e-06,
+      "loss": 0.5512,
+      "step": 1475
+    },
+    {
+      "epoch": 1.6530993278566095,
+      "grad_norm": 0.8280136585235596,
+      "learning_rate": 4.199909763343849e-06,
+      "loss": 0.5227,
+      "step": 1476
+    },
+    {
+      "epoch": 1.654219566840926,
+      "grad_norm": 0.8437508344650269,
+      "learning_rate": 4.198812944178889e-06,
+      "loss": 0.5635,
+      "step": 1477
+    },
+    {
+      "epoch": 1.6553398058252426,
+      "grad_norm": 0.8449869751930237,
+      "learning_rate": 4.197715517164716e-06,
+      "loss": 0.5127,
+      "step": 1478
+    },
+    {
+      "epoch": 1.6564600448095592,
+      "grad_norm": 0.8209198117256165,
+      "learning_rate": 4.196617482693996e-06,
+      "loss": 0.5483,
+      "step": 1479
+    },
+    {
+      "epoch": 1.657580283793876,
+      "grad_norm": 0.8873897194862366,
+      "learning_rate": 4.195518841159618e-06,
+      "loss": 0.5537,
+      "step": 1480
+    },
+    {
+      "epoch": 1.6587005227781928,
+      "grad_norm": 0.8520996570587158,
+      "learning_rate": 4.194419592954684e-06,
+      "loss": 0.5313,
+      "step": 1481
+    },
+    {
+      "epoch": 1.6598207617625094,
+      "grad_norm": 0.8172207474708557,
+      "learning_rate": 4.1933197384725135e-06,
+      "loss": 0.54,
+      "step": 1482
+    },
+    {
+      "epoch": 1.660941000746826,
+      "grad_norm": 0.8447046279907227,
+      "learning_rate": 4.192219278106643e-06,
+      "loss": 0.5205,
+      "step": 1483
+    },
+    {
+      "epoch": 1.6620612397311425,
+      "grad_norm": 0.8111680746078491,
+      "learning_rate": 4.191118212250827e-06,
+      "loss": 0.5247,
+      "step": 1484
+    },
+    {
+      "epoch": 1.6631814787154593,
+      "grad_norm": 0.8495075106620789,
+      "learning_rate": 4.190016541299036e-06,
+      "loss": 0.5079,
+      "step": 1485
+    },
+    {
+      "epoch": 1.664301717699776,
+      "grad_norm": 0.8128763437271118,
+      "learning_rate": 4.188914265645456e-06,
+      "loss": 0.5415,
+      "step": 1486
+    },
+    {
+      "epoch": 1.6654219566840927,
+      "grad_norm": 0.8739402890205383,
+      "learning_rate": 4.1878113856844925e-06,
+      "loss": 0.5658,
+      "step": 1487
+    },
+    {
+      "epoch": 1.6665421956684092,
+      "grad_norm": 0.8443946242332458,
+      "learning_rate": 4.186707901810762e-06,
+      "loss": 0.537,
+      "step": 1488
+    },
+    {
+      "epoch": 1.6676624346527258,
+      "grad_norm": 0.8422719836235046,
+      "learning_rate": 4.1856038144191024e-06,
+      "loss": 0.5278,
+      "step": 1489
+    },
+    {
+      "epoch": 1.6687826736370426,
+      "grad_norm": 0.8393810987472534,
+      "learning_rate": 4.184499123904564e-06,
+      "loss": 0.5241,
+      "step": 1490
+    },
+    {
+      "epoch": 1.6699029126213594,
+      "grad_norm": 0.8822279572486877,
+      "learning_rate": 4.183393830662415e-06,
+      "loss": 0.558,
+      "step": 1491
+    },
+    {
+      "epoch": 1.671023151605676,
+      "grad_norm": 0.8313372731208801,
+      "learning_rate": 4.182287935088138e-06,
+      "loss": 0.538,
+      "step": 1492
+    },
+    {
+      "epoch": 1.6721433905899925,
+      "grad_norm": 0.8423752188682556,
+      "learning_rate": 4.181181437577429e-06,
+      "loss": 0.5096,
+      "step": 1493
+    },
+    {
+      "epoch": 1.673263629574309,
+      "grad_norm": 0.7942532896995544,
+      "learning_rate": 4.180074338526208e-06,
+      "loss": 0.5513,
+      "step": 1494
+    },
+    {
+      "epoch": 1.6743838685586259,
+      "grad_norm": 0.843720555305481,
+      "learning_rate": 4.178966638330598e-06,
+      "loss": 0.5704,
+      "step": 1495
+    },
+    {
+      "epoch": 1.6755041075429427,
+      "grad_norm": 0.8642749786376953,
+      "learning_rate": 4.177858337386946e-06,
+      "loss": 0.5316,
+      "step": 1496
+    },
+    {
+      "epoch": 1.6766243465272592,
+      "grad_norm": 0.8187930583953857,
+      "learning_rate": 4.176749436091811e-06,
+      "loss": 0.5251,
+      "step": 1497
+    },
+    {
+      "epoch": 1.6777445855115758,
+      "grad_norm": 0.8171431422233582,
+      "learning_rate": 4.175639934841967e-06,
+      "loss": 0.5336,
+      "step": 1498
+    },
+    {
+      "epoch": 1.6788648244958924,
+      "grad_norm": 0.8119639158248901,
+      "learning_rate": 4.174529834034403e-06,
+      "loss": 0.5314,
+      "step": 1499
+    },
+    {
+      "epoch": 1.6799850634802092,
+      "grad_norm": 0.8608358502388,
+      "learning_rate": 4.173419134066321e-06,
+      "loss": 0.5341,
+      "step": 1500
+    },
+    {
+      "epoch": 1.6811053024645257,
+      "grad_norm": 0.8582233786582947,
+      "learning_rate": 4.17230783533514e-06,
+      "loss": 0.5484,
+      "step": 1501
+    },
+    {
+      "epoch": 1.6822255414488425,
+      "grad_norm": 0.8320852518081665,
+      "learning_rate": 4.17119593823849e-06,
+      "loss": 0.51,
+      "step": 1502
+    },
+    {
+      "epoch": 1.683345780433159,
+      "grad_norm": 0.8468641638755798,
+      "learning_rate": 4.170083443174218e-06,
+      "loss": 0.5572,
+      "step": 1503
+    },
+    {
+      "epoch": 1.6844660194174756,
+      "grad_norm": 0.8393252491950989,
+      "learning_rate": 4.168970350540385e-06,
+      "loss": 0.5348,
+      "step": 1504
+    },
+    {
+      "epoch": 1.6855862584017924,
+      "grad_norm": 0.8738113045692444,
+      "learning_rate": 4.167856660735262e-06,
+      "loss": 0.5524,
+      "step": 1505
+    },
+    {
+      "epoch": 1.686706497386109,
+      "grad_norm": 0.8472318649291992,
+      "learning_rate": 4.166742374157339e-06,
+      "loss": 0.5133,
+      "step": 1506
+    },
+    {
+      "epoch": 1.6878267363704258,
+      "grad_norm": 0.8055902123451233,
+      "learning_rate": 4.165627491205314e-06,
+      "loss": 0.511,
+      "step": 1507
+    },
+    {
+      "epoch": 1.6889469753547424,
+      "grad_norm": 0.8492250442504883,
+      "learning_rate": 4.164512012278104e-06,
+      "loss": 0.5621,
+      "step": 1508
+    },
+    {
+      "epoch": 1.690067214339059,
+      "grad_norm": 0.8291061520576477,
+      "learning_rate": 4.163395937774836e-06,
+      "loss": 0.5533,
+      "step": 1509
+    },
+    {
+      "epoch": 1.6911874533233755,
+      "grad_norm": 0.8331051468849182,
+      "learning_rate": 4.16227926809485e-06,
+      "loss": 0.5192,
+      "step": 1510
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.8490992188453674,
+      "learning_rate": 4.1611620036376985e-06,
+      "loss": 0.5357,
+      "step": 1511
+    },
+    {
+      "epoch": 1.693427931292009,
+      "grad_norm": 0.8369888663291931,
+      "learning_rate": 4.160044144803151e-06,
+      "loss": 0.5342,
+      "step": 1512
+    },
+    {
+      "epoch": 1.6945481702763256,
+      "grad_norm": 0.8247721195220947,
+      "learning_rate": 4.158925691991184e-06,
+      "loss": 0.5368,
+      "step": 1513
+    },
+    {
+      "epoch": 1.6956684092606422,
+      "grad_norm": 0.8113073706626892,
+      "learning_rate": 4.1578066456019885e-06,
+      "loss": 0.5555,
+      "step": 1514
+    },
+    {
+      "epoch": 1.6967886482449588,
+      "grad_norm": 0.8781044483184814,
+      "learning_rate": 4.15668700603597e-06,
+      "loss": 0.53,
+      "step": 1515
+    },
+    {
+      "epoch": 1.6979088872292756,
+      "grad_norm": 0.8180976510047913,
+      "learning_rate": 4.1555667736937455e-06,
+      "loss": 0.5474,
+      "step": 1516
+    },
+    {
+      "epoch": 1.6990291262135924,
+      "grad_norm": 0.8118385076522827,
+      "learning_rate": 4.154445948976141e-06,
+      "loss": 0.5292,
+      "step": 1517
+    },
+    {
+      "epoch": 1.700149365197909,
+      "grad_norm": 0.8206310272216797,
+      "learning_rate": 4.1533245322842e-06,
+      "loss": 0.5406,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7012696041822255,
+      "grad_norm": 0.8399862051010132,
+      "learning_rate": 4.152202524019171e-06,
+      "loss": 0.5417,
+      "step": 1519
+    },
+    {
+      "epoch": 1.702389843166542,
+      "grad_norm": 0.8626686334609985,
+      "learning_rate": 4.151079924582519e-06,
+      "loss": 0.5104,
+      "step": 1520
+    },
+    {
+      "epoch": 1.7035100821508589,
+      "grad_norm": 0.8447391986846924,
+      "learning_rate": 4.149956734375922e-06,
+      "loss": 0.5332,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7046303211351757,
+      "grad_norm": 0.8259961605072021,
+      "learning_rate": 4.148832953801262e-06,
+      "loss": 0.5242,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7057505601194922,
+      "grad_norm": 0.8265607357025146,
+      "learning_rate": 4.14770858326064e-06,
+      "loss": 0.5204,
+      "step": 1523
+    },
+    {
+      "epoch": 1.7068707991038088,
+      "grad_norm": 0.8091963529586792,
+      "learning_rate": 4.146583623156365e-06,
+      "loss": 0.5253,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7079910380881254,
+      "grad_norm": 0.830488920211792,
+      "learning_rate": 4.145458073890955e-06,
+      "loss": 0.5344,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7091112770724421,
+      "grad_norm": 0.8003833889961243,
+      "learning_rate": 4.144331935867144e-06,
+      "loss": 0.5294,
+      "step": 1526
+    },
+    {
+      "epoch": 1.710231516056759,
+      "grad_norm": 0.8505309224128723,
+      "learning_rate": 4.143205209487869e-06,
+      "loss": 0.531,
+      "step": 1527
+    },
+    {
+      "epoch": 1.7113517550410755,
+      "grad_norm": 0.8250753879547119,
+      "learning_rate": 4.1420778951562866e-06,
+      "loss": 0.5421,
+      "step": 1528
+    },
+    {
+      "epoch": 1.712471994025392,
+      "grad_norm": 0.8263447284698486,
+      "learning_rate": 4.140949993275756e-06,
+      "loss": 0.5461,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7135922330097086,
+      "grad_norm": 0.8574968576431274,
+      "learning_rate": 4.13982150424985e-06,
+      "loss": 0.5499,
+      "step": 1530
+    },
+    {
+      "epoch": 1.7147124719940254,
+      "grad_norm": 0.8031488060951233,
+      "learning_rate": 4.138692428482353e-06,
+      "loss": 0.5163,
+      "step": 1531
+    },
+    {
+      "epoch": 1.715832710978342,
+      "grad_norm": 0.8235338926315308,
+      "learning_rate": 4.137562766377256e-06,
+      "loss": 0.5346,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7169529499626588,
+      "grad_norm": 0.8231131434440613,
+      "learning_rate": 4.136432518338762e-06,
+      "loss": 0.5442,
+      "step": 1533
+    },
+    {
+      "epoch": 1.7180731889469754,
+      "grad_norm": 0.8224930763244629,
+      "learning_rate": 4.135301684771283e-06,
+      "loss": 0.545,
+      "step": 1534
+    },
+    {
+      "epoch": 1.719193427931292,
+      "grad_norm": 0.7999832034111023,
+      "learning_rate": 4.1341702660794395e-06,
+      "loss": 0.5226,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7203136669156085,
+      "grad_norm": 0.8502464294433594,
+      "learning_rate": 4.1330382626680635e-06,
+      "loss": 0.5513,
+      "step": 1536
+    },
+    {
+      "epoch": 1.7214339058999253,
+      "grad_norm": 0.8491352200508118,
+      "learning_rate": 4.131905674942196e-06,
+      "loss": 0.5178,
+      "step": 1537
+    },
+    {
+      "epoch": 1.722554144884242,
+      "grad_norm": 0.8294553756713867,
+      "learning_rate": 4.1307725033070835e-06,
+      "loss": 0.5415,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7236743838685586,
+      "grad_norm": 0.8187465667724609,
+      "learning_rate": 4.129638748168187e-06,
+      "loss": 0.5575,
+      "step": 1539
+    },
+    {
+      "epoch": 1.7247946228528752,
+      "grad_norm": 0.8333837985992432,
+      "learning_rate": 4.128504409931172e-06,
+      "loss": 0.5359,
+      "step": 1540
+    },
+    {
+      "epoch": 1.7259148618371918,
+      "grad_norm": 0.8331719636917114,
+      "learning_rate": 4.127369489001914e-06,
+      "loss": 0.5513,
+      "step": 1541
+    },
+    {
+      "epoch": 1.7270351008215086,
+      "grad_norm": 0.8731279969215393,
+      "learning_rate": 4.126233985786497e-06,
+      "loss": 0.5464,
+      "step": 1542
+    },
+    {
+      "epoch": 1.7281553398058254,
+      "grad_norm": 0.8208237886428833,
+      "learning_rate": 4.125097900691213e-06,
+      "loss": 0.5614,
+      "step": 1543
+    },
+    {
+      "epoch": 1.729275578790142,
+      "grad_norm": 0.8664867281913757,
+      "learning_rate": 4.1239612341225636e-06,
+      "loss": 0.5328,
+      "step": 1544
+    },
+    {
+      "epoch": 1.7303958177744585,
+      "grad_norm": 0.810331404209137,
+      "learning_rate": 4.122823986487257e-06,
+      "loss": 0.5303,
+      "step": 1545
+    },
+    {
+      "epoch": 1.731516056758775,
+      "grad_norm": 0.8227166533470154,
+      "learning_rate": 4.121686158192209e-06,
+      "loss": 0.5506,
+      "step": 1546
+    },
+    {
+      "epoch": 1.7326362957430919,
+      "grad_norm": 0.8276739716529846,
+      "learning_rate": 4.120547749644545e-06,
+      "loss": 0.5355,
+      "step": 1547
+    },
+    {
+      "epoch": 1.7337565347274086,
+      "grad_norm": 0.8266296982765198,
+      "learning_rate": 4.1194087612515965e-06,
+      "loss": 0.5661,
+      "step": 1548
+    },
+    {
+      "epoch": 1.7348767737117252,
+      "grad_norm": 0.8601844310760498,
+      "learning_rate": 4.118269193420903e-06,
+      "loss": 0.5089,
+      "step": 1549
+    },
+    {
+      "epoch": 1.7359970126960418,
+      "grad_norm": 0.8149077296257019,
+      "learning_rate": 4.117129046560211e-06,
+      "loss": 0.5679,
+      "step": 1550
+    },
+    {
+      "epoch": 1.7371172516803584,
+      "grad_norm": 0.8324862122535706,
+      "learning_rate": 4.115988321077474e-06,
+      "loss": 0.5541,
+      "step": 1551
+    },
+    {
+      "epoch": 1.7382374906646751,
+      "grad_norm": 0.8243516087532043,
+      "learning_rate": 4.114847017380854e-06,
+      "loss": 0.5364,
+      "step": 1552
+    },
+    {
+      "epoch": 1.739357729648992,
+      "grad_norm": 0.8432444930076599,
+      "learning_rate": 4.1137051358787185e-06,
+      "loss": 0.5574,
+      "step": 1553
+    },
+    {
+      "epoch": 1.7404779686333085,
+      "grad_norm": 0.8552629351615906,
+      "learning_rate": 4.112562676979641e-06,
+      "loss": 0.5233,
+      "step": 1554
+    },
+    {
+      "epoch": 1.741598207617625,
+      "grad_norm": 0.8027449250221252,
+      "learning_rate": 4.111419641092404e-06,
+      "loss": 0.565,
+      "step": 1555
+    },
+    {
+      "epoch": 1.7427184466019416,
+      "grad_norm": 0.8295990228652954,
+      "learning_rate": 4.110276028625995e-06,
+      "loss": 0.5056,
+      "step": 1556
+    },
+    {
+      "epoch": 1.7438386855862584,
+      "grad_norm": 0.8083453178405762,
+      "learning_rate": 4.1091318399896065e-06,
+      "loss": 0.5476,
+      "step": 1557
+    },
+    {
+      "epoch": 1.7449589245705752,
+      "grad_norm": 0.8240988254547119,
+      "learning_rate": 4.107987075592641e-06,
+      "loss": 0.5275,
+      "step": 1558
+    },
+    {
+      "epoch": 1.7460791635548918,
+      "grad_norm": 0.8225562572479248,
+      "learning_rate": 4.1068417358447024e-06,
+      "loss": 0.5498,
+      "step": 1559
+    },
+    {
+      "epoch": 1.7471994025392084,
+      "grad_norm": 0.8333079814910889,
+      "learning_rate": 4.105695821155604e-06,
+      "loss": 0.5372,
+      "step": 1560
+    },
+    {
+      "epoch": 1.748319641523525,
+      "grad_norm": 0.8118395209312439,
+      "learning_rate": 4.104549331935362e-06,
+      "loss": 0.5467,
+      "step": 1561
+    },
+    {
+      "epoch": 1.7494398805078417,
+      "grad_norm": 0.812199056148529,
+      "learning_rate": 4.103402268594202e-06,
+      "loss": 0.5412,
+      "step": 1562
+    },
+    {
+      "epoch": 1.7505601194921583,
+      "grad_norm": 0.8283849954605103,
+      "learning_rate": 4.102254631542549e-06,
+      "loss": 0.551,
+      "step": 1563
+    },
+    {
+      "epoch": 1.751680358476475,
+      "grad_norm": 0.8368584513664246,
+      "learning_rate": 4.101106421191041e-06,
+      "loss": 0.5585,
+      "step": 1564
+    },
+    {
+      "epoch": 1.7528005974607916,
+      "grad_norm": 0.824888288974762,
+      "learning_rate": 4.099957637950516e-06,
+      "loss": 0.5421,
+      "step": 1565
+    },
+    {
+      "epoch": 1.7539208364451082,
+      "grad_norm": 0.8353846669197083,
+      "learning_rate": 4.0988082822320165e-06,
+      "loss": 0.5398,
+      "step": 1566
+    },
+    {
+      "epoch": 1.7550410754294248,
+      "grad_norm": 0.8127483129501343,
+      "learning_rate": 4.097658354446792e-06,
+      "loss": 0.5484,
+      "step": 1567
+    },
+    {
+      "epoch": 1.7561613144137416,
+      "grad_norm": 0.83433598279953,
+      "learning_rate": 4.096507855006297e-06,
+      "loss": 0.52,
+      "step": 1568
+    },
+    {
+      "epoch": 1.7572815533980584,
+      "grad_norm": 0.8318457007408142,
+      "learning_rate": 4.095356784322189e-06,
+      "loss": 0.5435,
+      "step": 1569
+    },
+    {
+      "epoch": 1.758401792382375,
+      "grad_norm": 0.84352707862854,
+      "learning_rate": 4.09420514280633e-06,
+      "loss": 0.5318,
+      "step": 1570
+    },
+    {
+      "epoch": 1.7595220313666915,
+      "grad_norm": 0.8551493287086487,
+      "learning_rate": 4.093052930870788e-06,
+      "loss": 0.5502,
+      "step": 1571
+    },
+    {
+      "epoch": 1.760642270351008,
+      "grad_norm": 0.8463614583015442,
+      "learning_rate": 4.091900148927833e-06,
+      "loss": 0.546,
+      "step": 1572
+    },
+    {
+      "epoch": 1.7617625093353249,
+      "grad_norm": 0.8330878019332886,
+      "learning_rate": 4.090746797389941e-06,
+      "loss": 0.5365,
+      "step": 1573
+    },
+    {
+      "epoch": 1.7628827483196416,
+      "grad_norm": 0.8032608032226562,
+      "learning_rate": 4.0895928766697885e-06,
+      "loss": 0.5196,
+      "step": 1574
+    },
+    {
+      "epoch": 1.7640029873039582,
+      "grad_norm": 0.8519574403762817,
+      "learning_rate": 4.088438387180261e-06,
+      "loss": 0.5473,
+      "step": 1575
+    },
+    {
+      "epoch": 1.7651232262882748,
+      "grad_norm": 0.8020384907722473,
+      "learning_rate": 4.08728332933444e-06,
+      "loss": 0.5086,
+      "step": 1576
+    },
+    {
+      "epoch": 1.7662434652725914,
+      "grad_norm": 0.8228726387023926,
+      "learning_rate": 4.086127703545618e-06,
+      "loss": 0.5573,
+      "step": 1577
+    },
+    {
+      "epoch": 1.7673637042569081,
+      "grad_norm": 0.8214651346206665,
+      "learning_rate": 4.084971510227286e-06,
+      "loss": 0.5533,
+      "step": 1578
+    },
+    {
+      "epoch": 1.768483943241225,
+      "grad_norm": 0.8368930220603943,
+      "learning_rate": 4.08381474979314e-06,
+      "loss": 0.5333,
+      "step": 1579
+    },
+    {
+      "epoch": 1.7696041822255415,
+      "grad_norm": 0.8381990194320679,
+      "learning_rate": 4.082657422657079e-06,
+      "loss": 0.5473,
+      "step": 1580
+    },
+    {
+      "epoch": 1.770724421209858,
+      "grad_norm": 0.796229362487793,
+      "learning_rate": 4.0814995292332035e-06,
+      "loss": 0.5128,
+      "step": 1581
+    },
+    {
+      "epoch": 1.7718446601941746,
+      "grad_norm": 0.8201373815536499,
+      "learning_rate": 4.080341069935817e-06,
+      "loss": 0.5469,
+      "step": 1582
+    },
+    {
+      "epoch": 1.7729648991784914,
+      "grad_norm": 0.8479568958282471,
+      "learning_rate": 4.079182045179425e-06,
+      "loss": 0.5328,
+      "step": 1583
+    },
+    {
+      "epoch": 1.7740851381628082,
+      "grad_norm": 0.87519371509552,
+      "learning_rate": 4.078022455378738e-06,
+      "loss": 0.5291,
+      "step": 1584
+    },
+    {
+      "epoch": 1.7752053771471248,
+      "grad_norm": 0.8594712018966675,
+      "learning_rate": 4.076862300948665e-06,
+      "loss": 0.5579,
+      "step": 1585
+    },
+    {
+      "epoch": 1.7763256161314414,
+      "grad_norm": 0.8597174286842346,
+      "learning_rate": 4.075701582304321e-06,
+      "loss": 0.5325,
+      "step": 1586
+    },
+    {
+      "epoch": 1.777445855115758,
+      "grad_norm": 0.7930446863174438,
+      "learning_rate": 4.074540299861018e-06,
+      "loss": 0.53,
+      "step": 1587
+    },
+    {
+      "epoch": 1.7785660941000747,
+      "grad_norm": 0.8580162525177002,
+      "learning_rate": 4.073378454034273e-06,
+      "loss": 0.5043,
+      "step": 1588
+    },
+    {
+      "epoch": 1.7796863330843915,
+      "grad_norm": 0.8035150170326233,
+      "learning_rate": 4.0722160452398054e-06,
+      "loss": 0.546,
+      "step": 1589
+    },
+    {
+      "epoch": 1.780806572068708,
+      "grad_norm": 0.8199040293693542,
+      "learning_rate": 4.0710530738935336e-06,
+      "loss": 0.5545,
+      "step": 1590
+    },
+    {
+      "epoch": 1.7819268110530246,
+      "grad_norm": 0.8304824829101562,
+      "learning_rate": 4.069889540411579e-06,
+      "loss": 0.5173,
+      "step": 1591
+    },
+    {
+      "epoch": 1.7830470500373412,
+      "grad_norm": 0.8335815072059631,
+      "learning_rate": 4.068725445210263e-06,
+      "loss": 0.5091,
+      "step": 1592
+    },
+    {
+      "epoch": 1.784167289021658,
+      "grad_norm": 0.8112631440162659,
+      "learning_rate": 4.067560788706107e-06,
+      "loss": 0.5397,
+      "step": 1593
+    },
+    {
+      "epoch": 1.7852875280059746,
+      "grad_norm": 0.8443027138710022,
+      "learning_rate": 4.066395571315837e-06,
+      "loss": 0.5341,
+      "step": 1594
+    },
+    {
+      "epoch": 1.7864077669902914,
+      "grad_norm": 0.8510046005249023,
+      "learning_rate": 4.065229793456377e-06,
+      "loss": 0.5669,
+      "step": 1595
+    },
+    {
+      "epoch": 1.787528005974608,
+      "grad_norm": 0.8445265889167786,
+      "learning_rate": 4.0640634555448496e-06,
+      "loss": 0.5349,
+      "step": 1596
+    },
+    {
+      "epoch": 1.7886482449589245,
+      "grad_norm": 0.8498575687408447,
+      "learning_rate": 4.062896557998582e-06,
+      "loss": 0.5549,
+      "step": 1597
+    },
+    {
+      "epoch": 1.789768483943241,
+      "grad_norm": 0.8104247450828552,
+      "learning_rate": 4.0617291012351e-06,
+      "loss": 0.507,
+      "step": 1598
+    },
+    {
+      "epoch": 1.7908887229275579,
+      "grad_norm": 0.7849128842353821,
+      "learning_rate": 4.060561085672128e-06,
+      "loss": 0.5209,
+      "step": 1599
+    },
+    {
+      "epoch": 1.7920089619118746,
+      "grad_norm": 0.8311539888381958,
+      "learning_rate": 4.059392511727592e-06,
+      "loss": 0.5432,
+      "step": 1600
+    },
+    {
+      "epoch": 1.7931292008961912,
+      "grad_norm": 0.8311768174171448,
+      "learning_rate": 4.0582233798196185e-06,
+      "loss": 0.5458,
+      "step": 1601
+    },
+    {
+      "epoch": 1.7942494398805078,
+      "grad_norm": 0.8668813705444336,
+      "learning_rate": 4.057053690366531e-06,
+      "loss": 0.5252,
+      "step": 1602
+    },
+    {
+      "epoch": 1.7953696788648243,
+      "grad_norm": 0.8628605604171753,
+      "learning_rate": 4.0558834437868536e-06,
+      "loss": 0.5246,
+      "step": 1603
+    },
+    {
+      "epoch": 1.7964899178491411,
+      "grad_norm": 0.8353387117385864,
+      "learning_rate": 4.054712640499312e-06,
+      "loss": 0.5666,
+      "step": 1604
+    },
+    {
+      "epoch": 1.797610156833458,
+      "grad_norm": 0.8165217638015747,
+      "learning_rate": 4.053541280922828e-06,
+      "loss": 0.5054,
+      "step": 1605
+    },
+    {
+      "epoch": 1.7987303958177745,
+      "grad_norm": 0.8124107718467712,
+      "learning_rate": 4.052369365476524e-06,
+      "loss": 0.5548,
+      "step": 1606
+    },
+    {
+      "epoch": 1.799850634802091,
+      "grad_norm": 0.8459861874580383,
+      "learning_rate": 4.051196894579721e-06,
+      "loss": 0.5398,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8009708737864076,
+      "grad_norm": 0.8248313069343567,
+      "learning_rate": 4.050023868651938e-06,
+      "loss": 0.5538,
+      "step": 1608
+    },
+    {
+      "epoch": 1.8020911127707244,
+      "grad_norm": 0.8126117587089539,
+      "learning_rate": 4.048850288112895e-06,
+      "loss": 0.5538,
+      "step": 1609
+    },
+    {
+      "epoch": 1.8032113517550412,
+      "grad_norm": 0.8366143107414246,
+      "learning_rate": 4.047676153382506e-06,
+      "loss": 0.495,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8043315907393578,
+      "grad_norm": 0.7833600044250488,
+      "learning_rate": 4.046501464880888e-06,
+      "loss": 0.5553,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8054518297236744,
+      "grad_norm": 0.8468241095542908,
+      "learning_rate": 4.045326223028355e-06,
+      "loss": 0.5166,
+      "step": 1612
+    },
+    {
+      "epoch": 1.806572068707991,
+      "grad_norm": 0.8318121433258057,
+      "learning_rate": 4.044150428245416e-06,
+      "loss": 0.5451,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8076923076923077,
+      "grad_norm": 0.852816641330719,
+      "learning_rate": 4.042974080952782e-06,
+      "loss": 0.561,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8088125466766245,
+      "grad_norm": 0.8474749326705933,
+      "learning_rate": 4.041797181571358e-06,
+      "loss": 0.5503,
+      "step": 1615
+    },
+    {
+      "epoch": 1.809932785660941,
+      "grad_norm": 0.8417901992797852,
+      "learning_rate": 4.040619730522249e-06,
+      "loss": 0.5342,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8110530246452576,
+      "grad_norm": 0.791199266910553,
+      "learning_rate": 4.039441728226757e-06,
+      "loss": 0.5099,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8121732636295742,
+      "grad_norm": 0.8117905259132385,
+      "learning_rate": 4.0382631751063804e-06,
+      "loss": 0.5572,
+      "step": 1618
+    },
+    {
+      "epoch": 1.813293502613891,
+      "grad_norm": 0.8146982789039612,
+      "learning_rate": 4.037084071582816e-06,
+      "loss": 0.5264,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8144137415982076,
+      "grad_norm": 0.8424848914146423,
+      "learning_rate": 4.035904418077955e-06,
+      "loss": 0.5204,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8155339805825244,
+      "grad_norm": 0.8367023468017578,
+      "learning_rate": 4.034724215013889e-06,
+      "loss": 0.5185,
+      "step": 1621
+    },
+    {
+      "epoch": 1.816654219566841,
+      "grad_norm": 0.8297321200370789,
+      "learning_rate": 4.0335434628129025e-06,
+      "loss": 0.5693,
+      "step": 1622
+    },
+    {
+      "epoch": 1.8177744585511575,
+      "grad_norm": 0.8517071008682251,
+      "learning_rate": 4.03236216189748e-06,
+      "loss": 0.5769,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8188946975354743,
+      "grad_norm": 0.8563759326934814,
+      "learning_rate": 4.031180312690301e-06,
+      "loss": 0.5452,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8200149365197908,
+      "grad_norm": 0.8202133178710938,
+      "learning_rate": 4.029997915614238e-06,
+      "loss": 0.5263,
+      "step": 1625
+    },
+    {
+      "epoch": 1.8211351755041076,
+      "grad_norm": 0.8332318663597107,
+      "learning_rate": 4.028814971092365e-06,
+      "loss": 0.5736,
+      "step": 1626
+    },
+    {
+      "epoch": 1.8222554144884242,
+      "grad_norm": 0.8340821266174316,
+      "learning_rate": 4.027631479547949e-06,
+      "loss": 0.5138,
+      "step": 1627
+    },
+    {
+      "epoch": 1.8233756534727408,
+      "grad_norm": 0.7971771359443665,
+      "learning_rate": 4.026447441404452e-06,
+      "loss": 0.5253,
+      "step": 1628
+    },
+    {
+      "epoch": 1.8244958924570573,
+      "grad_norm": 0.8240051865577698,
+      "learning_rate": 4.025262857085532e-06,
+      "loss": 0.5102,
+      "step": 1629
+    },
+    {
+      "epoch": 1.8256161314413741,
+      "grad_norm": 0.8214726448059082,
+      "learning_rate": 4.024077727015045e-06,
+      "loss": 0.5374,
+      "step": 1630
+    },
+    {
+      "epoch": 1.826736370425691,
+      "grad_norm": 0.8141801953315735,
+      "learning_rate": 4.02289205161704e-06,
+      "loss": 0.5187,
+      "step": 1631
+    },
+    {
+      "epoch": 1.8278566094100075,
+      "grad_norm": 0.8089074492454529,
+      "learning_rate": 4.02170583131576e-06,
+      "loss": 0.552,
+      "step": 1632
+    },
+    {
+      "epoch": 1.828976848394324,
+      "grad_norm": 0.8477803468704224,
+      "learning_rate": 4.020519066535645e-06,
+      "loss": 0.5334,
+      "step": 1633
+    },
+    {
+      "epoch": 1.8300970873786406,
+      "grad_norm": 0.8391149044036865,
+      "learning_rate": 4.019331757701329e-06,
+      "loss": 0.5253,
+      "step": 1634
+    },
+    {
+      "epoch": 1.8312173263629574,
+      "grad_norm": 0.8368405103683472,
+      "learning_rate": 4.018143905237643e-06,
+      "loss": 0.5265,
+      "step": 1635
+    },
+    {
+      "epoch": 1.8323375653472742,
+      "grad_norm": 0.8110945224761963,
+      "learning_rate": 4.016955509569605e-06,
+      "loss": 0.5556,
+      "step": 1636
+    },
+    {
+      "epoch": 1.8334578043315908,
+      "grad_norm": 0.8260793089866638,
+      "learning_rate": 4.015766571122438e-06,
+      "loss": 0.5588,
+      "step": 1637
+    },
+    {
+      "epoch": 1.8345780433159073,
+      "grad_norm": 0.8496139645576477,
+      "learning_rate": 4.0145770903215525e-06,
+      "loss": 0.5297,
+      "step": 1638
+    },
+    {
+      "epoch": 1.835698282300224,
+      "grad_norm": 0.8399412035942078,
+      "learning_rate": 4.013387067592553e-06,
+      "loss": 0.5402,
+      "step": 1639
+    },
+    {
+      "epoch": 1.8368185212845407,
+      "grad_norm": 0.8328200578689575,
+      "learning_rate": 4.012196503361242e-06,
+      "loss": 0.525,
+      "step": 1640
+    },
+    {
+      "epoch": 1.8379387602688575,
+      "grad_norm": 0.8379107713699341,
+      "learning_rate": 4.01100539805361e-06,
+      "loss": 0.5741,
+      "step": 1641
+    },
+    {
+      "epoch": 1.839058999253174,
+      "grad_norm": 0.8842681646347046,
+      "learning_rate": 4.009813752095847e-06,
+      "loss": 0.5449,
+      "step": 1642
+    },
+    {
+      "epoch": 1.8401792382374906,
+      "grad_norm": 0.8259784579277039,
+      "learning_rate": 4.008621565914332e-06,
+      "loss": 0.545,
+      "step": 1643
+    },
+    {
+      "epoch": 1.8412994772218072,
+      "grad_norm": 0.8219314813613892,
+      "learning_rate": 4.0074288399356394e-06,
+      "loss": 0.5463,
+      "step": 1644
+    },
+    {
+      "epoch": 1.842419716206124,
+      "grad_norm": 0.8393583297729492,
+      "learning_rate": 4.006235574586537e-06,
+      "loss": 0.5381,
+      "step": 1645
+    },
+    {
+      "epoch": 1.8435399551904408,
+      "grad_norm": 0.8115900158882141,
+      "learning_rate": 4.005041770293985e-06,
+      "loss": 0.5264,
+      "step": 1646
+    },
+    {
+      "epoch": 1.8446601941747574,
+      "grad_norm": 0.7949982285499573,
+      "learning_rate": 4.003847427485135e-06,
+      "loss": 0.5305,
+      "step": 1647
+    },
+    {
+      "epoch": 1.845780433159074,
+      "grad_norm": 0.8323104381561279,
+      "learning_rate": 4.002652546587334e-06,
+      "loss": 0.5218,
+      "step": 1648
+    },
+    {
+      "epoch": 1.8469006721433905,
+      "grad_norm": 0.8135440349578857,
+      "learning_rate": 4.001457128028118e-06,
+      "loss": 0.5555,
+      "step": 1649
+    },
+    {
+      "epoch": 1.8480209111277073,
+      "grad_norm": 0.8250274062156677,
+      "learning_rate": 4.00026117223522e-06,
+      "loss": 0.5608,
+      "step": 1650
+    },
+    {
+      "epoch": 1.8491411501120238,
+      "grad_norm": 0.8387907147407532,
+      "learning_rate": 3.999064679636564e-06,
+      "loss": 0.5507,
+      "step": 1651
+    },
+    {
+      "epoch": 1.8502613890963406,
+      "grad_norm": 0.8458473682403564,
+      "learning_rate": 3.997867650660261e-06,
+      "loss": 0.5633,
+      "step": 1652
+    },
+    {
+      "epoch": 1.8513816280806572,
+      "grad_norm": 0.8356611132621765,
+      "learning_rate": 3.99667008573462e-06,
+      "loss": 0.5467,
+      "step": 1653
+    },
+    {
+      "epoch": 1.8525018670649738,
+      "grad_norm": 0.8319884538650513,
+      "learning_rate": 3.99547198528814e-06,
+      "loss": 0.5245,
+      "step": 1654
+    },
+    {
+      "epoch": 1.8536221060492906,
+      "grad_norm": 0.8177763223648071,
+      "learning_rate": 3.99427334974951e-06,
+      "loss": 0.5418,
+      "step": 1655
+    },
+    {
+      "epoch": 1.8547423450336071,
+      "grad_norm": 0.8282372951507568,
+      "learning_rate": 3.9930741795476114e-06,
+      "loss": 0.5273,
+      "step": 1656
+    },
+    {
+      "epoch": 1.855862584017924,
+      "grad_norm": 0.822182297706604,
+      "learning_rate": 3.991874475111518e-06,
+      "loss": 0.5441,
+      "step": 1657
+    },
+    {
+      "epoch": 1.8569828230022405,
+      "grad_norm": 0.8694227337837219,
+      "learning_rate": 3.990674236870494e-06,
+      "loss": 0.5492,
+      "step": 1658
+    },
+    {
+      "epoch": 1.858103061986557,
+      "grad_norm": 0.8233626484870911,
+      "learning_rate": 3.989473465253995e-06,
+      "loss": 0.5377,
+      "step": 1659
+    },
+    {
+      "epoch": 1.8592233009708736,
+      "grad_norm": 0.8524294495582581,
+      "learning_rate": 3.988272160691665e-06,
+      "loss": 0.5188,
+      "step": 1660
+    },
+    {
+      "epoch": 1.8603435399551904,
+      "grad_norm": 0.8165991306304932,
+      "learning_rate": 3.987070323613343e-06,
+      "loss": 0.5407,
+      "step": 1661
+    },
+    {
+      "epoch": 1.8614637789395072,
+      "grad_norm": 0.8025708198547363,
+      "learning_rate": 3.9858679544490535e-06,
+      "loss": 0.5267,
+      "step": 1662
+    },
+    {
+      "epoch": 1.8625840179238238,
+      "grad_norm": 0.7899367809295654,
+      "learning_rate": 3.984665053629016e-06,
+      "loss": 0.5652,
+      "step": 1663
+    },
+    {
+      "epoch": 1.8637042569081403,
+      "grad_norm": 0.8292130827903748,
+      "learning_rate": 3.983461621583638e-06,
+      "loss": 0.5166,
+      "step": 1664
+    },
+    {
+      "epoch": 1.864824495892457,
+      "grad_norm": 0.8118554949760437,
+      "learning_rate": 3.982257658743518e-06,
+      "loss": 0.5058,
+      "step": 1665
+    },
+    {
+      "epoch": 1.8659447348767737,
+      "grad_norm": 0.8407183289527893,
+      "learning_rate": 3.9810531655394415e-06,
+      "loss": 0.5383,
+      "step": 1666
+    },
+    {
+      "epoch": 1.8670649738610905,
+      "grad_norm": 0.8522183895111084,
+      "learning_rate": 3.979848142402388e-06,
+      "loss": 0.5174,
+      "step": 1667
+    },
+    {
+      "epoch": 1.868185212845407,
+      "grad_norm": 0.827687680721283,
+      "learning_rate": 3.978642589763525e-06,
+      "loss": 0.533,
+      "step": 1668
+    },
+    {
+      "epoch": 1.8693054518297236,
+      "grad_norm": 0.8289827704429626,
+      "learning_rate": 3.977436508054207e-06,
+      "loss": 0.5342,
+      "step": 1669
+    },
+    {
+      "epoch": 1.8704256908140402,
+      "grad_norm": 0.803909957408905,
+      "learning_rate": 3.976229897705981e-06,
+      "loss": 0.5312,
+      "step": 1670
+    },
+    {
+      "epoch": 1.871545929798357,
+      "grad_norm": 0.8095699548721313,
+      "learning_rate": 3.975022759150584e-06,
+      "loss": 0.4865,
+      "step": 1671
+    },
+    {
+      "epoch": 1.8726661687826738,
+      "grad_norm": 0.8200787305831909,
+      "learning_rate": 3.973815092819937e-06,
+      "loss": 0.5401,
+      "step": 1672
+    },
+    {
+      "epoch": 1.8737864077669903,
+      "grad_norm": 0.8103435635566711,
+      "learning_rate": 3.972606899146155e-06,
+      "loss": 0.5535,
+      "step": 1673
+    },
+    {
+      "epoch": 1.874906646751307,
+      "grad_norm": 0.8428035378456116,
+      "learning_rate": 3.9713981785615385e-06,
+      "loss": 0.5437,
+      "step": 1674
+    },
+    {
+      "epoch": 1.8760268857356235,
+      "grad_norm": 0.8183536529541016,
+      "learning_rate": 3.970188931498578e-06,
+      "loss": 0.5282,
+      "step": 1675
+    },
+    {
+      "epoch": 1.8771471247199403,
+      "grad_norm": 0.8357205390930176,
+      "learning_rate": 3.968979158389952e-06,
+      "loss": 0.5431,
+      "step": 1676
+    },
+    {
+      "epoch": 1.878267363704257,
+      "grad_norm": 0.8381440043449402,
+      "learning_rate": 3.9677688596685285e-06,
+      "loss": 0.5373,
+      "step": 1677
+    },
+    {
+      "epoch": 1.8793876026885736,
+      "grad_norm": 0.8339812159538269,
+      "learning_rate": 3.966558035767359e-06,
+      "loss": 0.5426,
+      "step": 1678
+    },
+    {
+      "epoch": 1.8805078416728902,
+      "grad_norm": 0.821755588054657,
+      "learning_rate": 3.96534668711969e-06,
+      "loss": 0.5035,
+      "step": 1679
+    },
+    {
+      "epoch": 1.8816280806572068,
+      "grad_norm": 0.8362539410591125,
+      "learning_rate": 3.96413481415895e-06,
+      "loss": 0.5392,
+      "step": 1680
+    },
+    {
+      "epoch": 1.8827483196415236,
+      "grad_norm": 0.8103832602500916,
+      "learning_rate": 3.962922417318758e-06,
+      "loss": 0.5537,
+      "step": 1681
+    },
+    {
+      "epoch": 1.8838685586258401,
+      "grad_norm": 0.8064588904380798,
+      "learning_rate": 3.961709497032918e-06,
+      "loss": 0.5397,
+      "step": 1682
+    },
+    {
+      "epoch": 1.884988797610157,
+      "grad_norm": 0.8112618923187256,
+      "learning_rate": 3.960496053735424e-06,
+      "loss": 0.5337,
+      "step": 1683
+    },
+    {
+      "epoch": 1.8861090365944735,
+      "grad_norm": 0.8196645975112915,
+      "learning_rate": 3.959282087860456e-06,
+      "loss": 0.5607,
+      "step": 1684
+    },
+    {
+      "epoch": 1.88722927557879,
+      "grad_norm": 0.8481451272964478,
+      "learning_rate": 3.958067599842381e-06,
+      "loss": 0.5439,
+      "step": 1685
+    },
+    {
+      "epoch": 1.8883495145631068,
+      "grad_norm": 0.8333935141563416,
+      "learning_rate": 3.9568525901157525e-06,
+      "loss": 0.5695,
+      "step": 1686
+    },
+    {
+      "epoch": 1.8894697535474234,
+      "grad_norm": 0.8329671621322632,
+      "learning_rate": 3.955637059115311e-06,
+      "loss": 0.526,
+      "step": 1687
+    },
+    {
+      "epoch": 1.8905899925317402,
+      "grad_norm": 0.8062114715576172,
+      "learning_rate": 3.954421007275984e-06,
+      "loss": 0.5299,
+      "step": 1688
+    },
+    {
+      "epoch": 1.8917102315160568,
+      "grad_norm": 0.8448646664619446,
+      "learning_rate": 3.9532044350328825e-06,
+      "loss": 0.5409,
+      "step": 1689
+    },
+    {
+      "epoch": 1.8928304705003733,
+      "grad_norm": 0.8065194487571716,
+      "learning_rate": 3.95198734282131e-06,
+      "loss": 0.5319,
+      "step": 1690
+    },
+    {
+      "epoch": 1.89395070948469,
+      "grad_norm": 0.8324974775314331,
+      "learning_rate": 3.950769731076748e-06,
+      "loss": 0.5018,
+      "step": 1691
+    },
+    {
+      "epoch": 1.8950709484690067,
+      "grad_norm": 0.8186408877372742,
+      "learning_rate": 3.949551600234871e-06,
+      "loss": 0.5073,
+      "step": 1692
+    },
+    {
+      "epoch": 1.8961911874533235,
+      "grad_norm": 0.8118941783905029,
+      "learning_rate": 3.948332950731535e-06,
+      "loss": 0.5357,
+      "step": 1693
+    },
+    {
+      "epoch": 1.89731142643764,
+      "grad_norm": 0.7879785895347595,
+      "learning_rate": 3.947113783002782e-06,
+      "loss": 0.5113,
+      "step": 1694
+    },
+    {
+      "epoch": 1.8984316654219566,
+      "grad_norm": 0.826366126537323,
+      "learning_rate": 3.9458940974848415e-06,
+      "loss": 0.5432,
+      "step": 1695
+    },
+    {
+      "epoch": 1.8995519044062732,
+      "grad_norm": 0.8095139265060425,
+      "learning_rate": 3.944673894614127e-06,
+      "loss": 0.5322,
+      "step": 1696
+    },
+    {
+      "epoch": 1.90067214339059,
+      "grad_norm": 0.8329222202301025,
+      "learning_rate": 3.943453174827236e-06,
+      "loss": 0.5352,
+      "step": 1697
+    },
+    {
+      "epoch": 1.9017923823749068,
+      "grad_norm": 0.8235338926315308,
+      "learning_rate": 3.942231938560953e-06,
+      "loss": 0.5404,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9029126213592233,
+      "grad_norm": 0.8362364768981934,
+      "learning_rate": 3.941010186252247e-06,
+      "loss": 0.5373,
+      "step": 1699
+    },
+    {
+      "epoch": 1.90403286034354,
+      "grad_norm": 0.8214544653892517,
+      "learning_rate": 3.939787918338269e-06,
+      "loss": 0.5123,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9051530993278565,
+      "grad_norm": 0.8045482635498047,
+      "learning_rate": 3.938565135256359e-06,
+      "loss": 0.4988,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9062733383121733,
+      "grad_norm": 0.7857670187950134,
+      "learning_rate": 3.9373418374440354e-06,
+      "loss": 0.524,
+      "step": 1702
+    },
+    {
+      "epoch": 1.90739357729649,
+      "grad_norm": 0.84732586145401,
+      "learning_rate": 3.936118025339007e-06,
+      "loss": 0.5146,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9085138162808066,
+      "grad_norm": 0.82671058177948,
+      "learning_rate": 3.9348936993791645e-06,
+      "loss": 0.5369,
+      "step": 1704
+    },
+    {
+      "epoch": 1.9096340552651232,
+      "grad_norm": 0.8431946039199829,
+      "learning_rate": 3.933668860002579e-06,
+      "loss": 0.5433,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9107542942494398,
+      "grad_norm": 0.8344864249229431,
+      "learning_rate": 3.932443507647509e-06,
+      "loss": 0.5607,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9118745332337566,
+      "grad_norm": 0.8002555966377258,
+      "learning_rate": 3.9312176427523965e-06,
+      "loss": 0.5378,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9129947722180733,
+      "grad_norm": 0.7870469093322754,
+      "learning_rate": 3.9299912657558656e-06,
+      "loss": 0.5147,
+      "step": 1708
+    },
+    {
+      "epoch": 1.91411501120239,
+      "grad_norm": 0.8041394948959351,
+      "learning_rate": 3.928764377096723e-06,
+      "loss": 0.5302,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9152352501867065,
+      "grad_norm": 0.8553950786590576,
+      "learning_rate": 3.927536977213962e-06,
+      "loss": 0.5389,
+      "step": 1710
+    },
+    {
+      "epoch": 1.916355489171023,
+      "grad_norm": 0.8105883002281189,
+      "learning_rate": 3.926309066546754e-06,
+      "loss": 0.5367,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9174757281553398,
+      "grad_norm": 0.8396196961402893,
+      "learning_rate": 3.925080645534457e-06,
+      "loss": 0.5253,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9185959671396564,
+      "grad_norm": 0.8068451881408691,
+      "learning_rate": 3.923851714616609e-06,
+      "loss": 0.537,
+      "step": 1713
+    },
+    {
+      "epoch": 1.9197162061239732,
+      "grad_norm": 0.8422108292579651,
+      "learning_rate": 3.922622274232933e-06,
+      "loss": 0.5659,
+      "step": 1714
+    },
+    {
+      "epoch": 1.9208364451082898,
+      "grad_norm": 0.8337264657020569,
+      "learning_rate": 3.921392324823331e-06,
+      "loss": 0.5298,
+      "step": 1715
+    },
+    {
+      "epoch": 1.9219566840926063,
+      "grad_norm": 0.8670448660850525,
+      "learning_rate": 3.92016186682789e-06,
+      "loss": 0.5484,
+      "step": 1716
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 0.8477528691291809,
+      "learning_rate": 3.918930900686878e-06,
+      "loss": 0.5279,
+      "step": 1717
+    },
+    {
+      "epoch": 1.9241971620612397,
+      "grad_norm": 0.8477908968925476,
+      "learning_rate": 3.917699426840745e-06,
+      "loss": 0.5633,
+      "step": 1718
+    },
+    {
+      "epoch": 1.9253174010455565,
+      "grad_norm": 0.813227117061615,
+      "learning_rate": 3.916467445730121e-06,
+      "loss": 0.5443,
+      "step": 1719
+    },
+    {
+      "epoch": 1.926437640029873,
+      "grad_norm": 0.821105420589447,
+      "learning_rate": 3.915234957795821e-06,
+      "loss": 0.5424,
+      "step": 1720
+    },
+    {
+      "epoch": 1.9275578790141896,
+      "grad_norm": 0.8325830101966858,
+      "learning_rate": 3.9140019634788385e-06,
+      "loss": 0.5218,
+      "step": 1721
+    },
+    {
+      "epoch": 1.9286781179985062,
+      "grad_norm": 0.8530997037887573,
+      "learning_rate": 3.9127684632203485e-06,
+      "loss": 0.5805,
+      "step": 1722
+    },
+    {
+      "epoch": 1.929798356982823,
+      "grad_norm": 0.836625874042511,
+      "learning_rate": 3.911534457461708e-06,
+      "loss": 0.5117,
+      "step": 1723
+    },
+    {
+      "epoch": 1.9309185959671398,
+      "grad_norm": 0.8190492391586304,
+      "learning_rate": 3.910299946644455e-06,
+      "loss": 0.5275,
+      "step": 1724
+    },
+    {
+      "epoch": 1.9320388349514563,
+      "grad_norm": 0.8174676299095154,
+      "learning_rate": 3.909064931210306e-06,
+      "loss": 0.5219,
+      "step": 1725
+    },
+    {
+      "epoch": 1.933159073935773,
+      "grad_norm": 0.8266441822052002,
+      "learning_rate": 3.907829411601161e-06,
+      "loss": 0.532,
+      "step": 1726
+    },
+    {
+      "epoch": 1.9342793129200895,
+      "grad_norm": 0.8342094421386719,
+      "learning_rate": 3.906593388259098e-06,
+      "loss": 0.5507,
+      "step": 1727
+    },
+    {
+      "epoch": 1.9353995519044063,
+      "grad_norm": 0.8714757561683655,
+      "learning_rate": 3.905356861626377e-06,
+      "loss": 0.5371,
+      "step": 1728
+    },
+    {
+      "epoch": 1.936519790888723,
+      "grad_norm": 0.8320139050483704,
+      "learning_rate": 3.904119832145437e-06,
+      "loss": 0.5143,
+      "step": 1729
+    },
+    {
+      "epoch": 1.9376400298730396,
+      "grad_norm": 0.7936547994613647,
+      "learning_rate": 3.902882300258898e-06,
+      "loss": 0.551,
+      "step": 1730
+    },
+    {
+      "epoch": 1.9387602688573562,
+      "grad_norm": 0.8392663598060608,
+      "learning_rate": 3.9016442664095585e-06,
+      "loss": 0.558,
+      "step": 1731
+    },
+    {
+      "epoch": 1.9398805078416728,
+      "grad_norm": 0.8122281432151794,
+      "learning_rate": 3.900405731040397e-06,
+      "loss": 0.5134,
+      "step": 1732
+    },
+    {
+      "epoch": 1.9410007468259896,
+      "grad_norm": 0.8271510601043701,
+      "learning_rate": 3.8991666945945735e-06,
+      "loss": 0.5211,
+      "step": 1733
+    },
+    {
+      "epoch": 1.9421209858103063,
+      "grad_norm": 0.8124710321426392,
+      "learning_rate": 3.897927157515423e-06,
+      "loss": 0.5532,
+      "step": 1734
+    },
+    {
+      "epoch": 1.943241224794623,
+      "grad_norm": 0.8266910314559937,
+      "learning_rate": 3.896687120246462e-06,
+      "loss": 0.5363,
+      "step": 1735
+    },
+    {
+      "epoch": 1.9443614637789395,
+      "grad_norm": 0.9358043074607849,
+      "learning_rate": 3.895446583231387e-06,
+      "loss": 0.5286,
+      "step": 1736
+    },
+    {
+      "epoch": 1.945481702763256,
+      "grad_norm": 0.8380877375602722,
+      "learning_rate": 3.8942055469140726e-06,
+      "loss": 0.5401,
+      "step": 1737
+    },
+    {
+      "epoch": 1.9466019417475728,
+      "grad_norm": 0.820033073425293,
+      "learning_rate": 3.892964011738571e-06,
+      "loss": 0.5398,
+      "step": 1738
+    },
+    {
+      "epoch": 1.9477221807318896,
+      "grad_norm": 0.8377394676208496,
+      "learning_rate": 3.891721978149115e-06,
+      "loss": 0.5261,
+      "step": 1739
+    },
+    {
+      "epoch": 1.9488424197162062,
+      "grad_norm": 0.8230284452438354,
+      "learning_rate": 3.890479446590113e-06,
+      "loss": 0.5449,
+      "step": 1740
+    },
+    {
+      "epoch": 1.9499626587005228,
+      "grad_norm": 0.8403946757316589,
+      "learning_rate": 3.889236417506153e-06,
+      "loss": 0.5317,
+      "step": 1741
+    },
+    {
+      "epoch": 1.9510828976848393,
+      "grad_norm": 0.8075201511383057,
+      "learning_rate": 3.887992891342002e-06,
+      "loss": 0.5229,
+      "step": 1742
+    },
+    {
+      "epoch": 1.9522031366691561,
+      "grad_norm": 0.8129234313964844,
+      "learning_rate": 3.886748868542604e-06,
+      "loss": 0.533,
+      "step": 1743
+    },
+    {
+      "epoch": 1.9533233756534727,
+      "grad_norm": 0.8530985713005066,
+      "learning_rate": 3.8855043495530785e-06,
+      "loss": 0.5253,
+      "step": 1744
+    },
+    {
+      "epoch": 1.9544436146377895,
+      "grad_norm": 0.8068393468856812,
+      "learning_rate": 3.884259334818726e-06,
+      "loss": 0.534,
+      "step": 1745
+    },
+    {
+      "epoch": 1.955563853622106,
+      "grad_norm": 0.8525707125663757,
+      "learning_rate": 3.883013824785024e-06,
+      "loss": 0.5254,
+      "step": 1746
+    },
+    {
+      "epoch": 1.9566840926064226,
+      "grad_norm": 0.8301211595535278,
+      "learning_rate": 3.881767819897625e-06,
+      "loss": 0.5394,
+      "step": 1747
+    },
+    {
+      "epoch": 1.9578043315907394,
+      "grad_norm": 0.8432576656341553,
+      "learning_rate": 3.88052132060236e-06,
+      "loss": 0.5368,
+      "step": 1748
+    },
+    {
+      "epoch": 1.958924570575056,
+      "grad_norm": 0.8383322954177856,
+      "learning_rate": 3.879274327345236e-06,
+      "loss": 0.5672,
+      "step": 1749
+    },
+    {
+      "epoch": 1.9600448095593728,
+      "grad_norm": 0.8277074098587036,
+      "learning_rate": 3.878026840572438e-06,
+      "loss": 0.5356,
+      "step": 1750
+    },
+    {
+      "epoch": 1.9611650485436893,
+      "grad_norm": 0.8360003232955933,
+      "learning_rate": 3.876778860730328e-06,
+      "loss": 0.5381,
+      "step": 1751
+    },
+    {
+      "epoch": 1.962285287528006,
+      "grad_norm": 0.8012877702713013,
+      "learning_rate": 3.8755303882654425e-06,
+      "loss": 0.5176,
+      "step": 1752
+    },
+    {
+      "epoch": 1.9634055265123225,
+      "grad_norm": 0.811970055103302,
+      "learning_rate": 3.874281423624495e-06,
+      "loss": 0.5243,
+      "step": 1753
+    },
+    {
+      "epoch": 1.9645257654966393,
+      "grad_norm": 0.8069865107536316,
+      "learning_rate": 3.873031967254375e-06,
+      "loss": 0.5329,
+      "step": 1754
+    },
+    {
+      "epoch": 1.965646004480956,
+      "grad_norm": 0.8421149253845215,
+      "learning_rate": 3.871782019602148e-06,
+      "loss": 0.4976,
+      "step": 1755
+    },
+    {
+      "epoch": 1.9667662434652726,
+      "grad_norm": 0.7935910224914551,
+      "learning_rate": 3.870531581115058e-06,
+      "loss": 0.4977,
+      "step": 1756
+    },
+    {
+      "epoch": 1.9678864824495892,
+      "grad_norm": 0.7953087091445923,
+      "learning_rate": 3.86928065224052e-06,
+      "loss": 0.5169,
+      "step": 1757
+    },
+    {
+      "epoch": 1.9690067214339058,
+      "grad_norm": 0.8351732492446899,
+      "learning_rate": 3.868029233426127e-06,
+      "loss": 0.5329,
+      "step": 1758
+    },
+    {
+      "epoch": 1.9701269604182226,
+      "grad_norm": 0.8786207437515259,
+      "learning_rate": 3.8667773251196466e-06,
+      "loss": 0.5645,
+      "step": 1759
+    },
+    {
+      "epoch": 1.9712471994025393,
+      "grad_norm": 0.8424807190895081,
+      "learning_rate": 3.865524927769023e-06,
+      "loss": 0.5236,
+      "step": 1760
+    },
+    {
+      "epoch": 1.972367438386856,
+      "grad_norm": 0.8212794661521912,
+      "learning_rate": 3.8642720418223725e-06,
+      "loss": 0.5448,
+      "step": 1761
+    },
+    {
+      "epoch": 1.9734876773711725,
+      "grad_norm": 0.8219057321548462,
+      "learning_rate": 3.863018667727989e-06,
+      "loss": 0.5176,
+      "step": 1762
+    },
+    {
+      "epoch": 1.974607916355489,
+      "grad_norm": 0.8447268009185791,
+      "learning_rate": 3.86176480593434e-06,
+      "loss": 0.5134,
+      "step": 1763
+    },
+    {
+      "epoch": 1.9757281553398058,
+      "grad_norm": 0.8190571665763855,
+      "learning_rate": 3.8605104568900685e-06,
+      "loss": 0.5135,
+      "step": 1764
+    },
+    {
+      "epoch": 1.9768483943241226,
+      "grad_norm": 0.8249821662902832,
+      "learning_rate": 3.859255621043989e-06,
+      "loss": 0.527,
+      "step": 1765
+    },
+    {
+      "epoch": 1.9779686333084392,
+      "grad_norm": 0.805282711982727,
+      "learning_rate": 3.8580002988450934e-06,
+      "loss": 0.5639,
+      "step": 1766
+    },
+    {
+      "epoch": 1.9790888722927558,
+      "grad_norm": 0.8726804256439209,
+      "learning_rate": 3.8567444907425464e-06,
+      "loss": 0.5074,
+      "step": 1767
+    },
+    {
+      "epoch": 1.9802091112770723,
+      "grad_norm": 0.8451085090637207,
+      "learning_rate": 3.855488197185685e-06,
+      "loss": 0.5287,
+      "step": 1768
+    },
+    {
+      "epoch": 1.9813293502613891,
+      "grad_norm": 0.8440846800804138,
+      "learning_rate": 3.8542314186240224e-06,
+      "loss": 0.5283,
+      "step": 1769
+    },
+    {
+      "epoch": 1.982449589245706,
+      "grad_norm": 0.8287026286125183,
+      "learning_rate": 3.8529741555072446e-06,
+      "loss": 0.5188,
+      "step": 1770
+    },
+    {
+      "epoch": 1.9835698282300225,
+      "grad_norm": 0.8392443656921387,
+      "learning_rate": 3.85171640828521e-06,
+      "loss": 0.5208,
+      "step": 1771
+    },
+    {
+      "epoch": 1.984690067214339,
+      "grad_norm": 0.8401972651481628,
+      "learning_rate": 3.8504581774079514e-06,
+      "loss": 0.5358,
+      "step": 1772
+    },
+    {
+      "epoch": 1.9858103061986556,
+      "grad_norm": 0.8330747485160828,
+      "learning_rate": 3.849199463325674e-06,
+      "loss": 0.5419,
+      "step": 1773
+    },
+    {
+      "epoch": 1.9869305451829724,
+      "grad_norm": 0.8241636753082275,
+      "learning_rate": 3.847940266488755e-06,
+      "loss": 0.5277,
+      "step": 1774
+    },
+    {
+      "epoch": 1.988050784167289,
+      "grad_norm": 0.868881344795227,
+      "learning_rate": 3.846680587347747e-06,
+      "loss": 0.5098,
+      "step": 1775
+    },
+    {
+      "epoch": 1.9891710231516058,
+      "grad_norm": 0.7896348237991333,
+      "learning_rate": 3.845420426353371e-06,
+      "loss": 0.5078,
+      "step": 1776
+    },
+    {
+      "epoch": 1.9902912621359223,
+      "grad_norm": 0.8297612071037292,
+      "learning_rate": 3.844159783956527e-06,
+      "loss": 0.5341,
+      "step": 1777
+    },
+    {
+      "epoch": 1.991411501120239,
+      "grad_norm": 0.828650176525116,
+      "learning_rate": 3.84289866060828e-06,
+      "loss": 0.54,
+      "step": 1778
+    },
+    {
+      "epoch": 1.9925317401045557,
+      "grad_norm": 0.8304623961448669,
+      "learning_rate": 3.84163705675987e-06,
+      "loss": 0.5255,
+      "step": 1779
+    },
+    {
+      "epoch": 1.9936519790888723,
+      "grad_norm": 0.7775478959083557,
+      "learning_rate": 3.840374972862712e-06,
+      "loss": 0.5496,
+      "step": 1780
+    },
+    {
+      "epoch": 1.994772218073189,
+      "grad_norm": 0.844093382358551,
+      "learning_rate": 3.839112409368389e-06,
+      "loss": 0.5448,
+      "step": 1781
+    },
+    {
+      "epoch": 1.9958924570575056,
+      "grad_norm": 0.8264373540878296,
+      "learning_rate": 3.837849366728653e-06,
+      "loss": 0.5093,
+      "step": 1782
+    },
+    {
+      "epoch": 1.9970126960418222,
+      "grad_norm": 0.8227883577346802,
+      "learning_rate": 3.8365858453954365e-06,
+      "loss": 0.5544,
+      "step": 1783
+    },
+    {
+      "epoch": 1.9981329350261388,
+      "grad_norm": 0.8628005385398865,
+      "learning_rate": 3.835321845820834e-06,
+      "loss": 0.5297,
+      "step": 1784
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 5352,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 6,
+  "save_steps": 892,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.505396951730094e+18,
+  "train_batch_size": 4,
+  "trial_name": null,
+  "trial_params": null
+}