diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14869 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.9964614295824488,
+  "eval_steps": 500,
+  "global_step": 2118,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0014154281670205238,
+      "grad_norm": 56.700288366666925,
+      "learning_rate": 0.0,
+      "loss": 11.0711,
+      "step": 1
+    },
+    {
+      "epoch": 0.0028308563340410475,
+      "grad_norm": 57.44157352268369,
+      "learning_rate": 2.3584905660377358e-07,
+      "loss": 10.9796,
+      "step": 2
+    },
+    {
+      "epoch": 0.004246284501061571,
+      "grad_norm": 53.372480303586045,
+      "learning_rate": 4.7169811320754717e-07,
+      "loss": 11.055,
+      "step": 3
+    },
+    {
+      "epoch": 0.005661712668082095,
+      "grad_norm": 54.58404851338792,
+      "learning_rate": 7.075471698113208e-07,
+      "loss": 11.1072,
+      "step": 4
+    },
+    {
+      "epoch": 0.007077140835102618,
+      "grad_norm": 56.233021204938744,
+      "learning_rate": 9.433962264150943e-07,
+      "loss": 11.1468,
+      "step": 5
+    },
+    {
+      "epoch": 0.008492569002123142,
+      "grad_norm": 58.256970362810634,
+      "learning_rate": 1.179245283018868e-06,
+      "loss": 11.0462,
+      "step": 6
+    },
+    {
+      "epoch": 0.009907997169143666,
+      "grad_norm": 56.64095298259024,
+      "learning_rate": 1.4150943396226415e-06,
+      "loss": 10.9947,
+      "step": 7
+    },
+    {
+      "epoch": 0.01132342533616419,
+      "grad_norm": 58.72592183949963,
+      "learning_rate": 1.650943396226415e-06,
+      "loss": 10.7203,
+      "step": 8
+    },
+    {
+      "epoch": 0.012738853503184714,
+      "grad_norm": 63.37966188745395,
+      "learning_rate": 1.8867924528301887e-06,
+      "loss": 10.6896,
+      "step": 9
+    },
+    {
+      "epoch": 0.014154281670205236,
+      "grad_norm": 86.37474442272796,
+      "learning_rate": 2.1226415094339624e-06,
+      "loss": 9.6251,
+      "step": 10
+    },
+    {
+      "epoch": 0.01556970983722576,
+      "grad_norm": 90.0472325162758,
+      "learning_rate": 2.358490566037736e-06,
+      "loss": 9.4276,
+      "step": 11
+    },
+    {
+      "epoch": 0.016985138004246284,
+      "grad_norm": 102.22797914866898,
+      "learning_rate": 2.5943396226415095e-06,
+      "loss": 8.8853,
+      "step": 12
+    },
+    {
+      "epoch": 0.01840056617126681,
+      "grad_norm": 86.241399089093,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 4.4159,
+      "step": 13
+    },
+    {
+      "epoch": 0.019815994338287332,
+      "grad_norm": 59.25457272201123,
+      "learning_rate": 3.0660377358490567e-06,
+      "loss": 3.3741,
+      "step": 14
+    },
+    {
+      "epoch": 0.021231422505307854,
+      "grad_norm": 55.03231585614142,
+      "learning_rate": 3.30188679245283e-06,
+      "loss": 3.1289,
+      "step": 15
+    },
+    {
+      "epoch": 0.02264685067232838,
+      "grad_norm": 35.542100873737745,
+      "learning_rate": 3.5377358490566038e-06,
+      "loss": 2.4912,
+      "step": 16
+    },
+    {
+      "epoch": 0.024062278839348902,
+      "grad_norm": 29.48191176082474,
+      "learning_rate": 3.7735849056603773e-06,
+      "loss": 2.2826,
+      "step": 17
+    },
+    {
+      "epoch": 0.025477707006369428,
+      "grad_norm": 7.025724837128468,
+      "learning_rate": 4.009433962264151e-06,
+      "loss": 1.4331,
+      "step": 18
+    },
+    {
+      "epoch": 0.02689313517338995,
+      "grad_norm": 5.084153252624878,
+      "learning_rate": 4.245283018867925e-06,
+      "loss": 1.3344,
+      "step": 19
+    },
+    {
+      "epoch": 0.028308563340410473,
+      "grad_norm": 4.10463053850957,
+      "learning_rate": 4.481132075471698e-06,
+      "loss": 1.2341,
+      "step": 20
+    },
+    {
+      "epoch": 0.029723991507430998,
+      "grad_norm": 3.239333220848038,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 1.2445,
+      "step": 21
+    },
+    {
+      "epoch": 0.03113941967445152,
+      "grad_norm": 2.4598734753960074,
+      "learning_rate": 4.952830188679246e-06,
+      "loss": 1.1433,
+      "step": 22
+    },
+    {
+      "epoch": 0.03255484784147204,
+      "grad_norm": 2.0603206837135257,
+      "learning_rate": 5.188679245283019e-06,
+      "loss": 1.106,
+      "step": 23
+    },
+    {
+      "epoch": 0.03397027600849257,
+      "grad_norm": 1.5119226227330653,
+      "learning_rate": 5.424528301886793e-06,
+      "loss": 1.0189,
+      "step": 24
+    },
+    {
+      "epoch": 0.035385704175513094,
+      "grad_norm": 38.97325572744693,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.952,
+      "step": 25
+    },
+    {
+      "epoch": 0.03680113234253362,
+      "grad_norm": 35.594655464652305,
+      "learning_rate": 5.89622641509434e-06,
+      "loss": 0.9317,
+      "step": 26
+    },
+    {
+      "epoch": 0.03821656050955414,
+      "grad_norm": 2.243266260915374,
+      "learning_rate": 6.132075471698113e-06,
+      "loss": 0.8992,
+      "step": 27
+    },
+    {
+      "epoch": 0.039631988676574664,
+      "grad_norm": 1.2834980121102546,
+      "learning_rate": 6.367924528301887e-06,
+      "loss": 0.8868,
+      "step": 28
+    },
+    {
+      "epoch": 0.04104741684359519,
+      "grad_norm": 1.0074041465604129,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.8448,
+      "step": 29
+    },
+    {
+      "epoch": 0.04246284501061571,
+      "grad_norm": 0.9346105193952094,
+      "learning_rate": 6.839622641509434e-06,
+      "loss": 0.8312,
+      "step": 30
+    },
+    {
+      "epoch": 0.043878273177636234,
+      "grad_norm": 0.8661479332986395,
+      "learning_rate": 7.0754716981132075e-06,
+      "loss": 0.794,
+      "step": 31
+    },
+    {
+      "epoch": 0.04529370134465676,
+      "grad_norm": 0.8137863776538741,
+      "learning_rate": 7.3113207547169815e-06,
+      "loss": 0.7864,
+      "step": 32
+    },
+    {
+      "epoch": 0.04670912951167728,
+      "grad_norm": 0.7938781033721152,
+      "learning_rate": 7.547169811320755e-06,
+      "loss": 0.7867,
+      "step": 33
+    },
+    {
+      "epoch": 0.048124557678697805,
+      "grad_norm": 1.0106701346359332,
+      "learning_rate": 7.783018867924528e-06,
+      "loss": 0.7753,
+      "step": 34
+    },
+    {
+      "epoch": 0.04953998584571833,
+      "grad_norm": 0.6556714750236174,
+      "learning_rate": 8.018867924528302e-06,
+      "loss": 0.7516,
+      "step": 35
+    },
+    {
+      "epoch": 0.050955414012738856,
+      "grad_norm": 0.6214486522550411,
+      "learning_rate": 8.254716981132076e-06,
+      "loss": 0.7006,
+      "step": 36
+    },
+    {
+      "epoch": 0.052370842179759375,
+      "grad_norm": 0.7730669399452611,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.7446,
+      "step": 37
+    },
+    {
+      "epoch": 0.0537862703467799,
+      "grad_norm": 0.7018102584659955,
+      "learning_rate": 8.726415094339622e-06,
+      "loss": 0.711,
+      "step": 38
+    },
+    {
+      "epoch": 0.055201698513800426,
+      "grad_norm": 0.5894767958912929,
+      "learning_rate": 8.962264150943396e-06,
+      "loss": 0.7113,
+      "step": 39
+    },
+    {
+      "epoch": 0.056617126680820945,
+      "grad_norm": 0.5133118675149501,
+      "learning_rate": 9.19811320754717e-06,
+      "loss": 0.6953,
+      "step": 40
+    },
+    {
+      "epoch": 0.05803255484784147,
+      "grad_norm": 0.6025515680004425,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.6876,
+      "step": 41
+    },
+    {
+      "epoch": 0.059447983014861996,
+      "grad_norm": 0.5831736200181878,
+      "learning_rate": 9.669811320754718e-06,
+      "loss": 0.7135,
+      "step": 42
+    },
+    {
+      "epoch": 0.06086341118188252,
+      "grad_norm": 0.5035341748650088,
+      "learning_rate": 9.905660377358492e-06,
+      "loss": 0.6603,
+      "step": 43
+    },
+    {
+      "epoch": 0.06227883934890304,
+      "grad_norm": 0.5102193306030185,
+      "learning_rate": 1.0141509433962266e-05,
+      "loss": 0.6497,
+      "step": 44
+    },
+    {
+      "epoch": 0.06369426751592357,
+      "grad_norm": 0.4668554672731941,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.6563,
+      "step": 45
+    },
+    {
+      "epoch": 0.06510969568294409,
+      "grad_norm": 0.39544745906790735,
+      "learning_rate": 1.0613207547169812e-05,
+      "loss": 0.6421,
+      "step": 46
+    },
+    {
+      "epoch": 0.06652512384996462,
+      "grad_norm": 0.4641271827730375,
+      "learning_rate": 1.0849056603773586e-05,
+      "loss": 0.6502,
+      "step": 47
+    },
+    {
+      "epoch": 0.06794055201698514,
+      "grad_norm": 0.5259678866370957,
+      "learning_rate": 1.108490566037736e-05,
+      "loss": 0.6534,
+      "step": 48
+    },
+    {
+      "epoch": 0.06935598018400566,
+      "grad_norm": 0.4245716432566208,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.6342,
+      "step": 49
+    },
+    {
+      "epoch": 0.07077140835102619,
+      "grad_norm": 0.3716803827062742,
+      "learning_rate": 1.1556603773584906e-05,
+      "loss": 0.581,
+      "step": 50
+    },
+    {
+      "epoch": 0.07218683651804671,
+      "grad_norm": 0.401007074045064,
+      "learning_rate": 1.179245283018868e-05,
+      "loss": 0.592,
+      "step": 51
+    },
+    {
+      "epoch": 0.07360226468506724,
+      "grad_norm": 0.36955059055091255,
+      "learning_rate": 1.2028301886792454e-05,
+      "loss": 0.5982,
+      "step": 52
+    },
+    {
+      "epoch": 0.07501769285208776,
+      "grad_norm": 0.37553426047306404,
+      "learning_rate": 1.2264150943396227e-05,
+      "loss": 0.6227,
+      "step": 53
+    },
+    {
+      "epoch": 0.07643312101910828,
+      "grad_norm": 0.38357989584669916,
+      "learning_rate": 1.25e-05,
+      "loss": 0.6383,
+      "step": 54
+    },
+    {
+      "epoch": 0.07784854918612881,
+      "grad_norm": 0.30675169193620705,
+      "learning_rate": 1.2735849056603775e-05,
+      "loss": 0.627,
+      "step": 55
+    },
+    {
+      "epoch": 0.07926397735314933,
+      "grad_norm": 0.29348505645605366,
+      "learning_rate": 1.2971698113207547e-05,
+      "loss": 0.6151,
+      "step": 56
+    },
+    {
+      "epoch": 0.08067940552016985,
+      "grad_norm": 0.35004341869750544,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.5992,
+      "step": 57
+    },
+    {
+      "epoch": 0.08209483368719038,
+      "grad_norm": 0.3279980602714091,
+      "learning_rate": 1.3443396226415095e-05,
+      "loss": 0.6399,
+      "step": 58
+    },
+    {
+      "epoch": 0.0835102618542109,
+      "grad_norm": 0.2777121636040088,
+      "learning_rate": 1.3679245283018869e-05,
+      "loss": 0.5908,
+      "step": 59
+    },
+    {
+      "epoch": 0.08492569002123142,
+      "grad_norm": 0.2646786874035676,
+      "learning_rate": 1.3915094339622641e-05,
+      "loss": 0.5783,
+      "step": 60
+    },
+    {
+      "epoch": 0.08634111818825195,
+      "grad_norm": 0.3153521431686913,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.599,
+      "step": 61
+    },
+    {
+      "epoch": 0.08775654635527247,
+      "grad_norm": 0.28886428433507566,
+      "learning_rate": 1.4386792452830189e-05,
+      "loss": 0.6061,
+      "step": 62
+    },
+    {
+      "epoch": 0.08917197452229299,
+      "grad_norm": 0.29782611977418977,
+      "learning_rate": 1.4622641509433963e-05,
+      "loss": 0.6184,
+      "step": 63
+    },
+    {
+      "epoch": 0.09058740268931352,
+      "grad_norm": 0.25742903864507843,
+      "learning_rate": 1.4858490566037735e-05,
+      "loss": 0.5752,
+      "step": 64
+    },
+    {
+      "epoch": 0.09200283085633404,
+      "grad_norm": 0.25729978441166224,
+      "learning_rate": 1.509433962264151e-05,
+      "loss": 0.5831,
+      "step": 65
+    },
+    {
+      "epoch": 0.09341825902335456,
+      "grad_norm": 0.2730860454565041,
+      "learning_rate": 1.5330188679245283e-05,
+      "loss": 0.5699,
+      "step": 66
+    },
+    {
+      "epoch": 0.09483368719037509,
+      "grad_norm": 0.2486573103575608,
+      "learning_rate": 1.5566037735849056e-05,
+      "loss": 0.6091,
+      "step": 67
+    },
+    {
+      "epoch": 0.09624911535739561,
+      "grad_norm": 0.2652550247675538,
+      "learning_rate": 1.580188679245283e-05,
+      "loss": 0.5901,
+      "step": 68
+    },
+    {
+      "epoch": 0.09766454352441614,
+      "grad_norm": 0.276533754994318,
+      "learning_rate": 1.6037735849056604e-05,
+      "loss": 0.5834,
+      "step": 69
+    },
+    {
+      "epoch": 0.09907997169143666,
+      "grad_norm": 0.23811332725591694,
+      "learning_rate": 1.6273584905660376e-05,
+      "loss": 0.5691,
+      "step": 70
+    },
+    {
+      "epoch": 0.10049539985845718,
+      "grad_norm": 0.2535636624906435,
+      "learning_rate": 1.650943396226415e-05,
+      "loss": 0.5669,
+      "step": 71
+    },
+    {
+      "epoch": 0.10191082802547771,
+      "grad_norm": 0.2844750862752134,
+      "learning_rate": 1.6745283018867924e-05,
+      "loss": 0.5832,
+      "step": 72
+    },
+    {
+      "epoch": 0.10332625619249823,
+      "grad_norm": 0.27334058079679957,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.5792,
+      "step": 73
+    },
+    {
+      "epoch": 0.10474168435951875,
+      "grad_norm": 0.2760364074849346,
+      "learning_rate": 1.7216981132075472e-05,
+      "loss": 0.5859,
+      "step": 74
+    },
+    {
+      "epoch": 0.10615711252653928,
+      "grad_norm": 0.24662318119066878,
+      "learning_rate": 1.7452830188679244e-05,
+      "loss": 0.5665,
+      "step": 75
+    },
+    {
+      "epoch": 0.1075725406935598,
+      "grad_norm": 0.25506885251253175,
+      "learning_rate": 1.768867924528302e-05,
+      "loss": 0.5584,
+      "step": 76
+    },
+    {
+      "epoch": 0.10898796886058032,
+      "grad_norm": 0.26773057280476475,
+      "learning_rate": 1.7924528301886792e-05,
+      "loss": 0.5547,
+      "step": 77
+    },
+    {
+      "epoch": 0.11040339702760085,
+      "grad_norm": 0.26103279361024406,
+      "learning_rate": 1.8160377358490564e-05,
+      "loss": 0.5539,
+      "step": 78
+    },
+    {
+      "epoch": 0.11181882519462137,
+      "grad_norm": 0.2535566104327732,
+      "learning_rate": 1.839622641509434e-05,
+      "loss": 0.5393,
+      "step": 79
+    },
+    {
+      "epoch": 0.11323425336164189,
+      "grad_norm": 0.25737562711758843,
+      "learning_rate": 1.8632075471698112e-05,
+      "loss": 0.5519,
+      "step": 80
+    },
+    {
+      "epoch": 0.11464968152866242,
+      "grad_norm": 0.2516666029791499,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.5586,
+      "step": 81
+    },
+    {
+      "epoch": 0.11606510969568294,
+      "grad_norm": 0.27538960051050937,
+      "learning_rate": 1.9103773584905664e-05,
+      "loss": 0.537,
+      "step": 82
+    },
+    {
+      "epoch": 0.11748053786270347,
+      "grad_norm": 0.25155459291746524,
+      "learning_rate": 1.9339622641509436e-05,
+      "loss": 0.5238,
+      "step": 83
+    },
+    {
+      "epoch": 0.11889596602972399,
+      "grad_norm": 0.2610631357388765,
+      "learning_rate": 1.9575471698113208e-05,
+      "loss": 0.5469,
+      "step": 84
+    },
+    {
+      "epoch": 0.12031139419674451,
+      "grad_norm": 0.2710744295346308,
+      "learning_rate": 1.9811320754716984e-05,
+      "loss": 0.5342,
+      "step": 85
+    },
+    {
+      "epoch": 0.12172682236376504,
+      "grad_norm": 0.2746000305881197,
+      "learning_rate": 2.0047169811320756e-05,
+      "loss": 0.5533,
+      "step": 86
+    },
+    {
+      "epoch": 0.12314225053078556,
+      "grad_norm": 0.2713072569437375,
+      "learning_rate": 2.0283018867924532e-05,
+      "loss": 0.564,
+      "step": 87
+    },
+    {
+      "epoch": 0.12455767869780608,
+      "grad_norm": 0.25200431770680803,
+      "learning_rate": 2.0518867924528304e-05,
+      "loss": 0.5308,
+      "step": 88
+    },
+    {
+      "epoch": 0.1259731068648266,
+      "grad_norm": 0.2904072154659877,
+      "learning_rate": 2.0754716981132076e-05,
+      "loss": 0.5561,
+      "step": 89
+    },
+    {
+      "epoch": 0.12738853503184713,
+      "grad_norm": 0.29066427628129154,
+      "learning_rate": 2.0990566037735852e-05,
+      "loss": 0.5419,
+      "step": 90
+    },
+    {
+      "epoch": 0.12880396319886767,
+      "grad_norm": 0.26449247738413356,
+      "learning_rate": 2.1226415094339624e-05,
+      "loss": 0.53,
+      "step": 91
+    },
+    {
+      "epoch": 0.13021939136588817,
+      "grad_norm": 0.28680674957429403,
+      "learning_rate": 2.1462264150943397e-05,
+      "loss": 0.5631,
+      "step": 92
+    },
+    {
+      "epoch": 0.1316348195329087,
+      "grad_norm": 0.29348512626053397,
+      "learning_rate": 2.1698113207547172e-05,
+      "loss": 0.5488,
+      "step": 93
+    },
+    {
+      "epoch": 0.13305024769992924,
+      "grad_norm": 0.2670829979081624,
+      "learning_rate": 2.1933962264150945e-05,
+      "loss": 0.5607,
+      "step": 94
+    },
+    {
+      "epoch": 0.13446567586694974,
+      "grad_norm": 0.2951742951553837,
+      "learning_rate": 2.216981132075472e-05,
+      "loss": 0.5264,
+      "step": 95
+    },
+    {
+      "epoch": 0.13588110403397027,
+      "grad_norm": 0.2738940637804454,
+      "learning_rate": 2.2405660377358493e-05,
+      "loss": 0.5497,
+      "step": 96
+    },
+    {
+      "epoch": 0.1372965322009908,
+      "grad_norm": 0.260689152524016,
+      "learning_rate": 2.2641509433962265e-05,
+      "loss": 0.529,
+      "step": 97
+    },
+    {
+      "epoch": 0.1387119603680113,
+      "grad_norm": 0.3460865251704031,
+      "learning_rate": 2.287735849056604e-05,
+      "loss": 0.5374,
+      "step": 98
+    },
+    {
+      "epoch": 0.14012738853503184,
+      "grad_norm": 0.26025504251336634,
+      "learning_rate": 2.3113207547169813e-05,
+      "loss": 0.5197,
+      "step": 99
+    },
+    {
+      "epoch": 0.14154281670205238,
+      "grad_norm": 0.32553833522366754,
+      "learning_rate": 2.3349056603773585e-05,
+      "loss": 0.5483,
+      "step": 100
+    },
+    {
+      "epoch": 0.14295824486907288,
+      "grad_norm": 0.28825995008378963,
+      "learning_rate": 2.358490566037736e-05,
+      "loss": 0.5206,
+      "step": 101
+    },
+    {
+      "epoch": 0.14437367303609341,
+      "grad_norm": 0.2873909845421847,
+      "learning_rate": 2.3820754716981133e-05,
+      "loss": 0.4962,
+      "step": 102
+    },
+    {
+      "epoch": 0.14578910120311395,
+      "grad_norm": 0.28026036525465287,
+      "learning_rate": 2.405660377358491e-05,
+      "loss": 0.5446,
+      "step": 103
+    },
+    {
+      "epoch": 0.14720452937013448,
+      "grad_norm": 0.3272353770003188,
+      "learning_rate": 2.429245283018868e-05,
+      "loss": 0.5166,
+      "step": 104
+    },
+    {
+      "epoch": 0.14861995753715498,
+      "grad_norm": 0.31747918749974297,
+      "learning_rate": 2.4528301886792453e-05,
+      "loss": 0.5468,
+      "step": 105
+    },
+    {
+      "epoch": 0.15003538570417552,
+      "grad_norm": 0.3248933376699994,
+      "learning_rate": 2.476415094339623e-05,
+      "loss": 0.5378,
+      "step": 106
+    },
+    {
+      "epoch": 0.15145081387119605,
+      "grad_norm": 0.34796809815050533,
+      "learning_rate": 2.5e-05,
+      "loss": 0.5336,
+      "step": 107
+    },
+    {
+      "epoch": 0.15286624203821655,
+      "grad_norm": 0.32031365578591203,
+      "learning_rate": 2.5235849056603777e-05,
+      "loss": 0.5265,
+      "step": 108
+    },
+    {
+      "epoch": 0.1542816702052371,
+      "grad_norm": 0.3290449991605199,
+      "learning_rate": 2.547169811320755e-05,
+      "loss": 0.5419,
+      "step": 109
+    },
+    {
+      "epoch": 0.15569709837225762,
+      "grad_norm": 0.31459646236712213,
+      "learning_rate": 2.5707547169811325e-05,
+      "loss": 0.5267,
+      "step": 110
+    },
+    {
+      "epoch": 0.15711252653927812,
+      "grad_norm": 0.31446522115607733,
+      "learning_rate": 2.5943396226415094e-05,
+      "loss": 0.5425,
+      "step": 111
+    },
+    {
+      "epoch": 0.15852795470629866,
+      "grad_norm": 0.314348472522342,
+      "learning_rate": 2.6179245283018873e-05,
+      "loss": 0.5194,
+      "step": 112
+    },
+    {
+      "epoch": 0.1599433828733192,
+      "grad_norm": 0.33590260833875185,
+      "learning_rate": 2.641509433962264e-05,
+      "loss": 0.5286,
+      "step": 113
+    },
+    {
+      "epoch": 0.1613588110403397,
+      "grad_norm": 0.293348617675266,
+      "learning_rate": 2.6650943396226417e-05,
+      "loss": 0.5227,
+      "step": 114
+    },
+    {
+      "epoch": 0.16277423920736023,
+      "grad_norm": 0.2908436470141757,
+      "learning_rate": 2.688679245283019e-05,
+      "loss": 0.5342,
+      "step": 115
+    },
+    {
+      "epoch": 0.16418966737438076,
+      "grad_norm": 0.2909829047502465,
+      "learning_rate": 2.7122641509433965e-05,
+      "loss": 0.5235,
+      "step": 116
+    },
+    {
+      "epoch": 0.16560509554140126,
+      "grad_norm": 0.30833330637724304,
+      "learning_rate": 2.7358490566037738e-05,
+      "loss": 0.519,
+      "step": 117
+    },
+    {
+      "epoch": 0.1670205237084218,
+      "grad_norm": 0.30762798006157255,
+      "learning_rate": 2.7594339622641513e-05,
+      "loss": 0.534,
+      "step": 118
+    },
+    {
+      "epoch": 0.16843595187544233,
+      "grad_norm": 0.2865739474038587,
+      "learning_rate": 2.7830188679245282e-05,
+      "loss": 0.517,
+      "step": 119
+    },
+    {
+      "epoch": 0.16985138004246284,
+      "grad_norm": 0.33305918186709865,
+      "learning_rate": 2.806603773584906e-05,
+      "loss": 0.5249,
+      "step": 120
+    },
+    {
+      "epoch": 0.17126680820948337,
+      "grad_norm": 0.28237709762325275,
+      "learning_rate": 2.830188679245283e-05,
+      "loss": 0.5034,
+      "step": 121
+    },
+    {
+      "epoch": 0.1726822363765039,
+      "grad_norm": 0.27928952164149773,
+      "learning_rate": 2.8537735849056606e-05,
+      "loss": 0.5375,
+      "step": 122
+    },
+    {
+      "epoch": 0.1740976645435244,
+      "grad_norm": 0.324115687867415,
+      "learning_rate": 2.8773584905660378e-05,
+      "loss": 0.5246,
+      "step": 123
+    },
+    {
+      "epoch": 0.17551309271054494,
+      "grad_norm": 0.2762247549280857,
+      "learning_rate": 2.9009433962264154e-05,
+      "loss": 0.5055,
+      "step": 124
+    },
+    {
+      "epoch": 0.17692852087756547,
+      "grad_norm": 0.32165873663895433,
+      "learning_rate": 2.9245283018867926e-05,
+      "loss": 0.5065,
+      "step": 125
+    },
+    {
+      "epoch": 0.17834394904458598,
+      "grad_norm": 0.3089669988568903,
+      "learning_rate": 2.9481132075471702e-05,
+      "loss": 0.5123,
+      "step": 126
+    },
+    {
+      "epoch": 0.1797593772116065,
+      "grad_norm": 0.2915336309538987,
+      "learning_rate": 2.971698113207547e-05,
+      "loss": 0.5103,
+      "step": 127
+    },
+    {
+      "epoch": 0.18117480537862704,
+      "grad_norm": 0.33637365443529005,
+      "learning_rate": 2.995283018867925e-05,
+      "loss": 0.5363,
+      "step": 128
+    },
+    {
+      "epoch": 0.18259023354564755,
+      "grad_norm": 0.3122972896517381,
+      "learning_rate": 3.018867924528302e-05,
+      "loss": 0.5194,
+      "step": 129
+    },
+    {
+      "epoch": 0.18400566171266808,
+      "grad_norm": 0.32543770734441557,
+      "learning_rate": 3.0424528301886794e-05,
+      "loss": 0.5135,
+      "step": 130
+    },
+    {
+      "epoch": 0.1854210898796886,
+      "grad_norm": 0.27252001697072026,
+      "learning_rate": 3.0660377358490567e-05,
+      "loss": 0.528,
+      "step": 131
+    },
+    {
+      "epoch": 0.18683651804670912,
+      "grad_norm": 0.34402518951819805,
+      "learning_rate": 3.0896226415094346e-05,
+      "loss": 0.5195,
+      "step": 132
+    },
+    {
+      "epoch": 0.18825194621372965,
+      "grad_norm": 0.3126019430630103,
+      "learning_rate": 3.113207547169811e-05,
+      "loss": 0.5491,
+      "step": 133
+    },
+    {
+      "epoch": 0.18966737438075018,
+      "grad_norm": 0.2688192270465016,
+      "learning_rate": 3.136792452830189e-05,
+      "loss": 0.5275,
+      "step": 134
+    },
+    {
+      "epoch": 0.1910828025477707,
+      "grad_norm": 0.3119675680452963,
+      "learning_rate": 3.160377358490566e-05,
+      "loss": 0.5277,
+      "step": 135
+    },
+    {
+      "epoch": 0.19249823071479122,
+      "grad_norm": 0.25218648912353075,
+      "learning_rate": 3.1839622641509435e-05,
+      "loss": 0.5024,
+      "step": 136
+    },
+    {
+      "epoch": 0.19391365888181175,
+      "grad_norm": 0.3018799406875476,
+      "learning_rate": 3.207547169811321e-05,
+      "loss": 0.5113,
+      "step": 137
+    },
+    {
+      "epoch": 0.19532908704883228,
+      "grad_norm": 0.2851896196357652,
+      "learning_rate": 3.2311320754716986e-05,
+      "loss": 0.4968,
+      "step": 138
+    },
+    {
+      "epoch": 0.1967445152158528,
+      "grad_norm": 0.31923660565775647,
+      "learning_rate": 3.254716981132075e-05,
+      "loss": 0.5179,
+      "step": 139
+    },
+    {
+      "epoch": 0.19815994338287332,
+      "grad_norm": 0.3376523343537797,
+      "learning_rate": 3.278301886792453e-05,
+      "loss": 0.4941,
+      "step": 140
+    },
+    {
+      "epoch": 0.19957537154989385,
+      "grad_norm": 0.3325849764665281,
+      "learning_rate": 3.30188679245283e-05,
+      "loss": 0.5348,
+      "step": 141
+    },
+    {
+      "epoch": 0.20099079971691436,
+      "grad_norm": 0.2951463049393858,
+      "learning_rate": 3.3254716981132075e-05,
+      "loss": 0.5089,
+      "step": 142
+    },
+    {
+      "epoch": 0.2024062278839349,
+      "grad_norm": 0.3368157389197186,
+      "learning_rate": 3.349056603773585e-05,
+      "loss": 0.5116,
+      "step": 143
+    },
+    {
+      "epoch": 0.20382165605095542,
+      "grad_norm": 0.3195128601211723,
+      "learning_rate": 3.3726415094339627e-05,
+      "loss": 0.5229,
+      "step": 144
+    },
+    {
+      "epoch": 0.20523708421797593,
+      "grad_norm": 0.36219481973365675,
+      "learning_rate": 3.39622641509434e-05,
+      "loss": 0.5122,
+      "step": 145
+    },
+    {
+      "epoch": 0.20665251238499646,
+      "grad_norm": 0.3172880541291816,
+      "learning_rate": 3.419811320754717e-05,
+      "loss": 0.5006,
+      "step": 146
+    },
+    {
+      "epoch": 0.208067940552017,
+      "grad_norm": 0.3789720604684654,
+      "learning_rate": 3.4433962264150943e-05,
+      "loss": 0.4973,
+      "step": 147
+    },
+    {
+      "epoch": 0.2094833687190375,
+      "grad_norm": 0.3455429168182563,
+      "learning_rate": 3.466981132075472e-05,
+      "loss": 0.5127,
+      "step": 148
+    },
+    {
+      "epoch": 0.21089879688605803,
+      "grad_norm": 0.30672119758259186,
+      "learning_rate": 3.490566037735849e-05,
+      "loss": 0.4918,
+      "step": 149
+    },
+    {
+      "epoch": 0.21231422505307856,
+      "grad_norm": 0.332798117726157,
+      "learning_rate": 3.514150943396227e-05,
+      "loss": 0.5,
+      "step": 150
+    },
+    {
+      "epoch": 0.21372965322009907,
+      "grad_norm": 0.33726327033143527,
+      "learning_rate": 3.537735849056604e-05,
+      "loss": 0.4936,
+      "step": 151
+    },
+    {
+      "epoch": 0.2151450813871196,
+      "grad_norm": 0.3032929045898552,
+      "learning_rate": 3.561320754716981e-05,
+      "loss": 0.4937,
+      "step": 152
+    },
+    {
+      "epoch": 0.21656050955414013,
+      "grad_norm": 0.3589340045521236,
+      "learning_rate": 3.5849056603773584e-05,
+      "loss": 0.4882,
+      "step": 153
+    },
+    {
+      "epoch": 0.21797593772116064,
+      "grad_norm": 0.31519815057694117,
+      "learning_rate": 3.608490566037736e-05,
+      "loss": 0.4884,
+      "step": 154
+    },
+    {
+      "epoch": 0.21939136588818117,
+      "grad_norm": 0.35459205720985826,
+      "learning_rate": 3.632075471698113e-05,
+      "loss": 0.5036,
+      "step": 155
+    },
+    {
+      "epoch": 0.2208067940552017,
+      "grad_norm": 0.43207177900202387,
+      "learning_rate": 3.655660377358491e-05,
+      "loss": 0.4988,
+      "step": 156
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 0.36421480300311665,
+      "learning_rate": 3.679245283018868e-05,
+      "loss": 0.5089,
+      "step": 157
+    },
+    {
+      "epoch": 0.22363765038924274,
+      "grad_norm": 0.3834592202853376,
+      "learning_rate": 3.702830188679245e-05,
+      "loss": 0.5143,
+      "step": 158
+    },
+    {
+      "epoch": 0.22505307855626328,
+      "grad_norm": 0.4048640876243969,
+      "learning_rate": 3.7264150943396224e-05,
+      "loss": 0.5204,
+      "step": 159
+    },
+    {
+      "epoch": 0.22646850672328378,
+      "grad_norm": 0.4624530002488823,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 0.5226,
+      "step": 160
+    },
+    {
+      "epoch": 0.2278839348903043,
+      "grad_norm": 0.3480343117860362,
+      "learning_rate": 3.7735849056603776e-05,
+      "loss": 0.5081,
+      "step": 161
+    },
+    {
+      "epoch": 0.22929936305732485,
+      "grad_norm": 0.42689915332306533,
+      "learning_rate": 3.797169811320755e-05,
+      "loss": 0.4847,
+      "step": 162
+    },
+    {
+      "epoch": 0.23071479122434538,
+      "grad_norm": 0.4105608858696282,
+      "learning_rate": 3.820754716981133e-05,
+      "loss": 0.497,
+      "step": 163
+    },
+    {
+      "epoch": 0.23213021939136588,
+      "grad_norm": 0.43465063717377017,
+      "learning_rate": 3.844339622641509e-05,
+      "loss": 0.4867,
+      "step": 164
+    },
+    {
+      "epoch": 0.23354564755838642,
+      "grad_norm": 0.4724016009185264,
+      "learning_rate": 3.867924528301887e-05,
+      "loss": 0.4869,
+      "step": 165
+    },
+    {
+      "epoch": 0.23496107572540695,
+      "grad_norm": 0.5181772638936092,
+      "learning_rate": 3.8915094339622644e-05,
+      "loss": 0.4916,
+      "step": 166
+    },
+    {
+      "epoch": 0.23637650389242745,
+      "grad_norm": 0.3856581389322893,
+      "learning_rate": 3.9150943396226416e-05,
+      "loss": 0.4942,
+      "step": 167
+    },
+    {
+      "epoch": 0.23779193205944799,
+      "grad_norm": 0.6128154044577643,
+      "learning_rate": 3.938679245283019e-05,
+      "loss": 0.5066,
+      "step": 168
+    },
+    {
+      "epoch": 0.23920736022646852,
+      "grad_norm": 0.5053186581991689,
+      "learning_rate": 3.962264150943397e-05,
+      "loss": 0.4964,
+      "step": 169
+    },
+    {
+      "epoch": 0.24062278839348902,
+      "grad_norm": 0.49986453289811006,
+      "learning_rate": 3.985849056603774e-05,
+      "loss": 0.5,
+      "step": 170
+    },
+    {
+      "epoch": 0.24203821656050956,
+      "grad_norm": 0.6792847076210307,
+      "learning_rate": 4.009433962264151e-05,
+      "loss": 0.5246,
+      "step": 171
+    },
+    {
+      "epoch": 0.2434536447275301,
+      "grad_norm": 0.6145656510061281,
+      "learning_rate": 4.0330188679245284e-05,
+      "loss": 0.4873,
+      "step": 172
+    },
+    {
+      "epoch": 0.2448690728945506,
+      "grad_norm": 0.4118351232509526,
+      "learning_rate": 4.0566037735849064e-05,
+      "loss": 0.4843,
+      "step": 173
+    },
+    {
+      "epoch": 0.24628450106157113,
+      "grad_norm": 0.6392886230615822,
+      "learning_rate": 4.080188679245283e-05,
+      "loss": 0.4967,
+      "step": 174
+    },
+    {
+      "epoch": 0.24769992922859166,
+      "grad_norm": 0.34893502077766275,
+      "learning_rate": 4.103773584905661e-05,
+      "loss": 0.4825,
+      "step": 175
+    },
+    {
+      "epoch": 0.24911535739561216,
+      "grad_norm": 0.5176109420998536,
+      "learning_rate": 4.127358490566038e-05,
+      "loss": 0.4744,
+      "step": 176
+    },
+    {
+      "epoch": 0.2505307855626327,
+      "grad_norm": 0.4123363340874883,
+      "learning_rate": 4.150943396226415e-05,
+      "loss": 0.4959,
+      "step": 177
+    },
+    {
+      "epoch": 0.2519462137296532,
+      "grad_norm": 0.450719206533509,
+      "learning_rate": 4.1745283018867925e-05,
+      "loss": 0.4881,
+      "step": 178
+    },
+    {
+      "epoch": 0.25336164189667376,
+      "grad_norm": 0.43351355624228327,
+      "learning_rate": 4.1981132075471704e-05,
+      "loss": 0.4999,
+      "step": 179
+    },
+    {
+      "epoch": 0.25477707006369427,
+      "grad_norm": 0.33664144300948656,
+      "learning_rate": 4.221698113207547e-05,
+      "loss": 0.5142,
+      "step": 180
+    },
+    {
+      "epoch": 0.25619249823071477,
+      "grad_norm": 0.3837742618615644,
+      "learning_rate": 4.245283018867925e-05,
+      "loss": 0.4733,
+      "step": 181
+    },
+    {
+      "epoch": 0.25760792639773533,
+      "grad_norm": 0.3151243355874945,
+      "learning_rate": 4.268867924528302e-05,
+      "loss": 0.4657,
+      "step": 182
+    },
+    {
+      "epoch": 0.25902335456475584,
+      "grad_norm": 0.3335622550176586,
+      "learning_rate": 4.292452830188679e-05,
+      "loss": 0.4772,
+      "step": 183
+    },
+    {
+      "epoch": 0.26043878273177634,
+      "grad_norm": 0.37628468547226257,
+      "learning_rate": 4.3160377358490565e-05,
+      "loss": 0.4848,
+      "step": 184
+    },
+    {
+      "epoch": 0.2618542108987969,
+      "grad_norm": 0.34250164420941787,
+      "learning_rate": 4.3396226415094345e-05,
+      "loss": 0.4963,
+      "step": 185
+    },
+    {
+      "epoch": 0.2632696390658174,
+      "grad_norm": 0.390192986807738,
+      "learning_rate": 4.363207547169812e-05,
+      "loss": 0.513,
+      "step": 186
+    },
+    {
+      "epoch": 0.2646850672328379,
+      "grad_norm": 0.3657597965323173,
+      "learning_rate": 4.386792452830189e-05,
+      "loss": 0.5028,
+      "step": 187
+    },
+    {
+      "epoch": 0.26610049539985847,
+      "grad_norm": 0.3552471961943396,
+      "learning_rate": 4.410377358490566e-05,
+      "loss": 0.4958,
+      "step": 188
+    },
+    {
+      "epoch": 0.267515923566879,
+      "grad_norm": 0.3751793282608708,
+      "learning_rate": 4.433962264150944e-05,
+      "loss": 0.5002,
+      "step": 189
+    },
+    {
+      "epoch": 0.2689313517338995,
+      "grad_norm": 0.31709289415813285,
+      "learning_rate": 4.4575471698113206e-05,
+      "loss": 0.485,
+      "step": 190
+    },
+    {
+      "epoch": 0.27034677990092004,
+      "grad_norm": 0.36930880358545004,
+      "learning_rate": 4.4811320754716985e-05,
+      "loss": 0.4763,
+      "step": 191
+    },
+    {
+      "epoch": 0.27176220806794055,
+      "grad_norm": 0.38164778972975844,
+      "learning_rate": 4.504716981132076e-05,
+      "loss": 0.4796,
+      "step": 192
+    },
+    {
+      "epoch": 0.27317763623496105,
+      "grad_norm": 0.3827965077528942,
+      "learning_rate": 4.528301886792453e-05,
+      "loss": 0.4978,
+      "step": 193
+    },
+    {
+      "epoch": 0.2745930644019816,
+      "grad_norm": 0.3892538141587405,
+      "learning_rate": 4.55188679245283e-05,
+      "loss": 0.4712,
+      "step": 194
+    },
+    {
+      "epoch": 0.2760084925690021,
+      "grad_norm": 0.3653595080312334,
+      "learning_rate": 4.575471698113208e-05,
+      "loss": 0.485,
+      "step": 195
+    },
+    {
+      "epoch": 0.2774239207360226,
+      "grad_norm": 0.3593560802498003,
+      "learning_rate": 4.5990566037735846e-05,
+      "loss": 0.4581,
+      "step": 196
+    },
+    {
+      "epoch": 0.2788393489030432,
+      "grad_norm": 0.3833174718177277,
+      "learning_rate": 4.6226415094339625e-05,
+      "loss": 0.4868,
+      "step": 197
+    },
+    {
+      "epoch": 0.2802547770700637,
+      "grad_norm": 0.3104445627049494,
+      "learning_rate": 4.64622641509434e-05,
+      "loss": 0.4939,
+      "step": 198
+    },
+    {
+      "epoch": 0.2816702052370842,
+      "grad_norm": 0.2925127148194562,
+      "learning_rate": 4.669811320754717e-05,
+      "loss": 0.4988,
+      "step": 199
+    },
+    {
+      "epoch": 0.28308563340410475,
+      "grad_norm": 0.34460020120708856,
+      "learning_rate": 4.693396226415094e-05,
+      "loss": 0.4918,
+      "step": 200
+    },
+    {
+      "epoch": 0.28450106157112526,
+      "grad_norm": 0.31227346468617606,
+      "learning_rate": 4.716981132075472e-05,
+      "loss": 0.4995,
+      "step": 201
+    },
+    {
+      "epoch": 0.28591648973814576,
+      "grad_norm": 0.3640317902107833,
+      "learning_rate": 4.7405660377358494e-05,
+      "loss": 0.4707,
+      "step": 202
+    },
+    {
+      "epoch": 0.2873319179051663,
+      "grad_norm": 0.3145870893766912,
+      "learning_rate": 4.7641509433962266e-05,
+      "loss": 0.4995,
+      "step": 203
+    },
+    {
+      "epoch": 0.28874734607218683,
+      "grad_norm": 0.32355667169539304,
+      "learning_rate": 4.787735849056604e-05,
+      "loss": 0.4844,
+      "step": 204
+    },
+    {
+      "epoch": 0.29016277423920733,
+      "grad_norm": 0.36886406958455736,
+      "learning_rate": 4.811320754716982e-05,
+      "loss": 0.5,
+      "step": 205
+    },
+    {
+      "epoch": 0.2915782024062279,
+      "grad_norm": 0.28157776049221367,
+      "learning_rate": 4.834905660377358e-05,
+      "loss": 0.4897,
+      "step": 206
+    },
+    {
+      "epoch": 0.2929936305732484,
+      "grad_norm": 0.7299868380151449,
+      "learning_rate": 4.858490566037736e-05,
+      "loss": 0.496,
+      "step": 207
+    },
+    {
+      "epoch": 0.29440905874026896,
+      "grad_norm": 0.3571637911618915,
+      "learning_rate": 4.8820754716981134e-05,
+      "loss": 0.4859,
+      "step": 208
+    },
+    {
+      "epoch": 0.29582448690728946,
+      "grad_norm": 0.31598318128466457,
+      "learning_rate": 4.9056603773584906e-05,
+      "loss": 0.4684,
+      "step": 209
+    },
+    {
+      "epoch": 0.29723991507430997,
+      "grad_norm": 0.35704049963583556,
+      "learning_rate": 4.929245283018868e-05,
+      "loss": 0.4653,
+      "step": 210
+    },
+    {
+      "epoch": 0.29865534324133053,
+      "grad_norm": 0.3380809563649105,
+      "learning_rate": 4.952830188679246e-05,
+      "loss": 0.4972,
+      "step": 211
+    },
+    {
+      "epoch": 0.30007077140835103,
+      "grad_norm": 0.2893108910416616,
+      "learning_rate": 4.976415094339622e-05,
+      "loss": 0.4672,
+      "step": 212
+    },
+    {
+      "epoch": 0.30148619957537154,
+      "grad_norm": 0.34105686642163857,
+      "learning_rate": 5e-05,
+      "loss": 0.4833,
+      "step": 213
+    },
+    {
+      "epoch": 0.3029016277423921,
+      "grad_norm": 0.3654037324084374,
+      "learning_rate": 4.997376705141658e-05,
+      "loss": 0.4787,
+      "step": 214
+    },
+    {
+      "epoch": 0.3043170559094126,
+      "grad_norm": 0.33962044855755386,
+      "learning_rate": 4.994753410283316e-05,
+      "loss": 0.4917,
+      "step": 215
+    },
+    {
+      "epoch": 0.3057324840764331,
+      "grad_norm": 0.3651658064563787,
+      "learning_rate": 4.9921301154249736e-05,
+      "loss": 0.4731,
+      "step": 216
+    },
+    {
+      "epoch": 0.30714791224345367,
+      "grad_norm": 0.39085977965098573,
+      "learning_rate": 4.989506820566632e-05,
+      "loss": 0.4966,
+      "step": 217
+    },
+    {
+      "epoch": 0.3085633404104742,
+      "grad_norm": 0.38749937841905047,
+      "learning_rate": 4.98688352570829e-05,
+      "loss": 0.4751,
+      "step": 218
+    },
+    {
+      "epoch": 0.3099787685774947,
+      "grad_norm": 0.31799760427494933,
+      "learning_rate": 4.984260230849948e-05,
+      "loss": 0.4733,
+      "step": 219
+    },
+    {
+      "epoch": 0.31139419674451524,
+      "grad_norm": 0.3359905056993565,
+      "learning_rate": 4.981636935991606e-05,
+      "loss": 0.467,
+      "step": 220
+    },
+    {
+      "epoch": 0.31280962491153574,
+      "grad_norm": 0.40511778428597134,
+      "learning_rate": 4.979013641133264e-05,
+      "loss": 0.5045,
+      "step": 221
+    },
+    {
+      "epoch": 0.31422505307855625,
+      "grad_norm": 0.379192341890831,
+      "learning_rate": 4.976390346274922e-05,
+      "loss": 0.4713,
+      "step": 222
+    },
+    {
+      "epoch": 0.3156404812455768,
+      "grad_norm": 0.38069431190869296,
+      "learning_rate": 4.97376705141658e-05,
+      "loss": 0.4691,
+      "step": 223
+    },
+    {
+      "epoch": 0.3170559094125973,
+      "grad_norm": 0.40512176731248734,
+      "learning_rate": 4.971143756558237e-05,
+      "loss": 0.4952,
+      "step": 224
+    },
+    {
+      "epoch": 0.3184713375796178,
+      "grad_norm": 0.42869246316273457,
+      "learning_rate": 4.968520461699895e-05,
+      "loss": 0.479,
+      "step": 225
+    },
+    {
+      "epoch": 0.3198867657466384,
+      "grad_norm": 0.34166560604627294,
+      "learning_rate": 4.965897166841553e-05,
+      "loss": 0.4616,
+      "step": 226
+    },
+    {
+      "epoch": 0.3213021939136589,
+      "grad_norm": 0.41668612296835184,
+      "learning_rate": 4.963273871983211e-05,
+      "loss": 0.476,
+      "step": 227
+    },
+    {
+      "epoch": 0.3227176220806794,
+      "grad_norm": 0.3444355148419626,
+      "learning_rate": 4.960650577124869e-05,
+      "loss": 0.4845,
+      "step": 228
+    },
+    {
+      "epoch": 0.32413305024769995,
+      "grad_norm": 0.4082393224656534,
+      "learning_rate": 4.958027282266527e-05,
+      "loss": 0.5083,
+      "step": 229
+    },
+    {
+      "epoch": 0.32554847841472045,
+      "grad_norm": 0.3581858800328615,
+      "learning_rate": 4.955403987408185e-05,
+      "loss": 0.4887,
+      "step": 230
+    },
+    {
+      "epoch": 0.32696390658174096,
+      "grad_norm": 0.3604962793037811,
+      "learning_rate": 4.952780692549843e-05,
+      "loss": 0.51,
+      "step": 231
+    },
+    {
+      "epoch": 0.3283793347487615,
+      "grad_norm": 0.3764941078516145,
+      "learning_rate": 4.950157397691501e-05,
+      "loss": 0.456,
+      "step": 232
+    },
+    {
+      "epoch": 0.329794762915782,
+      "grad_norm": 0.3111710179843355,
+      "learning_rate": 4.947534102833158e-05,
+      "loss": 0.4691,
+      "step": 233
+    },
+    {
+      "epoch": 0.33121019108280253,
+      "grad_norm": 0.35045316900302903,
+      "learning_rate": 4.944910807974817e-05,
+      "loss": 0.4875,
+      "step": 234
+    },
+    {
+      "epoch": 0.3326256192498231,
+      "grad_norm": 0.3162652367435176,
+      "learning_rate": 4.942287513116475e-05,
+      "loss": 0.4553,
+      "step": 235
+    },
+    {
+      "epoch": 0.3340410474168436,
+      "grad_norm": 0.3360681973991888,
+      "learning_rate": 4.939664218258133e-05,
+      "loss": 0.4796,
+      "step": 236
+    },
+    {
+      "epoch": 0.3354564755838641,
+      "grad_norm": 0.3741851712229335,
+      "learning_rate": 4.93704092339979e-05,
+      "loss": 0.4571,
+      "step": 237
+    },
+    {
+      "epoch": 0.33687190375088466,
+      "grad_norm": 0.3718199670391826,
+      "learning_rate": 4.934417628541448e-05,
+      "loss": 0.4841,
+      "step": 238
+    },
+    {
+      "epoch": 0.33828733191790517,
+      "grad_norm": 0.35326975836585,
+      "learning_rate": 4.931794333683106e-05,
+      "loss": 0.4838,
+      "step": 239
+    },
+    {
+      "epoch": 0.33970276008492567,
+      "grad_norm": 0.36254161920550493,
+      "learning_rate": 4.929171038824764e-05,
+      "loss": 0.4607,
+      "step": 240
+    },
+    {
+      "epoch": 0.34111818825194623,
+      "grad_norm": 0.4305497527280153,
+      "learning_rate": 4.926547743966422e-05,
+      "loss": 0.4808,
+      "step": 241
+    },
+    {
+      "epoch": 0.34253361641896674,
+      "grad_norm": 0.3854579649210384,
+      "learning_rate": 4.92392444910808e-05,
+      "loss": 0.4612,
+      "step": 242
+    },
+    {
+      "epoch": 0.34394904458598724,
+      "grad_norm": 0.37381038850584963,
+      "learning_rate": 4.921301154249738e-05,
+      "loss": 0.4681,
+      "step": 243
+    },
+    {
+      "epoch": 0.3453644727530078,
+      "grad_norm": 0.34094705578286977,
+      "learning_rate": 4.918677859391396e-05,
+      "loss": 0.4748,
+      "step": 244
+    },
+    {
+      "epoch": 0.3467799009200283,
+      "grad_norm": 0.43256806769218703,
+      "learning_rate": 4.916054564533054e-05,
+      "loss": 0.4869,
+      "step": 245
+    },
+    {
+      "epoch": 0.3481953290870488,
+      "grad_norm": 0.3512540885487744,
+      "learning_rate": 4.913431269674712e-05,
+      "loss": 0.4749,
+      "step": 246
+    },
+    {
+      "epoch": 0.34961075725406937,
+      "grad_norm": 0.4375420690405016,
+      "learning_rate": 4.910807974816369e-05,
+      "loss": 0.4567,
+      "step": 247
+    },
+    {
+      "epoch": 0.3510261854210899,
+      "grad_norm": 0.31331139412521325,
+      "learning_rate": 4.908184679958028e-05,
+      "loss": 0.4956,
+      "step": 248
+    },
+    {
+      "epoch": 0.3524416135881104,
+      "grad_norm": 0.40569597255126255,
+      "learning_rate": 4.905561385099686e-05,
+      "loss": 0.479,
+      "step": 249
+    },
+    {
+      "epoch": 0.35385704175513094,
+      "grad_norm": 0.3015217033905556,
+      "learning_rate": 4.902938090241343e-05,
+      "loss": 0.4631,
+      "step": 250
+    },
+    {
+      "epoch": 0.35527246992215145,
+      "grad_norm": 0.36234313075601515,
+      "learning_rate": 4.900314795383001e-05,
+      "loss": 0.4788,
+      "step": 251
+    },
+    {
+      "epoch": 0.35668789808917195,
+      "grad_norm": 0.3350987908357145,
+      "learning_rate": 4.897691500524659e-05,
+      "loss": 0.4666,
+      "step": 252
+    },
+    {
+      "epoch": 0.3581033262561925,
+      "grad_norm": 0.30055620666911115,
+      "learning_rate": 4.895068205666317e-05,
+      "loss": 0.4896,
+      "step": 253
+    },
+    {
+      "epoch": 0.359518754423213,
+      "grad_norm": 0.3224634511396762,
+      "learning_rate": 4.8924449108079753e-05,
+      "loss": 0.4689,
+      "step": 254
+    },
+    {
+      "epoch": 0.3609341825902335,
+      "grad_norm": 0.3523537727385361,
+      "learning_rate": 4.889821615949633e-05,
+      "loss": 0.474,
+      "step": 255
+    },
+    {
+      "epoch": 0.3623496107572541,
+      "grad_norm": 0.32606824732367184,
+      "learning_rate": 4.887198321091291e-05,
+      "loss": 0.483,
+      "step": 256
+    },
+    {
+      "epoch": 0.3637650389242746,
+      "grad_norm": 0.35312194584977225,
+      "learning_rate": 4.884575026232949e-05,
+      "loss": 0.4616,
+      "step": 257
+    },
+    {
+      "epoch": 0.3651804670912951,
+      "grad_norm": 0.2977039251083076,
+      "learning_rate": 4.881951731374607e-05,
+      "loss": 0.4893,
+      "step": 258
+    },
+    {
+      "epoch": 0.36659589525831565,
+      "grad_norm": 0.36965910622460707,
+      "learning_rate": 4.879328436516265e-05,
+      "loss": 0.484,
+      "step": 259
+    },
+    {
+      "epoch": 0.36801132342533616,
+      "grad_norm": 0.32694450731488317,
+      "learning_rate": 4.876705141657922e-05,
+      "loss": 0.4734,
+      "step": 260
+    },
+    {
+      "epoch": 0.36942675159235666,
+      "grad_norm": 0.33797975840114225,
+      "learning_rate": 4.87408184679958e-05,
+      "loss": 0.4609,
+      "step": 261
+    },
+    {
+      "epoch": 0.3708421797593772,
+      "grad_norm": 0.5188557226310996,
+      "learning_rate": 4.871458551941239e-05,
+      "loss": 0.4695,
+      "step": 262
+    },
+    {
+      "epoch": 0.3722576079263977,
+      "grad_norm": 0.28372641106405255,
+      "learning_rate": 4.868835257082896e-05,
+      "loss": 0.4505,
+      "step": 263
+    },
+    {
+      "epoch": 0.37367303609341823,
+      "grad_norm": 0.37728420455185746,
+      "learning_rate": 4.866211962224554e-05,
+      "loss": 0.4699,
+      "step": 264
+    },
+    {
+      "epoch": 0.3750884642604388,
+      "grad_norm": 0.27926539871155076,
+      "learning_rate": 4.863588667366212e-05,
+      "loss": 0.4497,
+      "step": 265
+    },
+    {
+      "epoch": 0.3765038924274593,
+      "grad_norm": 0.31523862716113915,
+      "learning_rate": 4.86096537250787e-05,
+      "loss": 0.4794,
+      "step": 266
+    },
+    {
+      "epoch": 0.37791932059447986,
+      "grad_norm": 0.3559960078668447,
+      "learning_rate": 4.8583420776495284e-05,
+      "loss": 0.4838,
+      "step": 267
+    },
+    {
+      "epoch": 0.37933474876150036,
+      "grad_norm": 0.26459931841726647,
+      "learning_rate": 4.855718782791186e-05,
+      "loss": 0.4701,
+      "step": 268
+    },
+    {
+      "epoch": 0.38075017692852087,
+      "grad_norm": 0.42754490475895474,
+      "learning_rate": 4.853095487932844e-05,
+      "loss": 0.4833,
+      "step": 269
+    },
+    {
+      "epoch": 0.3821656050955414,
+      "grad_norm": 0.31990202221223335,
+      "learning_rate": 4.850472193074502e-05,
+      "loss": 0.4664,
+      "step": 270
+    },
+    {
+      "epoch": 0.38358103326256193,
+      "grad_norm": 0.32284874424979937,
+      "learning_rate": 4.84784889821616e-05,
+      "loss": 0.4646,
+      "step": 271
+    },
+    {
+      "epoch": 0.38499646142958244,
+      "grad_norm": 0.3567088507467582,
+      "learning_rate": 4.845225603357818e-05,
+      "loss": 0.4899,
+      "step": 272
+    },
+    {
+      "epoch": 0.386411889596603,
+      "grad_norm": 0.28350442390303104,
+      "learning_rate": 4.842602308499475e-05,
+      "loss": 0.4628,
+      "step": 273
+    },
+    {
+      "epoch": 0.3878273177636235,
+      "grad_norm": 0.3380796093379581,
+      "learning_rate": 4.839979013641133e-05,
+      "loss": 0.4798,
+      "step": 274
+    },
+    {
+      "epoch": 0.389242745930644,
+      "grad_norm": 0.30402306763549686,
+      "learning_rate": 4.837355718782791e-05,
+      "loss": 0.478,
+      "step": 275
+    },
+    {
+      "epoch": 0.39065817409766457,
+      "grad_norm": 0.33347801143181516,
+      "learning_rate": 4.834732423924449e-05,
+      "loss": 0.4725,
+      "step": 276
+    },
+    {
+      "epoch": 0.3920736022646851,
+      "grad_norm": 0.3229830448029133,
+      "learning_rate": 4.832109129066107e-05,
+      "loss": 0.4661,
+      "step": 277
+    },
+    {
+      "epoch": 0.3934890304317056,
+      "grad_norm": 0.3214414708827648,
+      "learning_rate": 4.829485834207765e-05,
+      "loss": 0.4389,
+      "step": 278
+    },
+    {
+      "epoch": 0.39490445859872614,
+      "grad_norm": 0.2977414670901615,
+      "learning_rate": 4.8268625393494233e-05,
+      "loss": 0.4534,
+      "step": 279
+    },
+    {
+      "epoch": 0.39631988676574664,
+      "grad_norm": 0.3494964999721987,
+      "learning_rate": 4.8242392444910814e-05,
+      "loss": 0.4506,
+      "step": 280
+    },
+    {
+      "epoch": 0.39773531493276715,
+      "grad_norm": 0.33432250104369315,
+      "learning_rate": 4.821615949632739e-05,
+      "loss": 0.4794,
+      "step": 281
+    },
+    {
+      "epoch": 0.3991507430997877,
+      "grad_norm": 0.2989920920241222,
+      "learning_rate": 4.818992654774397e-05,
+      "loss": 0.4679,
+      "step": 282
+    },
+    {
+      "epoch": 0.4005661712668082,
+      "grad_norm": 0.33484591865165336,
+      "learning_rate": 4.816369359916055e-05,
+      "loss": 0.48,
+      "step": 283
+    },
+    {
+      "epoch": 0.4019815994338287,
+      "grad_norm": 0.308359897039709,
+      "learning_rate": 4.813746065057713e-05,
+      "loss": 0.4601,
+      "step": 284
+    },
+    {
+      "epoch": 0.4033970276008493,
+      "grad_norm": 0.34107458942761815,
+      "learning_rate": 4.811122770199371e-05,
+      "loss": 0.4775,
+      "step": 285
+    },
+    {
+      "epoch": 0.4048124557678698,
+      "grad_norm": 0.2790810489904784,
+      "learning_rate": 4.808499475341028e-05,
+      "loss": 0.4684,
+      "step": 286
+    },
+    {
+      "epoch": 0.4062278839348903,
+      "grad_norm": 0.31368051064270663,
+      "learning_rate": 4.805876180482686e-05,
+      "loss": 0.4524,
+      "step": 287
+    },
+    {
+      "epoch": 0.40764331210191085,
+      "grad_norm": 0.31954064273172456,
+      "learning_rate": 4.803252885624344e-05,
+      "loss": 0.4866,
+      "step": 288
+    },
+    {
+      "epoch": 0.40905874026893135,
+      "grad_norm": 0.3409185291839269,
+      "learning_rate": 4.800629590766002e-05,
+      "loss": 0.4696,
+      "step": 289
+    },
+    {
+      "epoch": 0.41047416843595186,
+      "grad_norm": 0.29737417322671006,
+      "learning_rate": 4.79800629590766e-05,
+      "loss": 0.4707,
+      "step": 290
+    },
+    {
+      "epoch": 0.4118895966029724,
+      "grad_norm": 0.3418785035707623,
+      "learning_rate": 4.795383001049318e-05,
+      "loss": 0.4494,
+      "step": 291
+    },
+    {
+      "epoch": 0.4133050247699929,
+      "grad_norm": 0.29733881088472347,
+      "learning_rate": 4.7927597061909764e-05,
+      "loss": 0.4741,
+      "step": 292
+    },
+    {
+      "epoch": 0.41472045293701343,
+      "grad_norm": 0.31999219722171973,
+      "learning_rate": 4.7901364113326344e-05,
+      "loss": 0.4565,
+      "step": 293
+    },
+    {
+      "epoch": 0.416135881104034,
+      "grad_norm": 0.2702853626810532,
+      "learning_rate": 4.787513116474292e-05,
+      "loss": 0.4789,
+      "step": 294
+    },
+    {
+      "epoch": 0.4175513092710545,
+      "grad_norm": 0.35698586881409045,
+      "learning_rate": 4.78488982161595e-05,
+      "loss": 0.4512,
+      "step": 295
+    },
+    {
+      "epoch": 0.418966737438075,
+      "grad_norm": 0.30683113400792583,
+      "learning_rate": 4.782266526757608e-05,
+      "loss": 0.4681,
+      "step": 296
+    },
+    {
+      "epoch": 0.42038216560509556,
+      "grad_norm": 0.3166932092062115,
+      "learning_rate": 4.779643231899266e-05,
+      "loss": 0.4611,
+      "step": 297
+    },
+    {
+      "epoch": 0.42179759377211606,
+      "grad_norm": 0.28440777282553725,
+      "learning_rate": 4.777019937040924e-05,
+      "loss": 0.4549,
+      "step": 298
+    },
+    {
+      "epoch": 0.42321302193913657,
+      "grad_norm": 0.30989709115616915,
+      "learning_rate": 4.774396642182581e-05,
+      "loss": 0.4538,
+      "step": 299
+    },
+    {
+      "epoch": 0.42462845010615713,
+      "grad_norm": 0.24260466757465912,
+      "learning_rate": 4.771773347324239e-05,
+      "loss": 0.4681,
+      "step": 300
+    },
+    {
+      "epoch": 0.42604387827317763,
+      "grad_norm": 0.3391264946077127,
+      "learning_rate": 4.769150052465897e-05,
+      "loss": 0.4617,
+      "step": 301
+    },
+    {
+      "epoch": 0.42745930644019814,
+      "grad_norm": 0.3008596481448339,
+      "learning_rate": 4.766526757607555e-05,
+      "loss": 0.4609,
+      "step": 302
+    },
+    {
+      "epoch": 0.4288747346072187,
+      "grad_norm": 0.33198306390049503,
+      "learning_rate": 4.763903462749213e-05,
+      "loss": 0.4776,
+      "step": 303
+    },
+    {
+      "epoch": 0.4302901627742392,
+      "grad_norm": 0.2962026562742108,
+      "learning_rate": 4.7612801678908713e-05,
+      "loss": 0.4617,
+      "step": 304
+    },
+    {
+      "epoch": 0.4317055909412597,
+      "grad_norm": 0.9650530020569035,
+      "learning_rate": 4.7586568730325294e-05,
+      "loss": 0.4646,
+      "step": 305
+    },
+    {
+      "epoch": 0.43312101910828027,
+      "grad_norm": 0.2876675793351413,
+      "learning_rate": 4.7560335781741874e-05,
+      "loss": 0.4564,
+      "step": 306
+    },
+    {
+      "epoch": 0.4345364472753008,
+      "grad_norm": 0.3037747439699752,
+      "learning_rate": 4.753410283315845e-05,
+      "loss": 0.487,
+      "step": 307
+    },
+    {
+      "epoch": 0.4359518754423213,
+      "grad_norm": 0.299098005284229,
+      "learning_rate": 4.750786988457503e-05,
+      "loss": 0.4568,
+      "step": 308
+    },
+    {
+      "epoch": 0.43736730360934184,
+      "grad_norm": 0.26342453389940673,
+      "learning_rate": 4.748163693599161e-05,
+      "loss": 0.4602,
+      "step": 309
+    },
+    {
+      "epoch": 0.43878273177636234,
+      "grad_norm": 0.26189351741312544,
+      "learning_rate": 4.745540398740819e-05,
+      "loss": 0.4575,
+      "step": 310
+    },
+    {
+      "epoch": 0.44019815994338285,
+      "grad_norm": 0.269146148673188,
+      "learning_rate": 4.742917103882477e-05,
+      "loss": 0.4472,
+      "step": 311
+    },
+    {
+      "epoch": 0.4416135881104034,
+      "grad_norm": 0.27774884805503863,
+      "learning_rate": 4.740293809024134e-05,
+      "loss": 0.4801,
+      "step": 312
+    },
+    {
+      "epoch": 0.4430290162774239,
+      "grad_norm": 0.2828678588034561,
+      "learning_rate": 4.737670514165792e-05,
+      "loss": 0.4476,
+      "step": 313
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 0.26381381025775535,
+      "learning_rate": 4.73504721930745e-05,
+      "loss": 0.4526,
+      "step": 314
+    },
+    {
+      "epoch": 0.445859872611465,
+      "grad_norm": 0.29545201089339096,
+      "learning_rate": 4.732423924449108e-05,
+      "loss": 0.4456,
+      "step": 315
+    },
+    {
+      "epoch": 0.4472753007784855,
+      "grad_norm": 0.31571268465596236,
+      "learning_rate": 4.7298006295907657e-05,
+      "loss": 0.4763,
+      "step": 316
+    },
+    {
+      "epoch": 0.448690728945506,
+      "grad_norm": 0.3139992834947803,
+      "learning_rate": 4.7271773347324244e-05,
+      "loss": 0.453,
+      "step": 317
+    },
+    {
+      "epoch": 0.45010615711252655,
+      "grad_norm": 0.2954480433149446,
+      "learning_rate": 4.7245540398740824e-05,
+      "loss": 0.4516,
+      "step": 318
+    },
+    {
+      "epoch": 0.45152158527954706,
+      "grad_norm": 0.3629808241350359,
+      "learning_rate": 4.7219307450157404e-05,
+      "loss": 0.4674,
+      "step": 319
+    },
+    {
+      "epoch": 0.45293701344656756,
+      "grad_norm": 0.3526526315963183,
+      "learning_rate": 4.719307450157398e-05,
+      "loss": 0.46,
+      "step": 320
+    },
+    {
+      "epoch": 0.4543524416135881,
+      "grad_norm": 0.42467086373484514,
+      "learning_rate": 4.716684155299056e-05,
+      "loss": 0.4722,
+      "step": 321
+    },
+    {
+      "epoch": 0.4557678697806086,
+      "grad_norm": 0.2848455631907692,
+      "learning_rate": 4.714060860440714e-05,
+      "loss": 0.4677,
+      "step": 322
+    },
+    {
+      "epoch": 0.45718329794762913,
+      "grad_norm": 0.3680582877057587,
+      "learning_rate": 4.711437565582372e-05,
+      "loss": 0.4626,
+      "step": 323
+    },
+    {
+      "epoch": 0.4585987261146497,
+      "grad_norm": 0.30480571190360467,
+      "learning_rate": 4.70881427072403e-05,
+      "loss": 0.4734,
+      "step": 324
+    },
+    {
+      "epoch": 0.4600141542816702,
+      "grad_norm": 0.3738153957980129,
+      "learning_rate": 4.706190975865687e-05,
+      "loss": 0.4672,
+      "step": 325
+    },
+    {
+      "epoch": 0.46142958244869076,
+      "grad_norm": 0.3202269092407385,
+      "learning_rate": 4.703567681007345e-05,
+      "loss": 0.4529,
+      "step": 326
+    },
+    {
+      "epoch": 0.46284501061571126,
+      "grad_norm": 0.3705235889244079,
+      "learning_rate": 4.700944386149003e-05,
+      "loss": 0.457,
+      "step": 327
+    },
+    {
+      "epoch": 0.46426043878273177,
+      "grad_norm": 0.3300580785020962,
+      "learning_rate": 4.698321091290661e-05,
+      "loss": 0.4634,
+      "step": 328
+    },
+    {
+      "epoch": 0.4656758669497523,
+      "grad_norm": 0.35469328535410066,
+      "learning_rate": 4.695697796432319e-05,
+      "loss": 0.4733,
+      "step": 329
+    },
+    {
+      "epoch": 0.46709129511677283,
+      "grad_norm": 0.3194681818790911,
+      "learning_rate": 4.693074501573977e-05,
+      "loss": 0.4427,
+      "step": 330
+    },
+    {
+      "epoch": 0.46850672328379334,
+      "grad_norm": 0.3613032196238565,
+      "learning_rate": 4.6904512067156354e-05,
+      "loss": 0.4556,
+      "step": 331
+    },
+    {
+      "epoch": 0.4699221514508139,
+      "grad_norm": 0.27287526828539865,
+      "learning_rate": 4.6878279118572934e-05,
+      "loss": 0.4531,
+      "step": 332
+    },
+    {
+      "epoch": 0.4713375796178344,
+      "grad_norm": 0.426853472134396,
+      "learning_rate": 4.685204616998951e-05,
+      "loss": 0.4856,
+      "step": 333
+    },
+    {
+      "epoch": 0.4727530077848549,
+      "grad_norm": 0.2764447274960176,
+      "learning_rate": 4.682581322140609e-05,
+      "loss": 0.4614,
+      "step": 334
+    },
+    {
+      "epoch": 0.47416843595187547,
+      "grad_norm": 0.3747012954551615,
+      "learning_rate": 4.679958027282267e-05,
+      "loss": 0.4484,
+      "step": 335
+    },
+    {
+      "epoch": 0.47558386411889597,
+      "grad_norm": 0.26796928979959844,
+      "learning_rate": 4.677334732423925e-05,
+      "loss": 0.4357,
+      "step": 336
+    },
+    {
+      "epoch": 0.4769992922859165,
+      "grad_norm": 0.3566137792471709,
+      "learning_rate": 4.674711437565583e-05,
+      "loss": 0.4731,
+      "step": 337
+    },
+    {
+      "epoch": 0.47841472045293704,
+      "grad_norm": 0.3475361430572997,
+      "learning_rate": 4.67208814270724e-05,
+      "loss": 0.4699,
+      "step": 338
+    },
+    {
+      "epoch": 0.47983014861995754,
+      "grad_norm": 0.4581592514232306,
+      "learning_rate": 4.669464847848898e-05,
+      "loss": 0.451,
+      "step": 339
+    },
+    {
+      "epoch": 0.48124557678697805,
+      "grad_norm": 0.3627293439819416,
+      "learning_rate": 4.666841552990556e-05,
+      "loss": 0.4625,
+      "step": 340
+    },
+    {
+      "epoch": 0.4826610049539986,
+      "grad_norm": 0.3511019572881394,
+      "learning_rate": 4.664218258132214e-05,
+      "loss": 0.4551,
+      "step": 341
+    },
+    {
+      "epoch": 0.4840764331210191,
+      "grad_norm": 0.3234996409960452,
+      "learning_rate": 4.661594963273872e-05,
+      "loss": 0.4761,
+      "step": 342
+    },
+    {
+      "epoch": 0.4854918612880396,
+      "grad_norm": 0.3915392794609834,
+      "learning_rate": 4.65897166841553e-05,
+      "loss": 0.4819,
+      "step": 343
+    },
+    {
+      "epoch": 0.4869072894550602,
+      "grad_norm": 0.30334442650643895,
+      "learning_rate": 4.656348373557188e-05,
+      "loss": 0.4617,
+      "step": 344
+    },
+    {
+      "epoch": 0.4883227176220807,
+      "grad_norm": 0.3434333598334738,
+      "learning_rate": 4.6537250786988465e-05,
+      "loss": 0.4591,
+      "step": 345
+    },
+    {
+      "epoch": 0.4897381457891012,
+      "grad_norm": 0.3011418095074162,
+      "learning_rate": 4.651101783840504e-05,
+      "loss": 0.4593,
+      "step": 346
+    },
+    {
+      "epoch": 0.49115357395612175,
+      "grad_norm": 0.2931682762521998,
+      "learning_rate": 4.648478488982162e-05,
+      "loss": 0.4534,
+      "step": 347
+    },
+    {
+      "epoch": 0.49256900212314225,
+      "grad_norm": 0.27983031997568253,
+      "learning_rate": 4.64585519412382e-05,
+      "loss": 0.456,
+      "step": 348
+    },
+    {
+      "epoch": 0.49398443029016276,
+      "grad_norm": 0.3330628925072195,
+      "learning_rate": 4.643231899265478e-05,
+      "loss": 0.4583,
+      "step": 349
+    },
+    {
+      "epoch": 0.4953998584571833,
+      "grad_norm": 0.24344080894387007,
+      "learning_rate": 4.640608604407136e-05,
+      "loss": 0.4444,
+      "step": 350
+    },
+    {
+      "epoch": 0.4968152866242038,
+      "grad_norm": 0.3149584883086908,
+      "learning_rate": 4.637985309548793e-05,
+      "loss": 0.4629,
+      "step": 351
+    },
+    {
+      "epoch": 0.4982307147912243,
+      "grad_norm": 0.25711306244469556,
+      "learning_rate": 4.635362014690451e-05,
+      "loss": 0.4353,
+      "step": 352
+    },
+    {
+      "epoch": 0.4996461429582449,
+      "grad_norm": 0.2828675944672857,
+      "learning_rate": 4.632738719832109e-05,
+      "loss": 0.4431,
+      "step": 353
+    },
+    {
+      "epoch": 0.5010615711252654,
+      "grad_norm": 0.31965497332448656,
+      "learning_rate": 4.6301154249737674e-05,
+      "loss": 0.4683,
+      "step": 354
+    },
+    {
+      "epoch": 0.502476999292286,
+      "grad_norm": 0.29420711453793835,
+      "learning_rate": 4.627492130115425e-05,
+      "loss": 0.4779,
+      "step": 355
+    },
+    {
+      "epoch": 0.5038924274593064,
+      "grad_norm": 0.3095981627932921,
+      "learning_rate": 4.624868835257083e-05,
+      "loss": 0.4809,
+      "step": 356
+    },
+    {
+      "epoch": 0.505307855626327,
+      "grad_norm": 0.2784944679427735,
+      "learning_rate": 4.622245540398741e-05,
+      "loss": 0.4608,
+      "step": 357
+    },
+    {
+      "epoch": 0.5067232837933475,
+      "grad_norm": 0.29624109800363596,
+      "learning_rate": 4.619622245540399e-05,
+      "loss": 0.4472,
+      "step": 358
+    },
+    {
+      "epoch": 0.508138711960368,
+      "grad_norm": 0.28349708245921784,
+      "learning_rate": 4.616998950682057e-05,
+      "loss": 0.4632,
+      "step": 359
+    },
+    {
+      "epoch": 0.5095541401273885,
+      "grad_norm": 0.296812331973689,
+      "learning_rate": 4.614375655823715e-05,
+      "loss": 0.4419,
+      "step": 360
+    },
+    {
+      "epoch": 0.5109695682944091,
+      "grad_norm": 0.29521985654259064,
+      "learning_rate": 4.611752360965373e-05,
+      "loss": 0.4404,
+      "step": 361
+    },
+    {
+      "epoch": 0.5123849964614295,
+      "grad_norm": 0.30198848925830873,
+      "learning_rate": 4.609129066107031e-05,
+      "loss": 0.4764,
+      "step": 362
+    },
+    {
+      "epoch": 0.5138004246284501,
+      "grad_norm": 0.2913858999538317,
+      "learning_rate": 4.606505771248689e-05,
+      "loss": 0.4656,
+      "step": 363
+    },
+    {
+      "epoch": 0.5152158527954707,
+      "grad_norm": 0.29621855684908976,
+      "learning_rate": 4.603882476390346e-05,
+      "loss": 0.449,
+      "step": 364
+    },
+    {
+      "epoch": 0.5166312809624911,
+      "grad_norm": 0.3060245599564631,
+      "learning_rate": 4.601259181532004e-05,
+      "loss": 0.4642,
+      "step": 365
+    },
+    {
+      "epoch": 0.5180467091295117,
+      "grad_norm": 0.2926543846950487,
+      "learning_rate": 4.5986358866736623e-05,
+      "loss": 0.4503,
+      "step": 366
+    },
+    {
+      "epoch": 0.5194621372965322,
+      "grad_norm": 0.26926515134699414,
+      "learning_rate": 4.5960125918153204e-05,
+      "loss": 0.4617,
+      "step": 367
+    },
+    {
+      "epoch": 0.5208775654635527,
+      "grad_norm": 0.27299414729049654,
+      "learning_rate": 4.5933892969569784e-05,
+      "loss": 0.45,
+      "step": 368
+    },
+    {
+      "epoch": 0.5222929936305732,
+      "grad_norm": 0.25833671342973175,
+      "learning_rate": 4.590766002098636e-05,
+      "loss": 0.4646,
+      "step": 369
+    },
+    {
+      "epoch": 0.5237084217975938,
+      "grad_norm": 0.2839236221161132,
+      "learning_rate": 4.588142707240294e-05,
+      "loss": 0.4302,
+      "step": 370
+    },
+    {
+      "epoch": 0.5251238499646143,
+      "grad_norm": 0.2786810220220998,
+      "learning_rate": 4.585519412381952e-05,
+      "loss": 0.4643,
+      "step": 371
+    },
+    {
+      "epoch": 0.5265392781316348,
+      "grad_norm": 0.2669933920291989,
+      "learning_rate": 4.58289611752361e-05,
+      "loss": 0.4624,
+      "step": 372
+    },
+    {
+      "epoch": 0.5279547062986554,
+      "grad_norm": 0.2858980696521966,
+      "learning_rate": 4.580272822665268e-05,
+      "loss": 0.4649,
+      "step": 373
+    },
+    {
+      "epoch": 0.5293701344656758,
+      "grad_norm": 0.2622927831062312,
+      "learning_rate": 4.577649527806926e-05,
+      "loss": 0.4557,
+      "step": 374
+    },
+    {
+      "epoch": 0.5307855626326964,
+      "grad_norm": 0.27472852707977724,
+      "learning_rate": 4.575026232948584e-05,
+      "loss": 0.4457,
+      "step": 375
+    },
+    {
+      "epoch": 0.5322009907997169,
+      "grad_norm": 0.2727301904161722,
+      "learning_rate": 4.572402938090242e-05,
+      "loss": 0.4634,
+      "step": 376
+    },
+    {
+      "epoch": 0.5336164189667374,
+      "grad_norm": 0.27932575044288194,
+      "learning_rate": 4.569779643231899e-05,
+      "loss": 0.4537,
+      "step": 377
+    },
+    {
+      "epoch": 0.535031847133758,
+      "grad_norm": 0.30699616443917316,
+      "learning_rate": 4.567156348373557e-05,
+      "loss": 0.452,
+      "step": 378
+    },
+    {
+      "epoch": 0.5364472753007785,
+      "grad_norm": 0.30299341502733096,
+      "learning_rate": 4.5645330535152154e-05,
+      "loss": 0.474,
+      "step": 379
+    },
+    {
+      "epoch": 0.537862703467799,
+      "grad_norm": 0.2919960224567498,
+      "learning_rate": 4.5619097586568734e-05,
+      "loss": 0.4625,
+      "step": 380
+    },
+    {
+      "epoch": 0.5392781316348195,
+      "grad_norm": 0.2990740185851751,
+      "learning_rate": 4.5592864637985314e-05,
+      "loss": 0.4617,
+      "step": 381
+    },
+    {
+      "epoch": 0.5406935598018401,
+      "grad_norm": 0.2659823960473287,
+      "learning_rate": 4.556663168940189e-05,
+      "loss": 0.4477,
+      "step": 382
+    },
+    {
+      "epoch": 0.5421089879688605,
+      "grad_norm": 0.26728861667810855,
+      "learning_rate": 4.554039874081847e-05,
+      "loss": 0.4592,
+      "step": 383
+    },
+    {
+      "epoch": 0.5435244161358811,
+      "grad_norm": 0.2853257514446039,
+      "learning_rate": 4.551416579223505e-05,
+      "loss": 0.4565,
+      "step": 384
+    },
+    {
+      "epoch": 0.5449398443029017,
+      "grad_norm": 0.2625814267311583,
+      "learning_rate": 4.548793284365163e-05,
+      "loss": 0.4389,
+      "step": 385
+    },
+    {
+      "epoch": 0.5463552724699221,
+      "grad_norm": 0.26885049478533035,
+      "learning_rate": 4.546169989506821e-05,
+      "loss": 0.4516,
+      "step": 386
+    },
+    {
+      "epoch": 0.5477707006369427,
+      "grad_norm": 0.26586269774742693,
+      "learning_rate": 4.543546694648479e-05,
+      "loss": 0.4537,
+      "step": 387
+    },
+    {
+      "epoch": 0.5491861288039632,
+      "grad_norm": 0.2888622602970239,
+      "learning_rate": 4.540923399790137e-05,
+      "loss": 0.4458,
+      "step": 388
+    },
+    {
+      "epoch": 0.5506015569709837,
+      "grad_norm": 0.2811624027110681,
+      "learning_rate": 4.538300104931795e-05,
+      "loss": 0.4546,
+      "step": 389
+    },
+    {
+      "epoch": 0.5520169851380042,
+      "grad_norm": 0.2741830826094752,
+      "learning_rate": 4.535676810073452e-05,
+      "loss": 0.4571,
+      "step": 390
+    },
+    {
+      "epoch": 0.5534324133050248,
+      "grad_norm": 0.2642055656361625,
+      "learning_rate": 4.5330535152151103e-05,
+      "loss": 0.4581,
+      "step": 391
+    },
+    {
+      "epoch": 0.5548478414720452,
+      "grad_norm": 0.31492914277218653,
+      "learning_rate": 4.5304302203567684e-05,
+      "loss": 0.4375,
+      "step": 392
+    },
+    {
+      "epoch": 0.5562632696390658,
+      "grad_norm": 0.25431995851103983,
+      "learning_rate": 4.5278069254984264e-05,
+      "loss": 0.4434,
+      "step": 393
+    },
+    {
+      "epoch": 0.5576786978060864,
+      "grad_norm": 0.26192007224916264,
+      "learning_rate": 4.5251836306400844e-05,
+      "loss": 0.4464,
+      "step": 394
+    },
+    {
+      "epoch": 0.5590941259731068,
+      "grad_norm": 0.24224205861266807,
+      "learning_rate": 4.522560335781742e-05,
+      "loss": 0.4426,
+      "step": 395
+    },
+    {
+      "epoch": 0.5605095541401274,
+      "grad_norm": 0.2672754644431964,
+      "learning_rate": 4.5199370409234e-05,
+      "loss": 0.4624,
+      "step": 396
+    },
+    {
+      "epoch": 0.5619249823071479,
+      "grad_norm": 0.244036715749111,
+      "learning_rate": 4.517313746065058e-05,
+      "loss": 0.4584,
+      "step": 397
+    },
+    {
+      "epoch": 0.5633404104741684,
+      "grad_norm": 0.27071584258359827,
+      "learning_rate": 4.514690451206716e-05,
+      "loss": 0.4485,
+      "step": 398
+    },
+    {
+      "epoch": 0.564755838641189,
+      "grad_norm": 0.2219741153795276,
+      "learning_rate": 4.512067156348373e-05,
+      "loss": 0.4479,
+      "step": 399
+    },
+    {
+      "epoch": 0.5661712668082095,
+      "grad_norm": 0.25037227044897187,
+      "learning_rate": 4.509443861490032e-05,
+      "loss": 0.4285,
+      "step": 400
+    },
+    {
+      "epoch": 0.56758669497523,
+      "grad_norm": 0.2605986634772363,
+      "learning_rate": 4.50682056663169e-05,
+      "loss": 0.4554,
+      "step": 401
+    },
+    {
+      "epoch": 0.5690021231422505,
+      "grad_norm": 0.22255212223784326,
+      "learning_rate": 4.504197271773348e-05,
+      "loss": 0.4428,
+      "step": 402
+    },
+    {
+      "epoch": 0.5704175513092711,
+      "grad_norm": 0.24900365207529354,
+      "learning_rate": 4.501573976915005e-05,
+      "loss": 0.4534,
+      "step": 403
+    },
+    {
+      "epoch": 0.5718329794762915,
+      "grad_norm": 0.24923821687732778,
+      "learning_rate": 4.4989506820566634e-05,
+      "loss": 0.4465,
+      "step": 404
+    },
+    {
+      "epoch": 0.5732484076433121,
+      "grad_norm": 0.24844544013157835,
+      "learning_rate": 4.4963273871983214e-05,
+      "loss": 0.4717,
+      "step": 405
+    },
+    {
+      "epoch": 0.5746638358103326,
+      "grad_norm": 0.2535482892944214,
+      "learning_rate": 4.4937040923399794e-05,
+      "loss": 0.4661,
+      "step": 406
+    },
+    {
+      "epoch": 0.5760792639773531,
+      "grad_norm": 0.2876988774674945,
+      "learning_rate": 4.4910807974816374e-05,
+      "loss": 0.4442,
+      "step": 407
+    },
+    {
+      "epoch": 0.5774946921443737,
+      "grad_norm": 0.26993016677284876,
+      "learning_rate": 4.488457502623295e-05,
+      "loss": 0.4552,
+      "step": 408
+    },
+    {
+      "epoch": 0.5789101203113942,
+      "grad_norm": 0.3358646447309973,
+      "learning_rate": 4.485834207764953e-05,
+      "loss": 0.4624,
+      "step": 409
+    },
+    {
+      "epoch": 0.5803255484784147,
+      "grad_norm": 0.309893397096339,
+      "learning_rate": 4.483210912906611e-05,
+      "loss": 0.4559,
+      "step": 410
+    },
+    {
+      "epoch": 0.5817409766454352,
+      "grad_norm": 0.3386303499703173,
+      "learning_rate": 4.480587618048269e-05,
+      "loss": 0.4514,
+      "step": 411
+    },
+    {
+      "epoch": 0.5831564048124558,
+      "grad_norm": 0.31112587874364034,
+      "learning_rate": 4.477964323189926e-05,
+      "loss": 0.4561,
+      "step": 412
+    },
+    {
+      "epoch": 0.5845718329794763,
+      "grad_norm": 0.30748628264223893,
+      "learning_rate": 4.475341028331584e-05,
+      "loss": 0.4196,
+      "step": 413
+    },
+    {
+      "epoch": 0.5859872611464968,
+      "grad_norm": 0.3668354913960175,
+      "learning_rate": 4.472717733473243e-05,
+      "loss": 0.4593,
+      "step": 414
+    },
+    {
+      "epoch": 0.5874026893135174,
+      "grad_norm": 0.3176035120916479,
+      "learning_rate": 4.470094438614901e-05,
+      "loss": 0.463,
+      "step": 415
+    },
+    {
+      "epoch": 0.5888181174805379,
+      "grad_norm": 0.3330738765417286,
+      "learning_rate": 4.4674711437565583e-05,
+      "loss": 0.4385,
+      "step": 416
+    },
+    {
+      "epoch": 0.5902335456475584,
+      "grad_norm": 0.2893855800235713,
+      "learning_rate": 4.4648478488982164e-05,
+      "loss": 0.4641,
+      "step": 417
+    },
+    {
+      "epoch": 0.5916489738145789,
+      "grad_norm": 0.31639363837672396,
+      "learning_rate": 4.4622245540398744e-05,
+      "loss": 0.4403,
+      "step": 418
+    },
+    {
+      "epoch": 0.5930644019815995,
+      "grad_norm": 0.22355317350787565,
+      "learning_rate": 4.4596012591815324e-05,
+      "loss": 0.4404,
+      "step": 419
+    },
+    {
+      "epoch": 0.5944798301486199,
+      "grad_norm": 0.3576283924175163,
+      "learning_rate": 4.4569779643231905e-05,
+      "loss": 0.4644,
+      "step": 420
+    },
+    {
+      "epoch": 0.5958952583156405,
+      "grad_norm": 0.25816574164368156,
+      "learning_rate": 4.454354669464848e-05,
+      "loss": 0.4558,
+      "step": 421
+    },
+    {
+      "epoch": 0.5973106864826611,
+      "grad_norm": 0.33002317077220583,
+      "learning_rate": 4.451731374606506e-05,
+      "loss": 0.4704,
+      "step": 422
+    },
+    {
+      "epoch": 0.5987261146496815,
+      "grad_norm": 0.2493116914048726,
+      "learning_rate": 4.449108079748164e-05,
+      "loss": 0.4247,
+      "step": 423
+    },
+    {
+      "epoch": 0.6001415428167021,
+      "grad_norm": 0.26699909660092275,
+      "learning_rate": 4.446484784889822e-05,
+      "loss": 0.4252,
+      "step": 424
+    },
+    {
+      "epoch": 0.6015569709837226,
+      "grad_norm": 0.32848504997824807,
+      "learning_rate": 4.443861490031479e-05,
+      "loss": 0.4343,
+      "step": 425
+    },
+    {
+      "epoch": 0.6029723991507431,
+      "grad_norm": 0.25097142910919984,
+      "learning_rate": 4.441238195173137e-05,
+      "loss": 0.4683,
+      "step": 426
+    },
+    {
+      "epoch": 0.6043878273177636,
+      "grad_norm": 0.2959450784149391,
+      "learning_rate": 4.438614900314795e-05,
+      "loss": 0.4701,
+      "step": 427
+    },
+    {
+      "epoch": 0.6058032554847842,
+      "grad_norm": 0.3311493732639423,
+      "learning_rate": 4.435991605456454e-05,
+      "loss": 0.4587,
+      "step": 428
+    },
+    {
+      "epoch": 0.6072186836518046,
+      "grad_norm": 0.3043125758832461,
+      "learning_rate": 4.4333683105981114e-05,
+      "loss": 0.4576,
+      "step": 429
+    },
+    {
+      "epoch": 0.6086341118188252,
+      "grad_norm": 0.3040835362722505,
+      "learning_rate": 4.4307450157397694e-05,
+      "loss": 0.4594,
+      "step": 430
+    },
+    {
+      "epoch": 0.6100495399858458,
+      "grad_norm": 0.30376844362999295,
+      "learning_rate": 4.4281217208814274e-05,
+      "loss": 0.4303,
+      "step": 431
+    },
+    {
+      "epoch": 0.6114649681528662,
+      "grad_norm": 0.36277328985368,
+      "learning_rate": 4.4254984260230854e-05,
+      "loss": 0.453,
+      "step": 432
+    },
+    {
+      "epoch": 0.6128803963198868,
+      "grad_norm": 0.3396860758261206,
+      "learning_rate": 4.4228751311647435e-05,
+      "loss": 0.4613,
+      "step": 433
+    },
+    {
+      "epoch": 0.6142958244869073,
+      "grad_norm": 0.32397361588158885,
+      "learning_rate": 4.420251836306401e-05,
+      "loss": 0.4392,
+      "step": 434
+    },
+    {
+      "epoch": 0.6157112526539278,
+      "grad_norm": 0.29303044508092985,
+      "learning_rate": 4.417628541448059e-05,
+      "loss": 0.4479,
+      "step": 435
+    },
+    {
+      "epoch": 0.6171266808209483,
+      "grad_norm": 0.2945470593351959,
+      "learning_rate": 4.415005246589717e-05,
+      "loss": 0.4577,
+      "step": 436
+    },
+    {
+      "epoch": 0.6185421089879689,
+      "grad_norm": 0.3092964622633318,
+      "learning_rate": 4.412381951731375e-05,
+      "loss": 0.4518,
+      "step": 437
+    },
+    {
+      "epoch": 0.6199575371549894,
+      "grad_norm": 0.28334203463423846,
+      "learning_rate": 4.409758656873032e-05,
+      "loss": 0.4657,
+      "step": 438
+    },
+    {
+      "epoch": 0.6213729653220099,
+      "grad_norm": 0.2877604209121691,
+      "learning_rate": 4.40713536201469e-05,
+      "loss": 0.4362,
+      "step": 439
+    },
+    {
+      "epoch": 0.6227883934890305,
+      "grad_norm": 0.2740454084168051,
+      "learning_rate": 4.404512067156348e-05,
+      "loss": 0.4485,
+      "step": 440
+    },
+    {
+      "epoch": 0.6242038216560509,
+      "grad_norm": 0.2877926081207299,
+      "learning_rate": 4.401888772298007e-05,
+      "loss": 0.4458,
+      "step": 441
+    },
+    {
+      "epoch": 0.6256192498230715,
+      "grad_norm": 0.31627814411582056,
+      "learning_rate": 4.3992654774396644e-05,
+      "loss": 0.4283,
+      "step": 442
+    },
+    {
+      "epoch": 0.627034677990092,
+      "grad_norm": 0.25505333075242553,
+      "learning_rate": 4.3966421825813224e-05,
+      "loss": 0.4389,
+      "step": 443
+    },
+    {
+      "epoch": 0.6284501061571125,
+      "grad_norm": 0.2735518374145452,
+      "learning_rate": 4.3940188877229804e-05,
+      "loss": 0.4331,
+      "step": 444
+    },
+    {
+      "epoch": 0.6298655343241331,
+      "grad_norm": 0.29238160109535305,
+      "learning_rate": 4.3913955928646385e-05,
+      "loss": 0.4472,
+      "step": 445
+    },
+    {
+      "epoch": 0.6312809624911536,
+      "grad_norm": 0.24821045828243918,
+      "learning_rate": 4.3887722980062965e-05,
+      "loss": 0.4608,
+      "step": 446
+    },
+    {
+      "epoch": 0.6326963906581741,
+      "grad_norm": 0.2516783612250606,
+      "learning_rate": 4.386149003147954e-05,
+      "loss": 0.4443,
+      "step": 447
+    },
+    {
+      "epoch": 0.6341118188251946,
+      "grad_norm": 0.2729001615032857,
+      "learning_rate": 4.383525708289612e-05,
+      "loss": 0.4439,
+      "step": 448
+    },
+    {
+      "epoch": 0.6355272469922152,
+      "grad_norm": 0.26562397446134683,
+      "learning_rate": 4.38090241343127e-05,
+      "loss": 0.4398,
+      "step": 449
+    },
+    {
+      "epoch": 0.6369426751592356,
+      "grad_norm": 0.27848528470733824,
+      "learning_rate": 4.378279118572928e-05,
+      "loss": 0.4725,
+      "step": 450
+    },
+    {
+      "epoch": 0.6383581033262562,
+      "grad_norm": 0.25076284762855006,
+      "learning_rate": 4.375655823714585e-05,
+      "loss": 0.4413,
+      "step": 451
+    },
+    {
+      "epoch": 0.6397735314932768,
+      "grad_norm": 0.2787634519406765,
+      "learning_rate": 4.373032528856243e-05,
+      "loss": 0.4446,
+      "step": 452
+    },
+    {
+      "epoch": 0.6411889596602972,
+      "grad_norm": 0.25957429540280813,
+      "learning_rate": 4.370409233997901e-05,
+      "loss": 0.4247,
+      "step": 453
+    },
+    {
+      "epoch": 0.6426043878273178,
+      "grad_norm": 0.29047238021603844,
+      "learning_rate": 4.3677859391395594e-05,
+      "loss": 0.4516,
+      "step": 454
+    },
+    {
+      "epoch": 0.6440198159943383,
+      "grad_norm": 0.27998705499552107,
+      "learning_rate": 4.3651626442812174e-05,
+      "loss": 0.4467,
+      "step": 455
+    },
+    {
+      "epoch": 0.6454352441613588,
+      "grad_norm": 0.29272174897988407,
+      "learning_rate": 4.3625393494228754e-05,
+      "loss": 0.4516,
+      "step": 456
+    },
+    {
+      "epoch": 0.6468506723283793,
+      "grad_norm": 0.27658532554543785,
+      "learning_rate": 4.3599160545645335e-05,
+      "loss": 0.4542,
+      "step": 457
+    },
+    {
+      "epoch": 0.6482661004953999,
+      "grad_norm": 0.2710425573124413,
+      "learning_rate": 4.3572927597061915e-05,
+      "loss": 0.4357,
+      "step": 458
+    },
+    {
+      "epoch": 0.6496815286624203,
+      "grad_norm": 0.2725022900196419,
+      "learning_rate": 4.3546694648478495e-05,
+      "loss": 0.4422,
+      "step": 459
+    },
+    {
+      "epoch": 0.6510969568294409,
+      "grad_norm": 0.2742452505894262,
+      "learning_rate": 4.352046169989507e-05,
+      "loss": 0.4539,
+      "step": 460
+    },
+    {
+      "epoch": 0.6525123849964615,
+      "grad_norm": 0.2587144103469113,
+      "learning_rate": 4.349422875131165e-05,
+      "loss": 0.4484,
+      "step": 461
+    },
+    {
+      "epoch": 0.6539278131634819,
+      "grad_norm": 0.22826321118851936,
+      "learning_rate": 4.346799580272823e-05,
+      "loss": 0.4575,
+      "step": 462
+    },
+    {
+      "epoch": 0.6553432413305025,
+      "grad_norm": 0.32262911700907015,
+      "learning_rate": 4.344176285414481e-05,
+      "loss": 0.462,
+      "step": 463
+    },
+    {
+      "epoch": 0.656758669497523,
+      "grad_norm": 0.26791948806923965,
+      "learning_rate": 4.341552990556138e-05,
+      "loss": 0.4463,
+      "step": 464
+    },
+    {
+      "epoch": 0.6581740976645435,
+      "grad_norm": 0.27271173640188784,
+      "learning_rate": 4.338929695697796e-05,
+      "loss": 0.4297,
+      "step": 465
+    },
+    {
+      "epoch": 0.659589525831564,
+      "grad_norm": 0.285366983143651,
+      "learning_rate": 4.3363064008394544e-05,
+      "loss": 0.4367,
+      "step": 466
+    },
+    {
+      "epoch": 0.6610049539985846,
+      "grad_norm": 0.2533623048370864,
+      "learning_rate": 4.3336831059811124e-05,
+      "loss": 0.4454,
+      "step": 467
+    },
+    {
+      "epoch": 0.6624203821656051,
+      "grad_norm": 0.26610790144962626,
+      "learning_rate": 4.3310598111227704e-05,
+      "loss": 0.4406,
+      "step": 468
+    },
+    {
+      "epoch": 0.6638358103326256,
+      "grad_norm": 0.27282829565311356,
+      "learning_rate": 4.3284365162644284e-05,
+      "loss": 0.425,
+      "step": 469
+    },
+    {
+      "epoch": 0.6652512384996462,
+      "grad_norm": 0.33805278799552196,
+      "learning_rate": 4.3258132214060865e-05,
+      "loss": 0.4392,
+      "step": 470
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.2924744394522499,
+      "learning_rate": 4.3231899265477445e-05,
+      "loss": 0.4459,
+      "step": 471
+    },
+    {
+      "epoch": 0.6680820948336872,
+      "grad_norm": 0.27584231458857017,
+      "learning_rate": 4.3205666316894025e-05,
+      "loss": 0.4394,
+      "step": 472
+    },
+    {
+      "epoch": 0.6694975230007078,
+      "grad_norm": 0.32991663219597556,
+      "learning_rate": 4.31794333683106e-05,
+      "loss": 0.4399,
+      "step": 473
+    },
+    {
+      "epoch": 0.6709129511677282,
+      "grad_norm": 0.3003211945829358,
+      "learning_rate": 4.315320041972718e-05,
+      "loss": 0.4436,
+      "step": 474
+    },
+    {
+      "epoch": 0.6723283793347488,
+      "grad_norm": 0.31229550572985726,
+      "learning_rate": 4.312696747114376e-05,
+      "loss": 0.438,
+      "step": 475
+    },
+    {
+      "epoch": 0.6737438075017693,
+      "grad_norm": 0.347527441930185,
+      "learning_rate": 4.310073452256034e-05,
+      "loss": 0.4784,
+      "step": 476
+    },
+    {
+      "epoch": 0.6751592356687898,
+      "grad_norm": 0.3532649194342591,
+      "learning_rate": 4.307450157397691e-05,
+      "loss": 0.462,
+      "step": 477
+    },
+    {
+      "epoch": 0.6765746638358103,
+      "grad_norm": 0.3440426012872478,
+      "learning_rate": 4.304826862539349e-05,
+      "loss": 0.4693,
+      "step": 478
+    },
+    {
+      "epoch": 0.6779900920028309,
+      "grad_norm": 0.2771881635662196,
+      "learning_rate": 4.3022035676810074e-05,
+      "loss": 0.4535,
+      "step": 479
+    },
+    {
+      "epoch": 0.6794055201698513,
+      "grad_norm": 0.33367051540770565,
+      "learning_rate": 4.2995802728226654e-05,
+      "loss": 0.4268,
+      "step": 480
+    },
+    {
+      "epoch": 0.6808209483368719,
+      "grad_norm": 0.3579079746745951,
+      "learning_rate": 4.2969569779643234e-05,
+      "loss": 0.4502,
+      "step": 481
+    },
+    {
+      "epoch": 0.6822363765038925,
+      "grad_norm": 0.26239819868050757,
+      "learning_rate": 4.294333683105981e-05,
+      "loss": 0.4232,
+      "step": 482
+    },
+    {
+      "epoch": 0.6836518046709129,
+      "grad_norm": 0.4129537053142466,
+      "learning_rate": 4.2917103882476395e-05,
+      "loss": 0.452,
+      "step": 483
+    },
+    {
+      "epoch": 0.6850672328379335,
+      "grad_norm": 0.26267539356220626,
+      "learning_rate": 4.2890870933892975e-05,
+      "loss": 0.4483,
+      "step": 484
+    },
+    {
+      "epoch": 0.686482661004954,
+      "grad_norm": 0.3353381200094175,
+      "learning_rate": 4.2864637985309555e-05,
+      "loss": 0.4377,
+      "step": 485
+    },
+    {
+      "epoch": 0.6878980891719745,
+      "grad_norm": 0.34902186848107175,
+      "learning_rate": 4.283840503672613e-05,
+      "loss": 0.4651,
+      "step": 486
+    },
+    {
+      "epoch": 0.689313517338995,
+      "grad_norm": 0.31170345792330595,
+      "learning_rate": 4.281217208814271e-05,
+      "loss": 0.4458,
+      "step": 487
+    },
+    {
+      "epoch": 0.6907289455060156,
+      "grad_norm": 0.3312254997595215,
+      "learning_rate": 4.278593913955929e-05,
+      "loss": 0.4551,
+      "step": 488
+    },
+    {
+      "epoch": 0.692144373673036,
+      "grad_norm": 0.25957333633861795,
+      "learning_rate": 4.275970619097587e-05,
+      "loss": 0.4422,
+      "step": 489
+    },
+    {
+      "epoch": 0.6935598018400566,
+      "grad_norm": 0.3555889031083105,
+      "learning_rate": 4.273347324239245e-05,
+      "loss": 0.4441,
+      "step": 490
+    },
+    {
+      "epoch": 0.6949752300070772,
+      "grad_norm": 0.2807205711336829,
+      "learning_rate": 4.2707240293809024e-05,
+      "loss": 0.4608,
+      "step": 491
+    },
+    {
+      "epoch": 0.6963906581740976,
+      "grad_norm": 0.26805299842014746,
+      "learning_rate": 4.2681007345225604e-05,
+      "loss": 0.459,
+      "step": 492
+    },
+    {
+      "epoch": 0.6978060863411182,
+      "grad_norm": 0.29511422434574813,
+      "learning_rate": 4.2654774396642184e-05,
+      "loss": 0.4341,
+      "step": 493
+    },
+    {
+      "epoch": 0.6992215145081387,
+      "grad_norm": 0.2908216275567921,
+      "learning_rate": 4.2628541448058764e-05,
+      "loss": 0.4426,
+      "step": 494
+    },
+    {
+      "epoch": 0.7006369426751592,
+      "grad_norm": 0.2815692706701559,
+      "learning_rate": 4.260230849947534e-05,
+      "loss": 0.4804,
+      "step": 495
+    },
+    {
+      "epoch": 0.7020523708421798,
+      "grad_norm": 0.2734488463505051,
+      "learning_rate": 4.257607555089192e-05,
+      "loss": 0.4264,
+      "step": 496
+    },
+    {
+      "epoch": 0.7034677990092003,
+      "grad_norm": 0.26114056612679676,
+      "learning_rate": 4.2549842602308505e-05,
+      "loss": 0.4479,
+      "step": 497
+    },
+    {
+      "epoch": 0.7048832271762208,
+      "grad_norm": 0.2985639437723813,
+      "learning_rate": 4.2523609653725086e-05,
+      "loss": 0.4454,
+      "step": 498
+    },
+    {
+      "epoch": 0.7062986553432413,
+      "grad_norm": 1.5976620033335418,
+      "learning_rate": 4.249737670514166e-05,
+      "loss": 0.4326,
+      "step": 499
+    },
+    {
+      "epoch": 0.7077140835102619,
+      "grad_norm": 0.32874439874375233,
+      "learning_rate": 4.247114375655824e-05,
+      "loss": 0.4279,
+      "step": 500
+    },
+    {
+      "epoch": 0.7091295116772823,
+      "grad_norm": 0.26165102344785035,
+      "learning_rate": 4.244491080797482e-05,
+      "loss": 0.43,
+      "step": 501
+    },
+    {
+      "epoch": 0.7105449398443029,
+      "grad_norm": 0.3553343056755559,
+      "learning_rate": 4.24186778593914e-05,
+      "loss": 0.4367,
+      "step": 502
+    },
+    {
+      "epoch": 0.7119603680113235,
+      "grad_norm": 0.2483556034336584,
+      "learning_rate": 4.239244491080798e-05,
+      "loss": 0.4404,
+      "step": 503
+    },
+    {
+      "epoch": 0.7133757961783439,
+      "grad_norm": 0.3278205650791823,
+      "learning_rate": 4.2366211962224554e-05,
+      "loss": 0.4611,
+      "step": 504
+    },
+    {
+      "epoch": 0.7147912243453645,
+      "grad_norm": 0.2690766337181193,
+      "learning_rate": 4.2339979013641134e-05,
+      "loss": 0.4498,
+      "step": 505
+    },
+    {
+      "epoch": 0.716206652512385,
+      "grad_norm": 0.28512054363268413,
+      "learning_rate": 4.2313746065057714e-05,
+      "loss": 0.4379,
+      "step": 506
+    },
+    {
+      "epoch": 0.7176220806794055,
+      "grad_norm": 0.3063504883919276,
+      "learning_rate": 4.2287513116474295e-05,
+      "loss": 0.4403,
+      "step": 507
+    },
+    {
+      "epoch": 0.719037508846426,
+      "grad_norm": 0.23758970561440687,
+      "learning_rate": 4.226128016789087e-05,
+      "loss": 0.4283,
+      "step": 508
+    },
+    {
+      "epoch": 0.7204529370134466,
+      "grad_norm": 0.27749982064530393,
+      "learning_rate": 4.223504721930745e-05,
+      "loss": 0.4632,
+      "step": 509
+    },
+    {
+      "epoch": 0.721868365180467,
+      "grad_norm": 0.2662941743049646,
+      "learning_rate": 4.2208814270724035e-05,
+      "loss": 0.4357,
+      "step": 510
+    },
+    {
+      "epoch": 0.7232837933474876,
+      "grad_norm": 0.2771224811633922,
+      "learning_rate": 4.2182581322140616e-05,
+      "loss": 0.4626,
+      "step": 511
+    },
+    {
+      "epoch": 0.7246992215145082,
+      "grad_norm": 0.25271868649341617,
+      "learning_rate": 4.215634837355719e-05,
+      "loss": 0.4311,
+      "step": 512
+    },
+    {
+      "epoch": 0.7261146496815286,
+      "grad_norm": 0.2889571898292604,
+      "learning_rate": 4.213011542497377e-05,
+      "loss": 0.4617,
+      "step": 513
+    },
+    {
+      "epoch": 0.7275300778485492,
+      "grad_norm": 0.26364411492956125,
+      "learning_rate": 4.210388247639035e-05,
+      "loss": 0.4415,
+      "step": 514
+    },
+    {
+      "epoch": 0.7289455060155697,
+      "grad_norm": 0.2931666870844134,
+      "learning_rate": 4.207764952780693e-05,
+      "loss": 0.4502,
+      "step": 515
+    },
+    {
+      "epoch": 0.7303609341825902,
+      "grad_norm": 0.258645009906452,
+      "learning_rate": 4.205141657922351e-05,
+      "loss": 0.4497,
+      "step": 516
+    },
+    {
+      "epoch": 0.7317763623496107,
+      "grad_norm": 0.24592005572202322,
+      "learning_rate": 4.2025183630640084e-05,
+      "loss": 0.4173,
+      "step": 517
+    },
+    {
+      "epoch": 0.7331917905166313,
+      "grad_norm": 0.2917455139254434,
+      "learning_rate": 4.1998950682056664e-05,
+      "loss": 0.4273,
+      "step": 518
+    },
+    {
+      "epoch": 0.7346072186836518,
+      "grad_norm": 0.26150067898128604,
+      "learning_rate": 4.1972717733473244e-05,
+      "loss": 0.4572,
+      "step": 519
+    },
+    {
+      "epoch": 0.7360226468506723,
+      "grad_norm": 0.3061571725238614,
+      "learning_rate": 4.1946484784889825e-05,
+      "loss": 0.4417,
+      "step": 520
+    },
+    {
+      "epoch": 0.7374380750176929,
+      "grad_norm": 0.2889083585606234,
+      "learning_rate": 4.19202518363064e-05,
+      "loss": 0.456,
+      "step": 521
+    },
+    {
+      "epoch": 0.7388535031847133,
+      "grad_norm": 0.24657878589038265,
+      "learning_rate": 4.189401888772298e-05,
+      "loss": 0.4351,
+      "step": 522
+    },
+    {
+      "epoch": 0.7402689313517339,
+      "grad_norm": 0.2961009417266799,
+      "learning_rate": 4.186778593913956e-05,
+      "loss": 0.4245,
+      "step": 523
+    },
+    {
+      "epoch": 0.7416843595187544,
+      "grad_norm": 0.29000447963640213,
+      "learning_rate": 4.1841552990556146e-05,
+      "loss": 0.4235,
+      "step": 524
+    },
+    {
+      "epoch": 0.7430997876857749,
+      "grad_norm": 0.24072534315865696,
+      "learning_rate": 4.181532004197272e-05,
+      "loss": 0.4555,
+      "step": 525
+    },
+    {
+      "epoch": 0.7445152158527955,
+      "grad_norm": 0.311490815899024,
+      "learning_rate": 4.17890870933893e-05,
+      "loss": 0.4392,
+      "step": 526
+    },
+    {
+      "epoch": 0.745930644019816,
+      "grad_norm": 0.25115626974902694,
+      "learning_rate": 4.176285414480588e-05,
+      "loss": 0.4452,
+      "step": 527
+    },
+    {
+      "epoch": 0.7473460721868365,
+      "grad_norm": 0.25670617969187803,
+      "learning_rate": 4.173662119622246e-05,
+      "loss": 0.4559,
+      "step": 528
+    },
+    {
+      "epoch": 0.748761500353857,
+      "grad_norm": 0.2562352740508127,
+      "learning_rate": 4.171038824763904e-05,
+      "loss": 0.4509,
+      "step": 529
+    },
+    {
+      "epoch": 0.7501769285208776,
+      "grad_norm": 0.26465908253857906,
+      "learning_rate": 4.1684155299055614e-05,
+      "loss": 0.4407,
+      "step": 530
+    },
+    {
+      "epoch": 0.7515923566878981,
+      "grad_norm": 0.2558921043589055,
+      "learning_rate": 4.1657922350472194e-05,
+      "loss": 0.4562,
+      "step": 531
+    },
+    {
+      "epoch": 0.7530077848549186,
+      "grad_norm": 0.2556217189091007,
+      "learning_rate": 4.1631689401888775e-05,
+      "loss": 0.4524,
+      "step": 532
+    },
+    {
+      "epoch": 0.7544232130219392,
+      "grad_norm": 0.2768146188546652,
+      "learning_rate": 4.1605456453305355e-05,
+      "loss": 0.433,
+      "step": 533
+    },
+    {
+      "epoch": 0.7558386411889597,
+      "grad_norm": 0.25924426442650855,
+      "learning_rate": 4.157922350472193e-05,
+      "loss": 0.4288,
+      "step": 534
+    },
+    {
+      "epoch": 0.7572540693559802,
+      "grad_norm": 0.3028465063884449,
+      "learning_rate": 4.155299055613851e-05,
+      "loss": 0.4655,
+      "step": 535
+    },
+    {
+      "epoch": 0.7586694975230007,
+      "grad_norm": 0.28990196149589315,
+      "learning_rate": 4.152675760755509e-05,
+      "loss": 0.4438,
+      "step": 536
+    },
+    {
+      "epoch": 0.7600849256900213,
+      "grad_norm": 0.29965783174238914,
+      "learning_rate": 4.150052465897167e-05,
+      "loss": 0.4321,
+      "step": 537
+    },
+    {
+      "epoch": 0.7615003538570417,
+      "grad_norm": 0.2704665797850079,
+      "learning_rate": 4.147429171038825e-05,
+      "loss": 0.442,
+      "step": 538
+    },
+    {
+      "epoch": 0.7629157820240623,
+      "grad_norm": 0.2846292282729222,
+      "learning_rate": 4.144805876180483e-05,
+      "loss": 0.4458,
+      "step": 539
+    },
+    {
+      "epoch": 0.7643312101910829,
+      "grad_norm": 0.3351484846406856,
+      "learning_rate": 4.142182581322141e-05,
+      "loss": 0.4483,
+      "step": 540
+    },
+    {
+      "epoch": 0.7657466383581033,
+      "grad_norm": 0.23511671534196746,
+      "learning_rate": 4.139559286463799e-05,
+      "loss": 0.4172,
+      "step": 541
+    },
+    {
+      "epoch": 0.7671620665251239,
+      "grad_norm": 0.29165081709888513,
+      "learning_rate": 4.136935991605457e-05,
+      "loss": 0.4479,
+      "step": 542
+    },
+    {
+      "epoch": 0.7685774946921444,
+      "grad_norm": 0.26675297246825325,
+      "learning_rate": 4.1343126967471144e-05,
+      "loss": 0.4436,
+      "step": 543
+    },
+    {
+      "epoch": 0.7699929228591649,
+      "grad_norm": 0.2543439669312433,
+      "learning_rate": 4.1316894018887724e-05,
+      "loss": 0.4325,
+      "step": 544
+    },
+    {
+      "epoch": 0.7714083510261854,
+      "grad_norm": 0.2647881854880466,
+      "learning_rate": 4.1290661070304305e-05,
+      "loss": 0.4334,
+      "step": 545
+    },
+    {
+      "epoch": 0.772823779193206,
+      "grad_norm": 0.2659462907807334,
+      "learning_rate": 4.1264428121720885e-05,
+      "loss": 0.4373,
+      "step": 546
+    },
+    {
+      "epoch": 0.7742392073602264,
+      "grad_norm": 0.31454427433265036,
+      "learning_rate": 4.123819517313746e-05,
+      "loss": 0.4356,
+      "step": 547
+    },
+    {
+      "epoch": 0.775654635527247,
+      "grad_norm": 0.23037668235704842,
+      "learning_rate": 4.121196222455404e-05,
+      "loss": 0.4223,
+      "step": 548
+    },
+    {
+      "epoch": 0.7770700636942676,
+      "grad_norm": 0.3038793619159252,
+      "learning_rate": 4.118572927597062e-05,
+      "loss": 0.4354,
+      "step": 549
+    },
+    {
+      "epoch": 0.778485491861288,
+      "grad_norm": 0.2503175095869354,
+      "learning_rate": 4.11594963273872e-05,
+      "loss": 0.4248,
+      "step": 550
+    },
+    {
+      "epoch": 0.7799009200283086,
+      "grad_norm": 0.23469319529918808,
+      "learning_rate": 4.113326337880378e-05,
+      "loss": 0.4359,
+      "step": 551
+    },
+    {
+      "epoch": 0.7813163481953291,
+      "grad_norm": 0.3377368475859871,
+      "learning_rate": 4.110703043022036e-05,
+      "loss": 0.4602,
+      "step": 552
+    },
+    {
+      "epoch": 0.7827317763623496,
+      "grad_norm": 0.26622970488718106,
+      "learning_rate": 4.108079748163694e-05,
+      "loss": 0.445,
+      "step": 553
+    },
+    {
+      "epoch": 0.7841472045293701,
+      "grad_norm": 0.284056764258355,
+      "learning_rate": 4.105456453305352e-05,
+      "loss": 0.4293,
+      "step": 554
+    },
+    {
+      "epoch": 0.7855626326963907,
+      "grad_norm": 0.2768862589880146,
+      "learning_rate": 4.10283315844701e-05,
+      "loss": 0.4501,
+      "step": 555
+    },
+    {
+      "epoch": 0.7869780608634112,
+      "grad_norm": 0.24905867903454612,
+      "learning_rate": 4.1002098635886674e-05,
+      "loss": 0.4329,
+      "step": 556
+    },
+    {
+      "epoch": 0.7883934890304317,
+      "grad_norm": 0.24635328000816292,
+      "learning_rate": 4.0975865687303255e-05,
+      "loss": 0.4364,
+      "step": 557
+    },
+    {
+      "epoch": 0.7898089171974523,
+      "grad_norm": 0.2746854679948567,
+      "learning_rate": 4.0949632738719835e-05,
+      "loss": 0.4619,
+      "step": 558
+    },
+    {
+      "epoch": 0.7912243453644727,
+      "grad_norm": 0.23544898533832812,
+      "learning_rate": 4.0923399790136415e-05,
+      "loss": 0.4235,
+      "step": 559
+    },
+    {
+      "epoch": 0.7926397735314933,
+      "grad_norm": 0.2893309340044802,
+      "learning_rate": 4.089716684155299e-05,
+      "loss": 0.4686,
+      "step": 560
+    },
+    {
+      "epoch": 0.7940552016985138,
+      "grad_norm": 0.25713408416266553,
+      "learning_rate": 4.087093389296957e-05,
+      "loss": 0.4613,
+      "step": 561
+    },
+    {
+      "epoch": 0.7954706298655343,
+      "grad_norm": 0.283343162310229,
+      "learning_rate": 4.084470094438615e-05,
+      "loss": 0.4652,
+      "step": 562
+    },
+    {
+      "epoch": 0.7968860580325549,
+      "grad_norm": 0.24124503356161994,
+      "learning_rate": 4.081846799580273e-05,
+      "loss": 0.4403,
+      "step": 563
+    },
+    {
+      "epoch": 0.7983014861995754,
+      "grad_norm": 0.2684086698707159,
+      "learning_rate": 4.079223504721931e-05,
+      "loss": 0.4398,
+      "step": 564
+    },
+    {
+      "epoch": 0.7997169143665959,
+      "grad_norm": 0.2467316807541914,
+      "learning_rate": 4.076600209863588e-05,
+      "loss": 0.4345,
+      "step": 565
+    },
+    {
+      "epoch": 0.8011323425336164,
+      "grad_norm": 0.24452592123968678,
+      "learning_rate": 4.073976915005247e-05,
+      "loss": 0.4254,
+      "step": 566
+    },
+    {
+      "epoch": 0.802547770700637,
+      "grad_norm": 0.2487243164721359,
+      "learning_rate": 4.071353620146905e-05,
+      "loss": 0.4287,
+      "step": 567
+    },
+    {
+      "epoch": 0.8039631988676574,
+      "grad_norm": 0.24641304492332408,
+      "learning_rate": 4.068730325288563e-05,
+      "loss": 0.4407,
+      "step": 568
+    },
+    {
+      "epoch": 0.805378627034678,
+      "grad_norm": 0.25730898378645617,
+      "learning_rate": 4.0661070304302204e-05,
+      "loss": 0.4455,
+      "step": 569
+    },
+    {
+      "epoch": 0.8067940552016986,
+      "grad_norm": 0.26003093220287893,
+      "learning_rate": 4.0634837355718785e-05,
+      "loss": 0.4304,
+      "step": 570
+    },
+    {
+      "epoch": 0.808209483368719,
+      "grad_norm": 0.2589801413670949,
+      "learning_rate": 4.0608604407135365e-05,
+      "loss": 0.4289,
+      "step": 571
+    },
+    {
+      "epoch": 0.8096249115357396,
+      "grad_norm": 0.2314003574658966,
+      "learning_rate": 4.0582371458551945e-05,
+      "loss": 0.4323,
+      "step": 572
+    },
+    {
+      "epoch": 0.8110403397027601,
+      "grad_norm": 0.25102135230481415,
+      "learning_rate": 4.055613850996852e-05,
+      "loss": 0.4302,
+      "step": 573
+    },
+    {
+      "epoch": 0.8124557678697806,
+      "grad_norm": 0.23891828131337306,
+      "learning_rate": 4.05299055613851e-05,
+      "loss": 0.4266,
+      "step": 574
+    },
+    {
+      "epoch": 0.8138711960368011,
+      "grad_norm": 0.2515143782722983,
+      "learning_rate": 4.050367261280168e-05,
+      "loss": 0.4401,
+      "step": 575
+    },
+    {
+      "epoch": 0.8152866242038217,
+      "grad_norm": 0.30978202190326437,
+      "learning_rate": 4.047743966421826e-05,
+      "loss": 0.4196,
+      "step": 576
+    },
+    {
+      "epoch": 0.8167020523708421,
+      "grad_norm": 0.2596078416967971,
+      "learning_rate": 4.045120671563484e-05,
+      "loss": 0.4193,
+      "step": 577
+    },
+    {
+      "epoch": 0.8181174805378627,
+      "grad_norm": 0.2648377726503376,
+      "learning_rate": 4.0424973767051413e-05,
+      "loss": 0.4334,
+      "step": 578
+    },
+    {
+      "epoch": 0.8195329087048833,
+      "grad_norm": 0.3033956340660034,
+      "learning_rate": 4.0398740818467994e-05,
+      "loss": 0.4491,
+      "step": 579
+    },
+    {
+      "epoch": 0.8209483368719037,
+      "grad_norm": 0.23790213249393122,
+      "learning_rate": 4.037250786988458e-05,
+      "loss": 0.4156,
+      "step": 580
+    },
+    {
+      "epoch": 0.8223637650389243,
+      "grad_norm": 0.2704426474419847,
+      "learning_rate": 4.034627492130116e-05,
+      "loss": 0.4409,
+      "step": 581
+    },
+    {
+      "epoch": 0.8237791932059448,
+      "grad_norm": 0.2525590290369702,
+      "learning_rate": 4.0320041972717735e-05,
+      "loss": 0.4552,
+      "step": 582
+    },
+    {
+      "epoch": 0.8251946213729653,
+      "grad_norm": 0.2820030156719851,
+      "learning_rate": 4.0293809024134315e-05,
+      "loss": 0.4334,
+      "step": 583
+    },
+    {
+      "epoch": 0.8266100495399858,
+      "grad_norm": 0.27633787683248595,
+      "learning_rate": 4.0267576075550895e-05,
+      "loss": 0.4287,
+      "step": 584
+    },
+    {
+      "epoch": 0.8280254777070064,
+      "grad_norm": 0.2591953413517348,
+      "learning_rate": 4.0241343126967475e-05,
+      "loss": 0.4416,
+      "step": 585
+    },
+    {
+      "epoch": 0.8294409058740269,
+      "grad_norm": 0.2575631752650505,
+      "learning_rate": 4.021511017838405e-05,
+      "loss": 0.4427,
+      "step": 586
+    },
+    {
+      "epoch": 0.8308563340410474,
+      "grad_norm": 0.233283484588631,
+      "learning_rate": 4.018887722980063e-05,
+      "loss": 0.4354,
+      "step": 587
+    },
+    {
+      "epoch": 0.832271762208068,
+      "grad_norm": 0.25412762816285034,
+      "learning_rate": 4.016264428121721e-05,
+      "loss": 0.4349,
+      "step": 588
+    },
+    {
+      "epoch": 0.8336871903750884,
+      "grad_norm": 0.25097603215877484,
+      "learning_rate": 4.013641133263379e-05,
+      "loss": 0.4576,
+      "step": 589
+    },
+    {
+      "epoch": 0.835102618542109,
+      "grad_norm": 0.2668490361059998,
+      "learning_rate": 4.011017838405037e-05,
+      "loss": 0.4166,
+      "step": 590
+    },
+    {
+      "epoch": 0.8365180467091295,
+      "grad_norm": 0.2615497111859223,
+      "learning_rate": 4.0083945435466944e-05,
+      "loss": 0.4461,
+      "step": 591
+    },
+    {
+      "epoch": 0.83793347487615,
+      "grad_norm": 0.2626945943986861,
+      "learning_rate": 4.0057712486883524e-05,
+      "loss": 0.4296,
+      "step": 592
+    },
+    {
+      "epoch": 0.8393489030431706,
+      "grad_norm": 0.24275266812210086,
+      "learning_rate": 4.003147953830011e-05,
+      "loss": 0.4191,
+      "step": 593
+    },
+    {
+      "epoch": 0.8407643312101911,
+      "grad_norm": 0.2580872822583186,
+      "learning_rate": 4.000524658971669e-05,
+      "loss": 0.4176,
+      "step": 594
+    },
+    {
+      "epoch": 0.8421797593772116,
+      "grad_norm": 0.21605702070130375,
+      "learning_rate": 3.9979013641133265e-05,
+      "loss": 0.4379,
+      "step": 595
+    },
+    {
+      "epoch": 0.8435951875442321,
+      "grad_norm": 0.2451930026518869,
+      "learning_rate": 3.9952780692549845e-05,
+      "loss": 0.4306,
+      "step": 596
+    },
+    {
+      "epoch": 0.8450106157112527,
+      "grad_norm": 0.22624512326095705,
+      "learning_rate": 3.9926547743966425e-05,
+      "loss": 0.4167,
+      "step": 597
+    },
+    {
+      "epoch": 0.8464260438782731,
+      "grad_norm": 0.26298017826623193,
+      "learning_rate": 3.9900314795383006e-05,
+      "loss": 0.4421,
+      "step": 598
+    },
+    {
+      "epoch": 0.8478414720452937,
+      "grad_norm": 0.2565705411358314,
+      "learning_rate": 3.987408184679958e-05,
+      "loss": 0.4543,
+      "step": 599
+    },
+    {
+      "epoch": 0.8492569002123143,
+      "grad_norm": 0.2836966769986463,
+      "learning_rate": 3.984784889821616e-05,
+      "loss": 0.449,
+      "step": 600
+    },
+    {
+      "epoch": 0.8506723283793347,
+      "grad_norm": 0.33121523517986445,
+      "learning_rate": 3.982161594963274e-05,
+      "loss": 0.4524,
+      "step": 601
+    },
+    {
+      "epoch": 0.8520877565463553,
+      "grad_norm": 0.2728582746620306,
+      "learning_rate": 3.979538300104932e-05,
+      "loss": 0.4296,
+      "step": 602
+    },
+    {
+      "epoch": 0.8535031847133758,
+      "grad_norm": 0.24504881169831788,
+      "learning_rate": 3.97691500524659e-05,
+      "loss": 0.4308,
+      "step": 603
+    },
+    {
+      "epoch": 0.8549186128803963,
+      "grad_norm": 0.2556664330044345,
+      "learning_rate": 3.9742917103882474e-05,
+      "loss": 0.4419,
+      "step": 604
+    },
+    {
+      "epoch": 0.8563340410474168,
+      "grad_norm": 0.249942745144725,
+      "learning_rate": 3.9716684155299054e-05,
+      "loss": 0.4316,
+      "step": 605
+    },
+    {
+      "epoch": 0.8577494692144374,
+      "grad_norm": 0.25277013855046737,
+      "learning_rate": 3.9690451206715634e-05,
+      "loss": 0.4177,
+      "step": 606
+    },
+    {
+      "epoch": 0.8591648973814578,
+      "grad_norm": 0.27490926708863284,
+      "learning_rate": 3.966421825813222e-05,
+      "loss": 0.4415,
+      "step": 607
+    },
+    {
+      "epoch": 0.8605803255484784,
+      "grad_norm": 0.25363680318434917,
+      "learning_rate": 3.9637985309548795e-05,
+      "loss": 0.4435,
+      "step": 608
+    },
+    {
+      "epoch": 0.861995753715499,
+      "grad_norm": 0.24172325545678805,
+      "learning_rate": 3.9611752360965375e-05,
+      "loss": 0.4441,
+      "step": 609
+    },
+    {
+      "epoch": 0.8634111818825194,
+      "grad_norm": 0.24821946310064713,
+      "learning_rate": 3.9585519412381956e-05,
+      "loss": 0.4262,
+      "step": 610
+    },
+    {
+      "epoch": 0.86482661004954,
+      "grad_norm": 0.2537246262286817,
+      "learning_rate": 3.9559286463798536e-05,
+      "loss": 0.4523,
+      "step": 611
+    },
+    {
+      "epoch": 0.8662420382165605,
+      "grad_norm": 0.25925970718854424,
+      "learning_rate": 3.953305351521511e-05,
+      "loss": 0.4515,
+      "step": 612
+    },
+    {
+      "epoch": 0.867657466383581,
+      "grad_norm": 0.23455080583476018,
+      "learning_rate": 3.950682056663169e-05,
+      "loss": 0.4141,
+      "step": 613
+    },
+    {
+      "epoch": 0.8690728945506015,
+      "grad_norm": 0.265557541743459,
+      "learning_rate": 3.948058761804827e-05,
+      "loss": 0.4524,
+      "step": 614
+    },
+    {
+      "epoch": 0.8704883227176221,
+      "grad_norm": 0.23864799283771135,
+      "learning_rate": 3.945435466946485e-05,
+      "loss": 0.4288,
+      "step": 615
+    },
+    {
+      "epoch": 0.8719037508846426,
+      "grad_norm": 0.2637664502630157,
+      "learning_rate": 3.942812172088143e-05,
+      "loss": 0.4677,
+      "step": 616
+    },
+    {
+      "epoch": 0.8733191790516631,
+      "grad_norm": 0.22574671608165228,
+      "learning_rate": 3.9401888772298004e-05,
+      "loss": 0.4587,
+      "step": 617
+    },
+    {
+      "epoch": 0.8747346072186837,
+      "grad_norm": 0.26419008526198395,
+      "learning_rate": 3.9375655823714584e-05,
+      "loss": 0.4483,
+      "step": 618
+    },
+    {
+      "epoch": 0.8761500353857041,
+      "grad_norm": 0.24585359511846885,
+      "learning_rate": 3.9349422875131165e-05,
+      "loss": 0.4284,
+      "step": 619
+    },
+    {
+      "epoch": 0.8775654635527247,
+      "grad_norm": 0.23905515655418283,
+      "learning_rate": 3.9323189926547745e-05,
+      "loss": 0.4312,
+      "step": 620
+    },
+    {
+      "epoch": 0.8789808917197452,
+      "grad_norm": 0.2728477728005644,
+      "learning_rate": 3.9296956977964325e-05,
+      "loss": 0.4227,
+      "step": 621
+    },
+    {
+      "epoch": 0.8803963198867657,
+      "grad_norm": 0.23520624771873033,
+      "learning_rate": 3.9270724029380905e-05,
+      "loss": 0.4326,
+      "step": 622
+    },
+    {
+      "epoch": 0.8818117480537863,
+      "grad_norm": 0.26074161845119215,
+      "learning_rate": 3.9244491080797486e-05,
+      "loss": 0.4234,
+      "step": 623
+    },
+    {
+      "epoch": 0.8832271762208068,
+      "grad_norm": 0.28690755639338716,
+      "learning_rate": 3.9218258132214066e-05,
+      "loss": 0.4135,
+      "step": 624
+    },
+    {
+      "epoch": 0.8846426043878273,
+      "grad_norm": 0.2787858417232892,
+      "learning_rate": 3.9192025183630646e-05,
+      "loss": 0.4399,
+      "step": 625
+    },
+    {
+      "epoch": 0.8860580325548478,
+      "grad_norm": 0.2650050206775565,
+      "learning_rate": 3.916579223504722e-05,
+      "loss": 0.4313,
+      "step": 626
+    },
+    {
+      "epoch": 0.8874734607218684,
+      "grad_norm": 0.28073341233108806,
+      "learning_rate": 3.91395592864638e-05,
+      "loss": 0.4358,
+      "step": 627
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 0.2903436499787625,
+      "learning_rate": 3.911332633788038e-05,
+      "loss": 0.4647,
+      "step": 628
+    },
+    {
+      "epoch": 0.8903043170559094,
+      "grad_norm": 0.25408490569862163,
+      "learning_rate": 3.908709338929696e-05,
+      "loss": 0.4158,
+      "step": 629
+    },
+    {
+      "epoch": 0.89171974522293,
+      "grad_norm": 0.2954359201711501,
+      "learning_rate": 3.9060860440713534e-05,
+      "loss": 0.4211,
+      "step": 630
+    },
+    {
+      "epoch": 0.8931351733899504,
+      "grad_norm": 0.2560466892940268,
+      "learning_rate": 3.9034627492130114e-05,
+      "loss": 0.4352,
+      "step": 631
+    },
+    {
+      "epoch": 0.894550601556971,
+      "grad_norm": 0.2695807117497625,
+      "learning_rate": 3.9008394543546695e-05,
+      "loss": 0.4399,
+      "step": 632
+    },
+    {
+      "epoch": 0.8959660297239915,
+      "grad_norm": 0.25329687351083074,
+      "learning_rate": 3.8982161594963275e-05,
+      "loss": 0.4069,
+      "step": 633
+    },
+    {
+      "epoch": 0.897381457891012,
+      "grad_norm": 0.2636093344080135,
+      "learning_rate": 3.8955928646379855e-05,
+      "loss": 0.4561,
+      "step": 634
+    },
+    {
+      "epoch": 0.8987968860580325,
+      "grad_norm": 0.2827350978156938,
+      "learning_rate": 3.8929695697796436e-05,
+      "loss": 0.4432,
+      "step": 635
+    },
+    {
+      "epoch": 0.9002123142250531,
+      "grad_norm": 0.22168217202680376,
+      "learning_rate": 3.8903462749213016e-05,
+      "loss": 0.4257,
+      "step": 636
+    },
+    {
+      "epoch": 0.9016277423920736,
+      "grad_norm": 0.2738450503933235,
+      "learning_rate": 3.8877229800629596e-05,
+      "loss": 0.4265,
+      "step": 637
+    },
+    {
+      "epoch": 0.9030431705590941,
+      "grad_norm": 0.24444045291322025,
+      "learning_rate": 3.8850996852046176e-05,
+      "loss": 0.4088,
+      "step": 638
+    },
+    {
+      "epoch": 0.9044585987261147,
+      "grad_norm": 0.26343060033893795,
+      "learning_rate": 3.882476390346275e-05,
+      "loss": 0.4331,
+      "step": 639
+    },
+    {
+      "epoch": 0.9058740268931351,
+      "grad_norm": 0.23493361383383915,
+      "learning_rate": 3.879853095487933e-05,
+      "loss": 0.4291,
+      "step": 640
+    },
+    {
+      "epoch": 0.9072894550601557,
+      "grad_norm": 0.22921948270380707,
+      "learning_rate": 3.877229800629591e-05,
+      "loss": 0.424,
+      "step": 641
+    },
+    {
+      "epoch": 0.9087048832271762,
+      "grad_norm": 0.23124241606530554,
+      "learning_rate": 3.874606505771249e-05,
+      "loss": 0.4268,
+      "step": 642
+    },
+    {
+      "epoch": 0.9101203113941967,
+      "grad_norm": 0.22397166748586314,
+      "learning_rate": 3.8719832109129064e-05,
+      "loss": 0.4356,
+      "step": 643
+    },
+    {
+      "epoch": 0.9115357395612173,
+      "grad_norm": 0.23526999597742354,
+      "learning_rate": 3.8693599160545645e-05,
+      "loss": 0.4405,
+      "step": 644
+    },
+    {
+      "epoch": 0.9129511677282378,
+      "grad_norm": 0.23143080534593588,
+      "learning_rate": 3.8667366211962225e-05,
+      "loss": 0.428,
+      "step": 645
+    },
+    {
+      "epoch": 0.9143665958952583,
+      "grad_norm": 0.2398969695399371,
+      "learning_rate": 3.8641133263378805e-05,
+      "loss": 0.413,
+      "step": 646
+    },
+    {
+      "epoch": 0.9157820240622788,
+      "grad_norm": 0.265000208973342,
+      "learning_rate": 3.8614900314795385e-05,
+      "loss": 0.4453,
+      "step": 647
+    },
+    {
+      "epoch": 0.9171974522292994,
+      "grad_norm": 0.2314365688408304,
+      "learning_rate": 3.858866736621196e-05,
+      "loss": 0.4324,
+      "step": 648
+    },
+    {
+      "epoch": 0.9186128803963199,
+      "grad_norm": 0.25430068981454346,
+      "learning_rate": 3.8562434417628546e-05,
+      "loss": 0.4375,
+      "step": 649
+    },
+    {
+      "epoch": 0.9200283085633404,
+      "grad_norm": 0.26196086355330445,
+      "learning_rate": 3.8536201469045126e-05,
+      "loss": 0.4449,
+      "step": 650
+    },
+    {
+      "epoch": 0.921443736730361,
+      "grad_norm": 0.24954221725629927,
+      "learning_rate": 3.8509968520461707e-05,
+      "loss": 0.4245,
+      "step": 651
+    },
+    {
+      "epoch": 0.9228591648973815,
+      "grad_norm": 0.27419111732405277,
+      "learning_rate": 3.848373557187828e-05,
+      "loss": 0.4369,
+      "step": 652
+    },
+    {
+      "epoch": 0.924274593064402,
+      "grad_norm": 0.22989973319759824,
+      "learning_rate": 3.845750262329486e-05,
+      "loss": 0.4089,
+      "step": 653
+    },
+    {
+      "epoch": 0.9256900212314225,
+      "grad_norm": 0.27756466437594846,
+      "learning_rate": 3.843126967471144e-05,
+      "loss": 0.4175,
+      "step": 654
+    },
+    {
+      "epoch": 0.9271054493984431,
+      "grad_norm": 0.23523394186249474,
+      "learning_rate": 3.840503672612802e-05,
+      "loss": 0.4359,
+      "step": 655
+    },
+    {
+      "epoch": 0.9285208775654635,
+      "grad_norm": 0.23806322529924434,
+      "learning_rate": 3.8378803777544594e-05,
+      "loss": 0.4304,
+      "step": 656
+    },
+    {
+      "epoch": 0.9299363057324841,
+      "grad_norm": 0.2251930698576581,
+      "learning_rate": 3.8352570828961175e-05,
+      "loss": 0.4351,
+      "step": 657
+    },
+    {
+      "epoch": 0.9313517338995047,
+      "grad_norm": 0.2336333323067211,
+      "learning_rate": 3.8326337880377755e-05,
+      "loss": 0.4159,
+      "step": 658
+    },
+    {
+      "epoch": 0.9327671620665251,
+      "grad_norm": 0.24200797436131163,
+      "learning_rate": 3.8300104931794335e-05,
+      "loss": 0.4309,
+      "step": 659
+    },
+    {
+      "epoch": 0.9341825902335457,
+      "grad_norm": 0.23043702492712964,
+      "learning_rate": 3.8273871983210916e-05,
+      "loss": 0.4388,
+      "step": 660
+    },
+    {
+      "epoch": 0.9355980184005662,
+      "grad_norm": 0.24789726741995202,
+      "learning_rate": 3.824763903462749e-05,
+      "loss": 0.4313,
+      "step": 661
+    },
+    {
+      "epoch": 0.9370134465675867,
+      "grad_norm": 0.22798645877919682,
+      "learning_rate": 3.8221406086044076e-05,
+      "loss": 0.4258,
+      "step": 662
+    },
+    {
+      "epoch": 0.9384288747346072,
+      "grad_norm": 0.27442869920495,
+      "learning_rate": 3.8195173137460656e-05,
+      "loss": 0.4223,
+      "step": 663
+    },
+    {
+      "epoch": 0.9398443029016278,
+      "grad_norm": 0.22800353019756675,
+      "learning_rate": 3.816894018887724e-05,
+      "loss": 0.444,
+      "step": 664
+    },
+    {
+      "epoch": 0.9412597310686482,
+      "grad_norm": 0.2962156144416487,
+      "learning_rate": 3.814270724029381e-05,
+      "loss": 0.4267,
+      "step": 665
+    },
+    {
+      "epoch": 0.9426751592356688,
+      "grad_norm": 0.23959166750933444,
+      "learning_rate": 3.811647429171039e-05,
+      "loss": 0.4372,
+      "step": 666
+    },
+    {
+      "epoch": 0.9440905874026894,
+      "grad_norm": 0.2455901443670002,
+      "learning_rate": 3.809024134312697e-05,
+      "loss": 0.4252,
+      "step": 667
+    },
+    {
+      "epoch": 0.9455060155697098,
+      "grad_norm": 0.3204269037975826,
+      "learning_rate": 3.806400839454355e-05,
+      "loss": 0.4338,
+      "step": 668
+    },
+    {
+      "epoch": 0.9469214437367304,
+      "grad_norm": 0.23500461276640994,
+      "learning_rate": 3.8037775445960125e-05,
+      "loss": 0.429,
+      "step": 669
+    },
+    {
+      "epoch": 0.9483368719037509,
+      "grad_norm": 0.27072050273435716,
+      "learning_rate": 3.8011542497376705e-05,
+      "loss": 0.4602,
+      "step": 670
+    },
+    {
+      "epoch": 0.9497523000707714,
+      "grad_norm": 0.30323135734042067,
+      "learning_rate": 3.7985309548793285e-05,
+      "loss": 0.4425,
+      "step": 671
+    },
+    {
+      "epoch": 0.9511677282377919,
+      "grad_norm": 0.2913058978530256,
+      "learning_rate": 3.7959076600209865e-05,
+      "loss": 0.4261,
+      "step": 672
+    },
+    {
+      "epoch": 0.9525831564048125,
+      "grad_norm": 0.30067495620914264,
+      "learning_rate": 3.7932843651626446e-05,
+      "loss": 0.4365,
+      "step": 673
+    },
+    {
+      "epoch": 0.953998584571833,
+      "grad_norm": 0.3342344094135603,
+      "learning_rate": 3.790661070304302e-05,
+      "loss": 0.409,
+      "step": 674
+    },
+    {
+      "epoch": 0.9554140127388535,
+      "grad_norm": 0.2245576022075877,
+      "learning_rate": 3.78803777544596e-05,
+      "loss": 0.4269,
+      "step": 675
+    },
+    {
+      "epoch": 0.9568294409058741,
+      "grad_norm": 0.30343794677876584,
+      "learning_rate": 3.7854144805876187e-05,
+      "loss": 0.4483,
+      "step": 676
+    },
+    {
+      "epoch": 0.9582448690728945,
+      "grad_norm": 0.3063562804828594,
+      "learning_rate": 3.782791185729277e-05,
+      "loss": 0.4299,
+      "step": 677
+    },
+    {
+      "epoch": 0.9596602972399151,
+      "grad_norm": 0.26645005492549423,
+      "learning_rate": 3.780167890870934e-05,
+      "loss": 0.4405,
+      "step": 678
+    },
+    {
+      "epoch": 0.9610757254069356,
+      "grad_norm": 0.3061269234294929,
+      "learning_rate": 3.777544596012592e-05,
+      "loss": 0.431,
+      "step": 679
+    },
+    {
+      "epoch": 0.9624911535739561,
+      "grad_norm": 0.28465396623067296,
+      "learning_rate": 3.77492130115425e-05,
+      "loss": 0.4476,
+      "step": 680
+    },
+    {
+      "epoch": 0.9639065817409767,
+      "grad_norm": 0.27440327110271684,
+      "learning_rate": 3.772298006295908e-05,
+      "loss": 0.4332,
+      "step": 681
+    },
+    {
+      "epoch": 0.9653220099079972,
+      "grad_norm": 0.3203946284360972,
+      "learning_rate": 3.7696747114375655e-05,
+      "loss": 0.4054,
+      "step": 682
+    },
+    {
+      "epoch": 0.9667374380750177,
+      "grad_norm": 0.2757673656158921,
+      "learning_rate": 3.7670514165792235e-05,
+      "loss": 0.4128,
+      "step": 683
+    },
+    {
+      "epoch": 0.9681528662420382,
+      "grad_norm": 0.23866159456452424,
+      "learning_rate": 3.7644281217208815e-05,
+      "loss": 0.4383,
+      "step": 684
+    },
+    {
+      "epoch": 0.9695682944090588,
+      "grad_norm": 0.3292539337191727,
+      "learning_rate": 3.7618048268625396e-05,
+      "loss": 0.4325,
+      "step": 685
+    },
+    {
+      "epoch": 0.9709837225760792,
+      "grad_norm": 0.26397915273129924,
+      "learning_rate": 3.7591815320041976e-05,
+      "loss": 0.4352,
+      "step": 686
+    },
+    {
+      "epoch": 0.9723991507430998,
+      "grad_norm": 0.30004166885793,
+      "learning_rate": 3.756558237145855e-05,
+      "loss": 0.4432,
+      "step": 687
+    },
+    {
+      "epoch": 0.9738145789101204,
+      "grad_norm": 0.33248547118937477,
+      "learning_rate": 3.753934942287513e-05,
+      "loss": 0.4097,
+      "step": 688
+    },
+    {
+      "epoch": 0.9752300070771408,
+      "grad_norm": 0.27787804232537733,
+      "learning_rate": 3.751311647429171e-05,
+      "loss": 0.4366,
+      "step": 689
+    },
+    {
+      "epoch": 0.9766454352441614,
+      "grad_norm": 0.2789774810347891,
+      "learning_rate": 3.74868835257083e-05,
+      "loss": 0.4287,
+      "step": 690
+    },
+    {
+      "epoch": 0.9780608634111819,
+      "grad_norm": 0.32553641228835045,
+      "learning_rate": 3.746065057712487e-05,
+      "loss": 0.4479,
+      "step": 691
+    },
+    {
+      "epoch": 0.9794762915782024,
+      "grad_norm": 0.2984969942135477,
+      "learning_rate": 3.743441762854145e-05,
+      "loss": 0.4449,
+      "step": 692
+    },
+    {
+      "epoch": 0.9808917197452229,
+      "grad_norm": 0.26360136789202876,
+      "learning_rate": 3.740818467995803e-05,
+      "loss": 0.4169,
+      "step": 693
+    },
+    {
+      "epoch": 0.9823071479122435,
+      "grad_norm": 0.3002600679766135,
+      "learning_rate": 3.738195173137461e-05,
+      "loss": 0.4327,
+      "step": 694
+    },
+    {
+      "epoch": 0.9837225760792639,
+      "grad_norm": 0.26761460969924294,
+      "learning_rate": 3.7355718782791185e-05,
+      "loss": 0.4295,
+      "step": 695
+    },
+    {
+      "epoch": 0.9851380042462845,
+      "grad_norm": 0.2774009798673986,
+      "learning_rate": 3.7329485834207765e-05,
+      "loss": 0.4321,
+      "step": 696
+    },
+    {
+      "epoch": 0.9865534324133051,
+      "grad_norm": 0.2963916085601546,
+      "learning_rate": 3.7303252885624345e-05,
+      "loss": 0.418,
+      "step": 697
+    },
+    {
+      "epoch": 0.9879688605803255,
+      "grad_norm": 0.22876594613173415,
+      "learning_rate": 3.7277019937040926e-05,
+      "loss": 0.4214,
+      "step": 698
+    },
+    {
+      "epoch": 0.9893842887473461,
+      "grad_norm": 0.2556840202116241,
+      "learning_rate": 3.7250786988457506e-05,
+      "loss": 0.4431,
+      "step": 699
+    },
+    {
+      "epoch": 0.9907997169143666,
+      "grad_norm": 0.2347036488614427,
+      "learning_rate": 3.722455403987408e-05,
+      "loss": 0.4317,
+      "step": 700
+    },
+    {
+      "epoch": 0.9922151450813871,
+      "grad_norm": 0.26576919366732943,
+      "learning_rate": 3.719832109129066e-05,
+      "loss": 0.4173,
+      "step": 701
+    },
+    {
+      "epoch": 0.9936305732484076,
+      "grad_norm": 0.24149398831154964,
+      "learning_rate": 3.717208814270724e-05,
+      "loss": 0.4498,
+      "step": 702
+    },
+    {
+      "epoch": 0.9950460014154282,
+      "grad_norm": 0.24034887998143845,
+      "learning_rate": 3.714585519412382e-05,
+      "loss": 0.4434,
+      "step": 703
+    },
+    {
+      "epoch": 0.9964614295824487,
+      "grad_norm": 0.25419984417706315,
+      "learning_rate": 3.71196222455404e-05,
+      "loss": 0.441,
+      "step": 704
+    },
+    {
+      "epoch": 0.9978768577494692,
+      "grad_norm": 0.22469194229972944,
+      "learning_rate": 3.709338929695698e-05,
+      "loss": 0.444,
+      "step": 705
+    },
+    {
+      "epoch": 0.9992922859164898,
+      "grad_norm": 0.25259247575525834,
+      "learning_rate": 3.706715634837356e-05,
+      "loss": 0.4148,
+      "step": 706
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.25259247575525834,
+      "learning_rate": 3.704092339979014e-05,
+      "loss": 0.4196,
+      "step": 707
+    },
+    {
+      "epoch": 1.0014154281670204,
+      "grad_norm": 0.40330885122332444,
+      "learning_rate": 3.7014690451206715e-05,
+      "loss": 0.3866,
+      "step": 708
+    },
+    {
+      "epoch": 1.0028308563340411,
+      "grad_norm": 0.25073682692017346,
+      "learning_rate": 3.6988457502623295e-05,
+      "loss": 0.3448,
+      "step": 709
+    },
+    {
+      "epoch": 1.0042462845010616,
+      "grad_norm": 0.2853163729929654,
+      "learning_rate": 3.6962224554039876e-05,
+      "loss": 0.3736,
+      "step": 710
+    },
+    {
+      "epoch": 1.005661712668082,
+      "grad_norm": 0.27424387603392747,
+      "learning_rate": 3.6935991605456456e-05,
+      "loss": 0.3744,
+      "step": 711
+    },
+    {
+      "epoch": 1.0070771408351027,
+      "grad_norm": 0.37706972872591604,
+      "learning_rate": 3.6909758656873036e-05,
+      "loss": 0.392,
+      "step": 712
+    },
+    {
+      "epoch": 1.0084925690021231,
+      "grad_norm": 0.26747508609564463,
+      "learning_rate": 3.688352570828961e-05,
+      "loss": 0.375,
+      "step": 713
+    },
+    {
+      "epoch": 1.0099079971691436,
+      "grad_norm": 0.2511475510195298,
+      "learning_rate": 3.685729275970619e-05,
+      "loss": 0.3614,
+      "step": 714
+    },
+    {
+      "epoch": 1.0113234253361643,
+      "grad_norm": 0.2874709688360536,
+      "learning_rate": 3.683105981112277e-05,
+      "loss": 0.3633,
+      "step": 715
+    },
+    {
+      "epoch": 1.0127388535031847,
+      "grad_norm": 0.2664187385028287,
+      "learning_rate": 3.680482686253935e-05,
+      "loss": 0.3637,
+      "step": 716
+    },
+    {
+      "epoch": 1.0141542816702052,
+      "grad_norm": 0.2849073293934639,
+      "learning_rate": 3.677859391395593e-05,
+      "loss": 0.3692,
+      "step": 717
+    },
+    {
+      "epoch": 1.0155697098372258,
+      "grad_norm": 0.24917420653285635,
+      "learning_rate": 3.675236096537251e-05,
+      "loss": 0.3767,
+      "step": 718
+    },
+    {
+      "epoch": 1.0169851380042463,
+      "grad_norm": 0.25923165709703305,
+      "learning_rate": 3.672612801678909e-05,
+      "loss": 0.3672,
+      "step": 719
+    },
+    {
+      "epoch": 1.0184005661712667,
+      "grad_norm": 0.24804588645407025,
+      "learning_rate": 3.669989506820567e-05,
+      "loss": 0.385,
+      "step": 720
+    },
+    {
+      "epoch": 1.0198159943382874,
+      "grad_norm": 0.27820354930045915,
+      "learning_rate": 3.6673662119622245e-05,
+      "loss": 0.3797,
+      "step": 721
+    },
+    {
+      "epoch": 1.0212314225053079,
+      "grad_norm": 0.2571375959428547,
+      "learning_rate": 3.6647429171038825e-05,
+      "loss": 0.371,
+      "step": 722
+    },
+    {
+      "epoch": 1.0226468506723283,
+      "grad_norm": 0.22299950541449304,
+      "learning_rate": 3.6621196222455406e-05,
+      "loss": 0.3729,
+      "step": 723
+    },
+    {
+      "epoch": 1.024062278839349,
+      "grad_norm": 0.2672635859495513,
+      "learning_rate": 3.6594963273871986e-05,
+      "loss": 0.3673,
+      "step": 724
+    },
+    {
+      "epoch": 1.0254777070063694,
+      "grad_norm": 0.2238148343196082,
+      "learning_rate": 3.6568730325288566e-05,
+      "loss": 0.3525,
+      "step": 725
+    },
+    {
+      "epoch": 1.0268931351733899,
+      "grad_norm": 0.21943937606585898,
+      "learning_rate": 3.654249737670514e-05,
+      "loss": 0.3448,
+      "step": 726
+    },
+    {
+      "epoch": 1.0283085633404105,
+      "grad_norm": 0.2286191309250446,
+      "learning_rate": 3.651626442812172e-05,
+      "loss": 0.3634,
+      "step": 727
+    },
+    {
+      "epoch": 1.029723991507431,
+      "grad_norm": 0.21770251056841974,
+      "learning_rate": 3.64900314795383e-05,
+      "loss": 0.3529,
+      "step": 728
+    },
+    {
+      "epoch": 1.0311394196744514,
+      "grad_norm": 0.23287135537136047,
+      "learning_rate": 3.646379853095488e-05,
+      "loss": 0.3526,
+      "step": 729
+    },
+    {
+      "epoch": 1.0325548478414721,
+      "grad_norm": 0.2770755741062495,
+      "learning_rate": 3.643756558237146e-05,
+      "loss": 0.3834,
+      "step": 730
+    },
+    {
+      "epoch": 1.0339702760084926,
+      "grad_norm": 0.24566882013501856,
+      "learning_rate": 3.641133263378804e-05,
+      "loss": 0.3758,
+      "step": 731
+    },
+    {
+      "epoch": 1.035385704175513,
+      "grad_norm": 0.24987011426830819,
+      "learning_rate": 3.638509968520462e-05,
+      "loss": 0.3651,
+      "step": 732
+    },
+    {
+      "epoch": 1.0368011323425337,
+      "grad_norm": 0.31041065460026135,
+      "learning_rate": 3.63588667366212e-05,
+      "loss": 0.3715,
+      "step": 733
+    },
+    {
+      "epoch": 1.0382165605095541,
+      "grad_norm": 0.24177259651765037,
+      "learning_rate": 3.6332633788037775e-05,
+      "loss": 0.3708,
+      "step": 734
+    },
+    {
+      "epoch": 1.0396319886765746,
+      "grad_norm": 0.21996306286337883,
+      "learning_rate": 3.6306400839454356e-05,
+      "loss": 0.3765,
+      "step": 735
+    },
+    {
+      "epoch": 1.0410474168435953,
+      "grad_norm": 0.25494992550486884,
+      "learning_rate": 3.6280167890870936e-05,
+      "loss": 0.3627,
+      "step": 736
+    },
+    {
+      "epoch": 1.0424628450106157,
+      "grad_norm": 0.21505088357504742,
+      "learning_rate": 3.6253934942287516e-05,
+      "loss": 0.3598,
+      "step": 737
+    },
+    {
+      "epoch": 1.0438782731776362,
+      "grad_norm": 0.2262185276339128,
+      "learning_rate": 3.6227701993704097e-05,
+      "loss": 0.3526,
+      "step": 738
+    },
+    {
+      "epoch": 1.0452937013446568,
+      "grad_norm": 0.21645046957698266,
+      "learning_rate": 3.620146904512067e-05,
+      "loss": 0.3669,
+      "step": 739
+    },
+    {
+      "epoch": 1.0467091295116773,
+      "grad_norm": 0.24673938200497808,
+      "learning_rate": 3.617523609653725e-05,
+      "loss": 0.3857,
+      "step": 740
+    },
+    {
+      "epoch": 1.0481245576786977,
+      "grad_norm": 0.2553921948350425,
+      "learning_rate": 3.614900314795383e-05,
+      "loss": 0.3694,
+      "step": 741
+    },
+    {
+      "epoch": 1.0495399858457184,
+      "grad_norm": 0.24339109836533926,
+      "learning_rate": 3.612277019937041e-05,
+      "loss": 0.3623,
+      "step": 742
+    },
+    {
+      "epoch": 1.0509554140127388,
+      "grad_norm": 0.24481636850870547,
+      "learning_rate": 3.609653725078699e-05,
+      "loss": 0.3691,
+      "step": 743
+    },
+    {
+      "epoch": 1.0523708421797593,
+      "grad_norm": 0.24077314804690694,
+      "learning_rate": 3.6070304302203565e-05,
+      "loss": 0.3884,
+      "step": 744
+    },
+    {
+      "epoch": 1.05378627034678,
+      "grad_norm": 0.26055509725576326,
+      "learning_rate": 3.604407135362015e-05,
+      "loss": 0.3675,
+      "step": 745
+    },
+    {
+      "epoch": 1.0552016985138004,
+      "grad_norm": 0.22320689999537213,
+      "learning_rate": 3.601783840503673e-05,
+      "loss": 0.3769,
+      "step": 746
+    },
+    {
+      "epoch": 1.0566171266808209,
+      "grad_norm": 0.20575120215372295,
+      "learning_rate": 3.599160545645331e-05,
+      "loss": 0.3589,
+      "step": 747
+    },
+    {
+      "epoch": 1.0580325548478415,
+      "grad_norm": 0.2258193270812288,
+      "learning_rate": 3.5965372507869886e-05,
+      "loss": 0.3655,
+      "step": 748
+    },
+    {
+      "epoch": 1.059447983014862,
+      "grad_norm": 0.22200653826263894,
+      "learning_rate": 3.5939139559286466e-05,
+      "loss": 0.3663,
+      "step": 749
+    },
+    {
+      "epoch": 1.0608634111818824,
+      "grad_norm": 0.23199985002095336,
+      "learning_rate": 3.5912906610703046e-05,
+      "loss": 0.3856,
+      "step": 750
+    },
+    {
+      "epoch": 1.062278839348903,
+      "grad_norm": 0.23475624219008442,
+      "learning_rate": 3.588667366211963e-05,
+      "loss": 0.3646,
+      "step": 751
+    },
+    {
+      "epoch": 1.0636942675159236,
+      "grad_norm": 0.2363507036132983,
+      "learning_rate": 3.58604407135362e-05,
+      "loss": 0.3475,
+      "step": 752
+    },
+    {
+      "epoch": 1.065109695682944,
+      "grad_norm": 0.23554394717989005,
+      "learning_rate": 3.583420776495278e-05,
+      "loss": 0.3621,
+      "step": 753
+    },
+    {
+      "epoch": 1.0665251238499647,
+      "grad_norm": 0.25712807151393957,
+      "learning_rate": 3.580797481636936e-05,
+      "loss": 0.3758,
+      "step": 754
+    },
+    {
+      "epoch": 1.0679405520169851,
+      "grad_norm": 0.21980062026864217,
+      "learning_rate": 3.578174186778594e-05,
+      "loss": 0.3637,
+      "step": 755
+    },
+    {
+      "epoch": 1.0693559801840056,
+      "grad_norm": 0.24587003401940713,
+      "learning_rate": 3.575550891920252e-05,
+      "loss": 0.3495,
+      "step": 756
+    },
+    {
+      "epoch": 1.0707714083510262,
+      "grad_norm": 0.24943392215339394,
+      "learning_rate": 3.5729275970619095e-05,
+      "loss": 0.3668,
+      "step": 757
+    },
+    {
+      "epoch": 1.0721868365180467,
+      "grad_norm": 0.2817878261350363,
+      "learning_rate": 3.5703043022035675e-05,
+      "loss": 0.3667,
+      "step": 758
+    },
+    {
+      "epoch": 1.0736022646850671,
+      "grad_norm": 0.21980883296838102,
+      "learning_rate": 3.567681007345226e-05,
+      "loss": 0.3757,
+      "step": 759
+    },
+    {
+      "epoch": 1.0750176928520878,
+      "grad_norm": 0.25569911421845076,
+      "learning_rate": 3.565057712486884e-05,
+      "loss": 0.3617,
+      "step": 760
+    },
+    {
+      "epoch": 1.0764331210191083,
+      "grad_norm": 0.2510990438768182,
+      "learning_rate": 3.5624344176285416e-05,
+      "loss": 0.3755,
+      "step": 761
+    },
+    {
+      "epoch": 1.0778485491861287,
+      "grad_norm": 0.24448758916499447,
+      "learning_rate": 3.5598111227701996e-05,
+      "loss": 0.3701,
+      "step": 762
+    },
+    {
+      "epoch": 1.0792639773531494,
+      "grad_norm": 0.21955348175980843,
+      "learning_rate": 3.5571878279118577e-05,
+      "loss": 0.3614,
+      "step": 763
+    },
+    {
+      "epoch": 1.0806794055201698,
+      "grad_norm": 0.2763996762537386,
+      "learning_rate": 3.554564533053516e-05,
+      "loss": 0.3725,
+      "step": 764
+    },
+    {
+      "epoch": 1.0820948336871903,
+      "grad_norm": 0.22060232690914258,
+      "learning_rate": 3.551941238195173e-05,
+      "loss": 0.3546,
+      "step": 765
+    },
+    {
+      "epoch": 1.083510261854211,
+      "grad_norm": 0.2690127459305308,
+      "learning_rate": 3.549317943336831e-05,
+      "loss": 0.3701,
+      "step": 766
+    },
+    {
+      "epoch": 1.0849256900212314,
+      "grad_norm": 0.23378521213147532,
+      "learning_rate": 3.546694648478489e-05,
+      "loss": 0.3553,
+      "step": 767
+    },
+    {
+      "epoch": 1.0863411181882519,
+      "grad_norm": 0.293776933751109,
+      "learning_rate": 3.544071353620147e-05,
+      "loss": 0.3867,
+      "step": 768
+    },
+    {
+      "epoch": 1.0877565463552725,
+      "grad_norm": 0.2769516330391029,
+      "learning_rate": 3.541448058761805e-05,
+      "loss": 0.3777,
+      "step": 769
+    },
+    {
+      "epoch": 1.089171974522293,
+      "grad_norm": 0.2766476933433498,
+      "learning_rate": 3.5388247639034625e-05,
+      "loss": 0.3654,
+      "step": 770
+    },
+    {
+      "epoch": 1.0905874026893134,
+      "grad_norm": 0.24123409080089156,
+      "learning_rate": 3.5362014690451205e-05,
+      "loss": 0.3706,
+      "step": 771
+    },
+    {
+      "epoch": 1.092002830856334,
+      "grad_norm": 0.24952465460683615,
+      "learning_rate": 3.5335781741867786e-05,
+      "loss": 0.3579,
+      "step": 772
+    },
+    {
+      "epoch": 1.0934182590233545,
+      "grad_norm": 0.25575201537604786,
+      "learning_rate": 3.530954879328437e-05,
+      "loss": 0.3673,
+      "step": 773
+    },
+    {
+      "epoch": 1.094833687190375,
+      "grad_norm": 0.23603855217803518,
+      "learning_rate": 3.5283315844700946e-05,
+      "loss": 0.3794,
+      "step": 774
+    },
+    {
+      "epoch": 1.0962491153573957,
+      "grad_norm": 0.23779241121404207,
+      "learning_rate": 3.5257082896117526e-05,
+      "loss": 0.3864,
+      "step": 775
+    },
+    {
+      "epoch": 1.0976645435244161,
+      "grad_norm": 0.25184642467770774,
+      "learning_rate": 3.523084994753411e-05,
+      "loss": 0.3533,
+      "step": 776
+    },
+    {
+      "epoch": 1.0990799716914366,
+      "grad_norm": 0.22707567886508429,
+      "learning_rate": 3.520461699895069e-05,
+      "loss": 0.381,
+      "step": 777
+    },
+    {
+      "epoch": 1.1004953998584572,
+      "grad_norm": 0.22382456529288564,
+      "learning_rate": 3.517838405036726e-05,
+      "loss": 0.3599,
+      "step": 778
+    },
+    {
+      "epoch": 1.1019108280254777,
+      "grad_norm": 0.22643392806140442,
+      "learning_rate": 3.515215110178384e-05,
+      "loss": 0.3549,
+      "step": 779
+    },
+    {
+      "epoch": 1.1033262561924981,
+      "grad_norm": 0.249318689583364,
+      "learning_rate": 3.512591815320042e-05,
+      "loss": 0.3767,
+      "step": 780
+    },
+    {
+      "epoch": 1.1047416843595188,
+      "grad_norm": 0.22818899513469243,
+      "learning_rate": 3.5099685204617e-05,
+      "loss": 0.3526,
+      "step": 781
+    },
+    {
+      "epoch": 1.1061571125265393,
+      "grad_norm": 0.22205582440125177,
+      "learning_rate": 3.507345225603358e-05,
+      "loss": 0.377,
+      "step": 782
+    },
+    {
+      "epoch": 1.1075725406935597,
+      "grad_norm": 0.23715401218273183,
+      "learning_rate": 3.5047219307450155e-05,
+      "loss": 0.3631,
+      "step": 783
+    },
+    {
+      "epoch": 1.1089879688605804,
+      "grad_norm": 0.24731495141449844,
+      "learning_rate": 3.5020986358866735e-05,
+      "loss": 0.357,
+      "step": 784
+    },
+    {
+      "epoch": 1.1104033970276008,
+      "grad_norm": 0.21043244736337724,
+      "learning_rate": 3.4994753410283316e-05,
+      "loss": 0.3758,
+      "step": 785
+    },
+    {
+      "epoch": 1.1118188251946213,
+      "grad_norm": 0.22034294434212712,
+      "learning_rate": 3.4968520461699896e-05,
+      "loss": 0.3656,
+      "step": 786
+    },
+    {
+      "epoch": 1.113234253361642,
+      "grad_norm": 0.22364850456219934,
+      "learning_rate": 3.4942287513116476e-05,
+      "loss": 0.3573,
+      "step": 787
+    },
+    {
+      "epoch": 1.1146496815286624,
+      "grad_norm": 0.2271183339148488,
+      "learning_rate": 3.4916054564533057e-05,
+      "loss": 0.3667,
+      "step": 788
+    },
+    {
+      "epoch": 1.1160651096956828,
+      "grad_norm": 0.2395804298830507,
+      "learning_rate": 3.488982161594964e-05,
+      "loss": 0.3815,
+      "step": 789
+    },
+    {
+      "epoch": 1.1174805378627035,
+      "grad_norm": 0.2280965467181193,
+      "learning_rate": 3.486358866736622e-05,
+      "loss": 0.3696,
+      "step": 790
+    },
+    {
+      "epoch": 1.118895966029724,
+      "grad_norm": 0.27142514033989723,
+      "learning_rate": 3.483735571878279e-05,
+      "loss": 0.3635,
+      "step": 791
+    },
+    {
+      "epoch": 1.1203113941967444,
+      "grad_norm": 0.25250123373484423,
+      "learning_rate": 3.481112277019937e-05,
+      "loss": 0.3637,
+      "step": 792
+    },
+    {
+      "epoch": 1.121726822363765,
+      "grad_norm": 0.2409488628984419,
+      "learning_rate": 3.478488982161595e-05,
+      "loss": 0.3767,
+      "step": 793
+    },
+    {
+      "epoch": 1.1231422505307855,
+      "grad_norm": 0.26703802478917377,
+      "learning_rate": 3.475865687303253e-05,
+      "loss": 0.3581,
+      "step": 794
+    },
+    {
+      "epoch": 1.124557678697806,
+      "grad_norm": 0.23190208676783128,
+      "learning_rate": 3.473242392444911e-05,
+      "loss": 0.3379,
+      "step": 795
+    },
+    {
+      "epoch": 1.1259731068648267,
+      "grad_norm": 0.26784040366950523,
+      "learning_rate": 3.4706190975865685e-05,
+      "loss": 0.367,
+      "step": 796
+    },
+    {
+      "epoch": 1.127388535031847,
+      "grad_norm": 0.2191740845730247,
+      "learning_rate": 3.4679958027282266e-05,
+      "loss": 0.3731,
+      "step": 797
+    },
+    {
+      "epoch": 1.1288039631988678,
+      "grad_norm": 0.26068999486444,
+      "learning_rate": 3.4653725078698846e-05,
+      "loss": 0.3631,
+      "step": 798
+    },
+    {
+      "epoch": 1.1302193913658882,
+      "grad_norm": 0.2311127862371228,
+      "learning_rate": 3.4627492130115426e-05,
+      "loss": 0.3435,
+      "step": 799
+    },
+    {
+      "epoch": 1.1316348195329087,
+      "grad_norm": 0.22156342608749344,
+      "learning_rate": 3.4601259181532006e-05,
+      "loss": 0.3791,
+      "step": 800
+    },
+    {
+      "epoch": 1.1330502476999293,
+      "grad_norm": 0.23528767608314413,
+      "learning_rate": 3.457502623294859e-05,
+      "loss": 0.379,
+      "step": 801
+    },
+    {
+      "epoch": 1.1344656758669498,
+      "grad_norm": 0.2588075699698271,
+      "learning_rate": 3.454879328436517e-05,
+      "loss": 0.387,
+      "step": 802
+    },
+    {
+      "epoch": 1.1358811040339702,
+      "grad_norm": 0.22063634101813656,
+      "learning_rate": 3.452256033578175e-05,
+      "loss": 0.3783,
+      "step": 803
+    },
+    {
+      "epoch": 1.137296532200991,
+      "grad_norm": 0.2759546955219502,
+      "learning_rate": 3.449632738719832e-05,
+      "loss": 0.353,
+      "step": 804
+    },
+    {
+      "epoch": 1.1387119603680114,
+      "grad_norm": 0.2590375007075574,
+      "learning_rate": 3.44700944386149e-05,
+      "loss": 0.372,
+      "step": 805
+    },
+    {
+      "epoch": 1.1401273885350318,
+      "grad_norm": 0.26235020909100687,
+      "learning_rate": 3.444386149003148e-05,
+      "loss": 0.3681,
+      "step": 806
+    },
+    {
+      "epoch": 1.1415428167020525,
+      "grad_norm": 0.25990600045468054,
+      "learning_rate": 3.441762854144806e-05,
+      "loss": 0.3734,
+      "step": 807
+    },
+    {
+      "epoch": 1.142958244869073,
+      "grad_norm": 0.25421808078588554,
+      "learning_rate": 3.439139559286464e-05,
+      "loss": 0.3643,
+      "step": 808
+    },
+    {
+      "epoch": 1.1443736730360934,
+      "grad_norm": 0.25954139055763814,
+      "learning_rate": 3.4365162644281215e-05,
+      "loss": 0.3828,
+      "step": 809
+    },
+    {
+      "epoch": 1.145789101203114,
+      "grad_norm": 0.2485386767677047,
+      "learning_rate": 3.4338929695697796e-05,
+      "loss": 0.3768,
+      "step": 810
+    },
+    {
+      "epoch": 1.1472045293701345,
+      "grad_norm": 0.22979787035800808,
+      "learning_rate": 3.4312696747114376e-05,
+      "loss": 0.3636,
+      "step": 811
+    },
+    {
+      "epoch": 1.148619957537155,
+      "grad_norm": 0.237713998539904,
+      "learning_rate": 3.4286463798530956e-05,
+      "loss": 0.3603,
+      "step": 812
+    },
+    {
+      "epoch": 1.1500353857041756,
+      "grad_norm": 0.21183509960801158,
+      "learning_rate": 3.4260230849947537e-05,
+      "loss": 0.366,
+      "step": 813
+    },
+    {
+      "epoch": 1.151450813871196,
+      "grad_norm": 0.2232984894405917,
+      "learning_rate": 3.423399790136412e-05,
+      "loss": 0.3552,
+      "step": 814
+    },
+    {
+      "epoch": 1.1528662420382165,
+      "grad_norm": 0.23826655047625447,
+      "learning_rate": 3.42077649527807e-05,
+      "loss": 0.3733,
+      "step": 815
+    },
+    {
+      "epoch": 1.1542816702052372,
+      "grad_norm": 0.23222605074737757,
+      "learning_rate": 3.418153200419728e-05,
+      "loss": 0.3855,
+      "step": 816
+    },
+    {
+      "epoch": 1.1556970983722576,
+      "grad_norm": 0.23808854243134703,
+      "learning_rate": 3.415529905561385e-05,
+      "loss": 0.3759,
+      "step": 817
+    },
+    {
+      "epoch": 1.157112526539278,
+      "grad_norm": 0.2682038429151991,
+      "learning_rate": 3.412906610703043e-05,
+      "loss": 0.3891,
+      "step": 818
+    },
+    {
+      "epoch": 1.1585279547062988,
+      "grad_norm": 0.25684259873831616,
+      "learning_rate": 3.410283315844701e-05,
+      "loss": 0.3671,
+      "step": 819
+    },
+    {
+      "epoch": 1.1599433828733192,
+      "grad_norm": 0.22366707096330654,
+      "learning_rate": 3.407660020986359e-05,
+      "loss": 0.3679,
+      "step": 820
+    },
+    {
+      "epoch": 1.1613588110403397,
+      "grad_norm": 0.23320918985975697,
+      "learning_rate": 3.405036726128017e-05,
+      "loss": 0.365,
+      "step": 821
+    },
+    {
+      "epoch": 1.1627742392073603,
+      "grad_norm": 0.2560345240813397,
+      "learning_rate": 3.4024134312696746e-05,
+      "loss": 0.3761,
+      "step": 822
+    },
+    {
+      "epoch": 1.1641896673743808,
+      "grad_norm": 0.22099107331766332,
+      "learning_rate": 3.3997901364113326e-05,
+      "loss": 0.3511,
+      "step": 823
+    },
+    {
+      "epoch": 1.1656050955414012,
+      "grad_norm": 0.23197104371717864,
+      "learning_rate": 3.3971668415529906e-05,
+      "loss": 0.3677,
+      "step": 824
+    },
+    {
+      "epoch": 1.167020523708422,
+      "grad_norm": 0.25715111525636897,
+      "learning_rate": 3.3945435466946486e-05,
+      "loss": 0.3597,
+      "step": 825
+    },
+    {
+      "epoch": 1.1684359518754424,
+      "grad_norm": 0.22692191767057515,
+      "learning_rate": 3.391920251836307e-05,
+      "loss": 0.3698,
+      "step": 826
+    },
+    {
+      "epoch": 1.1698513800424628,
+      "grad_norm": 0.2496074837513205,
+      "learning_rate": 3.389296956977964e-05,
+      "loss": 0.3659,
+      "step": 827
+    },
+    {
+      "epoch": 1.1712668082094835,
+      "grad_norm": 0.2502694016706319,
+      "learning_rate": 3.386673662119623e-05,
+      "loss": 0.3759,
+      "step": 828
+    },
+    {
+      "epoch": 1.172682236376504,
+      "grad_norm": 0.2532655245787918,
+      "learning_rate": 3.384050367261281e-05,
+      "loss": 0.3592,
+      "step": 829
+    },
+    {
+      "epoch": 1.1740976645435244,
+      "grad_norm": 0.2593297195760385,
+      "learning_rate": 3.381427072402938e-05,
+      "loss": 0.375,
+      "step": 830
+    },
+    {
+      "epoch": 1.175513092710545,
+      "grad_norm": 0.2359968199749705,
+      "learning_rate": 3.378803777544596e-05,
+      "loss": 0.3825,
+      "step": 831
+    },
+    {
+      "epoch": 1.1769285208775655,
+      "grad_norm": 0.25889001677763246,
+      "learning_rate": 3.376180482686254e-05,
+      "loss": 0.3717,
+      "step": 832
+    },
+    {
+      "epoch": 1.178343949044586,
+      "grad_norm": 0.23160734530040403,
+      "learning_rate": 3.373557187827912e-05,
+      "loss": 0.3428,
+      "step": 833
+    },
+    {
+      "epoch": 1.1797593772116066,
+      "grad_norm": 0.2219389279783254,
+      "learning_rate": 3.37093389296957e-05,
+      "loss": 0.4042,
+      "step": 834
+    },
+    {
+      "epoch": 1.181174805378627,
+      "grad_norm": 0.26257436972385445,
+      "learning_rate": 3.3683105981112276e-05,
+      "loss": 0.3754,
+      "step": 835
+    },
+    {
+      "epoch": 1.1825902335456475,
+      "grad_norm": 0.2747050720214299,
+      "learning_rate": 3.3656873032528856e-05,
+      "loss": 0.3639,
+      "step": 836
+    },
+    {
+      "epoch": 1.1840056617126682,
+      "grad_norm": 0.21820718002888326,
+      "learning_rate": 3.3630640083945436e-05,
+      "loss": 0.3603,
+      "step": 837
+    },
+    {
+      "epoch": 1.1854210898796886,
+      "grad_norm": 0.24426222094514102,
+      "learning_rate": 3.3604407135362017e-05,
+      "loss": 0.3909,
+      "step": 838
+    },
+    {
+      "epoch": 1.186836518046709,
+      "grad_norm": 0.2697440918806726,
+      "learning_rate": 3.35781741867786e-05,
+      "loss": 0.3722,
+      "step": 839
+    },
+    {
+      "epoch": 1.1882519462137298,
+      "grad_norm": 0.24904913274161672,
+      "learning_rate": 3.355194123819517e-05,
+      "loss": 0.3633,
+      "step": 840
+    },
+    {
+      "epoch": 1.1896673743807502,
+      "grad_norm": 0.22450159530467048,
+      "learning_rate": 3.352570828961175e-05,
+      "loss": 0.3727,
+      "step": 841
+    },
+    {
+      "epoch": 1.1910828025477707,
+      "grad_norm": 0.24682980603479437,
+      "learning_rate": 3.349947534102834e-05,
+      "loss": 0.3739,
+      "step": 842
+    },
+    {
+      "epoch": 1.1924982307147913,
+      "grad_norm": 0.22876377293747732,
+      "learning_rate": 3.347324239244491e-05,
+      "loss": 0.3589,
+      "step": 843
+    },
+    {
+      "epoch": 1.1939136588818118,
+      "grad_norm": 0.2485727894540362,
+      "learning_rate": 3.344700944386149e-05,
+      "loss": 0.3704,
+      "step": 844
+    },
+    {
+      "epoch": 1.1953290870488322,
+      "grad_norm": 0.24521133225067981,
+      "learning_rate": 3.342077649527807e-05,
+      "loss": 0.3826,
+      "step": 845
+    },
+    {
+      "epoch": 1.196744515215853,
+      "grad_norm": 0.24329138505723324,
+      "learning_rate": 3.339454354669465e-05,
+      "loss": 0.3636,
+      "step": 846
+    },
+    {
+      "epoch": 1.1981599433828733,
+      "grad_norm": 0.23853104947938453,
+      "learning_rate": 3.336831059811123e-05,
+      "loss": 0.3587,
+      "step": 847
+    },
+    {
+      "epoch": 1.1995753715498938,
+      "grad_norm": 0.24251162814105084,
+      "learning_rate": 3.3342077649527806e-05,
+      "loss": 0.3616,
+      "step": 848
+    },
+    {
+      "epoch": 1.2009907997169145,
+      "grad_norm": 0.26516103818641135,
+      "learning_rate": 3.3315844700944386e-05,
+      "loss": 0.3748,
+      "step": 849
+    },
+    {
+      "epoch": 1.202406227883935,
+      "grad_norm": 0.20688415308994873,
+      "learning_rate": 3.3289611752360966e-05,
+      "loss": 0.3566,
+      "step": 850
+    },
+    {
+      "epoch": 1.2038216560509554,
+      "grad_norm": 0.23466141155912668,
+      "learning_rate": 3.326337880377755e-05,
+      "loss": 0.3806,
+      "step": 851
+    },
+    {
+      "epoch": 1.205237084217976,
+      "grad_norm": 0.23207353563777747,
+      "learning_rate": 3.323714585519413e-05,
+      "loss": 0.3568,
+      "step": 852
+    },
+    {
+      "epoch": 1.2066525123849965,
+      "grad_norm": 0.22285080510272276,
+      "learning_rate": 3.32109129066107e-05,
+      "loss": 0.3782,
+      "step": 853
+    },
+    {
+      "epoch": 1.208067940552017,
+      "grad_norm": 0.21098164834539135,
+      "learning_rate": 3.318467995802728e-05,
+      "loss": 0.3632,
+      "step": 854
+    },
+    {
+      "epoch": 1.2094833687190376,
+      "grad_norm": 0.2158541149523356,
+      "learning_rate": 3.315844700944386e-05,
+      "loss": 0.3615,
+      "step": 855
+    },
+    {
+      "epoch": 1.210898796886058,
+      "grad_norm": 0.22165253427051132,
+      "learning_rate": 3.313221406086044e-05,
+      "loss": 0.3655,
+      "step": 856
+    },
+    {
+      "epoch": 1.2123142250530785,
+      "grad_norm": 0.21366192902117054,
+      "learning_rate": 3.310598111227702e-05,
+      "loss": 0.3647,
+      "step": 857
+    },
+    {
+      "epoch": 1.2137296532200992,
+      "grad_norm": 0.23320640229059145,
+      "learning_rate": 3.30797481636936e-05,
+      "loss": 0.3843,
+      "step": 858
+    },
+    {
+      "epoch": 1.2151450813871196,
+      "grad_norm": 0.22666909848091002,
+      "learning_rate": 3.305351521511018e-05,
+      "loss": 0.3575,
+      "step": 859
+    },
+    {
+      "epoch": 1.21656050955414,
+      "grad_norm": 0.21896446877425702,
+      "learning_rate": 3.302728226652676e-05,
+      "loss": 0.3669,
+      "step": 860
+    },
+    {
+      "epoch": 1.2179759377211608,
+      "grad_norm": 0.22732149506115526,
+      "learning_rate": 3.3001049317943336e-05,
+      "loss": 0.372,
+      "step": 861
+    },
+    {
+      "epoch": 1.2193913658881812,
+      "grad_norm": 0.2331464445769384,
+      "learning_rate": 3.2974816369359916e-05,
+      "loss": 0.3571,
+      "step": 862
+    },
+    {
+      "epoch": 1.2208067940552016,
+      "grad_norm": 0.2341286538913884,
+      "learning_rate": 3.29485834207765e-05,
+      "loss": 0.3619,
+      "step": 863
+    },
+    {
+      "epoch": 1.2222222222222223,
+      "grad_norm": 0.242917135991931,
+      "learning_rate": 3.292235047219308e-05,
+      "loss": 0.3836,
+      "step": 864
+    },
+    {
+      "epoch": 1.2236376503892428,
+      "grad_norm": 0.2377434057216746,
+      "learning_rate": 3.289611752360966e-05,
+      "loss": 0.3707,
+      "step": 865
+    },
+    {
+      "epoch": 1.2250530785562632,
+      "grad_norm": 0.23061035158384427,
+      "learning_rate": 3.286988457502623e-05,
+      "loss": 0.3649,
+      "step": 866
+    },
+    {
+      "epoch": 1.226468506723284,
+      "grad_norm": 0.26493593711379954,
+      "learning_rate": 3.284365162644281e-05,
+      "loss": 0.3547,
+      "step": 867
+    },
+    {
+      "epoch": 1.2278839348903043,
+      "grad_norm": 0.2817735921773945,
+      "learning_rate": 3.281741867785939e-05,
+      "loss": 0.3581,
+      "step": 868
+    },
+    {
+      "epoch": 1.2292993630573248,
+      "grad_norm": 0.24744764647826778,
+      "learning_rate": 3.279118572927597e-05,
+      "loss": 0.3572,
+      "step": 869
+    },
+    {
+      "epoch": 1.2307147912243455,
+      "grad_norm": 0.257164415071009,
+      "learning_rate": 3.276495278069255e-05,
+      "loss": 0.3772,
+      "step": 870
+    },
+    {
+      "epoch": 1.232130219391366,
+      "grad_norm": 0.2539971232037796,
+      "learning_rate": 3.273871983210913e-05,
+      "loss": 0.386,
+      "step": 871
+    },
+    {
+      "epoch": 1.2335456475583864,
+      "grad_norm": 0.25451197681884186,
+      "learning_rate": 3.271248688352571e-05,
+      "loss": 0.3741,
+      "step": 872
+    },
+    {
+      "epoch": 1.234961075725407,
+      "grad_norm": 0.23119645117655382,
+      "learning_rate": 3.268625393494229e-05,
+      "loss": 0.356,
+      "step": 873
+    },
+    {
+      "epoch": 1.2363765038924275,
+      "grad_norm": 0.22536736183119602,
+      "learning_rate": 3.2660020986358866e-05,
+      "loss": 0.353,
+      "step": 874
+    },
+    {
+      "epoch": 1.237791932059448,
+      "grad_norm": 0.228868007542863,
+      "learning_rate": 3.2633788037775447e-05,
+      "loss": 0.3678,
+      "step": 875
+    },
+    {
+      "epoch": 1.2392073602264686,
+      "grad_norm": 0.22422487236974076,
+      "learning_rate": 3.260755508919203e-05,
+      "loss": 0.3516,
+      "step": 876
+    },
+    {
+      "epoch": 1.240622788393489,
+      "grad_norm": 0.2273170241203407,
+      "learning_rate": 3.258132214060861e-05,
+      "loss": 0.3637,
+      "step": 877
+    },
+    {
+      "epoch": 1.2420382165605095,
+      "grad_norm": 0.2339973397823559,
+      "learning_rate": 3.255508919202519e-05,
+      "loss": 0.3419,
+      "step": 878
+    },
+    {
+      "epoch": 1.2434536447275302,
+      "grad_norm": 0.22059105204937277,
+      "learning_rate": 3.252885624344176e-05,
+      "loss": 0.3566,
+      "step": 879
+    },
+    {
+      "epoch": 1.2448690728945506,
+      "grad_norm": 0.21716790979082953,
+      "learning_rate": 3.250262329485834e-05,
+      "loss": 0.3496,
+      "step": 880
+    },
+    {
+      "epoch": 1.246284501061571,
+      "grad_norm": 0.22483525203992455,
+      "learning_rate": 3.247639034627492e-05,
+      "loss": 0.3602,
+      "step": 881
+    },
+    {
+      "epoch": 1.2476999292285917,
+      "grad_norm": 0.2205110253793736,
+      "learning_rate": 3.24501573976915e-05,
+      "loss": 0.3503,
+      "step": 882
+    },
+    {
+      "epoch": 1.2491153573956122,
+      "grad_norm": 0.2451649817997365,
+      "learning_rate": 3.242392444910808e-05,
+      "loss": 0.3869,
+      "step": 883
+    },
+    {
+      "epoch": 1.2505307855626326,
+      "grad_norm": 0.22327314044589933,
+      "learning_rate": 3.239769150052466e-05,
+      "loss": 0.3786,
+      "step": 884
+    },
+    {
+      "epoch": 1.2519462137296533,
+      "grad_norm": 0.2293582655618987,
+      "learning_rate": 3.237145855194124e-05,
+      "loss": 0.3582,
+      "step": 885
+    },
+    {
+      "epoch": 1.2533616418966738,
+      "grad_norm": 0.22115564542833394,
+      "learning_rate": 3.234522560335782e-05,
+      "loss": 0.3922,
+      "step": 886
+    },
+    {
+      "epoch": 1.2547770700636942,
+      "grad_norm": 0.2243335412240295,
+      "learning_rate": 3.2318992654774396e-05,
+      "loss": 0.3537,
+      "step": 887
+    },
+    {
+      "epoch": 1.2561924982307149,
+      "grad_norm": 0.2503356687519151,
+      "learning_rate": 3.229275970619098e-05,
+      "loss": 0.3772,
+      "step": 888
+    },
+    {
+      "epoch": 1.2576079263977353,
+      "grad_norm": 0.23483831990082277,
+      "learning_rate": 3.226652675760756e-05,
+      "loss": 0.3745,
+      "step": 889
+    },
+    {
+      "epoch": 1.2590233545647558,
+      "grad_norm": 0.2335033231898062,
+      "learning_rate": 3.224029380902414e-05,
+      "loss": 0.3437,
+      "step": 890
+    },
+    {
+      "epoch": 1.2604387827317765,
+      "grad_norm": 0.23691944679552432,
+      "learning_rate": 3.221406086044072e-05,
+      "loss": 0.3677,
+      "step": 891
+    },
+    {
+      "epoch": 1.261854210898797,
+      "grad_norm": 0.2181595939852755,
+      "learning_rate": 3.218782791185729e-05,
+      "loss": 0.3738,
+      "step": 892
+    },
+    {
+      "epoch": 1.2632696390658174,
+      "grad_norm": 0.22391237606573866,
+      "learning_rate": 3.216159496327387e-05,
+      "loss": 0.3631,
+      "step": 893
+    },
+    {
+      "epoch": 1.264685067232838,
+      "grad_norm": 0.233880286054612,
+      "learning_rate": 3.213536201469045e-05,
+      "loss": 0.3727,
+      "step": 894
+    },
+    {
+      "epoch": 1.2661004953998585,
+      "grad_norm": 0.23652089597038298,
+      "learning_rate": 3.210912906610703e-05,
+      "loss": 0.3751,
+      "step": 895
+    },
+    {
+      "epoch": 1.267515923566879,
+      "grad_norm": 0.21802470429593748,
+      "learning_rate": 3.2082896117523605e-05,
+      "loss": 0.3812,
+      "step": 896
+    },
+    {
+      "epoch": 1.2689313517338996,
+      "grad_norm": 0.5826761382172103,
+      "learning_rate": 3.205666316894019e-05,
+      "loss": 0.3636,
+      "step": 897
+    },
+    {
+      "epoch": 1.27034677990092,
+      "grad_norm": 0.2363193520278659,
+      "learning_rate": 3.203043022035677e-05,
+      "loss": 0.3604,
+      "step": 898
+    },
+    {
+      "epoch": 1.2717622080679405,
+      "grad_norm": 0.2363204162806771,
+      "learning_rate": 3.200419727177335e-05,
+      "loss": 0.3647,
+      "step": 899
+    },
+    {
+      "epoch": 1.2731776362349612,
+      "grad_norm": 0.25601617210524324,
+      "learning_rate": 3.1977964323189927e-05,
+      "loss": 0.3785,
+      "step": 900
+    },
+    {
+      "epoch": 1.2745930644019816,
+      "grad_norm": 0.21012097538566218,
+      "learning_rate": 3.195173137460651e-05,
+      "loss": 0.3672,
+      "step": 901
+    },
+    {
+      "epoch": 1.276008492569002,
+      "grad_norm": 0.2559276276421834,
+      "learning_rate": 3.192549842602309e-05,
+      "loss": 0.3651,
+      "step": 902
+    },
+    {
+      "epoch": 1.2774239207360227,
+      "grad_norm": 0.27317903610927285,
+      "learning_rate": 3.189926547743967e-05,
+      "loss": 0.3499,
+      "step": 903
+    },
+    {
+      "epoch": 1.2788393489030432,
+      "grad_norm": 0.23156491652563996,
+      "learning_rate": 3.187303252885625e-05,
+      "loss": 0.3631,
+      "step": 904
+    },
+    {
+      "epoch": 1.2802547770700636,
+      "grad_norm": 0.24253951927325784,
+      "learning_rate": 3.184679958027282e-05,
+      "loss": 0.354,
+      "step": 905
+    },
+    {
+      "epoch": 1.2816702052370843,
+      "grad_norm": 0.2278015940984733,
+      "learning_rate": 3.18205666316894e-05,
+      "loss": 0.3899,
+      "step": 906
+    },
+    {
+      "epoch": 1.2830856334041048,
+      "grad_norm": 0.26821722319228736,
+      "learning_rate": 3.179433368310598e-05,
+      "loss": 0.3591,
+      "step": 907
+    },
+    {
+      "epoch": 1.2845010615711252,
+      "grad_norm": 0.2201091318809155,
+      "learning_rate": 3.176810073452256e-05,
+      "loss": 0.3809,
+      "step": 908
+    },
+    {
+      "epoch": 1.2859164897381459,
+      "grad_norm": 0.22005725140665,
+      "learning_rate": 3.1741867785939136e-05,
+      "loss": 0.3537,
+      "step": 909
+    },
+    {
+      "epoch": 1.2873319179051663,
+      "grad_norm": 0.25439967947504993,
+      "learning_rate": 3.1715634837355716e-05,
+      "loss": 0.3802,
+      "step": 910
+    },
+    {
+      "epoch": 1.2887473460721868,
+      "grad_norm": 0.2077398037825165,
+      "learning_rate": 3.16894018887723e-05,
+      "loss": 0.3529,
+      "step": 911
+    },
+    {
+      "epoch": 1.2901627742392074,
+      "grad_norm": 0.2138094473716172,
+      "learning_rate": 3.166316894018888e-05,
+      "loss": 0.3541,
+      "step": 912
+    },
+    {
+      "epoch": 1.291578202406228,
+      "grad_norm": 0.274627449971989,
+      "learning_rate": 3.163693599160546e-05,
+      "loss": 0.3826,
+      "step": 913
+    },
+    {
+      "epoch": 1.2929936305732483,
+      "grad_norm": 0.23347713681140717,
+      "learning_rate": 3.161070304302204e-05,
+      "loss": 0.3716,
+      "step": 914
+    },
+    {
+      "epoch": 1.294409058740269,
+      "grad_norm": 0.22243696869454346,
+      "learning_rate": 3.158447009443862e-05,
+      "loss": 0.3542,
+      "step": 915
+    },
+    {
+      "epoch": 1.2958244869072895,
+      "grad_norm": 0.2482782039679565,
+      "learning_rate": 3.15582371458552e-05,
+      "loss": 0.3613,
+      "step": 916
+    },
+    {
+      "epoch": 1.29723991507431,
+      "grad_norm": 0.22891859018706157,
+      "learning_rate": 3.153200419727178e-05,
+      "loss": 0.3742,
+      "step": 917
+    },
+    {
+      "epoch": 1.2986553432413306,
+      "grad_norm": 0.2261013605218962,
+      "learning_rate": 3.150577124868835e-05,
+      "loss": 0.3581,
+      "step": 918
+    },
+    {
+      "epoch": 1.300070771408351,
+      "grad_norm": 0.23196626960986785,
+      "learning_rate": 3.147953830010493e-05,
+      "loss": 0.3679,
+      "step": 919
+    },
+    {
+      "epoch": 1.3014861995753715,
+      "grad_norm": 0.21890014293606355,
+      "learning_rate": 3.145330535152151e-05,
+      "loss": 0.3615,
+      "step": 920
+    },
+    {
+      "epoch": 1.3029016277423922,
+      "grad_norm": 0.2529204490877138,
+      "learning_rate": 3.142707240293809e-05,
+      "loss": 0.3786,
+      "step": 921
+    },
+    {
+      "epoch": 1.3043170559094126,
+      "grad_norm": 0.22439298889284814,
+      "learning_rate": 3.140083945435467e-05,
+      "loss": 0.3426,
+      "step": 922
+    },
+    {
+      "epoch": 1.305732484076433,
+      "grad_norm": 0.2204337693599633,
+      "learning_rate": 3.1374606505771246e-05,
+      "loss": 0.3955,
+      "step": 923
+    },
+    {
+      "epoch": 1.3071479122434537,
+      "grad_norm": 0.2598213431903822,
+      "learning_rate": 3.1348373557187826e-05,
+      "loss": 0.3778,
+      "step": 924
+    },
+    {
+      "epoch": 1.3085633404104742,
+      "grad_norm": 0.2042771559540871,
+      "learning_rate": 3.132214060860441e-05,
+      "loss": 0.3554,
+      "step": 925
+    },
+    {
+      "epoch": 1.3099787685774946,
+      "grad_norm": 0.2288031574818764,
+      "learning_rate": 3.129590766002099e-05,
+      "loss": 0.3687,
+      "step": 926
+    },
+    {
+      "epoch": 1.3113941967445153,
+      "grad_norm": 0.21371355059026345,
+      "learning_rate": 3.126967471143757e-05,
+      "loss": 0.3657,
+      "step": 927
+    },
+    {
+      "epoch": 1.3128096249115357,
+      "grad_norm": 0.21709384669920462,
+      "learning_rate": 3.124344176285415e-05,
+      "loss": 0.3764,
+      "step": 928
+    },
+    {
+      "epoch": 1.3142250530785562,
+      "grad_norm": 0.21520446958395642,
+      "learning_rate": 3.121720881427073e-05,
+      "loss": 0.3637,
+      "step": 929
+    },
+    {
+      "epoch": 1.3156404812455769,
+      "grad_norm": 0.20771227297372433,
+      "learning_rate": 3.119097586568731e-05,
+      "loss": 0.3625,
+      "step": 930
+    },
+    {
+      "epoch": 1.3170559094125973,
+      "grad_norm": 0.21476805571570684,
+      "learning_rate": 3.116474291710388e-05,
+      "loss": 0.3681,
+      "step": 931
+    },
+    {
+      "epoch": 1.3184713375796178,
+      "grad_norm": 0.22151467216274381,
+      "learning_rate": 3.113850996852046e-05,
+      "loss": 0.3694,
+      "step": 932
+    },
+    {
+      "epoch": 1.3198867657466384,
+      "grad_norm": 0.22748528401606127,
+      "learning_rate": 3.111227701993704e-05,
+      "loss": 0.3688,
+      "step": 933
+    },
+    {
+      "epoch": 1.3213021939136589,
+      "grad_norm": 0.19688475770704744,
+      "learning_rate": 3.108604407135362e-05,
+      "loss": 0.3722,
+      "step": 934
+    },
+    {
+      "epoch": 1.3227176220806793,
+      "grad_norm": 0.23937633238769224,
+      "learning_rate": 3.10598111227702e-05,
+      "loss": 0.3713,
+      "step": 935
+    },
+    {
+      "epoch": 1.3241330502477,
+      "grad_norm": 0.25045636062936005,
+      "learning_rate": 3.1033578174186776e-05,
+      "loss": 0.3608,
+      "step": 936
+    },
+    {
+      "epoch": 1.3255484784147205,
+      "grad_norm": 0.2226942847550117,
+      "learning_rate": 3.1007345225603356e-05,
+      "loss": 0.3812,
+      "step": 937
+    },
+    {
+      "epoch": 1.326963906581741,
+      "grad_norm": 0.2310460390764128,
+      "learning_rate": 3.098111227701994e-05,
+      "loss": 0.3709,
+      "step": 938
+    },
+    {
+      "epoch": 1.3283793347487616,
+      "grad_norm": 0.24080935704570205,
+      "learning_rate": 3.095487932843652e-05,
+      "loss": 0.3673,
+      "step": 939
+    },
+    {
+      "epoch": 1.329794762915782,
+      "grad_norm": 0.2342103443904471,
+      "learning_rate": 3.09286463798531e-05,
+      "loss": 0.3644,
+      "step": 940
+    },
+    {
+      "epoch": 1.3312101910828025,
+      "grad_norm": 0.2032076310557637,
+      "learning_rate": 3.090241343126968e-05,
+      "loss": 0.3722,
+      "step": 941
+    },
+    {
+      "epoch": 1.3326256192498231,
+      "grad_norm": 0.26965196767838706,
+      "learning_rate": 3.087618048268626e-05,
+      "loss": 0.3774,
+      "step": 942
+    },
+    {
+      "epoch": 1.3340410474168436,
+      "grad_norm": 0.2304105585688344,
+      "learning_rate": 3.084994753410284e-05,
+      "loss": 0.3716,
+      "step": 943
+    },
+    {
+      "epoch": 1.335456475583864,
+      "grad_norm": 0.23259721220450802,
+      "learning_rate": 3.082371458551941e-05,
+      "loss": 0.357,
+      "step": 944
+    },
+    {
+      "epoch": 1.3368719037508847,
+      "grad_norm": 0.2525325695276776,
+      "learning_rate": 3.079748163693599e-05,
+      "loss": 0.3669,
+      "step": 945
+    },
+    {
+      "epoch": 1.3382873319179052,
+      "grad_norm": 0.24111992235444446,
+      "learning_rate": 3.077124868835257e-05,
+      "loss": 0.3725,
+      "step": 946
+    },
+    {
+      "epoch": 1.3397027600849256,
+      "grad_norm": 0.2408990661356157,
+      "learning_rate": 3.074501573976915e-05,
+      "loss": 0.3603,
+      "step": 947
+    },
+    {
+      "epoch": 1.3411181882519463,
+      "grad_norm": 0.23285325781846208,
+      "learning_rate": 3.071878279118573e-05,
+      "loss": 0.3707,
+      "step": 948
+    },
+    {
+      "epoch": 1.3425336164189667,
+      "grad_norm": 0.21092131922844104,
+      "learning_rate": 3.0692549842602306e-05,
+      "loss": 0.3594,
+      "step": 949
+    },
+    {
+      "epoch": 1.3439490445859872,
+      "grad_norm": 0.23308712460130282,
+      "learning_rate": 3.0666316894018887e-05,
+      "loss": 0.3525,
+      "step": 950
+    },
+    {
+      "epoch": 1.3453644727530079,
+      "grad_norm": 0.6803790920448556,
+      "learning_rate": 3.064008394543547e-05,
+      "loss": 0.3603,
+      "step": 951
+    },
+    {
+      "epoch": 1.3467799009200283,
+      "grad_norm": 0.2018088093223601,
+      "learning_rate": 3.061385099685205e-05,
+      "loss": 0.3732,
+      "step": 952
+    },
+    {
+      "epoch": 1.3481953290870488,
+      "grad_norm": 0.2199294319998087,
+      "learning_rate": 3.058761804826863e-05,
+      "loss": 0.3892,
+      "step": 953
+    },
+    {
+      "epoch": 1.3496107572540694,
+      "grad_norm": 0.2327514036438414,
+      "learning_rate": 3.056138509968521e-05,
+      "loss": 0.3491,
+      "step": 954
+    },
+    {
+      "epoch": 1.3510261854210899,
+      "grad_norm": 0.25130993082161057,
+      "learning_rate": 3.053515215110179e-05,
+      "loss": 0.3851,
+      "step": 955
+    },
+    {
+      "epoch": 1.3524416135881103,
+      "grad_norm": 0.21527950192576611,
+      "learning_rate": 3.0508919202518365e-05,
+      "loss": 0.3627,
+      "step": 956
+    },
+    {
+      "epoch": 1.353857041755131,
+      "grad_norm": 0.25980533573369863,
+      "learning_rate": 3.0482686253934945e-05,
+      "loss": 0.3585,
+      "step": 957
+    },
+    {
+      "epoch": 1.3552724699221514,
+      "grad_norm": 0.2257822702451958,
+      "learning_rate": 3.0456453305351522e-05,
+      "loss": 0.3649,
+      "step": 958
+    },
+    {
+      "epoch": 1.356687898089172,
+      "grad_norm": 0.2386397562035341,
+      "learning_rate": 3.0430220356768102e-05,
+      "loss": 0.3745,
+      "step": 959
+    },
+    {
+      "epoch": 1.3581033262561926,
+      "grad_norm": 0.25076804426446075,
+      "learning_rate": 3.040398740818468e-05,
+      "loss": 0.3736,
+      "step": 960
+    },
+    {
+      "epoch": 1.359518754423213,
+      "grad_norm": 0.22408468395551037,
+      "learning_rate": 3.037775445960126e-05,
+      "loss": 0.3732,
+      "step": 961
+    },
+    {
+      "epoch": 1.3609341825902335,
+      "grad_norm": 0.24001877886842707,
+      "learning_rate": 3.035152151101784e-05,
+      "loss": 0.3596,
+      "step": 962
+    },
+    {
+      "epoch": 1.3623496107572541,
+      "grad_norm": 0.2343436942475651,
+      "learning_rate": 3.0325288562434417e-05,
+      "loss": 0.3728,
+      "step": 963
+    },
+    {
+      "epoch": 1.3637650389242746,
+      "grad_norm": 0.2360728276864227,
+      "learning_rate": 3.0299055613850997e-05,
+      "loss": 0.3791,
+      "step": 964
+    },
+    {
+      "epoch": 1.365180467091295,
+      "grad_norm": 0.20414828750484898,
+      "learning_rate": 3.0272822665267574e-05,
+      "loss": 0.3626,
+      "step": 965
+    },
+    {
+      "epoch": 1.3665958952583157,
+      "grad_norm": 0.26034336343202064,
+      "learning_rate": 3.024658971668416e-05,
+      "loss": 0.3653,
+      "step": 966
+    },
+    {
+      "epoch": 1.3680113234253362,
+      "grad_norm": 0.23404953257833314,
+      "learning_rate": 3.0220356768100738e-05,
+      "loss": 0.3721,
+      "step": 967
+    },
+    {
+      "epoch": 1.3694267515923566,
+      "grad_norm": 0.20222020504755647,
+      "learning_rate": 3.0194123819517318e-05,
+      "loss": 0.3823,
+      "step": 968
+    },
+    {
+      "epoch": 1.3708421797593773,
+      "grad_norm": 0.2711360684778627,
+      "learning_rate": 3.0167890870933895e-05,
+      "loss": 0.3923,
+      "step": 969
+    },
+    {
+      "epoch": 1.3722576079263977,
+      "grad_norm": 0.23346782220582887,
+      "learning_rate": 3.0141657922350475e-05,
+      "loss": 0.3675,
+      "step": 970
+    },
+    {
+      "epoch": 1.3736730360934182,
+      "grad_norm": 0.2178239812830853,
+      "learning_rate": 3.0115424973767052e-05,
+      "loss": 0.3632,
+      "step": 971
+    },
+    {
+      "epoch": 1.3750884642604388,
+      "grad_norm": 0.22909818108672744,
+      "learning_rate": 3.0089192025183633e-05,
+      "loss": 0.3699,
+      "step": 972
+    },
+    {
+      "epoch": 1.3765038924274593,
+      "grad_norm": 0.2187912992394647,
+      "learning_rate": 3.0062959076600213e-05,
+      "loss": 0.369,
+      "step": 973
+    },
+    {
+      "epoch": 1.3779193205944797,
+      "grad_norm": 0.23081880330934232,
+      "learning_rate": 3.003672612801679e-05,
+      "loss": 0.3512,
+      "step": 974
+    },
+    {
+      "epoch": 1.3793347487615004,
+      "grad_norm": 0.20996682928697005,
+      "learning_rate": 3.001049317943337e-05,
+      "loss": 0.3747,
+      "step": 975
+    },
+    {
+      "epoch": 1.3807501769285209,
+      "grad_norm": 0.22204530254709298,
+      "learning_rate": 2.9984260230849947e-05,
+      "loss": 0.3465,
+      "step": 976
+    },
+    {
+      "epoch": 1.3821656050955413,
+      "grad_norm": 0.23901856106889752,
+      "learning_rate": 2.9958027282266527e-05,
+      "loss": 0.3574,
+      "step": 977
+    },
+    {
+      "epoch": 1.383581033262562,
+      "grad_norm": 0.22443739525827147,
+      "learning_rate": 2.9931794333683104e-05,
+      "loss": 0.3612,
+      "step": 978
+    },
+    {
+      "epoch": 1.3849964614295824,
+      "grad_norm": 0.2112580984860598,
+      "learning_rate": 2.9905561385099684e-05,
+      "loss": 0.3716,
+      "step": 979
+    },
+    {
+      "epoch": 1.3864118895966029,
+      "grad_norm": 0.25386132304792175,
+      "learning_rate": 2.9879328436516268e-05,
+      "loss": 0.3771,
+      "step": 980
+    },
+    {
+      "epoch": 1.3878273177636236,
+      "grad_norm": 0.22408230073908217,
+      "learning_rate": 2.985309548793285e-05,
+      "loss": 0.3647,
+      "step": 981
+    },
+    {
+      "epoch": 1.389242745930644,
+      "grad_norm": 0.2400530598350793,
+      "learning_rate": 2.9826862539349425e-05,
+      "loss": 0.3738,
+      "step": 982
+    },
+    {
+      "epoch": 1.3906581740976645,
+      "grad_norm": 0.2490902161302705,
+      "learning_rate": 2.9800629590766006e-05,
+      "loss": 0.3831,
+      "step": 983
+    },
+    {
+      "epoch": 1.3920736022646851,
+      "grad_norm": 0.22605134604891303,
+      "learning_rate": 2.9774396642182582e-05,
+      "loss": 0.3637,
+      "step": 984
+    },
+    {
+      "epoch": 1.3934890304317056,
+      "grad_norm": 0.25582374149556475,
+      "learning_rate": 2.9748163693599163e-05,
+      "loss": 0.3604,
+      "step": 985
+    },
+    {
+      "epoch": 1.394904458598726,
+      "grad_norm": 0.2456668245143142,
+      "learning_rate": 2.9721930745015743e-05,
+      "loss": 0.3683,
+      "step": 986
+    },
+    {
+      "epoch": 1.3963198867657467,
+      "grad_norm": 0.22842410348728334,
+      "learning_rate": 2.969569779643232e-05,
+      "loss": 0.3738,
+      "step": 987
+    },
+    {
+      "epoch": 1.3977353149327671,
+      "grad_norm": 0.260567163270158,
+      "learning_rate": 2.96694648478489e-05,
+      "loss": 0.3571,
+      "step": 988
+    },
+    {
+      "epoch": 1.3991507430997876,
+      "grad_norm": 0.22473798072782286,
+      "learning_rate": 2.9643231899265477e-05,
+      "loss": 0.3647,
+      "step": 989
+    },
+    {
+      "epoch": 1.4005661712668083,
+      "grad_norm": 0.22814357369487426,
+      "learning_rate": 2.9616998950682057e-05,
+      "loss": 0.3507,
+      "step": 990
+    },
+    {
+      "epoch": 1.4019815994338287,
+      "grad_norm": 0.2442911897850581,
+      "learning_rate": 2.9590766002098634e-05,
+      "loss": 0.362,
+      "step": 991
+    },
+    {
+      "epoch": 1.4033970276008492,
+      "grad_norm": 0.24090280154357516,
+      "learning_rate": 2.9564533053515215e-05,
+      "loss": 0.3708,
+      "step": 992
+    },
+    {
+      "epoch": 1.4048124557678698,
+      "grad_norm": 0.20438235888587558,
+      "learning_rate": 2.9538300104931795e-05,
+      "loss": 0.3672,
+      "step": 993
+    },
+    {
+      "epoch": 1.4062278839348903,
+      "grad_norm": 0.2434081473944459,
+      "learning_rate": 2.951206715634838e-05,
+      "loss": 0.3742,
+      "step": 994
+    },
+    {
+      "epoch": 1.4076433121019107,
+      "grad_norm": 0.20574095461819145,
+      "learning_rate": 2.9485834207764955e-05,
+      "loss": 0.3537,
+      "step": 995
+    },
+    {
+      "epoch": 1.4090587402689314,
+      "grad_norm": 0.21182339581130577,
+      "learning_rate": 2.9459601259181536e-05,
+      "loss": 0.3565,
+      "step": 996
+    },
+    {
+      "epoch": 1.4104741684359519,
+      "grad_norm": 0.24332941408128245,
+      "learning_rate": 2.9433368310598113e-05,
+      "loss": 0.3788,
+      "step": 997
+    },
+    {
+      "epoch": 1.4118895966029723,
+      "grad_norm": 0.23924818096665304,
+      "learning_rate": 2.9407135362014693e-05,
+      "loss": 0.3768,
+      "step": 998
+    },
+    {
+      "epoch": 1.413305024769993,
+      "grad_norm": 0.2571126671988022,
+      "learning_rate": 2.9380902413431273e-05,
+      "loss": 0.3558,
+      "step": 999
+    },
+    {
+      "epoch": 1.4147204529370134,
+      "grad_norm": 0.23637822861459193,
+      "learning_rate": 2.935466946484785e-05,
+      "loss": 0.3672,
+      "step": 1000
+    },
+    {
+      "epoch": 1.4161358811040339,
+      "grad_norm": 0.2447175247640114,
+      "learning_rate": 2.932843651626443e-05,
+      "loss": 0.3732,
+      "step": 1001
+    },
+    {
+      "epoch": 1.4175513092710545,
+      "grad_norm": 0.24802730208698692,
+      "learning_rate": 2.9302203567681007e-05,
+      "loss": 0.3414,
+      "step": 1002
+    },
+    {
+      "epoch": 1.418966737438075,
+      "grad_norm": 0.20554775024342742,
+      "learning_rate": 2.9275970619097587e-05,
+      "loss": 0.3497,
+      "step": 1003
+    },
+    {
+      "epoch": 1.4203821656050954,
+      "grad_norm": 0.23003987528232994,
+      "learning_rate": 2.9249737670514164e-05,
+      "loss": 0.3691,
+      "step": 1004
+    },
+    {
+      "epoch": 1.4217975937721161,
+      "grad_norm": 0.23311796810883165,
+      "learning_rate": 2.9223504721930745e-05,
+      "loss": 0.3815,
+      "step": 1005
+    },
+    {
+      "epoch": 1.4232130219391366,
+      "grad_norm": 0.2562837129069495,
+      "learning_rate": 2.9197271773347325e-05,
+      "loss": 0.3864,
+      "step": 1006
+    },
+    {
+      "epoch": 1.424628450106157,
+      "grad_norm": 0.22296198226904215,
+      "learning_rate": 2.9171038824763902e-05,
+      "loss": 0.3701,
+      "step": 1007
+    },
+    {
+      "epoch": 1.4260438782731777,
+      "grad_norm": 0.23142877036513443,
+      "learning_rate": 2.9144805876180486e-05,
+      "loss": 0.3876,
+      "step": 1008
+    },
+    {
+      "epoch": 1.4274593064401981,
+      "grad_norm": 0.21364474385565857,
+      "learning_rate": 2.9118572927597066e-05,
+      "loss": 0.3583,
+      "step": 1009
+    },
+    {
+      "epoch": 1.4288747346072186,
+      "grad_norm": 0.23181334641384221,
+      "learning_rate": 2.9092339979013643e-05,
+      "loss": 0.3704,
+      "step": 1010
+    },
+    {
+      "epoch": 1.4302901627742393,
+      "grad_norm": 0.2193168659273958,
+      "learning_rate": 2.9066107030430223e-05,
+      "loss": 0.3671,
+      "step": 1011
+    },
+    {
+      "epoch": 1.4317055909412597,
+      "grad_norm": 0.20782991741805573,
+      "learning_rate": 2.9039874081846803e-05,
+      "loss": 0.3642,
+      "step": 1012
+    },
+    {
+      "epoch": 1.4331210191082802,
+      "grad_norm": 0.23108283829844503,
+      "learning_rate": 2.901364113326338e-05,
+      "loss": 0.3781,
+      "step": 1013
+    },
+    {
+      "epoch": 1.4345364472753008,
+      "grad_norm": 0.22638433711160158,
+      "learning_rate": 2.898740818467996e-05,
+      "loss": 0.3788,
+      "step": 1014
+    },
+    {
+      "epoch": 1.4359518754423213,
+      "grad_norm": 0.23004131381196977,
+      "learning_rate": 2.8961175236096537e-05,
+      "loss": 0.375,
+      "step": 1015
+    },
+    {
+      "epoch": 1.4373673036093417,
+      "grad_norm": 0.2405447495265062,
+      "learning_rate": 2.8934942287513118e-05,
+      "loss": 0.3737,
+      "step": 1016
+    },
+    {
+      "epoch": 1.4387827317763624,
+      "grad_norm": 0.24058873760567664,
+      "learning_rate": 2.8908709338929695e-05,
+      "loss": 0.3683,
+      "step": 1017
+    },
+    {
+      "epoch": 1.4401981599433828,
+      "grad_norm": 0.20937837279181093,
+      "learning_rate": 2.8882476390346275e-05,
+      "loss": 0.3727,
+      "step": 1018
+    },
+    {
+      "epoch": 1.4416135881104033,
+      "grad_norm": 0.23496697162074306,
+      "learning_rate": 2.8856243441762855e-05,
+      "loss": 0.3765,
+      "step": 1019
+    },
+    {
+      "epoch": 1.443029016277424,
+      "grad_norm": 0.22718655907664106,
+      "learning_rate": 2.8830010493179432e-05,
+      "loss": 0.3544,
+      "step": 1020
+    },
+    {
+      "epoch": 1.4444444444444444,
+      "grad_norm": 0.20884300023114202,
+      "learning_rate": 2.8803777544596016e-05,
+      "loss": 0.3586,
+      "step": 1021
+    },
+    {
+      "epoch": 1.4458598726114649,
+      "grad_norm": 0.25482456658386277,
+      "learning_rate": 2.8777544596012596e-05,
+      "loss": 0.3822,
+      "step": 1022
+    },
+    {
+      "epoch": 1.4472753007784855,
+      "grad_norm": 0.2180908582290951,
+      "learning_rate": 2.8751311647429173e-05,
+      "loss": 0.3618,
+      "step": 1023
+    },
+    {
+      "epoch": 1.448690728945506,
+      "grad_norm": 0.20662718782443495,
+      "learning_rate": 2.8725078698845753e-05,
+      "loss": 0.3782,
+      "step": 1024
+    },
+    {
+      "epoch": 1.4501061571125264,
+      "grad_norm": 0.2348809728633647,
+      "learning_rate": 2.8698845750262333e-05,
+      "loss": 0.363,
+      "step": 1025
+    },
+    {
+      "epoch": 1.451521585279547,
+      "grad_norm": 0.25302998104070096,
+      "learning_rate": 2.867261280167891e-05,
+      "loss": 0.367,
+      "step": 1026
+    },
+    {
+      "epoch": 1.4529370134465676,
+      "grad_norm": 0.2187538653616147,
+      "learning_rate": 2.864637985309549e-05,
+      "loss": 0.3584,
+      "step": 1027
+    },
+    {
+      "epoch": 1.454352441613588,
+      "grad_norm": 0.22053612475806544,
+      "learning_rate": 2.8620146904512068e-05,
+      "loss": 0.3784,
+      "step": 1028
+    },
+    {
+      "epoch": 1.4557678697806087,
+      "grad_norm": 0.2519320002641784,
+      "learning_rate": 2.8593913955928648e-05,
+      "loss": 0.3743,
+      "step": 1029
+    },
+    {
+      "epoch": 1.4571832979476291,
+      "grad_norm": 0.2201988589155671,
+      "learning_rate": 2.8567681007345225e-05,
+      "loss": 0.3676,
+      "step": 1030
+    },
+    {
+      "epoch": 1.4585987261146496,
+      "grad_norm": 0.22204855415179556,
+      "learning_rate": 2.8541448058761805e-05,
+      "loss": 0.3599,
+      "step": 1031
+    },
+    {
+      "epoch": 1.4600141542816703,
+      "grad_norm": 0.21904247423641438,
+      "learning_rate": 2.8515215110178385e-05,
+      "loss": 0.3658,
+      "step": 1032
+    },
+    {
+      "epoch": 1.4614295824486907,
+      "grad_norm": 0.25494283985786187,
+      "learning_rate": 2.8488982161594962e-05,
+      "loss": 0.3512,
+      "step": 1033
+    },
+    {
+      "epoch": 1.4628450106157111,
+      "grad_norm": 0.19494031012127083,
+      "learning_rate": 2.8462749213011542e-05,
+      "loss": 0.3663,
+      "step": 1034
+    },
+    {
+      "epoch": 1.4642604387827318,
+      "grad_norm": 0.22600307482991233,
+      "learning_rate": 2.8436516264428126e-05,
+      "loss": 0.3659,
+      "step": 1035
+    },
+    {
+      "epoch": 1.4656758669497523,
+      "grad_norm": 0.2368082752205335,
+      "learning_rate": 2.8410283315844703e-05,
+      "loss": 0.3531,
+      "step": 1036
+    },
+    {
+      "epoch": 1.4670912951167727,
+      "grad_norm": 0.19113265825566395,
+      "learning_rate": 2.8384050367261283e-05,
+      "loss": 0.3521,
+      "step": 1037
+    },
+    {
+      "epoch": 1.4685067232837934,
+      "grad_norm": 0.24235960233572124,
+      "learning_rate": 2.8357817418677864e-05,
+      "loss": 0.3846,
+      "step": 1038
+    },
+    {
+      "epoch": 1.4699221514508138,
+      "grad_norm": 0.21475512342055572,
+      "learning_rate": 2.833158447009444e-05,
+      "loss": 0.3738,
+      "step": 1039
+    },
+    {
+      "epoch": 1.4713375796178343,
+      "grad_norm": 0.24023920601917217,
+      "learning_rate": 2.830535152151102e-05,
+      "loss": 0.3663,
+      "step": 1040
+    },
+    {
+      "epoch": 1.472753007784855,
+      "grad_norm": 0.2529808662363619,
+      "learning_rate": 2.8279118572927598e-05,
+      "loss": 0.3756,
+      "step": 1041
+    },
+    {
+      "epoch": 1.4741684359518754,
+      "grad_norm": 0.1903618997021652,
+      "learning_rate": 2.8252885624344178e-05,
+      "loss": 0.3564,
+      "step": 1042
+    },
+    {
+      "epoch": 1.4755838641188959,
+      "grad_norm": 0.2358280615340478,
+      "learning_rate": 2.8226652675760755e-05,
+      "loss": 0.3654,
+      "step": 1043
+    },
+    {
+      "epoch": 1.4769992922859165,
+      "grad_norm": 0.26434316703105387,
+      "learning_rate": 2.8200419727177335e-05,
+      "loss": 0.3513,
+      "step": 1044
+    },
+    {
+      "epoch": 1.478414720452937,
+      "grad_norm": 0.2533092842511611,
+      "learning_rate": 2.8174186778593915e-05,
+      "loss": 0.3805,
+      "step": 1045
+    },
+    {
+      "epoch": 1.4798301486199574,
+      "grad_norm": 0.25468139675596396,
+      "learning_rate": 2.8147953830010492e-05,
+      "loss": 0.3549,
+      "step": 1046
+    },
+    {
+      "epoch": 1.481245576786978,
+      "grad_norm": 0.2754205033982748,
+      "learning_rate": 2.8121720881427073e-05,
+      "loss": 0.3652,
+      "step": 1047
+    },
+    {
+      "epoch": 1.4826610049539986,
+      "grad_norm": 0.2885005954030202,
+      "learning_rate": 2.809548793284365e-05,
+      "loss": 0.376,
+      "step": 1048
+    },
+    {
+      "epoch": 1.484076433121019,
+      "grad_norm": 0.19856502383557567,
+      "learning_rate": 2.8069254984260233e-05,
+      "loss": 0.3505,
+      "step": 1049
+    },
+    {
+      "epoch": 1.4854918612880397,
+      "grad_norm": 0.24301586162419772,
+      "learning_rate": 2.8043022035676813e-05,
+      "loss": 0.3528,
+      "step": 1050
+    },
+    {
+      "epoch": 1.4869072894550601,
+      "grad_norm": 0.3249546696287922,
+      "learning_rate": 2.8016789087093394e-05,
+      "loss": 0.3674,
+      "step": 1051
+    },
+    {
+      "epoch": 1.4883227176220806,
+      "grad_norm": 0.19954257697387418,
+      "learning_rate": 2.799055613850997e-05,
+      "loss": 0.3682,
+      "step": 1052
+    },
+    {
+      "epoch": 1.4897381457891012,
+      "grad_norm": 0.27863340166427214,
+      "learning_rate": 2.796432318992655e-05,
+      "loss": 0.3711,
+      "step": 1053
+    },
+    {
+      "epoch": 1.4911535739561217,
+      "grad_norm": 0.24351922549594743,
+      "learning_rate": 2.7938090241343128e-05,
+      "loss": 0.3658,
+      "step": 1054
+    },
+    {
+      "epoch": 1.4925690021231421,
+      "grad_norm": 0.2288387976793292,
+      "learning_rate": 2.7911857292759708e-05,
+      "loss": 0.3537,
+      "step": 1055
+    },
+    {
+      "epoch": 1.4939844302901628,
+      "grad_norm": 0.22026517679773322,
+      "learning_rate": 2.7885624344176285e-05,
+      "loss": 0.3604,
+      "step": 1056
+    },
+    {
+      "epoch": 1.4953998584571833,
+      "grad_norm": 0.23735410105725857,
+      "learning_rate": 2.7859391395592865e-05,
+      "loss": 0.365,
+      "step": 1057
+    },
+    {
+      "epoch": 1.4968152866242037,
+      "grad_norm": 0.21146990067723273,
+      "learning_rate": 2.7833158447009446e-05,
+      "loss": 0.3669,
+      "step": 1058
+    },
+    {
+      "epoch": 1.4982307147912244,
+      "grad_norm": 0.2144299492716616,
+      "learning_rate": 2.7806925498426022e-05,
+      "loss": 0.3437,
+      "step": 1059
+    },
+    {
+      "epoch": 1.4996461429582448,
+      "grad_norm": 0.21300314069787124,
+      "learning_rate": 2.7780692549842603e-05,
+      "loss": 0.3376,
+      "step": 1060
+    },
+    {
+      "epoch": 1.5010615711252653,
+      "grad_norm": 0.20151751265838325,
+      "learning_rate": 2.775445960125918e-05,
+      "loss": 0.3741,
+      "step": 1061
+    },
+    {
+      "epoch": 1.502476999292286,
+      "grad_norm": 0.21844206486312426,
+      "learning_rate": 2.772822665267576e-05,
+      "loss": 0.3639,
+      "step": 1062
+    },
+    {
+      "epoch": 1.5038924274593064,
+      "grad_norm": 0.20312416602967337,
+      "learning_rate": 2.7701993704092344e-05,
+      "loss": 0.3432,
+      "step": 1063
+    },
+    {
+      "epoch": 1.5053078556263269,
+      "grad_norm": 0.18205707926616682,
+      "learning_rate": 2.7675760755508924e-05,
+      "loss": 0.345,
+      "step": 1064
+    },
+    {
+      "epoch": 1.5067232837933475,
+      "grad_norm": 0.21890121527783746,
+      "learning_rate": 2.76495278069255e-05,
+      "loss": 0.3578,
+      "step": 1065
+    },
+    {
+      "epoch": 1.508138711960368,
+      "grad_norm": 0.20010961881540978,
+      "learning_rate": 2.762329485834208e-05,
+      "loss": 0.3771,
+      "step": 1066
+    },
+    {
+      "epoch": 1.5095541401273884,
+      "grad_norm": 0.23165164400389923,
+      "learning_rate": 2.7597061909758658e-05,
+      "loss": 0.3765,
+      "step": 1067
+    },
+    {
+      "epoch": 1.510969568294409,
+      "grad_norm": 0.19652743167014758,
+      "learning_rate": 2.7570828961175238e-05,
+      "loss": 0.3564,
+      "step": 1068
+    },
+    {
+      "epoch": 1.5123849964614295,
+      "grad_norm": 0.2209830707675103,
+      "learning_rate": 2.7544596012591815e-05,
+      "loss": 0.3612,
+      "step": 1069
+    },
+    {
+      "epoch": 1.51380042462845,
+      "grad_norm": 0.2137239523923151,
+      "learning_rate": 2.7518363064008395e-05,
+      "loss": 0.3525,
+      "step": 1070
+    },
+    {
+      "epoch": 1.5152158527954707,
+      "grad_norm": 0.19768010727047206,
+      "learning_rate": 2.7492130115424976e-05,
+      "loss": 0.37,
+      "step": 1071
+    },
+    {
+      "epoch": 1.5166312809624911,
+      "grad_norm": 0.24691943698866267,
+      "learning_rate": 2.7465897166841553e-05,
+      "loss": 0.3595,
+      "step": 1072
+    },
+    {
+      "epoch": 1.5180467091295116,
+      "grad_norm": 0.2225667560651534,
+      "learning_rate": 2.7439664218258133e-05,
+      "loss": 0.3555,
+      "step": 1073
+    },
+    {
+      "epoch": 1.5194621372965322,
+      "grad_norm": 0.20554886183569684,
+      "learning_rate": 2.741343126967471e-05,
+      "loss": 0.3588,
+      "step": 1074
+    },
+    {
+      "epoch": 1.5208775654635527,
+      "grad_norm": 0.2505762201576205,
+      "learning_rate": 2.738719832109129e-05,
+      "loss": 0.377,
+      "step": 1075
+    },
+    {
+      "epoch": 1.5222929936305731,
+      "grad_norm": 0.19715242913361825,
+      "learning_rate": 2.7360965372507867e-05,
+      "loss": 0.3516,
+      "step": 1076
+    },
+    {
+      "epoch": 1.5237084217975938,
+      "grad_norm": 0.22718238951725306,
+      "learning_rate": 2.7334732423924454e-05,
+      "loss": 0.3594,
+      "step": 1077
+    },
+    {
+      "epoch": 1.5251238499646143,
+      "grad_norm": 0.2279872658425017,
+      "learning_rate": 2.730849947534103e-05,
+      "loss": 0.3738,
+      "step": 1078
+    },
+    {
+      "epoch": 1.5265392781316347,
+      "grad_norm": 0.20948469884476686,
+      "learning_rate": 2.728226652675761e-05,
+      "loss": 0.3682,
+      "step": 1079
+    },
+    {
+      "epoch": 1.5279547062986554,
+      "grad_norm": 0.2463602564574548,
+      "learning_rate": 2.7256033578174188e-05,
+      "loss": 0.3686,
+      "step": 1080
+    },
+    {
+      "epoch": 1.5293701344656758,
+      "grad_norm": 0.21928438259582247,
+      "learning_rate": 2.722980062959077e-05,
+      "loss": 0.3836,
+      "step": 1081
+    },
+    {
+      "epoch": 1.5307855626326963,
+      "grad_norm": 0.20912170994070226,
+      "learning_rate": 2.7203567681007345e-05,
+      "loss": 0.3717,
+      "step": 1082
+    },
+    {
+      "epoch": 1.532200990799717,
+      "grad_norm": 0.20838477992310525,
+      "learning_rate": 2.7177334732423926e-05,
+      "loss": 0.3681,
+      "step": 1083
+    },
+    {
+      "epoch": 1.5336164189667374,
+      "grad_norm": 0.21226189954948013,
+      "learning_rate": 2.7151101783840506e-05,
+      "loss": 0.3806,
+      "step": 1084
+    },
+    {
+      "epoch": 1.5350318471337578,
+      "grad_norm": 0.2090527854066501,
+      "learning_rate": 2.7124868835257083e-05,
+      "loss": 0.3813,
+      "step": 1085
+    },
+    {
+      "epoch": 1.5364472753007785,
+      "grad_norm": 0.20644505102138733,
+      "learning_rate": 2.7098635886673663e-05,
+      "loss": 0.3703,
+      "step": 1086
+    },
+    {
+      "epoch": 1.537862703467799,
+      "grad_norm": 0.22661853567657378,
+      "learning_rate": 2.707240293809024e-05,
+      "loss": 0.363,
+      "step": 1087
+    },
+    {
+      "epoch": 1.5392781316348194,
+      "grad_norm": 0.20511838703888072,
+      "learning_rate": 2.704616998950682e-05,
+      "loss": 0.3785,
+      "step": 1088
+    },
+    {
+      "epoch": 1.54069355980184,
+      "grad_norm": 0.21196080769048828,
+      "learning_rate": 2.7019937040923397e-05,
+      "loss": 0.3579,
+      "step": 1089
+    },
+    {
+      "epoch": 1.5421089879688605,
+      "grad_norm": 0.20735173163276482,
+      "learning_rate": 2.6993704092339977e-05,
+      "loss": 0.358,
+      "step": 1090
+    },
+    {
+      "epoch": 1.543524416135881,
+      "grad_norm": 0.224486876680772,
+      "learning_rate": 2.696747114375656e-05,
+      "loss": 0.3589,
+      "step": 1091
+    },
+    {
+      "epoch": 1.5449398443029017,
+      "grad_norm": 0.2176990110171203,
+      "learning_rate": 2.694123819517314e-05,
+      "loss": 0.3684,
+      "step": 1092
+    },
+    {
+      "epoch": 1.546355272469922,
+      "grad_norm": 0.24338052069509808,
+      "learning_rate": 2.6915005246589718e-05,
+      "loss": 0.3568,
+      "step": 1093
+    },
+    {
+      "epoch": 1.5477707006369426,
+      "grad_norm": 0.2064407637815234,
+      "learning_rate": 2.68887722980063e-05,
+      "loss": 0.36,
+      "step": 1094
+    },
+    {
+      "epoch": 1.5491861288039632,
+      "grad_norm": 0.22650926555396364,
+      "learning_rate": 2.686253934942288e-05,
+      "loss": 0.3454,
+      "step": 1095
+    },
+    {
+      "epoch": 1.5506015569709837,
+      "grad_norm": 0.22872938113961477,
+      "learning_rate": 2.6836306400839456e-05,
+      "loss": 0.3609,
+      "step": 1096
+    },
+    {
+      "epoch": 1.5520169851380041,
+      "grad_norm": 0.20175298543747824,
+      "learning_rate": 2.6810073452256036e-05,
+      "loss": 0.3902,
+      "step": 1097
+    },
+    {
+      "epoch": 1.5534324133050248,
+      "grad_norm": 0.20200444825151267,
+      "learning_rate": 2.6783840503672613e-05,
+      "loss": 0.3524,
+      "step": 1098
+    },
+    {
+      "epoch": 1.5548478414720452,
+      "grad_norm": 0.2043927249238711,
+      "learning_rate": 2.6757607555089193e-05,
+      "loss": 0.375,
+      "step": 1099
+    },
+    {
+      "epoch": 1.5562632696390657,
+      "grad_norm": 0.20839038411254784,
+      "learning_rate": 2.673137460650577e-05,
+      "loss": 0.3572,
+      "step": 1100
+    },
+    {
+      "epoch": 1.5576786978060864,
+      "grad_norm": 0.20193426343055904,
+      "learning_rate": 2.670514165792235e-05,
+      "loss": 0.3592,
+      "step": 1101
+    },
+    {
+      "epoch": 1.5590941259731068,
+      "grad_norm": 0.2181862764076479,
+      "learning_rate": 2.6678908709338927e-05,
+      "loss": 0.3781,
+      "step": 1102
+    },
+    {
+      "epoch": 1.5605095541401273,
+      "grad_norm": 0.23147357207714458,
+      "learning_rate": 2.6652675760755508e-05,
+      "loss": 0.3754,
+      "step": 1103
+    },
+    {
+      "epoch": 1.561924982307148,
+      "grad_norm": 0.21284670672504122,
+      "learning_rate": 2.662644281217209e-05,
+      "loss": 0.3632,
+      "step": 1104
+    },
+    {
+      "epoch": 1.5633404104741684,
+      "grad_norm": 0.22805341924589922,
+      "learning_rate": 2.660020986358867e-05,
+      "loss": 0.3509,
+      "step": 1105
+    },
+    {
+      "epoch": 1.5647558386411888,
+      "grad_norm": 0.2345407386045021,
+      "learning_rate": 2.657397691500525e-05,
+      "loss": 0.372,
+      "step": 1106
+    },
+    {
+      "epoch": 1.5661712668082095,
+      "grad_norm": 0.20720783219676922,
+      "learning_rate": 2.654774396642183e-05,
+      "loss": 0.353,
+      "step": 1107
+    },
+    {
+      "epoch": 1.56758669497523,
+      "grad_norm": 0.21894913401677554,
+      "learning_rate": 2.652151101783841e-05,
+      "loss": 0.3694,
+      "step": 1108
+    },
+    {
+      "epoch": 1.5690021231422504,
+      "grad_norm": 0.22380173004214715,
+      "learning_rate": 2.6495278069254986e-05,
+      "loss": 0.3606,
+      "step": 1109
+    },
+    {
+      "epoch": 1.570417551309271,
+      "grad_norm": 0.20898100258282237,
+      "learning_rate": 2.6469045120671566e-05,
+      "loss": 0.3841,
+      "step": 1110
+    },
+    {
+      "epoch": 1.5718329794762915,
+      "grad_norm": 0.23414724230585565,
+      "learning_rate": 2.6442812172088143e-05,
+      "loss": 0.3566,
+      "step": 1111
+    },
+    {
+      "epoch": 1.573248407643312,
+      "grad_norm": 0.2341494998744966,
+      "learning_rate": 2.6416579223504723e-05,
+      "loss": 0.3853,
+      "step": 1112
+    },
+    {
+      "epoch": 1.5746638358103326,
+      "grad_norm": 0.20981474741627684,
+      "learning_rate": 2.63903462749213e-05,
+      "loss": 0.3728,
+      "step": 1113
+    },
+    {
+      "epoch": 1.576079263977353,
+      "grad_norm": 0.20158378943617117,
+      "learning_rate": 2.636411332633788e-05,
+      "loss": 0.3503,
+      "step": 1114
+    },
+    {
+      "epoch": 1.5774946921443735,
+      "grad_norm": 0.21060938604199136,
+      "learning_rate": 2.6337880377754457e-05,
+      "loss": 0.3367,
+      "step": 1115
+    },
+    {
+      "epoch": 1.5789101203113942,
+      "grad_norm": 0.20811196733887938,
+      "learning_rate": 2.6311647429171038e-05,
+      "loss": 0.3451,
+      "step": 1116
+    },
+    {
+      "epoch": 1.5803255484784147,
+      "grad_norm": 0.2140577294474795,
+      "learning_rate": 2.6285414480587618e-05,
+      "loss": 0.3583,
+      "step": 1117
+    },
+    {
+      "epoch": 1.5817409766454351,
+      "grad_norm": 0.21441477409429785,
+      "learning_rate": 2.6259181532004202e-05,
+      "loss": 0.3668,
+      "step": 1118
+    },
+    {
+      "epoch": 1.5831564048124558,
+      "grad_norm": 0.23804938645493584,
+      "learning_rate": 2.623294858342078e-05,
+      "loss": 0.3669,
+      "step": 1119
+    },
+    {
+      "epoch": 1.5845718329794765,
+      "grad_norm": 0.23920713672808078,
+      "learning_rate": 2.620671563483736e-05,
+      "loss": 0.3487,
+      "step": 1120
+    },
+    {
+      "epoch": 1.5859872611464967,
+      "grad_norm": 0.24404548961406083,
+      "learning_rate": 2.618048268625394e-05,
+      "loss": 0.3703,
+      "step": 1121
+    },
+    {
+      "epoch": 1.5874026893135174,
+      "grad_norm": 0.2002433289115946,
+      "learning_rate": 2.6154249737670516e-05,
+      "loss": 0.3687,
+      "step": 1122
+    },
+    {
+      "epoch": 1.588818117480538,
+      "grad_norm": 0.24372277480931426,
+      "learning_rate": 2.6128016789087096e-05,
+      "loss": 0.3611,
+      "step": 1123
+    },
+    {
+      "epoch": 1.5902335456475583,
+      "grad_norm": 0.26436244382151886,
+      "learning_rate": 2.6101783840503673e-05,
+      "loss": 0.3559,
+      "step": 1124
+    },
+    {
+      "epoch": 1.591648973814579,
+      "grad_norm": 0.21150768737427175,
+      "learning_rate": 2.6075550891920254e-05,
+      "loss": 0.3595,
+      "step": 1125
+    },
+    {
+      "epoch": 1.5930644019815996,
+      "grad_norm": 0.2259403164389297,
+      "learning_rate": 2.604931794333683e-05,
+      "loss": 0.3848,
+      "step": 1126
+    },
+    {
+      "epoch": 1.5944798301486198,
+      "grad_norm": 0.28490702017390696,
+      "learning_rate": 2.602308499475341e-05,
+      "loss": 0.358,
+      "step": 1127
+    },
+    {
+      "epoch": 1.5958952583156405,
+      "grad_norm": 0.22472009084309402,
+      "learning_rate": 2.599685204616999e-05,
+      "loss": 0.3471,
+      "step": 1128
+    },
+    {
+      "epoch": 1.5973106864826612,
+      "grad_norm": 0.2130869303407535,
+      "learning_rate": 2.5970619097586568e-05,
+      "loss": 0.3762,
+      "step": 1129
+    },
+    {
+      "epoch": 1.5987261146496814,
+      "grad_norm": 0.24686052819553847,
+      "learning_rate": 2.5944386149003148e-05,
+      "loss": 0.361,
+      "step": 1130
+    },
+    {
+      "epoch": 1.600141542816702,
+      "grad_norm": 0.6512322260544745,
+      "learning_rate": 2.5918153200419725e-05,
+      "loss": 0.3913,
+      "step": 1131
+    },
+    {
+      "epoch": 1.6015569709837227,
+      "grad_norm": 0.22223853192188883,
+      "learning_rate": 2.589192025183631e-05,
+      "loss": 0.3709,
+      "step": 1132
+    },
+    {
+      "epoch": 1.602972399150743,
+      "grad_norm": 0.20356356200066156,
+      "learning_rate": 2.586568730325289e-05,
+      "loss": 0.3568,
+      "step": 1133
+    },
+    {
+      "epoch": 1.6043878273177636,
+      "grad_norm": 0.21906757130975804,
+      "learning_rate": 2.583945435466947e-05,
+      "loss": 0.3673,
+      "step": 1134
+    },
+    {
+      "epoch": 1.6058032554847843,
+      "grad_norm": 0.21619433178114403,
+      "learning_rate": 2.5813221406086046e-05,
+      "loss": 0.365,
+      "step": 1135
+    },
+    {
+      "epoch": 1.6072186836518045,
+      "grad_norm": 0.20454654351321744,
+      "learning_rate": 2.5786988457502627e-05,
+      "loss": 0.368,
+      "step": 1136
+    },
+    {
+      "epoch": 1.6086341118188252,
+      "grad_norm": 0.2315571747498643,
+      "learning_rate": 2.5760755508919203e-05,
+      "loss": 0.3648,
+      "step": 1137
+    },
+    {
+      "epoch": 1.6100495399858459,
+      "grad_norm": 0.21673882039259132,
+      "learning_rate": 2.5734522560335784e-05,
+      "loss": 0.3628,
+      "step": 1138
+    },
+    {
+      "epoch": 1.611464968152866,
+      "grad_norm": 0.2068740654670679,
+      "learning_rate": 2.570828961175236e-05,
+      "loss": 0.3618,
+      "step": 1139
+    },
+    {
+      "epoch": 1.6128803963198868,
+      "grad_norm": 0.20087506797620958,
+      "learning_rate": 2.568205666316894e-05,
+      "loss": 0.3762,
+      "step": 1140
+    },
+    {
+      "epoch": 1.6142958244869074,
+      "grad_norm": 0.2123128231858163,
+      "learning_rate": 2.565582371458552e-05,
+      "loss": 0.3773,
+      "step": 1141
+    },
+    {
+      "epoch": 1.6157112526539277,
+      "grad_norm": 0.21442235677515326,
+      "learning_rate": 2.5629590766002098e-05,
+      "loss": 0.3777,
+      "step": 1142
+    },
+    {
+      "epoch": 1.6171266808209483,
+      "grad_norm": 0.20873851860300455,
+      "learning_rate": 2.560335781741868e-05,
+      "loss": 0.3619,
+      "step": 1143
+    },
+    {
+      "epoch": 1.618542108987969,
+      "grad_norm": 0.189086589071609,
+      "learning_rate": 2.5577124868835255e-05,
+      "loss": 0.3566,
+      "step": 1144
+    },
+    {
+      "epoch": 1.6199575371549892,
+      "grad_norm": 0.2166760306608455,
+      "learning_rate": 2.5550891920251836e-05,
+      "loss": 0.3627,
+      "step": 1145
+    },
+    {
+      "epoch": 1.62137296532201,
+      "grad_norm": 0.20996874666882176,
+      "learning_rate": 2.552465897166842e-05,
+      "loss": 0.3711,
+      "step": 1146
+    },
+    {
+      "epoch": 1.6227883934890306,
+      "grad_norm": 0.22105885419371793,
+      "learning_rate": 2.5498426023085e-05,
+      "loss": 0.3667,
+      "step": 1147
+    },
+    {
+      "epoch": 1.6242038216560508,
+      "grad_norm": 0.22800684499540488,
+      "learning_rate": 2.5472193074501576e-05,
+      "loss": 0.3739,
+      "step": 1148
+    },
+    {
+      "epoch": 1.6256192498230715,
+      "grad_norm": 0.22698536578933734,
+      "learning_rate": 2.5445960125918157e-05,
+      "loss": 0.36,
+      "step": 1149
+    },
+    {
+      "epoch": 1.6270346779900922,
+      "grad_norm": 0.18825409854030373,
+      "learning_rate": 2.5419727177334734e-05,
+      "loss": 0.3693,
+      "step": 1150
+    },
+    {
+      "epoch": 1.6284501061571124,
+      "grad_norm": 0.2281011630827293,
+      "learning_rate": 2.5393494228751314e-05,
+      "loss": 0.3567,
+      "step": 1151
+    },
+    {
+      "epoch": 1.629865534324133,
+      "grad_norm": 0.19742550616229176,
+      "learning_rate": 2.536726128016789e-05,
+      "loss": 0.3693,
+      "step": 1152
+    },
+    {
+      "epoch": 1.6312809624911537,
+      "grad_norm": 0.20797475786931,
+      "learning_rate": 2.534102833158447e-05,
+      "loss": 0.3692,
+      "step": 1153
+    },
+    {
+      "epoch": 1.632696390658174,
+      "grad_norm": 0.2138619317417531,
+      "learning_rate": 2.531479538300105e-05,
+      "loss": 0.3701,
+      "step": 1154
+    },
+    {
+      "epoch": 1.6341118188251946,
+      "grad_norm": 0.2069543005190604,
+      "learning_rate": 2.5288562434417628e-05,
+      "loss": 0.3705,
+      "step": 1155
+    },
+    {
+      "epoch": 1.6355272469922153,
+      "grad_norm": 0.21017404070821266,
+      "learning_rate": 2.526232948583421e-05,
+      "loss": 0.3569,
+      "step": 1156
+    },
+    {
+      "epoch": 1.6369426751592355,
+      "grad_norm": 0.1991742248894112,
+      "learning_rate": 2.5236096537250785e-05,
+      "loss": 0.3553,
+      "step": 1157
+    },
+    {
+      "epoch": 1.6383581033262562,
+      "grad_norm": 0.21323664699460512,
+      "learning_rate": 2.5209863588667366e-05,
+      "loss": 0.3605,
+      "step": 1158
+    },
+    {
+      "epoch": 1.6397735314932769,
+      "grad_norm": 0.21728754436493367,
+      "learning_rate": 2.5183630640083943e-05,
+      "loss": 0.3561,
+      "step": 1159
+    },
+    {
+      "epoch": 1.641188959660297,
+      "grad_norm": 0.23906888466302859,
+      "learning_rate": 2.515739769150053e-05,
+      "loss": 0.3526,
+      "step": 1160
+    },
+    {
+      "epoch": 1.6426043878273178,
+      "grad_norm": 0.2252437181573515,
+      "learning_rate": 2.5131164742917107e-05,
+      "loss": 0.3748,
+      "step": 1161
+    },
+    {
+      "epoch": 1.6440198159943384,
+      "grad_norm": 0.2701946642824661,
+      "learning_rate": 2.5104931794333687e-05,
+      "loss": 0.3807,
+      "step": 1162
+    },
+    {
+      "epoch": 1.6454352441613587,
+      "grad_norm": 0.24750083676506437,
+      "learning_rate": 2.5078698845750264e-05,
+      "loss": 0.3857,
+      "step": 1163
+    },
+    {
+      "epoch": 1.6468506723283793,
+      "grad_norm": 0.21529557169837596,
+      "learning_rate": 2.5052465897166844e-05,
+      "loss": 0.3665,
+      "step": 1164
+    },
+    {
+      "epoch": 1.6482661004954,
+      "grad_norm": 0.2258206847573629,
+      "learning_rate": 2.502623294858342e-05,
+      "loss": 0.3574,
+      "step": 1165
+    },
+    {
+      "epoch": 1.6496815286624202,
+      "grad_norm": 0.2069826183446378,
+      "learning_rate": 2.5e-05,
+      "loss": 0.3665,
+      "step": 1166
+    },
+    {
+      "epoch": 1.651096956829441,
+      "grad_norm": 0.2169513049293553,
+      "learning_rate": 2.497376705141658e-05,
+      "loss": 0.3732,
+      "step": 1167
+    },
+    {
+      "epoch": 1.6525123849964616,
+      "grad_norm": 0.212906101059109,
+      "learning_rate": 2.494753410283316e-05,
+      "loss": 0.3534,
+      "step": 1168
+    },
+    {
+      "epoch": 1.6539278131634818,
+      "grad_norm": 0.2105570290098452,
+      "learning_rate": 2.492130115424974e-05,
+      "loss": 0.3622,
+      "step": 1169
+    },
+    {
+      "epoch": 1.6553432413305025,
+      "grad_norm": 0.21546636807833525,
+      "learning_rate": 2.489506820566632e-05,
+      "loss": 0.3673,
+      "step": 1170
+    },
+    {
+      "epoch": 1.6567586694975232,
+      "grad_norm": 0.22406045718755155,
+      "learning_rate": 2.48688352570829e-05,
+      "loss": 0.3465,
+      "step": 1171
+    },
+    {
+      "epoch": 1.6581740976645434,
+      "grad_norm": 0.21206566096914026,
+      "learning_rate": 2.4842602308499476e-05,
+      "loss": 0.3701,
+      "step": 1172
+    },
+    {
+      "epoch": 1.659589525831564,
+      "grad_norm": 0.2178940497544651,
+      "learning_rate": 2.4816369359916056e-05,
+      "loss": 0.381,
+      "step": 1173
+    },
+    {
+      "epoch": 1.6610049539985847,
+      "grad_norm": 0.2261416831205509,
+      "learning_rate": 2.4790136411332633e-05,
+      "loss": 0.372,
+      "step": 1174
+    },
+    {
+      "epoch": 1.662420382165605,
+      "grad_norm": 0.214874681558974,
+      "learning_rate": 2.4763903462749214e-05,
+      "loss": 0.3717,
+      "step": 1175
+    },
+    {
+      "epoch": 1.6638358103326256,
+      "grad_norm": 0.2313551183517155,
+      "learning_rate": 2.473767051416579e-05,
+      "loss": 0.3752,
+      "step": 1176
+    },
+    {
+      "epoch": 1.6652512384996463,
+      "grad_norm": 0.22257778337500586,
+      "learning_rate": 2.4711437565582374e-05,
+      "loss": 0.3648,
+      "step": 1177
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 0.2106210720419038,
+      "learning_rate": 2.468520461699895e-05,
+      "loss": 0.3683,
+      "step": 1178
+    },
+    {
+      "epoch": 1.6680820948336872,
+      "grad_norm": 0.20845895875371392,
+      "learning_rate": 2.465897166841553e-05,
+      "loss": 0.3777,
+      "step": 1179
+    },
+    {
+      "epoch": 1.6694975230007079,
+      "grad_norm": 0.21186402961211692,
+      "learning_rate": 2.463273871983211e-05,
+      "loss": 0.3698,
+      "step": 1180
+    },
+    {
+      "epoch": 1.670912951167728,
+      "grad_norm": 0.1959558234904786,
+      "learning_rate": 2.460650577124869e-05,
+      "loss": 0.3511,
+      "step": 1181
+    },
+    {
+      "epoch": 1.6723283793347488,
+      "grad_norm": 0.20690244788270776,
+      "learning_rate": 2.458027282266527e-05,
+      "loss": 0.3678,
+      "step": 1182
+    },
+    {
+      "epoch": 1.6737438075017694,
+      "grad_norm": 0.23862560377417902,
+      "learning_rate": 2.4554039874081846e-05,
+      "loss": 0.3713,
+      "step": 1183
+    },
+    {
+      "epoch": 1.6751592356687897,
+      "grad_norm": 0.19173248146815988,
+      "learning_rate": 2.452780692549843e-05,
+      "loss": 0.3437,
+      "step": 1184
+    },
+    {
+      "epoch": 1.6765746638358103,
+      "grad_norm": 0.21354581946825002,
+      "learning_rate": 2.4501573976915006e-05,
+      "loss": 0.3783,
+      "step": 1185
+    },
+    {
+      "epoch": 1.677990092002831,
+      "grad_norm": 0.21694878267745185,
+      "learning_rate": 2.4475341028331587e-05,
+      "loss": 0.3669,
+      "step": 1186
+    },
+    {
+      "epoch": 1.6794055201698512,
+      "grad_norm": 0.20270673295676334,
+      "learning_rate": 2.4449108079748163e-05,
+      "loss": 0.3713,
+      "step": 1187
+    },
+    {
+      "epoch": 1.680820948336872,
+      "grad_norm": 0.2297035807958727,
+      "learning_rate": 2.4422875131164744e-05,
+      "loss": 0.358,
+      "step": 1188
+    },
+    {
+      "epoch": 1.6822363765038926,
+      "grad_norm": 0.19338680798787286,
+      "learning_rate": 2.4396642182581324e-05,
+      "loss": 0.3638,
+      "step": 1189
+    },
+    {
+      "epoch": 1.6836518046709128,
+      "grad_norm": 0.21068205711888074,
+      "learning_rate": 2.43704092339979e-05,
+      "loss": 0.3821,
+      "step": 1190
+    },
+    {
+      "epoch": 1.6850672328379335,
+      "grad_norm": 0.20778466400564496,
+      "learning_rate": 2.434417628541448e-05,
+      "loss": 0.3555,
+      "step": 1191
+    },
+    {
+      "epoch": 1.6864826610049541,
+      "grad_norm": 0.1966167293258372,
+      "learning_rate": 2.431794333683106e-05,
+      "loss": 0.3566,
+      "step": 1192
+    },
+    {
+      "epoch": 1.6878980891719744,
+      "grad_norm": 0.1961489333151967,
+      "learning_rate": 2.4291710388247642e-05,
+      "loss": 0.3614,
+      "step": 1193
+    },
+    {
+      "epoch": 1.689313517338995,
+      "grad_norm": 0.19471307863244142,
+      "learning_rate": 2.426547743966422e-05,
+      "loss": 0.358,
+      "step": 1194
+    },
+    {
+      "epoch": 1.6907289455060157,
+      "grad_norm": 0.20180964645438249,
+      "learning_rate": 2.42392444910808e-05,
+      "loss": 0.3479,
+      "step": 1195
+    },
+    {
+      "epoch": 1.692144373673036,
+      "grad_norm": 0.19026046834471852,
+      "learning_rate": 2.4213011542497376e-05,
+      "loss": 0.3582,
+      "step": 1196
+    },
+    {
+      "epoch": 1.6935598018400566,
+      "grad_norm": 0.1903543816431227,
+      "learning_rate": 2.4186778593913956e-05,
+      "loss": 0.3709,
+      "step": 1197
+    },
+    {
+      "epoch": 1.6949752300070773,
+      "grad_norm": 0.2092728212533432,
+      "learning_rate": 2.4160545645330536e-05,
+      "loss": 0.3725,
+      "step": 1198
+    },
+    {
+      "epoch": 1.6963906581740975,
+      "grad_norm": 0.21226469419764257,
+      "learning_rate": 2.4134312696747117e-05,
+      "loss": 0.3587,
+      "step": 1199
+    },
+    {
+      "epoch": 1.6978060863411182,
+      "grad_norm": 0.2044049166977702,
+      "learning_rate": 2.4108079748163694e-05,
+      "loss": 0.3591,
+      "step": 1200
+    },
+    {
+      "epoch": 1.6992215145081389,
+      "grad_norm": 0.20048213843328103,
+      "learning_rate": 2.4081846799580274e-05,
+      "loss": 0.3627,
+      "step": 1201
+    },
+    {
+      "epoch": 1.700636942675159,
+      "grad_norm": 0.19580747071870352,
+      "learning_rate": 2.4055613850996854e-05,
+      "loss": 0.3502,
+      "step": 1202
+    },
+    {
+      "epoch": 1.7020523708421798,
+      "grad_norm": 0.20105011429903724,
+      "learning_rate": 2.402938090241343e-05,
+      "loss": 0.3441,
+      "step": 1203
+    },
+    {
+      "epoch": 1.7034677990092004,
+      "grad_norm": 0.2018537985655408,
+      "learning_rate": 2.400314795383001e-05,
+      "loss": 0.3556,
+      "step": 1204
+    },
+    {
+      "epoch": 1.7048832271762207,
+      "grad_norm": 0.2044453087083259,
+      "learning_rate": 2.397691500524659e-05,
+      "loss": 0.3558,
+      "step": 1205
+    },
+    {
+      "epoch": 1.7062986553432413,
+      "grad_norm": 0.22086503599296572,
+      "learning_rate": 2.3950682056663172e-05,
+      "loss": 0.3602,
+      "step": 1206
+    },
+    {
+      "epoch": 1.707714083510262,
+      "grad_norm": 0.19641136659257846,
+      "learning_rate": 2.392444910807975e-05,
+      "loss": 0.3743,
+      "step": 1207
+    },
+    {
+      "epoch": 1.7091295116772822,
+      "grad_norm": 0.20260136418832903,
+      "learning_rate": 2.389821615949633e-05,
+      "loss": 0.378,
+      "step": 1208
+    },
+    {
+      "epoch": 1.710544939844303,
+      "grad_norm": 0.9136989290123337,
+      "learning_rate": 2.3871983210912906e-05,
+      "loss": 0.3782,
+      "step": 1209
+    },
+    {
+      "epoch": 1.7119603680113236,
+      "grad_norm": 0.22703872536118613,
+      "learning_rate": 2.3845750262329486e-05,
+      "loss": 0.352,
+      "step": 1210
+    },
+    {
+      "epoch": 1.7133757961783438,
+      "grad_norm": 0.28524183327470354,
+      "learning_rate": 2.3819517313746067e-05,
+      "loss": 0.3421,
+      "step": 1211
+    },
+    {
+      "epoch": 1.7147912243453645,
+      "grad_norm": 0.20985868088696918,
+      "learning_rate": 2.3793284365162647e-05,
+      "loss": 0.3641,
+      "step": 1212
+    },
+    {
+      "epoch": 1.7162066525123851,
+      "grad_norm": 0.2096411624630168,
+      "learning_rate": 2.3767051416579224e-05,
+      "loss": 0.3627,
+      "step": 1213
+    },
+    {
+      "epoch": 1.7176220806794054,
+      "grad_norm": 0.2289168020447138,
+      "learning_rate": 2.3740818467995804e-05,
+      "loss": 0.3701,
+      "step": 1214
+    },
+    {
+      "epoch": 1.719037508846426,
+      "grad_norm": 0.23972621162129334,
+      "learning_rate": 2.3714585519412384e-05,
+      "loss": 0.3528,
+      "step": 1215
+    },
+    {
+      "epoch": 1.7204529370134467,
+      "grad_norm": 0.2047453070334584,
+      "learning_rate": 2.368835257082896e-05,
+      "loss": 0.3594,
+      "step": 1216
+    },
+    {
+      "epoch": 1.721868365180467,
+      "grad_norm": 0.2080153214325207,
+      "learning_rate": 2.366211962224554e-05,
+      "loss": 0.3766,
+      "step": 1217
+    },
+    {
+      "epoch": 1.7232837933474876,
+      "grad_norm": 0.25294177188787764,
+      "learning_rate": 2.3635886673662122e-05,
+      "loss": 0.3601,
+      "step": 1218
+    },
+    {
+      "epoch": 1.7246992215145083,
+      "grad_norm": 0.20454836431609313,
+      "learning_rate": 2.3609653725078702e-05,
+      "loss": 0.3588,
+      "step": 1219
+    },
+    {
+      "epoch": 1.7261146496815285,
+      "grad_norm": 0.19248036034295865,
+      "learning_rate": 2.358342077649528e-05,
+      "loss": 0.3663,
+      "step": 1220
+    },
+    {
+      "epoch": 1.7275300778485492,
+      "grad_norm": 0.24928635062323073,
+      "learning_rate": 2.355718782791186e-05,
+      "loss": 0.3543,
+      "step": 1221
+    },
+    {
+      "epoch": 1.7289455060155698,
+      "grad_norm": 0.22813261657970457,
+      "learning_rate": 2.3530954879328436e-05,
+      "loss": 0.3837,
+      "step": 1222
+    },
+    {
+      "epoch": 1.73036093418259,
+      "grad_norm": 0.21568958693996085,
+      "learning_rate": 2.3504721930745016e-05,
+      "loss": 0.3725,
+      "step": 1223
+    },
+    {
+      "epoch": 1.7317763623496107,
+      "grad_norm": 0.23325960207870589,
+      "learning_rate": 2.3478488982161593e-05,
+      "loss": 0.3717,
+      "step": 1224
+    },
+    {
+      "epoch": 1.7331917905166314,
+      "grad_norm": 0.23918959953390725,
+      "learning_rate": 2.3452256033578177e-05,
+      "loss": 0.3698,
+      "step": 1225
+    },
+    {
+      "epoch": 1.7346072186836516,
+      "grad_norm": 0.20037216180436881,
+      "learning_rate": 2.3426023084994754e-05,
+      "loss": 0.3628,
+      "step": 1226
+    },
+    {
+      "epoch": 1.7360226468506723,
+      "grad_norm": 0.22088144016706737,
+      "learning_rate": 2.3399790136411334e-05,
+      "loss": 0.3686,
+      "step": 1227
+    },
+    {
+      "epoch": 1.737438075017693,
+      "grad_norm": 0.22951569956332904,
+      "learning_rate": 2.3373557187827915e-05,
+      "loss": 0.3777,
+      "step": 1228
+    },
+    {
+      "epoch": 1.7388535031847132,
+      "grad_norm": 0.20888138755141997,
+      "learning_rate": 2.334732423924449e-05,
+      "loss": 0.3697,
+      "step": 1229
+    },
+    {
+      "epoch": 1.7402689313517339,
+      "grad_norm": 0.20044970774934878,
+      "learning_rate": 2.332109129066107e-05,
+      "loss": 0.3624,
+      "step": 1230
+    },
+    {
+      "epoch": 1.7416843595187546,
+      "grad_norm": 0.2119562447197313,
+      "learning_rate": 2.329485834207765e-05,
+      "loss": 0.3649,
+      "step": 1231
+    },
+    {
+      "epoch": 1.7430997876857748,
+      "grad_norm": 0.2122586202791008,
+      "learning_rate": 2.3268625393494232e-05,
+      "loss": 0.3707,
+      "step": 1232
+    },
+    {
+      "epoch": 1.7445152158527955,
+      "grad_norm": 0.21687015008306162,
+      "learning_rate": 2.324239244491081e-05,
+      "loss": 0.3556,
+      "step": 1233
+    },
+    {
+      "epoch": 1.7459306440198161,
+      "grad_norm": 0.2108980964868219,
+      "learning_rate": 2.321615949632739e-05,
+      "loss": 0.3748,
+      "step": 1234
+    },
+    {
+      "epoch": 1.7473460721868364,
+      "grad_norm": 1.0913112858035567,
+      "learning_rate": 2.3189926547743966e-05,
+      "loss": 0.3518,
+      "step": 1235
+    },
+    {
+      "epoch": 1.748761500353857,
+      "grad_norm": 0.23288980005086307,
+      "learning_rate": 2.3163693599160547e-05,
+      "loss": 0.3638,
+      "step": 1236
+    },
+    {
+      "epoch": 1.7501769285208777,
+      "grad_norm": 0.21201148675782364,
+      "learning_rate": 2.3137460650577124e-05,
+      "loss": 0.3502,
+      "step": 1237
+    },
+    {
+      "epoch": 1.7515923566878981,
+      "grad_norm": 0.20520118643995194,
+      "learning_rate": 2.3111227701993704e-05,
+      "loss": 0.3695,
+      "step": 1238
+    },
+    {
+      "epoch": 1.7530077848549186,
+      "grad_norm": 0.23805035582970135,
+      "learning_rate": 2.3084994753410284e-05,
+      "loss": 0.3885,
+      "step": 1239
+    },
+    {
+      "epoch": 1.7544232130219393,
+      "grad_norm": 0.23750328263631534,
+      "learning_rate": 2.3058761804826864e-05,
+      "loss": 0.3617,
+      "step": 1240
+    },
+    {
+      "epoch": 1.7558386411889597,
+      "grad_norm": 0.2146087119167009,
+      "learning_rate": 2.3032528856243445e-05,
+      "loss": 0.364,
+      "step": 1241
+    },
+    {
+      "epoch": 1.7572540693559802,
+      "grad_norm": 0.20739386731977583,
+      "learning_rate": 2.300629590766002e-05,
+      "loss": 0.3537,
+      "step": 1242
+    },
+    {
+      "epoch": 1.7586694975230008,
+      "grad_norm": 0.2633167341683389,
+      "learning_rate": 2.2980062959076602e-05,
+      "loss": 0.3695,
+      "step": 1243
+    },
+    {
+      "epoch": 1.7600849256900213,
+      "grad_norm": 0.2121191502498069,
+      "learning_rate": 2.295383001049318e-05,
+      "loss": 0.3589,
+      "step": 1244
+    },
+    {
+      "epoch": 1.7615003538570417,
+      "grad_norm": 0.19799416955148808,
+      "learning_rate": 2.292759706190976e-05,
+      "loss": 0.3731,
+      "step": 1245
+    },
+    {
+      "epoch": 1.7629157820240624,
+      "grad_norm": 0.23500520602126604,
+      "learning_rate": 2.290136411332634e-05,
+      "loss": 0.3698,
+      "step": 1246
+    },
+    {
+      "epoch": 1.7643312101910829,
+      "grad_norm": 0.22391141058695835,
+      "learning_rate": 2.287513116474292e-05,
+      "loss": 0.3513,
+      "step": 1247
+    },
+    {
+      "epoch": 1.7657466383581033,
+      "grad_norm": 0.21100652384119065,
+      "learning_rate": 2.2848898216159496e-05,
+      "loss": 0.3626,
+      "step": 1248
+    },
+    {
+      "epoch": 1.767162066525124,
+      "grad_norm": 0.21100168937294683,
+      "learning_rate": 2.2822665267576077e-05,
+      "loss": 0.3664,
+      "step": 1249
+    },
+    {
+      "epoch": 1.7685774946921444,
+      "grad_norm": 0.22302360086013803,
+      "learning_rate": 2.2796432318992657e-05,
+      "loss": 0.3742,
+      "step": 1250
+    },
+    {
+      "epoch": 1.7699929228591649,
+      "grad_norm": 0.22908994489609763,
+      "learning_rate": 2.2770199370409234e-05,
+      "loss": 0.3694,
+      "step": 1251
+    },
+    {
+      "epoch": 1.7714083510261855,
+      "grad_norm": 0.2316123139862592,
+      "learning_rate": 2.2743966421825814e-05,
+      "loss": 0.3533,
+      "step": 1252
+    },
+    {
+      "epoch": 1.772823779193206,
+      "grad_norm": 0.21261889885287916,
+      "learning_rate": 2.2717733473242395e-05,
+      "loss": 0.334,
+      "step": 1253
+    },
+    {
+      "epoch": 1.7742392073602264,
+      "grad_norm": 0.2423707099137475,
+      "learning_rate": 2.2691500524658975e-05,
+      "loss": 0.3604,
+      "step": 1254
+    },
+    {
+      "epoch": 1.7756546355272471,
+      "grad_norm": 0.20433184791489278,
+      "learning_rate": 2.2665267576075552e-05,
+      "loss": 0.3651,
+      "step": 1255
+    },
+    {
+      "epoch": 1.7770700636942676,
+      "grad_norm": 0.22088648020444507,
+      "learning_rate": 2.2639034627492132e-05,
+      "loss": 0.3911,
+      "step": 1256
+    },
+    {
+      "epoch": 1.778485491861288,
+      "grad_norm": 0.255384104535801,
+      "learning_rate": 2.261280167890871e-05,
+      "loss": 0.3359,
+      "step": 1257
+    },
+    {
+      "epoch": 1.7799009200283087,
+      "grad_norm": 0.21730056521809238,
+      "learning_rate": 2.258656873032529e-05,
+      "loss": 0.3452,
+      "step": 1258
+    },
+    {
+      "epoch": 1.7813163481953291,
+      "grad_norm": 0.23095327884837383,
+      "learning_rate": 2.2560335781741866e-05,
+      "loss": 0.379,
+      "step": 1259
+    },
+    {
+      "epoch": 1.7827317763623496,
+      "grad_norm": 0.23134452811986084,
+      "learning_rate": 2.253410283315845e-05,
+      "loss": 0.3871,
+      "step": 1260
+    },
+    {
+      "epoch": 1.7841472045293703,
+      "grad_norm": 0.21404094506623178,
+      "learning_rate": 2.2507869884575027e-05,
+      "loss": 0.3555,
+      "step": 1261
+    },
+    {
+      "epoch": 1.7855626326963907,
+      "grad_norm": 0.3466551124045913,
+      "learning_rate": 2.2481636935991607e-05,
+      "loss": 0.3697,
+      "step": 1262
+    },
+    {
+      "epoch": 1.7869780608634112,
+      "grad_norm": 0.20695439752442338,
+      "learning_rate": 2.2455403987408187e-05,
+      "loss": 0.3656,
+      "step": 1263
+    },
+    {
+      "epoch": 1.7883934890304318,
+      "grad_norm": 0.20396098860354842,
+      "learning_rate": 2.2429171038824764e-05,
+      "loss": 0.3884,
+      "step": 1264
+    },
+    {
+      "epoch": 1.7898089171974523,
+      "grad_norm": 0.25635116226554294,
+      "learning_rate": 2.2402938090241344e-05,
+      "loss": 0.3817,
+      "step": 1265
+    },
+    {
+      "epoch": 1.7912243453644727,
+      "grad_norm": 0.2115951041264189,
+      "learning_rate": 2.237670514165792e-05,
+      "loss": 0.3685,
+      "step": 1266
+    },
+    {
+      "epoch": 1.7926397735314934,
+      "grad_norm": 0.20319279122789133,
+      "learning_rate": 2.2350472193074505e-05,
+      "loss": 0.369,
+      "step": 1267
+    },
+    {
+      "epoch": 1.7940552016985138,
+      "grad_norm": 0.2274827066830239,
+      "learning_rate": 2.2324239244491082e-05,
+      "loss": 0.3468,
+      "step": 1268
+    },
+    {
+      "epoch": 1.7954706298655343,
+      "grad_norm": 0.2046829944609587,
+      "learning_rate": 2.2298006295907662e-05,
+      "loss": 0.3668,
+      "step": 1269
+    },
+    {
+      "epoch": 1.796886058032555,
+      "grad_norm": 0.19991518384288134,
+      "learning_rate": 2.227177334732424e-05,
+      "loss": 0.3504,
+      "step": 1270
+    },
+    {
+      "epoch": 1.7983014861995754,
+      "grad_norm": 0.22501211515746133,
+      "learning_rate": 2.224554039874082e-05,
+      "loss": 0.358,
+      "step": 1271
+    },
+    {
+      "epoch": 1.7997169143665959,
+      "grad_norm": 0.21863795747539036,
+      "learning_rate": 2.2219307450157396e-05,
+      "loss": 0.3423,
+      "step": 1272
+    },
+    {
+      "epoch": 1.8011323425336165,
+      "grad_norm": 0.19463870238231556,
+      "learning_rate": 2.2193074501573977e-05,
+      "loss": 0.38,
+      "step": 1273
+    },
+    {
+      "epoch": 1.802547770700637,
+      "grad_norm": 0.20625397008836294,
+      "learning_rate": 2.2166841552990557e-05,
+      "loss": 0.363,
+      "step": 1274
+    },
+    {
+      "epoch": 1.8039631988676574,
+      "grad_norm": 0.22540358011888795,
+      "learning_rate": 2.2140608604407137e-05,
+      "loss": 0.3766,
+      "step": 1275
+    },
+    {
+      "epoch": 1.805378627034678,
+      "grad_norm": 0.2056579233410256,
+      "learning_rate": 2.2114375655823717e-05,
+      "loss": 0.3727,
+      "step": 1276
+    },
+    {
+      "epoch": 1.8067940552016986,
+      "grad_norm": 0.23614851830523279,
+      "learning_rate": 2.2088142707240294e-05,
+      "loss": 0.3598,
+      "step": 1277
+    },
+    {
+      "epoch": 1.808209483368719,
+      "grad_norm": 0.20523123531066292,
+      "learning_rate": 2.2061909758656875e-05,
+      "loss": 0.3718,
+      "step": 1278
+    },
+    {
+      "epoch": 1.8096249115357397,
+      "grad_norm": 0.2369288444316511,
+      "learning_rate": 2.203567681007345e-05,
+      "loss": 0.3814,
+      "step": 1279
+    },
+    {
+      "epoch": 1.8110403397027601,
+      "grad_norm": 0.21759488565678942,
+      "learning_rate": 2.2009443861490035e-05,
+      "loss": 0.3825,
+      "step": 1280
+    },
+    {
+      "epoch": 1.8124557678697806,
+      "grad_norm": 0.20459769459772856,
+      "learning_rate": 2.1983210912906612e-05,
+      "loss": 0.3861,
+      "step": 1281
+    },
+    {
+      "epoch": 1.8138711960368012,
+      "grad_norm": 0.23408130549169665,
+      "learning_rate": 2.1956977964323192e-05,
+      "loss": 0.365,
+      "step": 1282
+    },
+    {
+      "epoch": 1.8152866242038217,
+      "grad_norm": 0.19418467624251456,
+      "learning_rate": 2.193074501573977e-05,
+      "loss": 0.3412,
+      "step": 1283
+    },
+    {
+      "epoch": 1.8167020523708421,
+      "grad_norm": 0.19656100787624234,
+      "learning_rate": 2.190451206715635e-05,
+      "loss": 0.3695,
+      "step": 1284
+    },
+    {
+      "epoch": 1.8181174805378628,
+      "grad_norm": 0.20293810051390856,
+      "learning_rate": 2.1878279118572926e-05,
+      "loss": 0.3637,
+      "step": 1285
+    },
+    {
+      "epoch": 1.8195329087048833,
+      "grad_norm": 0.23094731602569574,
+      "learning_rate": 2.1852046169989507e-05,
+      "loss": 0.3631,
+      "step": 1286
+    },
+    {
+      "epoch": 1.8209483368719037,
+      "grad_norm": 0.23025613270928239,
+      "learning_rate": 2.1825813221406087e-05,
+      "loss": 0.3674,
+      "step": 1287
+    },
+    {
+      "epoch": 1.8223637650389244,
+      "grad_norm": 0.2226632950521876,
+      "learning_rate": 2.1799580272822667e-05,
+      "loss": 0.3743,
+      "step": 1288
+    },
+    {
+      "epoch": 1.8237791932059448,
+      "grad_norm": 0.21775390538899728,
+      "learning_rate": 2.1773347324239248e-05,
+      "loss": 0.3566,
+      "step": 1289
+    },
+    {
+      "epoch": 1.8251946213729653,
+      "grad_norm": 0.26983059901854184,
+      "learning_rate": 2.1747114375655824e-05,
+      "loss": 0.347,
+      "step": 1290
+    },
+    {
+      "epoch": 1.826610049539986,
+      "grad_norm": 0.1987013924047395,
+      "learning_rate": 2.1720881427072405e-05,
+      "loss": 0.3558,
+      "step": 1291
+    },
+    {
+      "epoch": 1.8280254777070064,
+      "grad_norm": 0.2051541900220511,
+      "learning_rate": 2.169464847848898e-05,
+      "loss": 0.3711,
+      "step": 1292
+    },
+    {
+      "epoch": 1.8294409058740269,
+      "grad_norm": 0.25572075355083324,
+      "learning_rate": 2.1668415529905562e-05,
+      "loss": 0.3536,
+      "step": 1293
+    },
+    {
+      "epoch": 1.8308563340410475,
+      "grad_norm": 0.19893210566212471,
+      "learning_rate": 2.1642182581322142e-05,
+      "loss": 0.3611,
+      "step": 1294
+    },
+    {
+      "epoch": 1.832271762208068,
+      "grad_norm": 0.1994034930117717,
+      "learning_rate": 2.1615949632738722e-05,
+      "loss": 0.3825,
+      "step": 1295
+    },
+    {
+      "epoch": 1.8336871903750884,
+      "grad_norm": 0.20061911877028749,
+      "learning_rate": 2.15897166841553e-05,
+      "loss": 0.3612,
+      "step": 1296
+    },
+    {
+      "epoch": 1.835102618542109,
+      "grad_norm": 0.20471315318320596,
+      "learning_rate": 2.156348373557188e-05,
+      "loss": 0.3477,
+      "step": 1297
+    },
+    {
+      "epoch": 1.8365180467091295,
+      "grad_norm": 0.19983725111724615,
+      "learning_rate": 2.1537250786988457e-05,
+      "loss": 0.3546,
+      "step": 1298
+    },
+    {
+      "epoch": 1.83793347487615,
+      "grad_norm": 0.19499829531879675,
+      "learning_rate": 2.1511017838405037e-05,
+      "loss": 0.3473,
+      "step": 1299
+    },
+    {
+      "epoch": 1.8393489030431707,
+      "grad_norm": 0.2055781282636102,
+      "learning_rate": 2.1484784889821617e-05,
+      "loss": 0.3484,
+      "step": 1300
+    },
+    {
+      "epoch": 1.8407643312101911,
+      "grad_norm": 0.22577798017905634,
+      "learning_rate": 2.1458551941238197e-05,
+      "loss": 0.3647,
+      "step": 1301
+    },
+    {
+      "epoch": 1.8421797593772116,
+      "grad_norm": 0.19401973361309152,
+      "learning_rate": 2.1432318992654778e-05,
+      "loss": 0.369,
+      "step": 1302
+    },
+    {
+      "epoch": 1.8435951875442322,
+      "grad_norm": 0.23586191495152506,
+      "learning_rate": 2.1406086044071355e-05,
+      "loss": 0.3633,
+      "step": 1303
+    },
+    {
+      "epoch": 1.8450106157112527,
+      "grad_norm": 0.2137537597890475,
+      "learning_rate": 2.1379853095487935e-05,
+      "loss": 0.3564,
+      "step": 1304
+    },
+    {
+      "epoch": 1.8464260438782731,
+      "grad_norm": 0.19896395757117963,
+      "learning_rate": 2.1353620146904512e-05,
+      "loss": 0.3557,
+      "step": 1305
+    },
+    {
+      "epoch": 1.8478414720452938,
+      "grad_norm": 0.20906941348976985,
+      "learning_rate": 2.1327387198321092e-05,
+      "loss": 0.3596,
+      "step": 1306
+    },
+    {
+      "epoch": 1.8492569002123143,
+      "grad_norm": 0.22159183190357726,
+      "learning_rate": 2.130115424973767e-05,
+      "loss": 0.3562,
+      "step": 1307
+    },
+    {
+      "epoch": 1.8506723283793347,
+      "grad_norm": 0.18295249145000686,
+      "learning_rate": 2.1274921301154253e-05,
+      "loss": 0.3551,
+      "step": 1308
+    },
+    {
+      "epoch": 1.8520877565463554,
+      "grad_norm": 0.21844066556104139,
+      "learning_rate": 2.124868835257083e-05,
+      "loss": 0.3675,
+      "step": 1309
+    },
+    {
+      "epoch": 1.8535031847133758,
+      "grad_norm": 0.1957473561795513,
+      "learning_rate": 2.122245540398741e-05,
+      "loss": 0.3383,
+      "step": 1310
+    },
+    {
+      "epoch": 1.8549186128803963,
+      "grad_norm": 0.4087020703926344,
+      "learning_rate": 2.119622245540399e-05,
+      "loss": 0.3873,
+      "step": 1311
+    },
+    {
+      "epoch": 1.856334041047417,
+      "grad_norm": 0.20848052268632755,
+      "learning_rate": 2.1169989506820567e-05,
+      "loss": 0.3553,
+      "step": 1312
+    },
+    {
+      "epoch": 1.8577494692144374,
+      "grad_norm": 0.20476002741793628,
+      "learning_rate": 2.1143756558237147e-05,
+      "loss": 0.3654,
+      "step": 1313
+    },
+    {
+      "epoch": 1.8591648973814578,
+      "grad_norm": 0.22650052167287157,
+      "learning_rate": 2.1117523609653724e-05,
+      "loss": 0.3688,
+      "step": 1314
+    },
+    {
+      "epoch": 1.8605803255484785,
+      "grad_norm": 0.21967639631932895,
+      "learning_rate": 2.1091290661070308e-05,
+      "loss": 0.3659,
+      "step": 1315
+    },
+    {
+      "epoch": 1.861995753715499,
+      "grad_norm": 0.19573350674272505,
+      "learning_rate": 2.1065057712486885e-05,
+      "loss": 0.3518,
+      "step": 1316
+    },
+    {
+      "epoch": 1.8634111818825194,
+      "grad_norm": 0.21566711659758614,
+      "learning_rate": 2.1038824763903465e-05,
+      "loss": 0.3737,
+      "step": 1317
+    },
+    {
+      "epoch": 1.86482661004954,
+      "grad_norm": 0.2106286767662346,
+      "learning_rate": 2.1012591815320042e-05,
+      "loss": 0.3535,
+      "step": 1318
+    },
+    {
+      "epoch": 1.8662420382165605,
+      "grad_norm": 0.20610275935482994,
+      "learning_rate": 2.0986358866736622e-05,
+      "loss": 0.3706,
+      "step": 1319
+    },
+    {
+      "epoch": 1.867657466383581,
+      "grad_norm": 0.2094944978172916,
+      "learning_rate": 2.09601259181532e-05,
+      "loss": 0.3773,
+      "step": 1320
+    },
+    {
+      "epoch": 1.8690728945506017,
+      "grad_norm": 0.2009512235832306,
+      "learning_rate": 2.093389296956978e-05,
+      "loss": 0.3651,
+      "step": 1321
+    },
+    {
+      "epoch": 1.870488322717622,
+      "grad_norm": 0.19933772882372747,
+      "learning_rate": 2.090766002098636e-05,
+      "loss": 0.3707,
+      "step": 1322
+    },
+    {
+      "epoch": 1.8719037508846426,
+      "grad_norm": 0.20506289727859822,
+      "learning_rate": 2.088142707240294e-05,
+      "loss": 0.349,
+      "step": 1323
+    },
+    {
+      "epoch": 1.8733191790516632,
+      "grad_norm": 0.21095123180435418,
+      "learning_rate": 2.085519412381952e-05,
+      "loss": 0.3573,
+      "step": 1324
+    },
+    {
+      "epoch": 1.8747346072186837,
+      "grad_norm": 0.19076272836405586,
+      "learning_rate": 2.0828961175236097e-05,
+      "loss": 0.3607,
+      "step": 1325
+    },
+    {
+      "epoch": 1.8761500353857041,
+      "grad_norm": 0.19288645362192516,
+      "learning_rate": 2.0802728226652677e-05,
+      "loss": 0.3697,
+      "step": 1326
+    },
+    {
+      "epoch": 1.8775654635527248,
+      "grad_norm": 0.2003841123589038,
+      "learning_rate": 2.0776495278069254e-05,
+      "loss": 0.3556,
+      "step": 1327
+    },
+    {
+      "epoch": 1.8789808917197452,
+      "grad_norm": 0.20691879619101566,
+      "learning_rate": 2.0750262329485835e-05,
+      "loss": 0.3662,
+      "step": 1328
+    },
+    {
+      "epoch": 1.8803963198867657,
+      "grad_norm": 0.18551678784944658,
+      "learning_rate": 2.0724029380902415e-05,
+      "loss": 0.3589,
+      "step": 1329
+    },
+    {
+      "epoch": 1.8818117480537864,
+      "grad_norm": 0.21042470211569628,
+      "learning_rate": 2.0697796432318995e-05,
+      "loss": 0.3593,
+      "step": 1330
+    },
+    {
+      "epoch": 1.8832271762208068,
+      "grad_norm": 0.2243848627008227,
+      "learning_rate": 2.0671563483735572e-05,
+      "loss": 0.3631,
+      "step": 1331
+    },
+    {
+      "epoch": 1.8846426043878273,
+      "grad_norm": 0.20252637607681076,
+      "learning_rate": 2.0645330535152152e-05,
+      "loss": 0.3596,
+      "step": 1332
+    },
+    {
+      "epoch": 1.886058032554848,
+      "grad_norm": 0.20201222497778787,
+      "learning_rate": 2.061909758656873e-05,
+      "loss": 0.3734,
+      "step": 1333
+    },
+    {
+      "epoch": 1.8874734607218684,
+      "grad_norm": 0.21797866269539595,
+      "learning_rate": 2.059286463798531e-05,
+      "loss": 0.368,
+      "step": 1334
+    },
+    {
+      "epoch": 1.8888888888888888,
+      "grad_norm": 0.19661726770953397,
+      "learning_rate": 2.056663168940189e-05,
+      "loss": 0.3446,
+      "step": 1335
+    },
+    {
+      "epoch": 1.8903043170559095,
+      "grad_norm": 0.20102166812135902,
+      "learning_rate": 2.054039874081847e-05,
+      "loss": 0.354,
+      "step": 1336
+    },
+    {
+      "epoch": 1.89171974522293,
+      "grad_norm": 0.21327874311920472,
+      "learning_rate": 2.051416579223505e-05,
+      "loss": 0.3687,
+      "step": 1337
+    },
+    {
+      "epoch": 1.8931351733899504,
+      "grad_norm": 0.19691581170479022,
+      "learning_rate": 2.0487932843651627e-05,
+      "loss": 0.345,
+      "step": 1338
+    },
+    {
+      "epoch": 1.894550601556971,
+      "grad_norm": 0.20012006527780385,
+      "learning_rate": 2.0461699895068208e-05,
+      "loss": 0.3512,
+      "step": 1339
+    },
+    {
+      "epoch": 1.8959660297239915,
+      "grad_norm": 0.224910501612043,
+      "learning_rate": 2.0435466946484784e-05,
+      "loss": 0.3613,
+      "step": 1340
+    },
+    {
+      "epoch": 1.897381457891012,
+      "grad_norm": 0.19538273423275948,
+      "learning_rate": 2.0409233997901365e-05,
+      "loss": 0.3499,
+      "step": 1341
+    },
+    {
+      "epoch": 1.8987968860580327,
+      "grad_norm": 0.19171272752787255,
+      "learning_rate": 2.038300104931794e-05,
+      "loss": 0.368,
+      "step": 1342
+    },
+    {
+      "epoch": 1.900212314225053,
+      "grad_norm": 0.1983476334140531,
+      "learning_rate": 2.0356768100734525e-05,
+      "loss": 0.3609,
+      "step": 1343
+    },
+    {
+      "epoch": 1.9016277423920736,
+      "grad_norm": 0.20005901565438508,
+      "learning_rate": 2.0330535152151102e-05,
+      "loss": 0.3458,
+      "step": 1344
+    },
+    {
+      "epoch": 1.9030431705590942,
+      "grad_norm": 0.2291395605242673,
+      "learning_rate": 2.0304302203567683e-05,
+      "loss": 0.3463,
+      "step": 1345
+    },
+    {
+      "epoch": 1.9044585987261147,
+      "grad_norm": 0.1924624447422567,
+      "learning_rate": 2.027806925498426e-05,
+      "loss": 0.3668,
+      "step": 1346
+    },
+    {
+      "epoch": 1.9058740268931351,
+      "grad_norm": 0.22436482558939622,
+      "learning_rate": 2.025183630640084e-05,
+      "loss": 0.3556,
+      "step": 1347
+    },
+    {
+      "epoch": 1.9072894550601558,
+      "grad_norm": 0.20674060334416514,
+      "learning_rate": 2.022560335781742e-05,
+      "loss": 0.3775,
+      "step": 1348
+    },
+    {
+      "epoch": 1.9087048832271762,
+      "grad_norm": 0.21241679987762357,
+      "learning_rate": 2.0199370409233997e-05,
+      "loss": 0.3706,
+      "step": 1349
+    },
+    {
+      "epoch": 1.9101203113941967,
+      "grad_norm": 0.2035827700235495,
+      "learning_rate": 2.017313746065058e-05,
+      "loss": 0.3682,
+      "step": 1350
+    },
+    {
+      "epoch": 1.9115357395612174,
+      "grad_norm": 0.2102112730791476,
+      "learning_rate": 2.0146904512067157e-05,
+      "loss": 0.3533,
+      "step": 1351
+    },
+    {
+      "epoch": 1.9129511677282378,
+      "grad_norm": 0.1977936871380788,
+      "learning_rate": 2.0120671563483738e-05,
+      "loss": 0.3659,
+      "step": 1352
+    },
+    {
+      "epoch": 1.9143665958952583,
+      "grad_norm": 0.2073046648660033,
+      "learning_rate": 2.0094438614900315e-05,
+      "loss": 0.3494,
+      "step": 1353
+    },
+    {
+      "epoch": 1.915782024062279,
+      "grad_norm": 0.21001193349632688,
+      "learning_rate": 2.0068205666316895e-05,
+      "loss": 0.3546,
+      "step": 1354
+    },
+    {
+      "epoch": 1.9171974522292994,
+      "grad_norm": 0.21798544832491265,
+      "learning_rate": 2.0041972717733472e-05,
+      "loss": 0.3566,
+      "step": 1355
+    },
+    {
+      "epoch": 1.9186128803963198,
+      "grad_norm": 0.21290936487815346,
+      "learning_rate": 2.0015739769150055e-05,
+      "loss": 0.3632,
+      "step": 1356
+    },
+    {
+      "epoch": 1.9200283085633405,
+      "grad_norm": 0.20721108578002567,
+      "learning_rate": 1.9989506820566632e-05,
+      "loss": 0.3605,
+      "step": 1357
+    },
+    {
+      "epoch": 1.921443736730361,
+      "grad_norm": 0.21341532178889183,
+      "learning_rate": 1.9963273871983213e-05,
+      "loss": 0.3714,
+      "step": 1358
+    },
+    {
+      "epoch": 1.9228591648973814,
+      "grad_norm": 0.20509852001562356,
+      "learning_rate": 1.993704092339979e-05,
+      "loss": 0.3639,
+      "step": 1359
+    },
+    {
+      "epoch": 1.924274593064402,
+      "grad_norm": 0.18996587530747674,
+      "learning_rate": 1.991080797481637e-05,
+      "loss": 0.3491,
+      "step": 1360
+    },
+    {
+      "epoch": 1.9256900212314225,
+      "grad_norm": 0.22186916332584036,
+      "learning_rate": 1.988457502623295e-05,
+      "loss": 0.3584,
+      "step": 1361
+    },
+    {
+      "epoch": 1.927105449398443,
+      "grad_norm": 0.19859813027365922,
+      "learning_rate": 1.9858342077649527e-05,
+      "loss": 0.3507,
+      "step": 1362
+    },
+    {
+      "epoch": 1.9285208775654636,
+      "grad_norm": 0.19952702214893042,
+      "learning_rate": 1.983210912906611e-05,
+      "loss": 0.3639,
+      "step": 1363
+    },
+    {
+      "epoch": 1.929936305732484,
+      "grad_norm": 0.20121877277304315,
+      "learning_rate": 1.9805876180482688e-05,
+      "loss": 0.374,
+      "step": 1364
+    },
+    {
+      "epoch": 1.9313517338995045,
+      "grad_norm": 0.19745024884216902,
+      "learning_rate": 1.9779643231899268e-05,
+      "loss": 0.3627,
+      "step": 1365
+    },
+    {
+      "epoch": 1.9327671620665252,
+      "grad_norm": 0.20892556854783773,
+      "learning_rate": 1.9753410283315845e-05,
+      "loss": 0.356,
+      "step": 1366
+    },
+    {
+      "epoch": 1.9341825902335457,
+      "grad_norm": 0.20820929191051837,
+      "learning_rate": 1.9727177334732425e-05,
+      "loss": 0.363,
+      "step": 1367
+    },
+    {
+      "epoch": 1.9355980184005661,
+      "grad_norm": 0.19872227127810696,
+      "learning_rate": 1.9700944386149002e-05,
+      "loss": 0.3788,
+      "step": 1368
+    },
+    {
+      "epoch": 1.9370134465675868,
+      "grad_norm": 0.20891761384090757,
+      "learning_rate": 1.9674711437565582e-05,
+      "loss": 0.3648,
+      "step": 1369
+    },
+    {
+      "epoch": 1.9384288747346072,
+      "grad_norm": 0.19845543506864766,
+      "learning_rate": 1.9648478488982163e-05,
+      "loss": 0.369,
+      "step": 1370
+    },
+    {
+      "epoch": 1.9398443029016277,
+      "grad_norm": 0.19113509199428175,
+      "learning_rate": 1.9622245540398743e-05,
+      "loss": 0.364,
+      "step": 1371
+    },
+    {
+      "epoch": 1.9412597310686484,
+      "grad_norm": 0.21549384559079116,
+      "learning_rate": 1.9596012591815323e-05,
+      "loss": 0.3625,
+      "step": 1372
+    },
+    {
+      "epoch": 1.9426751592356688,
+      "grad_norm": 0.22030815818854466,
+      "learning_rate": 1.95697796432319e-05,
+      "loss": 0.3745,
+      "step": 1373
+    },
+    {
+      "epoch": 1.9440905874026893,
+      "grad_norm": 0.19402336745232818,
+      "learning_rate": 1.954354669464848e-05,
+      "loss": 0.3534,
+      "step": 1374
+    },
+    {
+      "epoch": 1.94550601556971,
+      "grad_norm": 0.21121560897394442,
+      "learning_rate": 1.9517313746065057e-05,
+      "loss": 0.3591,
+      "step": 1375
+    },
+    {
+      "epoch": 1.9469214437367304,
+      "grad_norm": 0.20599072914025068,
+      "learning_rate": 1.9491080797481637e-05,
+      "loss": 0.3482,
+      "step": 1376
+    },
+    {
+      "epoch": 1.9483368719037508,
+      "grad_norm": 0.2101078777457639,
+      "learning_rate": 1.9464847848898218e-05,
+      "loss": 0.3666,
+      "step": 1377
+    },
+    {
+      "epoch": 1.9497523000707715,
+      "grad_norm": 0.18892328435730557,
+      "learning_rate": 1.9438614900314798e-05,
+      "loss": 0.3628,
+      "step": 1378
+    },
+    {
+      "epoch": 1.951167728237792,
+      "grad_norm": 0.2131996417180784,
+      "learning_rate": 1.9412381951731375e-05,
+      "loss": 0.3635,
+      "step": 1379
+    },
+    {
+      "epoch": 1.9525831564048124,
+      "grad_norm": 0.20128606932865298,
+      "learning_rate": 1.9386149003147955e-05,
+      "loss": 0.3638,
+      "step": 1380
+    },
+    {
+      "epoch": 1.953998584571833,
+      "grad_norm": 0.21900352568972972,
+      "learning_rate": 1.9359916054564532e-05,
+      "loss": 0.3638,
+      "step": 1381
+    },
+    {
+      "epoch": 1.9554140127388535,
+      "grad_norm": 0.1956686768669383,
+      "learning_rate": 1.9333683105981112e-05,
+      "loss": 0.3625,
+      "step": 1382
+    },
+    {
+      "epoch": 1.956829440905874,
+      "grad_norm": 0.19093257924184567,
+      "learning_rate": 1.9307450157397693e-05,
+      "loss": 0.3684,
+      "step": 1383
+    },
+    {
+      "epoch": 1.9582448690728946,
+      "grad_norm": 0.20479976158398785,
+      "learning_rate": 1.9281217208814273e-05,
+      "loss": 0.3625,
+      "step": 1384
+    },
+    {
+      "epoch": 1.959660297239915,
+      "grad_norm": 0.20885102906024197,
+      "learning_rate": 1.9254984260230853e-05,
+      "loss": 0.3579,
+      "step": 1385
+    },
+    {
+      "epoch": 1.9610757254069355,
+      "grad_norm": 0.17258172072412006,
+      "learning_rate": 1.922875131164743e-05,
+      "loss": 0.3565,
+      "step": 1386
+    },
+    {
+      "epoch": 1.9624911535739562,
+      "grad_norm": 0.2178182423526484,
+      "learning_rate": 1.920251836306401e-05,
+      "loss": 0.3698,
+      "step": 1387
+    },
+    {
+      "epoch": 1.9639065817409767,
+      "grad_norm": 0.19896161831763218,
+      "learning_rate": 1.9176285414480587e-05,
+      "loss": 0.3502,
+      "step": 1388
+    },
+    {
+      "epoch": 1.965322009907997,
+      "grad_norm": 0.18768357869637123,
+      "learning_rate": 1.9150052465897168e-05,
+      "loss": 0.3668,
+      "step": 1389
+    },
+    {
+      "epoch": 1.9667374380750178,
+      "grad_norm": 0.19297739606631034,
+      "learning_rate": 1.9123819517313745e-05,
+      "loss": 0.365,
+      "step": 1390
+    },
+    {
+      "epoch": 1.9681528662420382,
+      "grad_norm": 0.22915739722100426,
+      "learning_rate": 1.9097586568730328e-05,
+      "loss": 0.374,
+      "step": 1391
+    },
+    {
+      "epoch": 1.9695682944090587,
+      "grad_norm": 0.21354617076364507,
+      "learning_rate": 1.9071353620146905e-05,
+      "loss": 0.3504,
+      "step": 1392
+    },
+    {
+      "epoch": 1.9709837225760793,
+      "grad_norm": 0.1978571548565372,
+      "learning_rate": 1.9045120671563485e-05,
+      "loss": 0.333,
+      "step": 1393
+    },
+    {
+      "epoch": 1.9723991507430998,
+      "grad_norm": 0.19721042495263594,
+      "learning_rate": 1.9018887722980062e-05,
+      "loss": 0.3448,
+      "step": 1394
+    },
+    {
+      "epoch": 1.9738145789101202,
+      "grad_norm": 0.19005416630899646,
+      "learning_rate": 1.8992654774396643e-05,
+      "loss": 0.3771,
+      "step": 1395
+    },
+    {
+      "epoch": 1.975230007077141,
+      "grad_norm": 0.2133322733488417,
+      "learning_rate": 1.8966421825813223e-05,
+      "loss": 0.357,
+      "step": 1396
+    },
+    {
+      "epoch": 1.9766454352441614,
+      "grad_norm": 0.1894050791328522,
+      "learning_rate": 1.89401888772298e-05,
+      "loss": 0.3533,
+      "step": 1397
+    },
+    {
+      "epoch": 1.9780608634111818,
+      "grad_norm": 0.20012948540827005,
+      "learning_rate": 1.8913955928646383e-05,
+      "loss": 0.3681,
+      "step": 1398
+    },
+    {
+      "epoch": 1.9794762915782025,
+      "grad_norm": 0.20502607195124412,
+      "learning_rate": 1.888772298006296e-05,
+      "loss": 0.3688,
+      "step": 1399
+    },
+    {
+      "epoch": 1.980891719745223,
+      "grad_norm": 0.19446905713223603,
+      "learning_rate": 1.886149003147954e-05,
+      "loss": 0.3488,
+      "step": 1400
+    },
+    {
+      "epoch": 1.9823071479122434,
+      "grad_norm": 0.19833011267832945,
+      "learning_rate": 1.8835257082896118e-05,
+      "loss": 0.3752,
+      "step": 1401
+    },
+    {
+      "epoch": 1.983722576079264,
+      "grad_norm": 0.19273254120634167,
+      "learning_rate": 1.8809024134312698e-05,
+      "loss": 0.3621,
+      "step": 1402
+    },
+    {
+      "epoch": 1.9851380042462845,
+      "grad_norm": 0.20749102898898195,
+      "learning_rate": 1.8782791185729275e-05,
+      "loss": 0.3509,
+      "step": 1403
+    },
+    {
+      "epoch": 1.986553432413305,
+      "grad_norm": 0.20304238601187583,
+      "learning_rate": 1.8756558237145855e-05,
+      "loss": 0.3514,
+      "step": 1404
+    },
+    {
+      "epoch": 1.9879688605803256,
+      "grad_norm": 0.1855035727758363,
+      "learning_rate": 1.8730325288562435e-05,
+      "loss": 0.3559,
+      "step": 1405
+    },
+    {
+      "epoch": 1.989384288747346,
+      "grad_norm": 0.1900038858199582,
+      "learning_rate": 1.8704092339979016e-05,
+      "loss": 0.3593,
+      "step": 1406
+    },
+    {
+      "epoch": 1.9907997169143665,
+      "grad_norm": 0.26751614648177435,
+      "learning_rate": 1.8677859391395592e-05,
+      "loss": 0.3584,
+      "step": 1407
+    },
+    {
+      "epoch": 1.9922151450813872,
+      "grad_norm": 0.1895064557610461,
+      "learning_rate": 1.8651626442812173e-05,
+      "loss": 0.3389,
+      "step": 1408
+    },
+    {
+      "epoch": 1.9936305732484076,
+      "grad_norm": 0.20279845259843857,
+      "learning_rate": 1.8625393494228753e-05,
+      "loss": 0.363,
+      "step": 1409
+    },
+    {
+      "epoch": 1.995046001415428,
+      "grad_norm": 0.2039992306549258,
+      "learning_rate": 1.859916054564533e-05,
+      "loss": 0.3554,
+      "step": 1410
+    },
+    {
+      "epoch": 1.9964614295824488,
+      "grad_norm": 0.18649910427325964,
+      "learning_rate": 1.857292759706191e-05,
+      "loss": 0.3424,
+      "step": 1411
+    },
+    {
+      "epoch": 1.9978768577494692,
+      "grad_norm": 0.20570040271922255,
+      "learning_rate": 1.854669464847849e-05,
+      "loss": 0.3775,
+      "step": 1412
+    },
+    {
+      "epoch": 1.9992922859164897,
+      "grad_norm": 0.19929248975834807,
+      "learning_rate": 1.852046169989507e-05,
+      "loss": 0.3617,
+      "step": 1413
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.31812015380648995,
+      "learning_rate": 1.8494228751311648e-05,
+      "loss": 0.3737,
+      "step": 1414
+    },
+    {
+      "epoch": 2.0014154281670207,
+      "grad_norm": 0.2618432835627456,
+      "learning_rate": 1.8467995802728228e-05,
+      "loss": 0.3034,
+      "step": 1415
+    },
+    {
+      "epoch": 2.002830856334041,
+      "grad_norm": 0.22735472922428987,
+      "learning_rate": 1.8441762854144805e-05,
+      "loss": 0.3015,
+      "step": 1416
+    },
+    {
+      "epoch": 2.0042462845010616,
+      "grad_norm": 0.2304654888028929,
+      "learning_rate": 1.8415529905561385e-05,
+      "loss": 0.2929,
+      "step": 1417
+    },
+    {
+      "epoch": 2.0056617126680822,
+      "grad_norm": 0.3077158798541765,
+      "learning_rate": 1.8389296956977965e-05,
+      "loss": 0.2998,
+      "step": 1418
+    },
+    {
+      "epoch": 2.0070771408351025,
+      "grad_norm": 0.22354884644249307,
+      "learning_rate": 1.8363064008394546e-05,
+      "loss": 0.2892,
+      "step": 1419
+    },
+    {
+      "epoch": 2.008492569002123,
+      "grad_norm": 0.2307122902012782,
+      "learning_rate": 1.8336831059811123e-05,
+      "loss": 0.288,
+      "step": 1420
+    },
+    {
+      "epoch": 2.009907997169144,
+      "grad_norm": 0.3004574962063541,
+      "learning_rate": 1.8310598111227703e-05,
+      "loss": 0.2954,
+      "step": 1421
+    },
+    {
+      "epoch": 2.011323425336164,
+      "grad_norm": 0.25742558193381443,
+      "learning_rate": 1.8284365162644283e-05,
+      "loss": 0.2904,
+      "step": 1422
+    },
+    {
+      "epoch": 2.0127388535031847,
+      "grad_norm": 0.22091196543564764,
+      "learning_rate": 1.825813221406086e-05,
+      "loss": 0.2984,
+      "step": 1423
+    },
+    {
+      "epoch": 2.0141542816702054,
+      "grad_norm": 0.2693133643238991,
+      "learning_rate": 1.823189926547744e-05,
+      "loss": 0.2809,
+      "step": 1424
+    },
+    {
+      "epoch": 2.0155697098372256,
+      "grad_norm": 0.2946008164449425,
+      "learning_rate": 1.820566631689402e-05,
+      "loss": 0.3134,
+      "step": 1425
+    },
+    {
+      "epoch": 2.0169851380042463,
+      "grad_norm": 0.22329898493932543,
+      "learning_rate": 1.81794333683106e-05,
+      "loss": 0.2792,
+      "step": 1426
+    },
+    {
+      "epoch": 2.018400566171267,
+      "grad_norm": 0.23521702050759785,
+      "learning_rate": 1.8153200419727178e-05,
+      "loss": 0.2834,
+      "step": 1427
+    },
+    {
+      "epoch": 2.019815994338287,
+      "grad_norm": 0.2301865651909983,
+      "learning_rate": 1.8126967471143758e-05,
+      "loss": 0.2833,
+      "step": 1428
+    },
+    {
+      "epoch": 2.021231422505308,
+      "grad_norm": 0.23805812832238346,
+      "learning_rate": 1.8100734522560335e-05,
+      "loss": 0.2948,
+      "step": 1429
+    },
+    {
+      "epoch": 2.0226468506723285,
+      "grad_norm": 0.2204999436021115,
+      "learning_rate": 1.8074501573976915e-05,
+      "loss": 0.2925,
+      "step": 1430
+    },
+    {
+      "epoch": 2.0240622788393487,
+      "grad_norm": 0.2099342934273649,
+      "learning_rate": 1.8048268625393496e-05,
+      "loss": 0.2917,
+      "step": 1431
+    },
+    {
+      "epoch": 2.0254777070063694,
+      "grad_norm": 0.21316955775068464,
+      "learning_rate": 1.8022035676810076e-05,
+      "loss": 0.2911,
+      "step": 1432
+    },
+    {
+      "epoch": 2.02689313517339,
+      "grad_norm": 0.203591778774364,
+      "learning_rate": 1.7995802728226656e-05,
+      "loss": 0.284,
+      "step": 1433
+    },
+    {
+      "epoch": 2.0283085633404103,
+      "grad_norm": 0.19857682958767808,
+      "learning_rate": 1.7969569779643233e-05,
+      "loss": 0.2682,
+      "step": 1434
+    },
+    {
+      "epoch": 2.029723991507431,
+      "grad_norm": 0.20738110099610785,
+      "learning_rate": 1.7943336831059813e-05,
+      "loss": 0.2856,
+      "step": 1435
+    },
+    {
+      "epoch": 2.0311394196744517,
+      "grad_norm": 0.20879171795426643,
+      "learning_rate": 1.791710388247639e-05,
+      "loss": 0.2898,
+      "step": 1436
+    },
+    {
+      "epoch": 2.032554847841472,
+      "grad_norm": 0.2100994408662993,
+      "learning_rate": 1.789087093389297e-05,
+      "loss": 0.2853,
+      "step": 1437
+    },
+    {
+      "epoch": 2.0339702760084926,
+      "grad_norm": 0.21861018036891264,
+      "learning_rate": 1.7864637985309547e-05,
+      "loss": 0.2842,
+      "step": 1438
+    },
+    {
+      "epoch": 2.0353857041755132,
+      "grad_norm": 0.1998874003435123,
+      "learning_rate": 1.783840503672613e-05,
+      "loss": 0.2877,
+      "step": 1439
+    },
+    {
+      "epoch": 2.0368011323425335,
+      "grad_norm": 0.19733644838624692,
+      "learning_rate": 1.7812172088142708e-05,
+      "loss": 0.2805,
+      "step": 1440
+    },
+    {
+      "epoch": 2.038216560509554,
+      "grad_norm": 0.2102345477010653,
+      "learning_rate": 1.7785939139559288e-05,
+      "loss": 0.2893,
+      "step": 1441
+    },
+    {
+      "epoch": 2.039631988676575,
+      "grad_norm": 0.22967771644312862,
+      "learning_rate": 1.7759706190975865e-05,
+      "loss": 0.2897,
+      "step": 1442
+    },
+    {
+      "epoch": 2.041047416843595,
+      "grad_norm": 0.21654862978318307,
+      "learning_rate": 1.7733473242392445e-05,
+      "loss": 0.283,
+      "step": 1443
+    },
+    {
+      "epoch": 2.0424628450106157,
+      "grad_norm": 0.2069240069772118,
+      "learning_rate": 1.7707240293809026e-05,
+      "loss": 0.2932,
+      "step": 1444
+    },
+    {
+      "epoch": 2.0438782731776364,
+      "grad_norm": 0.20251972837802149,
+      "learning_rate": 1.7681007345225603e-05,
+      "loss": 0.2823,
+      "step": 1445
+    },
+    {
+      "epoch": 2.0452937013446566,
+      "grad_norm": 0.1985558114956269,
+      "learning_rate": 1.7654774396642186e-05,
+      "loss": 0.2774,
+      "step": 1446
+    },
+    {
+      "epoch": 2.0467091295116773,
+      "grad_norm": 0.21672062844894024,
+      "learning_rate": 1.7628541448058763e-05,
+      "loss": 0.2801,
+      "step": 1447
+    },
+    {
+      "epoch": 2.048124557678698,
+      "grad_norm": 0.1868319256602976,
+      "learning_rate": 1.7602308499475343e-05,
+      "loss": 0.285,
+      "step": 1448
+    },
+    {
+      "epoch": 2.049539985845718,
+      "grad_norm": 0.20743295673083878,
+      "learning_rate": 1.757607555089192e-05,
+      "loss": 0.278,
+      "step": 1449
+    },
+    {
+      "epoch": 2.050955414012739,
+      "grad_norm": 0.22077380379857767,
+      "learning_rate": 1.75498426023085e-05,
+      "loss": 0.3091,
+      "step": 1450
+    },
+    {
+      "epoch": 2.0523708421797595,
+      "grad_norm": 0.2034193756295043,
+      "learning_rate": 1.7523609653725078e-05,
+      "loss": 0.2831,
+      "step": 1451
+    },
+    {
+      "epoch": 2.0537862703467797,
+      "grad_norm": 0.1977083504359432,
+      "learning_rate": 1.7497376705141658e-05,
+      "loss": 0.2934,
+      "step": 1452
+    },
+    {
+      "epoch": 2.0552016985138004,
+      "grad_norm": 0.21295222552781645,
+      "learning_rate": 1.7471143756558238e-05,
+      "loss": 0.2801,
+      "step": 1453
+    },
+    {
+      "epoch": 2.056617126680821,
+      "grad_norm": 0.20122015505948554,
+      "learning_rate": 1.744491080797482e-05,
+      "loss": 0.2948,
+      "step": 1454
+    },
+    {
+      "epoch": 2.0580325548478413,
+      "grad_norm": 0.19851809324642486,
+      "learning_rate": 1.7418677859391395e-05,
+      "loss": 0.288,
+      "step": 1455
+    },
+    {
+      "epoch": 2.059447983014862,
+      "grad_norm": 0.2084094141218526,
+      "learning_rate": 1.7392444910807976e-05,
+      "loss": 0.3048,
+      "step": 1456
+    },
+    {
+      "epoch": 2.0608634111818827,
+      "grad_norm": 0.18950859697388175,
+      "learning_rate": 1.7366211962224556e-05,
+      "loss": 0.2781,
+      "step": 1457
+    },
+    {
+      "epoch": 2.062278839348903,
+      "grad_norm": 0.17827020683407357,
+      "learning_rate": 1.7339979013641133e-05,
+      "loss": 0.2872,
+      "step": 1458
+    },
+    {
+      "epoch": 2.0636942675159236,
+      "grad_norm": 0.20524943163518478,
+      "learning_rate": 1.7313746065057713e-05,
+      "loss": 0.2844,
+      "step": 1459
+    },
+    {
+      "epoch": 2.0651096956829442,
+      "grad_norm": 0.20828751157840567,
+      "learning_rate": 1.7287513116474293e-05,
+      "loss": 0.2911,
+      "step": 1460
+    },
+    {
+      "epoch": 2.0665251238499645,
+      "grad_norm": 0.19230496490646315,
+      "learning_rate": 1.7261280167890874e-05,
+      "loss": 0.2997,
+      "step": 1461
+    },
+    {
+      "epoch": 2.067940552016985,
+      "grad_norm": 0.18348519918314377,
+      "learning_rate": 1.723504721930745e-05,
+      "loss": 0.2772,
+      "step": 1462
+    },
+    {
+      "epoch": 2.069355980184006,
+      "grad_norm": 0.194299296268244,
+      "learning_rate": 1.720881427072403e-05,
+      "loss": 0.2987,
+      "step": 1463
+    },
+    {
+      "epoch": 2.070771408351026,
+      "grad_norm": 0.19044245083990521,
+      "learning_rate": 1.7182581322140608e-05,
+      "loss": 0.2886,
+      "step": 1464
+    },
+    {
+      "epoch": 2.0721868365180467,
+      "grad_norm": 0.209393739696165,
+      "learning_rate": 1.7156348373557188e-05,
+      "loss": 0.2848,
+      "step": 1465
+    },
+    {
+      "epoch": 2.0736022646850674,
+      "grad_norm": 0.20147250073789505,
+      "learning_rate": 1.7130115424973768e-05,
+      "loss": 0.2771,
+      "step": 1466
+    },
+    {
+      "epoch": 2.0750176928520876,
+      "grad_norm": 0.20558457802376123,
+      "learning_rate": 1.710388247639035e-05,
+      "loss": 0.2948,
+      "step": 1467
+    },
+    {
+      "epoch": 2.0764331210191083,
+      "grad_norm": 0.19255025397263117,
+      "learning_rate": 1.7077649527806925e-05,
+      "loss": 0.2833,
+      "step": 1468
+    },
+    {
+      "epoch": 2.077848549186129,
+      "grad_norm": 0.19957625475208307,
+      "learning_rate": 1.7051416579223506e-05,
+      "loss": 0.3008,
+      "step": 1469
+    },
+    {
+      "epoch": 2.079263977353149,
+      "grad_norm": 0.2166469264979805,
+      "learning_rate": 1.7025183630640086e-05,
+      "loss": 0.2824,
+      "step": 1470
+    },
+    {
+      "epoch": 2.08067940552017,
+      "grad_norm": 0.2013825115658045,
+      "learning_rate": 1.6998950682056663e-05,
+      "loss": 0.2886,
+      "step": 1471
+    },
+    {
+      "epoch": 2.0820948336871905,
+      "grad_norm": 0.2099458003952584,
+      "learning_rate": 1.6972717733473243e-05,
+      "loss": 0.2835,
+      "step": 1472
+    },
+    {
+      "epoch": 2.0835102618542107,
+      "grad_norm": 0.1937562307282105,
+      "learning_rate": 1.694648478488982e-05,
+      "loss": 0.2777,
+      "step": 1473
+    },
+    {
+      "epoch": 2.0849256900212314,
+      "grad_norm": 0.22143430661074096,
+      "learning_rate": 1.6920251836306404e-05,
+      "loss": 0.295,
+      "step": 1474
+    },
+    {
+      "epoch": 2.086341118188252,
+      "grad_norm": 0.19961487797679497,
+      "learning_rate": 1.689401888772298e-05,
+      "loss": 0.2893,
+      "step": 1475
+    },
+    {
+      "epoch": 2.0877565463552723,
+      "grad_norm": 0.21740387957685733,
+      "learning_rate": 1.686778593913956e-05,
+      "loss": 0.2869,
+      "step": 1476
+    },
+    {
+      "epoch": 2.089171974522293,
+      "grad_norm": 0.1984607922205683,
+      "learning_rate": 1.6841552990556138e-05,
+      "loss": 0.2914,
+      "step": 1477
+    },
+    {
+      "epoch": 2.0905874026893136,
+      "grad_norm": 0.1970620738011852,
+      "learning_rate": 1.6815320041972718e-05,
+      "loss": 0.2872,
+      "step": 1478
+    },
+    {
+      "epoch": 2.092002830856334,
+      "grad_norm": 0.20203859593307708,
+      "learning_rate": 1.67890870933893e-05,
+      "loss": 0.2935,
+      "step": 1479
+    },
+    {
+      "epoch": 2.0934182590233545,
+      "grad_norm": 0.19345474961814704,
+      "learning_rate": 1.6762854144805875e-05,
+      "loss": 0.2919,
+      "step": 1480
+    },
+    {
+      "epoch": 2.094833687190375,
+      "grad_norm": 0.19683053511901935,
+      "learning_rate": 1.6736621196222456e-05,
+      "loss": 0.2902,
+      "step": 1481
+    },
+    {
+      "epoch": 2.0962491153573954,
+      "grad_norm": 0.20117757322148833,
+      "learning_rate": 1.6710388247639036e-05,
+      "loss": 0.2907,
+      "step": 1482
+    },
+    {
+      "epoch": 2.097664543524416,
+      "grad_norm": 0.19094219547799363,
+      "learning_rate": 1.6684155299055616e-05,
+      "loss": 0.2771,
+      "step": 1483
+    },
+    {
+      "epoch": 2.099079971691437,
+      "grad_norm": 0.1917696231454564,
+      "learning_rate": 1.6657922350472193e-05,
+      "loss": 0.2791,
+      "step": 1484
+    },
+    {
+      "epoch": 2.100495399858457,
+      "grad_norm": 0.20887738097023587,
+      "learning_rate": 1.6631689401888773e-05,
+      "loss": 0.2895,
+      "step": 1485
+    },
+    {
+      "epoch": 2.1019108280254777,
+      "grad_norm": 0.20145195422505788,
+      "learning_rate": 1.660545645330535e-05,
+      "loss": 0.277,
+      "step": 1486
+    },
+    {
+      "epoch": 2.1033262561924984,
+      "grad_norm": 0.21906519913947312,
+      "learning_rate": 1.657922350472193e-05,
+      "loss": 0.3038,
+      "step": 1487
+    },
+    {
+      "epoch": 2.1047416843595186,
+      "grad_norm": 0.2032232846451326,
+      "learning_rate": 1.655299055613851e-05,
+      "loss": 0.2772,
+      "step": 1488
+    },
+    {
+      "epoch": 2.1061571125265393,
+      "grad_norm": 0.19785341115033714,
+      "learning_rate": 1.652675760755509e-05,
+      "loss": 0.2867,
+      "step": 1489
+    },
+    {
+      "epoch": 2.10757254069356,
+      "grad_norm": 0.2049090742872932,
+      "learning_rate": 1.6500524658971668e-05,
+      "loss": 0.2868,
+      "step": 1490
+    },
+    {
+      "epoch": 2.10898796886058,
+      "grad_norm": 0.2026792528862842,
+      "learning_rate": 1.647429171038825e-05,
+      "loss": 0.2923,
+      "step": 1491
+    },
+    {
+      "epoch": 2.110403397027601,
+      "grad_norm": 0.21064805818044663,
+      "learning_rate": 1.644805876180483e-05,
+      "loss": 0.2933,
+      "step": 1492
+    },
+    {
+      "epoch": 2.1118188251946215,
+      "grad_norm": 0.21631163550771182,
+      "learning_rate": 1.6421825813221406e-05,
+      "loss": 0.2794,
+      "step": 1493
+    },
+    {
+      "epoch": 2.1132342533616417,
+      "grad_norm": 0.20710600992144443,
+      "learning_rate": 1.6395592864637986e-05,
+      "loss": 0.2731,
+      "step": 1494
+    },
+    {
+      "epoch": 2.1146496815286624,
+      "grad_norm": 0.2030260738287538,
+      "learning_rate": 1.6369359916054566e-05,
+      "loss": 0.2753,
+      "step": 1495
+    },
+    {
+      "epoch": 2.116065109695683,
+      "grad_norm": 0.2384668341939414,
+      "learning_rate": 1.6343126967471146e-05,
+      "loss": 0.2786,
+      "step": 1496
+    },
+    {
+      "epoch": 2.1174805378627033,
+      "grad_norm": 0.2151123848622728,
+      "learning_rate": 1.6316894018887723e-05,
+      "loss": 0.2877,
+      "step": 1497
+    },
+    {
+      "epoch": 2.118895966029724,
+      "grad_norm": 0.21168635358673377,
+      "learning_rate": 1.6290661070304304e-05,
+      "loss": 0.2972,
+      "step": 1498
+    },
+    {
+      "epoch": 2.1203113941967446,
+      "grad_norm": 0.19550048749813878,
+      "learning_rate": 1.626442812172088e-05,
+      "loss": 0.2818,
+      "step": 1499
+    },
+    {
+      "epoch": 2.121726822363765,
+      "grad_norm": 0.19624090199655572,
+      "learning_rate": 1.623819517313746e-05,
+      "loss": 0.285,
+      "step": 1500
+    },
+    {
+      "epoch": 2.1231422505307855,
+      "grad_norm": 0.2089549740401407,
+      "learning_rate": 1.621196222455404e-05,
+      "loss": 0.294,
+      "step": 1501
+    },
+    {
+      "epoch": 2.124557678697806,
+      "grad_norm": 0.21856180217421553,
+      "learning_rate": 1.618572927597062e-05,
+      "loss": 0.2834,
+      "step": 1502
+    },
+    {
+      "epoch": 2.1259731068648264,
+      "grad_norm": 0.19946569492829458,
+      "learning_rate": 1.6159496327387198e-05,
+      "loss": 0.2988,
+      "step": 1503
+    },
+    {
+      "epoch": 2.127388535031847,
+      "grad_norm": 0.18882579304258193,
+      "learning_rate": 1.613326337880378e-05,
+      "loss": 0.2768,
+      "step": 1504
+    },
+    {
+      "epoch": 2.1288039631988678,
+      "grad_norm": 0.21404486282736535,
+      "learning_rate": 1.610703043022036e-05,
+      "loss": 0.2804,
+      "step": 1505
+    },
+    {
+      "epoch": 2.130219391365888,
+      "grad_norm": 0.20090952321056332,
+      "learning_rate": 1.6080797481636936e-05,
+      "loss": 0.2756,
+      "step": 1506
+    },
+    {
+      "epoch": 2.1316348195329087,
+      "grad_norm": 0.1943995402386731,
+      "learning_rate": 1.6054564533053516e-05,
+      "loss": 0.2937,
+      "step": 1507
+    },
+    {
+      "epoch": 2.1330502476999293,
+      "grad_norm": 0.21000063080660478,
+      "learning_rate": 1.6028331584470096e-05,
+      "loss": 0.2989,
+      "step": 1508
+    },
+    {
+      "epoch": 2.1344656758669496,
+      "grad_norm": 0.19677351524392991,
+      "learning_rate": 1.6002098635886677e-05,
+      "loss": 0.2794,
+      "step": 1509
+    },
+    {
+      "epoch": 2.1358811040339702,
+      "grad_norm": 0.18794381684145336,
+      "learning_rate": 1.5975865687303253e-05,
+      "loss": 0.2789,
+      "step": 1510
+    },
+    {
+      "epoch": 2.137296532200991,
+      "grad_norm": 0.20567790184577053,
+      "learning_rate": 1.5949632738719834e-05,
+      "loss": 0.2872,
+      "step": 1511
+    },
+    {
+      "epoch": 2.138711960368011,
+      "grad_norm": 0.18201871625247118,
+      "learning_rate": 1.592339979013641e-05,
+      "loss": 0.2771,
+      "step": 1512
+    },
+    {
+      "epoch": 2.140127388535032,
+      "grad_norm": 0.19584814842799103,
+      "learning_rate": 1.589716684155299e-05,
+      "loss": 0.2831,
+      "step": 1513
+    },
+    {
+      "epoch": 2.1415428167020525,
+      "grad_norm": 0.1903199666447714,
+      "learning_rate": 1.5870933892969568e-05,
+      "loss": 0.2929,
+      "step": 1514
+    },
+    {
+      "epoch": 2.1429582448690727,
+      "grad_norm": 0.18679516628993567,
+      "learning_rate": 1.584470094438615e-05,
+      "loss": 0.2925,
+      "step": 1515
+    },
+    {
+      "epoch": 2.1443736730360934,
+      "grad_norm": 0.18996749381846695,
+      "learning_rate": 1.581846799580273e-05,
+      "loss": 0.2944,
+      "step": 1516
+    },
+    {
+      "epoch": 2.145789101203114,
+      "grad_norm": 0.19897930810530637,
+      "learning_rate": 1.579223504721931e-05,
+      "loss": 0.2916,
+      "step": 1517
+    },
+    {
+      "epoch": 2.1472045293701343,
+      "grad_norm": 0.19099277382658197,
+      "learning_rate": 1.576600209863589e-05,
+      "loss": 0.2924,
+      "step": 1518
+    },
+    {
+      "epoch": 2.148619957537155,
+      "grad_norm": 0.18179013442370356,
+      "learning_rate": 1.5739769150052466e-05,
+      "loss": 0.2847,
+      "step": 1519
+    },
+    {
+      "epoch": 2.1500353857041756,
+      "grad_norm": 0.191157987934032,
+      "learning_rate": 1.5713536201469046e-05,
+      "loss": 0.2788,
+      "step": 1520
+    },
+    {
+      "epoch": 2.151450813871196,
+      "grad_norm": 0.20928856949707,
+      "learning_rate": 1.5687303252885623e-05,
+      "loss": 0.281,
+      "step": 1521
+    },
+    {
+      "epoch": 2.1528662420382165,
+      "grad_norm": 0.18497058614375653,
+      "learning_rate": 1.5661070304302207e-05,
+      "loss": 0.2795,
+      "step": 1522
+    },
+    {
+      "epoch": 2.154281670205237,
+      "grad_norm": 0.18326141099307156,
+      "learning_rate": 1.5634837355718784e-05,
+      "loss": 0.2886,
+      "step": 1523
+    },
+    {
+      "epoch": 2.1556970983722574,
+      "grad_norm": 0.20853024897514022,
+      "learning_rate": 1.5608604407135364e-05,
+      "loss": 0.2889,
+      "step": 1524
+    },
+    {
+      "epoch": 2.157112526539278,
+      "grad_norm": 0.20387139051690897,
+      "learning_rate": 1.558237145855194e-05,
+      "loss": 0.2975,
+      "step": 1525
+    },
+    {
+      "epoch": 2.1585279547062988,
+      "grad_norm": 0.18843544363156378,
+      "learning_rate": 1.555613850996852e-05,
+      "loss": 0.2839,
+      "step": 1526
+    },
+    {
+      "epoch": 2.159943382873319,
+      "grad_norm": 0.18707200958840547,
+      "learning_rate": 1.55299055613851e-05,
+      "loss": 0.2687,
+      "step": 1527
+    },
+    {
+      "epoch": 2.1613588110403397,
+      "grad_norm": 0.21495673801240114,
+      "learning_rate": 1.5503672612801678e-05,
+      "loss": 0.2852,
+      "step": 1528
+    },
+    {
+      "epoch": 2.1627742392073603,
+      "grad_norm": 0.18758479870398373,
+      "learning_rate": 1.547743966421826e-05,
+      "loss": 0.289,
+      "step": 1529
+    },
+    {
+      "epoch": 2.1641896673743806,
+      "grad_norm": 0.19202383784848073,
+      "learning_rate": 1.545120671563484e-05,
+      "loss": 0.2847,
+      "step": 1530
+    },
+    {
+      "epoch": 2.1656050955414012,
+      "grad_norm": 0.20312171765896048,
+      "learning_rate": 1.542497376705142e-05,
+      "loss": 0.2901,
+      "step": 1531
+    },
+    {
+      "epoch": 2.167020523708422,
+      "grad_norm": 0.20618143284667947,
+      "learning_rate": 1.5398740818467996e-05,
+      "loss": 0.2831,
+      "step": 1532
+    },
+    {
+      "epoch": 2.168435951875442,
+      "grad_norm": 0.196538483127226,
+      "learning_rate": 1.5372507869884576e-05,
+      "loss": 0.2787,
+      "step": 1533
+    },
+    {
+      "epoch": 2.169851380042463,
+      "grad_norm": 0.19945378439158584,
+      "learning_rate": 1.5346274921301153e-05,
+      "loss": 0.2894,
+      "step": 1534
+    },
+    {
+      "epoch": 2.1712668082094835,
+      "grad_norm": 0.2097338644680216,
+      "learning_rate": 1.5320041972717733e-05,
+      "loss": 0.2855,
+      "step": 1535
+    },
+    {
+      "epoch": 2.1726822363765037,
+      "grad_norm": 0.20428088695183094,
+      "learning_rate": 1.5293809024134314e-05,
+      "loss": 0.2785,
+      "step": 1536
+    },
+    {
+      "epoch": 2.1740976645435244,
+      "grad_norm": 0.19921228761845436,
+      "learning_rate": 1.5267576075550894e-05,
+      "loss": 0.2858,
+      "step": 1537
+    },
+    {
+      "epoch": 2.175513092710545,
+      "grad_norm": 0.2019443429253189,
+      "learning_rate": 1.5241343126967473e-05,
+      "loss": 0.2841,
+      "step": 1538
+    },
+    {
+      "epoch": 2.1769285208775653,
+      "grad_norm": 0.21353560230811622,
+      "learning_rate": 1.5215110178384051e-05,
+      "loss": 0.2857,
+      "step": 1539
+    },
+    {
+      "epoch": 2.178343949044586,
+      "grad_norm": 0.20643770177726536,
+      "learning_rate": 1.518887722980063e-05,
+      "loss": 0.2835,
+      "step": 1540
+    },
+    {
+      "epoch": 2.1797593772116066,
+      "grad_norm": 0.20839208391175837,
+      "learning_rate": 1.5162644281217208e-05,
+      "loss": 0.2777,
+      "step": 1541
+    },
+    {
+      "epoch": 2.181174805378627,
+      "grad_norm": 0.19473202944524465,
+      "learning_rate": 1.5136411332633787e-05,
+      "loss": 0.283,
+      "step": 1542
+    },
+    {
+      "epoch": 2.1825902335456475,
+      "grad_norm": 0.22478890469445573,
+      "learning_rate": 1.5110178384050369e-05,
+      "loss": 0.2784,
+      "step": 1543
+    },
+    {
+      "epoch": 2.184005661712668,
+      "grad_norm": 0.20242980114206965,
+      "learning_rate": 1.5083945435466948e-05,
+      "loss": 0.287,
+      "step": 1544
+    },
+    {
+      "epoch": 2.1854210898796884,
+      "grad_norm": 0.20043047241395306,
+      "learning_rate": 1.5057712486883526e-05,
+      "loss": 0.2873,
+      "step": 1545
+    },
+    {
+      "epoch": 2.186836518046709,
+      "grad_norm": 0.20468599869509715,
+      "learning_rate": 1.5031479538300106e-05,
+      "loss": 0.292,
+      "step": 1546
+    },
+    {
+      "epoch": 2.1882519462137298,
+      "grad_norm": 0.20662087487734626,
+      "learning_rate": 1.5005246589716685e-05,
+      "loss": 0.2941,
+      "step": 1547
+    },
+    {
+      "epoch": 2.18966737438075,
+      "grad_norm": 0.21542200419312985,
+      "learning_rate": 1.4979013641133264e-05,
+      "loss": 0.2834,
+      "step": 1548
+    },
+    {
+      "epoch": 2.1910828025477707,
+      "grad_norm": 0.1863274592133216,
+      "learning_rate": 1.4952780692549842e-05,
+      "loss": 0.2863,
+      "step": 1549
+    },
+    {
+      "epoch": 2.1924982307147913,
+      "grad_norm": 0.19340686279433708,
+      "learning_rate": 1.4926547743966424e-05,
+      "loss": 0.2985,
+      "step": 1550
+    },
+    {
+      "epoch": 2.1939136588818116,
+      "grad_norm": 0.19711186712331613,
+      "learning_rate": 1.4900314795383003e-05,
+      "loss": 0.2883,
+      "step": 1551
+    },
+    {
+      "epoch": 2.1953290870488322,
+      "grad_norm": 0.21712207731412586,
+      "learning_rate": 1.4874081846799581e-05,
+      "loss": 0.279,
+      "step": 1552
+    },
+    {
+      "epoch": 2.196744515215853,
+      "grad_norm": 0.2052315554511649,
+      "learning_rate": 1.484784889821616e-05,
+      "loss": 0.2965,
+      "step": 1553
+    },
+    {
+      "epoch": 2.198159943382873,
+      "grad_norm": 0.23380152085394898,
+      "learning_rate": 1.4821615949632739e-05,
+      "loss": 0.2723,
+      "step": 1554
+    },
+    {
+      "epoch": 2.199575371549894,
+      "grad_norm": 0.19439749822742092,
+      "learning_rate": 1.4795383001049317e-05,
+      "loss": 0.2885,
+      "step": 1555
+    },
+    {
+      "epoch": 2.2009907997169145,
+      "grad_norm": 0.2208178838349699,
+      "learning_rate": 1.4769150052465897e-05,
+      "loss": 0.2922,
+      "step": 1556
+    },
+    {
+      "epoch": 2.2024062278839347,
+      "grad_norm": 0.18946801359847978,
+      "learning_rate": 1.4742917103882478e-05,
+      "loss": 0.2816,
+      "step": 1557
+    },
+    {
+      "epoch": 2.2038216560509554,
+      "grad_norm": 0.24766768111708512,
+      "learning_rate": 1.4716684155299056e-05,
+      "loss": 0.2803,
+      "step": 1558
+    },
+    {
+      "epoch": 2.205237084217976,
+      "grad_norm": 0.19010237210609238,
+      "learning_rate": 1.4690451206715637e-05,
+      "loss": 0.2882,
+      "step": 1559
+    },
+    {
+      "epoch": 2.2066525123849963,
+      "grad_norm": 0.2079249719794975,
+      "learning_rate": 1.4664218258132215e-05,
+      "loss": 0.2882,
+      "step": 1560
+    },
+    {
+      "epoch": 2.208067940552017,
+      "grad_norm": 0.1886838074370487,
+      "learning_rate": 1.4637985309548794e-05,
+      "loss": 0.3127,
+      "step": 1561
+    },
+    {
+      "epoch": 2.2094833687190376,
+      "grad_norm": 0.20139912485284628,
+      "learning_rate": 1.4611752360965372e-05,
+      "loss": 0.313,
+      "step": 1562
+    },
+    {
+      "epoch": 2.210898796886058,
+      "grad_norm": 0.19938539644589226,
+      "learning_rate": 1.4585519412381951e-05,
+      "loss": 0.2859,
+      "step": 1563
+    },
+    {
+      "epoch": 2.2123142250530785,
+      "grad_norm": 0.191817556837941,
+      "learning_rate": 1.4559286463798533e-05,
+      "loss": 0.2841,
+      "step": 1564
+    },
+    {
+      "epoch": 2.213729653220099,
+      "grad_norm": 0.19815427633012597,
+      "learning_rate": 1.4533053515215112e-05,
+      "loss": 0.2806,
+      "step": 1565
+    },
+    {
+      "epoch": 2.2151450813871194,
+      "grad_norm": 0.18164242679367584,
+      "learning_rate": 1.450682056663169e-05,
+      "loss": 0.2743,
+      "step": 1566
+    },
+    {
+      "epoch": 2.21656050955414,
+      "grad_norm": 0.2032830666017277,
+      "learning_rate": 1.4480587618048269e-05,
+      "loss": 0.2957,
+      "step": 1567
+    },
+    {
+      "epoch": 2.2179759377211608,
+      "grad_norm": 0.19073119961731047,
+      "learning_rate": 1.4454354669464847e-05,
+      "loss": 0.2887,
+      "step": 1568
+    },
+    {
+      "epoch": 2.219391365888181,
+      "grad_norm": 0.1847595315159744,
+      "learning_rate": 1.4428121720881428e-05,
+      "loss": 0.2919,
+      "step": 1569
+    },
+    {
+      "epoch": 2.2208067940552016,
+      "grad_norm": 0.1917366937540336,
+      "learning_rate": 1.4401888772298008e-05,
+      "loss": 0.294,
+      "step": 1570
+    },
+    {
+      "epoch": 2.2222222222222223,
+      "grad_norm": 0.20226926426518502,
+      "learning_rate": 1.4375655823714586e-05,
+      "loss": 0.2882,
+      "step": 1571
+    },
+    {
+      "epoch": 2.2236376503892425,
+      "grad_norm": 0.201197899603766,
+      "learning_rate": 1.4349422875131167e-05,
+      "loss": 0.2828,
+      "step": 1572
+    },
+    {
+      "epoch": 2.225053078556263,
+      "grad_norm": 0.18603936110751054,
+      "learning_rate": 1.4323189926547745e-05,
+      "loss": 0.2827,
+      "step": 1573
+    },
+    {
+      "epoch": 2.226468506723284,
+      "grad_norm": 0.19669786047984944,
+      "learning_rate": 1.4296956977964324e-05,
+      "loss": 0.288,
+      "step": 1574
+    },
+    {
+      "epoch": 2.227883934890304,
+      "grad_norm": 0.19406508734392347,
+      "learning_rate": 1.4270724029380902e-05,
+      "loss": 0.2812,
+      "step": 1575
+    },
+    {
+      "epoch": 2.229299363057325,
+      "grad_norm": 0.18691275882541675,
+      "learning_rate": 1.4244491080797481e-05,
+      "loss": 0.2815,
+      "step": 1576
+    },
+    {
+      "epoch": 2.2307147912243455,
+      "grad_norm": 0.2178364402165408,
+      "learning_rate": 1.4218258132214063e-05,
+      "loss": 0.3096,
+      "step": 1577
+    },
+    {
+      "epoch": 2.2321302193913657,
+      "grad_norm": 0.19373188157453802,
+      "learning_rate": 1.4192025183630642e-05,
+      "loss": 0.301,
+      "step": 1578
+    },
+    {
+      "epoch": 2.2335456475583864,
+      "grad_norm": 0.2056928076228888,
+      "learning_rate": 1.416579223504722e-05,
+      "loss": 0.2857,
+      "step": 1579
+    },
+    {
+      "epoch": 2.234961075725407,
+      "grad_norm": 0.18382361352956086,
+      "learning_rate": 1.4139559286463799e-05,
+      "loss": 0.2864,
+      "step": 1580
+    },
+    {
+      "epoch": 2.2363765038924273,
+      "grad_norm": 0.20618476503814526,
+      "learning_rate": 1.4113326337880377e-05,
+      "loss": 0.2878,
+      "step": 1581
+    },
+    {
+      "epoch": 2.237791932059448,
+      "grad_norm": 0.1937950024125009,
+      "learning_rate": 1.4087093389296958e-05,
+      "loss": 0.268,
+      "step": 1582
+    },
+    {
+      "epoch": 2.2392073602264686,
+      "grad_norm": 0.19092954796711575,
+      "learning_rate": 1.4060860440713536e-05,
+      "loss": 0.288,
+      "step": 1583
+    },
+    {
+      "epoch": 2.240622788393489,
+      "grad_norm": 0.19717194852662742,
+      "learning_rate": 1.4034627492130117e-05,
+      "loss": 0.3031,
+      "step": 1584
+    },
+    {
+      "epoch": 2.2420382165605095,
+      "grad_norm": 0.1902434038668803,
+      "learning_rate": 1.4008394543546697e-05,
+      "loss": 0.2884,
+      "step": 1585
+    },
+    {
+      "epoch": 2.24345364472753,
+      "grad_norm": 0.18215832190034617,
+      "learning_rate": 1.3982161594963275e-05,
+      "loss": 0.2759,
+      "step": 1586
+    },
+    {
+      "epoch": 2.2448690728945504,
+      "grad_norm": 0.18167957909640053,
+      "learning_rate": 1.3955928646379854e-05,
+      "loss": 0.2719,
+      "step": 1587
+    },
+    {
+      "epoch": 2.246284501061571,
+      "grad_norm": 0.20532991110898174,
+      "learning_rate": 1.3929695697796433e-05,
+      "loss": 0.2805,
+      "step": 1588
+    },
+    {
+      "epoch": 2.2476999292285917,
+      "grad_norm": 0.20280748433545553,
+      "learning_rate": 1.3903462749213011e-05,
+      "loss": 0.2945,
+      "step": 1589
+    },
+    {
+      "epoch": 2.249115357395612,
+      "grad_norm": 0.2102414685159178,
+      "learning_rate": 1.387722980062959e-05,
+      "loss": 0.2879,
+      "step": 1590
+    },
+    {
+      "epoch": 2.2505307855626326,
+      "grad_norm": 0.2069270659647197,
+      "learning_rate": 1.3850996852046172e-05,
+      "loss": 0.2771,
+      "step": 1591
+    },
+    {
+      "epoch": 2.2519462137296533,
+      "grad_norm": 0.21114683188258085,
+      "learning_rate": 1.382476390346275e-05,
+      "loss": 0.2864,
+      "step": 1592
+    },
+    {
+      "epoch": 2.2533616418966735,
+      "grad_norm": 0.1907649784482529,
+      "learning_rate": 1.3798530954879329e-05,
+      "loss": 0.2933,
+      "step": 1593
+    },
+    {
+      "epoch": 2.254777070063694,
+      "grad_norm": 0.18513807889980596,
+      "learning_rate": 1.3772298006295908e-05,
+      "loss": 0.2662,
+      "step": 1594
+    },
+    {
+      "epoch": 2.256192498230715,
+      "grad_norm": 0.20949970921326494,
+      "learning_rate": 1.3746065057712488e-05,
+      "loss": 0.2854,
+      "step": 1595
+    },
+    {
+      "epoch": 2.2576079263977356,
+      "grad_norm": 0.20091183098409918,
+      "learning_rate": 1.3719832109129066e-05,
+      "loss": 0.2994,
+      "step": 1596
+    },
+    {
+      "epoch": 2.259023354564756,
+      "grad_norm": 0.1859277391528771,
+      "learning_rate": 1.3693599160545645e-05,
+      "loss": 0.2808,
+      "step": 1597
+    },
+    {
+      "epoch": 2.2604387827317765,
+      "grad_norm": 0.18520117324638943,
+      "learning_rate": 1.3667366211962227e-05,
+      "loss": 0.2831,
+      "step": 1598
+    },
+    {
+      "epoch": 2.2618542108987967,
+      "grad_norm": 0.18833415298828882,
+      "learning_rate": 1.3641133263378806e-05,
+      "loss": 0.2871,
+      "step": 1599
+    },
+    {
+      "epoch": 2.2632696390658174,
+      "grad_norm": 0.18848821621206596,
+      "learning_rate": 1.3614900314795384e-05,
+      "loss": 0.2766,
+      "step": 1600
+    },
+    {
+      "epoch": 2.264685067232838,
+      "grad_norm": 0.2024115235104692,
+      "learning_rate": 1.3588667366211963e-05,
+      "loss": 0.2922,
+      "step": 1601
+    },
+    {
+      "epoch": 2.2661004953998587,
+      "grad_norm": 0.2008973713899374,
+      "learning_rate": 1.3562434417628541e-05,
+      "loss": 0.2962,
+      "step": 1602
+    },
+    {
+      "epoch": 2.267515923566879,
+      "grad_norm": 0.20729689011139363,
+      "learning_rate": 1.353620146904512e-05,
+      "loss": 0.2843,
+      "step": 1603
+    },
+    {
+      "epoch": 2.2689313517338996,
+      "grad_norm": 0.21523153306999607,
+      "learning_rate": 1.3509968520461699e-05,
+      "loss": 0.2742,
+      "step": 1604
+    },
+    {
+      "epoch": 2.27034677990092,
+      "grad_norm": 0.18912304205705838,
+      "learning_rate": 1.348373557187828e-05,
+      "loss": 0.2778,
+      "step": 1605
+    },
+    {
+      "epoch": 2.2717622080679405,
+      "grad_norm": 0.1974380153859812,
+      "learning_rate": 1.3457502623294859e-05,
+      "loss": 0.2868,
+      "step": 1606
+    },
+    {
+      "epoch": 2.273177636234961,
+      "grad_norm": 0.20180879535619137,
+      "learning_rate": 1.343126967471144e-05,
+      "loss": 0.2934,
+      "step": 1607
+    },
+    {
+      "epoch": 2.274593064401982,
+      "grad_norm": 0.19163665717120043,
+      "learning_rate": 1.3405036726128018e-05,
+      "loss": 0.2856,
+      "step": 1608
+    },
+    {
+      "epoch": 2.276008492569002,
+      "grad_norm": 0.22024265924133277,
+      "learning_rate": 1.3378803777544597e-05,
+      "loss": 0.2899,
+      "step": 1609
+    },
+    {
+      "epoch": 2.2774239207360227,
+      "grad_norm": 0.19289156194429993,
+      "learning_rate": 1.3352570828961175e-05,
+      "loss": 0.2793,
+      "step": 1610
+    },
+    {
+      "epoch": 2.278839348903043,
+      "grad_norm": 0.19600954795335204,
+      "learning_rate": 1.3326337880377754e-05,
+      "loss": 0.2766,
+      "step": 1611
+    },
+    {
+      "epoch": 2.2802547770700636,
+      "grad_norm": 0.20924587152335908,
+      "learning_rate": 1.3300104931794336e-05,
+      "loss": 0.2948,
+      "step": 1612
+    },
+    {
+      "epoch": 2.2816702052370843,
+      "grad_norm": 0.19653291000591008,
+      "learning_rate": 1.3273871983210914e-05,
+      "loss": 0.2939,
+      "step": 1613
+    },
+    {
+      "epoch": 2.283085633404105,
+      "grad_norm": 0.1910918798595604,
+      "learning_rate": 1.3247639034627493e-05,
+      "loss": 0.2743,
+      "step": 1614
+    },
+    {
+      "epoch": 2.284501061571125,
+      "grad_norm": 0.20205394590303571,
+      "learning_rate": 1.3221406086044072e-05,
+      "loss": 0.3006,
+      "step": 1615
+    },
+    {
+      "epoch": 2.285916489738146,
+      "grad_norm": 0.2060297460654065,
+      "learning_rate": 1.319517313746065e-05,
+      "loss": 0.2865,
+      "step": 1616
+    },
+    {
+      "epoch": 2.287331917905166,
+      "grad_norm": 0.18498862483625148,
+      "learning_rate": 1.3168940188877229e-05,
+      "loss": 0.2745,
+      "step": 1617
+    },
+    {
+      "epoch": 2.2887473460721868,
+      "grad_norm": 0.19030723917171546,
+      "learning_rate": 1.3142707240293809e-05,
+      "loss": 0.2827,
+      "step": 1618
+    },
+    {
+      "epoch": 2.2901627742392074,
+      "grad_norm": 0.18388847175027578,
+      "learning_rate": 1.311647429171039e-05,
+      "loss": 0.2726,
+      "step": 1619
+    },
+    {
+      "epoch": 2.291578202406228,
+      "grad_norm": 0.21625030859649072,
+      "learning_rate": 1.309024134312697e-05,
+      "loss": 0.2859,
+      "step": 1620
+    },
+    {
+      "epoch": 2.2929936305732483,
+      "grad_norm": 0.20451720977139845,
+      "learning_rate": 1.3064008394543548e-05,
+      "loss": 0.2933,
+      "step": 1621
+    },
+    {
+      "epoch": 2.294409058740269,
+      "grad_norm": 0.20094353353440728,
+      "learning_rate": 1.3037775445960127e-05,
+      "loss": 0.2929,
+      "step": 1622
+    },
+    {
+      "epoch": 2.2958244869072892,
+      "grad_norm": 0.19957600133289297,
+      "learning_rate": 1.3011542497376705e-05,
+      "loss": 0.2934,
+      "step": 1623
+    },
+    {
+      "epoch": 2.29723991507431,
+      "grad_norm": 0.18883357366458575,
+      "learning_rate": 1.2985309548793284e-05,
+      "loss": 0.2744,
+      "step": 1624
+    },
+    {
+      "epoch": 2.2986553432413306,
+      "grad_norm": 0.20674024088486972,
+      "learning_rate": 1.2959076600209863e-05,
+      "loss": 0.2998,
+      "step": 1625
+    },
+    {
+      "epoch": 2.3000707714083513,
+      "grad_norm": 0.21630882145739538,
+      "learning_rate": 1.2932843651626445e-05,
+      "loss": 0.3008,
+      "step": 1626
+    },
+    {
+      "epoch": 2.3014861995753715,
+      "grad_norm": 0.18992728343898063,
+      "learning_rate": 1.2906610703043023e-05,
+      "loss": 0.2934,
+      "step": 1627
+    },
+    {
+      "epoch": 2.302901627742392,
+      "grad_norm": 0.19675282780344314,
+      "learning_rate": 1.2880377754459602e-05,
+      "loss": 0.2974,
+      "step": 1628
+    },
+    {
+      "epoch": 2.3043170559094124,
+      "grad_norm": 0.23728782889287953,
+      "learning_rate": 1.285414480587618e-05,
+      "loss": 0.2844,
+      "step": 1629
+    },
+    {
+      "epoch": 2.305732484076433,
+      "grad_norm": 0.18387211272773954,
+      "learning_rate": 1.282791185729276e-05,
+      "loss": 0.2732,
+      "step": 1630
+    },
+    {
+      "epoch": 2.3071479122434537,
+      "grad_norm": 0.18857157563443108,
+      "learning_rate": 1.280167890870934e-05,
+      "loss": 0.284,
+      "step": 1631
+    },
+    {
+      "epoch": 2.3085633404104744,
+      "grad_norm": 0.19182401108083405,
+      "learning_rate": 1.2775445960125918e-05,
+      "loss": 0.28,
+      "step": 1632
+    },
+    {
+      "epoch": 2.3099787685774946,
+      "grad_norm": 0.18806438910368767,
+      "learning_rate": 1.27492130115425e-05,
+      "loss": 0.2669,
+      "step": 1633
+    },
+    {
+      "epoch": 2.3113941967445153,
+      "grad_norm": 0.19240347359549476,
+      "learning_rate": 1.2722980062959078e-05,
+      "loss": 0.2786,
+      "step": 1634
+    },
+    {
+      "epoch": 2.3128096249115355,
+      "grad_norm": 0.18641527282755707,
+      "learning_rate": 1.2696747114375657e-05,
+      "loss": 0.2711,
+      "step": 1635
+    },
+    {
+      "epoch": 2.314225053078556,
+      "grad_norm": 0.18234644719608498,
+      "learning_rate": 1.2670514165792236e-05,
+      "loss": 0.2801,
+      "step": 1636
+    },
+    {
+      "epoch": 2.315640481245577,
+      "grad_norm": 0.20968979176034708,
+      "learning_rate": 1.2644281217208814e-05,
+      "loss": 0.289,
+      "step": 1637
+    },
+    {
+      "epoch": 2.3170559094125975,
+      "grad_norm": 0.18806510420148745,
+      "learning_rate": 1.2618048268625393e-05,
+      "loss": 0.2917,
+      "step": 1638
+    },
+    {
+      "epoch": 2.3184713375796178,
+      "grad_norm": 0.2029698510668608,
+      "learning_rate": 1.2591815320041971e-05,
+      "loss": 0.2953,
+      "step": 1639
+    },
+    {
+      "epoch": 2.3198867657466384,
+      "grad_norm": 0.19660392567887017,
+      "learning_rate": 1.2565582371458553e-05,
+      "loss": 0.2721,
+      "step": 1640
+    },
+    {
+      "epoch": 2.3213021939136587,
+      "grad_norm": 0.19707921251142746,
+      "learning_rate": 1.2539349422875132e-05,
+      "loss": 0.2911,
+      "step": 1641
+    },
+    {
+      "epoch": 2.3227176220806793,
+      "grad_norm": 0.18014931048280203,
+      "learning_rate": 1.251311647429171e-05,
+      "loss": 0.2837,
+      "step": 1642
+    },
+    {
+      "epoch": 2.3241330502477,
+      "grad_norm": 0.20230753017131037,
+      "learning_rate": 1.248688352570829e-05,
+      "loss": 0.2878,
+      "step": 1643
+    },
+    {
+      "epoch": 2.3255484784147207,
+      "grad_norm": 0.1986302628772248,
+      "learning_rate": 1.246065057712487e-05,
+      "loss": 0.2833,
+      "step": 1644
+    },
+    {
+      "epoch": 2.326963906581741,
+      "grad_norm": 0.19288103271468032,
+      "learning_rate": 1.243441762854145e-05,
+      "loss": 0.2886,
+      "step": 1645
+    },
+    {
+      "epoch": 2.3283793347487616,
+      "grad_norm": 0.19483944520797117,
+      "learning_rate": 1.2408184679958028e-05,
+      "loss": 0.2877,
+      "step": 1646
+    },
+    {
+      "epoch": 2.329794762915782,
+      "grad_norm": 0.18771203694792818,
+      "learning_rate": 1.2381951731374607e-05,
+      "loss": 0.2875,
+      "step": 1647
+    },
+    {
+      "epoch": 2.3312101910828025,
+      "grad_norm": 0.19312612453223374,
+      "learning_rate": 1.2355718782791187e-05,
+      "loss": 0.2855,
+      "step": 1648
+    },
+    {
+      "epoch": 2.332625619249823,
+      "grad_norm": 0.18881242383849123,
+      "learning_rate": 1.2329485834207766e-05,
+      "loss": 0.2764,
+      "step": 1649
+    },
+    {
+      "epoch": 2.334041047416844,
+      "grad_norm": 0.18098008583019026,
+      "learning_rate": 1.2303252885624344e-05,
+      "loss": 0.2786,
+      "step": 1650
+    },
+    {
+      "epoch": 2.335456475583864,
+      "grad_norm": 0.19492465692066005,
+      "learning_rate": 1.2277019937040923e-05,
+      "loss": 0.2912,
+      "step": 1651
+    },
+    {
+      "epoch": 2.3368719037508847,
+      "grad_norm": 0.20935236412990635,
+      "learning_rate": 1.2250786988457503e-05,
+      "loss": 0.2895,
+      "step": 1652
+    },
+    {
+      "epoch": 2.338287331917905,
+      "grad_norm": 0.18797260869911742,
+      "learning_rate": 1.2224554039874082e-05,
+      "loss": 0.2892,
+      "step": 1653
+    },
+    {
+      "epoch": 2.3397027600849256,
+      "grad_norm": 0.1991671401529202,
+      "learning_rate": 1.2198321091290662e-05,
+      "loss": 0.2942,
+      "step": 1654
+    },
+    {
+      "epoch": 2.3411181882519463,
+      "grad_norm": 0.18970334539112932,
+      "learning_rate": 1.217208814270724e-05,
+      "loss": 0.2774,
+      "step": 1655
+    },
+    {
+      "epoch": 2.342533616418967,
+      "grad_norm": 0.1872611883631897,
+      "learning_rate": 1.2145855194123821e-05,
+      "loss": 0.2861,
+      "step": 1656
+    },
+    {
+      "epoch": 2.343949044585987,
+      "grad_norm": 0.20505699637426708,
+      "learning_rate": 1.21196222455404e-05,
+      "loss": 0.3072,
+      "step": 1657
+    },
+    {
+      "epoch": 2.345364472753008,
+      "grad_norm": 0.19186506797113723,
+      "learning_rate": 1.2093389296956978e-05,
+      "loss": 0.2971,
+      "step": 1658
+    },
+    {
+      "epoch": 2.346779900920028,
+      "grad_norm": 0.18294014123423893,
+      "learning_rate": 1.2067156348373558e-05,
+      "loss": 0.2836,
+      "step": 1659
+    },
+    {
+      "epoch": 2.3481953290870488,
+      "grad_norm": 0.18493772369567185,
+      "learning_rate": 1.2040923399790137e-05,
+      "loss": 0.2751,
+      "step": 1660
+    },
+    {
+      "epoch": 2.3496107572540694,
+      "grad_norm": 0.20501585148695503,
+      "learning_rate": 1.2014690451206716e-05,
+      "loss": 0.2854,
+      "step": 1661
+    },
+    {
+      "epoch": 2.35102618542109,
+      "grad_norm": 0.18909638186884367,
+      "learning_rate": 1.1988457502623296e-05,
+      "loss": 0.2816,
+      "step": 1662
+    },
+    {
+      "epoch": 2.3524416135881103,
+      "grad_norm": 0.18298287687612652,
+      "learning_rate": 1.1962224554039874e-05,
+      "loss": 0.2691,
+      "step": 1663
+    },
+    {
+      "epoch": 2.353857041755131,
+      "grad_norm": 0.195585784316558,
+      "learning_rate": 1.1935991605456453e-05,
+      "loss": 0.2806,
+      "step": 1664
+    },
+    {
+      "epoch": 2.3552724699221512,
+      "grad_norm": 0.20023242229909904,
+      "learning_rate": 1.1909758656873033e-05,
+      "loss": 0.2874,
+      "step": 1665
+    },
+    {
+      "epoch": 2.356687898089172,
+      "grad_norm": 0.20043783231352735,
+      "learning_rate": 1.1883525708289612e-05,
+      "loss": 0.292,
+      "step": 1666
+    },
+    {
+      "epoch": 2.3581033262561926,
+      "grad_norm": 0.21082009566883989,
+      "learning_rate": 1.1857292759706192e-05,
+      "loss": 0.2885,
+      "step": 1667
+    },
+    {
+      "epoch": 2.3595187544232132,
+      "grad_norm": 0.2045543901472315,
+      "learning_rate": 1.183105981112277e-05,
+      "loss": 0.2948,
+      "step": 1668
+    },
+    {
+      "epoch": 2.3609341825902335,
+      "grad_norm": 0.19600728280903276,
+      "learning_rate": 1.1804826862539351e-05,
+      "loss": 0.3041,
+      "step": 1669
+    },
+    {
+      "epoch": 2.362349610757254,
+      "grad_norm": 0.19437448899997456,
+      "learning_rate": 1.177859391395593e-05,
+      "loss": 0.2829,
+      "step": 1670
+    },
+    {
+      "epoch": 2.3637650389242744,
+      "grad_norm": 0.18823136977189336,
+      "learning_rate": 1.1752360965372508e-05,
+      "loss": 0.2818,
+      "step": 1671
+    },
+    {
+      "epoch": 2.365180467091295,
+      "grad_norm": 0.20404660050537488,
+      "learning_rate": 1.1726128016789089e-05,
+      "loss": 0.2957,
+      "step": 1672
+    },
+    {
+      "epoch": 2.3665958952583157,
+      "grad_norm": 0.18792494181919675,
+      "learning_rate": 1.1699895068205667e-05,
+      "loss": 0.2809,
+      "step": 1673
+    },
+    {
+      "epoch": 2.3680113234253364,
+      "grad_norm": 0.18945851267858987,
+      "learning_rate": 1.1673662119622246e-05,
+      "loss": 0.2946,
+      "step": 1674
+    },
+    {
+      "epoch": 2.3694267515923566,
+      "grad_norm": 0.1975555826468101,
+      "learning_rate": 1.1647429171038824e-05,
+      "loss": 0.2844,
+      "step": 1675
+    },
+    {
+      "epoch": 2.3708421797593773,
+      "grad_norm": 0.19905077106155683,
+      "learning_rate": 1.1621196222455405e-05,
+      "loss": 0.2728,
+      "step": 1676
+    },
+    {
+      "epoch": 2.3722576079263975,
+      "grad_norm": 0.2052488670029528,
+      "learning_rate": 1.1594963273871983e-05,
+      "loss": 0.2962,
+      "step": 1677
+    },
+    {
+      "epoch": 2.373673036093418,
+      "grad_norm": 0.1899601465728956,
+      "learning_rate": 1.1568730325288562e-05,
+      "loss": 0.2802,
+      "step": 1678
+    },
+    {
+      "epoch": 2.375088464260439,
+      "grad_norm": 0.18865472944370884,
+      "learning_rate": 1.1542497376705142e-05,
+      "loss": 0.2928,
+      "step": 1679
+    },
+    {
+      "epoch": 2.3765038924274595,
+      "grad_norm": 0.20141404594053566,
+      "learning_rate": 1.1516264428121722e-05,
+      "loss": 0.2827,
+      "step": 1680
+    },
+    {
+      "epoch": 2.3779193205944797,
+      "grad_norm": 0.2316866248634522,
+      "learning_rate": 1.1490031479538301e-05,
+      "loss": 0.2864,
+      "step": 1681
+    },
+    {
+      "epoch": 2.3793347487615004,
+      "grad_norm": 0.189109282917738,
+      "learning_rate": 1.146379853095488e-05,
+      "loss": 0.2947,
+      "step": 1682
+    },
+    {
+      "epoch": 2.3807501769285206,
+      "grad_norm": 0.18663462596385907,
+      "learning_rate": 1.143756558237146e-05,
+      "loss": 0.2883,
+      "step": 1683
+    },
+    {
+      "epoch": 2.3821656050955413,
+      "grad_norm": 0.24963661794508527,
+      "learning_rate": 1.1411332633788038e-05,
+      "loss": 0.2976,
+      "step": 1684
+    },
+    {
+      "epoch": 2.383581033262562,
+      "grad_norm": 0.20908853758881932,
+      "learning_rate": 1.1385099685204617e-05,
+      "loss": 0.3017,
+      "step": 1685
+    },
+    {
+      "epoch": 2.3849964614295827,
+      "grad_norm": 0.2020369520638962,
+      "learning_rate": 1.1358866736621197e-05,
+      "loss": 0.3039,
+      "step": 1686
+    },
+    {
+      "epoch": 2.386411889596603,
+      "grad_norm": 0.1841556772303126,
+      "learning_rate": 1.1332633788037776e-05,
+      "loss": 0.2749,
+      "step": 1687
+    },
+    {
+      "epoch": 2.3878273177636236,
+      "grad_norm": 0.21285164297603593,
+      "learning_rate": 1.1306400839454354e-05,
+      "loss": 0.2887,
+      "step": 1688
+    },
+    {
+      "epoch": 2.389242745930644,
+      "grad_norm": 0.20583511099567137,
+      "learning_rate": 1.1280167890870933e-05,
+      "loss": 0.2788,
+      "step": 1689
+    },
+    {
+      "epoch": 2.3906581740976645,
+      "grad_norm": 0.20708969381375833,
+      "learning_rate": 1.1253934942287513e-05,
+      "loss": 0.2765,
+      "step": 1690
+    },
+    {
+      "epoch": 2.392073602264685,
+      "grad_norm": 0.19666467582284977,
+      "learning_rate": 1.1227701993704094e-05,
+      "loss": 0.2926,
+      "step": 1691
+    },
+    {
+      "epoch": 2.393489030431706,
+      "grad_norm": 0.20505521755851502,
+      "learning_rate": 1.1201469045120672e-05,
+      "loss": 0.2841,
+      "step": 1692
+    },
+    {
+      "epoch": 2.394904458598726,
+      "grad_norm": 0.21731096035292308,
+      "learning_rate": 1.1175236096537252e-05,
+      "loss": 0.2868,
+      "step": 1693
+    },
+    {
+      "epoch": 2.3963198867657467,
+      "grad_norm": 0.18965142028058407,
+      "learning_rate": 1.1149003147953831e-05,
+      "loss": 0.2978,
+      "step": 1694
+    },
+    {
+      "epoch": 2.397735314932767,
+      "grad_norm": 0.2058365142452598,
+      "learning_rate": 1.112277019937041e-05,
+      "loss": 0.2912,
+      "step": 1695
+    },
+    {
+      "epoch": 2.3991507430997876,
+      "grad_norm": 0.20940864332061368,
+      "learning_rate": 1.1096537250786988e-05,
+      "loss": 0.3,
+      "step": 1696
+    },
+    {
+      "epoch": 2.4005661712668083,
+      "grad_norm": 0.20291991491379344,
+      "learning_rate": 1.1070304302203569e-05,
+      "loss": 0.2831,
+      "step": 1697
+    },
+    {
+      "epoch": 2.401981599433829,
+      "grad_norm": 0.20138578529128853,
+      "learning_rate": 1.1044071353620147e-05,
+      "loss": 0.2968,
+      "step": 1698
+    },
+    {
+      "epoch": 2.403397027600849,
+      "grad_norm": 0.19224993627720924,
+      "learning_rate": 1.1017838405036726e-05,
+      "loss": 0.2655,
+      "step": 1699
+    },
+    {
+      "epoch": 2.40481245576787,
+      "grad_norm": 0.19719007475383227,
+      "learning_rate": 1.0991605456453306e-05,
+      "loss": 0.2792,
+      "step": 1700
+    },
+    {
+      "epoch": 2.40622788393489,
+      "grad_norm": 0.21831494195525009,
+      "learning_rate": 1.0965372507869885e-05,
+      "loss": 0.2985,
+      "step": 1701
+    },
+    {
+      "epoch": 2.4076433121019107,
+      "grad_norm": 0.19020846891844465,
+      "learning_rate": 1.0939139559286463e-05,
+      "loss": 0.2893,
+      "step": 1702
+    },
+    {
+      "epoch": 2.4090587402689314,
+      "grad_norm": 0.18289524936418528,
+      "learning_rate": 1.0912906610703043e-05,
+      "loss": 0.2851,
+      "step": 1703
+    },
+    {
+      "epoch": 2.410474168435952,
+      "grad_norm": 0.1974186204137678,
+      "learning_rate": 1.0886673662119624e-05,
+      "loss": 0.2874,
+      "step": 1704
+    },
+    {
+      "epoch": 2.4118895966029723,
+      "grad_norm": 0.18629571329619682,
+      "learning_rate": 1.0860440713536202e-05,
+      "loss": 0.2865,
+      "step": 1705
+    },
+    {
+      "epoch": 2.413305024769993,
+      "grad_norm": 0.19587811877631692,
+      "learning_rate": 1.0834207764952781e-05,
+      "loss": 0.3,
+      "step": 1706
+    },
+    {
+      "epoch": 2.414720452937013,
+      "grad_norm": 0.19078404188550765,
+      "learning_rate": 1.0807974816369361e-05,
+      "loss": 0.2805,
+      "step": 1707
+    },
+    {
+      "epoch": 2.416135881104034,
+      "grad_norm": 0.18334271635230978,
+      "learning_rate": 1.078174186778594e-05,
+      "loss": 0.2879,
+      "step": 1708
+    },
+    {
+      "epoch": 2.4175513092710545,
+      "grad_norm": 0.2099964101240209,
+      "learning_rate": 1.0755508919202518e-05,
+      "loss": 0.2964,
+      "step": 1709
+    },
+    {
+      "epoch": 2.418966737438075,
+      "grad_norm": 0.18778422256175833,
+      "learning_rate": 1.0729275970619099e-05,
+      "loss": 0.2811,
+      "step": 1710
+    },
+    {
+      "epoch": 2.4203821656050954,
+      "grad_norm": 0.19030408167368482,
+      "learning_rate": 1.0703043022035677e-05,
+      "loss": 0.2849,
+      "step": 1711
+    },
+    {
+      "epoch": 2.421797593772116,
+      "grad_norm": 0.3319292426618223,
+      "learning_rate": 1.0676810073452256e-05,
+      "loss": 0.2809,
+      "step": 1712
+    },
+    {
+      "epoch": 2.4232130219391363,
+      "grad_norm": 0.20408958072566255,
+      "learning_rate": 1.0650577124868834e-05,
+      "loss": 0.2842,
+      "step": 1713
+    },
+    {
+      "epoch": 2.424628450106157,
+      "grad_norm": 0.20386837831449744,
+      "learning_rate": 1.0624344176285415e-05,
+      "loss": 0.2868,
+      "step": 1714
+    },
+    {
+      "epoch": 2.4260438782731777,
+      "grad_norm": 0.18784192812808573,
+      "learning_rate": 1.0598111227701995e-05,
+      "loss": 0.2794,
+      "step": 1715
+    },
+    {
+      "epoch": 2.4274593064401984,
+      "grad_norm": 0.1925027128688008,
+      "learning_rate": 1.0571878279118574e-05,
+      "loss": 0.2854,
+      "step": 1716
+    },
+    {
+      "epoch": 2.4288747346072186,
+      "grad_norm": 0.1865697450118313,
+      "learning_rate": 1.0545645330535154e-05,
+      "loss": 0.2825,
+      "step": 1717
+    },
+    {
+      "epoch": 2.4302901627742393,
+      "grad_norm": 0.19960749213574797,
+      "learning_rate": 1.0519412381951733e-05,
+      "loss": 0.2879,
+      "step": 1718
+    },
+    {
+      "epoch": 2.4317055909412595,
+      "grad_norm": 0.20678638934566496,
+      "learning_rate": 1.0493179433368311e-05,
+      "loss": 0.2828,
+      "step": 1719
+    },
+    {
+      "epoch": 2.43312101910828,
+      "grad_norm": 0.19170488624048732,
+      "learning_rate": 1.046694648478489e-05,
+      "loss": 0.2825,
+      "step": 1720
+    },
+    {
+      "epoch": 2.434536447275301,
+      "grad_norm": 0.21365876250706559,
+      "learning_rate": 1.044071353620147e-05,
+      "loss": 0.2855,
+      "step": 1721
+    },
+    {
+      "epoch": 2.4359518754423215,
+      "grad_norm": 0.20008633956473132,
+      "learning_rate": 1.0414480587618049e-05,
+      "loss": 0.2826,
+      "step": 1722
+    },
+    {
+      "epoch": 2.4373673036093417,
+      "grad_norm": 0.2033193173356251,
+      "learning_rate": 1.0388247639034627e-05,
+      "loss": 0.2825,
+      "step": 1723
+    },
+    {
+      "epoch": 2.4387827317763624,
+      "grad_norm": 0.18756102875592565,
+      "learning_rate": 1.0362014690451207e-05,
+      "loss": 0.2983,
+      "step": 1724
+    },
+    {
+      "epoch": 2.4401981599433826,
+      "grad_norm": 0.2177383215119078,
+      "learning_rate": 1.0335781741867786e-05,
+      "loss": 0.2867,
+      "step": 1725
+    },
+    {
+      "epoch": 2.4416135881104033,
+      "grad_norm": 0.19994004702139706,
+      "learning_rate": 1.0309548793284365e-05,
+      "loss": 0.2782,
+      "step": 1726
+    },
+    {
+      "epoch": 2.443029016277424,
+      "grad_norm": 0.19165837212395614,
+      "learning_rate": 1.0283315844700945e-05,
+      "loss": 0.2776,
+      "step": 1727
+    },
+    {
+      "epoch": 2.4444444444444446,
+      "grad_norm": 0.19445758983775668,
+      "learning_rate": 1.0257082896117525e-05,
+      "loss": 0.3013,
+      "step": 1728
+    },
+    {
+      "epoch": 2.445859872611465,
+      "grad_norm": 0.2007145447916944,
+      "learning_rate": 1.0230849947534104e-05,
+      "loss": 0.2962,
+      "step": 1729
+    },
+    {
+      "epoch": 2.4472753007784855,
+      "grad_norm": 0.21240393799613172,
+      "learning_rate": 1.0204616998950682e-05,
+      "loss": 0.2841,
+      "step": 1730
+    },
+    {
+      "epoch": 2.4486907289455058,
+      "grad_norm": 0.19228717160213668,
+      "learning_rate": 1.0178384050367263e-05,
+      "loss": 0.2866,
+      "step": 1731
+    },
+    {
+      "epoch": 2.4501061571125264,
+      "grad_norm": 0.1854903433710835,
+      "learning_rate": 1.0152151101783841e-05,
+      "loss": 0.2769,
+      "step": 1732
+    },
+    {
+      "epoch": 2.451521585279547,
+      "grad_norm": 0.20102508537794403,
+      "learning_rate": 1.012591815320042e-05,
+      "loss": 0.2861,
+      "step": 1733
+    },
+    {
+      "epoch": 2.452937013446568,
+      "grad_norm": 0.1941395854546883,
+      "learning_rate": 1.0099685204616998e-05,
+      "loss": 0.2836,
+      "step": 1734
+    },
+    {
+      "epoch": 2.454352441613588,
+      "grad_norm": 0.2028174742597168,
+      "learning_rate": 1.0073452256033579e-05,
+      "loss": 0.294,
+      "step": 1735
+    },
+    {
+      "epoch": 2.4557678697806087,
+      "grad_norm": 0.1981830261166878,
+      "learning_rate": 1.0047219307450157e-05,
+      "loss": 0.277,
+      "step": 1736
+    },
+    {
+      "epoch": 2.457183297947629,
+      "grad_norm": 0.20093369675985923,
+      "learning_rate": 1.0020986358866736e-05,
+      "loss": 0.2955,
+      "step": 1737
+    },
+    {
+      "epoch": 2.4585987261146496,
+      "grad_norm": 0.2001686805200334,
+      "learning_rate": 9.994753410283316e-06,
+      "loss": 0.299,
+      "step": 1738
+    },
+    {
+      "epoch": 2.4600141542816703,
+      "grad_norm": 0.19054589682313114,
+      "learning_rate": 9.968520461699895e-06,
+      "loss": 0.2722,
+      "step": 1739
+    },
+    {
+      "epoch": 2.461429582448691,
+      "grad_norm": 0.17379130432862766,
+      "learning_rate": 9.942287513116475e-06,
+      "loss": 0.2694,
+      "step": 1740
+    },
+    {
+      "epoch": 2.462845010615711,
+      "grad_norm": 0.18743548991095602,
+      "learning_rate": 9.916054564533055e-06,
+      "loss": 0.2811,
+      "step": 1741
+    },
+    {
+      "epoch": 2.464260438782732,
+      "grad_norm": 0.19612228488550978,
+      "learning_rate": 9.889821615949634e-06,
+      "loss": 0.2828,
+      "step": 1742
+    },
+    {
+      "epoch": 2.4656758669497525,
+      "grad_norm": 0.19257011195751622,
+      "learning_rate": 9.863588667366213e-06,
+      "loss": 0.2823,
+      "step": 1743
+    },
+    {
+      "epoch": 2.4670912951167727,
+      "grad_norm": 0.19551678907161937,
+      "learning_rate": 9.837355718782791e-06,
+      "loss": 0.2702,
+      "step": 1744
+    },
+    {
+      "epoch": 2.4685067232837934,
+      "grad_norm": 0.19172581322744886,
+      "learning_rate": 9.811122770199371e-06,
+      "loss": 0.288,
+      "step": 1745
+    },
+    {
+      "epoch": 2.469922151450814,
+      "grad_norm": 0.1774729926191194,
+      "learning_rate": 9.78488982161595e-06,
+      "loss": 0.2771,
+      "step": 1746
+    },
+    {
+      "epoch": 2.4713375796178343,
+      "grad_norm": 0.18709764194534617,
+      "learning_rate": 9.758656873032529e-06,
+      "loss": 0.279,
+      "step": 1747
+    },
+    {
+      "epoch": 2.472753007784855,
+      "grad_norm": 0.1783885544112287,
+      "learning_rate": 9.732423924449109e-06,
+      "loss": 0.2776,
+      "step": 1748
+    },
+    {
+      "epoch": 2.4741684359518756,
+      "grad_norm": 0.19330769319927216,
+      "learning_rate": 9.706190975865687e-06,
+      "loss": 0.3033,
+      "step": 1749
+    },
+    {
+      "epoch": 2.475583864118896,
+      "grad_norm": 0.1951366237517273,
+      "learning_rate": 9.679958027282266e-06,
+      "loss": 0.2829,
+      "step": 1750
+    },
+    {
+      "epoch": 2.4769992922859165,
+      "grad_norm": 0.18665621657459827,
+      "learning_rate": 9.653725078698846e-06,
+      "loss": 0.2905,
+      "step": 1751
+    },
+    {
+      "epoch": 2.478414720452937,
+      "grad_norm": 0.18424743703731458,
+      "learning_rate": 9.627492130115427e-06,
+      "loss": 0.2693,
+      "step": 1752
+    },
+    {
+      "epoch": 2.4798301486199574,
+      "grad_norm": 0.19089555732541547,
+      "learning_rate": 9.601259181532005e-06,
+      "loss": 0.2867,
+      "step": 1753
+    },
+    {
+      "epoch": 2.481245576786978,
+      "grad_norm": 0.17861854716249345,
+      "learning_rate": 9.575026232948584e-06,
+      "loss": 0.288,
+      "step": 1754
+    },
+    {
+      "epoch": 2.4826610049539988,
+      "grad_norm": 0.1721315445775327,
+      "learning_rate": 9.548793284365164e-06,
+      "loss": 0.2746,
+      "step": 1755
+    },
+    {
+      "epoch": 2.484076433121019,
+      "grad_norm": 0.187265615065389,
+      "learning_rate": 9.522560335781743e-06,
+      "loss": 0.2776,
+      "step": 1756
+    },
+    {
+      "epoch": 2.4854918612880397,
+      "grad_norm": 0.1906652371264402,
+      "learning_rate": 9.496327387198321e-06,
+      "loss": 0.2834,
+      "step": 1757
+    },
+    {
+      "epoch": 2.4869072894550603,
+      "grad_norm": 0.18973806755256659,
+      "learning_rate": 9.4700944386149e-06,
+      "loss": 0.2734,
+      "step": 1758
+    },
+    {
+      "epoch": 2.4883227176220806,
+      "grad_norm": 0.18680172089985692,
+      "learning_rate": 9.44386149003148e-06,
+      "loss": 0.278,
+      "step": 1759
+    },
+    {
+      "epoch": 2.4897381457891012,
+      "grad_norm": 0.18989404461557566,
+      "learning_rate": 9.417628541448059e-06,
+      "loss": 0.282,
+      "step": 1760
+    },
+    {
+      "epoch": 2.491153573956122,
+      "grad_norm": 0.1947137202882916,
+      "learning_rate": 9.391395592864637e-06,
+      "loss": 0.2874,
+      "step": 1761
+    },
+    {
+      "epoch": 2.492569002123142,
+      "grad_norm": 0.18019379349151343,
+      "learning_rate": 9.365162644281218e-06,
+      "loss": 0.2825,
+      "step": 1762
+    },
+    {
+      "epoch": 2.493984430290163,
+      "grad_norm": 0.18529175062442796,
+      "learning_rate": 9.338929695697796e-06,
+      "loss": 0.2697,
+      "step": 1763
+    },
+    {
+      "epoch": 2.4953998584571835,
+      "grad_norm": 0.19215544617973443,
+      "learning_rate": 9.312696747114377e-06,
+      "loss": 0.2839,
+      "step": 1764
+    },
+    {
+      "epoch": 2.4968152866242037,
+      "grad_norm": 0.18458485693360022,
+      "learning_rate": 9.286463798530955e-06,
+      "loss": 0.2752,
+      "step": 1765
+    },
+    {
+      "epoch": 2.4982307147912244,
+      "grad_norm": 0.1947758838586988,
+      "learning_rate": 9.260230849947535e-06,
+      "loss": 0.2835,
+      "step": 1766
+    },
+    {
+      "epoch": 2.499646142958245,
+      "grad_norm": 0.1989466258135159,
+      "learning_rate": 9.233997901364114e-06,
+      "loss": 0.31,
+      "step": 1767
+    },
+    {
+      "epoch": 2.5010615711252653,
+      "grad_norm": 0.19676251384745466,
+      "learning_rate": 9.207764952780693e-06,
+      "loss": 0.2937,
+      "step": 1768
+    },
+    {
+      "epoch": 2.502476999292286,
+      "grad_norm": 0.2013785603803912,
+      "learning_rate": 9.181532004197273e-06,
+      "loss": 0.2927,
+      "step": 1769
+    },
+    {
+      "epoch": 2.5038924274593066,
+      "grad_norm": 0.1907466490422871,
+      "learning_rate": 9.155299055613851e-06,
+      "loss": 0.2915,
+      "step": 1770
+    },
+    {
+      "epoch": 2.505307855626327,
+      "grad_norm": 0.18137168182898722,
+      "learning_rate": 9.12906610703043e-06,
+      "loss": 0.2846,
+      "step": 1771
+    },
+    {
+      "epoch": 2.5067232837933475,
+      "grad_norm": 0.19295147501733936,
+      "learning_rate": 9.10283315844701e-06,
+      "loss": 0.2879,
+      "step": 1772
+    },
+    {
+      "epoch": 2.5081387119603678,
+      "grad_norm": 0.1956545760836981,
+      "learning_rate": 9.076600209863589e-06,
+      "loss": 0.2784,
+      "step": 1773
+    },
+    {
+      "epoch": 2.5095541401273884,
+      "grad_norm": 0.1813866948626585,
+      "learning_rate": 9.050367261280168e-06,
+      "loss": 0.275,
+      "step": 1774
+    },
+    {
+      "epoch": 2.510969568294409,
+      "grad_norm": 0.1889014444987169,
+      "learning_rate": 9.024134312696748e-06,
+      "loss": 0.2794,
+      "step": 1775
+    },
+    {
+      "epoch": 2.5123849964614298,
+      "grad_norm": 0.17823356530652573,
+      "learning_rate": 8.997901364113328e-06,
+      "loss": 0.2755,
+      "step": 1776
+    },
+    {
+      "epoch": 2.51380042462845,
+      "grad_norm": 0.20194077440692024,
+      "learning_rate": 8.971668415529907e-06,
+      "loss": 0.2939,
+      "step": 1777
+    },
+    {
+      "epoch": 2.5152158527954707,
+      "grad_norm": 0.19442054303306802,
+      "learning_rate": 8.945435466946485e-06,
+      "loss": 0.2916,
+      "step": 1778
+    },
+    {
+      "epoch": 2.516631280962491,
+      "grad_norm": 0.18917905930574533,
+      "learning_rate": 8.919202518363066e-06,
+      "loss": 0.2954,
+      "step": 1779
+    },
+    {
+      "epoch": 2.5180467091295116,
+      "grad_norm": 0.17858071110586585,
+      "learning_rate": 8.892969569779644e-06,
+      "loss": 0.2915,
+      "step": 1780
+    },
+    {
+      "epoch": 2.5194621372965322,
+      "grad_norm": 0.1737991334501271,
+      "learning_rate": 8.866736621196223e-06,
+      "loss": 0.2884,
+      "step": 1781
+    },
+    {
+      "epoch": 2.520877565463553,
+      "grad_norm": 0.1787321629946045,
+      "learning_rate": 8.840503672612801e-06,
+      "loss": 0.2781,
+      "step": 1782
+    },
+    {
+      "epoch": 2.522292993630573,
+      "grad_norm": 0.18487310998562898,
+      "learning_rate": 8.814270724029382e-06,
+      "loss": 0.2876,
+      "step": 1783
+    },
+    {
+      "epoch": 2.523708421797594,
+      "grad_norm": 0.19011181620372405,
+      "learning_rate": 8.78803777544596e-06,
+      "loss": 0.2827,
+      "step": 1784
+    },
+    {
+      "epoch": 2.525123849964614,
+      "grad_norm": 0.18458161214077176,
+      "learning_rate": 8.761804826862539e-06,
+      "loss": 0.2888,
+      "step": 1785
+    },
+    {
+      "epoch": 2.5265392781316347,
+      "grad_norm": 0.19197191697122068,
+      "learning_rate": 8.735571878279119e-06,
+      "loss": 0.2889,
+      "step": 1786
+    },
+    {
+      "epoch": 2.5279547062986554,
+      "grad_norm": 0.18520101528729924,
+      "learning_rate": 8.709338929695698e-06,
+      "loss": 0.2923,
+      "step": 1787
+    },
+    {
+      "epoch": 2.529370134465676,
+      "grad_norm": 0.19625949244610194,
+      "learning_rate": 8.683105981112278e-06,
+      "loss": 0.2921,
+      "step": 1788
+    },
+    {
+      "epoch": 2.5307855626326963,
+      "grad_norm": 0.18483412184029158,
+      "learning_rate": 8.656873032528857e-06,
+      "loss": 0.2821,
+      "step": 1789
+    },
+    {
+      "epoch": 2.532200990799717,
+      "grad_norm": 0.1913109404131173,
+      "learning_rate": 8.630640083945437e-06,
+      "loss": 0.2971,
+      "step": 1790
+    },
+    {
+      "epoch": 2.533616418966737,
+      "grad_norm": 0.18453330441031862,
+      "learning_rate": 8.604407135362015e-06,
+      "loss": 0.2936,
+      "step": 1791
+    },
+    {
+      "epoch": 2.535031847133758,
+      "grad_norm": 0.18572328873024224,
+      "learning_rate": 8.578174186778594e-06,
+      "loss": 0.2849,
+      "step": 1792
+    },
+    {
+      "epoch": 2.5364472753007785,
+      "grad_norm": 0.17746223874171665,
+      "learning_rate": 8.551941238195174e-06,
+      "loss": 0.2737,
+      "step": 1793
+    },
+    {
+      "epoch": 2.537862703467799,
+      "grad_norm": 0.1948047969977976,
+      "learning_rate": 8.525708289611753e-06,
+      "loss": 0.2929,
+      "step": 1794
+    },
+    {
+      "epoch": 2.5392781316348194,
+      "grad_norm": 0.1812728957254174,
+      "learning_rate": 8.499475341028331e-06,
+      "loss": 0.2831,
+      "step": 1795
+    },
+    {
+      "epoch": 2.54069355980184,
+      "grad_norm": 0.17664206051208176,
+      "learning_rate": 8.47324239244491e-06,
+      "loss": 0.2932,
+      "step": 1796
+    },
+    {
+      "epoch": 2.5421089879688603,
+      "grad_norm": 0.19364101658694155,
+      "learning_rate": 8.44700944386149e-06,
+      "loss": 0.2814,
+      "step": 1797
+    },
+    {
+      "epoch": 2.543524416135881,
+      "grad_norm": 0.19920710210422615,
+      "learning_rate": 8.420776495278069e-06,
+      "loss": 0.2902,
+      "step": 1798
+    },
+    {
+      "epoch": 2.5449398443029017,
+      "grad_norm": 0.19178070631225227,
+      "learning_rate": 8.39454354669465e-06,
+      "loss": 0.2889,
+      "step": 1799
+    },
+    {
+      "epoch": 2.5463552724699223,
+      "grad_norm": 0.19420951217064594,
+      "learning_rate": 8.368310598111228e-06,
+      "loss": 0.2986,
+      "step": 1800
+    },
+    {
+      "epoch": 2.5477707006369426,
+      "grad_norm": 0.20395759310322037,
+      "learning_rate": 8.342077649527808e-06,
+      "loss": 0.3069,
+      "step": 1801
+    },
+    {
+      "epoch": 2.5491861288039632,
+      "grad_norm": 0.19759484712292424,
+      "learning_rate": 8.315844700944387e-06,
+      "loss": 0.2883,
+      "step": 1802
+    },
+    {
+      "epoch": 2.5506015569709835,
+      "grad_norm": 0.1790110297887989,
+      "learning_rate": 8.289611752360965e-06,
+      "loss": 0.2837,
+      "step": 1803
+    },
+    {
+      "epoch": 2.552016985138004,
+      "grad_norm": 0.19674318163208754,
+      "learning_rate": 8.263378803777546e-06,
+      "loss": 0.2864,
+      "step": 1804
+    },
+    {
+      "epoch": 2.553432413305025,
+      "grad_norm": 0.18444891681361106,
+      "learning_rate": 8.237145855194124e-06,
+      "loss": 0.2925,
+      "step": 1805
+    },
+    {
+      "epoch": 2.5548478414720455,
+      "grad_norm": 0.17804797600094827,
+      "learning_rate": 8.210912906610703e-06,
+      "loss": 0.294,
+      "step": 1806
+    },
+    {
+      "epoch": 2.5562632696390657,
+      "grad_norm": 0.18656197397743954,
+      "learning_rate": 8.184679958027283e-06,
+      "loss": 0.2892,
+      "step": 1807
+    },
+    {
+      "epoch": 2.5576786978060864,
+      "grad_norm": 0.18207746960029497,
+      "learning_rate": 8.158447009443862e-06,
+      "loss": 0.2881,
+      "step": 1808
+    },
+    {
+      "epoch": 2.5590941259731066,
+      "grad_norm": 0.18558652365935788,
+      "learning_rate": 8.13221406086044e-06,
+      "loss": 0.2925,
+      "step": 1809
+    },
+    {
+      "epoch": 2.5605095541401273,
+      "grad_norm": 0.1806740707048797,
+      "learning_rate": 8.10598111227702e-06,
+      "loss": 0.2978,
+      "step": 1810
+    },
+    {
+      "epoch": 2.561924982307148,
+      "grad_norm": 0.18513945362124973,
+      "learning_rate": 8.079748163693599e-06,
+      "loss": 0.2819,
+      "step": 1811
+    },
+    {
+      "epoch": 2.5633404104741686,
+      "grad_norm": 0.19188799707417797,
+      "learning_rate": 8.05351521511018e-06,
+      "loss": 0.3017,
+      "step": 1812
+    },
+    {
+      "epoch": 2.564755838641189,
+      "grad_norm": 0.18467584871999182,
+      "learning_rate": 8.027282266526758e-06,
+      "loss": 0.2801,
+      "step": 1813
+    },
+    {
+      "epoch": 2.5661712668082095,
+      "grad_norm": 0.17336259123772044,
+      "learning_rate": 8.001049317943338e-06,
+      "loss": 0.2818,
+      "step": 1814
+    },
+    {
+      "epoch": 2.5675866949752297,
+      "grad_norm": 0.18518477739723296,
+      "learning_rate": 7.974816369359917e-06,
+      "loss": 0.2744,
+      "step": 1815
+    },
+    {
+      "epoch": 2.5690021231422504,
+      "grad_norm": 0.184374472353793,
+      "learning_rate": 7.948583420776495e-06,
+      "loss": 0.2962,
+      "step": 1816
+    },
+    {
+      "epoch": 2.570417551309271,
+      "grad_norm": 0.18811552381445062,
+      "learning_rate": 7.922350472193076e-06,
+      "loss": 0.286,
+      "step": 1817
+    },
+    {
+      "epoch": 2.5718329794762917,
+      "grad_norm": 0.1850722906635902,
+      "learning_rate": 7.896117523609654e-06,
+      "loss": 0.2795,
+      "step": 1818
+    },
+    {
+      "epoch": 2.573248407643312,
+      "grad_norm": 0.1872232911377214,
+      "learning_rate": 7.869884575026233e-06,
+      "loss": 0.2805,
+      "step": 1819
+    },
+    {
+      "epoch": 2.5746638358103326,
+      "grad_norm": 0.17884119975176793,
+      "learning_rate": 7.843651626442811e-06,
+      "loss": 0.281,
+      "step": 1820
+    },
+    {
+      "epoch": 2.576079263977353,
+      "grad_norm": 0.18130817208207814,
+      "learning_rate": 7.817418677859392e-06,
+      "loss": 0.2771,
+      "step": 1821
+    },
+    {
+      "epoch": 2.5774946921443735,
+      "grad_norm": 0.1798684636187942,
+      "learning_rate": 7.79118572927597e-06,
+      "loss": 0.2848,
+      "step": 1822
+    },
+    {
+      "epoch": 2.578910120311394,
+      "grad_norm": 0.17825232480124747,
+      "learning_rate": 7.76495278069255e-06,
+      "loss": 0.2736,
+      "step": 1823
+    },
+    {
+      "epoch": 2.580325548478415,
+      "grad_norm": 0.18390615555070572,
+      "learning_rate": 7.73871983210913e-06,
+      "loss": 0.2823,
+      "step": 1824
+    },
+    {
+      "epoch": 2.581740976645435,
+      "grad_norm": 0.1953773122629701,
+      "learning_rate": 7.71248688352571e-06,
+      "loss": 0.2958,
+      "step": 1825
+    },
+    {
+      "epoch": 2.583156404812456,
+      "grad_norm": 0.1976335826419702,
+      "learning_rate": 7.686253934942288e-06,
+      "loss": 0.2879,
+      "step": 1826
+    },
+    {
+      "epoch": 2.5845718329794765,
+      "grad_norm": 0.19169910287214764,
+      "learning_rate": 7.660020986358867e-06,
+      "loss": 0.2873,
+      "step": 1827
+    },
+    {
+      "epoch": 2.5859872611464967,
+      "grad_norm": 0.1786869386826253,
+      "learning_rate": 7.633788037775447e-06,
+      "loss": 0.2725,
+      "step": 1828
+    },
+    {
+      "epoch": 2.5874026893135174,
+      "grad_norm": 0.18415450922430598,
+      "learning_rate": 7.607555089192026e-06,
+      "loss": 0.2865,
+      "step": 1829
+    },
+    {
+      "epoch": 2.588818117480538,
+      "grad_norm": 0.18500654761858334,
+      "learning_rate": 7.581322140608604e-06,
+      "loss": 0.3046,
+      "step": 1830
+    },
+    {
+      "epoch": 2.5902335456475583,
+      "grad_norm": 0.18940801943582591,
+      "learning_rate": 7.5550891920251845e-06,
+      "loss": 0.2983,
+      "step": 1831
+    },
+    {
+      "epoch": 2.591648973814579,
+      "grad_norm": 0.19171549447045327,
+      "learning_rate": 7.528856243441763e-06,
+      "loss": 0.2933,
+      "step": 1832
+    },
+    {
+      "epoch": 2.5930644019815996,
+      "grad_norm": 0.17434424948793265,
+      "learning_rate": 7.5026232948583425e-06,
+      "loss": 0.2847,
+      "step": 1833
+    },
+    {
+      "epoch": 2.59447983014862,
+      "grad_norm": 0.1956790129381184,
+      "learning_rate": 7.476390346274921e-06,
+      "loss": 0.2908,
+      "step": 1834
+    },
+    {
+      "epoch": 2.5958952583156405,
+      "grad_norm": 0.191386449253825,
+      "learning_rate": 7.450157397691501e-06,
+      "loss": 0.2866,
+      "step": 1835
+    },
+    {
+      "epoch": 2.597310686482661,
+      "grad_norm": 0.19813004589544073,
+      "learning_rate": 7.42392444910808e-06,
+      "loss": 0.2841,
+      "step": 1836
+    },
+    {
+      "epoch": 2.5987261146496814,
+      "grad_norm": 0.1832862502583108,
+      "learning_rate": 7.3976915005246586e-06,
+      "loss": 0.2962,
+      "step": 1837
+    },
+    {
+      "epoch": 2.600141542816702,
+      "grad_norm": 0.1925834407695736,
+      "learning_rate": 7.371458551941239e-06,
+      "loss": 0.2901,
+      "step": 1838
+    },
+    {
+      "epoch": 2.6015569709837227,
+      "grad_norm": 0.19516481876501107,
+      "learning_rate": 7.345225603357818e-06,
+      "loss": 0.2819,
+      "step": 1839
+    },
+    {
+      "epoch": 2.602972399150743,
+      "grad_norm": 0.1766600106048532,
+      "learning_rate": 7.318992654774397e-06,
+      "loss": 0.2691,
+      "step": 1840
+    },
+    {
+      "epoch": 2.6043878273177636,
+      "grad_norm": 0.19019561454678408,
+      "learning_rate": 7.2927597061909755e-06,
+      "loss": 0.2911,
+      "step": 1841
+    },
+    {
+      "epoch": 2.6058032554847843,
+      "grad_norm": 0.18281180121904445,
+      "learning_rate": 7.266526757607556e-06,
+      "loss": 0.2992,
+      "step": 1842
+    },
+    {
+      "epoch": 2.6072186836518045,
+      "grad_norm": 0.18956913138535492,
+      "learning_rate": 7.240293809024134e-06,
+      "loss": 0.2899,
+      "step": 1843
+    },
+    {
+      "epoch": 2.608634111818825,
+      "grad_norm": 0.18216296469351154,
+      "learning_rate": 7.214060860440714e-06,
+      "loss": 0.2846,
+      "step": 1844
+    },
+    {
+      "epoch": 2.610049539985846,
+      "grad_norm": 0.1864964375261479,
+      "learning_rate": 7.187827911857293e-06,
+      "loss": 0.2862,
+      "step": 1845
+    },
+    {
+      "epoch": 2.611464968152866,
+      "grad_norm": 0.17721283274245875,
+      "learning_rate": 7.161594963273873e-06,
+      "loss": 0.2924,
+      "step": 1846
+    },
+    {
+      "epoch": 2.6128803963198868,
+      "grad_norm": 0.1838086426191061,
+      "learning_rate": 7.135362014690451e-06,
+      "loss": 0.2913,
+      "step": 1847
+    },
+    {
+      "epoch": 2.6142958244869074,
+      "grad_norm": 0.1913187649654714,
+      "learning_rate": 7.1091290661070315e-06,
+      "loss": 0.2908,
+      "step": 1848
+    },
+    {
+      "epoch": 2.6157112526539277,
+      "grad_norm": 0.18822206075681172,
+      "learning_rate": 7.08289611752361e-06,
+      "loss": 0.2916,
+      "step": 1849
+    },
+    {
+      "epoch": 2.6171266808209483,
+      "grad_norm": 0.1827073739968568,
+      "learning_rate": 7.056663168940189e-06,
+      "loss": 0.3065,
+      "step": 1850
+    },
+    {
+      "epoch": 2.618542108987969,
+      "grad_norm": 0.19522047436385434,
+      "learning_rate": 7.030430220356768e-06,
+      "loss": 0.2955,
+      "step": 1851
+    },
+    {
+      "epoch": 2.6199575371549892,
+      "grad_norm": 0.18022638550628511,
+      "learning_rate": 7.0041972717733484e-06,
+      "loss": 0.2875,
+      "step": 1852
+    },
+    {
+      "epoch": 2.62137296532201,
+      "grad_norm": 0.21508080661024662,
+      "learning_rate": 6.977964323189927e-06,
+      "loss": 0.2855,
+      "step": 1853
+    },
+    {
+      "epoch": 2.6227883934890306,
+      "grad_norm": 0.1747002634441143,
+      "learning_rate": 6.951731374606506e-06,
+      "loss": 0.2793,
+      "step": 1854
+    },
+    {
+      "epoch": 2.624203821656051,
+      "grad_norm": 0.18125593556978656,
+      "learning_rate": 6.925498426023086e-06,
+      "loss": 0.2849,
+      "step": 1855
+    },
+    {
+      "epoch": 2.6256192498230715,
+      "grad_norm": 0.18951545885331525,
+      "learning_rate": 6.8992654774396645e-06,
+      "loss": 0.2937,
+      "step": 1856
+    },
+    {
+      "epoch": 2.627034677990092,
+      "grad_norm": 0.1855323965413897,
+      "learning_rate": 6.873032528856244e-06,
+      "loss": 0.2935,
+      "step": 1857
+    },
+    {
+      "epoch": 2.6284501061571124,
+      "grad_norm": 0.17210946333714922,
+      "learning_rate": 6.8467995802728225e-06,
+      "loss": 0.2741,
+      "step": 1858
+    },
+    {
+      "epoch": 2.629865534324133,
+      "grad_norm": 0.18252709891494256,
+      "learning_rate": 6.820566631689403e-06,
+      "loss": 0.2877,
+      "step": 1859
+    },
+    {
+      "epoch": 2.6312809624911537,
+      "grad_norm": 0.18013341848747635,
+      "learning_rate": 6.794333683105981e-06,
+      "loss": 0.2742,
+      "step": 1860
+    },
+    {
+      "epoch": 2.632696390658174,
+      "grad_norm": 0.18568461347186016,
+      "learning_rate": 6.76810073452256e-06,
+      "loss": 0.3012,
+      "step": 1861
+    },
+    {
+      "epoch": 2.6341118188251946,
+      "grad_norm": 0.19734367213602771,
+      "learning_rate": 6.74186778593914e-06,
+      "loss": 0.3028,
+      "step": 1862
+    },
+    {
+      "epoch": 2.6355272469922153,
+      "grad_norm": 0.18309094784045296,
+      "learning_rate": 6.71563483735572e-06,
+      "loss": 0.2839,
+      "step": 1863
+    },
+    {
+      "epoch": 2.6369426751592355,
+      "grad_norm": 0.17072703179456888,
+      "learning_rate": 6.689401888772298e-06,
+      "loss": 0.2784,
+      "step": 1864
+    },
+    {
+      "epoch": 2.638358103326256,
+      "grad_norm": 0.18582678531448524,
+      "learning_rate": 6.663168940188877e-06,
+      "loss": 0.2815,
+      "step": 1865
+    },
+    {
+      "epoch": 2.639773531493277,
+      "grad_norm": 0.1924396916188198,
+      "learning_rate": 6.636935991605457e-06,
+      "loss": 0.2853,
+      "step": 1866
+    },
+    {
+      "epoch": 2.641188959660297,
+      "grad_norm": 0.18207577185241333,
+      "learning_rate": 6.610703043022036e-06,
+      "loss": 0.2894,
+      "step": 1867
+    },
+    {
+      "epoch": 2.6426043878273178,
+      "grad_norm": 0.19551476595141212,
+      "learning_rate": 6.584470094438614e-06,
+      "loss": 0.2873,
+      "step": 1868
+    },
+    {
+      "epoch": 2.6440198159943384,
+      "grad_norm": 0.16976395480923093,
+      "learning_rate": 6.558237145855195e-06,
+      "loss": 0.2845,
+      "step": 1869
+    },
+    {
+      "epoch": 2.6454352441613587,
+      "grad_norm": 0.1794667115720783,
+      "learning_rate": 6.532004197271774e-06,
+      "loss": 0.2861,
+      "step": 1870
+    },
+    {
+      "epoch": 2.6468506723283793,
+      "grad_norm": 0.17986008928354258,
+      "learning_rate": 6.505771248688353e-06,
+      "loss": 0.2781,
+      "step": 1871
+    },
+    {
+      "epoch": 2.6482661004954,
+      "grad_norm": 0.19873992835420157,
+      "learning_rate": 6.479538300104931e-06,
+      "loss": 0.2813,
+      "step": 1872
+    },
+    {
+      "epoch": 2.6496815286624202,
+      "grad_norm": 0.21479283680707595,
+      "learning_rate": 6.4533053515215116e-06,
+      "loss": 0.2912,
+      "step": 1873
+    },
+    {
+      "epoch": 2.651096956829441,
+      "grad_norm": 0.18424082693483232,
+      "learning_rate": 6.42707240293809e-06,
+      "loss": 0.2777,
+      "step": 1874
+    },
+    {
+      "epoch": 2.6525123849964616,
+      "grad_norm": 0.18232234087078958,
+      "learning_rate": 6.40083945435467e-06,
+      "loss": 0.2985,
+      "step": 1875
+    },
+    {
+      "epoch": 2.653927813163482,
+      "grad_norm": 0.1848666204181274,
+      "learning_rate": 6.37460650577125e-06,
+      "loss": 0.2823,
+      "step": 1876
+    },
+    {
+      "epoch": 2.6553432413305025,
+      "grad_norm": 0.18856393761385729,
+      "learning_rate": 6.3483735571878285e-06,
+      "loss": 0.2932,
+      "step": 1877
+    },
+    {
+      "epoch": 2.656758669497523,
+      "grad_norm": 0.17971858451183984,
+      "learning_rate": 6.322140608604407e-06,
+      "loss": 0.293,
+      "step": 1878
+    },
+    {
+      "epoch": 2.6581740976645434,
+      "grad_norm": 0.17066349292261504,
+      "learning_rate": 6.295907660020986e-06,
+      "loss": 0.2845,
+      "step": 1879
+    },
+    {
+      "epoch": 2.659589525831564,
+      "grad_norm": 0.23408224675290515,
+      "learning_rate": 6.269674711437566e-06,
+      "loss": 0.2834,
+      "step": 1880
+    },
+    {
+      "epoch": 2.6610049539985847,
+      "grad_norm": 0.18778554084997837,
+      "learning_rate": 6.243441762854145e-06,
+      "loss": 0.3077,
+      "step": 1881
+    },
+    {
+      "epoch": 2.662420382165605,
+      "grad_norm": 0.18402603177297536,
+      "learning_rate": 6.217208814270725e-06,
+      "loss": 0.2923,
+      "step": 1882
+    },
+    {
+      "epoch": 2.6638358103326256,
+      "grad_norm": 0.1790808754907691,
+      "learning_rate": 6.190975865687303e-06,
+      "loss": 0.2877,
+      "step": 1883
+    },
+    {
+      "epoch": 2.6652512384996463,
+      "grad_norm": 0.18117254341267883,
+      "learning_rate": 6.164742917103883e-06,
+      "loss": 0.298,
+      "step": 1884
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.18133341165545064,
+      "learning_rate": 6.1385099685204614e-06,
+      "loss": 0.285,
+      "step": 1885
+    },
+    {
+      "epoch": 2.668082094833687,
+      "grad_norm": 0.18694177265565673,
+      "learning_rate": 6.112277019937041e-06,
+      "loss": 0.2888,
+      "step": 1886
+    },
+    {
+      "epoch": 2.669497523000708,
+      "grad_norm": 0.17536912203009425,
+      "learning_rate": 6.08604407135362e-06,
+      "loss": 0.2879,
+      "step": 1887
+    },
+    {
+      "epoch": 2.670912951167728,
+      "grad_norm": 0.1776523188094862,
+      "learning_rate": 6.0598111227702e-06,
+      "loss": 0.2815,
+      "step": 1888
+    },
+    {
+      "epoch": 2.6723283793347488,
+      "grad_norm": 0.17913360450470825,
+      "learning_rate": 6.033578174186779e-06,
+      "loss": 0.2915,
+      "step": 1889
+    },
+    {
+      "epoch": 2.6737438075017694,
+      "grad_norm": 0.19344378917017407,
+      "learning_rate": 6.007345225603358e-06,
+      "loss": 0.2891,
+      "step": 1890
+    },
+    {
+      "epoch": 2.6751592356687897,
+      "grad_norm": 0.18365487875162972,
+      "learning_rate": 5.981112277019937e-06,
+      "loss": 0.285,
+      "step": 1891
+    },
+    {
+      "epoch": 2.6765746638358103,
+      "grad_norm": 0.18008101002043897,
+      "learning_rate": 5.954879328436517e-06,
+      "loss": 0.2919,
+      "step": 1892
+    },
+    {
+      "epoch": 2.677990092002831,
+      "grad_norm": 0.18236055900255962,
+      "learning_rate": 5.928646379853096e-06,
+      "loss": 0.2879,
+      "step": 1893
+    },
+    {
+      "epoch": 2.6794055201698512,
+      "grad_norm": 0.1782327309337629,
+      "learning_rate": 5.9024134312696755e-06,
+      "loss": 0.287,
+      "step": 1894
+    },
+    {
+      "epoch": 2.680820948336872,
+      "grad_norm": 0.18375618710095973,
+      "learning_rate": 5.876180482686254e-06,
+      "loss": 0.2883,
+      "step": 1895
+    },
+    {
+      "epoch": 2.6822363765038926,
+      "grad_norm": 0.1802953324961345,
+      "learning_rate": 5.8499475341028336e-06,
+      "loss": 0.2815,
+      "step": 1896
+    },
+    {
+      "epoch": 2.683651804670913,
+      "grad_norm": 0.18838302801802298,
+      "learning_rate": 5.823714585519412e-06,
+      "loss": 0.3063,
+      "step": 1897
+    },
+    {
+      "epoch": 2.6850672328379335,
+      "grad_norm": 0.1850222907843767,
+      "learning_rate": 5.797481636935992e-06,
+      "loss": 0.2917,
+      "step": 1898
+    },
+    {
+      "epoch": 2.686482661004954,
+      "grad_norm": 0.17707903423488489,
+      "learning_rate": 5.771248688352571e-06,
+      "loss": 0.2993,
+      "step": 1899
+    },
+    {
+      "epoch": 2.6878980891719744,
+      "grad_norm": 0.17923211297671582,
+      "learning_rate": 5.7450157397691505e-06,
+      "loss": 0.2841,
+      "step": 1900
+    },
+    {
+      "epoch": 2.689313517338995,
+      "grad_norm": 0.18554093813240305,
+      "learning_rate": 5.71878279118573e-06,
+      "loss": 0.277,
+      "step": 1901
+    },
+    {
+      "epoch": 2.6907289455060157,
+      "grad_norm": 0.18768414892060284,
+      "learning_rate": 5.6925498426023085e-06,
+      "loss": 0.2865,
+      "step": 1902
+    },
+    {
+      "epoch": 2.692144373673036,
+      "grad_norm": 0.1857402006462927,
+      "learning_rate": 5.666316894018888e-06,
+      "loss": 0.2973,
+      "step": 1903
+    },
+    {
+      "epoch": 2.6935598018400566,
+      "grad_norm": 0.17132007260096363,
+      "learning_rate": 5.6400839454354665e-06,
+      "loss": 0.2776,
+      "step": 1904
+    },
+    {
+      "epoch": 2.6949752300070773,
+      "grad_norm": 0.17720084601471012,
+      "learning_rate": 5.613850996852047e-06,
+      "loss": 0.2701,
+      "step": 1905
+    },
+    {
+      "epoch": 2.6963906581740975,
+      "grad_norm": 0.20231472960856142,
+      "learning_rate": 5.587618048268626e-06,
+      "loss": 0.2879,
+      "step": 1906
+    },
+    {
+      "epoch": 2.697806086341118,
+      "grad_norm": 0.1948076173661505,
+      "learning_rate": 5.561385099685205e-06,
+      "loss": 0.3071,
+      "step": 1907
+    },
+    {
+      "epoch": 2.699221514508139,
+      "grad_norm": 0.19277975165339786,
+      "learning_rate": 5.535152151101784e-06,
+      "loss": 0.288,
+      "step": 1908
+    },
+    {
+      "epoch": 2.700636942675159,
+      "grad_norm": 0.18941875379657241,
+      "learning_rate": 5.508919202518363e-06,
+      "loss": 0.3003,
+      "step": 1909
+    },
+    {
+      "epoch": 2.7020523708421798,
+      "grad_norm": 0.17808309503947614,
+      "learning_rate": 5.482686253934942e-06,
+      "loss": 0.285,
+      "step": 1910
+    },
+    {
+      "epoch": 2.7034677990092004,
+      "grad_norm": 0.17738627483023298,
+      "learning_rate": 5.456453305351522e-06,
+      "loss": 0.2726,
+      "step": 1911
+    },
+    {
+      "epoch": 2.7048832271762207,
+      "grad_norm": 0.1870343549277759,
+      "learning_rate": 5.430220356768101e-06,
+      "loss": 0.2847,
+      "step": 1912
+    },
+    {
+      "epoch": 2.7062986553432413,
+      "grad_norm": 0.18336127786092873,
+      "learning_rate": 5.403987408184681e-06,
+      "loss": 0.2894,
+      "step": 1913
+    },
+    {
+      "epoch": 2.707714083510262,
+      "grad_norm": 0.18002245752166193,
+      "learning_rate": 5.377754459601259e-06,
+      "loss": 0.2987,
+      "step": 1914
+    },
+    {
+      "epoch": 2.709129511677282,
+      "grad_norm": 0.17536975786409212,
+      "learning_rate": 5.351521511017839e-06,
+      "loss": 0.2812,
+      "step": 1915
+    },
+    {
+      "epoch": 2.710544939844303,
+      "grad_norm": 0.17030964776321728,
+      "learning_rate": 5.325288562434417e-06,
+      "loss": 0.2845,
+      "step": 1916
+    },
+    {
+      "epoch": 2.7119603680113236,
+      "grad_norm": 0.17298737950738244,
+      "learning_rate": 5.2990556138509975e-06,
+      "loss": 0.276,
+      "step": 1917
+    },
+    {
+      "epoch": 2.713375796178344,
+      "grad_norm": 0.18792212139908102,
+      "learning_rate": 5.272822665267577e-06,
+      "loss": 0.2901,
+      "step": 1918
+    },
+    {
+      "epoch": 2.7147912243453645,
+      "grad_norm": 0.17302303670598543,
+      "learning_rate": 5.2465897166841556e-06,
+      "loss": 0.2866,
+      "step": 1919
+    },
+    {
+      "epoch": 2.716206652512385,
+      "grad_norm": 0.29187006960632644,
+      "learning_rate": 5.220356768100735e-06,
+      "loss": 0.3027,
+      "step": 1920
+    },
+    {
+      "epoch": 2.7176220806794054,
+      "grad_norm": 0.18511802906000652,
+      "learning_rate": 5.194123819517314e-06,
+      "loss": 0.2853,
+      "step": 1921
+    },
+    {
+      "epoch": 2.719037508846426,
+      "grad_norm": 0.1800424289490244,
+      "learning_rate": 5.167890870933893e-06,
+      "loss": 0.2707,
+      "step": 1922
+    },
+    {
+      "epoch": 2.7204529370134467,
+      "grad_norm": 0.18131224935839205,
+      "learning_rate": 5.1416579223504725e-06,
+      "loss": 0.2754,
+      "step": 1923
+    },
+    {
+      "epoch": 2.721868365180467,
+      "grad_norm": 0.1923264676927521,
+      "learning_rate": 5.115424973767052e-06,
+      "loss": 0.2795,
+      "step": 1924
+    },
+    {
+      "epoch": 2.7232837933474876,
+      "grad_norm": 0.1741853780638869,
+      "learning_rate": 5.089192025183631e-06,
+      "loss": 0.2762,
+      "step": 1925
+    },
+    {
+      "epoch": 2.7246992215145083,
+      "grad_norm": 0.1872875625453876,
+      "learning_rate": 5.06295907660021e-06,
+      "loss": 0.2895,
+      "step": 1926
+    },
+    {
+      "epoch": 2.7261146496815285,
+      "grad_norm": 0.17331130432224612,
+      "learning_rate": 5.036726128016789e-06,
+      "loss": 0.2806,
+      "step": 1927
+    },
+    {
+      "epoch": 2.727530077848549,
+      "grad_norm": 0.18165497532503466,
+      "learning_rate": 5.010493179433368e-06,
+      "loss": 0.2821,
+      "step": 1928
+    },
+    {
+      "epoch": 2.72894550601557,
+      "grad_norm": 0.18825902755069426,
+      "learning_rate": 4.984260230849947e-06,
+      "loss": 0.2912,
+      "step": 1929
+    },
+    {
+      "epoch": 2.73036093418259,
+      "grad_norm": 0.1856572192632043,
+      "learning_rate": 4.958027282266528e-06,
+      "loss": 0.2808,
+      "step": 1930
+    },
+    {
+      "epoch": 2.7317763623496107,
+      "grad_norm": 0.18010845986339513,
+      "learning_rate": 4.931794333683106e-06,
+      "loss": 0.2922,
+      "step": 1931
+    },
+    {
+      "epoch": 2.7331917905166314,
+      "grad_norm": 0.1766883022479923,
+      "learning_rate": 4.905561385099686e-06,
+      "loss": 0.2875,
+      "step": 1932
+    },
+    {
+      "epoch": 2.7346072186836516,
+      "grad_norm": 0.18382795448047617,
+      "learning_rate": 4.879328436516264e-06,
+      "loss": 0.302,
+      "step": 1933
+    },
+    {
+      "epoch": 2.7360226468506723,
+      "grad_norm": 0.1825510343747918,
+      "learning_rate": 4.853095487932844e-06,
+      "loss": 0.2892,
+      "step": 1934
+    },
+    {
+      "epoch": 2.737438075017693,
+      "grad_norm": 0.17940605284014863,
+      "learning_rate": 4.826862539349423e-06,
+      "loss": 0.2918,
+      "step": 1935
+    },
+    {
+      "epoch": 2.738853503184713,
+      "grad_norm": 0.18781804698819057,
+      "learning_rate": 4.800629590766003e-06,
+      "loss": 0.2856,
+      "step": 1936
+    },
+    {
+      "epoch": 2.740268931351734,
+      "grad_norm": 0.17054774170847606,
+      "learning_rate": 4.774396642182582e-06,
+      "loss": 0.2856,
+      "step": 1937
+    },
+    {
+      "epoch": 2.7416843595187546,
+      "grad_norm": 0.1730177383569187,
+      "learning_rate": 4.748163693599161e-06,
+      "loss": 0.2862,
+      "step": 1938
+    },
+    {
+      "epoch": 2.743099787685775,
+      "grad_norm": 0.17655836178940787,
+      "learning_rate": 4.72193074501574e-06,
+      "loss": 0.2804,
+      "step": 1939
+    },
+    {
+      "epoch": 2.7445152158527955,
+      "grad_norm": 0.18955604901001924,
+      "learning_rate": 4.695697796432319e-06,
+      "loss": 0.2867,
+      "step": 1940
+    },
+    {
+      "epoch": 2.745930644019816,
+      "grad_norm": 0.17533273828919024,
+      "learning_rate": 4.669464847848898e-06,
+      "loss": 0.2855,
+      "step": 1941
+    },
+    {
+      "epoch": 2.7473460721868364,
+      "grad_norm": 0.18292716308367937,
+      "learning_rate": 4.6432318992654776e-06,
+      "loss": 0.2938,
+      "step": 1942
+    },
+    {
+      "epoch": 2.748761500353857,
+      "grad_norm": 0.1857497386208829,
+      "learning_rate": 4.616998950682057e-06,
+      "loss": 0.2995,
+      "step": 1943
+    },
+    {
+      "epoch": 2.7501769285208777,
+      "grad_norm": 0.18732398873952333,
+      "learning_rate": 4.5907660020986364e-06,
+      "loss": 0.2892,
+      "step": 1944
+    },
+    {
+      "epoch": 2.7515923566878984,
+      "grad_norm": 0.18475276685355677,
+      "learning_rate": 4.564533053515215e-06,
+      "loss": 0.2973,
+      "step": 1945
+    },
+    {
+      "epoch": 2.7530077848549186,
+      "grad_norm": 0.1838772729638837,
+      "learning_rate": 4.5383001049317945e-06,
+      "loss": 0.2938,
+      "step": 1946
+    },
+    {
+      "epoch": 2.7544232130219393,
+      "grad_norm": 0.17828112084647768,
+      "learning_rate": 4.512067156348374e-06,
+      "loss": 0.2808,
+      "step": 1947
+    },
+    {
+      "epoch": 2.7558386411889595,
+      "grad_norm": 0.17472294949009795,
+      "learning_rate": 4.485834207764953e-06,
+      "loss": 0.2855,
+      "step": 1948
+    },
+    {
+      "epoch": 2.75725406935598,
+      "grad_norm": 0.19968711710429163,
+      "learning_rate": 4.459601259181533e-06,
+      "loss": 0.2859,
+      "step": 1949
+    },
+    {
+      "epoch": 2.758669497523001,
+      "grad_norm": 0.18481764504677684,
+      "learning_rate": 4.433368310598111e-06,
+      "loss": 0.2875,
+      "step": 1950
+    },
+    {
+      "epoch": 2.7600849256900215,
+      "grad_norm": 0.18806160443408587,
+      "learning_rate": 4.407135362014691e-06,
+      "loss": 0.2774,
+      "step": 1951
+    },
+    {
+      "epoch": 2.7615003538570417,
+      "grad_norm": 0.1868899501278574,
+      "learning_rate": 4.380902413431269e-06,
+      "loss": 0.2962,
+      "step": 1952
+    },
+    {
+      "epoch": 2.7629157820240624,
+      "grad_norm": 0.19836249564818045,
+      "learning_rate": 4.354669464847849e-06,
+      "loss": 0.2958,
+      "step": 1953
+    },
+    {
+      "epoch": 2.7643312101910826,
+      "grad_norm": 0.18067134134728374,
+      "learning_rate": 4.328436516264428e-06,
+      "loss": 0.2851,
+      "step": 1954
+    },
+    {
+      "epoch": 2.7657466383581033,
+      "grad_norm": 0.18152054180820446,
+      "learning_rate": 4.302203567681008e-06,
+      "loss": 0.2817,
+      "step": 1955
+    },
+    {
+      "epoch": 2.767162066525124,
+      "grad_norm": 0.17651808878217853,
+      "learning_rate": 4.275970619097587e-06,
+      "loss": 0.283,
+      "step": 1956
+    },
+    {
+      "epoch": 2.7685774946921446,
+      "grad_norm": 0.16912999089035635,
+      "learning_rate": 4.249737670514166e-06,
+      "loss": 0.2776,
+      "step": 1957
+    },
+    {
+      "epoch": 2.769992922859165,
+      "grad_norm": 0.18341343085799097,
+      "learning_rate": 4.223504721930745e-06,
+      "loss": 0.2934,
+      "step": 1958
+    },
+    {
+      "epoch": 2.7714083510261855,
+      "grad_norm": 0.19078243306578607,
+      "learning_rate": 4.197271773347325e-06,
+      "loss": 0.2841,
+      "step": 1959
+    },
+    {
+      "epoch": 2.7728237791932058,
+      "grad_norm": 0.17903784125053218,
+      "learning_rate": 4.171038824763904e-06,
+      "loss": 0.2847,
+      "step": 1960
+    },
+    {
+      "epoch": 2.7742392073602264,
+      "grad_norm": 0.18382338913876073,
+      "learning_rate": 4.144805876180483e-06,
+      "loss": 0.2857,
+      "step": 1961
+    },
+    {
+      "epoch": 2.775654635527247,
+      "grad_norm": 0.17684932630528774,
+      "learning_rate": 4.118572927597062e-06,
+      "loss": 0.2791,
+      "step": 1962
+    },
+    {
+      "epoch": 2.777070063694268,
+      "grad_norm": 0.1744776834452007,
+      "learning_rate": 4.0923399790136415e-06,
+      "loss": 0.2786,
+      "step": 1963
+    },
+    {
+      "epoch": 2.778485491861288,
+      "grad_norm": 0.17909606156140384,
+      "learning_rate": 4.06610703043022e-06,
+      "loss": 0.284,
+      "step": 1964
+    },
+    {
+      "epoch": 2.7799009200283087,
+      "grad_norm": 0.18050373442345014,
+      "learning_rate": 4.0398740818467995e-06,
+      "loss": 0.2852,
+      "step": 1965
+    },
+    {
+      "epoch": 2.781316348195329,
+      "grad_norm": 0.17504325599037854,
+      "learning_rate": 4.013641133263379e-06,
+      "loss": 0.2925,
+      "step": 1966
+    },
+    {
+      "epoch": 2.7827317763623496,
+      "grad_norm": 0.1783461491146812,
+      "learning_rate": 3.9874081846799584e-06,
+      "loss": 0.2859,
+      "step": 1967
+    },
+    {
+      "epoch": 2.7841472045293703,
+      "grad_norm": 0.1793103082043295,
+      "learning_rate": 3.961175236096538e-06,
+      "loss": 0.2649,
+      "step": 1968
+    },
+    {
+      "epoch": 2.785562632696391,
+      "grad_norm": 0.1693521387678535,
+      "learning_rate": 3.9349422875131165e-06,
+      "loss": 0.2871,
+      "step": 1969
+    },
+    {
+      "epoch": 2.786978060863411,
+      "grad_norm": 0.17525165546726384,
+      "learning_rate": 3.908709338929696e-06,
+      "loss": 0.2798,
+      "step": 1970
+    },
+    {
+      "epoch": 2.788393489030432,
+      "grad_norm": 0.17652384705179844,
+      "learning_rate": 3.882476390346275e-06,
+      "loss": 0.2842,
+      "step": 1971
+    },
+    {
+      "epoch": 2.789808917197452,
+      "grad_norm": 0.17048800995777114,
+      "learning_rate": 3.856243441762855e-06,
+      "loss": 0.2797,
+      "step": 1972
+    },
+    {
+      "epoch": 2.7912243453644727,
+      "grad_norm": 0.19232225549379658,
+      "learning_rate": 3.830010493179433e-06,
+      "loss": 0.2949,
+      "step": 1973
+    },
+    {
+      "epoch": 2.7926397735314934,
+      "grad_norm": 0.18158427098508625,
+      "learning_rate": 3.803777544596013e-06,
+      "loss": 0.2919,
+      "step": 1974
+    },
+    {
+      "epoch": 2.794055201698514,
+      "grad_norm": 0.17039888361757377,
+      "learning_rate": 3.7775445960125922e-06,
+      "loss": 0.2789,
+      "step": 1975
+    },
+    {
+      "epoch": 2.7954706298655343,
+      "grad_norm": 0.1710460774216996,
+      "learning_rate": 3.7513116474291713e-06,
+      "loss": 0.283,
+      "step": 1976
+    },
+    {
+      "epoch": 2.796886058032555,
+      "grad_norm": 0.18678374857417665,
+      "learning_rate": 3.7250786988457507e-06,
+      "loss": 0.2865,
+      "step": 1977
+    },
+    {
+      "epoch": 2.798301486199575,
+      "grad_norm": 0.17676886545179707,
+      "learning_rate": 3.6988457502623293e-06,
+      "loss": 0.2928,
+      "step": 1978
+    },
+    {
+      "epoch": 2.799716914366596,
+      "grad_norm": 0.17821278496624815,
+      "learning_rate": 3.672612801678909e-06,
+      "loss": 0.2825,
+      "step": 1979
+    },
+    {
+      "epoch": 2.8011323425336165,
+      "grad_norm": 0.1807914249066901,
+      "learning_rate": 3.6463798530954877e-06,
+      "loss": 0.2713,
+      "step": 1980
+    },
+    {
+      "epoch": 2.802547770700637,
+      "grad_norm": 0.18384108470044716,
+      "learning_rate": 3.620146904512067e-06,
+      "loss": 0.273,
+      "step": 1981
+    },
+    {
+      "epoch": 2.8039631988676574,
+      "grad_norm": 0.17353001772169335,
+      "learning_rate": 3.5939139559286466e-06,
+      "loss": 0.2825,
+      "step": 1982
+    },
+    {
+      "epoch": 2.805378627034678,
+      "grad_norm": 0.1840707179807425,
+      "learning_rate": 3.5676810073452256e-06,
+      "loss": 0.2872,
+      "step": 1983
+    },
+    {
+      "epoch": 2.8067940552016983,
+      "grad_norm": 0.17514367701031278,
+      "learning_rate": 3.541448058761805e-06,
+      "loss": 0.2869,
+      "step": 1984
+    },
+    {
+      "epoch": 2.808209483368719,
+      "grad_norm": 0.17928029059399161,
+      "learning_rate": 3.515215110178384e-06,
+      "loss": 0.2831,
+      "step": 1985
+    },
+    {
+      "epoch": 2.8096249115357397,
+      "grad_norm": 0.17848612415415777,
+      "learning_rate": 3.4889821615949635e-06,
+      "loss": 0.2797,
+      "step": 1986
+    },
+    {
+      "epoch": 2.8110403397027603,
+      "grad_norm": 0.17737711539709516,
+      "learning_rate": 3.462749213011543e-06,
+      "loss": 0.2949,
+      "step": 1987
+    },
+    {
+      "epoch": 2.8124557678697806,
+      "grad_norm": 0.17399136200106763,
+      "learning_rate": 3.436516264428122e-06,
+      "loss": 0.2937,
+      "step": 1988
+    },
+    {
+      "epoch": 2.8138711960368012,
+      "grad_norm": 0.18242700749152266,
+      "learning_rate": 3.4102833158447014e-06,
+      "loss": 0.2868,
+      "step": 1989
+    },
+    {
+      "epoch": 2.8152866242038215,
+      "grad_norm": 0.1771122346213806,
+      "learning_rate": 3.38405036726128e-06,
+      "loss": 0.2884,
+      "step": 1990
+    },
+    {
+      "epoch": 2.816702052370842,
+      "grad_norm": 0.17652808422220528,
+      "learning_rate": 3.35781741867786e-06,
+      "loss": 0.2836,
+      "step": 1991
+    },
+    {
+      "epoch": 2.818117480537863,
+      "grad_norm": 0.17830076323575594,
+      "learning_rate": 3.3315844700944385e-06,
+      "loss": 0.2817,
+      "step": 1992
+    },
+    {
+      "epoch": 2.8195329087048835,
+      "grad_norm": 0.18223032359971642,
+      "learning_rate": 3.305351521511018e-06,
+      "loss": 0.2943,
+      "step": 1993
+    },
+    {
+      "epoch": 2.8209483368719037,
+      "grad_norm": 0.17657096233286845,
+      "learning_rate": 3.2791185729275973e-06,
+      "loss": 0.2907,
+      "step": 1994
+    },
+    {
+      "epoch": 2.8223637650389244,
+      "grad_norm": 0.17075541903115724,
+      "learning_rate": 3.2528856243441763e-06,
+      "loss": 0.2749,
+      "step": 1995
+    },
+    {
+      "epoch": 2.8237791932059446,
+      "grad_norm": 0.17384883418147812,
+      "learning_rate": 3.2266526757607558e-06,
+      "loss": 0.294,
+      "step": 1996
+    },
+    {
+      "epoch": 2.8251946213729653,
+      "grad_norm": 0.18292539916744055,
+      "learning_rate": 3.200419727177335e-06,
+      "loss": 0.2781,
+      "step": 1997
+    },
+    {
+      "epoch": 2.826610049539986,
+      "grad_norm": 0.17339033962692266,
+      "learning_rate": 3.1741867785939142e-06,
+      "loss": 0.2765,
+      "step": 1998
+    },
+    {
+      "epoch": 2.8280254777070066,
+      "grad_norm": 0.17691626215998835,
+      "learning_rate": 3.147953830010493e-06,
+      "loss": 0.2962,
+      "step": 1999
+    },
+    {
+      "epoch": 2.829440905874027,
+      "grad_norm": 0.1766712989401266,
+      "learning_rate": 3.1217208814270727e-06,
+      "loss": 0.2796,
+      "step": 2000
+    },
+    {
+      "epoch": 2.8308563340410475,
+      "grad_norm": 0.18888976742039038,
+      "learning_rate": 3.0954879328436517e-06,
+      "loss": 0.2871,
+      "step": 2001
+    },
+    {
+      "epoch": 2.8322717622080678,
+      "grad_norm": 0.18792724511405656,
+      "learning_rate": 3.0692549842602307e-06,
+      "loss": 0.2902,
+      "step": 2002
+    },
+    {
+      "epoch": 2.8336871903750884,
+      "grad_norm": 0.16425281618847395,
+      "learning_rate": 3.04302203567681e-06,
+      "loss": 0.273,
+      "step": 2003
+    },
+    {
+      "epoch": 2.835102618542109,
+      "grad_norm": 0.1746146438734197,
+      "learning_rate": 3.0167890870933896e-06,
+      "loss": 0.2826,
+      "step": 2004
+    },
+    {
+      "epoch": 2.8365180467091298,
+      "grad_norm": 0.19294227092682006,
+      "learning_rate": 2.9905561385099686e-06,
+      "loss": 0.2871,
+      "step": 2005
+    },
+    {
+      "epoch": 2.83793347487615,
+      "grad_norm": 0.1776014193267871,
+      "learning_rate": 2.964323189926548e-06,
+      "loss": 0.2886,
+      "step": 2006
+    },
+    {
+      "epoch": 2.8393489030431707,
+      "grad_norm": 0.1798288722394478,
+      "learning_rate": 2.938090241343127e-06,
+      "loss": 0.2929,
+      "step": 2007
+    },
+    {
+      "epoch": 2.840764331210191,
+      "grad_norm": 0.18006032181040652,
+      "learning_rate": 2.911857292759706e-06,
+      "loss": 0.2716,
+      "step": 2008
+    },
+    {
+      "epoch": 2.8421797593772116,
+      "grad_norm": 0.18136764338845685,
+      "learning_rate": 2.8856243441762855e-06,
+      "loss": 0.2804,
+      "step": 2009
+    },
+    {
+      "epoch": 2.8435951875442322,
+      "grad_norm": 0.1702942902210057,
+      "learning_rate": 2.859391395592865e-06,
+      "loss": 0.283,
+      "step": 2010
+    },
+    {
+      "epoch": 2.845010615711253,
+      "grad_norm": 0.16858535738863575,
+      "learning_rate": 2.833158447009444e-06,
+      "loss": 0.2715,
+      "step": 2011
+    },
+    {
+      "epoch": 2.846426043878273,
+      "grad_norm": 0.17249587169123212,
+      "learning_rate": 2.8069254984260234e-06,
+      "loss": 0.2727,
+      "step": 2012
+    },
+    {
+      "epoch": 2.847841472045294,
+      "grad_norm": 0.1854098483141046,
+      "learning_rate": 2.7806925498426024e-06,
+      "loss": 0.2958,
+      "step": 2013
+    },
+    {
+      "epoch": 2.849256900212314,
+      "grad_norm": 0.18003906868221864,
+      "learning_rate": 2.7544596012591814e-06,
+      "loss": 0.2899,
+      "step": 2014
+    },
+    {
+      "epoch": 2.8506723283793347,
+      "grad_norm": 0.1848795205341624,
+      "learning_rate": 2.728226652675761e-06,
+      "loss": 0.2906,
+      "step": 2015
+    },
+    {
+      "epoch": 2.8520877565463554,
+      "grad_norm": 0.18063285127831286,
+      "learning_rate": 2.7019937040923403e-06,
+      "loss": 0.2775,
+      "step": 2016
+    },
+    {
+      "epoch": 2.853503184713376,
+      "grad_norm": 0.17428351528968997,
+      "learning_rate": 2.6757607555089193e-06,
+      "loss": 0.2847,
+      "step": 2017
+    },
+    {
+      "epoch": 2.8549186128803963,
+      "grad_norm": 0.1948698715579363,
+      "learning_rate": 2.6495278069254988e-06,
+      "loss": 0.2971,
+      "step": 2018
+    },
+    {
+      "epoch": 2.856334041047417,
+      "grad_norm": 0.17185235351608957,
+      "learning_rate": 2.6232948583420778e-06,
+      "loss": 0.2815,
+      "step": 2019
+    },
+    {
+      "epoch": 2.857749469214437,
+      "grad_norm": 0.16970422017649608,
+      "learning_rate": 2.597061909758657e-06,
+      "loss": 0.271,
+      "step": 2020
+    },
+    {
+      "epoch": 2.859164897381458,
+      "grad_norm": 0.18560638216370043,
+      "learning_rate": 2.5708289611752362e-06,
+      "loss": 0.2865,
+      "step": 2021
+    },
+    {
+      "epoch": 2.8605803255484785,
+      "grad_norm": 0.18118828426589775,
+      "learning_rate": 2.5445960125918157e-06,
+      "loss": 0.2957,
+      "step": 2022
+    },
+    {
+      "epoch": 2.861995753715499,
+      "grad_norm": 0.1766585834822367,
+      "learning_rate": 2.5183630640083947e-06,
+      "loss": 0.2756,
+      "step": 2023
+    },
+    {
+      "epoch": 2.8634111818825194,
+      "grad_norm": 0.1717726611092561,
+      "learning_rate": 2.4921301154249737e-06,
+      "loss": 0.2811,
+      "step": 2024
+    },
+    {
+      "epoch": 2.86482661004954,
+      "grad_norm": 0.17868368522308714,
+      "learning_rate": 2.465897166841553e-06,
+      "loss": 0.2955,
+      "step": 2025
+    },
+    {
+      "epoch": 2.8662420382165603,
+      "grad_norm": 0.17728811150135537,
+      "learning_rate": 2.439664218258132e-06,
+      "loss": 0.2925,
+      "step": 2026
+    },
+    {
+      "epoch": 2.867657466383581,
+      "grad_norm": 0.1690730724900412,
+      "learning_rate": 2.4134312696747116e-06,
+      "loss": 0.284,
+      "step": 2027
+    },
+    {
+      "epoch": 2.8690728945506017,
+      "grad_norm": 0.17088543157293817,
+      "learning_rate": 2.387198321091291e-06,
+      "loss": 0.2959,
+      "step": 2028
+    },
+    {
+      "epoch": 2.8704883227176223,
+      "grad_norm": 0.17345729443031038,
+      "learning_rate": 2.36096537250787e-06,
+      "loss": 0.2806,
+      "step": 2029
+    },
+    {
+      "epoch": 2.8719037508846426,
+      "grad_norm": 0.17168864907383077,
+      "learning_rate": 2.334732423924449e-06,
+      "loss": 0.2899,
+      "step": 2030
+    },
+    {
+      "epoch": 2.8733191790516632,
+      "grad_norm": 0.16892911484230724,
+      "learning_rate": 2.3084994753410285e-06,
+      "loss": 0.2868,
+      "step": 2031
+    },
+    {
+      "epoch": 2.8747346072186835,
+      "grad_norm": 0.1870298401137464,
+      "learning_rate": 2.2822665267576075e-06,
+      "loss": 0.2787,
+      "step": 2032
+    },
+    {
+      "epoch": 2.876150035385704,
+      "grad_norm": 0.17388283426752,
+      "learning_rate": 2.256033578174187e-06,
+      "loss": 0.2784,
+      "step": 2033
+    },
+    {
+      "epoch": 2.877565463552725,
+      "grad_norm": 0.18035167623450257,
+      "learning_rate": 2.2298006295907664e-06,
+      "loss": 0.2962,
+      "step": 2034
+    },
+    {
+      "epoch": 2.8789808917197455,
+      "grad_norm": 0.17859035676811588,
+      "learning_rate": 2.2035676810073454e-06,
+      "loss": 0.2866,
+      "step": 2035
+    },
+    {
+      "epoch": 2.8803963198867657,
+      "grad_norm": 0.17367103802023898,
+      "learning_rate": 2.1773347324239244e-06,
+      "loss": 0.295,
+      "step": 2036
+    },
+    {
+      "epoch": 2.8818117480537864,
+      "grad_norm": 0.1735270033159891,
+      "learning_rate": 2.151101783840504e-06,
+      "loss": 0.277,
+      "step": 2037
+    },
+    {
+      "epoch": 2.8832271762208066,
+      "grad_norm": 0.17076648095639954,
+      "learning_rate": 2.124868835257083e-06,
+      "loss": 0.2779,
+      "step": 2038
+    },
+    {
+      "epoch": 2.8846426043878273,
+      "grad_norm": 0.17769982420123154,
+      "learning_rate": 2.0986358866736623e-06,
+      "loss": 0.2773,
+      "step": 2039
+    },
+    {
+      "epoch": 2.886058032554848,
+      "grad_norm": 0.17716108684876158,
+      "learning_rate": 2.0724029380902413e-06,
+      "loss": 0.2899,
+      "step": 2040
+    },
+    {
+      "epoch": 2.8874734607218686,
+      "grad_norm": 0.20152356634552215,
+      "learning_rate": 2.0461699895068208e-06,
+      "loss": 0.3025,
+      "step": 2041
+    },
+    {
+      "epoch": 2.888888888888889,
+      "grad_norm": 0.18105580605108734,
+      "learning_rate": 2.0199370409233998e-06,
+      "loss": 0.2854,
+      "step": 2042
+    },
+    {
+      "epoch": 2.8903043170559095,
+      "grad_norm": 0.1701357726201586,
+      "learning_rate": 1.9937040923399792e-06,
+      "loss": 0.2747,
+      "step": 2043
+    },
+    {
+      "epoch": 2.8917197452229297,
+      "grad_norm": 0.17670981194170432,
+      "learning_rate": 1.9674711437565582e-06,
+      "loss": 0.2811,
+      "step": 2044
+    },
+    {
+      "epoch": 2.8931351733899504,
+      "grad_norm": 0.18421121623096698,
+      "learning_rate": 1.9412381951731377e-06,
+      "loss": 0.2819,
+      "step": 2045
+    },
+    {
+      "epoch": 2.894550601556971,
+      "grad_norm": 0.18030333227492373,
+      "learning_rate": 1.9150052465897167e-06,
+      "loss": 0.2819,
+      "step": 2046
+    },
+    {
+      "epoch": 2.8959660297239918,
+      "grad_norm": 0.17001697349326303,
+      "learning_rate": 1.8887722980062961e-06,
+      "loss": 0.2768,
+      "step": 2047
+    },
+    {
+      "epoch": 2.897381457891012,
+      "grad_norm": 0.17961651630100461,
+      "learning_rate": 1.8625393494228753e-06,
+      "loss": 0.2985,
+      "step": 2048
+    },
+    {
+      "epoch": 2.8987968860580327,
+      "grad_norm": 0.17265583195377293,
+      "learning_rate": 1.8363064008394546e-06,
+      "loss": 0.2845,
+      "step": 2049
+    },
+    {
+      "epoch": 2.900212314225053,
+      "grad_norm": 0.17269811076363084,
+      "learning_rate": 1.8100734522560336e-06,
+      "loss": 0.2775,
+      "step": 2050
+    },
+    {
+      "epoch": 2.9016277423920736,
+      "grad_norm": 0.18776579560972845,
+      "learning_rate": 1.7838405036726128e-06,
+      "loss": 0.2838,
+      "step": 2051
+    },
+    {
+      "epoch": 2.903043170559094,
+      "grad_norm": 0.17867741388267788,
+      "learning_rate": 1.757607555089192e-06,
+      "loss": 0.2891,
+      "step": 2052
+    },
+    {
+      "epoch": 2.904458598726115,
+      "grad_norm": 0.17846859530717932,
+      "learning_rate": 1.7313746065057715e-06,
+      "loss": 0.3017,
+      "step": 2053
+    },
+    {
+      "epoch": 2.905874026893135,
+      "grad_norm": 0.18241316656654646,
+      "learning_rate": 1.7051416579223507e-06,
+      "loss": 0.306,
+      "step": 2054
+    },
+    {
+      "epoch": 2.907289455060156,
+      "grad_norm": 0.17810706516304647,
+      "learning_rate": 1.67890870933893e-06,
+      "loss": 0.2936,
+      "step": 2055
+    },
+    {
+      "epoch": 2.908704883227176,
+      "grad_norm": 0.17545134113009844,
+      "learning_rate": 1.652675760755509e-06,
+      "loss": 0.2759,
+      "step": 2056
+    },
+    {
+      "epoch": 2.9101203113941967,
+      "grad_norm": 0.16912469809935698,
+      "learning_rate": 1.6264428121720882e-06,
+      "loss": 0.2697,
+      "step": 2057
+    },
+    {
+      "epoch": 2.9115357395612174,
+      "grad_norm": 0.17386103881783685,
+      "learning_rate": 1.6002098635886674e-06,
+      "loss": 0.2802,
+      "step": 2058
+    },
+    {
+      "epoch": 2.912951167728238,
+      "grad_norm": 0.16935816411411495,
+      "learning_rate": 1.5739769150052464e-06,
+      "loss": 0.2733,
+      "step": 2059
+    },
+    {
+      "epoch": 2.9143665958952583,
+      "grad_norm": 0.17409526578818196,
+      "learning_rate": 1.5477439664218259e-06,
+      "loss": 0.2806,
+      "step": 2060
+    },
+    {
+      "epoch": 2.915782024062279,
+      "grad_norm": 0.17528174653087739,
+      "learning_rate": 1.521511017838405e-06,
+      "loss": 0.2805,
+      "step": 2061
+    },
+    {
+      "epoch": 2.917197452229299,
+      "grad_norm": 0.17415870629933414,
+      "learning_rate": 1.4952780692549843e-06,
+      "loss": 0.2721,
+      "step": 2062
+    },
+    {
+      "epoch": 2.91861288039632,
+      "grad_norm": 0.1674190257626401,
+      "learning_rate": 1.4690451206715635e-06,
+      "loss": 0.2777,
+      "step": 2063
+    },
+    {
+      "epoch": 2.9200283085633405,
+      "grad_norm": 0.18890631457900856,
+      "learning_rate": 1.4428121720881428e-06,
+      "loss": 0.2888,
+      "step": 2064
+    },
+    {
+      "epoch": 2.921443736730361,
+      "grad_norm": 0.17650409521486077,
+      "learning_rate": 1.416579223504722e-06,
+      "loss": 0.2848,
+      "step": 2065
+    },
+    {
+      "epoch": 2.9228591648973814,
+      "grad_norm": 0.17415093004971588,
+      "learning_rate": 1.3903462749213012e-06,
+      "loss": 0.266,
+      "step": 2066
+    },
+    {
+      "epoch": 2.924274593064402,
+      "grad_norm": 0.17889620724483848,
+      "learning_rate": 1.3641133263378804e-06,
+      "loss": 0.2856,
+      "step": 2067
+    },
+    {
+      "epoch": 2.9256900212314223,
+      "grad_norm": 0.1747134497606467,
+      "learning_rate": 1.3378803777544597e-06,
+      "loss": 0.2939,
+      "step": 2068
+    },
+    {
+      "epoch": 2.927105449398443,
+      "grad_norm": 0.16967566671723144,
+      "learning_rate": 1.3116474291710389e-06,
+      "loss": 0.2892,
+      "step": 2069
+    },
+    {
+      "epoch": 2.9285208775654636,
+      "grad_norm": 0.17429815642737306,
+      "learning_rate": 1.2854144805876181e-06,
+      "loss": 0.2825,
+      "step": 2070
+    },
+    {
+      "epoch": 2.9299363057324843,
+      "grad_norm": 0.17334968710760035,
+      "learning_rate": 1.2591815320041973e-06,
+      "loss": 0.2878,
+      "step": 2071
+    },
+    {
+      "epoch": 2.9313517338995045,
+      "grad_norm": 0.17764237449975134,
+      "learning_rate": 1.2329485834207766e-06,
+      "loss": 0.2822,
+      "step": 2072
+    },
+    {
+      "epoch": 2.932767162066525,
+      "grad_norm": 0.1742807974773749,
+      "learning_rate": 1.2067156348373558e-06,
+      "loss": 0.2999,
+      "step": 2073
+    },
+    {
+      "epoch": 2.9341825902335454,
+      "grad_norm": 0.1720479125811444,
+      "learning_rate": 1.180482686253935e-06,
+      "loss": 0.2843,
+      "step": 2074
+    },
+    {
+      "epoch": 2.935598018400566,
+      "grad_norm": 0.17424562122848383,
+      "learning_rate": 1.1542497376705142e-06,
+      "loss": 0.2824,
+      "step": 2075
+    },
+    {
+      "epoch": 2.937013446567587,
+      "grad_norm": 0.1702802994215502,
+      "learning_rate": 1.1280167890870935e-06,
+      "loss": 0.2827,
+      "step": 2076
+    },
+    {
+      "epoch": 2.9384288747346075,
+      "grad_norm": 0.17724938854296238,
+      "learning_rate": 1.1017838405036727e-06,
+      "loss": 0.2665,
+      "step": 2077
+    },
+    {
+      "epoch": 2.9398443029016277,
+      "grad_norm": 0.16881102462161476,
+      "learning_rate": 1.075550891920252e-06,
+      "loss": 0.2875,
+      "step": 2078
+    },
+    {
+      "epoch": 2.9412597310686484,
+      "grad_norm": 0.17586128182531433,
+      "learning_rate": 1.0493179433368312e-06,
+      "loss": 0.2918,
+      "step": 2079
+    },
+    {
+      "epoch": 2.9426751592356686,
+      "grad_norm": 0.17880544066965168,
+      "learning_rate": 1.0230849947534104e-06,
+      "loss": 0.2848,
+      "step": 2080
+    },
+    {
+      "epoch": 2.9440905874026893,
+      "grad_norm": 0.17612939411552903,
+      "learning_rate": 9.968520461699896e-07,
+      "loss": 0.2989,
+      "step": 2081
+    },
+    {
+      "epoch": 2.94550601556971,
+      "grad_norm": 0.17079837970890843,
+      "learning_rate": 9.706190975865688e-07,
+      "loss": 0.2826,
+      "step": 2082
+    },
+    {
+      "epoch": 2.9469214437367306,
+      "grad_norm": 0.17576729615022918,
+      "learning_rate": 9.443861490031481e-07,
+      "loss": 0.294,
+      "step": 2083
+    },
+    {
+      "epoch": 2.948336871903751,
+      "grad_norm": 0.1759809699335125,
+      "learning_rate": 9.181532004197273e-07,
+      "loss": 0.2708,
+      "step": 2084
+    },
+    {
+      "epoch": 2.9497523000707715,
+      "grad_norm": 0.17217197283212832,
+      "learning_rate": 8.919202518363064e-07,
+      "loss": 0.2884,
+      "step": 2085
+    },
+    {
+      "epoch": 2.9511677282377917,
+      "grad_norm": 0.17418729257127177,
+      "learning_rate": 8.656873032528857e-07,
+      "loss": 0.2817,
+      "step": 2086
+    },
+    {
+      "epoch": 2.9525831564048124,
+      "grad_norm": 0.17445904450889047,
+      "learning_rate": 8.39454354669465e-07,
+      "loss": 0.2913,
+      "step": 2087
+    },
+    {
+      "epoch": 2.953998584571833,
+      "grad_norm": 0.17258868036349787,
+      "learning_rate": 8.132214060860441e-07,
+      "loss": 0.2907,
+      "step": 2088
+    },
+    {
+      "epoch": 2.9554140127388537,
+      "grad_norm": 0.17168645858086345,
+      "learning_rate": 7.869884575026232e-07,
+      "loss": 0.2807,
+      "step": 2089
+    },
+    {
+      "epoch": 2.956829440905874,
+      "grad_norm": 0.1709694601992102,
+      "learning_rate": 7.607555089192025e-07,
+      "loss": 0.286,
+      "step": 2090
+    },
+    {
+      "epoch": 2.9582448690728946,
+      "grad_norm": 0.18076593969351598,
+      "learning_rate": 7.345225603357818e-07,
+      "loss": 0.2935,
+      "step": 2091
+    },
+    {
+      "epoch": 2.959660297239915,
+      "grad_norm": 0.16122100248321902,
+      "learning_rate": 7.08289611752361e-07,
+      "loss": 0.2537,
+      "step": 2092
+    },
+    {
+      "epoch": 2.9610757254069355,
+      "grad_norm": 0.17460588873108918,
+      "learning_rate": 6.820566631689402e-07,
+      "loss": 0.288,
+      "step": 2093
+    },
+    {
+      "epoch": 2.962491153573956,
+      "grad_norm": 0.17928398628197456,
+      "learning_rate": 6.558237145855194e-07,
+      "loss": 0.2734,
+      "step": 2094
+    },
+    {
+      "epoch": 2.963906581740977,
+      "grad_norm": 0.17212227085116813,
+      "learning_rate": 6.295907660020987e-07,
+      "loss": 0.271,
+      "step": 2095
+    },
+    {
+      "epoch": 2.965322009907997,
+      "grad_norm": 0.16846021432755692,
+      "learning_rate": 6.033578174186779e-07,
+      "loss": 0.2882,
+      "step": 2096
+    },
+    {
+      "epoch": 2.9667374380750178,
+      "grad_norm": 0.16842469893670636,
+      "learning_rate": 5.771248688352571e-07,
+      "loss": 0.2709,
+      "step": 2097
+    },
+    {
+      "epoch": 2.968152866242038,
+      "grad_norm": 0.17418546182047925,
+      "learning_rate": 5.508919202518364e-07,
+      "loss": 0.2824,
+      "step": 2098
+    },
+    {
+      "epoch": 2.9695682944090587,
+      "grad_norm": 0.1755673694382143,
+      "learning_rate": 5.246589716684156e-07,
+      "loss": 0.2913,
+      "step": 2099
+    },
+    {
+      "epoch": 2.9709837225760793,
+      "grad_norm": 0.1739443277389086,
+      "learning_rate": 4.984260230849948e-07,
+      "loss": 0.2842,
+      "step": 2100
+    },
+    {
+      "epoch": 2.9723991507431,
+      "grad_norm": 0.16962826263262715,
+      "learning_rate": 4.7219307450157403e-07,
+      "loss": 0.2786,
+      "step": 2101
+    },
+    {
+      "epoch": 2.9738145789101202,
+      "grad_norm": 0.1844668822431492,
+      "learning_rate": 4.459601259181532e-07,
+      "loss": 0.3197,
+      "step": 2102
+    },
+    {
+      "epoch": 2.975230007077141,
+      "grad_norm": 0.17038166826214368,
+      "learning_rate": 4.197271773347325e-07,
+      "loss": 0.2863,
+      "step": 2103
+    },
+    {
+      "epoch": 2.976645435244161,
+      "grad_norm": 0.1654973461554926,
+      "learning_rate": 3.934942287513116e-07,
+      "loss": 0.2733,
+      "step": 2104
+    },
+    {
+      "epoch": 2.978060863411182,
+      "grad_norm": 0.17118209967441136,
+      "learning_rate": 3.672612801678909e-07,
+      "loss": 0.2771,
+      "step": 2105
+    },
+    {
+      "epoch": 2.9794762915782025,
+      "grad_norm": 0.17596578468972426,
+      "learning_rate": 3.410283315844701e-07,
+      "loss": 0.2783,
+      "step": 2106
+    },
+    {
+      "epoch": 2.980891719745223,
+      "grad_norm": 0.1683430471571756,
+      "learning_rate": 3.1479538300104934e-07,
+      "loss": 0.2735,
+      "step": 2107
+    },
+    {
+      "epoch": 2.9823071479122434,
+      "grad_norm": 0.16686184103722593,
+      "learning_rate": 2.8856243441762856e-07,
+      "loss": 0.2687,
+      "step": 2108
+    },
+    {
+      "epoch": 2.983722576079264,
+      "grad_norm": 0.17598665900195548,
+      "learning_rate": 2.623294858342078e-07,
+      "loss": 0.2975,
+      "step": 2109
+    },
+    {
+      "epoch": 2.9851380042462843,
+      "grad_norm": 0.19097911943519666,
+      "learning_rate": 2.3609653725078701e-07,
+      "loss": 0.3143,
+      "step": 2110
+    },
+    {
+      "epoch": 2.986553432413305,
+      "grad_norm": 0.18505478775984915,
+      "learning_rate": 2.0986358866736624e-07,
+      "loss": 0.267,
+      "step": 2111
+    },
+    {
+      "epoch": 2.9879688605803256,
+      "grad_norm": 0.16498905166042235,
+      "learning_rate": 1.8363064008394544e-07,
+      "loss": 0.2798,
+      "step": 2112
+    },
+    {
+      "epoch": 2.9893842887473463,
+      "grad_norm": 0.17010577207806246,
+      "learning_rate": 1.5739769150052467e-07,
+      "loss": 0.2779,
+      "step": 2113
+    },
+    {
+      "epoch": 2.9907997169143665,
+      "grad_norm": 0.1670793382426729,
+      "learning_rate": 1.311647429171039e-07,
+      "loss": 0.2849,
+      "step": 2114
+    },
+    {
+      "epoch": 2.992215145081387,
+      "grad_norm": 0.17153417278807456,
+      "learning_rate": 1.0493179433368312e-07,
+      "loss": 0.2767,
+      "step": 2115
+    },
+    {
+      "epoch": 2.9936305732484074,
+      "grad_norm": 0.16578146300653185,
+      "learning_rate": 7.869884575026233e-08,
+      "loss": 0.2841,
+      "step": 2116
+    },
+    {
+      "epoch": 2.995046001415428,
+      "grad_norm": 0.16854873982311497,
+      "learning_rate": 5.246589716684156e-08,
+      "loss": 0.2929,
+      "step": 2117
+    },
+    {
+      "epoch": 2.9964614295824488,
+      "grad_norm": 0.18959798822634213,
+      "learning_rate": 2.623294858342078e-08,
+      "loss": 0.2741,
+      "step": 2118
+    },
+    {
+      "epoch": 2.9964614295824488,
+      "step": 2118,
+      "total_flos": 2.3544511201506492e+19,
+      "train_loss": 0.4426687185470511,
+      "train_runtime": 62711.6278,
+      "train_samples_per_second": 0.541,
+      "train_steps_per_second": 0.034
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2118,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.3544511201506492e+19,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}