diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7406 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 2.0,
+  "eval_steps": 500,
+  "global_step": 1052,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0019011406844106464,
+      "grad_norm": 3.6272729960894416,
+      "learning_rate": 1.886792452830189e-07,
+      "loss": 0.6953,
+      "step": 1
+    },
+    {
+      "epoch": 0.0038022813688212928,
+      "grad_norm": 3.2468947214808432,
+      "learning_rate": 3.773584905660378e-07,
+      "loss": 0.7108,
+      "step": 2
+    },
+    {
+      "epoch": 0.005703422053231939,
+      "grad_norm": 3.486445528309055,
+      "learning_rate": 5.660377358490567e-07,
+      "loss": 0.7032,
+      "step": 3
+    },
+    {
+      "epoch": 0.0076045627376425855,
+      "grad_norm": 3.0122196989099987,
+      "learning_rate": 7.547169811320755e-07,
+      "loss": 0.6852,
+      "step": 4
+    },
+    {
+      "epoch": 0.009505703422053232,
+      "grad_norm": 3.4205582033691924,
+      "learning_rate": 9.433962264150944e-07,
+      "loss": 0.7375,
+      "step": 5
+    },
+    {
+      "epoch": 0.011406844106463879,
+      "grad_norm": 3.006486996025197,
+      "learning_rate": 1.1320754716981133e-06,
+      "loss": 0.6784,
+      "step": 6
+    },
+    {
+      "epoch": 0.013307984790874524,
+      "grad_norm": 3.1315959019738853,
+      "learning_rate": 1.3207547169811322e-06,
+      "loss": 0.6847,
+      "step": 7
+    },
+    {
+      "epoch": 0.015209125475285171,
+      "grad_norm": 3.1088855374359716,
+      "learning_rate": 1.509433962264151e-06,
+      "loss": 0.7266,
+      "step": 8
+    },
+    {
+      "epoch": 0.017110266159695818,
+      "grad_norm": 2.641777348986974,
+      "learning_rate": 1.6981132075471698e-06,
+      "loss": 0.6968,
+      "step": 9
+    },
+    {
+      "epoch": 0.019011406844106463,
+      "grad_norm": 2.378691549997581,
+      "learning_rate": 1.8867924528301889e-06,
+      "loss": 0.6771,
+      "step": 10
+    },
+    {
+      "epoch": 0.02091254752851711,
+      "grad_norm": 1.8184950846143932,
+      "learning_rate": 2.075471698113208e-06,
+      "loss": 0.6635,
+      "step": 11
+    },
+    {
+      "epoch": 0.022813688212927757,
+      "grad_norm": 1.6179320724283004,
+      "learning_rate": 2.2641509433962266e-06,
+      "loss": 0.5912,
+      "step": 12
+    },
+    {
+      "epoch": 0.024714828897338403,
+      "grad_norm": 1.5781638900111103,
+      "learning_rate": 2.4528301886792453e-06,
+      "loss": 0.5932,
+      "step": 13
+    },
+    {
+      "epoch": 0.026615969581749048,
+      "grad_norm": 1.3880101183259435,
+      "learning_rate": 2.6415094339622644e-06,
+      "loss": 0.5926,
+      "step": 14
+    },
+    {
+      "epoch": 0.028517110266159697,
+      "grad_norm": 1.9026105993842082,
+      "learning_rate": 2.830188679245283e-06,
+      "loss": 0.6004,
+      "step": 15
+    },
+    {
+      "epoch": 0.030418250950570342,
+      "grad_norm": 2.219843038788561,
+      "learning_rate": 3.018867924528302e-06,
+      "loss": 0.5789,
+      "step": 16
+    },
+    {
+      "epoch": 0.03231939163498099,
+      "grad_norm": 2.2375054102375085,
+      "learning_rate": 3.207547169811321e-06,
+      "loss": 0.5751,
+      "step": 17
+    },
+    {
+      "epoch": 0.034220532319391636,
+      "grad_norm": 1.7476948346847236,
+      "learning_rate": 3.3962264150943395e-06,
+      "loss": 0.5749,
+      "step": 18
+    },
+    {
+      "epoch": 0.03612167300380228,
+      "grad_norm": 1.4624815634258725,
+      "learning_rate": 3.5849056603773586e-06,
+      "loss": 0.5455,
+      "step": 19
+    },
+    {
+      "epoch": 0.03802281368821293,
+      "grad_norm": 1.2973743959859565,
+      "learning_rate": 3.7735849056603777e-06,
+      "loss": 0.5639,
+      "step": 20
+    },
+    {
+      "epoch": 0.039923954372623575,
+      "grad_norm": 1.4289102123524366,
+      "learning_rate": 3.962264150943396e-06,
+      "loss": 0.543,
+      "step": 21
+    },
+    {
+      "epoch": 0.04182509505703422,
+      "grad_norm": 1.4043133089171078,
+      "learning_rate": 4.150943396226416e-06,
+      "loss": 0.584,
+      "step": 22
+    },
+    {
+      "epoch": 0.043726235741444866,
+      "grad_norm": 1.3778691083913985,
+      "learning_rate": 4.339622641509435e-06,
+      "loss": 0.5602,
+      "step": 23
+    },
+    {
+      "epoch": 0.045627376425855515,
+      "grad_norm": 1.1215774936807987,
+      "learning_rate": 4.528301886792453e-06,
+      "loss": 0.5432,
+      "step": 24
+    },
+    {
+      "epoch": 0.04752851711026616,
+      "grad_norm": 1.1029418346946827,
+      "learning_rate": 4.716981132075472e-06,
+      "loss": 0.5506,
+      "step": 25
+    },
+    {
+      "epoch": 0.049429657794676805,
+      "grad_norm": 1.012651977101632,
+      "learning_rate": 4.905660377358491e-06,
+      "loss": 0.5474,
+      "step": 26
+    },
+    {
+      "epoch": 0.051330798479087454,
+      "grad_norm": 1.0242097598078577,
+      "learning_rate": 5.09433962264151e-06,
+      "loss": 0.5348,
+      "step": 27
+    },
+    {
+      "epoch": 0.053231939163498096,
+      "grad_norm": 1.0560515495477218,
+      "learning_rate": 5.283018867924529e-06,
+      "loss": 0.5383,
+      "step": 28
+    },
+    {
+      "epoch": 0.055133079847908745,
+      "grad_norm": 0.9905358172203085,
+      "learning_rate": 5.4716981132075475e-06,
+      "loss": 0.528,
+      "step": 29
+    },
+    {
+      "epoch": 0.057034220532319393,
+      "grad_norm": 0.85672422158207,
+      "learning_rate": 5.660377358490566e-06,
+      "loss": 0.5094,
+      "step": 30
+    },
+    {
+      "epoch": 0.058935361216730035,
+      "grad_norm": 0.9239353784760164,
+      "learning_rate": 5.849056603773585e-06,
+      "loss": 0.5098,
+      "step": 31
+    },
+    {
+      "epoch": 0.060836501901140684,
+      "grad_norm": 0.8719701698834567,
+      "learning_rate": 6.037735849056604e-06,
+      "loss": 0.4943,
+      "step": 32
+    },
+    {
+      "epoch": 0.06273764258555133,
+      "grad_norm": 0.9045228037202876,
+      "learning_rate": 6.226415094339623e-06,
+      "loss": 0.5118,
+      "step": 33
+    },
+    {
+      "epoch": 0.06463878326996197,
+      "grad_norm": 0.961340523762788,
+      "learning_rate": 6.415094339622642e-06,
+      "loss": 0.5244,
+      "step": 34
+    },
+    {
+      "epoch": 0.06653992395437262,
+      "grad_norm": 0.9250528699050148,
+      "learning_rate": 6.60377358490566e-06,
+      "loss": 0.4979,
+      "step": 35
+    },
+    {
+      "epoch": 0.06844106463878327,
+      "grad_norm": 0.8112089041298184,
+      "learning_rate": 6.792452830188679e-06,
+      "loss": 0.487,
+      "step": 36
+    },
+    {
+      "epoch": 0.07034220532319392,
+      "grad_norm": 0.8175588297328692,
+      "learning_rate": 6.981132075471699e-06,
+      "loss": 0.4572,
+      "step": 37
+    },
+    {
+      "epoch": 0.07224334600760456,
+      "grad_norm": 0.9607557144392525,
+      "learning_rate": 7.169811320754717e-06,
+      "loss": 0.5254,
+      "step": 38
+    },
+    {
+      "epoch": 0.0741444866920152,
+      "grad_norm": 0.9717917789725954,
+      "learning_rate": 7.358490566037736e-06,
+      "loss": 0.4999,
+      "step": 39
+    },
+    {
+      "epoch": 0.07604562737642585,
+      "grad_norm": 0.8661871379846303,
+      "learning_rate": 7.5471698113207555e-06,
+      "loss": 0.4934,
+      "step": 40
+    },
+    {
+      "epoch": 0.0779467680608365,
+      "grad_norm": 0.9458603773931039,
+      "learning_rate": 7.735849056603775e-06,
+      "loss": 0.5053,
+      "step": 41
+    },
+    {
+      "epoch": 0.07984790874524715,
+      "grad_norm": 0.9128831658150581,
+      "learning_rate": 7.924528301886793e-06,
+      "loss": 0.477,
+      "step": 42
+    },
+    {
+      "epoch": 0.0817490494296578,
+      "grad_norm": 0.9859516293969818,
+      "learning_rate": 8.113207547169812e-06,
+      "loss": 0.4723,
+      "step": 43
+    },
+    {
+      "epoch": 0.08365019011406843,
+      "grad_norm": 0.9338198215791357,
+      "learning_rate": 8.301886792452832e-06,
+      "loss": 0.4753,
+      "step": 44
+    },
+    {
+      "epoch": 0.08555133079847908,
+      "grad_norm": 0.8642903487309621,
+      "learning_rate": 8.49056603773585e-06,
+      "loss": 0.4883,
+      "step": 45
+    },
+    {
+      "epoch": 0.08745247148288973,
+      "grad_norm": 0.8950472855881554,
+      "learning_rate": 8.67924528301887e-06,
+      "loss": 0.4837,
+      "step": 46
+    },
+    {
+      "epoch": 0.08935361216730038,
+      "grad_norm": 0.9263025521354988,
+      "learning_rate": 8.867924528301887e-06,
+      "loss": 0.4885,
+      "step": 47
+    },
+    {
+      "epoch": 0.09125475285171103,
+      "grad_norm": 0.8709492927995927,
+      "learning_rate": 9.056603773584907e-06,
+      "loss": 0.4729,
+      "step": 48
+    },
+    {
+      "epoch": 0.09315589353612168,
+      "grad_norm": 0.9495561491207107,
+      "learning_rate": 9.245283018867926e-06,
+      "loss": 0.4809,
+      "step": 49
+    },
+    {
+      "epoch": 0.09505703422053231,
+      "grad_norm": 0.9212430557085315,
+      "learning_rate": 9.433962264150944e-06,
+      "loss": 0.4949,
+      "step": 50
+    },
+    {
+      "epoch": 0.09695817490494296,
+      "grad_norm": 0.8697264217509816,
+      "learning_rate": 9.622641509433963e-06,
+      "loss": 0.4837,
+      "step": 51
+    },
+    {
+      "epoch": 0.09885931558935361,
+      "grad_norm": 0.9473649077013891,
+      "learning_rate": 9.811320754716981e-06,
+      "loss": 0.4894,
+      "step": 52
+    },
+    {
+      "epoch": 0.10076045627376426,
+      "grad_norm": 0.8727902750529724,
+      "learning_rate": 1e-05,
+      "loss": 0.4717,
+      "step": 53
+    },
+    {
+      "epoch": 0.10266159695817491,
+      "grad_norm": 0.9313428764692041,
+      "learning_rate": 1.018867924528302e-05,
+      "loss": 0.4919,
+      "step": 54
+    },
+    {
+      "epoch": 0.10456273764258556,
+      "grad_norm": 0.9704067365451778,
+      "learning_rate": 1.0377358490566038e-05,
+      "loss": 0.4926,
+      "step": 55
+    },
+    {
+      "epoch": 0.10646387832699619,
+      "grad_norm": 0.8061053116487912,
+      "learning_rate": 1.0566037735849058e-05,
+      "loss": 0.4657,
+      "step": 56
+    },
+    {
+      "epoch": 0.10836501901140684,
+      "grad_norm": 0.9380225481319396,
+      "learning_rate": 1.0754716981132076e-05,
+      "loss": 0.4554,
+      "step": 57
+    },
+    {
+      "epoch": 0.11026615969581749,
+      "grad_norm": 0.9962433970680612,
+      "learning_rate": 1.0943396226415095e-05,
+      "loss": 0.4915,
+      "step": 58
+    },
+    {
+      "epoch": 0.11216730038022814,
+      "grad_norm": 0.8442436129591926,
+      "learning_rate": 1.1132075471698115e-05,
+      "loss": 0.4768,
+      "step": 59
+    },
+    {
+      "epoch": 0.11406844106463879,
+      "grad_norm": 0.9891151685844161,
+      "learning_rate": 1.1320754716981132e-05,
+      "loss": 0.5032,
+      "step": 60
+    },
+    {
+      "epoch": 0.11596958174904944,
+      "grad_norm": 0.8972626315731577,
+      "learning_rate": 1.1509433962264152e-05,
+      "loss": 0.4471,
+      "step": 61
+    },
+    {
+      "epoch": 0.11787072243346007,
+      "grad_norm": 0.9044173073797717,
+      "learning_rate": 1.169811320754717e-05,
+      "loss": 0.4917,
+      "step": 62
+    },
+    {
+      "epoch": 0.11977186311787072,
+      "grad_norm": 0.8142713373282179,
+      "learning_rate": 1.188679245283019e-05,
+      "loss": 0.4603,
+      "step": 63
+    },
+    {
+      "epoch": 0.12167300380228137,
+      "grad_norm": 0.9740779273383415,
+      "learning_rate": 1.2075471698113209e-05,
+      "loss": 0.5005,
+      "step": 64
+    },
+    {
+      "epoch": 0.12357414448669202,
+      "grad_norm": 0.9101639423652284,
+      "learning_rate": 1.2264150943396227e-05,
+      "loss": 0.4765,
+      "step": 65
+    },
+    {
+      "epoch": 0.12547528517110265,
+      "grad_norm": 1.059970826214118,
+      "learning_rate": 1.2452830188679246e-05,
+      "loss": 0.4647,
+      "step": 66
+    },
+    {
+      "epoch": 0.12737642585551331,
+      "grad_norm": 0.912478903454531,
+      "learning_rate": 1.2641509433962264e-05,
+      "loss": 0.4688,
+      "step": 67
+    },
+    {
+      "epoch": 0.12927756653992395,
+      "grad_norm": 1.134588272279839,
+      "learning_rate": 1.2830188679245283e-05,
+      "loss": 0.4745,
+      "step": 68
+    },
+    {
+      "epoch": 0.1311787072243346,
+      "grad_norm": 1.0220782107537978,
+      "learning_rate": 1.3018867924528303e-05,
+      "loss": 0.5045,
+      "step": 69
+    },
+    {
+      "epoch": 0.13307984790874525,
+      "grad_norm": 0.8931530729975117,
+      "learning_rate": 1.320754716981132e-05,
+      "loss": 0.4751,
+      "step": 70
+    },
+    {
+      "epoch": 0.13498098859315588,
+      "grad_norm": 0.9861586298421906,
+      "learning_rate": 1.339622641509434e-05,
+      "loss": 0.466,
+      "step": 71
+    },
+    {
+      "epoch": 0.13688212927756654,
+      "grad_norm": 0.9363062644224878,
+      "learning_rate": 1.3584905660377358e-05,
+      "loss": 0.4854,
+      "step": 72
+    },
+    {
+      "epoch": 0.13878326996197718,
+      "grad_norm": 0.937339381555883,
+      "learning_rate": 1.3773584905660378e-05,
+      "loss": 0.4681,
+      "step": 73
+    },
+    {
+      "epoch": 0.14068441064638784,
+      "grad_norm": 0.8591870712230874,
+      "learning_rate": 1.3962264150943397e-05,
+      "loss": 0.458,
+      "step": 74
+    },
+    {
+      "epoch": 0.14258555133079848,
+      "grad_norm": 0.858981228894339,
+      "learning_rate": 1.4150943396226415e-05,
+      "loss": 0.4603,
+      "step": 75
+    },
+    {
+      "epoch": 0.1444866920152091,
+      "grad_norm": 0.9186212567787977,
+      "learning_rate": 1.4339622641509435e-05,
+      "loss": 0.4769,
+      "step": 76
+    },
+    {
+      "epoch": 0.14638783269961977,
+      "grad_norm": 0.9022274880100807,
+      "learning_rate": 1.4528301886792452e-05,
+      "loss": 0.4877,
+      "step": 77
+    },
+    {
+      "epoch": 0.1482889733840304,
+      "grad_norm": 0.930988554652446,
+      "learning_rate": 1.4716981132075472e-05,
+      "loss": 0.4563,
+      "step": 78
+    },
+    {
+      "epoch": 0.15019011406844107,
+      "grad_norm": 0.9259357506630179,
+      "learning_rate": 1.4905660377358491e-05,
+      "loss": 0.473,
+      "step": 79
+    },
+    {
+      "epoch": 0.1520912547528517,
+      "grad_norm": 0.8874955393900509,
+      "learning_rate": 1.5094339622641511e-05,
+      "loss": 0.4845,
+      "step": 80
+    },
+    {
+      "epoch": 0.15399239543726237,
+      "grad_norm": 0.8318666493017952,
+      "learning_rate": 1.5283018867924532e-05,
+      "loss": 0.4872,
+      "step": 81
+    },
+    {
+      "epoch": 0.155893536121673,
+      "grad_norm": 0.9566814808261703,
+      "learning_rate": 1.547169811320755e-05,
+      "loss": 0.4595,
+      "step": 82
+    },
+    {
+      "epoch": 0.15779467680608364,
+      "grad_norm": 0.8510629259210857,
+      "learning_rate": 1.5660377358490568e-05,
+      "loss": 0.4739,
+      "step": 83
+    },
+    {
+      "epoch": 0.1596958174904943,
+      "grad_norm": 0.9116392611478177,
+      "learning_rate": 1.5849056603773586e-05,
+      "loss": 0.4691,
+      "step": 84
+    },
+    {
+      "epoch": 0.16159695817490494,
+      "grad_norm": 0.8877199897313627,
+      "learning_rate": 1.6037735849056607e-05,
+      "loss": 0.4639,
+      "step": 85
+    },
+    {
+      "epoch": 0.1634980988593156,
+      "grad_norm": 0.9772044064892068,
+      "learning_rate": 1.6226415094339625e-05,
+      "loss": 0.448,
+      "step": 86
+    },
+    {
+      "epoch": 0.16539923954372623,
+      "grad_norm": 0.9076758707884138,
+      "learning_rate": 1.6415094339622643e-05,
+      "loss": 0.4794,
+      "step": 87
+    },
+    {
+      "epoch": 0.16730038022813687,
+      "grad_norm": 0.9275505025705646,
+      "learning_rate": 1.6603773584905664e-05,
+      "loss": 0.4889,
+      "step": 88
+    },
+    {
+      "epoch": 0.16920152091254753,
+      "grad_norm": 0.9073442193453491,
+      "learning_rate": 1.679245283018868e-05,
+      "loss": 0.4835,
+      "step": 89
+    },
+    {
+      "epoch": 0.17110266159695817,
+      "grad_norm": 0.8835871761554153,
+      "learning_rate": 1.69811320754717e-05,
+      "loss": 0.477,
+      "step": 90
+    },
+    {
+      "epoch": 0.17300380228136883,
+      "grad_norm": 0.9343484409659698,
+      "learning_rate": 1.716981132075472e-05,
+      "loss": 0.4828,
+      "step": 91
+    },
+    {
+      "epoch": 0.17490494296577946,
+      "grad_norm": 0.9323492506921999,
+      "learning_rate": 1.735849056603774e-05,
+      "loss": 0.48,
+      "step": 92
+    },
+    {
+      "epoch": 0.17680608365019013,
+      "grad_norm": 0.823159847578494,
+      "learning_rate": 1.7547169811320756e-05,
+      "loss": 0.4825,
+      "step": 93
+    },
+    {
+      "epoch": 0.17870722433460076,
+      "grad_norm": 1.0011551534304184,
+      "learning_rate": 1.7735849056603774e-05,
+      "loss": 0.481,
+      "step": 94
+    },
+    {
+      "epoch": 0.1806083650190114,
+      "grad_norm": 0.9800957243928676,
+      "learning_rate": 1.7924528301886795e-05,
+      "loss": 0.4889,
+      "step": 95
+    },
+    {
+      "epoch": 0.18250950570342206,
+      "grad_norm": 0.9067461263250202,
+      "learning_rate": 1.8113207547169813e-05,
+      "loss": 0.4745,
+      "step": 96
+    },
+    {
+      "epoch": 0.1844106463878327,
+      "grad_norm": 0.8524675459989356,
+      "learning_rate": 1.830188679245283e-05,
+      "loss": 0.4771,
+      "step": 97
+    },
+    {
+      "epoch": 0.18631178707224336,
+      "grad_norm": 0.8483896132239259,
+      "learning_rate": 1.8490566037735852e-05,
+      "loss": 0.435,
+      "step": 98
+    },
+    {
+      "epoch": 0.188212927756654,
+      "grad_norm": 0.8627467149752069,
+      "learning_rate": 1.867924528301887e-05,
+      "loss": 0.4605,
+      "step": 99
+    },
+    {
+      "epoch": 0.19011406844106463,
+      "grad_norm": 0.8757480457482022,
+      "learning_rate": 1.8867924528301888e-05,
+      "loss": 0.4707,
+      "step": 100
+    },
+    {
+      "epoch": 0.1920152091254753,
+      "grad_norm": 0.8777210441434259,
+      "learning_rate": 1.905660377358491e-05,
+      "loss": 0.4921,
+      "step": 101
+    },
+    {
+      "epoch": 0.19391634980988592,
+      "grad_norm": 0.833085371904221,
+      "learning_rate": 1.9245283018867927e-05,
+      "loss": 0.4562,
+      "step": 102
+    },
+    {
+      "epoch": 0.1958174904942966,
+      "grad_norm": 0.8733936553420234,
+      "learning_rate": 1.9433962264150945e-05,
+      "loss": 0.4826,
+      "step": 103
+    },
+    {
+      "epoch": 0.19771863117870722,
+      "grad_norm": 0.8673465708505467,
+      "learning_rate": 1.9622641509433963e-05,
+      "loss": 0.4561,
+      "step": 104
+    },
+    {
+      "epoch": 0.19961977186311788,
+      "grad_norm": 0.8945220805737648,
+      "learning_rate": 1.9811320754716984e-05,
+      "loss": 0.4733,
+      "step": 105
+    },
+    {
+      "epoch": 0.20152091254752852,
+      "grad_norm": 0.8783327528452897,
+      "learning_rate": 2e-05,
+      "loss": 0.4687,
+      "step": 106
+    },
+    {
+      "epoch": 0.20342205323193915,
+      "grad_norm": 1.0258227764751655,
+      "learning_rate": 1.9999944857420527e-05,
+      "loss": 0.4462,
+      "step": 107
+    },
+    {
+      "epoch": 0.20532319391634982,
+      "grad_norm": 0.8205906087475225,
+      "learning_rate": 1.9999779430290247e-05,
+      "loss": 0.4441,
+      "step": 108
+    },
+    {
+      "epoch": 0.20722433460076045,
+      "grad_norm": 1.0972288838300526,
+      "learning_rate": 1.9999503720433575e-05,
+      "loss": 0.4702,
+      "step": 109
+    },
+    {
+      "epoch": 0.20912547528517111,
+      "grad_norm": 0.8510661639759426,
+      "learning_rate": 1.999911773089118e-05,
+      "loss": 0.4465,
+      "step": 110
+    },
+    {
+      "epoch": 0.21102661596958175,
+      "grad_norm": 1.0262015955138777,
+      "learning_rate": 1.999862146591996e-05,
+      "loss": 0.4472,
+      "step": 111
+    },
+    {
+      "epoch": 0.21292775665399238,
+      "grad_norm": 0.89809224510171,
+      "learning_rate": 1.9998014930992976e-05,
+      "loss": 0.4485,
+      "step": 112
+    },
+    {
+      "epoch": 0.21482889733840305,
+      "grad_norm": 0.9667696534742204,
+      "learning_rate": 1.9997298132799408e-05,
+      "loss": 0.4742,
+      "step": 113
+    },
+    {
+      "epoch": 0.21673003802281368,
+      "grad_norm": 0.932486995734174,
+      "learning_rate": 1.9996471079244477e-05,
+      "loss": 0.4845,
+      "step": 114
+    },
+    {
+      "epoch": 0.21863117870722434,
+      "grad_norm": 0.9561888147147497,
+      "learning_rate": 1.999553377944936e-05,
+      "loss": 0.4613,
+      "step": 115
+    },
+    {
+      "epoch": 0.22053231939163498,
+      "grad_norm": 0.8444181377731163,
+      "learning_rate": 1.9994486243751076e-05,
+      "loss": 0.4789,
+      "step": 116
+    },
+    {
+      "epoch": 0.2224334600760456,
+      "grad_norm": 1.013400111358396,
+      "learning_rate": 1.9993328483702393e-05,
+      "loss": 0.4599,
+      "step": 117
+    },
+    {
+      "epoch": 0.22433460076045628,
+      "grad_norm": 0.8962337870344801,
+      "learning_rate": 1.999206051207169e-05,
+      "loss": 0.4577,
+      "step": 118
+    },
+    {
+      "epoch": 0.2262357414448669,
+      "grad_norm": 0.9552812909767043,
+      "learning_rate": 1.9990682342842805e-05,
+      "loss": 0.4675,
+      "step": 119
+    },
+    {
+      "epoch": 0.22813688212927757,
+      "grad_norm": 0.8292883832051765,
+      "learning_rate": 1.99891939912149e-05,
+      "loss": 0.443,
+      "step": 120
+    },
+    {
+      "epoch": 0.2300380228136882,
+      "grad_norm": 0.916665304798377,
+      "learning_rate": 1.9987595473602292e-05,
+      "loss": 0.482,
+      "step": 121
+    },
+    {
+      "epoch": 0.23193916349809887,
+      "grad_norm": 0.8654854491066889,
+      "learning_rate": 1.9985886807634246e-05,
+      "loss": 0.4944,
+      "step": 122
+    },
+    {
+      "epoch": 0.2338403041825095,
+      "grad_norm": 0.9419054895779853,
+      "learning_rate": 1.9984068012154824e-05,
+      "loss": 0.4598,
+      "step": 123
+    },
+    {
+      "epoch": 0.23574144486692014,
+      "grad_norm": 0.850714932775501,
+      "learning_rate": 1.9982139107222634e-05,
+      "loss": 0.4571,
+      "step": 124
+    },
+    {
+      "epoch": 0.2376425855513308,
+      "grad_norm": 1.1712515079525407,
+      "learning_rate": 1.9980100114110637e-05,
+      "loss": 0.4528,
+      "step": 125
+    },
+    {
+      "epoch": 0.23954372623574144,
+      "grad_norm": 0.890698353051668,
+      "learning_rate": 1.99779510553059e-05,
+      "loss": 0.4526,
+      "step": 126
+    },
+    {
+      "epoch": 0.2414448669201521,
+      "grad_norm": 0.8203643637993776,
+      "learning_rate": 1.9975691954509347e-05,
+      "loss": 0.4852,
+      "step": 127
+    },
+    {
+      "epoch": 0.24334600760456274,
+      "grad_norm": 0.8586124391337356,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 0.4638,
+      "step": 128
+    },
+    {
+      "epoch": 0.24524714828897337,
+      "grad_norm": 0.8687339538949509,
+      "learning_rate": 1.997084372781226e-05,
+      "loss": 0.4951,
+      "step": 129
+    },
+    {
+      "epoch": 0.24714828897338403,
+      "grad_norm": 0.8668279034750476,
+      "learning_rate": 1.9968254655380465e-05,
+      "loss": 0.4705,
+      "step": 130
+    },
+    {
+      "epoch": 0.24904942965779467,
+      "grad_norm": 0.8826342460039781,
+      "learning_rate": 1.996555564789376e-05,
+      "loss": 0.4655,
+      "step": 131
+    },
+    {
+      "epoch": 0.2509505703422053,
+      "grad_norm": 0.9181556661159211,
+      "learning_rate": 1.996274673511819e-05,
+      "loss": 0.4456,
+      "step": 132
+    },
+    {
+      "epoch": 0.25285171102661597,
+      "grad_norm": 0.9006089045342128,
+      "learning_rate": 1.99598279480319e-05,
+      "loss": 0.4578,
+      "step": 133
+    },
+    {
+      "epoch": 0.25475285171102663,
+      "grad_norm": 0.822084907065894,
+      "learning_rate": 1.9956799318824776e-05,
+      "loss": 0.4516,
+      "step": 134
+    },
+    {
+      "epoch": 0.25665399239543724,
+      "grad_norm": 0.9532529261996935,
+      "learning_rate": 1.99536608808981e-05,
+      "loss": 0.4679,
+      "step": 135
+    },
+    {
+      "epoch": 0.2585551330798479,
+      "grad_norm": 1.027938333625226,
+      "learning_rate": 1.995041266886419e-05,
+      "loss": 0.4692,
+      "step": 136
+    },
+    {
+      "epoch": 0.26045627376425856,
+      "grad_norm": 0.8561410518984488,
+      "learning_rate": 1.9947054718545996e-05,
+      "loss": 0.4757,
+      "step": 137
+    },
+    {
+      "epoch": 0.2623574144486692,
+      "grad_norm": 0.7768657212182566,
+      "learning_rate": 1.994358706697674e-05,
+      "loss": 0.4578,
+      "step": 138
+    },
+    {
+      "epoch": 0.26425855513307983,
+      "grad_norm": 0.8248455920856633,
+      "learning_rate": 1.9940009752399462e-05,
+      "loss": 0.4721,
+      "step": 139
+    },
+    {
+      "epoch": 0.2661596958174905,
+      "grad_norm": 0.8484957479670967,
+      "learning_rate": 1.9936322814266634e-05,
+      "loss": 0.4724,
+      "step": 140
+    },
+    {
+      "epoch": 0.26806083650190116,
+      "grad_norm": 0.7791590140923764,
+      "learning_rate": 1.9932526293239713e-05,
+      "loss": 0.4654,
+      "step": 141
+    },
+    {
+      "epoch": 0.26996197718631176,
+      "grad_norm": 0.8509870152648292,
+      "learning_rate": 1.9928620231188694e-05,
+      "loss": 0.4388,
+      "step": 142
+    },
+    {
+      "epoch": 0.2718631178707224,
+      "grad_norm": 0.8826600108119385,
+      "learning_rate": 1.992460467119164e-05,
+      "loss": 0.4717,
+      "step": 143
+    },
+    {
+      "epoch": 0.2737642585551331,
+      "grad_norm": 0.8064189732997133,
+      "learning_rate": 1.992047965753422e-05,
+      "loss": 0.493,
+      "step": 144
+    },
+    {
+      "epoch": 0.27566539923954375,
+      "grad_norm": 0.8810288028194865,
+      "learning_rate": 1.991624523570922e-05,
+      "loss": 0.4797,
+      "step": 145
+    },
+    {
+      "epoch": 0.27756653992395436,
+      "grad_norm": 0.8159312384199753,
+      "learning_rate": 1.9911901452416012e-05,
+      "loss": 0.447,
+      "step": 146
+    },
+    {
+      "epoch": 0.279467680608365,
+      "grad_norm": 0.808570167531465,
+      "learning_rate": 1.9907448355560094e-05,
+      "loss": 0.4545,
+      "step": 147
+    },
+    {
+      "epoch": 0.2813688212927757,
+      "grad_norm": 0.8306291251414123,
+      "learning_rate": 1.9902885994252506e-05,
+      "loss": 0.4506,
+      "step": 148
+    },
+    {
+      "epoch": 0.2832699619771863,
+      "grad_norm": 0.8087386410173785,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4644,
+      "step": 149
+    },
+    {
+      "epoch": 0.28517110266159695,
+      "grad_norm": 0.7769981049076707,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 0.46,
+      "step": 150
+    },
+    {
+      "epoch": 0.2870722433460076,
+      "grad_norm": 0.7954940978098926,
+      "learning_rate": 1.9888543832802277e-05,
+      "loss": 0.443,
+      "step": 151
+    },
+    {
+      "epoch": 0.2889733840304182,
+      "grad_norm": 0.7756673238006605,
+      "learning_rate": 1.9883544928890612e-05,
+      "loss": 0.4582,
+      "step": 152
+    },
+    {
+      "epoch": 0.2908745247148289,
+      "grad_norm": 0.8645285972546879,
+      "learning_rate": 1.9878437024146603e-05,
+      "loss": 0.4817,
+      "step": 153
+    },
+    {
+      "epoch": 0.29277566539923955,
+      "grad_norm": 0.8748541887408302,
+      "learning_rate": 1.9873220174902857e-05,
+      "loss": 0.4565,
+      "step": 154
+    },
+    {
+      "epoch": 0.2946768060836502,
+      "grad_norm": 0.8074671173089965,
+      "learning_rate": 1.986789443869348e-05,
+      "loss": 0.4643,
+      "step": 155
+    },
+    {
+      "epoch": 0.2965779467680608,
+      "grad_norm": 0.8499264086322205,
+      "learning_rate": 1.9862459874253438e-05,
+      "loss": 0.4391,
+      "step": 156
+    },
+    {
+      "epoch": 0.2984790874524715,
+      "grad_norm": 0.8234452054430735,
+      "learning_rate": 1.985691654151791e-05,
+      "loss": 0.4443,
+      "step": 157
+    },
+    {
+      "epoch": 0.30038022813688214,
+      "grad_norm": 0.78948106804705,
+      "learning_rate": 1.9851264501621635e-05,
+      "loss": 0.4669,
+      "step": 158
+    },
+    {
+      "epoch": 0.30228136882129275,
+      "grad_norm": 0.8989430162569413,
+      "learning_rate": 1.984550381689822e-05,
+      "loss": 0.4617,
+      "step": 159
+    },
+    {
+      "epoch": 0.3041825095057034,
+      "grad_norm": 0.7708840089085938,
+      "learning_rate": 1.983963455087946e-05,
+      "loss": 0.4681,
+      "step": 160
+    },
+    {
+      "epoch": 0.3060836501901141,
+      "grad_norm": 0.9075413956094507,
+      "learning_rate": 1.983365676829466e-05,
+      "loss": 0.4607,
+      "step": 161
+    },
+    {
+      "epoch": 0.30798479087452474,
+      "grad_norm": 0.8186387533529618,
+      "learning_rate": 1.982757053506989e-05,
+      "loss": 0.4526,
+      "step": 162
+    },
+    {
+      "epoch": 0.30988593155893535,
+      "grad_norm": 0.760923420825655,
+      "learning_rate": 1.9821375918327268e-05,
+      "loss": 0.4614,
+      "step": 163
+    },
+    {
+      "epoch": 0.311787072243346,
+      "grad_norm": 0.8549987427322232,
+      "learning_rate": 1.981507298638422e-05,
+      "loss": 0.4398,
+      "step": 164
+    },
+    {
+      "epoch": 0.31368821292775667,
+      "grad_norm": 0.8365330410831996,
+      "learning_rate": 1.9808661808752735e-05,
+      "loss": 0.4627,
+      "step": 165
+    },
+    {
+      "epoch": 0.3155893536121673,
+      "grad_norm": 0.9396947999489904,
+      "learning_rate": 1.980214245613858e-05,
+      "loss": 0.4441,
+      "step": 166
+    },
+    {
+      "epoch": 0.31749049429657794,
+      "grad_norm": 0.8628933390920055,
+      "learning_rate": 1.979551500044055e-05,
+      "loss": 0.4678,
+      "step": 167
+    },
+    {
+      "epoch": 0.3193916349809886,
+      "grad_norm": 0.9295443838829754,
+      "learning_rate": 1.9788779514749635e-05,
+      "loss": 0.4852,
+      "step": 168
+    },
+    {
+      "epoch": 0.32129277566539927,
+      "grad_norm": 1.0001657796279226,
+      "learning_rate": 1.978193607334826e-05,
+      "loss": 0.4454,
+      "step": 169
+    },
+    {
+      "epoch": 0.3231939163498099,
+      "grad_norm": 0.8440603443494984,
+      "learning_rate": 1.977498475170941e-05,
+      "loss": 0.4688,
+      "step": 170
+    },
+    {
+      "epoch": 0.32509505703422054,
+      "grad_norm": 0.8425207483539459,
+      "learning_rate": 1.9767925626495857e-05,
+      "loss": 0.4522,
+      "step": 171
+    },
+    {
+      "epoch": 0.3269961977186312,
+      "grad_norm": 1.1156428715940345,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 0.4692,
+      "step": 172
+    },
+    {
+      "epoch": 0.3288973384030418,
+      "grad_norm": 0.9061054250308335,
+      "learning_rate": 1.975348427793939e-05,
+      "loss": 0.4682,
+      "step": 173
+    },
+    {
+      "epoch": 0.33079847908745247,
+      "grad_norm": 1.0059198122851127,
+      "learning_rate": 1.9746102213863113e-05,
+      "loss": 0.447,
+      "step": 174
+    },
+    {
+      "epoch": 0.33269961977186313,
+      "grad_norm": 1.0001415770301876,
+      "learning_rate": 1.973861266474366e-05,
+      "loss": 0.4776,
+      "step": 175
+    },
+    {
+      "epoch": 0.33460076045627374,
+      "grad_norm": 1.1253314031960093,
+      "learning_rate": 1.9731015713179643e-05,
+      "loss": 0.4671,
+      "step": 176
+    },
+    {
+      "epoch": 0.3365019011406844,
+      "grad_norm": 0.8734608186816246,
+      "learning_rate": 1.9723311442954163e-05,
+      "loss": 0.4514,
+      "step": 177
+    },
+    {
+      "epoch": 0.33840304182509506,
+      "grad_norm": 0.8462412034425375,
+      "learning_rate": 1.9715499939033883e-05,
+      "loss": 0.4357,
+      "step": 178
+    },
+    {
+      "epoch": 0.3403041825095057,
+      "grad_norm": 1.0092983397685686,
+      "learning_rate": 1.9707581287568094e-05,
+      "loss": 0.4403,
+      "step": 179
+    },
+    {
+      "epoch": 0.34220532319391633,
+      "grad_norm": 0.8031555525141562,
+      "learning_rate": 1.969955557588778e-05,
+      "loss": 0.4436,
+      "step": 180
+    },
+    {
+      "epoch": 0.344106463878327,
+      "grad_norm": 1.0703381925344335,
+      "learning_rate": 1.9691422892504626e-05,
+      "loss": 0.4564,
+      "step": 181
+    },
+    {
+      "epoch": 0.34600760456273766,
+      "grad_norm": 0.836414112631248,
+      "learning_rate": 1.968318332711006e-05,
+      "loss": 0.4791,
+      "step": 182
+    },
+    {
+      "epoch": 0.34790874524714827,
+      "grad_norm": 1.0729233471390593,
+      "learning_rate": 1.9674836970574253e-05,
+      "loss": 0.461,
+      "step": 183
+    },
+    {
+      "epoch": 0.34980988593155893,
+      "grad_norm": 0.8781330933950663,
+      "learning_rate": 1.966638391494514e-05,
+      "loss": 0.4405,
+      "step": 184
+    },
+    {
+      "epoch": 0.3517110266159696,
+      "grad_norm": 0.8642400437178476,
+      "learning_rate": 1.9657824253447378e-05,
+      "loss": 0.4452,
+      "step": 185
+    },
+    {
+      "epoch": 0.35361216730038025,
+      "grad_norm": 0.8374822558787207,
+      "learning_rate": 1.9649158080481327e-05,
+      "loss": 0.4681,
+      "step": 186
+    },
+    {
+      "epoch": 0.35551330798479086,
+      "grad_norm": 0.868779246808885,
+      "learning_rate": 1.964038549162201e-05,
+      "loss": 0.4387,
+      "step": 187
+    },
+    {
+      "epoch": 0.3574144486692015,
+      "grad_norm": 0.8822515188210716,
+      "learning_rate": 1.963150658361807e-05,
+      "loss": 0.4507,
+      "step": 188
+    },
+    {
+      "epoch": 0.3593155893536122,
+      "grad_norm": 0.8229664193183039,
+      "learning_rate": 1.962252145439068e-05,
+      "loss": 0.4241,
+      "step": 189
+    },
+    {
+      "epoch": 0.3612167300380228,
+      "grad_norm": 0.8845480397390805,
+      "learning_rate": 1.9613430203032486e-05,
+      "loss": 0.4425,
+      "step": 190
+    },
+    {
+      "epoch": 0.36311787072243346,
+      "grad_norm": 0.8497929554493379,
+      "learning_rate": 1.9604232929806493e-05,
+      "loss": 0.4619,
+      "step": 191
+    },
+    {
+      "epoch": 0.3650190114068441,
+      "grad_norm": 0.8910077069561141,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.4507,
+      "step": 192
+    },
+    {
+      "epoch": 0.3669201520912547,
+      "grad_norm": 0.8225617516361949,
+      "learning_rate": 1.9585520724648354e-05,
+      "loss": 0.4558,
+      "step": 193
+    },
+    {
+      "epoch": 0.3688212927756654,
+      "grad_norm": 0.7376446537690261,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 0.4348,
+      "step": 194
+    },
+    {
+      "epoch": 0.37072243346007605,
+      "grad_norm": 1.034854610415476,
+      "learning_rate": 1.95663856643854e-05,
+      "loss": 0.4539,
+      "step": 195
+    },
+    {
+      "epoch": 0.3726235741444867,
+      "grad_norm": 0.8202104781025685,
+      "learning_rate": 1.955665982665038e-05,
+      "loss": 0.4397,
+      "step": 196
+    },
+    {
+      "epoch": 0.3745247148288973,
+      "grad_norm": 0.7516360121525889,
+      "learning_rate": 1.9546828593140565e-05,
+      "loss": 0.4595,
+      "step": 197
+    },
+    {
+      "epoch": 0.376425855513308,
+      "grad_norm": 0.8176066544592406,
+      "learning_rate": 1.9536892072279863e-05,
+      "loss": 0.458,
+      "step": 198
+    },
+    {
+      "epoch": 0.37832699619771865,
+      "grad_norm": 0.7730338071670566,
+      "learning_rate": 1.9526850373653356e-05,
+      "loss": 0.4701,
+      "step": 199
+    },
+    {
+      "epoch": 0.38022813688212925,
+      "grad_norm": 0.7491229405592764,
+      "learning_rate": 1.9516703608006074e-05,
+      "loss": 0.411,
+      "step": 200
+    },
+    {
+      "epoch": 0.3821292775665399,
+      "grad_norm": 0.7679836570728917,
+      "learning_rate": 1.9506451887241787e-05,
+      "loss": 0.4516,
+      "step": 201
+    },
+    {
+      "epoch": 0.3840304182509506,
+      "grad_norm": 0.8669970707876079,
+      "learning_rate": 1.949609532442176e-05,
+      "loss": 0.4544,
+      "step": 202
+    },
+    {
+      "epoch": 0.38593155893536124,
+      "grad_norm": 0.7952503365423778,
+      "learning_rate": 1.9485634033763507e-05,
+      "loss": 0.4491,
+      "step": 203
+    },
+    {
+      "epoch": 0.38783269961977185,
+      "grad_norm": 0.948045152202822,
+      "learning_rate": 1.9475068130639543e-05,
+      "loss": 0.4329,
+      "step": 204
+    },
+    {
+      "epoch": 0.3897338403041825,
+      "grad_norm": 0.8316252682821497,
+      "learning_rate": 1.9464397731576093e-05,
+      "loss": 0.4524,
+      "step": 205
+    },
+    {
+      "epoch": 0.3916349809885932,
+      "grad_norm": 0.9264128825386274,
+      "learning_rate": 1.945362295425183e-05,
+      "loss": 0.4561,
+      "step": 206
+    },
+    {
+      "epoch": 0.3935361216730038,
+      "grad_norm": 0.7316168173583714,
+      "learning_rate": 1.944274391749655e-05,
+      "loss": 0.4359,
+      "step": 207
+    },
+    {
+      "epoch": 0.39543726235741444,
+      "grad_norm": 0.8596477880886351,
+      "learning_rate": 1.9431760741289886e-05,
+      "loss": 0.445,
+      "step": 208
+    },
+    {
+      "epoch": 0.3973384030418251,
+      "grad_norm": 0.7767198277535254,
+      "learning_rate": 1.942067354675997e-05,
+      "loss": 0.4577,
+      "step": 209
+    },
+    {
+      "epoch": 0.39923954372623577,
+      "grad_norm": 0.7332243452861559,
+      "learning_rate": 1.9409482456182105e-05,
+      "loss": 0.4343,
+      "step": 210
+    },
+    {
+      "epoch": 0.4011406844106464,
+      "grad_norm": 0.7763141803675838,
+      "learning_rate": 1.939818759297741e-05,
+      "loss": 0.4492,
+      "step": 211
+    },
+    {
+      "epoch": 0.40304182509505704,
+      "grad_norm": 0.8176992678699004,
+      "learning_rate": 1.9386789081711465e-05,
+      "loss": 0.4487,
+      "step": 212
+    },
+    {
+      "epoch": 0.4049429657794677,
+      "grad_norm": 0.7371890070028103,
+      "learning_rate": 1.9375287048092927e-05,
+      "loss": 0.4624,
+      "step": 213
+    },
+    {
+      "epoch": 0.4068441064638783,
+      "grad_norm": 0.745484739594109,
+      "learning_rate": 1.9363681618972166e-05,
+      "loss": 0.4421,
+      "step": 214
+    },
+    {
+      "epoch": 0.40874524714828897,
+      "grad_norm": 0.7834012305574243,
+      "learning_rate": 1.9351972922339835e-05,
+      "loss": 0.4473,
+      "step": 215
+    },
+    {
+      "epoch": 0.41064638783269963,
+      "grad_norm": 0.8098619011368847,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 0.4573,
+      "step": 216
+    },
+    {
+      "epoch": 0.41254752851711024,
+      "grad_norm": 0.7423621989392688,
+      "learning_rate": 1.9328246244196117e-05,
+      "loss": 0.4349,
+      "step": 217
+    },
+    {
+      "epoch": 0.4144486692015209,
+      "grad_norm": 0.8229993807147958,
+      "learning_rate": 1.931622852435478e-05,
+      "loss": 0.4482,
+      "step": 218
+    },
+    {
+      "epoch": 0.41634980988593157,
+      "grad_norm": 0.7789581652872024,
+      "learning_rate": 1.930410806033908e-05,
+      "loss": 0.4279,
+      "step": 219
+    },
+    {
+      "epoch": 0.41825095057034223,
+      "grad_norm": 0.8304122146086419,
+      "learning_rate": 1.929188498581975e-05,
+      "loss": 0.4413,
+      "step": 220
+    },
+    {
+      "epoch": 0.42015209125475284,
+      "grad_norm": 0.765023384412332,
+      "learning_rate": 1.9279559435599164e-05,
+      "loss": 0.4511,
+      "step": 221
+    },
+    {
+      "epoch": 0.4220532319391635,
+      "grad_norm": 0.7691633516875152,
+      "learning_rate": 1.926713154560984e-05,
+      "loss": 0.4583,
+      "step": 222
+    },
+    {
+      "epoch": 0.42395437262357416,
+      "grad_norm": 0.8937160367772905,
+      "learning_rate": 1.9254601452912972e-05,
+      "loss": 0.4677,
+      "step": 223
+    },
+    {
+      "epoch": 0.42585551330798477,
+      "grad_norm": 0.6990728450459365,
+      "learning_rate": 1.924196929569688e-05,
+      "loss": 0.436,
+      "step": 224
+    },
+    {
+      "epoch": 0.42775665399239543,
+      "grad_norm": 0.7649799938510272,
+      "learning_rate": 1.922923521327551e-05,
+      "loss": 0.472,
+      "step": 225
+    },
+    {
+      "epoch": 0.4296577946768061,
+      "grad_norm": 0.7208999957024362,
+      "learning_rate": 1.9216399346086893e-05,
+      "loss": 0.4393,
+      "step": 226
+    },
+    {
+      "epoch": 0.43155893536121676,
+      "grad_norm": 0.7579693582095106,
+      "learning_rate": 1.9203461835691596e-05,
+      "loss": 0.4564,
+      "step": 227
+    },
+    {
+      "epoch": 0.43346007604562736,
+      "grad_norm": 0.7595182677948493,
+      "learning_rate": 1.9190422824771158e-05,
+      "loss": 0.454,
+      "step": 228
+    },
+    {
+      "epoch": 0.435361216730038,
+      "grad_norm": 0.7190831370460424,
+      "learning_rate": 1.9177282457126515e-05,
+      "loss": 0.4438,
+      "step": 229
+    },
+    {
+      "epoch": 0.4372623574144487,
+      "grad_norm": 0.7569221306264814,
+      "learning_rate": 1.9164040877676425e-05,
+      "loss": 0.4384,
+      "step": 230
+    },
+    {
+      "epoch": 0.4391634980988593,
+      "grad_norm": 0.757615263304378,
+      "learning_rate": 1.9150698232455853e-05,
+      "loss": 0.4136,
+      "step": 231
+    },
+    {
+      "epoch": 0.44106463878326996,
+      "grad_norm": 0.7294839901287309,
+      "learning_rate": 1.913725466861438e-05,
+      "loss": 0.4427,
+      "step": 232
+    },
+    {
+      "epoch": 0.4429657794676806,
+      "grad_norm": 0.8029313636007074,
+      "learning_rate": 1.9123710334414552e-05,
+      "loss": 0.4518,
+      "step": 233
+    },
+    {
+      "epoch": 0.4448669201520912,
+      "grad_norm": 0.8058268457160375,
+      "learning_rate": 1.911006537923029e-05,
+      "loss": 0.4295,
+      "step": 234
+    },
+    {
+      "epoch": 0.4467680608365019,
+      "grad_norm": 0.7156394314598286,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.4355,
+      "step": 235
+    },
+    {
+      "epoch": 0.44866920152091255,
+      "grad_norm": 0.7549241517192204,
+      "learning_rate": 1.908247420895089e-05,
+      "loss": 0.4438,
+      "step": 236
+    },
+    {
+      "epoch": 0.4505703422053232,
+      "grad_norm": 0.8191961798805029,
+      "learning_rate": 1.9068528298145418e-05,
+      "loss": 0.451,
+      "step": 237
+    },
+    {
+      "epoch": 0.4524714828897338,
+      "grad_norm": 0.7244507562689031,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 0.448,
+      "step": 238
+    },
+    {
+      "epoch": 0.4543726235741445,
+      "grad_norm": 0.8301746297059636,
+      "learning_rate": 1.9040336594214727e-05,
+      "loss": 0.4345,
+      "step": 239
+    },
+    {
+      "epoch": 0.45627376425855515,
+      "grad_norm": 0.6952867731769793,
+      "learning_rate": 1.9026091112002163e-05,
+      "loss": 0.4378,
+      "step": 240
+    },
+    {
+      "epoch": 0.45817490494296575,
+      "grad_norm": 0.8093615106712789,
+      "learning_rate": 1.90117460854003e-05,
+      "loss": 0.4329,
+      "step": 241
+    },
+    {
+      "epoch": 0.4600760456273764,
+      "grad_norm": 0.8030130407774219,
+      "learning_rate": 1.8997301672613496e-05,
+      "loss": 0.4834,
+      "step": 242
+    },
+    {
+      "epoch": 0.4619771863117871,
+      "grad_norm": 0.7863338343494916,
+      "learning_rate": 1.8982758032942184e-05,
+      "loss": 0.457,
+      "step": 243
+    },
+    {
+      "epoch": 0.46387832699619774,
+      "grad_norm": 0.791371018554521,
+      "learning_rate": 1.896811532678113e-05,
+      "loss": 0.4468,
+      "step": 244
+    },
+    {
+      "epoch": 0.46577946768060835,
+      "grad_norm": 0.7911431316618127,
+      "learning_rate": 1.8953373715617646e-05,
+      "loss": 0.4398,
+      "step": 245
+    },
+    {
+      "epoch": 0.467680608365019,
+      "grad_norm": 0.7045115590464698,
+      "learning_rate": 1.893853336202983e-05,
+      "loss": 0.4277,
+      "step": 246
+    },
+    {
+      "epoch": 0.4695817490494297,
+      "grad_norm": 0.8827929270463941,
+      "learning_rate": 1.892359442968475e-05,
+      "loss": 0.4429,
+      "step": 247
+    },
+    {
+      "epoch": 0.4714828897338403,
+      "grad_norm": 0.7061677517559865,
+      "learning_rate": 1.8908557083336668e-05,
+      "loss": 0.4309,
+      "step": 248
+    },
+    {
+      "epoch": 0.47338403041825095,
+      "grad_norm": 0.8467962180578004,
+      "learning_rate": 1.889342148882519e-05,
+      "loss": 0.4358,
+      "step": 249
+    },
+    {
+      "epoch": 0.4752851711026616,
+      "grad_norm": 0.7333371982658439,
+      "learning_rate": 1.8878187813073465e-05,
+      "loss": 0.4562,
+      "step": 250
+    },
+    {
+      "epoch": 0.47718631178707227,
+      "grad_norm": 0.7240022145139627,
+      "learning_rate": 1.886285622408633e-05,
+      "loss": 0.4188,
+      "step": 251
+    },
+    {
+      "epoch": 0.4790874524714829,
+      "grad_norm": 0.7534176031970651,
+      "learning_rate": 1.8847426890948447e-05,
+      "loss": 0.451,
+      "step": 252
+    },
+    {
+      "epoch": 0.48098859315589354,
+      "grad_norm": 0.7518736759983994,
+      "learning_rate": 1.8831899983822475e-05,
+      "loss": 0.4372,
+      "step": 253
+    },
+    {
+      "epoch": 0.4828897338403042,
+      "grad_norm": 0.7207304372714498,
+      "learning_rate": 1.8816275673947148e-05,
+      "loss": 0.4317,
+      "step": 254
+    },
+    {
+      "epoch": 0.4847908745247148,
+      "grad_norm": 0.7155716703912828,
+      "learning_rate": 1.8800554133635417e-05,
+      "loss": 0.4341,
+      "step": 255
+    },
+    {
+      "epoch": 0.4866920152091255,
+      "grad_norm": 0.7260707266482074,
+      "learning_rate": 1.8784735536272543e-05,
+      "loss": 0.422,
+      "step": 256
+    },
+    {
+      "epoch": 0.48859315589353614,
+      "grad_norm": 0.7385535421524079,
+      "learning_rate": 1.8768820056314173e-05,
+      "loss": 0.4272,
+      "step": 257
+    },
+    {
+      "epoch": 0.49049429657794674,
+      "grad_norm": 0.8230003216984235,
+      "learning_rate": 1.875280786928444e-05,
+      "loss": 0.4465,
+      "step": 258
+    },
+    {
+      "epoch": 0.4923954372623574,
+      "grad_norm": 0.7638799382740075,
+      "learning_rate": 1.873669915177399e-05,
+      "loss": 0.4507,
+      "step": 259
+    },
+    {
+      "epoch": 0.49429657794676807,
+      "grad_norm": 0.7398114103105022,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 0.4192,
+      "step": 260
+    },
+    {
+      "epoch": 0.49619771863117873,
+      "grad_norm": 0.7278491492655724,
+      "learning_rate": 1.8704192836994578e-05,
+      "loss": 0.4155,
+      "step": 261
+    },
+    {
+      "epoch": 0.49809885931558934,
+      "grad_norm": 0.8383089527657179,
+      "learning_rate": 1.8687795598222024e-05,
+      "loss": 0.4377,
+      "step": 262
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9022173759120395,
+      "learning_rate": 1.8671302545957628e-05,
+      "loss": 0.4411,
+      "step": 263
+    },
+    {
+      "epoch": 0.5019011406844106,
+      "grad_norm": 0.775460559768517,
+      "learning_rate": 1.8654713862095272e-05,
+      "loss": 0.4388,
+      "step": 264
+    },
+    {
+      "epoch": 0.5038022813688213,
+      "grad_norm": 0.7997135022039303,
+      "learning_rate": 1.8638029729583524e-05,
+      "loss": 0.4583,
+      "step": 265
+    },
+    {
+      "epoch": 0.5057034220532319,
+      "grad_norm": 0.8256378028356478,
+      "learning_rate": 1.8621250332423603e-05,
+      "loss": 0.4466,
+      "step": 266
+    },
+    {
+      "epoch": 0.5076045627376425,
+      "grad_norm": 0.7504908459637162,
+      "learning_rate": 1.860437585566736e-05,
+      "loss": 0.4413,
+      "step": 267
+    },
+    {
+      "epoch": 0.5095057034220533,
+      "grad_norm": 0.7949080612719884,
+      "learning_rate": 1.8587406485415226e-05,
+      "loss": 0.4385,
+      "step": 268
+    },
+    {
+      "epoch": 0.5114068441064639,
+      "grad_norm": 0.818806015218277,
+      "learning_rate": 1.8570342408814173e-05,
+      "loss": 0.4492,
+      "step": 269
+    },
+    {
+      "epoch": 0.5133079847908745,
+      "grad_norm": 0.7670987393079959,
+      "learning_rate": 1.855318381405564e-05,
+      "loss": 0.4647,
+      "step": 270
+    },
+    {
+      "epoch": 0.5152091254752852,
+      "grad_norm": 0.8405798384900733,
+      "learning_rate": 1.8535930890373467e-05,
+      "loss": 0.4431,
+      "step": 271
+    },
+    {
+      "epoch": 0.5171102661596958,
+      "grad_norm": 0.8310376851182427,
+      "learning_rate": 1.8518583828041787e-05,
+      "loss": 0.4274,
+      "step": 272
+    },
+    {
+      "epoch": 0.5190114068441065,
+      "grad_norm": 0.7742500774542205,
+      "learning_rate": 1.8501142818372964e-05,
+      "loss": 0.4307,
+      "step": 273
+    },
+    {
+      "epoch": 0.5209125475285171,
+      "grad_norm": 0.8998499399110481,
+      "learning_rate": 1.848360805371544e-05,
+      "loss": 0.4406,
+      "step": 274
+    },
+    {
+      "epoch": 0.5228136882129277,
+      "grad_norm": 0.8059160027278037,
+      "learning_rate": 1.8465979727451653e-05,
+      "loss": 0.4476,
+      "step": 275
+    },
+    {
+      "epoch": 0.5247148288973384,
+      "grad_norm": 0.9586842188224283,
+      "learning_rate": 1.8448258033995877e-05,
+      "loss": 0.4263,
+      "step": 276
+    },
+    {
+      "epoch": 0.526615969581749,
+      "grad_norm": 0.8890845063380293,
+      "learning_rate": 1.8430443168792087e-05,
+      "loss": 0.4184,
+      "step": 277
+    },
+    {
+      "epoch": 0.5285171102661597,
+      "grad_norm": 0.7688350950287203,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.4495,
+      "step": 278
+    },
+    {
+      "epoch": 0.5304182509505704,
+      "grad_norm": 0.961472135246734,
+      "learning_rate": 1.8394534710051956e-05,
+      "loss": 0.448,
+      "step": 279
+    },
+    {
+      "epoch": 0.532319391634981,
+      "grad_norm": 0.7545199706098225,
+      "learning_rate": 1.8376441512532617e-05,
+      "loss": 0.4409,
+      "step": 280
+    },
+    {
+      "epoch": 0.5342205323193916,
+      "grad_norm": 0.7678528475035395,
+      "learning_rate": 1.835825593529492e-05,
+      "loss": 0.422,
+      "step": 281
+    },
+    {
+      "epoch": 0.5361216730038023,
+      "grad_norm": 0.7095069040341553,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 0.4249,
+      "step": 282
+    },
+    {
+      "epoch": 0.5380228136882129,
+      "grad_norm": 0.7254843703014583,
+      "learning_rate": 1.8321608444920738e-05,
+      "loss": 0.4208,
+      "step": 283
+    },
+    {
+      "epoch": 0.5399239543726235,
+      "grad_norm": 0.702128916453244,
+      "learning_rate": 1.830314693595169e-05,
+      "loss": 0.4297,
+      "step": 284
+    },
+    {
+      "epoch": 0.5418250950570342,
+      "grad_norm": 0.7229121265950238,
+      "learning_rate": 1.828459385559468e-05,
+      "loss": 0.4295,
+      "step": 285
+    },
+    {
+      "epoch": 0.5437262357414449,
+      "grad_norm": 0.7191249366876319,
+      "learning_rate": 1.8265949408462657e-05,
+      "loss": 0.4307,
+      "step": 286
+    },
+    {
+      "epoch": 0.5456273764258555,
+      "grad_norm": 0.7060200556410641,
+      "learning_rate": 1.8247213800176192e-05,
+      "loss": 0.4375,
+      "step": 287
+    },
+    {
+      "epoch": 0.5475285171102662,
+      "grad_norm": 0.7252634361077612,
+      "learning_rate": 1.8228387237361245e-05,
+      "loss": 0.4321,
+      "step": 288
+    },
+    {
+      "epoch": 0.5494296577946768,
+      "grad_norm": 0.7252531696504165,
+      "learning_rate": 1.8209469927646863e-05,
+      "loss": 0.4336,
+      "step": 289
+    },
+    {
+      "epoch": 0.5513307984790875,
+      "grad_norm": 0.7131800869830593,
+      "learning_rate": 1.8190462079662897e-05,
+      "loss": 0.4545,
+      "step": 290
+    },
+    {
+      "epoch": 0.5532319391634981,
+      "grad_norm": 0.6988423569695901,
+      "learning_rate": 1.81713639030377e-05,
+      "loss": 0.4294,
+      "step": 291
+    },
+    {
+      "epoch": 0.5551330798479087,
+      "grad_norm": 0.7142797939687778,
+      "learning_rate": 1.8152175608395814e-05,
+      "loss": 0.4241,
+      "step": 292
+    },
+    {
+      "epoch": 0.5570342205323194,
+      "grad_norm": 0.765845198381463,
+      "learning_rate": 1.8132897407355657e-05,
+      "loss": 0.4435,
+      "step": 293
+    },
+    {
+      "epoch": 0.55893536121673,
+      "grad_norm": 1.005647182386002,
+      "learning_rate": 1.811352951252717e-05,
+      "loss": 0.4089,
+      "step": 294
+    },
+    {
+      "epoch": 0.5608365019011406,
+      "grad_norm": 1.050397886305871,
+      "learning_rate": 1.809407213750949e-05,
+      "loss": 0.4314,
+      "step": 295
+    },
+    {
+      "epoch": 0.5627376425855514,
+      "grad_norm": 1.0469939942046973,
+      "learning_rate": 1.807452549688859e-05,
+      "loss": 0.4279,
+      "step": 296
+    },
+    {
+      "epoch": 0.564638783269962,
+      "grad_norm": 1.0171725382364882,
+      "learning_rate": 1.8054889806234906e-05,
+      "loss": 0.4739,
+      "step": 297
+    },
+    {
+      "epoch": 0.5665399239543726,
+      "grad_norm": 1.0298849829352918,
+      "learning_rate": 1.8035165282100963e-05,
+      "loss": 0.4274,
+      "step": 298
+    },
+    {
+      "epoch": 0.5684410646387833,
+      "grad_norm": 0.936062659229467,
+      "learning_rate": 1.8015352142018984e-05,
+      "loss": 0.4273,
+      "step": 299
+    },
+    {
+      "epoch": 0.5703422053231939,
+      "grad_norm": 0.9328587495407188,
+      "learning_rate": 1.799545060449851e-05,
+      "loss": 0.4218,
+      "step": 300
+    },
+    {
+      "epoch": 0.5722433460076045,
+      "grad_norm": 0.747991749541965,
+      "learning_rate": 1.797546088902396e-05,
+      "loss": 0.4208,
+      "step": 301
+    },
+    {
+      "epoch": 0.5741444866920152,
+      "grad_norm": 1.1054072075939436,
+      "learning_rate": 1.7955383216052224e-05,
+      "loss": 0.4276,
+      "step": 302
+    },
+    {
+      "epoch": 0.5760456273764258,
+      "grad_norm": 0.7487149131472096,
+      "learning_rate": 1.7935217807010238e-05,
+      "loss": 0.4445,
+      "step": 303
+    },
+    {
+      "epoch": 0.5779467680608364,
+      "grad_norm": 0.9054358930276075,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 0.4367,
+      "step": 304
+    },
+    {
+      "epoch": 0.5798479087452472,
+      "grad_norm": 0.7980799495199992,
+      "learning_rate": 1.7894624671258813e-05,
+      "loss": 0.4476,
+      "step": 305
+    },
+    {
+      "epoch": 0.5817490494296578,
+      "grad_norm": 0.7156177614982158,
+      "learning_rate": 1.7874197392231414e-05,
+      "loss": 0.4181,
+      "step": 306
+    },
+    {
+      "epoch": 0.5836501901140685,
+      "grad_norm": 0.8539455081753311,
+      "learning_rate": 1.7853683272492913e-05,
+      "loss": 0.4446,
+      "step": 307
+    },
+    {
+      "epoch": 0.5855513307984791,
+      "grad_norm": 0.734122875086201,
+      "learning_rate": 1.7833082538283615e-05,
+      "loss": 0.4309,
+      "step": 308
+    },
+    {
+      "epoch": 0.5874524714828897,
+      "grad_norm": 0.8508683328618183,
+      "learning_rate": 1.7812395416799034e-05,
+      "loss": 0.4193,
+      "step": 309
+    },
+    {
+      "epoch": 0.5893536121673004,
+      "grad_norm": 0.7629211534488731,
+      "learning_rate": 1.7791622136187422e-05,
+      "loss": 0.4608,
+      "step": 310
+    },
+    {
+      "epoch": 0.591254752851711,
+      "grad_norm": 0.7549594454329173,
+      "learning_rate": 1.7770762925547235e-05,
+      "loss": 0.4288,
+      "step": 311
+    },
+    {
+      "epoch": 0.5931558935361216,
+      "grad_norm": 0.712900145204785,
+      "learning_rate": 1.7749818014924612e-05,
+      "loss": 0.4312,
+      "step": 312
+    },
+    {
+      "epoch": 0.5950570342205324,
+      "grad_norm": 0.6893233338048916,
+      "learning_rate": 1.7728787635310828e-05,
+      "loss": 0.4089,
+      "step": 313
+    },
+    {
+      "epoch": 0.596958174904943,
+      "grad_norm": 0.6825814872096522,
+      "learning_rate": 1.770767201863976e-05,
+      "loss": 0.4327,
+      "step": 314
+    },
+    {
+      "epoch": 0.5988593155893536,
+      "grad_norm": 0.7438134148894817,
+      "learning_rate": 1.7686471397785322e-05,
+      "loss": 0.418,
+      "step": 315
+    },
+    {
+      "epoch": 0.6007604562737643,
+      "grad_norm": 0.6691879303250395,
+      "learning_rate": 1.76651860065589e-05,
+      "loss": 0.4142,
+      "step": 316
+    },
+    {
+      "epoch": 0.6026615969581749,
+      "grad_norm": 0.7530541859359616,
+      "learning_rate": 1.764381607970677e-05,
+      "loss": 0.4289,
+      "step": 317
+    },
+    {
+      "epoch": 0.6045627376425855,
+      "grad_norm": 0.7779174466685825,
+      "learning_rate": 1.7622361852907506e-05,
+      "loss": 0.4067,
+      "step": 318
+    },
+    {
+      "epoch": 0.6064638783269962,
+      "grad_norm": 0.729919516135425,
+      "learning_rate": 1.760082356276939e-05,
+      "loss": 0.439,
+      "step": 319
+    },
+    {
+      "epoch": 0.6083650190114068,
+      "grad_norm": 0.6714255709395164,
+      "learning_rate": 1.75792014468278e-05,
+      "loss": 0.4151,
+      "step": 320
+    },
+    {
+      "epoch": 0.6102661596958175,
+      "grad_norm": 0.6808133488887521,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.4376,
+      "step": 321
+    },
+    {
+      "epoch": 0.6121673003802282,
+      "grad_norm": 0.7256708969678756,
+      "learning_rate": 1.7535706692295436e-05,
+      "loss": 0.4428,
+      "step": 322
+    },
+    {
+      "epoch": 0.6140684410646388,
+      "grad_norm": 0.6848328910295242,
+      "learning_rate": 1.7513834533387256e-05,
+      "loss": 0.4209,
+      "step": 323
+    },
+    {
+      "epoch": 0.6159695817490495,
+      "grad_norm": 0.6944098338400391,
+      "learning_rate": 1.7491879508035488e-05,
+      "loss": 0.4416,
+      "step": 324
+    },
+    {
+      "epoch": 0.6178707224334601,
+      "grad_norm": 0.6654502350636388,
+      "learning_rate": 1.746984185837149e-05,
+      "loss": 0.4215,
+      "step": 325
+    },
+    {
+      "epoch": 0.6197718631178707,
+      "grad_norm": 0.7018651310753525,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 0.4273,
+      "step": 326
+    },
+    {
+      "epoch": 0.6216730038022814,
+      "grad_norm": 0.700045533506836,
+      "learning_rate": 1.7425519659185596e-05,
+      "loss": 0.4252,
+      "step": 327
+    },
+    {
+      "epoch": 0.623574144486692,
+      "grad_norm": 0.6978565551886846,
+      "learning_rate": 1.740323559847179e-05,
+      "loss": 0.4343,
+      "step": 328
+    },
+    {
+      "epoch": 0.6254752851711026,
+      "grad_norm": 0.685470796874879,
+      "learning_rate": 1.738086989105651e-05,
+      "loss": 0.4081,
+      "step": 329
+    },
+    {
+      "epoch": 0.6273764258555133,
+      "grad_norm": 0.7913435737861949,
+      "learning_rate": 1.735842278360032e-05,
+      "loss": 0.4557,
+      "step": 330
+    },
+    {
+      "epoch": 0.629277566539924,
+      "grad_norm": 0.6588906764258432,
+      "learning_rate": 1.73358945236615e-05,
+      "loss": 0.4226,
+      "step": 331
+    },
+    {
+      "epoch": 0.6311787072243346,
+      "grad_norm": 0.661951355060734,
+      "learning_rate": 1.7313285359693322e-05,
+      "loss": 0.4376,
+      "step": 332
+    },
+    {
+      "epoch": 0.6330798479087453,
+      "grad_norm": 0.7424890959424313,
+      "learning_rate": 1.7290595541041312e-05,
+      "loss": 0.3997,
+      "step": 333
+    },
+    {
+      "epoch": 0.6349809885931559,
+      "grad_norm": 0.6896349177232862,
+      "learning_rate": 1.7267825317940494e-05,
+      "loss": 0.4449,
+      "step": 334
+    },
+    {
+      "epoch": 0.6368821292775665,
+      "grad_norm": 0.7345447695117684,
+      "learning_rate": 1.724497494151264e-05,
+      "loss": 0.4126,
+      "step": 335
+    },
+    {
+      "epoch": 0.6387832699619772,
+      "grad_norm": 0.7059812106257065,
+      "learning_rate": 1.7222044663763484e-05,
+      "loss": 0.4263,
+      "step": 336
+    },
+    {
+      "epoch": 0.6406844106463878,
+      "grad_norm": 0.7674748088738242,
+      "learning_rate": 1.7199034737579962e-05,
+      "loss": 0.4272,
+      "step": 337
+    },
+    {
+      "epoch": 0.6425855513307985,
+      "grad_norm": 0.711118520403189,
+      "learning_rate": 1.7175945416727405e-05,
+      "loss": 0.4466,
+      "step": 338
+    },
+    {
+      "epoch": 0.6444866920152091,
+      "grad_norm": 0.6994780768037421,
+      "learning_rate": 1.7152776955846768e-05,
+      "loss": 0.4145,
+      "step": 339
+    },
+    {
+      "epoch": 0.6463878326996197,
+      "grad_norm": 0.7344154680610566,
+      "learning_rate": 1.7129529610451775e-05,
+      "loss": 0.4132,
+      "step": 340
+    },
+    {
+      "epoch": 0.6482889733840305,
+      "grad_norm": 0.6980718305036266,
+      "learning_rate": 1.7106203636926154e-05,
+      "loss": 0.436,
+      "step": 341
+    },
+    {
+      "epoch": 0.6501901140684411,
+      "grad_norm": 0.9889787096786525,
+      "learning_rate": 1.7082799292520767e-05,
+      "loss": 0.4168,
+      "step": 342
+    },
+    {
+      "epoch": 0.6520912547528517,
+      "grad_norm": 0.7183484141891322,
+      "learning_rate": 1.7059316835350806e-05,
+      "loss": 0.4236,
+      "step": 343
+    },
+    {
+      "epoch": 0.6539923954372624,
+      "grad_norm": 0.9350374129787101,
+      "learning_rate": 1.7035756524392924e-05,
+      "loss": 0.4248,
+      "step": 344
+    },
+    {
+      "epoch": 0.655893536121673,
+      "grad_norm": 0.7724280603714035,
+      "learning_rate": 1.7012118619482376e-05,
+      "loss": 0.437,
+      "step": 345
+    },
+    {
+      "epoch": 0.6577946768060836,
+      "grad_norm": 0.8494440151857628,
+      "learning_rate": 1.6988403381310177e-05,
+      "loss": 0.4463,
+      "step": 346
+    },
+    {
+      "epoch": 0.6596958174904943,
+      "grad_norm": 0.762413468427766,
+      "learning_rate": 1.696461107142021e-05,
+      "loss": 0.4329,
+      "step": 347
+    },
+    {
+      "epoch": 0.6615969581749049,
+      "grad_norm": 0.6530153008143537,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 0.4225,
+      "step": 348
+    },
+    {
+      "epoch": 0.6634980988593155,
+      "grad_norm": 0.8759213333612214,
+      "learning_rate": 1.691679628690953e-05,
+      "loss": 0.4402,
+      "step": 349
+    },
+    {
+      "epoch": 0.6653992395437263,
+      "grad_norm": 0.696926269352893,
+      "learning_rate": 1.6892774339614927e-05,
+      "loss": 0.4143,
+      "step": 350
+    },
+    {
+      "epoch": 0.6673003802281369,
+      "grad_norm": 0.8456193483234181,
+      "learning_rate": 1.686867637524896e-05,
+      "loss": 0.4441,
+      "step": 351
+    },
+    {
+      "epoch": 0.6692015209125475,
+      "grad_norm": 0.7686957706257199,
+      "learning_rate": 1.6844502659576414e-05,
+      "loss": 0.4267,
+      "step": 352
+    },
+    {
+      "epoch": 0.6711026615969582,
+      "grad_norm": 0.6654202422811619,
+      "learning_rate": 1.6820253459197493e-05,
+      "loss": 0.4377,
+      "step": 353
+    },
+    {
+      "epoch": 0.6730038022813688,
+      "grad_norm": 0.9687415337995555,
+      "learning_rate": 1.679592904154489e-05,
+      "loss": 0.4381,
+      "step": 354
+    },
+    {
+      "epoch": 0.6749049429657795,
+      "grad_norm": 0.6694556328170388,
+      "learning_rate": 1.677152967488084e-05,
+      "loss": 0.4383,
+      "step": 355
+    },
+    {
+      "epoch": 0.6768060836501901,
+      "grad_norm": 0.7378088429178773,
+      "learning_rate": 1.6747055628294134e-05,
+      "loss": 0.4336,
+      "step": 356
+    },
+    {
+      "epoch": 0.6787072243346007,
+      "grad_norm": 0.7970643687049169,
+      "learning_rate": 1.6722507171697184e-05,
+      "loss": 0.4239,
+      "step": 357
+    },
+    {
+      "epoch": 0.6806083650190115,
+      "grad_norm": 0.6946104424266204,
+      "learning_rate": 1.669788457582304e-05,
+      "loss": 0.4478,
+      "step": 358
+    },
+    {
+      "epoch": 0.6825095057034221,
+      "grad_norm": 0.7851236530252377,
+      "learning_rate": 1.6673188112222394e-05,
+      "loss": 0.419,
+      "step": 359
+    },
+    {
+      "epoch": 0.6844106463878327,
+      "grad_norm": 0.8071504085130161,
+      "learning_rate": 1.6648418053260585e-05,
+      "loss": 0.4406,
+      "step": 360
+    },
+    {
+      "epoch": 0.6863117870722434,
+      "grad_norm": 0.6812745148344767,
+      "learning_rate": 1.6623574672114596e-05,
+      "loss": 0.4122,
+      "step": 361
+    },
+    {
+      "epoch": 0.688212927756654,
+      "grad_norm": 0.772846162563241,
+      "learning_rate": 1.6598658242770054e-05,
+      "loss": 0.4352,
+      "step": 362
+    },
+    {
+      "epoch": 0.6901140684410646,
+      "grad_norm": 0.7336843469011123,
+      "learning_rate": 1.6573669040018202e-05,
+      "loss": 0.4255,
+      "step": 363
+    },
+    {
+      "epoch": 0.6920152091254753,
+      "grad_norm": 0.7116217185816214,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.4149,
+      "step": 364
+    },
+    {
+      "epoch": 0.6939163498098859,
+      "grad_norm": 0.8475425527465938,
+      "learning_rate": 1.652347341746737e-05,
+      "loss": 0.4269,
+      "step": 365
+    },
+    {
+      "epoch": 0.6958174904942965,
+      "grad_norm": 0.7068154412947997,
+      "learning_rate": 1.6498267551251618e-05,
+      "loss": 0.4191,
+      "step": 366
+    },
+    {
+      "epoch": 0.6977186311787072,
+      "grad_norm": 0.8026030499241988,
+      "learning_rate": 1.6472990018788884e-05,
+      "loss": 0.4061,
+      "step": 367
+    },
+    {
+      "epoch": 0.6996197718631179,
+      "grad_norm": 0.7293620095428041,
+      "learning_rate": 1.644764109885284e-05,
+      "loss": 0.4419,
+      "step": 368
+    },
+    {
+      "epoch": 0.7015209125475285,
+      "grad_norm": 0.7903767231518011,
+      "learning_rate": 1.642222107100446e-05,
+      "loss": 0.4242,
+      "step": 369
+    },
+    {
+      "epoch": 0.7034220532319392,
+      "grad_norm": 0.7428664679528834,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 0.4204,
+      "step": 370
+    },
+    {
+      "epoch": 0.7053231939163498,
+      "grad_norm": 0.7265102623479446,
+      "learning_rate": 1.6371168813732514e-05,
+      "loss": 0.4212,
+      "step": 371
+    },
+    {
+      "epoch": 0.7072243346007605,
+      "grad_norm": 0.8407209953347514,
+      "learning_rate": 1.6345537147339578e-05,
+      "loss": 0.4301,
+      "step": 372
+    },
+    {
+      "epoch": 0.7091254752851711,
+      "grad_norm": 0.6705756367173369,
+      "learning_rate": 1.6319835499089358e-05,
+      "loss": 0.4197,
+      "step": 373
+    },
+    {
+      "epoch": 0.7110266159695817,
+      "grad_norm": 0.7497329251626604,
+      "learning_rate": 1.6294064152432878e-05,
+      "loss": 0.4118,
+      "step": 374
+    },
+    {
+      "epoch": 0.7129277566539924,
+      "grad_norm": 0.7105139730721509,
+      "learning_rate": 1.626822339158985e-05,
+      "loss": 0.4161,
+      "step": 375
+    },
+    {
+      "epoch": 0.714828897338403,
+      "grad_norm": 0.711497334411663,
+      "learning_rate": 1.6242313501545522e-05,
+      "loss": 0.4169,
+      "step": 376
+    },
+    {
+      "epoch": 0.7167300380228137,
+      "grad_norm": 0.7522296007691398,
+      "learning_rate": 1.621633476804752e-05,
+      "loss": 0.4386,
+      "step": 377
+    },
+    {
+      "epoch": 0.7186311787072244,
+      "grad_norm": 0.7983681131507673,
+      "learning_rate": 1.6190287477602716e-05,
+      "loss": 0.4375,
+      "step": 378
+    },
+    {
+      "epoch": 0.720532319391635,
+      "grad_norm": 0.6839271230510038,
+      "learning_rate": 1.6164171917474078e-05,
+      "loss": 0.4208,
+      "step": 379
+    },
+    {
+      "epoch": 0.7224334600760456,
+      "grad_norm": 0.7321185457540013,
+      "learning_rate": 1.6137988375677466e-05,
+      "loss": 0.4665,
+      "step": 380
+    },
+    {
+      "epoch": 0.7243346007604563,
+      "grad_norm": 0.7053799603886048,
+      "learning_rate": 1.6111737140978495e-05,
+      "loss": 0.4083,
+      "step": 381
+    },
+    {
+      "epoch": 0.7262357414448669,
+      "grad_norm": 0.6894943613796017,
+      "learning_rate": 1.6085418502889315e-05,
+      "loss": 0.4208,
+      "step": 382
+    },
+    {
+      "epoch": 0.7281368821292775,
+      "grad_norm": 0.7482115631072515,
+      "learning_rate": 1.6059032751665454e-05,
+      "loss": 0.4449,
+      "step": 383
+    },
+    {
+      "epoch": 0.7300380228136882,
+      "grad_norm": 0.7312175060795585,
+      "learning_rate": 1.6032580178302585e-05,
+      "loss": 0.4491,
+      "step": 384
+    },
+    {
+      "epoch": 0.7319391634980988,
+      "grad_norm": 0.7036819720430177,
+      "learning_rate": 1.600606107453333e-05,
+      "loss": 0.4402,
+      "step": 385
+    },
+    {
+      "epoch": 0.7338403041825095,
+      "grad_norm": 0.7143560438458977,
+      "learning_rate": 1.597947573282405e-05,
+      "loss": 0.4293,
+      "step": 386
+    },
+    {
+      "epoch": 0.7357414448669202,
+      "grad_norm": 0.6898327204445933,
+      "learning_rate": 1.5952824446371608e-05,
+      "loss": 0.4501,
+      "step": 387
+    },
+    {
+      "epoch": 0.7376425855513308,
+      "grad_norm": 0.6864240736559833,
+      "learning_rate": 1.592610750910014e-05,
+      "loss": 0.415,
+      "step": 388
+    },
+    {
+      "epoch": 0.7395437262357415,
+      "grad_norm": 0.6744977552810958,
+      "learning_rate": 1.589932521565781e-05,
+      "loss": 0.4286,
+      "step": 389
+    },
+    {
+      "epoch": 0.7414448669201521,
+      "grad_norm": 0.7165569986474158,
+      "learning_rate": 1.587247786141358e-05,
+      "loss": 0.4194,
+      "step": 390
+    },
+    {
+      "epoch": 0.7433460076045627,
+      "grad_norm": 0.7144470279954032,
+      "learning_rate": 1.5845565742453906e-05,
+      "loss": 0.4409,
+      "step": 391
+    },
+    {
+      "epoch": 0.7452471482889734,
+      "grad_norm": 0.7362490448857079,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 0.423,
+      "step": 392
+    },
+    {
+      "epoch": 0.747148288973384,
+      "grad_norm": 0.7162007431267081,
+      "learning_rate": 1.5791548398302167e-05,
+      "loss": 0.4411,
+      "step": 393
+    },
+    {
+      "epoch": 0.7490494296577946,
+      "grad_norm": 0.6507891613824422,
+      "learning_rate": 1.5764443768841234e-05,
+      "loss": 0.399,
+      "step": 394
+    },
+    {
+      "epoch": 0.7509505703422054,
+      "grad_norm": 0.6478017561460849,
+      "learning_rate": 1.5737275566120577e-05,
+      "loss": 0.4254,
+      "step": 395
+    },
+    {
+      "epoch": 0.752851711026616,
+      "grad_norm": 0.6392644009475273,
+      "learning_rate": 1.5710044089765144e-05,
+      "loss": 0.4246,
+      "step": 396
+    },
+    {
+      "epoch": 0.7547528517110266,
+      "grad_norm": 0.690859818302916,
+      "learning_rate": 1.5682749640097708e-05,
+      "loss": 0.4012,
+      "step": 397
+    },
+    {
+      "epoch": 0.7566539923954373,
+      "grad_norm": 0.6637356238751483,
+      "learning_rate": 1.565539251813554e-05,
+      "loss": 0.396,
+      "step": 398
+    },
+    {
+      "epoch": 0.7585551330798479,
+      "grad_norm": 0.7322017503927534,
+      "learning_rate": 1.5627973025587093e-05,
+      "loss": 0.4354,
+      "step": 399
+    },
+    {
+      "epoch": 0.7604562737642585,
+      "grad_norm": 0.6746692530126546,
+      "learning_rate": 1.560049146484868e-05,
+      "loss": 0.412,
+      "step": 400
+    },
+    {
+      "epoch": 0.7623574144486692,
+      "grad_norm": 0.7492842417494959,
+      "learning_rate": 1.5572948139001128e-05,
+      "loss": 0.445,
+      "step": 401
+    },
+    {
+      "epoch": 0.7642585551330798,
+      "grad_norm": 0.6904203102291823,
+      "learning_rate": 1.5545343351806443e-05,
+      "loss": 0.4039,
+      "step": 402
+    },
+    {
+      "epoch": 0.7661596958174905,
+      "grad_norm": 0.7623602220363814,
+      "learning_rate": 1.551767740770446e-05,
+      "loss": 0.4092,
+      "step": 403
+    },
+    {
+      "epoch": 0.7680608365019012,
+      "grad_norm": 0.7647807838405704,
+      "learning_rate": 1.5489950611809484e-05,
+      "loss": 0.4145,
+      "step": 404
+    },
+    {
+      "epoch": 0.7699619771863118,
+      "grad_norm": 0.7415420747755611,
+      "learning_rate": 1.5462163269906928e-05,
+      "loss": 0.4224,
+      "step": 405
+    },
+    {
+      "epoch": 0.7718631178707225,
+      "grad_norm": 0.7482606678642779,
+      "learning_rate": 1.5434315688449924e-05,
+      "loss": 0.4317,
+      "step": 406
+    },
+    {
+      "epoch": 0.7737642585551331,
+      "grad_norm": 0.6913822949781429,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.4054,
+      "step": 407
+    },
+    {
+      "epoch": 0.7756653992395437,
+      "grad_norm": 0.7156376527897887,
+      "learning_rate": 1.5378441036003543e-05,
+      "loss": 0.4477,
+      "step": 408
+    },
+    {
+      "epoch": 0.7775665399239544,
+      "grad_norm": 0.6882963696638815,
+      "learning_rate": 1.535041458122865e-05,
+      "loss": 0.4063,
+      "step": 409
+    },
+    {
+      "epoch": 0.779467680608365,
+      "grad_norm": 0.6816034157462327,
+      "learning_rate": 1.5322329119321508e-05,
+      "loss": 0.4171,
+      "step": 410
+    },
+    {
+      "epoch": 0.7813688212927756,
+      "grad_norm": 0.6937176509233782,
+      "learning_rate": 1.529418496002308e-05,
+      "loss": 0.4164,
+      "step": 411
+    },
+    {
+      "epoch": 0.7832699619771863,
+      "grad_norm": 0.6686609147429702,
+      "learning_rate": 1.5265982413721662e-05,
+      "loss": 0.3962,
+      "step": 412
+    },
+    {
+      "epoch": 0.785171102661597,
+      "grad_norm": 0.7217285772997423,
+      "learning_rate": 1.5237721791449497e-05,
+      "loss": 0.428,
+      "step": 413
+    },
+    {
+      "epoch": 0.7870722433460076,
+      "grad_norm": 0.7218632785325884,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 0.4175,
+      "step": 414
+    },
+    {
+      "epoch": 0.7889733840304183,
+      "grad_norm": 0.6024709473117515,
+      "learning_rate": 1.5181027566320858e-05,
+      "loss": 0.4102,
+      "step": 415
+    },
+    {
+      "epoch": 0.7908745247148289,
+      "grad_norm": 0.7054997399086531,
+      "learning_rate": 1.5152594588717544e-05,
+      "loss": 0.4305,
+      "step": 416
+    },
+    {
+      "epoch": 0.7927756653992395,
+      "grad_norm": 0.6825783621606308,
+      "learning_rate": 1.5124104785642909e-05,
+      "loss": 0.4241,
+      "step": 417
+    },
+    {
+      "epoch": 0.7946768060836502,
+      "grad_norm": 0.7267953388305626,
+      "learning_rate": 1.5095558471297196e-05,
+      "loss": 0.4199,
+      "step": 418
+    },
+    {
+      "epoch": 0.7965779467680608,
+      "grad_norm": 0.6382095119192498,
+      "learning_rate": 1.5066955960503893e-05,
+      "loss": 0.4169,
+      "step": 419
+    },
+    {
+      "epoch": 0.7984790874524715,
+      "grad_norm": 0.7145201555635163,
+      "learning_rate": 1.5038297568706244e-05,
+      "loss": 0.444,
+      "step": 420
+    },
+    {
+      "epoch": 0.8003802281368821,
+      "grad_norm": 0.6938113513341884,
+      "learning_rate": 1.5009583611963772e-05,
+      "loss": 0.42,
+      "step": 421
+    },
+    {
+      "epoch": 0.8022813688212928,
+      "grad_norm": 0.6340581529912502,
+      "learning_rate": 1.4980814406948806e-05,
+      "loss": 0.4136,
+      "step": 422
+    },
+    {
+      "epoch": 0.8041825095057035,
+      "grad_norm": 0.6658823774122303,
+      "learning_rate": 1.4951990270942991e-05,
+      "loss": 0.3994,
+      "step": 423
+    },
+    {
+      "epoch": 0.8060836501901141,
+      "grad_norm": 0.6705619036501977,
+      "learning_rate": 1.492311152183376e-05,
+      "loss": 0.3967,
+      "step": 424
+    },
+    {
+      "epoch": 0.8079847908745247,
+      "grad_norm": 0.6964488087372772,
+      "learning_rate": 1.4894178478110856e-05,
+      "loss": 0.43,
+      "step": 425
+    },
+    {
+      "epoch": 0.8098859315589354,
+      "grad_norm": 0.6451331845406248,
+      "learning_rate": 1.4865191458862816e-05,
+      "loss": 0.4044,
+      "step": 426
+    },
+    {
+      "epoch": 0.811787072243346,
+      "grad_norm": 0.7430789787491603,
+      "learning_rate": 1.4836150783773442e-05,
+      "loss": 0.4029,
+      "step": 427
+    },
+    {
+      "epoch": 0.8136882129277566,
+      "grad_norm": 0.7041897470949452,
+      "learning_rate": 1.4807056773118276e-05,
+      "loss": 0.4402,
+      "step": 428
+    },
+    {
+      "epoch": 0.8155893536121673,
+      "grad_norm": 0.6036106557881402,
+      "learning_rate": 1.4777909747761085e-05,
+      "loss": 0.3768,
+      "step": 429
+    },
+    {
+      "epoch": 0.8174904942965779,
+      "grad_norm": 0.7135732599367652,
+      "learning_rate": 1.4748710029150296e-05,
+      "loss": 0.4298,
+      "step": 430
+    },
+    {
+      "epoch": 0.8193916349809885,
+      "grad_norm": 0.6942497407500151,
+      "learning_rate": 1.4719457939315468e-05,
+      "loss": 0.4259,
+      "step": 431
+    },
+    {
+      "epoch": 0.8212927756653993,
+      "grad_norm": 0.6750746339690238,
+      "learning_rate": 1.4690153800863743e-05,
+      "loss": 0.4135,
+      "step": 432
+    },
+    {
+      "epoch": 0.8231939163498099,
+      "grad_norm": 0.7974241177699236,
+      "learning_rate": 1.4660797936976278e-05,
+      "loss": 0.4055,
+      "step": 433
+    },
+    {
+      "epoch": 0.8250950570342205,
+      "grad_norm": 0.6966474517472557,
+      "learning_rate": 1.4631390671404682e-05,
+      "loss": 0.4145,
+      "step": 434
+    },
+    {
+      "epoch": 0.8269961977186312,
+      "grad_norm": 0.6958050069539574,
+      "learning_rate": 1.460193232846745e-05,
+      "loss": 0.4088,
+      "step": 435
+    },
+    {
+      "epoch": 0.8288973384030418,
+      "grad_norm": 0.8461081012871562,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 0.444,
+      "step": 436
+    },
+    {
+      "epoch": 0.8307984790874525,
+      "grad_norm": 0.750880681049732,
+      "learning_rate": 1.4542863710583022e-05,
+      "loss": 0.4363,
+      "step": 437
+    },
+    {
+      "epoch": 0.8326996197718631,
+      "grad_norm": 0.7788256285888076,
+      "learning_rate": 1.4513254087075015e-05,
+      "loss": 0.42,
+      "step": 438
+    },
+    {
+      "epoch": 0.8346007604562737,
+      "grad_norm": 0.649175258098246,
+      "learning_rate": 1.4483594689072571e-05,
+      "loss": 0.4097,
+      "step": 439
+    },
+    {
+      "epoch": 0.8365019011406845,
+      "grad_norm": 0.7668411520179457,
+      "learning_rate": 1.4453885843674837e-05,
+      "loss": 0.4112,
+      "step": 440
+    },
+    {
+      "epoch": 0.8384030418250951,
+      "grad_norm": 0.6539260565291376,
+      "learning_rate": 1.4424127878526278e-05,
+      "loss": 0.4227,
+      "step": 441
+    },
+    {
+      "epoch": 0.8403041825095057,
+      "grad_norm": 0.7428601378603686,
+      "learning_rate": 1.4394321121813093e-05,
+      "loss": 0.413,
+      "step": 442
+    },
+    {
+      "epoch": 0.8422053231939164,
+      "grad_norm": 0.7088978711396908,
+      "learning_rate": 1.436446590225957e-05,
+      "loss": 0.4288,
+      "step": 443
+    },
+    {
+      "epoch": 0.844106463878327,
+      "grad_norm": 0.6254495483866781,
+      "learning_rate": 1.433456254912447e-05,
+      "loss": 0.4034,
+      "step": 444
+    },
+    {
+      "epoch": 0.8460076045627376,
+      "grad_norm": 0.7395571305821382,
+      "learning_rate": 1.4304611392197399e-05,
+      "loss": 0.4188,
+      "step": 445
+    },
+    {
+      "epoch": 0.8479087452471483,
+      "grad_norm": 0.6669766854021364,
+      "learning_rate": 1.427461276179517e-05,
+      "loss": 0.4185,
+      "step": 446
+    },
+    {
+      "epoch": 0.8498098859315589,
+      "grad_norm": 0.6447879931458861,
+      "learning_rate": 1.4244566988758152e-05,
+      "loss": 0.4043,
+      "step": 447
+    },
+    {
+      "epoch": 0.8517110266159695,
+      "grad_norm": 0.6711757818041713,
+      "learning_rate": 1.4214474404446633e-05,
+      "loss": 0.4181,
+      "step": 448
+    },
+    {
+      "epoch": 0.8536121673003803,
+      "grad_norm": 0.6688065238182859,
+      "learning_rate": 1.4184335340737158e-05,
+      "loss": 0.4155,
+      "step": 449
+    },
+    {
+      "epoch": 0.8555133079847909,
+      "grad_norm": 0.6983026223100461,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.4221,
+      "step": 450
+    },
+    {
+      "epoch": 0.8574144486692015,
+      "grad_norm": 0.7431474290668802,
+      "learning_rate": 1.4123919105189836e-05,
+      "loss": 0.4298,
+      "step": 451
+    },
+    {
+      "epoch": 0.8593155893536122,
+      "grad_norm": 0.649594518526619,
+      "learning_rate": 1.4093642599653406e-05,
+      "loss": 0.4125,
+      "step": 452
+    },
+    {
+      "epoch": 0.8612167300380228,
+      "grad_norm": 0.7317869936146294,
+      "learning_rate": 1.40633209473145e-05,
+      "loss": 0.4353,
+      "step": 453
+    },
+    {
+      "epoch": 0.8631178707224335,
+      "grad_norm": 0.7362724141787009,
+      "learning_rate": 1.4032954482575938e-05,
+      "loss": 0.4193,
+      "step": 454
+    },
+    {
+      "epoch": 0.8650190114068441,
+      "grad_norm": 0.6669535303620008,
+      "learning_rate": 1.4002543540334766e-05,
+      "loss": 0.4209,
+      "step": 455
+    },
+    {
+      "epoch": 0.8669201520912547,
+      "grad_norm": 0.8014446637371387,
+      "learning_rate": 1.3972088455978537e-05,
+      "loss": 0.4007,
+      "step": 456
+    },
+    {
+      "epoch": 0.8688212927756654,
+      "grad_norm": 0.6830164460800686,
+      "learning_rate": 1.3941589565381635e-05,
+      "loss": 0.4087,
+      "step": 457
+    },
+    {
+      "epoch": 0.870722433460076,
+      "grad_norm": 0.747393751188262,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 0.4427,
+      "step": 458
+    },
+    {
+      "epoch": 0.8726235741444867,
+      "grad_norm": 0.7240004489943094,
+      "learning_rate": 1.3880461711375224e-05,
+      "loss": 0.3907,
+      "step": 459
+    },
+    {
+      "epoch": 0.8745247148288974,
+      "grad_norm": 0.6991190446193921,
+      "learning_rate": 1.3849833422115221e-05,
+      "loss": 0.4167,
+      "step": 460
+    },
+    {
+      "epoch": 0.876425855513308,
+      "grad_norm": 0.6709088702982998,
+      "learning_rate": 1.3819162674906134e-05,
+      "loss": 0.4109,
+      "step": 461
+    },
+    {
+      "epoch": 0.8783269961977186,
+      "grad_norm": 0.6383674256937936,
+      "learning_rate": 1.378844980800078e-05,
+      "loss": 0.4016,
+      "step": 462
+    },
+    {
+      "epoch": 0.8802281368821293,
+      "grad_norm": 0.6780481175141266,
+      "learning_rate": 1.3757695160116502e-05,
+      "loss": 0.4415,
+      "step": 463
+    },
+    {
+      "epoch": 0.8821292775665399,
+      "grad_norm": 0.7545521572421436,
+      "learning_rate": 1.3726899070431423e-05,
+      "loss": 0.4114,
+      "step": 464
+    },
+    {
+      "epoch": 0.8840304182509505,
+      "grad_norm": 0.685275430836102,
+      "learning_rate": 1.3696061878580707e-05,
+      "loss": 0.4293,
+      "step": 465
+    },
+    {
+      "epoch": 0.8859315589353612,
+      "grad_norm": 0.7592398931712849,
+      "learning_rate": 1.3665183924652817e-05,
+      "loss": 0.4269,
+      "step": 466
+    },
+    {
+      "epoch": 0.8878326996197718,
+      "grad_norm": 0.7078761240018746,
+      "learning_rate": 1.3634265549185755e-05,
+      "loss": 0.3981,
+      "step": 467
+    },
+    {
+      "epoch": 0.8897338403041825,
+      "grad_norm": 0.6616593551003727,
+      "learning_rate": 1.3603307093163319e-05,
+      "loss": 0.4324,
+      "step": 468
+    },
+    {
+      "epoch": 0.8916349809885932,
+      "grad_norm": 0.8105017143028723,
+      "learning_rate": 1.3572308898011328e-05,
+      "loss": 0.4272,
+      "step": 469
+    },
+    {
+      "epoch": 0.8935361216730038,
+      "grad_norm": 0.7109008731480415,
+      "learning_rate": 1.3541271305593878e-05,
+      "loss": 0.4299,
+      "step": 470
+    },
+    {
+      "epoch": 0.8954372623574145,
+      "grad_norm": 0.729636475666318,
+      "learning_rate": 1.3510194658209547e-05,
+      "loss": 0.4128,
+      "step": 471
+    },
+    {
+      "epoch": 0.8973384030418251,
+      "grad_norm": 0.7329667216343763,
+      "learning_rate": 1.3479079298587634e-05,
+      "loss": 0.4028,
+      "step": 472
+    },
+    {
+      "epoch": 0.8992395437262357,
+      "grad_norm": 0.6731304236118901,
+      "learning_rate": 1.3447925569884374e-05,
+      "loss": 0.4071,
+      "step": 473
+    },
+    {
+      "epoch": 0.9011406844106464,
+      "grad_norm": 0.6658011874277965,
+      "learning_rate": 1.3416733815679166e-05,
+      "loss": 0.4209,
+      "step": 474
+    },
+    {
+      "epoch": 0.903041825095057,
+      "grad_norm": 0.9353383386030673,
+      "learning_rate": 1.3385504379970764e-05,
+      "loss": 0.4343,
+      "step": 475
+    },
+    {
+      "epoch": 0.9049429657794676,
+      "grad_norm": 0.6896577730866533,
+      "learning_rate": 1.3354237607173494e-05,
+      "loss": 0.408,
+      "step": 476
+    },
+    {
+      "epoch": 0.9068441064638784,
+      "grad_norm": 0.7799019761172136,
+      "learning_rate": 1.3322933842113457e-05,
+      "loss": 0.4196,
+      "step": 477
+    },
+    {
+      "epoch": 0.908745247148289,
+      "grad_norm": 0.6980303765076431,
+      "learning_rate": 1.3291593430024727e-05,
+      "loss": 0.4028,
+      "step": 478
+    },
+    {
+      "epoch": 0.9106463878326996,
+      "grad_norm": 0.6425197088129551,
+      "learning_rate": 1.3260216716545534e-05,
+      "loss": 0.4141,
+      "step": 479
+    },
+    {
+      "epoch": 0.9125475285171103,
+      "grad_norm": 0.7098001869403527,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 0.3955,
+      "step": 480
+    },
+    {
+      "epoch": 0.9144486692015209,
+      "grad_norm": 0.7337594792132631,
+      "learning_rate": 1.319735576996663e-05,
+      "loss": 0.4409,
+      "step": 481
+    },
+    {
+      "epoch": 0.9163498098859315,
+      "grad_norm": 0.6734545627310691,
+      "learning_rate": 1.3165872230129869e-05,
+      "loss": 0.4153,
+      "step": 482
+    },
+    {
+      "epoch": 0.9182509505703422,
+      "grad_norm": 0.6236575476961407,
+      "learning_rate": 1.3134353775420895e-05,
+      "loss": 0.4048,
+      "step": 483
+    },
+    {
+      "epoch": 0.9201520912547528,
+      "grad_norm": 0.6288528776701348,
+      "learning_rate": 1.3102800753441488e-05,
+      "loss": 0.4232,
+      "step": 484
+    },
+    {
+      "epoch": 0.9220532319391636,
+      "grad_norm": 0.639800204132476,
+      "learning_rate": 1.3071213512174655e-05,
+      "loss": 0.3898,
+      "step": 485
+    },
+    {
+      "epoch": 0.9239543726235742,
+      "grad_norm": 0.6524562190867899,
+      "learning_rate": 1.3039592399980785e-05,
+      "loss": 0.3912,
+      "step": 486
+    },
+    {
+      "epoch": 0.9258555133079848,
+      "grad_norm": 0.6824650027018551,
+      "learning_rate": 1.3007937765593818e-05,
+      "loss": 0.4333,
+      "step": 487
+    },
+    {
+      "epoch": 0.9277566539923955,
+      "grad_norm": 0.6552292730322581,
+      "learning_rate": 1.2976249958117395e-05,
+      "loss": 0.4189,
+      "step": 488
+    },
+    {
+      "epoch": 0.9296577946768061,
+      "grad_norm": 0.7133658504688607,
+      "learning_rate": 1.2944529327021002e-05,
+      "loss": 0.4142,
+      "step": 489
+    },
+    {
+      "epoch": 0.9315589353612167,
+      "grad_norm": 0.6589823172228413,
+      "learning_rate": 1.291277622213612e-05,
+      "loss": 0.4051,
+      "step": 490
+    },
+    {
+      "epoch": 0.9334600760456274,
+      "grad_norm": 0.6232312823034041,
+      "learning_rate": 1.2880990993652379e-05,
+      "loss": 0.4127,
+      "step": 491
+    },
+    {
+      "epoch": 0.935361216730038,
+      "grad_norm": 0.6199813634957917,
+      "learning_rate": 1.2849173992113669e-05,
+      "loss": 0.4233,
+      "step": 492
+    },
+    {
+      "epoch": 0.9372623574144486,
+      "grad_norm": 0.6349721432411458,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.4226,
+      "step": 493
+    },
+    {
+      "epoch": 0.9391634980988594,
+      "grad_norm": 0.651392535293986,
+      "learning_rate": 1.2785446073795118e-05,
+      "loss": 0.406,
+      "step": 494
+    },
+    {
+      "epoch": 0.94106463878327,
+      "grad_norm": 0.6810702917011978,
+      "learning_rate": 1.2753535859839638e-05,
+      "loss": 0.4158,
+      "step": 495
+    },
+    {
+      "epoch": 0.9429657794676806,
+      "grad_norm": 0.6080923737443408,
+      "learning_rate": 1.272159527847016e-05,
+      "loss": 0.3883,
+      "step": 496
+    },
+    {
+      "epoch": 0.9448669201520913,
+      "grad_norm": 0.6742221447430864,
+      "learning_rate": 1.2689624681943897e-05,
+      "loss": 0.4321,
+      "step": 497
+    },
+    {
+      "epoch": 0.9467680608365019,
+      "grad_norm": 0.6605724695771594,
+      "learning_rate": 1.2657624422849077e-05,
+      "loss": 0.4143,
+      "step": 498
+    },
+    {
+      "epoch": 0.9486692015209125,
+      "grad_norm": 0.6981620511024048,
+      "learning_rate": 1.2625594854101066e-05,
+      "loss": 0.4268,
+      "step": 499
+    },
+    {
+      "epoch": 0.9505703422053232,
+      "grad_norm": 0.733925946123189,
+      "learning_rate": 1.2593536328938471e-05,
+      "loss": 0.4195,
+      "step": 500
+    },
+    {
+      "epoch": 0.9524714828897338,
+      "grad_norm": 0.6411640372135279,
+      "learning_rate": 1.2561449200919253e-05,
+      "loss": 0.3944,
+      "step": 501
+    },
+    {
+      "epoch": 0.9543726235741445,
+      "grad_norm": 0.6886217336459237,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 0.3955,
+      "step": 502
+    },
+    {
+      "epoch": 0.9562737642585551,
+      "grad_norm": 0.71390769453024,
+      "learning_rate": 1.2497190552116082e-05,
+      "loss": 0.4382,
+      "step": 503
+    },
+    {
+      "epoch": 0.9581749049429658,
+      "grad_norm": 0.7074739965381804,
+      "learning_rate": 1.2465019740009662e-05,
+      "loss": 0.4018,
+      "step": 504
+    },
+    {
+      "epoch": 0.9600760456273765,
+      "grad_norm": 0.6524668330978645,
+      "learning_rate": 1.2432821742393854e-05,
+      "loss": 0.4168,
+      "step": 505
+    },
+    {
+      "epoch": 0.9619771863117871,
+      "grad_norm": 0.6118329112192084,
+      "learning_rate": 1.2400596914364792e-05,
+      "loss": 0.4059,
+      "step": 506
+    },
+    {
+      "epoch": 0.9638783269961977,
+      "grad_norm": 0.721209861200704,
+      "learning_rate": 1.2368345611314508e-05,
+      "loss": 0.4199,
+      "step": 507
+    },
+    {
+      "epoch": 0.9657794676806084,
+      "grad_norm": 0.7079320417430721,
+      "learning_rate": 1.2336068188927002e-05,
+      "loss": 0.3906,
+      "step": 508
+    },
+    {
+      "epoch": 0.967680608365019,
+      "grad_norm": 0.6476464948749265,
+      "learning_rate": 1.2303765003174342e-05,
+      "loss": 0.415,
+      "step": 509
+    },
+    {
+      "epoch": 0.9695817490494296,
+      "grad_norm": 0.6601627816624424,
+      "learning_rate": 1.2271436410312727e-05,
+      "loss": 0.4267,
+      "step": 510
+    },
+    {
+      "epoch": 0.9714828897338403,
+      "grad_norm": 0.6900159137712154,
+      "learning_rate": 1.2239082766878557e-05,
+      "loss": 0.43,
+      "step": 511
+    },
+    {
+      "epoch": 0.973384030418251,
+      "grad_norm": 0.712273024419061,
+      "learning_rate": 1.2206704429684504e-05,
+      "loss": 0.4225,
+      "step": 512
+    },
+    {
+      "epoch": 0.9752851711026616,
+      "grad_norm": 0.6385658937621382,
+      "learning_rate": 1.2174301755815572e-05,
+      "loss": 0.4175,
+      "step": 513
+    },
+    {
+      "epoch": 0.9771863117870723,
+      "grad_norm": 0.6355968844808271,
+      "learning_rate": 1.2141875102625166e-05,
+      "loss": 0.4017,
+      "step": 514
+    },
+    {
+      "epoch": 0.9790874524714829,
+      "grad_norm": 0.6470140865764977,
+      "learning_rate": 1.2109424827731144e-05,
+      "loss": 0.4285,
+      "step": 515
+    },
+    {
+      "epoch": 0.9809885931558935,
+      "grad_norm": 0.6750781376433499,
+      "learning_rate": 1.2076951289011884e-05,
+      "loss": 0.4266,
+      "step": 516
+    },
+    {
+      "epoch": 0.9828897338403042,
+      "grad_norm": 0.6570847600473237,
+      "learning_rate": 1.204445484460232e-05,
+      "loss": 0.4047,
+      "step": 517
+    },
+    {
+      "epoch": 0.9847908745247148,
+      "grad_norm": 0.7352187375826484,
+      "learning_rate": 1.2011935852890004e-05,
+      "loss": 0.4271,
+      "step": 518
+    },
+    {
+      "epoch": 0.9866920152091255,
+      "grad_norm": 0.6931413338165016,
+      "learning_rate": 1.1979394672511156e-05,
+      "loss": 0.3927,
+      "step": 519
+    },
+    {
+      "epoch": 0.9885931558935361,
+      "grad_norm": 0.7235855925121836,
+      "learning_rate": 1.19468316623467e-05,
+      "loss": 0.4325,
+      "step": 520
+    },
+    {
+      "epoch": 0.9904942965779467,
+      "grad_norm": 0.6904833269511884,
+      "learning_rate": 1.1914247181518312e-05,
+      "loss": 0.41,
+      "step": 521
+    },
+    {
+      "epoch": 0.9923954372623575,
+      "grad_norm": 0.657682870441457,
+      "learning_rate": 1.1881641589384456e-05,
+      "loss": 0.4188,
+      "step": 522
+    },
+    {
+      "epoch": 0.9942965779467681,
+      "grad_norm": 0.7231004563775573,
+      "learning_rate": 1.1849015245536424e-05,
+      "loss": 0.4149,
+      "step": 523
+    },
+    {
+      "epoch": 0.9961977186311787,
+      "grad_norm": 0.6711094774061731,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 0.4153,
+      "step": 524
+    },
+    {
+      "epoch": 0.9980988593155894,
+      "grad_norm": 0.6202574132265527,
+      "learning_rate": 1.1783701742203326e-05,
+      "loss": 0.3902,
+      "step": 525
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6617349524192438,
+      "learning_rate": 1.1751015303029272e-05,
+      "loss": 0.4016,
+      "step": 526
+    },
+    {
+      "epoch": 1.0019011406844107,
+      "grad_norm": 0.8057778037455804,
+      "learning_rate": 1.1718309552755118e-05,
+      "loss": 0.3187,
+      "step": 527
+    },
+    {
+      "epoch": 1.0038022813688212,
+      "grad_norm": 0.7412544361222694,
+      "learning_rate": 1.1685584852076746e-05,
+      "loss": 0.3223,
+      "step": 528
+    },
+    {
+      "epoch": 1.005703422053232,
+      "grad_norm": 0.6759672537804652,
+      "learning_rate": 1.1652841561899042e-05,
+      "loss": 0.3155,
+      "step": 529
+    },
+    {
+      "epoch": 1.0076045627376427,
+      "grad_norm": 0.6748507054648835,
+      "learning_rate": 1.1620080043331901e-05,
+      "loss": 0.3138,
+      "step": 530
+    },
+    {
+      "epoch": 1.0095057034220531,
+      "grad_norm": 0.7942237928890288,
+      "learning_rate": 1.1587300657686254e-05,
+      "loss": 0.3157,
+      "step": 531
+    },
+    {
+      "epoch": 1.0114068441064639,
+      "grad_norm": 0.8846230582484171,
+      "learning_rate": 1.1554503766470069e-05,
+      "loss": 0.3246,
+      "step": 532
+    },
+    {
+      "epoch": 1.0133079847908746,
+      "grad_norm": 0.8119792512496486,
+      "learning_rate": 1.1521689731384391e-05,
+      "loss": 0.329,
+      "step": 533
+    },
+    {
+      "epoch": 1.015209125475285,
+      "grad_norm": 0.7923034561211636,
+      "learning_rate": 1.1488858914319321e-05,
+      "loss": 0.3258,
+      "step": 534
+    },
+    {
+      "epoch": 1.0171102661596958,
+      "grad_norm": 0.7160112870308005,
+      "learning_rate": 1.1456011677350052e-05,
+      "loss": 0.2977,
+      "step": 535
+    },
+    {
+      "epoch": 1.0190114068441065,
+      "grad_norm": 0.7609695163084949,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.3204,
+      "step": 536
+    },
+    {
+      "epoch": 1.020912547528517,
+      "grad_norm": 0.8209869237400179,
+      "learning_rate": 1.1390269392901096e-05,
+      "loss": 0.3086,
+      "step": 537
+    },
+    {
+      "epoch": 1.0228136882129277,
+      "grad_norm": 0.6678411394993501,
+      "learning_rate": 1.1357375070461241e-05,
+      "loss": 0.3067,
+      "step": 538
+    },
+    {
+      "epoch": 1.0247148288973384,
+      "grad_norm": 0.717364650068582,
+      "learning_rate": 1.1324465778188846e-05,
+      "loss": 0.3082,
+      "step": 539
+    },
+    {
+      "epoch": 1.026615969581749,
+      "grad_norm": 0.7308942230333575,
+      "learning_rate": 1.1291541879024568e-05,
+      "loss": 0.3226,
+      "step": 540
+    },
+    {
+      "epoch": 1.0285171102661597,
+      "grad_norm": 0.6456359302870989,
+      "learning_rate": 1.1258603736070145e-05,
+      "loss": 0.2919,
+      "step": 541
+    },
+    {
+      "epoch": 1.0304182509505704,
+      "grad_norm": 0.710931548787735,
+      "learning_rate": 1.1225651712584413e-05,
+      "loss": 0.3099,
+      "step": 542
+    },
+    {
+      "epoch": 1.032319391634981,
+      "grad_norm": 0.709785246141236,
+      "learning_rate": 1.1192686171979288e-05,
+      "loss": 0.3099,
+      "step": 543
+    },
+    {
+      "epoch": 1.0342205323193916,
+      "grad_norm": 0.6981605328190861,
+      "learning_rate": 1.1159707477815756e-05,
+      "loss": 0.2905,
+      "step": 544
+    },
+    {
+      "epoch": 1.0361216730038023,
+      "grad_norm": 0.7170936209092622,
+      "learning_rate": 1.1126715993799875e-05,
+      "loss": 0.3083,
+      "step": 545
+    },
+    {
+      "epoch": 1.038022813688213,
+      "grad_norm": 0.7004979207806828,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 0.2983,
+      "step": 546
+    },
+    {
+      "epoch": 1.0399239543726235,
+      "grad_norm": 0.7458192529204735,
+      "learning_rate": 1.1060696111736515e-05,
+      "loss": 0.3127,
+      "step": 547
+    },
+    {
+      "epoch": 1.0418250950570342,
+      "grad_norm": 0.6809556078021375,
+      "learning_rate": 1.1027668441790358e-05,
+      "loss": 0.3101,
+      "step": 548
+    },
+    {
+      "epoch": 1.043726235741445,
+      "grad_norm": 0.7008889965178042,
+      "learning_rate": 1.099462943818646e-05,
+      "loss": 0.3134,
+      "step": 549
+    },
+    {
+      "epoch": 1.0456273764258555,
+      "grad_norm": 0.7110253244715296,
+      "learning_rate": 1.0961579465295987e-05,
+      "loss": 0.3049,
+      "step": 550
+    },
+    {
+      "epoch": 1.0475285171102662,
+      "grad_norm": 0.7227214280209071,
+      "learning_rate": 1.0928518887611099e-05,
+      "loss": 0.3207,
+      "step": 551
+    },
+    {
+      "epoch": 1.049429657794677,
+      "grad_norm": 0.7033243845953586,
+      "learning_rate": 1.0895448069740902e-05,
+      "loss": 0.3135,
+      "step": 552
+    },
+    {
+      "epoch": 1.0513307984790874,
+      "grad_norm": 0.6612026608935847,
+      "learning_rate": 1.0862367376407433e-05,
+      "loss": 0.2978,
+      "step": 553
+    },
+    {
+      "epoch": 1.053231939163498,
+      "grad_norm": 0.7131320163927526,
+      "learning_rate": 1.0829277172441648e-05,
+      "loss": 0.3062,
+      "step": 554
+    },
+    {
+      "epoch": 1.0551330798479088,
+      "grad_norm": 0.664005238252551,
+      "learning_rate": 1.0796177822779384e-05,
+      "loss": 0.2869,
+      "step": 555
+    },
+    {
+      "epoch": 1.0570342205323193,
+      "grad_norm": 0.710722241160985,
+      "learning_rate": 1.0763069692457346e-05,
+      "loss": 0.305,
+      "step": 556
+    },
+    {
+      "epoch": 1.05893536121673,
+      "grad_norm": 0.7385112551354895,
+      "learning_rate": 1.0729953146609076e-05,
+      "loss": 0.3093,
+      "step": 557
+    },
+    {
+      "epoch": 1.0608365019011408,
+      "grad_norm": 0.7376687467676918,
+      "learning_rate": 1.0696828550460928e-05,
+      "loss": 0.3052,
+      "step": 558
+    },
+    {
+      "epoch": 1.0627376425855513,
+      "grad_norm": 0.7303436532923512,
+      "learning_rate": 1.0663696269328034e-05,
+      "loss": 0.3112,
+      "step": 559
+    },
+    {
+      "epoch": 1.064638783269962,
+      "grad_norm": 0.7359671930671974,
+      "learning_rate": 1.0630556668610286e-05,
+      "loss": 0.297,
+      "step": 560
+    },
+    {
+      "epoch": 1.0665399239543727,
+      "grad_norm": 0.6970657334385685,
+      "learning_rate": 1.059741011378829e-05,
+      "loss": 0.2971,
+      "step": 561
+    },
+    {
+      "epoch": 1.0684410646387832,
+      "grad_norm": 0.8377867831844311,
+      "learning_rate": 1.0564256970419367e-05,
+      "loss": 0.308,
+      "step": 562
+    },
+    {
+      "epoch": 1.070342205323194,
+      "grad_norm": 0.7630566782196169,
+      "learning_rate": 1.0531097604133473e-05,
+      "loss": 0.3187,
+      "step": 563
+    },
+    {
+      "epoch": 1.0722433460076046,
+      "grad_norm": 0.7216293225087548,
+      "learning_rate": 1.0497932380629207e-05,
+      "loss": 0.3048,
+      "step": 564
+    },
+    {
+      "epoch": 1.0741444866920151,
+      "grad_norm": 0.6991326562987921,
+      "learning_rate": 1.0464761665669771e-05,
+      "loss": 0.3048,
+      "step": 565
+    },
+    {
+      "epoch": 1.0760456273764258,
+      "grad_norm": 0.7401360213714921,
+      "learning_rate": 1.0431585825078916e-05,
+      "loss": 0.3,
+      "step": 566
+    },
+    {
+      "epoch": 1.0779467680608366,
+      "grad_norm": 0.7119732505776465,
+      "learning_rate": 1.0398405224736927e-05,
+      "loss": 0.3079,
+      "step": 567
+    },
+    {
+      "epoch": 1.079847908745247,
+      "grad_norm": 0.6702960479673759,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 0.2958,
+      "step": 568
+    },
+    {
+      "epoch": 1.0817490494296578,
+      "grad_norm": 0.7760420851621637,
+      "learning_rate": 1.0332031208579133e-05,
+      "loss": 0.3184,
+      "step": 569
+    },
+    {
+      "epoch": 1.0836501901140685,
+      "grad_norm": 0.7040112198402739,
+      "learning_rate": 1.0298838524770212e-05,
+      "loss": 0.2977,
+      "step": 570
+    },
+    {
+      "epoch": 1.085551330798479,
+      "grad_norm": 0.8414411669683015,
+      "learning_rate": 1.0265642545215872e-05,
+      "loss": 0.3058,
+      "step": 571
+    },
+    {
+      "epoch": 1.0874524714828897,
+      "grad_norm": 0.7133845249788654,
+      "learning_rate": 1.0232443636018502e-05,
+      "loss": 0.3194,
+      "step": 572
+    },
+    {
+      "epoch": 1.0893536121673004,
+      "grad_norm": 0.7206672256867014,
+      "learning_rate": 1.0199242163312794e-05,
+      "loss": 0.2917,
+      "step": 573
+    },
+    {
+      "epoch": 1.091254752851711,
+      "grad_norm": 0.7211983943219695,
+      "learning_rate": 1.0166038493261723e-05,
+      "loss": 0.2984,
+      "step": 574
+    },
+    {
+      "epoch": 1.0931558935361216,
+      "grad_norm": 0.7848732891201756,
+      "learning_rate": 1.013283299205249e-05,
+      "loss": 0.3047,
+      "step": 575
+    },
+    {
+      "epoch": 1.0950570342205324,
+      "grad_norm": 0.7107276063558873,
+      "learning_rate": 1.0099626025892491e-05,
+      "loss": 0.2959,
+      "step": 576
+    },
+    {
+      "epoch": 1.0969581749049429,
+      "grad_norm": 0.7819807312757199,
+      "learning_rate": 1.0066417961005283e-05,
+      "loss": 0.3048,
+      "step": 577
+    },
+    {
+      "epoch": 1.0988593155893536,
+      "grad_norm": 0.7701205978214283,
+      "learning_rate": 1.0033209163626539e-05,
+      "loss": 0.3097,
+      "step": 578
+    },
+    {
+      "epoch": 1.1007604562737643,
+      "grad_norm": 0.7143090612824017,
+      "learning_rate": 1e-05,
+      "loss": 0.3047,
+      "step": 579
+    },
+    {
+      "epoch": 1.102661596958175,
+      "grad_norm": 0.6977877502703614,
+      "learning_rate": 9.966790836373465e-06,
+      "loss": 0.2967,
+      "step": 580
+    },
+    {
+      "epoch": 1.1045627376425855,
+      "grad_norm": 0.7325169357445763,
+      "learning_rate": 9.933582038994719e-06,
+      "loss": 0.3089,
+      "step": 581
+    },
+    {
+      "epoch": 1.1064638783269962,
+      "grad_norm": 0.7797264597989865,
+      "learning_rate": 9.90037397410751e-06,
+      "loss": 0.3182,
+      "step": 582
+    },
+    {
+      "epoch": 1.108365019011407,
+      "grad_norm": 0.691480912342788,
+      "learning_rate": 9.867167007947511e-06,
+      "loss": 0.3061,
+      "step": 583
+    },
+    {
+      "epoch": 1.1102661596958174,
+      "grad_norm": 0.7647706154047912,
+      "learning_rate": 9.833961506738282e-06,
+      "loss": 0.3098,
+      "step": 584
+    },
+    {
+      "epoch": 1.1121673003802282,
+      "grad_norm": 0.7408774853222746,
+      "learning_rate": 9.80075783668721e-06,
+      "loss": 0.3033,
+      "step": 585
+    },
+    {
+      "epoch": 1.1140684410646389,
+      "grad_norm": 0.706168988750185,
+      "learning_rate": 9.767556363981503e-06,
+      "loss": 0.3132,
+      "step": 586
+    },
+    {
+      "epoch": 1.1159695817490494,
+      "grad_norm": 0.6845698316441193,
+      "learning_rate": 9.734357454784131e-06,
+      "loss": 0.2956,
+      "step": 587
+    },
+    {
+      "epoch": 1.11787072243346,
+      "grad_norm": 0.694173196508064,
+      "learning_rate": 9.701161475229791e-06,
+      "loss": 0.2985,
+      "step": 588
+    },
+    {
+      "epoch": 1.1197718631178708,
+      "grad_norm": 0.7016885297787291,
+      "learning_rate": 9.66796879142087e-06,
+      "loss": 0.2982,
+      "step": 589
+    },
+    {
+      "epoch": 1.1216730038022813,
+      "grad_norm": 0.671261859470996,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 0.2882,
+      "step": 590
+    },
+    {
+      "epoch": 1.123574144486692,
+      "grad_norm": 0.6854874511537371,
+      "learning_rate": 9.601594775263073e-06,
+      "loss": 0.2983,
+      "step": 591
+    },
+    {
+      "epoch": 1.1254752851711027,
+      "grad_norm": 0.7821572101680239,
+      "learning_rate": 9.568414174921085e-06,
+      "loss": 0.3194,
+      "step": 592
+    },
+    {
+      "epoch": 1.1273764258555132,
+      "grad_norm": 0.6835974832450535,
+      "learning_rate": 9.535238334330234e-06,
+      "loss": 0.3013,
+      "step": 593
+    },
+    {
+      "epoch": 1.129277566539924,
+      "grad_norm": 0.7334415902163285,
+      "learning_rate": 9.502067619370794e-06,
+      "loss": 0.3072,
+      "step": 594
+    },
+    {
+      "epoch": 1.1311787072243347,
+      "grad_norm": 0.7023121640065046,
+      "learning_rate": 9.468902395866532e-06,
+      "loss": 0.2983,
+      "step": 595
+    },
+    {
+      "epoch": 1.1330798479087452,
+      "grad_norm": 0.710876287872261,
+      "learning_rate": 9.435743029580638e-06,
+      "loss": 0.2965,
+      "step": 596
+    },
+    {
+      "epoch": 1.1349809885931559,
+      "grad_norm": 0.6770817288289162,
+      "learning_rate": 9.402589886211711e-06,
+      "loss": 0.299,
+      "step": 597
+    },
+    {
+      "epoch": 1.1368821292775666,
+      "grad_norm": 0.7171551789560907,
+      "learning_rate": 9.369443331389718e-06,
+      "loss": 0.307,
+      "step": 598
+    },
+    {
+      "epoch": 1.138783269961977,
+      "grad_norm": 0.7696635987347414,
+      "learning_rate": 9.336303730671968e-06,
+      "loss": 0.3021,
+      "step": 599
+    },
+    {
+      "epoch": 1.1406844106463878,
+      "grad_norm": 0.7307664549367932,
+      "learning_rate": 9.303171449539074e-06,
+      "loss": 0.3225,
+      "step": 600
+    },
+    {
+      "epoch": 1.1425855513307985,
+      "grad_norm": 0.6670746579644035,
+      "learning_rate": 9.270046853390924e-06,
+      "loss": 0.2949,
+      "step": 601
+    },
+    {
+      "epoch": 1.144486692015209,
+      "grad_norm": 0.7336106974079709,
+      "learning_rate": 9.236930307542654e-06,
+      "loss": 0.2993,
+      "step": 602
+    },
+    {
+      "epoch": 1.1463878326996197,
+      "grad_norm": 0.7094233409607221,
+      "learning_rate": 9.203822177220621e-06,
+      "loss": 0.2877,
+      "step": 603
+    },
+    {
+      "epoch": 1.1482889733840305,
+      "grad_norm": 0.7982822716997209,
+      "learning_rate": 9.170722827558357e-06,
+      "loss": 0.3039,
+      "step": 604
+    },
+    {
+      "epoch": 1.1501901140684412,
+      "grad_norm": 0.7184238380620656,
+      "learning_rate": 9.13763262359257e-06,
+      "loss": 0.316,
+      "step": 605
+    },
+    {
+      "epoch": 1.1520912547528517,
+      "grad_norm": 0.7690930696895716,
+      "learning_rate": 9.104551930259101e-06,
+      "loss": 0.3076,
+      "step": 606
+    },
+    {
+      "epoch": 1.1539923954372624,
+      "grad_norm": 0.7418742541581451,
+      "learning_rate": 9.071481112388905e-06,
+      "loss": 0.3135,
+      "step": 607
+    },
+    {
+      "epoch": 1.1558935361216731,
+      "grad_norm": 0.6613978970230133,
+      "learning_rate": 9.038420534704015e-06,
+      "loss": 0.2972,
+      "step": 608
+    },
+    {
+      "epoch": 1.1577946768060836,
+      "grad_norm": 0.8092191953773207,
+      "learning_rate": 9.005370561813545e-06,
+      "loss": 0.3195,
+      "step": 609
+    },
+    {
+      "epoch": 1.1596958174904943,
+      "grad_norm": 0.7189809866365593,
+      "learning_rate": 8.972331558209644e-06,
+      "loss": 0.3097,
+      "step": 610
+    },
+    {
+      "epoch": 1.161596958174905,
+      "grad_norm": 0.6821296213837992,
+      "learning_rate": 8.939303888263485e-06,
+      "loss": 0.3026,
+      "step": 611
+    },
+    {
+      "epoch": 1.1634980988593155,
+      "grad_norm": 0.7512975816311926,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 0.3063,
+      "step": 612
+    },
+    {
+      "epoch": 1.1653992395437263,
+      "grad_norm": 0.777050746487349,
+      "learning_rate": 8.873284006200129e-06,
+      "loss": 0.3079,
+      "step": 613
+    },
+    {
+      "epoch": 1.167300380228137,
+      "grad_norm": 0.7646375380503718,
+      "learning_rate": 8.840292522184247e-06,
+      "loss": 0.2944,
+      "step": 614
+    },
+    {
+      "epoch": 1.1692015209125475,
+      "grad_norm": 0.8221348922593711,
+      "learning_rate": 8.807313828020715e-06,
+      "loss": 0.3196,
+      "step": 615
+    },
+    {
+      "epoch": 1.1711026615969582,
+      "grad_norm": 0.7522833675209694,
+      "learning_rate": 8.774348287415589e-06,
+      "loss": 0.316,
+      "step": 616
+    },
+    {
+      "epoch": 1.173003802281369,
+      "grad_norm": 0.8154761032618316,
+      "learning_rate": 8.74139626392986e-06,
+      "loss": 0.3217,
+      "step": 617
+    },
+    {
+      "epoch": 1.1749049429657794,
+      "grad_norm": 0.8688973359280431,
+      "learning_rate": 8.708458120975436e-06,
+      "loss": 0.3064,
+      "step": 618
+    },
+    {
+      "epoch": 1.1768060836501901,
+      "grad_norm": 0.7893218637035048,
+      "learning_rate": 8.675534221811156e-06,
+      "loss": 0.2924,
+      "step": 619
+    },
+    {
+      "epoch": 1.1787072243346008,
+      "grad_norm": 0.7593175709146341,
+      "learning_rate": 8.64262492953876e-06,
+      "loss": 0.2875,
+      "step": 620
+    },
+    {
+      "epoch": 1.1806083650190113,
+      "grad_norm": 0.8980922553762621,
+      "learning_rate": 8.60973060709891e-06,
+      "loss": 0.3258,
+      "step": 621
+    },
+    {
+      "epoch": 1.182509505703422,
+      "grad_norm": 0.7304724224689862,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.3032,
+      "step": 622
+    },
+    {
+      "epoch": 1.1844106463878328,
+      "grad_norm": 0.8233977368402046,
+      "learning_rate": 8.543988322649954e-06,
+      "loss": 0.3366,
+      "step": 623
+    },
+    {
+      "epoch": 1.1863117870722433,
+      "grad_norm": 0.86410400296458,
+      "learning_rate": 8.511141085680684e-06,
+      "loss": 0.3061,
+      "step": 624
+    },
+    {
+      "epoch": 1.188212927756654,
+      "grad_norm": 0.7817658025550973,
+      "learning_rate": 8.478310268615612e-06,
+      "loss": 0.3153,
+      "step": 625
+    },
+    {
+      "epoch": 1.1901140684410647,
+      "grad_norm": 0.694440893239606,
+      "learning_rate": 8.445496233529934e-06,
+      "loss": 0.302,
+      "step": 626
+    },
+    {
+      "epoch": 1.1920152091254752,
+      "grad_norm": 0.8118544238253723,
+      "learning_rate": 8.41269934231375e-06,
+      "loss": 0.3164,
+      "step": 627
+    },
+    {
+      "epoch": 1.193916349809886,
+      "grad_norm": 0.7335133540539855,
+      "learning_rate": 8.3799199566681e-06,
+      "loss": 0.2839,
+      "step": 628
+    },
+    {
+      "epoch": 1.1958174904942966,
+      "grad_norm": 0.6745989690142823,
+      "learning_rate": 8.34715843810096e-06,
+      "loss": 0.2995,
+      "step": 629
+    },
+    {
+      "epoch": 1.1977186311787071,
+      "grad_norm": 0.8381464598125277,
+      "learning_rate": 8.314415147923254e-06,
+      "loss": 0.3058,
+      "step": 630
+    },
+    {
+      "epoch": 1.1996197718631179,
+      "grad_norm": 0.7575646214083855,
+      "learning_rate": 8.281690447244887e-06,
+      "loss": 0.3019,
+      "step": 631
+    },
+    {
+      "epoch": 1.2015209125475286,
+      "grad_norm": 0.7435604650444324,
+      "learning_rate": 8.248984696970732e-06,
+      "loss": 0.3098,
+      "step": 632
+    },
+    {
+      "epoch": 1.203422053231939,
+      "grad_norm": 0.7167627487804589,
+      "learning_rate": 8.216298257796677e-06,
+      "loss": 0.3171,
+      "step": 633
+    },
+    {
+      "epoch": 1.2053231939163498,
+      "grad_norm": 0.7462508343456407,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 0.2934,
+      "step": 634
+    },
+    {
+      "epoch": 1.2072243346007605,
+      "grad_norm": 0.7154974983480035,
+      "learning_rate": 8.150984754463578e-06,
+      "loss": 0.3052,
+      "step": 635
+    },
+    {
+      "epoch": 1.209125475285171,
+      "grad_norm": 0.6898714738388552,
+      "learning_rate": 8.118358410615545e-06,
+      "loss": 0.2986,
+      "step": 636
+    },
+    {
+      "epoch": 1.2110266159695817,
+      "grad_norm": 0.6530010169738768,
+      "learning_rate": 8.08575281848169e-06,
+      "loss": 0.2966,
+      "step": 637
+    },
+    {
+      "epoch": 1.2129277566539924,
+      "grad_norm": 0.8505752537538249,
+      "learning_rate": 8.0531683376533e-06,
+      "loss": 0.2921,
+      "step": 638
+    },
+    {
+      "epoch": 1.214828897338403,
+      "grad_norm": 0.7330205322721064,
+      "learning_rate": 8.020605327488846e-06,
+      "loss": 0.3058,
+      "step": 639
+    },
+    {
+      "epoch": 1.2167300380228137,
+      "grad_norm": 0.7267873762157572,
+      "learning_rate": 7.988064147110001e-06,
+      "loss": 0.3081,
+      "step": 640
+    },
+    {
+      "epoch": 1.2186311787072244,
+      "grad_norm": 0.6966326168204822,
+      "learning_rate": 7.955545155397684e-06,
+      "loss": 0.3145,
+      "step": 641
+    },
+    {
+      "epoch": 1.2205323193916349,
+      "grad_norm": 0.9097554173815025,
+      "learning_rate": 7.923048710988119e-06,
+      "loss": 0.3009,
+      "step": 642
+    },
+    {
+      "epoch": 1.2224334600760456,
+      "grad_norm": 0.7049115761028015,
+      "learning_rate": 7.890575172268858e-06,
+      "loss": 0.3218,
+      "step": 643
+    },
+    {
+      "epoch": 1.2243346007604563,
+      "grad_norm": 0.7129588056676595,
+      "learning_rate": 7.858124897374837e-06,
+      "loss": 0.2984,
+      "step": 644
+    },
+    {
+      "epoch": 1.2262357414448668,
+      "grad_norm": 0.7963167616618384,
+      "learning_rate": 7.825698244184432e-06,
+      "loss": 0.2949,
+      "step": 645
+    },
+    {
+      "epoch": 1.2281368821292775,
+      "grad_norm": 0.6759179001734439,
+      "learning_rate": 7.7932955703155e-06,
+      "loss": 0.2982,
+      "step": 646
+    },
+    {
+      "epoch": 1.2300380228136882,
+      "grad_norm": 0.7573382452877564,
+      "learning_rate": 7.760917233121443e-06,
+      "loss": 0.3083,
+      "step": 647
+    },
+    {
+      "epoch": 1.231939163498099,
+      "grad_norm": 0.6940656389472214,
+      "learning_rate": 7.728563589687275e-06,
+      "loss": 0.292,
+      "step": 648
+    },
+    {
+      "epoch": 1.2338403041825095,
+      "grad_norm": 0.7056323073620804,
+      "learning_rate": 7.696234996825663e-06,
+      "loss": 0.3163,
+      "step": 649
+    },
+    {
+      "epoch": 1.2357414448669202,
+      "grad_norm": 0.6974380407254572,
+      "learning_rate": 7.663931811073003e-06,
+      "loss": 0.2928,
+      "step": 650
+    },
+    {
+      "epoch": 1.2376425855513309,
+      "grad_norm": 0.7473566725707713,
+      "learning_rate": 7.631654388685496e-06,
+      "loss": 0.3041,
+      "step": 651
+    },
+    {
+      "epoch": 1.2395437262357414,
+      "grad_norm": 0.7183279057715524,
+      "learning_rate": 7.599403085635208e-06,
+      "loss": 0.3049,
+      "step": 652
+    },
+    {
+      "epoch": 1.241444866920152,
+      "grad_norm": 0.7046573585481826,
+      "learning_rate": 7.567178257606147e-06,
+      "loss": 0.3112,
+      "step": 653
+    },
+    {
+      "epoch": 1.2433460076045628,
+      "grad_norm": 0.6756411836040294,
+      "learning_rate": 7.534980259990341e-06,
+      "loss": 0.2894,
+      "step": 654
+    },
+    {
+      "epoch": 1.2452471482889733,
+      "grad_norm": 0.7608364012151083,
+      "learning_rate": 7.50280944788392e-06,
+      "loss": 0.2899,
+      "step": 655
+    },
+    {
+      "epoch": 1.247148288973384,
+      "grad_norm": 0.7426259981955133,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 0.3245,
+      "step": 656
+    },
+    {
+      "epoch": 1.2490494296577948,
+      "grad_norm": 0.7713819448065024,
+      "learning_rate": 7.438550799080746e-06,
+      "loss": 0.3127,
+      "step": 657
+    },
+    {
+      "epoch": 1.2509505703422052,
+      "grad_norm": 0.6642612786558842,
+      "learning_rate": 7.40646367106153e-06,
+      "loss": 0.2938,
+      "step": 658
+    },
+    {
+      "epoch": 1.252851711026616,
+      "grad_norm": 0.7015151327909261,
+      "learning_rate": 7.3744051458989395e-06,
+      "loss": 0.3032,
+      "step": 659
+    },
+    {
+      "epoch": 1.2547528517110267,
+      "grad_norm": 0.6794629279222151,
+      "learning_rate": 7.342375577150928e-06,
+      "loss": 0.3053,
+      "step": 660
+    },
+    {
+      "epoch": 1.2566539923954372,
+      "grad_norm": 0.7081692819708384,
+      "learning_rate": 7.310375318056107e-06,
+      "loss": 0.3093,
+      "step": 661
+    },
+    {
+      "epoch": 1.258555133079848,
+      "grad_norm": 0.7096565729039336,
+      "learning_rate": 7.278404721529843e-06,
+      "loss": 0.3039,
+      "step": 662
+    },
+    {
+      "epoch": 1.2604562737642586,
+      "grad_norm": 0.6639021034892703,
+      "learning_rate": 7.246464140160365e-06,
+      "loss": 0.295,
+      "step": 663
+    },
+    {
+      "epoch": 1.2623574144486693,
+      "grad_norm": 0.7307583245777143,
+      "learning_rate": 7.214553926204884e-06,
+      "loss": 0.299,
+      "step": 664
+    },
+    {
+      "epoch": 1.2642585551330798,
+      "grad_norm": 0.7082822508579133,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.311,
+      "step": 665
+    },
+    {
+      "epoch": 1.2661596958174905,
+      "grad_norm": 0.735574814254584,
+      "learning_rate": 7.150826007886334e-06,
+      "loss": 0.3014,
+      "step": 666
+    },
+    {
+      "epoch": 1.2680608365019013,
+      "grad_norm": 0.7155412097523799,
+      "learning_rate": 7.119009006347625e-06,
+      "loss": 0.2934,
+      "step": 667
+    },
+    {
+      "epoch": 1.2699619771863118,
+      "grad_norm": 0.8503906914668881,
+      "learning_rate": 7.087223777863883e-06,
+      "loss": 0.3002,
+      "step": 668
+    },
+    {
+      "epoch": 1.2718631178707225,
+      "grad_norm": 0.7317049648164148,
+      "learning_rate": 7.055470672979003e-06,
+      "loss": 0.3119,
+      "step": 669
+    },
+    {
+      "epoch": 1.2737642585551332,
+      "grad_norm": 0.7526826129990981,
+      "learning_rate": 7.023750041882609e-06,
+      "loss": 0.3034,
+      "step": 670
+    },
+    {
+      "epoch": 1.2756653992395437,
+      "grad_norm": 0.7179155137470628,
+      "learning_rate": 6.992062234406185e-06,
+      "loss": 0.2979,
+      "step": 671
+    },
+    {
+      "epoch": 1.2775665399239544,
+      "grad_norm": 0.7304272247414944,
+      "learning_rate": 6.960407600019217e-06,
+      "loss": 0.2917,
+      "step": 672
+    },
+    {
+      "epoch": 1.2794676806083651,
+      "grad_norm": 0.666440944008466,
+      "learning_rate": 6.9287864878253475e-06,
+      "loss": 0.304,
+      "step": 673
+    },
+    {
+      "epoch": 1.2813688212927756,
+      "grad_norm": 0.7567439567280311,
+      "learning_rate": 6.897199246558515e-06,
+      "loss": 0.2993,
+      "step": 674
+    },
+    {
+      "epoch": 1.2832699619771863,
+      "grad_norm": 0.7406618429380093,
+      "learning_rate": 6.865646224579108e-06,
+      "loss": 0.3043,
+      "step": 675
+    },
+    {
+      "epoch": 1.285171102661597,
+      "grad_norm": 0.7099246090446435,
+      "learning_rate": 6.834127769870134e-06,
+      "loss": 0.3055,
+      "step": 676
+    },
+    {
+      "epoch": 1.2870722433460076,
+      "grad_norm": 0.6813660057296255,
+      "learning_rate": 6.802644230033373e-06,
+      "loss": 0.3011,
+      "step": 677
+    },
+    {
+      "epoch": 1.2889733840304183,
+      "grad_norm": 0.853660463302684,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 0.289,
+      "step": 678
+    },
+    {
+      "epoch": 1.290874524714829,
+      "grad_norm": 0.7242910664133385,
+      "learning_rate": 6.739783283454469e-06,
+      "loss": 0.3112,
+      "step": 679
+    },
+    {
+      "epoch": 1.2927756653992395,
+      "grad_norm": 0.6443547749166783,
+      "learning_rate": 6.708406569975274e-06,
+      "loss": 0.2833,
+      "step": 680
+    },
+    {
+      "epoch": 1.2946768060836502,
+      "grad_norm": 0.762825451488999,
+      "learning_rate": 6.6770661578865444e-06,
+      "loss": 0.2944,
+      "step": 681
+    },
+    {
+      "epoch": 1.296577946768061,
+      "grad_norm": 0.6937039172160832,
+      "learning_rate": 6.645762392826509e-06,
+      "loss": 0.3039,
+      "step": 682
+    },
+    {
+      "epoch": 1.2984790874524714,
+      "grad_norm": 0.7383138000780872,
+      "learning_rate": 6.614495620029238e-06,
+      "loss": 0.3024,
+      "step": 683
+    },
+    {
+      "epoch": 1.3003802281368821,
+      "grad_norm": 0.7236208230569622,
+      "learning_rate": 6.583266184320836e-06,
+      "loss": 0.2987,
+      "step": 684
+    },
+    {
+      "epoch": 1.3022813688212929,
+      "grad_norm": 0.8299080227359633,
+      "learning_rate": 6.552074430115624e-06,
+      "loss": 0.309,
+      "step": 685
+    },
+    {
+      "epoch": 1.3041825095057034,
+      "grad_norm": 0.6856495650584966,
+      "learning_rate": 6.520920701412371e-06,
+      "loss": 0.2903,
+      "step": 686
+    },
+    {
+      "epoch": 1.306083650190114,
+      "grad_norm": 0.7207431004345501,
+      "learning_rate": 6.489805341790456e-06,
+      "loss": 0.3076,
+      "step": 687
+    },
+    {
+      "epoch": 1.3079847908745248,
+      "grad_norm": 0.7410055528146177,
+      "learning_rate": 6.458728694406124e-06,
+      "loss": 0.3146,
+      "step": 688
+    },
+    {
+      "epoch": 1.3098859315589353,
+      "grad_norm": 0.6688534205235728,
+      "learning_rate": 6.427691101988673e-06,
+      "loss": 0.2979,
+      "step": 689
+    },
+    {
+      "epoch": 1.311787072243346,
+      "grad_norm": 0.7198478551346036,
+      "learning_rate": 6.396692906836686e-06,
+      "loss": 0.3182,
+      "step": 690
+    },
+    {
+      "epoch": 1.3136882129277567,
+      "grad_norm": 0.682083131568912,
+      "learning_rate": 6.3657344508142495e-06,
+      "loss": 0.2979,
+      "step": 691
+    },
+    {
+      "epoch": 1.3155893536121672,
+      "grad_norm": 0.6714840051274571,
+      "learning_rate": 6.334816075347185e-06,
+      "loss": 0.2861,
+      "step": 692
+    },
+    {
+      "epoch": 1.317490494296578,
+      "grad_norm": 0.689820875056291,
+      "learning_rate": 6.303938121419295e-06,
+      "loss": 0.294,
+      "step": 693
+    },
+    {
+      "epoch": 1.3193916349809887,
+      "grad_norm": 0.7147144533712376,
+      "learning_rate": 6.273100929568579e-06,
+      "loss": 0.3075,
+      "step": 694
+    },
+    {
+      "epoch": 1.3212927756653992,
+      "grad_norm": 0.7067187637747501,
+      "learning_rate": 6.242304839883502e-06,
+      "loss": 0.2904,
+      "step": 695
+    },
+    {
+      "epoch": 1.3231939163498099,
+      "grad_norm": 0.6776239862253831,
+      "learning_rate": 6.211550191999223e-06,
+      "loss": 0.3018,
+      "step": 696
+    },
+    {
+      "epoch": 1.3250950570342206,
+      "grad_norm": 0.730336121728125,
+      "learning_rate": 6.18083732509387e-06,
+      "loss": 0.3177,
+      "step": 697
+    },
+    {
+      "epoch": 1.326996197718631,
+      "grad_norm": 0.7125893145513373,
+      "learning_rate": 6.150166577884781e-06,
+      "loss": 0.3036,
+      "step": 698
+    },
+    {
+      "epoch": 1.3288973384030418,
+      "grad_norm": 0.7489275099996262,
+      "learning_rate": 6.119538288624778e-06,
+      "loss": 0.2921,
+      "step": 699
+    },
+    {
+      "epoch": 1.3307984790874525,
+      "grad_norm": 0.7024670259807829,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 0.3008,
+      "step": 700
+    },
+    {
+      "epoch": 1.332699619771863,
+      "grad_norm": 0.7617691476961448,
+      "learning_rate": 6.058410434618367e-06,
+      "loss": 0.3009,
+      "step": 701
+    },
+    {
+      "epoch": 1.3346007604562737,
+      "grad_norm": 0.6601387274434397,
+      "learning_rate": 6.027911544021465e-06,
+      "loss": 0.2883,
+      "step": 702
+    },
+    {
+      "epoch": 1.3365019011406845,
+      "grad_norm": 0.689628809732374,
+      "learning_rate": 5.997456459665237e-06,
+      "loss": 0.3042,
+      "step": 703
+    },
+    {
+      "epoch": 1.338403041825095,
+      "grad_norm": 0.6895061062232118,
+      "learning_rate": 5.967045517424062e-06,
+      "loss": 0.2984,
+      "step": 704
+    },
+    {
+      "epoch": 1.3403041825095057,
+      "grad_norm": 0.8624163150166511,
+      "learning_rate": 5.936679052685505e-06,
+      "loss": 0.3095,
+      "step": 705
+    },
+    {
+      "epoch": 1.3422053231939164,
+      "grad_norm": 0.7472202016568489,
+      "learning_rate": 5.906357400346596e-06,
+      "loss": 0.3074,
+      "step": 706
+    },
+    {
+      "epoch": 1.3441064638783269,
+      "grad_norm": 0.6837283010758579,
+      "learning_rate": 5.876080894810167e-06,
+      "loss": 0.2974,
+      "step": 707
+    },
+    {
+      "epoch": 1.3460076045627376,
+      "grad_norm": 0.8696890589855144,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.3019,
+      "step": 708
+    },
+    {
+      "epoch": 1.3479087452471483,
+      "grad_norm": 0.8784651669872162,
+      "learning_rate": 5.815664659262845e-06,
+      "loss": 0.3123,
+      "step": 709
+    },
+    {
+      "epoch": 1.3498098859315588,
+      "grad_norm": 0.741061794873379,
+      "learning_rate": 5.78552559555337e-06,
+      "loss": 0.3007,
+      "step": 710
+    },
+    {
+      "epoch": 1.3517110266159695,
+      "grad_norm": 0.7928470557663179,
+      "learning_rate": 5.755433011241851e-06,
+      "loss": 0.3002,
+      "step": 711
+    },
+    {
+      "epoch": 1.3536121673003803,
+      "grad_norm": 0.9313650144368278,
+      "learning_rate": 5.725387238204831e-06,
+      "loss": 0.3108,
+      "step": 712
+    },
+    {
+      "epoch": 1.3555133079847907,
+      "grad_norm": 0.7692850992504119,
+      "learning_rate": 5.695388607802603e-06,
+      "loss": 0.3141,
+      "step": 713
+    },
+    {
+      "epoch": 1.3574144486692015,
+      "grad_norm": 0.7153346529811843,
+      "learning_rate": 5.665437450875534e-06,
+      "loss": 0.3054,
+      "step": 714
+    },
+    {
+      "epoch": 1.3593155893536122,
+      "grad_norm": 0.8879627334944136,
+      "learning_rate": 5.635534097740435e-06,
+      "loss": 0.3066,
+      "step": 715
+    },
+    {
+      "epoch": 1.3612167300380227,
+      "grad_norm": 0.7640888913791557,
+      "learning_rate": 5.605678878186911e-06,
+      "loss": 0.2982,
+      "step": 716
+    },
+    {
+      "epoch": 1.3631178707224334,
+      "grad_norm": 0.741269410070457,
+      "learning_rate": 5.575872121473722e-06,
+      "loss": 0.2821,
+      "step": 717
+    },
+    {
+      "epoch": 1.3650190114068441,
+      "grad_norm": 0.7316557351244668,
+      "learning_rate": 5.546114156325166e-06,
+      "loss": 0.3001,
+      "step": 718
+    },
+    {
+      "epoch": 1.3669201520912546,
+      "grad_norm": 0.7049024907065368,
+      "learning_rate": 5.516405310927431e-06,
+      "loss": 0.2925,
+      "step": 719
+    },
+    {
+      "epoch": 1.3688212927756653,
+      "grad_norm": 0.7271822649880768,
+      "learning_rate": 5.4867459129249846e-06,
+      "loss": 0.2863,
+      "step": 720
+    },
+    {
+      "epoch": 1.370722433460076,
+      "grad_norm": 0.7797619996579619,
+      "learning_rate": 5.4571362894169795e-06,
+      "loss": 0.3025,
+      "step": 721
+    },
+    {
+      "epoch": 1.3726235741444868,
+      "grad_norm": 0.7189337871152031,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 0.3016,
+      "step": 722
+    },
+    {
+      "epoch": 1.3745247148288973,
+      "grad_norm": 0.6836738934045961,
+      "learning_rate": 5.398067671532554e-06,
+      "loss": 0.2906,
+      "step": 723
+    },
+    {
+      "epoch": 1.376425855513308,
+      "grad_norm": 0.6878232469470896,
+      "learning_rate": 5.368609328595323e-06,
+      "loss": 0.3025,
+      "step": 724
+    },
+    {
+      "epoch": 1.3783269961977187,
+      "grad_norm": 0.6545903730512905,
+      "learning_rate": 5.339202063023727e-06,
+      "loss": 0.2834,
+      "step": 725
+    },
+    {
+      "epoch": 1.3802281368821292,
+      "grad_norm": 0.6891971992294891,
+      "learning_rate": 5.309846199136258e-06,
+      "loss": 0.2875,
+      "step": 726
+    },
+    {
+      "epoch": 1.38212927756654,
+      "grad_norm": 0.6626599460402622,
+      "learning_rate": 5.280542060684535e-06,
+      "loss": 0.3039,
+      "step": 727
+    },
+    {
+      "epoch": 1.3840304182509506,
+      "grad_norm": 0.7072952699809278,
+      "learning_rate": 5.2512899708497086e-06,
+      "loss": 0.2952,
+      "step": 728
+    },
+    {
+      "epoch": 1.3859315589353614,
+      "grad_norm": 0.689988642582098,
+      "learning_rate": 5.222090252238916e-06,
+      "loss": 0.2825,
+      "step": 729
+    },
+    {
+      "epoch": 1.3878326996197718,
+      "grad_norm": 0.7051359764494823,
+      "learning_rate": 5.192943226881724e-06,
+      "loss": 0.2733,
+      "step": 730
+    },
+    {
+      "epoch": 1.3897338403041826,
+      "grad_norm": 0.7517787459132846,
+      "learning_rate": 5.163849216226562e-06,
+      "loss": 0.3023,
+      "step": 731
+    },
+    {
+      "epoch": 1.3916349809885933,
+      "grad_norm": 0.682390732635279,
+      "learning_rate": 5.134808541137183e-06,
+      "loss": 0.3052,
+      "step": 732
+    },
+    {
+      "epoch": 1.3935361216730038,
+      "grad_norm": 0.6881593159245546,
+      "learning_rate": 5.105821521889147e-06,
+      "loss": 0.3059,
+      "step": 733
+    },
+    {
+      "epoch": 1.3954372623574145,
+      "grad_norm": 0.7612087385242844,
+      "learning_rate": 5.076888478166247e-06,
+      "loss": 0.3057,
+      "step": 734
+    },
+    {
+      "epoch": 1.3973384030418252,
+      "grad_norm": 0.7609153974737631,
+      "learning_rate": 5.048009729057012e-06,
+      "loss": 0.2964,
+      "step": 735
+    },
+    {
+      "epoch": 1.3992395437262357,
+      "grad_norm": 0.7783101982651986,
+      "learning_rate": 5.0191855930511946e-06,
+      "loss": 0.3057,
+      "step": 736
+    },
+    {
+      "epoch": 1.4011406844106464,
+      "grad_norm": 0.6851425513747134,
+      "learning_rate": 4.990416388036233e-06,
+      "loss": 0.2834,
+      "step": 737
+    },
+    {
+      "epoch": 1.4030418250950571,
+      "grad_norm": 0.707074102927044,
+      "learning_rate": 4.961702431293759e-06,
+      "loss": 0.3013,
+      "step": 738
+    },
+    {
+      "epoch": 1.4049429657794676,
+      "grad_norm": 0.7447729289865167,
+      "learning_rate": 4.933044039496107e-06,
+      "loss": 0.3038,
+      "step": 739
+    },
+    {
+      "epoch": 1.4068441064638784,
+      "grad_norm": 0.6726888571543709,
+      "learning_rate": 4.904441528702806e-06,
+      "loss": 0.2933,
+      "step": 740
+    },
+    {
+      "epoch": 1.408745247148289,
+      "grad_norm": 0.7250569654006208,
+      "learning_rate": 4.875895214357093e-06,
+      "loss": 0.296,
+      "step": 741
+    },
+    {
+      "epoch": 1.4106463878326996,
+      "grad_norm": 0.7173409400277954,
+      "learning_rate": 4.847405411282462e-06,
+      "loss": 0.3119,
+      "step": 742
+    },
+    {
+      "epoch": 1.4125475285171103,
+      "grad_norm": 0.690859779724277,
+      "learning_rate": 4.818972433679145e-06,
+      "loss": 0.2917,
+      "step": 743
+    },
+    {
+      "epoch": 1.414448669201521,
+      "grad_norm": 0.6874427322784807,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 0.2895,
+      "step": 744
+    },
+    {
+      "epoch": 1.4163498098859315,
+      "grad_norm": 0.6446492262759717,
+      "learning_rate": 4.762278208550505e-06,
+      "loss": 0.2768,
+      "step": 745
+    },
+    {
+      "epoch": 1.4182509505703422,
+      "grad_norm": 0.701566418607254,
+      "learning_rate": 4.734017586278337e-06,
+      "loss": 0.2947,
+      "step": 746
+    },
+    {
+      "epoch": 1.420152091254753,
+      "grad_norm": 0.7127180159280057,
+      "learning_rate": 4.7058150399769245e-06,
+      "loss": 0.3026,
+      "step": 747
+    },
+    {
+      "epoch": 1.4220532319391634,
+      "grad_norm": 0.6965841731104763,
+      "learning_rate": 4.677670880678493e-06,
+      "loss": 0.3046,
+      "step": 748
+    },
+    {
+      "epoch": 1.4239543726235742,
+      "grad_norm": 0.6968052889962494,
+      "learning_rate": 4.649585418771348e-06,
+      "loss": 0.3023,
+      "step": 749
+    },
+    {
+      "epoch": 1.4258555133079849,
+      "grad_norm": 0.8254584396917158,
+      "learning_rate": 4.621558963996458e-06,
+      "loss": 0.307,
+      "step": 750
+    },
+    {
+      "epoch": 1.4277566539923954,
+      "grad_norm": 0.6906504151361319,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2934,
+      "step": 751
+    },
+    {
+      "epoch": 1.429657794676806,
+      "grad_norm": 0.6791924209726916,
+      "learning_rate": 4.565684311550077e-06,
+      "loss": 0.304,
+      "step": 752
+    },
+    {
+      "epoch": 1.4315589353612168,
+      "grad_norm": 0.739686315717149,
+      "learning_rate": 4.537836730093077e-06,
+      "loss": 0.3119,
+      "step": 753
+    },
+    {
+      "epoch": 1.4334600760456273,
+      "grad_norm": 0.7341206935674344,
+      "learning_rate": 4.510049388190518e-06,
+      "loss": 0.3011,
+      "step": 754
+    },
+    {
+      "epoch": 1.435361216730038,
+      "grad_norm": 0.678774005206085,
+      "learning_rate": 4.482322592295541e-06,
+      "loss": 0.2889,
+      "step": 755
+    },
+    {
+      "epoch": 1.4372623574144487,
+      "grad_norm": 0.6790931545633795,
+      "learning_rate": 4.454656648193559e-06,
+      "loss": 0.276,
+      "step": 756
+    },
+    {
+      "epoch": 1.4391634980988592,
+      "grad_norm": 0.6544087813444031,
+      "learning_rate": 4.427051860998877e-06,
+      "loss": 0.2953,
+      "step": 757
+    },
+    {
+      "epoch": 1.44106463878327,
+      "grad_norm": 0.6979819878136554,
+      "learning_rate": 4.399508535151321e-06,
+      "loss": 0.2879,
+      "step": 758
+    },
+    {
+      "epoch": 1.4429657794676807,
+      "grad_norm": 0.6704041962251459,
+      "learning_rate": 4.372026974412907e-06,
+      "loss": 0.2869,
+      "step": 759
+    },
+    {
+      "epoch": 1.4448669201520912,
+      "grad_norm": 0.701340267320709,
+      "learning_rate": 4.344607481864466e-06,
+      "loss": 0.2899,
+      "step": 760
+    },
+    {
+      "epoch": 1.446768060836502,
+      "grad_norm": 0.6816788646971017,
+      "learning_rate": 4.317250359902295e-06,
+      "loss": 0.2939,
+      "step": 761
+    },
+    {
+      "epoch": 1.4486692015209126,
+      "grad_norm": 0.6864627615760209,
+      "learning_rate": 4.2899559102348585e-06,
+      "loss": 0.3071,
+      "step": 762
+    },
+    {
+      "epoch": 1.450570342205323,
+      "grad_norm": 0.6845593995667809,
+      "learning_rate": 4.262724433879427e-06,
+      "loss": 0.3025,
+      "step": 763
+    },
+    {
+      "epoch": 1.4524714828897338,
+      "grad_norm": 0.7157573868901977,
+      "learning_rate": 4.235556231158765e-06,
+      "loss": 0.3039,
+      "step": 764
+    },
+    {
+      "epoch": 1.4543726235741445,
+      "grad_norm": 0.6729161150145657,
+      "learning_rate": 4.208451601697836e-06,
+      "loss": 0.2905,
+      "step": 765
+    },
+    {
+      "epoch": 1.456273764258555,
+      "grad_norm": 0.668037211038967,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.3008,
+      "step": 766
+    },
+    {
+      "epoch": 1.4581749049429658,
+      "grad_norm": 0.6995806036205511,
+      "learning_rate": 4.154434257546095e-06,
+      "loss": 0.2852,
+      "step": 767
+    },
+    {
+      "epoch": 1.4600760456273765,
+      "grad_norm": 0.6565165473905802,
+      "learning_rate": 4.127522138586424e-06,
+      "loss": 0.2945,
+      "step": 768
+    },
+    {
+      "epoch": 1.461977186311787,
+      "grad_norm": 0.6988360073249953,
+      "learning_rate": 4.10067478434219e-06,
+      "loss": 0.2937,
+      "step": 769
+    },
+    {
+      "epoch": 1.4638783269961977,
+      "grad_norm": 0.6562873363126441,
+      "learning_rate": 4.073892490899865e-06,
+      "loss": 0.2881,
+      "step": 770
+    },
+    {
+      "epoch": 1.4657794676806084,
+      "grad_norm": 0.6802214616906704,
+      "learning_rate": 4.047175553628397e-06,
+      "loss": 0.2918,
+      "step": 771
+    },
+    {
+      "epoch": 1.467680608365019,
+      "grad_norm": 0.7248271078964494,
+      "learning_rate": 4.020524267175954e-06,
+      "loss": 0.3003,
+      "step": 772
+    },
+    {
+      "epoch": 1.4695817490494296,
+      "grad_norm": 0.6760399191164651,
+      "learning_rate": 3.993938925466674e-06,
+      "loss": 0.2919,
+      "step": 773
+    },
+    {
+      "epoch": 1.4714828897338403,
+      "grad_norm": 0.6820832267114996,
+      "learning_rate": 3.96741982169742e-06,
+      "loss": 0.2922,
+      "step": 774
+    },
+    {
+      "epoch": 1.4733840304182508,
+      "grad_norm": 0.705860196243192,
+      "learning_rate": 3.9409672483345465e-06,
+      "loss": 0.2934,
+      "step": 775
+    },
+    {
+      "epoch": 1.4752851711026616,
+      "grad_norm": 0.6958197432943308,
+      "learning_rate": 3.914581497110684e-06,
+      "loss": 0.2846,
+      "step": 776
+    },
+    {
+      "epoch": 1.4771863117870723,
+      "grad_norm": 0.7264244216271954,
+      "learning_rate": 3.888262859021508e-06,
+      "loss": 0.2883,
+      "step": 777
+    },
+    {
+      "epoch": 1.4790874524714828,
+      "grad_norm": 0.6844971542360613,
+      "learning_rate": 3.862011624322534e-06,
+      "loss": 0.2873,
+      "step": 778
+    },
+    {
+      "epoch": 1.4809885931558935,
+      "grad_norm": 0.7338462399547395,
+      "learning_rate": 3.835828082525925e-06,
+      "loss": 0.3098,
+      "step": 779
+    },
+    {
+      "epoch": 1.4828897338403042,
+      "grad_norm": 0.7175856031970519,
+      "learning_rate": 3.8097125223972864e-06,
+      "loss": 0.2904,
+      "step": 780
+    },
+    {
+      "epoch": 1.4847908745247147,
+      "grad_norm": 0.7363454640587259,
+      "learning_rate": 3.7836652319524835e-06,
+      "loss": 0.2941,
+      "step": 781
+    },
+    {
+      "epoch": 1.4866920152091254,
+      "grad_norm": 0.7273028394879475,
+      "learning_rate": 3.7576864984544814e-06,
+      "loss": 0.2967,
+      "step": 782
+    },
+    {
+      "epoch": 1.4885931558935361,
+      "grad_norm": 0.7707075635033304,
+      "learning_rate": 3.73177660841015e-06,
+      "loss": 0.3099,
+      "step": 783
+    },
+    {
+      "epoch": 1.4904942965779466,
+      "grad_norm": 0.7387058410165958,
+      "learning_rate": 3.7059358475671225e-06,
+      "loss": 0.2862,
+      "step": 784
+    },
+    {
+      "epoch": 1.4923954372623573,
+      "grad_norm": 0.6755462419822337,
+      "learning_rate": 3.680164500910646e-06,
+      "loss": 0.285,
+      "step": 785
+    },
+    {
+      "epoch": 1.494296577946768,
+      "grad_norm": 0.6822866461237879,
+      "learning_rate": 3.654462852660423e-06,
+      "loss": 0.2975,
+      "step": 786
+    },
+    {
+      "epoch": 1.4961977186311788,
+      "grad_norm": 0.6765268845784645,
+      "learning_rate": 3.6288311862674885e-06,
+      "loss": 0.3069,
+      "step": 787
+    },
+    {
+      "epoch": 1.4980988593155893,
+      "grad_norm": 0.7122937790553711,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 0.3091,
+      "step": 788
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.6880175477835273,
+      "learning_rate": 3.5777789289955454e-06,
+      "loss": 0.2997,
+      "step": 789
+    },
+    {
+      "epoch": 1.5019011406844105,
+      "grad_norm": 0.6348257629401532,
+      "learning_rate": 3.5523589011471592e-06,
+      "loss": 0.2769,
+      "step": 790
+    },
+    {
+      "epoch": 1.5038022813688214,
+      "grad_norm": 0.7596240780310816,
+      "learning_rate": 3.527009981211119e-06,
+      "loss": 0.3108,
+      "step": 791
+    },
+    {
+      "epoch": 1.505703422053232,
+      "grad_norm": 0.6706974669685762,
+      "learning_rate": 3.5017324487483873e-06,
+      "loss": 0.2954,
+      "step": 792
+    },
+    {
+      "epoch": 1.5076045627376424,
+      "grad_norm": 0.6683196247454588,
+      "learning_rate": 3.47652658253263e-06,
+      "loss": 0.2829,
+      "step": 793
+    },
+    {
+      "epoch": 1.5095057034220534,
+      "grad_norm": 0.6865526391071624,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.2844,
+      "step": 794
+    },
+    {
+      "epoch": 1.5114068441064639,
+      "grad_norm": 0.6750930069919843,
+      "learning_rate": 3.4263309599818017e-06,
+      "loss": 0.2938,
+      "step": 795
+    },
+    {
+      "epoch": 1.5133079847908744,
+      "grad_norm": 0.6906878894157301,
+      "learning_rate": 3.4013417572299446e-06,
+      "loss": 0.31,
+      "step": 796
+    },
+    {
+      "epoch": 1.5152091254752853,
+      "grad_norm": 0.6994865595835779,
+      "learning_rate": 3.37642532788541e-06,
+      "loss": 0.299,
+      "step": 797
+    },
+    {
+      "epoch": 1.5171102661596958,
+      "grad_norm": 0.7152957731754468,
+      "learning_rate": 3.3515819467394184e-06,
+      "loss": 0.3008,
+      "step": 798
+    },
+    {
+      "epoch": 1.5190114068441065,
+      "grad_norm": 0.6924070697595933,
+      "learning_rate": 3.326811887777607e-06,
+      "loss": 0.3011,
+      "step": 799
+    },
+    {
+      "epoch": 1.5209125475285172,
+      "grad_norm": 0.7271353830675523,
+      "learning_rate": 3.3021154241769606e-06,
+      "loss": 0.2896,
+      "step": 800
+    },
+    {
+      "epoch": 1.5228136882129277,
+      "grad_norm": 0.7239257344134434,
+      "learning_rate": 3.2774928283028153e-06,
+      "loss": 0.2957,
+      "step": 801
+    },
+    {
+      "epoch": 1.5247148288973384,
+      "grad_norm": 0.7593880502038796,
+      "learning_rate": 3.2529443717058693e-06,
+      "loss": 0.3004,
+      "step": 802
+    },
+    {
+      "epoch": 1.5266159695817492,
+      "grad_norm": 0.6692616262503324,
+      "learning_rate": 3.228470325119164e-06,
+      "loss": 0.2886,
+      "step": 803
+    },
+    {
+      "epoch": 1.5285171102661597,
+      "grad_norm": 0.6840957962380307,
+      "learning_rate": 3.20407095845511e-06,
+      "loss": 0.2897,
+      "step": 804
+    },
+    {
+      "epoch": 1.5304182509505704,
+      "grad_norm": 0.7605874500982063,
+      "learning_rate": 3.179746540802506e-06,
+      "loss": 0.3201,
+      "step": 805
+    },
+    {
+      "epoch": 1.532319391634981,
+      "grad_norm": 0.6714918792837926,
+      "learning_rate": 3.155497340423588e-06,
+      "loss": 0.2899,
+      "step": 806
+    },
+    {
+      "epoch": 1.5342205323193916,
+      "grad_norm": 0.7042603977301997,
+      "learning_rate": 3.1313236247510414e-06,
+      "loss": 0.2944,
+      "step": 807
+    },
+    {
+      "epoch": 1.5361216730038023,
+      "grad_norm": 0.7219145727524959,
+      "learning_rate": 3.107225660385077e-06,
+      "loss": 0.3089,
+      "step": 808
+    },
+    {
+      "epoch": 1.538022813688213,
+      "grad_norm": 0.681419061133769,
+      "learning_rate": 3.0832037130904748e-06,
+      "loss": 0.2945,
+      "step": 809
+    },
+    {
+      "epoch": 1.5399239543726235,
+      "grad_norm": 0.7457600216546459,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.2903,
+      "step": 810
+    },
+    {
+      "epoch": 1.5418250950570342,
+      "grad_norm": 0.788060535268472,
+      "learning_rate": 3.035388928579792e-06,
+      "loss": 0.3,
+      "step": 811
+    },
+    {
+      "epoch": 1.543726235741445,
+      "grad_norm": 0.6987516619726815,
+      "learning_rate": 3.011596618689825e-06,
+      "loss": 0.2907,
+      "step": 812
+    },
+    {
+      "epoch": 1.5456273764258555,
+      "grad_norm": 0.6757206146668999,
+      "learning_rate": 2.9878813805176252e-06,
+      "loss": 0.2849,
+      "step": 813
+    },
+    {
+      "epoch": 1.5475285171102662,
+      "grad_norm": 0.7241016707932825,
+      "learning_rate": 2.9642434756070793e-06,
+      "loss": 0.2903,
+      "step": 814
+    },
+    {
+      "epoch": 1.549429657794677,
+      "grad_norm": 0.6898554037825949,
+      "learning_rate": 2.940683164649194e-06,
+      "loss": 0.289,
+      "step": 815
+    },
+    {
+      "epoch": 1.5513307984790874,
+      "grad_norm": 0.7115316303907544,
+      "learning_rate": 2.9172007074792342e-06,
+      "loss": 0.2858,
+      "step": 816
+    },
+    {
+      "epoch": 1.553231939163498,
+      "grad_norm": 0.6664130828801642,
+      "learning_rate": 2.8937963630738517e-06,
+      "loss": 0.2955,
+      "step": 817
+    },
+    {
+      "epoch": 1.5551330798479088,
+      "grad_norm": 0.6633934717830003,
+      "learning_rate": 2.87047038954823e-06,
+      "loss": 0.2946,
+      "step": 818
+    },
+    {
+      "epoch": 1.5570342205323193,
+      "grad_norm": 0.7246399644237944,
+      "learning_rate": 2.8472230441532365e-06,
+      "loss": 0.2929,
+      "step": 819
+    },
+    {
+      "epoch": 1.55893536121673,
+      "grad_norm": 0.6725206337241619,
+      "learning_rate": 2.8240545832725963e-06,
+      "loss": 0.277,
+      "step": 820
+    },
+    {
+      "epoch": 1.5608365019011408,
+      "grad_norm": 0.7521031934202115,
+      "learning_rate": 2.8009652624200436e-06,
+      "loss": 0.296,
+      "step": 821
+    },
+    {
+      "epoch": 1.5627376425855513,
+      "grad_norm": 0.6649490791653402,
+      "learning_rate": 2.7779553362365184e-06,
+      "loss": 0.2691,
+      "step": 822
+    },
+    {
+      "epoch": 1.564638783269962,
+      "grad_norm": 0.7416418679669249,
+      "learning_rate": 2.755025058487364e-06,
+      "loss": 0.2947,
+      "step": 823
+    },
+    {
+      "epoch": 1.5665399239543727,
+      "grad_norm": 0.748717562939948,
+      "learning_rate": 2.7321746820595084e-06,
+      "loss": 0.2922,
+      "step": 824
+    },
+    {
+      "epoch": 1.5684410646387832,
+      "grad_norm": 0.7189896283757032,
+      "learning_rate": 2.709404458958693e-06,
+      "loss": 0.3,
+      "step": 825
+    },
+    {
+      "epoch": 1.570342205323194,
+      "grad_norm": 0.6778151212846256,
+      "learning_rate": 2.6867146403066833e-06,
+      "loss": 0.2795,
+      "step": 826
+    },
+    {
+      "epoch": 1.5722433460076046,
+      "grad_norm": 0.6482856460156998,
+      "learning_rate": 2.6641054763385044e-06,
+      "loss": 0.2727,
+      "step": 827
+    },
+    {
+      "epoch": 1.5741444866920151,
+      "grad_norm": 0.7178101261839865,
+      "learning_rate": 2.6415772163996845e-06,
+      "loss": 0.288,
+      "step": 828
+    },
+    {
+      "epoch": 1.5760456273764258,
+      "grad_norm": 0.7548748726915699,
+      "learning_rate": 2.619130108943494e-06,
+      "loss": 0.3015,
+      "step": 829
+    },
+    {
+      "epoch": 1.5779467680608366,
+      "grad_norm": 0.6697971227113539,
+      "learning_rate": 2.5967644015282146e-06,
+      "loss": 0.2724,
+      "step": 830
+    },
+    {
+      "epoch": 1.579847908745247,
+      "grad_norm": 0.705543895211343,
+      "learning_rate": 2.5744803408144026e-06,
+      "loss": 0.2958,
+      "step": 831
+    },
+    {
+      "epoch": 1.5817490494296578,
+      "grad_norm": 0.7466399982164742,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.3134,
+      "step": 832
+    },
+    {
+      "epoch": 1.5836501901140685,
+      "grad_norm": 0.6834710574328275,
+      "learning_rate": 2.530158141628515e-06,
+      "loss": 0.2853,
+      "step": 833
+    },
+    {
+      "epoch": 1.585551330798479,
+      "grad_norm": 0.6904210294961538,
+      "learning_rate": 2.508120491964512e-06,
+      "loss": 0.2981,
+      "step": 834
+    },
+    {
+      "epoch": 1.5874524714828897,
+      "grad_norm": 0.7134507041675866,
+      "learning_rate": 2.486165466612751e-06,
+      "loss": 0.2928,
+      "step": 835
+    },
+    {
+      "epoch": 1.5893536121673004,
+      "grad_norm": 0.7163069727636016,
+      "learning_rate": 2.464293307704566e-06,
+      "loss": 0.284,
+      "step": 836
+    },
+    {
+      "epoch": 1.591254752851711,
+      "grad_norm": 0.729799632475782,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.288,
+      "step": 837
+    },
+    {
+      "epoch": 1.5931558935361216,
+      "grad_norm": 0.6504224710355665,
+      "learning_rate": 2.4207985531722034e-06,
+      "loss": 0.2871,
+      "step": 838
+    },
+    {
+      "epoch": 1.5950570342205324,
+      "grad_norm": 0.7162572870841332,
+      "learning_rate": 2.3991764372306113e-06,
+      "loss": 0.2983,
+      "step": 839
+    },
+    {
+      "epoch": 1.5969581749049429,
+      "grad_norm": 0.6528631590867244,
+      "learning_rate": 2.377638147092497e-06,
+      "loss": 0.2767,
+      "step": 840
+    },
+    {
+      "epoch": 1.5988593155893536,
+      "grad_norm": 0.6756702527323629,
+      "learning_rate": 2.3561839202932344e-06,
+      "loss": 0.2979,
+      "step": 841
+    },
+    {
+      "epoch": 1.6007604562737643,
+      "grad_norm": 0.7026114119035859,
+      "learning_rate": 2.3348139934411008e-06,
+      "loss": 0.2913,
+      "step": 842
+    },
+    {
+      "epoch": 1.6026615969581748,
+      "grad_norm": 0.6837243131624721,
+      "learning_rate": 2.3135286022146785e-06,
+      "loss": 0.2915,
+      "step": 843
+    },
+    {
+      "epoch": 1.6045627376425855,
+      "grad_norm": 0.6369470043332507,
+      "learning_rate": 2.292327981360245e-06,
+      "loss": 0.2742,
+      "step": 844
+    },
+    {
+      "epoch": 1.6064638783269962,
+      "grad_norm": 0.679446992341736,
+      "learning_rate": 2.271212364689176e-06,
+      "loss": 0.288,
+      "step": 845
+    },
+    {
+      "epoch": 1.6083650190114067,
+      "grad_norm": 0.7419636867126453,
+      "learning_rate": 2.2501819850753925e-06,
+      "loss": 0.3037,
+      "step": 846
+    },
+    {
+      "epoch": 1.6102661596958177,
+      "grad_norm": 0.7254877844175798,
+      "learning_rate": 2.229237074452768e-06,
+      "loss": 0.2912,
+      "step": 847
+    },
+    {
+      "epoch": 1.6121673003802282,
+      "grad_norm": 0.6745766452178334,
+      "learning_rate": 2.2083778638125796e-06,
+      "loss": 0.2826,
+      "step": 848
+    },
+    {
+      "epoch": 1.6140684410646386,
+      "grad_norm": 0.7031417636695934,
+      "learning_rate": 2.1876045832009694e-06,
+      "loss": 0.2819,
+      "step": 849
+    },
+    {
+      "epoch": 1.6159695817490496,
+      "grad_norm": 0.6840837286650099,
+      "learning_rate": 2.16691746171639e-06,
+      "loss": 0.2896,
+      "step": 850
+    },
+    {
+      "epoch": 1.61787072243346,
+      "grad_norm": 0.6541199220384035,
+      "learning_rate": 2.1463167275070863e-06,
+      "loss": 0.2945,
+      "step": 851
+    },
+    {
+      "epoch": 1.6197718631178706,
+      "grad_norm": 0.7044290926674595,
+      "learning_rate": 2.125802607768588e-06,
+      "loss": 0.2912,
+      "step": 852
+    },
+    {
+      "epoch": 1.6216730038022815,
+      "grad_norm": 0.7001253202350683,
+      "learning_rate": 2.1053753287411895e-06,
+      "loss": 0.2805,
+      "step": 853
+    },
+    {
+      "epoch": 1.623574144486692,
+      "grad_norm": 0.693844774218909,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 0.2937,
+      "step": 854
+    },
+    {
+      "epoch": 1.6254752851711025,
+      "grad_norm": 0.6411951499437264,
+      "learning_rate": 2.064782192989765e-06,
+      "loss": 0.2817,
+      "step": 855
+    },
+    {
+      "epoch": 1.6273764258555135,
+      "grad_norm": 0.7222151978506833,
+      "learning_rate": 2.0446167839477815e-06,
+      "loss": 0.2852,
+      "step": 856
+    },
+    {
+      "epoch": 1.629277566539924,
+      "grad_norm": 0.6913652763684746,
+      "learning_rate": 2.0245391109760437e-06,
+      "loss": 0.2892,
+      "step": 857
+    },
+    {
+      "epoch": 1.6311787072243344,
+      "grad_norm": 0.6817511270368097,
+      "learning_rate": 2.0045493955014915e-06,
+      "loss": 0.2924,
+      "step": 858
+    },
+    {
+      "epoch": 1.6330798479087454,
+      "grad_norm": 0.6992206308391961,
+      "learning_rate": 1.984647857981017e-06,
+      "loss": 0.2928,
+      "step": 859
+    },
+    {
+      "epoch": 1.6349809885931559,
+      "grad_norm": 0.6709723525872011,
+      "learning_rate": 1.96483471789904e-06,
+      "loss": 0.2971,
+      "step": 860
+    },
+    {
+      "epoch": 1.6368821292775664,
+      "grad_norm": 0.675625645397297,
+      "learning_rate": 1.9451101937650963e-06,
+      "loss": 0.2943,
+      "step": 861
+    },
+    {
+      "epoch": 1.6387832699619773,
+      "grad_norm": 0.7004993337399331,
+      "learning_rate": 1.925474503111412e-06,
+      "loss": 0.2958,
+      "step": 862
+    },
+    {
+      "epoch": 1.6406844106463878,
+      "grad_norm": 0.6846562668268429,
+      "learning_rate": 1.905927862490512e-06,
+      "loss": 0.2993,
+      "step": 863
+    },
+    {
+      "epoch": 1.6425855513307985,
+      "grad_norm": 0.6847912208747811,
+      "learning_rate": 1.8864704874728346e-06,
+      "loss": 0.2806,
+      "step": 864
+    },
+    {
+      "epoch": 1.6444866920152093,
+      "grad_norm": 0.7469718363268735,
+      "learning_rate": 1.8671025926443464e-06,
+      "loss": 0.2929,
+      "step": 865
+    },
+    {
+      "epoch": 1.6463878326996197,
+      "grad_norm": 0.6622353022237473,
+      "learning_rate": 1.8478243916041882e-06,
+      "loss": 0.2809,
+      "step": 866
+    },
+    {
+      "epoch": 1.6482889733840305,
+      "grad_norm": 0.6726742666857457,
+      "learning_rate": 1.828636096962304e-06,
+      "loss": 0.2802,
+      "step": 867
+    },
+    {
+      "epoch": 1.6501901140684412,
+      "grad_norm": 0.6693324793206099,
+      "learning_rate": 1.8095379203371044e-06,
+      "loss": 0.2929,
+      "step": 868
+    },
+    {
+      "epoch": 1.6520912547528517,
+      "grad_norm": 0.6958850572764865,
+      "learning_rate": 1.7905300723531393e-06,
+      "loss": 0.295,
+      "step": 869
+    },
+    {
+      "epoch": 1.6539923954372624,
+      "grad_norm": 0.675344081946386,
+      "learning_rate": 1.771612762638758e-06,
+      "loss": 0.2993,
+      "step": 870
+    },
+    {
+      "epoch": 1.6558935361216731,
+      "grad_norm": 0.6799343388122288,
+      "learning_rate": 1.7527861998238094e-06,
+      "loss": 0.2915,
+      "step": 871
+    },
+    {
+      "epoch": 1.6577946768060836,
+      "grad_norm": 0.6970248770782166,
+      "learning_rate": 1.7340505915373495e-06,
+      "loss": 0.2776,
+      "step": 872
+    },
+    {
+      "epoch": 1.6596958174904943,
+      "grad_norm": 0.6869407111074151,
+      "learning_rate": 1.7154061444053239e-06,
+      "loss": 0.2961,
+      "step": 873
+    },
+    {
+      "epoch": 1.661596958174905,
+      "grad_norm": 0.6739278465492032,
+      "learning_rate": 1.6968530640483126e-06,
+      "loss": 0.2826,
+      "step": 874
+    },
+    {
+      "epoch": 1.6634980988593155,
+      "grad_norm": 0.6627832667904867,
+      "learning_rate": 1.6783915550792652e-06,
+      "loss": 0.2768,
+      "step": 875
+    },
+    {
+      "epoch": 1.6653992395437263,
+      "grad_norm": 0.672471061873299,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.2754,
+      "step": 876
+    },
+    {
+      "epoch": 1.667300380228137,
+      "grad_norm": 0.6323725629430416,
+      "learning_rate": 1.6417440647050853e-06,
+      "loss": 0.2717,
+      "step": 877
+    },
+    {
+      "epoch": 1.6692015209125475,
+      "grad_norm": 0.6687314839508512,
+      "learning_rate": 1.6235584874673848e-06,
+      "loss": 0.2886,
+      "step": 878
+    },
+    {
+      "epoch": 1.6711026615969582,
+      "grad_norm": 0.6789022682971786,
+      "learning_rate": 1.6054652899480472e-06,
+      "loss": 0.2785,
+      "step": 879
+    },
+    {
+      "epoch": 1.673003802281369,
+      "grad_norm": 0.6522004011838884,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2852,
+      "step": 880
+    },
+    {
+      "epoch": 1.6749049429657794,
+      "grad_norm": 0.6990396993019006,
+      "learning_rate": 1.5695568312079156e-06,
+      "loss": 0.2858,
+      "step": 881
+    },
+    {
+      "epoch": 1.6768060836501901,
+      "grad_norm": 0.6788115285368762,
+      "learning_rate": 1.5517419660041277e-06,
+      "loss": 0.2805,
+      "step": 882
+    },
+    {
+      "epoch": 1.6787072243346008,
+      "grad_norm": 0.7059493806598247,
+      "learning_rate": 1.534020272548349e-06,
+      "loss": 0.2919,
+      "step": 883
+    },
+    {
+      "epoch": 1.6806083650190113,
+      "grad_norm": 0.729505603314673,
+      "learning_rate": 1.5163919462845622e-06,
+      "loss": 0.2997,
+      "step": 884
+    },
+    {
+      "epoch": 1.682509505703422,
+      "grad_norm": 0.7061297241950061,
+      "learning_rate": 1.4988571816270402e-06,
+      "loss": 0.3094,
+      "step": 885
+    },
+    {
+      "epoch": 1.6844106463878328,
+      "grad_norm": 0.7158955694146114,
+      "learning_rate": 1.4814161719582132e-06,
+      "loss": 0.275,
+      "step": 886
+    },
+    {
+      "epoch": 1.6863117870722433,
+      "grad_norm": 0.6923077993490927,
+      "learning_rate": 1.4640691096265358e-06,
+      "loss": 0.2992,
+      "step": 887
+    },
+    {
+      "epoch": 1.688212927756654,
+      "grad_norm": 0.6757717346433129,
+      "learning_rate": 1.4468161859443609e-06,
+      "loss": 0.2992,
+      "step": 888
+    },
+    {
+      "epoch": 1.6901140684410647,
+      "grad_norm": 0.6502118172310717,
+      "learning_rate": 1.4296575911858268e-06,
+      "loss": 0.2775,
+      "step": 889
+    },
+    {
+      "epoch": 1.6920152091254752,
+      "grad_norm": 0.7212326057206864,
+      "learning_rate": 1.412593514584777e-06,
+      "loss": 0.2914,
+      "step": 890
+    },
+    {
+      "epoch": 1.693916349809886,
+      "grad_norm": 0.717814982268225,
+      "learning_rate": 1.3956241443326423e-06,
+      "loss": 0.2834,
+      "step": 891
+    },
+    {
+      "epoch": 1.6958174904942966,
+      "grad_norm": 0.6931153598256459,
+      "learning_rate": 1.378749667576399e-06,
+      "loss": 0.3052,
+      "step": 892
+    },
+    {
+      "epoch": 1.6977186311787071,
+      "grad_norm": 0.675812444258002,
+      "learning_rate": 1.3619702704164783e-06,
+      "loss": 0.2985,
+      "step": 893
+    },
+    {
+      "epoch": 1.6996197718631179,
+      "grad_norm": 0.7132124946133126,
+      "learning_rate": 1.3452861379047289e-06,
+      "loss": 0.2948,
+      "step": 894
+    },
+    {
+      "epoch": 1.7015209125475286,
+      "grad_norm": 0.6722753001803864,
+      "learning_rate": 1.3286974540423747e-06,
+      "loss": 0.2933,
+      "step": 895
+    },
+    {
+      "epoch": 1.703422053231939,
+      "grad_norm": 0.7033374763702748,
+      "learning_rate": 1.3122044017779768e-06,
+      "loss": 0.3063,
+      "step": 896
+    },
+    {
+      "epoch": 1.7053231939163498,
+      "grad_norm": 0.6621570193052906,
+      "learning_rate": 1.2958071630054214e-06,
+      "loss": 0.2951,
+      "step": 897
+    },
+    {
+      "epoch": 1.7072243346007605,
+      "grad_norm": 0.6684466391239013,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 0.3014,
+      "step": 898
+    },
+    {
+      "epoch": 1.709125475285171,
+      "grad_norm": 0.698684262422898,
+      "learning_rate": 1.2633008482260146e-06,
+      "loss": 0.3014,
+      "step": 899
+    },
+    {
+      "epoch": 1.7110266159695817,
+      "grad_norm": 0.6821766878872,
+      "learning_rate": 1.2471921307155655e-06,
+      "loss": 0.2816,
+      "step": 900
+    },
+    {
+      "epoch": 1.7129277566539924,
+      "grad_norm": 0.7144314228372851,
+      "learning_rate": 1.2311799436858275e-06,
+      "loss": 0.3029,
+      "step": 901
+    },
+    {
+      "epoch": 1.714828897338403,
+      "grad_norm": 0.6812939383131124,
+      "learning_rate": 1.2152644637274603e-06,
+      "loss": 0.2942,
+      "step": 902
+    },
+    {
+      "epoch": 1.7167300380228137,
+      "grad_norm": 0.7181420014741079,
+      "learning_rate": 1.1994458663645836e-06,
+      "loss": 0.3054,
+      "step": 903
+    },
+    {
+      "epoch": 1.7186311787072244,
+      "grad_norm": 0.7125081890078074,
+      "learning_rate": 1.1837243260528542e-06,
+      "loss": 0.2868,
+      "step": 904
+    },
+    {
+      "epoch": 1.7205323193916349,
+      "grad_norm": 0.679870957560119,
+      "learning_rate": 1.168100016177528e-06,
+      "loss": 0.2829,
+      "step": 905
+    },
+    {
+      "epoch": 1.7224334600760456,
+      "grad_norm": 0.6888510167626353,
+      "learning_rate": 1.1525731090515536e-06,
+      "loss": 0.2784,
+      "step": 906
+    },
+    {
+      "epoch": 1.7243346007604563,
+      "grad_norm": 0.6825513781242533,
+      "learning_rate": 1.137143775913675e-06,
+      "loss": 0.2895,
+      "step": 907
+    },
+    {
+      "epoch": 1.7262357414448668,
+      "grad_norm": 0.6848285002276066,
+      "learning_rate": 1.1218121869265365e-06,
+      "loss": 0.288,
+      "step": 908
+    },
+    {
+      "epoch": 1.7281368821292775,
+      "grad_norm": 0.6486000726132766,
+      "learning_rate": 1.1065785111748117e-06,
+      "loss": 0.2645,
+      "step": 909
+    },
+    {
+      "epoch": 1.7300380228136882,
+      "grad_norm": 0.6840951182507553,
+      "learning_rate": 1.0914429166633355e-06,
+      "loss": 0.2856,
+      "step": 910
+    },
+    {
+      "epoch": 1.7319391634980987,
+      "grad_norm": 0.7153472336382996,
+      "learning_rate": 1.076405570315252e-06,
+      "loss": 0.2951,
+      "step": 911
+    },
+    {
+      "epoch": 1.7338403041825095,
+      "grad_norm": 0.7283218855744864,
+      "learning_rate": 1.0614666379701732e-06,
+      "loss": 0.2836,
+      "step": 912
+    },
+    {
+      "epoch": 1.7357414448669202,
+      "grad_norm": 0.705903557606627,
+      "learning_rate": 1.046626284382356e-06,
+      "loss": 0.2858,
+      "step": 913
+    },
+    {
+      "epoch": 1.7376425855513307,
+      "grad_norm": 0.7254500135586339,
+      "learning_rate": 1.0318846732188737e-06,
+      "loss": 0.2879,
+      "step": 914
+    },
+    {
+      "epoch": 1.7395437262357416,
+      "grad_norm": 0.69108479862469,
+      "learning_rate": 1.017241967057816e-06,
+      "loss": 0.2971,
+      "step": 915
+    },
+    {
+      "epoch": 1.741444866920152,
+      "grad_norm": 0.6980084362756945,
+      "learning_rate": 1.0026983273865055e-06,
+      "loss": 0.2897,
+      "step": 916
+    },
+    {
+      "epoch": 1.7433460076045626,
+      "grad_norm": 0.6478640429907006,
+      "learning_rate": 9.882539145997027e-07,
+      "loss": 0.2866,
+      "step": 917
+    },
+    {
+      "epoch": 1.7452471482889735,
+      "grad_norm": 0.6632357501598067,
+      "learning_rate": 9.739088879978409e-07,
+      "loss": 0.277,
+      "step": 918
+    },
+    {
+      "epoch": 1.747148288973384,
+      "grad_norm": 0.6835575170452318,
+      "learning_rate": 9.59663405785277e-07,
+      "loss": 0.2804,
+      "step": 919
+    },
+    {
+      "epoch": 1.7490494296577945,
+      "grad_norm": 0.7190067228379501,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 0.2984,
+      "step": 920
+    },
+    {
+      "epoch": 1.7509505703422055,
+      "grad_norm": 0.6826836040756071,
+      "learning_rate": 9.314717018545838e-07,
+      "loss": 0.2773,
+      "step": 921
+    },
+    {
+      "epoch": 1.752851711026616,
+      "grad_norm": 0.7168449504045971,
+      "learning_rate": 9.17525791049112e-07,
+      "loss": 0.2826,
+      "step": 922
+    },
+    {
+      "epoch": 1.7547528517110265,
+      "grad_norm": 0.718823129658248,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.2827,
+      "step": 923
+    },
+    {
+      "epoch": 1.7566539923954374,
+      "grad_norm": 0.738208768209008,
+      "learning_rate": 8.899346207697135e-07,
+      "loss": 0.2863,
+      "step": 924
+    },
+    {
+      "epoch": 1.758555133079848,
+      "grad_norm": 0.7235904635704915,
+      "learning_rate": 8.762896655854481e-07,
+      "loss": 0.316,
+      "step": 925
+    },
+    {
+      "epoch": 1.7604562737642584,
+      "grad_norm": 0.6949700608571605,
+      "learning_rate": 8.627453313856249e-07,
+      "loss": 0.288,
+      "step": 926
+    },
+    {
+      "epoch": 1.7623574144486693,
+      "grad_norm": 0.6850281941736235,
+      "learning_rate": 8.493017675441495e-07,
+      "loss": 0.2826,
+      "step": 927
+    },
+    {
+      "epoch": 1.7642585551330798,
+      "grad_norm": 0.7404300976673953,
+      "learning_rate": 8.359591223235785e-07,
+      "loss": 0.3063,
+      "step": 928
+    },
+    {
+      "epoch": 1.7661596958174905,
+      "grad_norm": 0.7098710465061595,
+      "learning_rate": 8.227175428734868e-07,
+      "loss": 0.3008,
+      "step": 929
+    },
+    {
+      "epoch": 1.7680608365019013,
+      "grad_norm": 0.663076881926223,
+      "learning_rate": 8.095771752288451e-07,
+      "loss": 0.2825,
+      "step": 930
+    },
+    {
+      "epoch": 1.7699619771863118,
+      "grad_norm": 0.6669534974032036,
+      "learning_rate": 7.965381643084069e-07,
+      "loss": 0.2767,
+      "step": 931
+    },
+    {
+      "epoch": 1.7718631178707225,
+      "grad_norm": 0.6952805310127537,
+      "learning_rate": 7.83600653913108e-07,
+      "loss": 0.2854,
+      "step": 932
+    },
+    {
+      "epoch": 1.7737642585551332,
+      "grad_norm": 0.6307524932955819,
+      "learning_rate": 7.707647867244927e-07,
+      "loss": 0.2701,
+      "step": 933
+    },
+    {
+      "epoch": 1.7756653992395437,
+      "grad_norm": 0.7236358928829301,
+      "learning_rate": 7.580307043031232e-07,
+      "loss": 0.2958,
+      "step": 934
+    },
+    {
+      "epoch": 1.7775665399239544,
+      "grad_norm": 0.7262100345608755,
+      "learning_rate": 7.453985470870284e-07,
+      "loss": 0.2971,
+      "step": 935
+    },
+    {
+      "epoch": 1.7794676806083651,
+      "grad_norm": 0.6925182741390906,
+      "learning_rate": 7.328684543901598e-07,
+      "loss": 0.2937,
+      "step": 936
+    },
+    {
+      "epoch": 1.7813688212927756,
+      "grad_norm": 0.6517802018489877,
+      "learning_rate": 7.204405644008416e-07,
+      "loss": 0.289,
+      "step": 937
+    },
+    {
+      "epoch": 1.7832699619771863,
+      "grad_norm": 0.6888567826520788,
+      "learning_rate": 7.081150141802518e-07,
+      "loss": 0.2883,
+      "step": 938
+    },
+    {
+      "epoch": 1.785171102661597,
+      "grad_norm": 0.6660323665115999,
+      "learning_rate": 6.958919396609231e-07,
+      "loss": 0.2876,
+      "step": 939
+    },
+    {
+      "epoch": 1.7870722433460076,
+      "grad_norm": 0.6991581288383165,
+      "learning_rate": 6.837714756452241e-07,
+      "loss": 0.2861,
+      "step": 940
+    },
+    {
+      "epoch": 1.7889733840304183,
+      "grad_norm": 0.698795359283034,
+      "learning_rate": 6.717537558038845e-07,
+      "loss": 0.2837,
+      "step": 941
+    },
+    {
+      "epoch": 1.790874524714829,
+      "grad_norm": 0.6995147313549952,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 0.3046,
+      "step": 942
+    },
+    {
+      "epoch": 1.7927756653992395,
+      "grad_norm": 0.6886100459018475,
+      "learning_rate": 6.480270776601682e-07,
+      "loss": 0.2913,
+      "step": 943
+    },
+    {
+      "epoch": 1.7946768060836502,
+      "grad_norm": 0.6784034851430106,
+      "learning_rate": 6.36318381027835e-07,
+      "loss": 0.2782,
+      "step": 944
+    },
+    {
+      "epoch": 1.796577946768061,
+      "grad_norm": 0.6517285087488358,
+      "learning_rate": 6.247129519070728e-07,
+      "loss": 0.2782,
+      "step": 945
+    },
+    {
+      "epoch": 1.7984790874524714,
+      "grad_norm": 0.683977014721598,
+      "learning_rate": 6.132109182885382e-07,
+      "loss": 0.2941,
+      "step": 946
+    },
+    {
+      "epoch": 1.8003802281368821,
+      "grad_norm": 0.6539845255421082,
+      "learning_rate": 6.018124070225928e-07,
+      "loss": 0.2762,
+      "step": 947
+    },
+    {
+      "epoch": 1.8022813688212929,
+      "grad_norm": 0.6728878417139437,
+      "learning_rate": 5.905175438178979e-07,
+      "loss": 0.2992,
+      "step": 948
+    },
+    {
+      "epoch": 1.8041825095057034,
+      "grad_norm": 0.661190868288062,
+      "learning_rate": 5.793264532400311e-07,
+      "loss": 0.2929,
+      "step": 949
+    },
+    {
+      "epoch": 1.806083650190114,
+      "grad_norm": 0.667845559612317,
+      "learning_rate": 5.68239258710116e-07,
+      "loss": 0.2813,
+      "step": 950
+    },
+    {
+      "epoch": 1.8079847908745248,
+      "grad_norm": 0.6937455417455977,
+      "learning_rate": 5.572560825034523e-07,
+      "loss": 0.2926,
+      "step": 951
+    },
+    {
+      "epoch": 1.8098859315589353,
+      "grad_norm": 0.7354015367534048,
+      "learning_rate": 5.463770457481732e-07,
+      "loss": 0.2886,
+      "step": 952
+    },
+    {
+      "epoch": 1.811787072243346,
+      "grad_norm": 0.6502145664807248,
+      "learning_rate": 5.35602268423906e-07,
+      "loss": 0.2917,
+      "step": 953
+    },
+    {
+      "epoch": 1.8136882129277567,
+      "grad_norm": 0.7426612735167499,
+      "learning_rate": 5.249318693604577e-07,
+      "loss": 0.3021,
+      "step": 954
+    },
+    {
+      "epoch": 1.8155893536121672,
+      "grad_norm": 0.7412243667388426,
+      "learning_rate": 5.143659662364931e-07,
+      "loss": 0.2993,
+      "step": 955
+    },
+    {
+      "epoch": 1.817490494296578,
+      "grad_norm": 0.6803520627935851,
+      "learning_rate": 5.039046755782417e-07,
+      "loss": 0.2829,
+      "step": 956
+    },
+    {
+      "epoch": 1.8193916349809887,
+      "grad_norm": 0.6590870600210572,
+      "learning_rate": 4.935481127582131e-07,
+      "loss": 0.2676,
+      "step": 957
+    },
+    {
+      "epoch": 1.8212927756653992,
+      "grad_norm": 0.7410240006441985,
+      "learning_rate": 4.83296391993926e-07,
+      "loss": 0.3089,
+      "step": 958
+    },
+    {
+      "epoch": 1.8231939163498099,
+      "grad_norm": 0.6771559292620978,
+      "learning_rate": 4.7314962634664616e-07,
+      "loss": 0.2875,
+      "step": 959
+    },
+    {
+      "epoch": 1.8250950570342206,
+      "grad_norm": 0.6480600103714194,
+      "learning_rate": 4.631079277201389e-07,
+      "loss": 0.2744,
+      "step": 960
+    },
+    {
+      "epoch": 1.826996197718631,
+      "grad_norm": 0.6578552716511652,
+      "learning_rate": 4.5317140685943726e-07,
+      "loss": 0.2854,
+      "step": 961
+    },
+    {
+      "epoch": 1.8288973384030418,
+      "grad_norm": 0.6673271758337779,
+      "learning_rate": 4.433401733496201e-07,
+      "loss": 0.2842,
+      "step": 962
+    },
+    {
+      "epoch": 1.8307984790874525,
+      "grad_norm": 0.7100014285314042,
+      "learning_rate": 4.3361433561460274e-07,
+      "loss": 0.3039,
+      "step": 963
+    },
+    {
+      "epoch": 1.832699619771863,
+      "grad_norm": 0.6908336280614016,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 0.2793,
+      "step": 964
+    },
+    {
+      "epoch": 1.8346007604562737,
+      "grad_norm": 0.6769729562900517,
+      "learning_rate": 4.14479275351648e-07,
+      "loss": 0.2849,
+      "step": 965
+    },
+    {
+      "epoch": 1.8365019011406845,
+      "grad_norm": 0.6562578982643045,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2841,
+      "step": 966
+    },
+    {
+      "epoch": 1.838403041825095,
+      "grad_norm": 0.6692296664919734,
+      "learning_rate": 3.9576707019350903e-07,
+      "loss": 0.2945,
+      "step": 967
+    },
+    {
+      "epoch": 1.8403041825095057,
+      "grad_norm": 0.7094006805263994,
+      "learning_rate": 3.865697969675164e-07,
+      "loss": 0.2989,
+      "step": 968
+    },
+    {
+      "epoch": 1.8422053231939164,
+      "grad_norm": 0.6840714003695744,
+      "learning_rate": 3.7747854560931996e-07,
+      "loss": 0.287,
+      "step": 969
+    },
+    {
+      "epoch": 1.8441064638783269,
+      "grad_norm": 0.7004254880143974,
+      "learning_rate": 3.684934163819309e-07,
+      "loss": 0.2865,
+      "step": 970
+    },
+    {
+      "epoch": 1.8460076045627376,
+      "grad_norm": 0.6835045862244365,
+      "learning_rate": 3.596145083779912e-07,
+      "loss": 0.2851,
+      "step": 971
+    },
+    {
+      "epoch": 1.8479087452471483,
+      "grad_norm": 0.6707584333176398,
+      "learning_rate": 3.508419195186774e-07,
+      "loss": 0.2737,
+      "step": 972
+    },
+    {
+      "epoch": 1.8498098859315588,
+      "grad_norm": 0.7040951982078943,
+      "learning_rate": 3.421757465526243e-07,
+      "loss": 0.2856,
+      "step": 973
+    },
+    {
+      "epoch": 1.8517110266159695,
+      "grad_norm": 0.7080484534082977,
+      "learning_rate": 3.33616085054862e-07,
+      "loss": 0.2781,
+      "step": 974
+    },
+    {
+      "epoch": 1.8536121673003803,
+      "grad_norm": 0.6973642285563096,
+      "learning_rate": 3.2516302942574794e-07,
+      "loss": 0.2821,
+      "step": 975
+    },
+    {
+      "epoch": 1.8555133079847907,
+      "grad_norm": 0.692423714500858,
+      "learning_rate": 3.1681667288994353e-07,
+      "loss": 0.298,
+      "step": 976
+    },
+    {
+      "epoch": 1.8574144486692015,
+      "grad_norm": 0.678448689923694,
+      "learning_rate": 3.0857710749537585e-07,
+      "loss": 0.2843,
+      "step": 977
+    },
+    {
+      "epoch": 1.8593155893536122,
+      "grad_norm": 0.64263271778872,
+      "learning_rate": 3.0044442411222066e-07,
+      "loss": 0.2772,
+      "step": 978
+    },
+    {
+      "epoch": 1.8612167300380227,
+      "grad_norm": 0.6772732129193577,
+      "learning_rate": 2.9241871243190555e-07,
+      "loss": 0.29,
+      "step": 979
+    },
+    {
+      "epoch": 1.8631178707224336,
+      "grad_norm": 0.6641464050682951,
+      "learning_rate": 2.845000609661208e-07,
+      "loss": 0.2772,
+      "step": 980
+    },
+    {
+      "epoch": 1.8650190114068441,
+      "grad_norm": 0.6744772778122544,
+      "learning_rate": 2.7668855704583997e-07,
+      "loss": 0.2842,
+      "step": 981
+    },
+    {
+      "epoch": 1.8669201520912546,
+      "grad_norm": 0.6671204539831328,
+      "learning_rate": 2.689842868203563e-07,
+      "loss": 0.2795,
+      "step": 982
+    },
+    {
+      "epoch": 1.8688212927756656,
+      "grad_norm": 0.7005735235273514,
+      "learning_rate": 2.6138733525633896e-07,
+      "loss": 0.2825,
+      "step": 983
+    },
+    {
+      "epoch": 1.870722433460076,
+      "grad_norm": 0.7041142234185361,
+      "learning_rate": 2.5389778613688744e-07,
+      "loss": 0.3038,
+      "step": 984
+    },
+    {
+      "epoch": 1.8726235741444865,
+      "grad_norm": 0.6610607243551558,
+      "learning_rate": 2.46515722060614e-07,
+      "loss": 0.2875,
+      "step": 985
+    },
+    {
+      "epoch": 1.8745247148288975,
+      "grad_norm": 0.6872731425895331,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.2693,
+      "step": 986
+    },
+    {
+      "epoch": 1.876425855513308,
+      "grad_norm": 0.6890900850341375,
+      "learning_rate": 2.3207437350414418e-07,
+      "loss": 0.2852,
+      "step": 987
+    },
+    {
+      "epoch": 1.8783269961977185,
+      "grad_norm": 0.679111042440927,
+      "learning_rate": 2.2501524829059208e-07,
+      "loss": 0.2834,
+      "step": 988
+    },
+    {
+      "epoch": 1.8802281368821294,
+      "grad_norm": 0.7240616737983802,
+      "learning_rate": 2.180639266517448e-07,
+      "loss": 0.2913,
+      "step": 989
+    },
+    {
+      "epoch": 1.88212927756654,
+      "grad_norm": 0.7193328850424774,
+      "learning_rate": 2.1122048525036409e-07,
+      "loss": 0.2874,
+      "step": 990
+    },
+    {
+      "epoch": 1.8840304182509504,
+      "grad_norm": 0.677687575856997,
+      "learning_rate": 2.0448499955945223e-07,
+      "loss": 0.2851,
+      "step": 991
+    },
+    {
+      "epoch": 1.8859315589353614,
+      "grad_norm": 0.6924566450044342,
+      "learning_rate": 1.9785754386142164e-07,
+      "loss": 0.289,
+      "step": 992
+    },
+    {
+      "epoch": 1.8878326996197718,
+      "grad_norm": 0.674607345915512,
+      "learning_rate": 1.9133819124727003e-07,
+      "loss": 0.2893,
+      "step": 993
+    },
+    {
+      "epoch": 1.8897338403041823,
+      "grad_norm": 0.7209849872160335,
+      "learning_rate": 1.8492701361578326e-07,
+      "loss": 0.2932,
+      "step": 994
+    },
+    {
+      "epoch": 1.8916349809885933,
+      "grad_norm": 0.6831355382123203,
+      "learning_rate": 1.7862408167273472e-07,
+      "loss": 0.2932,
+      "step": 995
+    },
+    {
+      "epoch": 1.8935361216730038,
+      "grad_norm": 0.6953634149735767,
+      "learning_rate": 1.724294649301095e-07,
+      "loss": 0.286,
+      "step": 996
+    },
+    {
+      "epoch": 1.8954372623574145,
+      "grad_norm": 0.6913834587275323,
+      "learning_rate": 1.6634323170533928e-07,
+      "loss": 0.29,
+      "step": 997
+    },
+    {
+      "epoch": 1.8973384030418252,
+      "grad_norm": 0.6602953452308825,
+      "learning_rate": 1.6036544912054087e-07,
+      "loss": 0.2816,
+      "step": 998
+    },
+    {
+      "epoch": 1.8992395437262357,
+      "grad_norm": 0.6805301949373087,
+      "learning_rate": 1.544961831017855e-07,
+      "loss": 0.2966,
+      "step": 999
+    },
+    {
+      "epoch": 1.9011406844106464,
+      "grad_norm": 0.7019539921184856,
+      "learning_rate": 1.487354983783673e-07,
+      "loss": 0.2868,
+      "step": 1000
+    },
+    {
+      "epoch": 1.9030418250950571,
+      "grad_norm": 0.6547373408336923,
+      "learning_rate": 1.430834584820895e-07,
+      "loss": 0.2844,
+      "step": 1001
+    },
+    {
+      "epoch": 1.9049429657794676,
+      "grad_norm": 0.6854587160034771,
+      "learning_rate": 1.375401257465625e-07,
+      "loss": 0.2846,
+      "step": 1002
+    },
+    {
+      "epoch": 1.9068441064638784,
+      "grad_norm": 0.6680108541741815,
+      "learning_rate": 1.3210556130652031e-07,
+      "loss": 0.2827,
+      "step": 1003
+    },
+    {
+      "epoch": 1.908745247148289,
+      "grad_norm": 0.6753853014028878,
+      "learning_rate": 1.2677982509714415e-07,
+      "loss": 0.2782,
+      "step": 1004
+    },
+    {
+      "epoch": 1.9106463878326996,
+      "grad_norm": 0.703216680369279,
+      "learning_rate": 1.2156297585339872e-07,
+      "loss": 0.282,
+      "step": 1005
+    },
+    {
+      "epoch": 1.9125475285171103,
+      "grad_norm": 0.7221226047083721,
+      "learning_rate": 1.1645507110938925e-07,
+      "loss": 0.2919,
+      "step": 1006
+    },
+    {
+      "epoch": 1.914448669201521,
+      "grad_norm": 0.6726153834884064,
+      "learning_rate": 1.1145616719772545e-07,
+      "loss": 0.3011,
+      "step": 1007
+    },
+    {
+      "epoch": 1.9163498098859315,
+      "grad_norm": 0.6640673476821523,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.2783,
+      "step": 1008
+    },
+    {
+      "epoch": 1.9182509505703422,
+      "grad_norm": 0.6726493934897552,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2831,
+      "step": 1009
+    },
+    {
+      "epoch": 1.920152091254753,
+      "grad_norm": 0.6855639947595032,
+      "learning_rate": 9.711400574749507e-08,
+      "loss": 0.2824,
+      "step": 1010
+    },
+    {
+      "epoch": 1.9220532319391634,
+      "grad_norm": 0.6432583692832848,
+      "learning_rate": 9.255164443990994e-08,
+      "loss": 0.2867,
+      "step": 1011
+    },
+    {
+      "epoch": 1.9239543726235742,
+      "grad_norm": 0.7078077377520312,
+      "learning_rate": 8.809854758399017e-08,
+      "loss": 0.2918,
+      "step": 1012
+    },
+    {
+      "epoch": 1.9258555133079849,
+      "grad_norm": 0.660901140817257,
+      "learning_rate": 8.375476429078543e-08,
+      "loss": 0.2875,
+      "step": 1013
+    },
+    {
+      "epoch": 1.9277566539923954,
+      "grad_norm": 0.6860940210532674,
+      "learning_rate": 7.952034246577977e-08,
+      "loss": 0.2825,
+      "step": 1014
+    },
+    {
+      "epoch": 1.929657794676806,
+      "grad_norm": 0.6498169252729303,
+      "learning_rate": 7.539532880836087e-08,
+      "loss": 0.278,
+      "step": 1015
+    },
+    {
+      "epoch": 1.9315589353612168,
+      "grad_norm": 0.6357966300013042,
+      "learning_rate": 7.137976881130826e-08,
+      "loss": 0.2784,
+      "step": 1016
+    },
+    {
+      "epoch": 1.9334600760456273,
+      "grad_norm": 0.6861147896273655,
+      "learning_rate": 6.747370676028819e-08,
+      "loss": 0.2845,
+      "step": 1017
+    },
+    {
+      "epoch": 1.935361216730038,
+      "grad_norm": 0.700374051250366,
+      "learning_rate": 6.367718573336845e-08,
+      "loss": 0.2831,
+      "step": 1018
+    },
+    {
+      "epoch": 1.9372623574144487,
+      "grad_norm": 0.7249233057770381,
+      "learning_rate": 5.999024760054095e-08,
+      "loss": 0.3007,
+      "step": 1019
+    },
+    {
+      "epoch": 1.9391634980988592,
+      "grad_norm": 0.6740399420167456,
+      "learning_rate": 5.641293302326323e-08,
+      "loss": 0.2914,
+      "step": 1020
+    },
+    {
+      "epoch": 1.94106463878327,
+      "grad_norm": 0.6831861473311723,
+      "learning_rate": 5.2945281454003236e-08,
+      "loss": 0.2836,
+      "step": 1021
+    },
+    {
+      "epoch": 1.9429657794676807,
+      "grad_norm": 0.6692750270704219,
+      "learning_rate": 4.958733113581415e-08,
+      "loss": 0.29,
+      "step": 1022
+    },
+    {
+      "epoch": 1.9448669201520912,
+      "grad_norm": 0.6803361684070416,
+      "learning_rate": 4.6339119101902475e-08,
+      "loss": 0.2864,
+      "step": 1023
+    },
+    {
+      "epoch": 1.946768060836502,
+      "grad_norm": 0.6835393511708907,
+      "learning_rate": 4.320068117522835e-08,
+      "loss": 0.2931,
+      "step": 1024
+    },
+    {
+      "epoch": 1.9486692015209126,
+      "grad_norm": 0.6904115210272912,
+      "learning_rate": 4.0172051968101474e-08,
+      "loss": 0.2872,
+      "step": 1025
+    },
+    {
+      "epoch": 1.950570342205323,
+      "grad_norm": 0.6330189661558953,
+      "learning_rate": 3.7253264881809137e-08,
+      "loss": 0.2789,
+      "step": 1026
+    },
+    {
+      "epoch": 1.9524714828897338,
+      "grad_norm": 0.6652552000909566,
+      "learning_rate": 3.4444352106242086e-08,
+      "loss": 0.2949,
+      "step": 1027
+    },
+    {
+      "epoch": 1.9543726235741445,
+      "grad_norm": 0.6835903830000685,
+      "learning_rate": 3.174534461953593e-08,
+      "loss": 0.297,
+      "step": 1028
+    },
+    {
+      "epoch": 1.956273764258555,
+      "grad_norm": 0.6721191967262458,
+      "learning_rate": 2.915627218774142e-08,
+      "loss": 0.2903,
+      "step": 1029
+    },
+    {
+      "epoch": 1.9581749049429658,
+      "grad_norm": 0.7001954816115531,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.299,
+      "step": 1030
+    },
+    {
+      "epoch": 1.9600760456273765,
+      "grad_norm": 0.6588780264825448,
+      "learning_rate": 2.430804549065302e-08,
+      "loss": 0.2925,
+      "step": 1031
+    },
+    {
+      "epoch": 1.961977186311787,
+      "grad_norm": 0.6750856277947396,
+      "learning_rate": 2.2048944694104123e-08,
+      "loss": 0.2901,
+      "step": 1032
+    },
+    {
+      "epoch": 1.9638783269961977,
+      "grad_norm": 0.7044566136465064,
+      "learning_rate": 1.989988588936509e-08,
+      "loss": 0.2935,
+      "step": 1033
+    },
+    {
+      "epoch": 1.9657794676806084,
+      "grad_norm": 0.6758762388127414,
+      "learning_rate": 1.7860892777367133e-08,
+      "loss": 0.2857,
+      "step": 1034
+    },
+    {
+      "epoch": 1.967680608365019,
+      "grad_norm": 0.6423479164680296,
+      "learning_rate": 1.5931987845176912e-08,
+      "loss": 0.2716,
+      "step": 1035
+    },
+    {
+      "epoch": 1.9695817490494296,
+      "grad_norm": 0.6583540521286209,
+      "learning_rate": 1.411319236575337e-08,
+      "loss": 0.2817,
+      "step": 1036
+    },
+    {
+      "epoch": 1.9714828897338403,
+      "grad_norm": 0.7234443830067822,
+      "learning_rate": 1.2404526397711281e-08,
+      "loss": 0.3049,
+      "step": 1037
+    },
+    {
+      "epoch": 1.9733840304182508,
+      "grad_norm": 0.6687518470436962,
+      "learning_rate": 1.0806008785100297e-08,
+      "loss": 0.2705,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9752851711026616,
+      "grad_norm": 0.7844975620686251,
+      "learning_rate": 9.317657157197347e-09,
+      "loss": 0.3016,
+      "step": 1039
+    },
+    {
+      "epoch": 1.9771863117870723,
+      "grad_norm": 0.7144191818427592,
+      "learning_rate": 7.93948792831234e-09,
+      "loss": 0.2938,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9790874524714828,
+      "grad_norm": 0.6732522305643281,
+      "learning_rate": 6.671516297606095e-09,
+      "loss": 0.2865,
+      "step": 1041
+    },
+    {
+      "epoch": 1.9809885931558935,
+      "grad_norm": 0.6576157461437023,
+      "learning_rate": 5.513756248924917e-09,
+      "loss": 0.2922,
+      "step": 1042
+    },
+    {
+      "epoch": 1.9828897338403042,
+      "grad_norm": 0.6830289546875221,
+      "learning_rate": 4.466220550641831e-09,
+      "loss": 0.2801,
+      "step": 1043
+    },
+    {
+      "epoch": 1.9847908745247147,
+      "grad_norm": 0.6826786269023402,
+      "learning_rate": 3.528920755523357e-09,
+      "loss": 0.288,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9866920152091256,
+      "grad_norm": 0.6952271637512921,
+      "learning_rate": 2.701867200592956e-09,
+      "loss": 0.2829,
+      "step": 1045
+    },
+    {
+      "epoch": 1.9885931558935361,
+      "grad_norm": 0.6708828460313605,
+      "learning_rate": 1.9850690070266633e-09,
+      "loss": 0.2933,
+      "step": 1046
+    },
+    {
+      "epoch": 1.9904942965779466,
+      "grad_norm": 0.7006126020844139,
+      "learning_rate": 1.378534080042071e-09,
+      "loss": 0.2937,
+      "step": 1047
+    },
+    {
+      "epoch": 1.9923954372623576,
+      "grad_norm": 0.6810197712497644,
+      "learning_rate": 8.822691088195001e-10,
+      "loss": 0.2896,
+      "step": 1048
+    },
+    {
+      "epoch": 1.994296577946768,
+      "grad_norm": 0.6634019955658176,
+      "learning_rate": 4.962795664265052e-10,
+      "loss": 0.28,
+      "step": 1049
+    },
+    {
+      "epoch": 1.9961977186311786,
+      "grad_norm": 0.6783418657468078,
+      "learning_rate": 2.2056970975459223e-10,
+      "loss": 0.2838,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9980988593155895,
+      "grad_norm": 0.8501728865841743,
+      "learning_rate": 5.514257947369928e-11,
+      "loss": 0.3105,
+      "step": 1051
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.647174266772532,
+      "learning_rate": 0.0,
+      "loss": 0.2831,
+      "step": 1052
+    },
+    {
+      "epoch": 2.0,
+      "step": 1052,
+      "total_flos": 110069417902080.0,
+      "train_loss": 0.3732619928656422,
+      "train_runtime": 3524.051,
+      "train_samples_per_second": 38.211,
+      "train_steps_per_second": 0.299
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1052,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "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": 110069417902080.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}