diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,6385 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1812,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0033112582781456954,
+      "grad_norm": 0.609375,
+      "learning_rate": 1.3186813186813187e-07,
+      "loss": 1.9402873516082764,
+      "step": 2
+    },
+    {
+      "epoch": 0.006622516556291391,
+      "grad_norm": 0.7890625,
+      "learning_rate": 3.956043956043956e-07,
+      "loss": 1.9065133333206177,
+      "step": 4
+    },
+    {
+      "epoch": 0.009933774834437087,
+      "grad_norm": 0.984375,
+      "learning_rate": 6.593406593406594e-07,
+      "loss": 2.009326696395874,
+      "step": 6
+    },
+    {
+      "epoch": 0.013245033112582781,
+      "grad_norm": 0.69921875,
+      "learning_rate": 9.230769230769231e-07,
+      "loss": 2.020029306411743,
+      "step": 8
+    },
+    {
+      "epoch": 0.016556291390728478,
+      "grad_norm": 0.98828125,
+      "learning_rate": 1.1868131868131868e-06,
+      "loss": 2.017151355743408,
+      "step": 10
+    },
+    {
+      "epoch": 0.019867549668874173,
+      "grad_norm": 0.5859375,
+      "learning_rate": 1.4505494505494505e-06,
+      "loss": 1.8570733070373535,
+      "step": 12
+    },
+    {
+      "epoch": 0.023178807947019868,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.7142857142857143e-06,
+      "loss": 1.8826595544815063,
+      "step": 14
+    },
+    {
+      "epoch": 0.026490066225165563,
+      "grad_norm": 0.97265625,
+      "learning_rate": 1.9780219780219782e-06,
+      "loss": 1.8526196479797363,
+      "step": 16
+    },
+    {
+      "epoch": 0.029801324503311258,
+      "grad_norm": 0.59765625,
+      "learning_rate": 2.2417582417582418e-06,
+      "loss": 1.8230018615722656,
+      "step": 18
+    },
+    {
+      "epoch": 0.033112582781456956,
+      "grad_norm": 0.7578125,
+      "learning_rate": 2.5054945054945057e-06,
+      "loss": 2.0143635272979736,
+      "step": 20
+    },
+    {
+      "epoch": 0.03642384105960265,
+      "grad_norm": 0.6015625,
+      "learning_rate": 2.7692307692307693e-06,
+      "loss": 1.9134697914123535,
+      "step": 22
+    },
+    {
+      "epoch": 0.039735099337748346,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.032967032967033e-06,
+      "loss": 1.945981740951538,
+      "step": 24
+    },
+    {
+      "epoch": 0.04304635761589404,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.296703296703297e-06,
+      "loss": 1.8832194805145264,
+      "step": 26
+    },
+    {
+      "epoch": 0.046357615894039736,
+      "grad_norm": 0.5703125,
+      "learning_rate": 3.5604395604395607e-06,
+      "loss": 1.7985222339630127,
+      "step": 28
+    },
+    {
+      "epoch": 0.04966887417218543,
+      "grad_norm": 0.56640625,
+      "learning_rate": 3.824175824175824e-06,
+      "loss": 1.8062254190444946,
+      "step": 30
+    },
+    {
+      "epoch": 0.052980132450331126,
+      "grad_norm": 4.5,
+      "learning_rate": 4.087912087912088e-06,
+      "loss": 1.8083665370941162,
+      "step": 32
+    },
+    {
+      "epoch": 0.056291390728476824,
+      "grad_norm": 0.8046875,
+      "learning_rate": 4.351648351648352e-06,
+      "loss": 1.9273862838745117,
+      "step": 34
+    },
+    {
+      "epoch": 0.059602649006622516,
+      "grad_norm": 1.046875,
+      "learning_rate": 4.615384615384616e-06,
+      "loss": 1.8486627340316772,
+      "step": 36
+    },
+    {
+      "epoch": 0.06291390728476821,
+      "grad_norm": 0.65625,
+      "learning_rate": 4.879120879120879e-06,
+      "loss": 1.798123836517334,
+      "step": 38
+    },
+    {
+      "epoch": 0.06622516556291391,
+      "grad_norm": 0.53125,
+      "learning_rate": 5.142857142857142e-06,
+      "loss": 1.829748272895813,
+      "step": 40
+    },
+    {
+      "epoch": 0.0695364238410596,
+      "grad_norm": 0.62109375,
+      "learning_rate": 5.406593406593407e-06,
+      "loss": 1.7815451622009277,
+      "step": 42
+    },
+    {
+      "epoch": 0.0728476821192053,
+      "grad_norm": 0.484375,
+      "learning_rate": 5.67032967032967e-06,
+      "loss": 1.847506046295166,
+      "step": 44
+    },
+    {
+      "epoch": 0.076158940397351,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.934065934065935e-06,
+      "loss": 1.8606122732162476,
+      "step": 46
+    },
+    {
+      "epoch": 0.07947019867549669,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.197802197802198e-06,
+      "loss": 1.9397691488265991,
+      "step": 48
+    },
+    {
+      "epoch": 0.08278145695364239,
+      "grad_norm": 0.8359375,
+      "learning_rate": 6.461538461538462e-06,
+      "loss": 1.9331084489822388,
+      "step": 50
+    },
+    {
+      "epoch": 0.08609271523178808,
+      "grad_norm": 0.56640625,
+      "learning_rate": 6.725274725274725e-06,
+      "loss": 1.732893943786621,
+      "step": 52
+    },
+    {
+      "epoch": 0.08940397350993377,
+      "grad_norm": 0.419921875,
+      "learning_rate": 6.98901098901099e-06,
+      "loss": 1.8520346879959106,
+      "step": 54
+    },
+    {
+      "epoch": 0.09271523178807947,
+      "grad_norm": 0.36328125,
+      "learning_rate": 7.252747252747252e-06,
+      "loss": 1.7379193305969238,
+      "step": 56
+    },
+    {
+      "epoch": 0.09602649006622517,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.516483516483517e-06,
+      "loss": 1.6829403638839722,
+      "step": 58
+    },
+    {
+      "epoch": 0.09933774834437085,
+      "grad_norm": 0.384765625,
+      "learning_rate": 7.78021978021978e-06,
+      "loss": 1.695286512374878,
+      "step": 60
+    },
+    {
+      "epoch": 0.10264900662251655,
+      "grad_norm": 0.41796875,
+      "learning_rate": 8.043956043956044e-06,
+      "loss": 1.7086611986160278,
+      "step": 62
+    },
+    {
+      "epoch": 0.10596026490066225,
+      "grad_norm": 0.373046875,
+      "learning_rate": 8.307692307692307e-06,
+      "loss": 1.696899175643921,
+      "step": 64
+    },
+    {
+      "epoch": 0.10927152317880795,
+      "grad_norm": 0.5,
+      "learning_rate": 8.571428571428573e-06,
+      "loss": 1.7970926761627197,
+      "step": 66
+    },
+    {
+      "epoch": 0.11258278145695365,
+      "grad_norm": 11.25,
+      "learning_rate": 8.835164835164836e-06,
+      "loss": 1.7659003734588623,
+      "step": 68
+    },
+    {
+      "epoch": 0.11589403973509933,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.098901098901098e-06,
+      "loss": 1.73307204246521,
+      "step": 70
+    },
+    {
+      "epoch": 0.11920529801324503,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.362637362637363e-06,
+      "loss": 1.716630220413208,
+      "step": 72
+    },
+    {
+      "epoch": 0.12251655629139073,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.626373626373627e-06,
+      "loss": 1.699965000152588,
+      "step": 74
+    },
+    {
+      "epoch": 0.12582781456953643,
+      "grad_norm": 0.73046875,
+      "learning_rate": 9.89010989010989e-06,
+      "loss": 1.7758498191833496,
+      "step": 76
+    },
+    {
+      "epoch": 0.1291390728476821,
+      "grad_norm": 0.421875,
+      "learning_rate": 1.0153846153846154e-05,
+      "loss": 1.640932559967041,
+      "step": 78
+    },
+    {
+      "epoch": 0.13245033112582782,
+      "grad_norm": 2.109375,
+      "learning_rate": 1.0417582417582417e-05,
+      "loss": 1.6744168996810913,
+      "step": 80
+    },
+    {
+      "epoch": 0.1357615894039735,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.0681318681318681e-05,
+      "loss": 1.6837286949157715,
+      "step": 82
+    },
+    {
+      "epoch": 0.1390728476821192,
+      "grad_norm": 0.8046875,
+      "learning_rate": 1.0945054945054944e-05,
+      "loss": 1.5954879522323608,
+      "step": 84
+    },
+    {
+      "epoch": 0.1423841059602649,
+      "grad_norm": 0.6796875,
+      "learning_rate": 1.120879120879121e-05,
+      "loss": 1.6993297338485718,
+      "step": 86
+    },
+    {
+      "epoch": 0.1456953642384106,
+      "grad_norm": 0.42578125,
+      "learning_rate": 1.1472527472527473e-05,
+      "loss": 1.50092351436615,
+      "step": 88
+    },
+    {
+      "epoch": 0.1490066225165563,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.1736263736263737e-05,
+      "loss": 1.4782135486602783,
+      "step": 90
+    },
+    {
+      "epoch": 0.152317880794702,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.2e-05,
+      "loss": 1.655004858970642,
+      "step": 92
+    },
+    {
+      "epoch": 0.15562913907284767,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.1999968010404002e-05,
+      "loss": 1.4930201768875122,
+      "step": 94
+    },
+    {
+      "epoch": 0.15894039735099338,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.1999872042042387e-05,
+      "loss": 1.5215712785720825,
+      "step": 96
+    },
+    {
+      "epoch": 0.16225165562913907,
+      "grad_norm": 0.71875,
+      "learning_rate": 1.1999712096194322e-05,
+      "loss": 1.5142920017242432,
+      "step": 98
+    },
+    {
+      "epoch": 0.16556291390728478,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.1999488174991726e-05,
+      "loss": 1.5328994989395142,
+      "step": 100
+    },
+    {
+      "epoch": 0.16887417218543047,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1999200281419243e-05,
+      "loss": 1.573561429977417,
+      "step": 102
+    },
+    {
+      "epoch": 0.17218543046357615,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1998848419314207e-05,
+      "loss": 1.3963382244110107,
+      "step": 104
+    },
+    {
+      "epoch": 0.17549668874172186,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.1998432593366584e-05,
+      "loss": 1.4885351657867432,
+      "step": 106
+    },
+    {
+      "epoch": 0.17880794701986755,
+      "grad_norm": 0.7109375,
+      "learning_rate": 1.199795280911892e-05,
+      "loss": 1.509068489074707,
+      "step": 108
+    },
+    {
+      "epoch": 0.18211920529801323,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.199740907296626e-05,
+      "loss": 1.3925909996032715,
+      "step": 110
+    },
+    {
+      "epoch": 0.18543046357615894,
+      "grad_norm": 0.61328125,
+      "learning_rate": 1.1996801392156057e-05,
+      "loss": 1.4695332050323486,
+      "step": 112
+    },
+    {
+      "epoch": 0.18874172185430463,
+      "grad_norm": 0.470703125,
+      "learning_rate": 1.1996129774788096e-05,
+      "loss": 1.4908018112182617,
+      "step": 114
+    },
+    {
+      "epoch": 0.19205298013245034,
+      "grad_norm": 0.314453125,
+      "learning_rate": 1.1995394229814352e-05,
+      "loss": 1.3704346418380737,
+      "step": 116
+    },
+    {
+      "epoch": 0.19536423841059603,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1994594767038916e-05,
+      "loss": 1.5010247230529785,
+      "step": 118
+    },
+    {
+      "epoch": 0.1986754966887417,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.1993731397117814e-05,
+      "loss": 1.406250238418579,
+      "step": 120
+    },
+    {
+      "epoch": 0.20198675496688742,
+      "grad_norm": 0.84765625,
+      "learning_rate": 1.1992804131558912e-05,
+      "loss": 1.3863422870635986,
+      "step": 122
+    },
+    {
+      "epoch": 0.2052980132450331,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.1991812982721724e-05,
+      "loss": 1.2978781461715698,
+      "step": 124
+    },
+    {
+      "epoch": 0.20860927152317882,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.1990757963817274e-05,
+      "loss": 1.4624699354171753,
+      "step": 126
+    },
+    {
+      "epoch": 0.2119205298013245,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.198963908890791e-05,
+      "loss": 1.436671495437622,
+      "step": 128
+    },
+    {
+      "epoch": 0.2152317880794702,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1988456372907107e-05,
+      "loss": 1.4213171005249023,
+      "step": 130
+    },
+    {
+      "epoch": 0.2185430463576159,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.1987209831579291e-05,
+      "loss": 1.3802193403244019,
+      "step": 132
+    },
+    {
+      "epoch": 0.22185430463576158,
+      "grad_norm": 3.734375,
+      "learning_rate": 1.1985899481539603e-05,
+      "loss": 1.4687957763671875,
+      "step": 134
+    },
+    {
+      "epoch": 0.2251655629139073,
+      "grad_norm": 0.275390625,
+      "learning_rate": 1.1984525340253706e-05,
+      "loss": 1.3982669115066528,
+      "step": 136
+    },
+    {
+      "epoch": 0.22847682119205298,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.1983087426037516e-05,
+      "loss": 1.4475197792053223,
+      "step": 138
+    },
+    {
+      "epoch": 0.23178807947019867,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.1981585758057e-05,
+      "loss": 1.3907548189163208,
+      "step": 140
+    },
+    {
+      "epoch": 0.23509933774834438,
+      "grad_norm": 0.28125,
+      "learning_rate": 1.1980020356327879e-05,
+      "loss": 1.4539854526519775,
+      "step": 142
+    },
+    {
+      "epoch": 0.23841059602649006,
+      "grad_norm": 0.8203125,
+      "learning_rate": 1.1978391241715392e-05,
+      "loss": 1.3030190467834473,
+      "step": 144
+    },
+    {
+      "epoch": 0.24172185430463577,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.1976698435934008e-05,
+      "loss": 1.4244205951690674,
+      "step": 146
+    },
+    {
+      "epoch": 0.24503311258278146,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.1974941961547131e-05,
+      "loss": 1.4479776620864868,
+      "step": 148
+    },
+    {
+      "epoch": 0.24834437086092714,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.1973121841966806e-05,
+      "loss": 1.4405760765075684,
+      "step": 150
+    },
+    {
+      "epoch": 0.25165562913907286,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.19712381014534e-05,
+      "loss": 1.3684296607971191,
+      "step": 152
+    },
+    {
+      "epoch": 0.25496688741721857,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.1969290765115295e-05,
+      "loss": 1.3721959590911865,
+      "step": 154
+    },
+    {
+      "epoch": 0.2582781456953642,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.1967279858908524e-05,
+      "loss": 1.2832247018814087,
+      "step": 156
+    },
+    {
+      "epoch": 0.26158940397350994,
+      "grad_norm": 0.37890625,
+      "learning_rate": 1.196520540963646e-05,
+      "loss": 1.3485926389694214,
+      "step": 158
+    },
+    {
+      "epoch": 0.26490066225165565,
+      "grad_norm": 0.48046875,
+      "learning_rate": 1.1963067444949428e-05,
+      "loss": 1.3671433925628662,
+      "step": 160
+    },
+    {
+      "epoch": 0.2682119205298013,
+      "grad_norm": 1.34375,
+      "learning_rate": 1.1960865993344364e-05,
+      "loss": 1.3472628593444824,
+      "step": 162
+    },
+    {
+      "epoch": 0.271523178807947,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.1958601084164404e-05,
+      "loss": 1.3688595294952393,
+      "step": 164
+    },
+    {
+      "epoch": 0.27483443708609273,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.1956272747598524e-05,
+      "loss": 1.3366729021072388,
+      "step": 166
+    },
+    {
+      "epoch": 0.2781456953642384,
+      "grad_norm": 0.466796875,
+      "learning_rate": 1.1953881014681118e-05,
+      "loss": 1.2818353176116943,
+      "step": 168
+    },
+    {
+      "epoch": 0.2814569536423841,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1951425917291586e-05,
+      "loss": 1.3180421590805054,
+      "step": 170
+    },
+    {
+      "epoch": 0.2847682119205298,
+      "grad_norm": 0.478515625,
+      "learning_rate": 1.194890748815392e-05,
+      "loss": 1.4252783060073853,
+      "step": 172
+    },
+    {
+      "epoch": 0.28807947019867547,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1946325760836258e-05,
+      "loss": 1.3658868074417114,
+      "step": 174
+    },
+    {
+      "epoch": 0.2913907284768212,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1943680769750438e-05,
+      "loss": 1.3427958488464355,
+      "step": 176
+    },
+    {
+      "epoch": 0.2947019867549669,
+      "grad_norm": 0.6171875,
+      "learning_rate": 1.1940972550151545e-05,
+      "loss": 1.3762881755828857,
+      "step": 178
+    },
+    {
+      "epoch": 0.2980132450331126,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.1938201138137428e-05,
+      "loss": 1.3542152643203735,
+      "step": 180
+    },
+    {
+      "epoch": 0.30132450331125826,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.1935366570648237e-05,
+      "loss": 1.25732421875,
+      "step": 182
+    },
+    {
+      "epoch": 0.304635761589404,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1932468885465917e-05,
+      "loss": 1.4017279148101807,
+      "step": 184
+    },
+    {
+      "epoch": 0.3079470198675497,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1929508121213708e-05,
+      "loss": 1.3300182819366455,
+      "step": 186
+    },
+    {
+      "epoch": 0.31125827814569534,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.1926484317355632e-05,
+      "loss": 1.3051140308380127,
+      "step": 188
+    },
+    {
+      "epoch": 0.31456953642384106,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.1923397514195965e-05,
+      "loss": 1.349334955215454,
+      "step": 190
+    },
+    {
+      "epoch": 0.31788079470198677,
+      "grad_norm": 0.5546875,
+      "learning_rate": 1.1920247752878701e-05,
+      "loss": 1.3017675876617432,
+      "step": 192
+    },
+    {
+      "epoch": 0.3211920529801324,
+      "grad_norm": 0.40625,
+      "learning_rate": 1.1917035075387008e-05,
+      "loss": 1.325446367263794,
+      "step": 194
+    },
+    {
+      "epoch": 0.32450331125827814,
+      "grad_norm": 0.44921875,
+      "learning_rate": 1.1913759524542653e-05,
+      "loss": 1.217170238494873,
+      "step": 196
+    },
+    {
+      "epoch": 0.32781456953642385,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.191042114400545e-05,
+      "loss": 1.3321378231048584,
+      "step": 198
+    },
+    {
+      "epoch": 0.33112582781456956,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.1907019978272667e-05,
+      "loss": 1.2979496717453003,
+      "step": 200
+    },
+    {
+      "epoch": 0.3344370860927152,
+      "grad_norm": 0.859375,
+      "learning_rate": 1.190355607267844e-05,
+      "loss": 1.2881971597671509,
+      "step": 202
+    },
+    {
+      "epoch": 0.33774834437086093,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1900029473393154e-05,
+      "loss": 1.3050187826156616,
+      "step": 204
+    },
+    {
+      "epoch": 0.34105960264900664,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.1896440227422847e-05,
+      "loss": 1.2708481550216675,
+      "step": 206
+    },
+    {
+      "epoch": 0.3443708609271523,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.1892788382608577e-05,
+      "loss": 1.2817398309707642,
+      "step": 208
+    },
+    {
+      "epoch": 0.347682119205298,
+      "grad_norm": 0.26953125,
+      "learning_rate": 1.1889073987625775e-05,
+      "loss": 1.3485984802246094,
+      "step": 210
+    },
+    {
+      "epoch": 0.3509933774834437,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.1885297091983607e-05,
+      "loss": 1.3059399127960205,
+      "step": 212
+    },
+    {
+      "epoch": 0.3543046357615894,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.1881457746024304e-05,
+      "loss": 1.2323215007781982,
+      "step": 214
+    },
+    {
+      "epoch": 0.3576158940397351,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.1877556000922508e-05,
+      "loss": 1.3672224283218384,
+      "step": 216
+    },
+    {
+      "epoch": 0.3609271523178808,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1873591908684567e-05,
+      "loss": 1.4012298583984375,
+      "step": 218
+    },
+    {
+      "epoch": 0.36423841059602646,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.1869565522147865e-05,
+      "loss": 1.3226133584976196,
+      "step": 220
+    },
+    {
+      "epoch": 0.3675496688741722,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1865476894980099e-05,
+      "loss": 1.2324888706207275,
+      "step": 222
+    },
+    {
+      "epoch": 0.3708609271523179,
+      "grad_norm": 0.6953125,
+      "learning_rate": 1.1861326081678575e-05,
+      "loss": 1.3415271043777466,
+      "step": 224
+    },
+    {
+      "epoch": 0.3741721854304636,
+      "grad_norm": 0.326171875,
+      "learning_rate": 1.1857113137569471e-05,
+      "loss": 1.2043918371200562,
+      "step": 226
+    },
+    {
+      "epoch": 0.37748344370860926,
+      "grad_norm": 0.70703125,
+      "learning_rate": 1.185283811880712e-05,
+      "loss": 1.2683188915252686,
+      "step": 228
+    },
+    {
+      "epoch": 0.38079470198675497,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.1848501082373235e-05,
+      "loss": 1.2712855339050293,
+      "step": 230
+    },
+    {
+      "epoch": 0.3841059602649007,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.184410208607617e-05,
+      "loss": 1.3452003002166748,
+      "step": 232
+    },
+    {
+      "epoch": 0.38741721854304634,
+      "grad_norm": 0.53125,
+      "learning_rate": 1.1839641188550139e-05,
+      "loss": 1.3081663846969604,
+      "step": 234
+    },
+    {
+      "epoch": 0.39072847682119205,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.1835118449254436e-05,
+      "loss": 1.3071247339248657,
+      "step": 236
+    },
+    {
+      "epoch": 0.39403973509933776,
+      "grad_norm": 0.6640625,
+      "learning_rate": 1.1830533928472643e-05,
+      "loss": 1.3188387155532837,
+      "step": 238
+    },
+    {
+      "epoch": 0.3973509933774834,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.1825887687311838e-05,
+      "loss": 1.3684465885162354,
+      "step": 240
+    },
+    {
+      "epoch": 0.40066225165562913,
+      "grad_norm": 0.34765625,
+      "learning_rate": 1.1821179787701752e-05,
+      "loss": 1.3333898782730103,
+      "step": 242
+    },
+    {
+      "epoch": 0.40397350993377484,
+      "grad_norm": 0.96484375,
+      "learning_rate": 1.1816410292393981e-05,
+      "loss": 1.2793607711791992,
+      "step": 244
+    },
+    {
+      "epoch": 0.40728476821192056,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.1811579264961115e-05,
+      "loss": 1.279396891593933,
+      "step": 246
+    },
+    {
+      "epoch": 0.4105960264900662,
+      "grad_norm": 0.451171875,
+      "learning_rate": 1.1806686769795907e-05,
+      "loss": 1.215667724609375,
+      "step": 248
+    },
+    {
+      "epoch": 0.4139072847682119,
+      "grad_norm": 1.15625,
+      "learning_rate": 1.1801732872110422e-05,
+      "loss": 1.313541293144226,
+      "step": 250
+    },
+    {
+      "epoch": 0.41721854304635764,
+      "grad_norm": 0.76171875,
+      "learning_rate": 1.1796717637935153e-05,
+      "loss": 1.2715359926223755,
+      "step": 252
+    },
+    {
+      "epoch": 0.4205298013245033,
+      "grad_norm": 0.5625,
+      "learning_rate": 1.1791641134118148e-05,
+      "loss": 1.309295892715454,
+      "step": 254
+    },
+    {
+      "epoch": 0.423841059602649,
+      "grad_norm": 0.259765625,
+      "learning_rate": 1.1786503428324115e-05,
+      "loss": 1.2521069049835205,
+      "step": 256
+    },
+    {
+      "epoch": 0.4271523178807947,
+      "grad_norm": 0.89453125,
+      "learning_rate": 1.1781304589033533e-05,
+      "loss": 1.140201210975647,
+      "step": 258
+    },
+    {
+      "epoch": 0.4304635761589404,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1776044685541714e-05,
+      "loss": 1.2161259651184082,
+      "step": 260
+    },
+    {
+      "epoch": 0.4337748344370861,
+      "grad_norm": 0.43359375,
+      "learning_rate": 1.1770723787957905e-05,
+      "loss": 1.326733112335205,
+      "step": 262
+    },
+    {
+      "epoch": 0.4370860927152318,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.1765341967204343e-05,
+      "loss": 1.2771416902542114,
+      "step": 264
+    },
+    {
+      "epoch": 0.44039735099337746,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.175989929501531e-05,
+      "loss": 1.2998336553573608,
+      "step": 266
+    },
+    {
+      "epoch": 0.44370860927152317,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.1754395843936171e-05,
+      "loss": 1.3085970878601074,
+      "step": 268
+    },
+    {
+      "epoch": 0.4470198675496689,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1748831687322418e-05,
+      "loss": 1.307025671005249,
+      "step": 270
+    },
+    {
+      "epoch": 0.4503311258278146,
+      "grad_norm": 0.349609375,
+      "learning_rate": 1.1743206899338683e-05,
+      "loss": 1.2538138628005981,
+      "step": 272
+    },
+    {
+      "epoch": 0.45364238410596025,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1737521554957757e-05,
+      "loss": 1.3346917629241943,
+      "step": 274
+    },
+    {
+      "epoch": 0.45695364238410596,
+      "grad_norm": 0.3671875,
+      "learning_rate": 1.1731775729959587e-05,
+      "loss": 1.2633906602859497,
+      "step": 276
+    },
+    {
+      "epoch": 0.4602649006622517,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1725969500930265e-05,
+      "loss": 1.2056422233581543,
+      "step": 278
+    },
+    {
+      "epoch": 0.46357615894039733,
+      "grad_norm": 0.48828125,
+      "learning_rate": 1.1720102945261004e-05,
+      "loss": 1.2352993488311768,
+      "step": 280
+    },
+    {
+      "epoch": 0.46688741721854304,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.1714176141147119e-05,
+      "loss": 1.1110432147979736,
+      "step": 282
+    },
+    {
+      "epoch": 0.47019867549668876,
+      "grad_norm": 0.412109375,
+      "learning_rate": 1.1708189167586969e-05,
+      "loss": 1.3148404359817505,
+      "step": 284
+    },
+    {
+      "epoch": 0.4735099337748344,
+      "grad_norm": 0.95703125,
+      "learning_rate": 1.1702142104380916e-05,
+      "loss": 1.3085788488388062,
+      "step": 286
+    },
+    {
+      "epoch": 0.4768211920529801,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.1696035032130257e-05,
+      "loss": 1.276586651802063,
+      "step": 288
+    },
+    {
+      "epoch": 0.48013245033112584,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.1689868032236147e-05,
+      "loss": 1.1529457569122314,
+      "step": 290
+    },
+    {
+      "epoch": 0.48344370860927155,
+      "grad_norm": 0.373046875,
+      "learning_rate": 1.1683641186898514e-05,
+      "loss": 1.2723958492279053,
+      "step": 292
+    },
+    {
+      "epoch": 0.4867549668874172,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.1677354579114973e-05,
+      "loss": 1.2373274564743042,
+      "step": 294
+    },
+    {
+      "epoch": 0.4900662251655629,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.1671008292679711e-05,
+      "loss": 1.264053463935852,
+      "step": 296
+    },
+    {
+      "epoch": 0.49337748344370863,
+      "grad_norm": 0.625,
+      "learning_rate": 1.1664602412182362e-05,
+      "loss": 1.2861313819885254,
+      "step": 298
+    },
+    {
+      "epoch": 0.4966887417218543,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.1658137023006901e-05,
+      "loss": 1.246382236480713,
+      "step": 300
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.1651612211330492e-05,
+      "loss": 1.278993844985962,
+      "step": 302
+    },
+    {
+      "epoch": 0.5033112582781457,
+      "grad_norm": 0.3984375,
+      "learning_rate": 1.1645028064122333e-05,
+      "loss": 1.3227043151855469,
+      "step": 304
+    },
+    {
+      "epoch": 0.5066225165562914,
+      "grad_norm": 0.58984375,
+      "learning_rate": 1.1638384669142515e-05,
+      "loss": 1.2452068328857422,
+      "step": 306
+    },
+    {
+      "epoch": 0.5099337748344371,
+      "grad_norm": 0.55859375,
+      "learning_rate": 1.1631682114940838e-05,
+      "loss": 1.2535678148269653,
+      "step": 308
+    },
+    {
+      "epoch": 0.5132450331125827,
+      "grad_norm": 0.57421875,
+      "learning_rate": 1.1624920490855633e-05,
+      "loss": 1.276329755783081,
+      "step": 310
+    },
+    {
+      "epoch": 0.5165562913907285,
+      "grad_norm": 0.53515625,
+      "learning_rate": 1.1618099887012581e-05,
+      "loss": 1.2224030494689941,
+      "step": 312
+    },
+    {
+      "epoch": 0.5198675496688742,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.1611220394323498e-05,
+      "loss": 1.2366602420806885,
+      "step": 314
+    },
+    {
+      "epoch": 0.5231788079470199,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.1604282104485128e-05,
+      "loss": 1.2945020198822021,
+      "step": 316
+    },
+    {
+      "epoch": 0.5264900662251656,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.1597285109977927e-05,
+      "loss": 1.216389536857605,
+      "step": 318
+    },
+    {
+      "epoch": 0.5298013245033113,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.159022950406482e-05,
+      "loss": 1.2610507011413574,
+      "step": 320
+    },
+    {
+      "epoch": 0.5331125827814569,
+      "grad_norm": 0.31640625,
+      "learning_rate": 1.158311538078997e-05,
+      "loss": 1.2242460250854492,
+      "step": 322
+    },
+    {
+      "epoch": 0.5364238410596026,
+      "grad_norm": 0.6015625,
+      "learning_rate": 1.1575942834977512e-05,
+      "loss": 1.2804948091506958,
+      "step": 324
+    },
+    {
+      "epoch": 0.5397350993377483,
+      "grad_norm": 0.46484375,
+      "learning_rate": 1.1568711962230302e-05,
+      "loss": 1.2407575845718384,
+      "step": 326
+    },
+    {
+      "epoch": 0.543046357615894,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.1561422858928625e-05,
+      "loss": 1.2760370969772339,
+      "step": 328
+    },
+    {
+      "epoch": 0.5463576158940397,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.1554075622228926e-05,
+      "loss": 1.2385847568511963,
+      "step": 330
+    },
+    {
+      "epoch": 0.5496688741721855,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.1546670350062515e-05,
+      "loss": 1.271156668663025,
+      "step": 332
+    },
+    {
+      "epoch": 0.5529801324503312,
+      "grad_norm": 0.369140625,
+      "learning_rate": 1.1539207141134253e-05,
+      "loss": 1.288719654083252,
+      "step": 334
+    },
+    {
+      "epoch": 0.5562913907284768,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.1531686094921236e-05,
+      "loss": 1.2762104272842407,
+      "step": 336
+    },
+    {
+      "epoch": 0.5596026490066225,
+      "grad_norm": 0.796875,
+      "learning_rate": 1.1524107311671476e-05,
+      "loss": 1.1551339626312256,
+      "step": 338
+    },
+    {
+      "epoch": 0.5629139072847682,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.1516470892402566e-05,
+      "loss": 1.2796684503555298,
+      "step": 340
+    },
+    {
+      "epoch": 0.5662251655629139,
+      "grad_norm": 0.74609375,
+      "learning_rate": 1.1508776938900323e-05,
+      "loss": 1.2220215797424316,
+      "step": 342
+    },
+    {
+      "epoch": 0.5695364238410596,
+      "grad_norm": 0.2255859375,
+      "learning_rate": 1.150102555371744e-05,
+      "loss": 1.2903807163238525,
+      "step": 344
+    },
+    {
+      "epoch": 0.5728476821192053,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.1493216840172122e-05,
+      "loss": 1.2266340255737305,
+      "step": 346
+    },
+    {
+      "epoch": 0.5761589403973509,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.148535090234669e-05,
+      "loss": 1.2798501253128052,
+      "step": 348
+    },
+    {
+      "epoch": 0.5794701986754967,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.1477427845086227e-05,
+      "loss": 1.2531522512435913,
+      "step": 350
+    },
+    {
+      "epoch": 0.5827814569536424,
+      "grad_norm": 0.3125,
+      "learning_rate": 1.1469447773997141e-05,
+      "loss": 1.237846851348877,
+      "step": 352
+    },
+    {
+      "epoch": 0.5860927152317881,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.1461410795445787e-05,
+      "loss": 1.2226859331130981,
+      "step": 354
+    },
+    {
+      "epoch": 0.5894039735099338,
+      "grad_norm": 0.310546875,
+      "learning_rate": 1.1453317016557045e-05,
+      "loss": 1.2930818796157837,
+      "step": 356
+    },
+    {
+      "epoch": 0.5927152317880795,
+      "grad_norm": 0.703125,
+      "learning_rate": 1.1445166545212877e-05,
+      "loss": 1.3712222576141357,
+      "step": 358
+    },
+    {
+      "epoch": 0.5960264900662252,
+      "grad_norm": 0.59765625,
+      "learning_rate": 1.14369594900509e-05,
+      "loss": 1.2117972373962402,
+      "step": 360
+    },
+    {
+      "epoch": 0.5993377483443708,
+      "grad_norm": 0.875,
+      "learning_rate": 1.1428695960462942e-05,
+      "loss": 1.1898581981658936,
+      "step": 362
+    },
+    {
+      "epoch": 0.6026490066225165,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.142037606659357e-05,
+      "loss": 1.2242567539215088,
+      "step": 364
+    },
+    {
+      "epoch": 0.6059602649006622,
+      "grad_norm": 0.3515625,
+      "learning_rate": 1.1411999919338643e-05,
+      "loss": 1.3001854419708252,
+      "step": 366
+    },
+    {
+      "epoch": 0.609271523178808,
+      "grad_norm": 0.353515625,
+      "learning_rate": 1.1403567630343807e-05,
+      "loss": 1.33438241481781,
+      "step": 368
+    },
+    {
+      "epoch": 0.6125827814569537,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.139507931200303e-05,
+      "loss": 1.2074322700500488,
+      "step": 370
+    },
+    {
+      "epoch": 0.6158940397350994,
+      "grad_norm": 0.75,
+      "learning_rate": 1.1386535077457095e-05,
+      "loss": 1.1700530052185059,
+      "step": 372
+    },
+    {
+      "epoch": 0.6192052980132451,
+      "grad_norm": 0.416015625,
+      "learning_rate": 1.1377935040592088e-05,
+      "loss": 1.1994613409042358,
+      "step": 374
+    },
+    {
+      "epoch": 0.6225165562913907,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.1369279316037887e-05,
+      "loss": 1.3158299922943115,
+      "step": 376
+    },
+    {
+      "epoch": 0.6258278145695364,
+      "grad_norm": 0.4453125,
+      "learning_rate": 1.136056801916663e-05,
+      "loss": 1.2066657543182373,
+      "step": 378
+    },
+    {
+      "epoch": 0.6291390728476821,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.1351801266091177e-05,
+      "loss": 1.2929366827011108,
+      "step": 380
+    },
+    {
+      "epoch": 0.6324503311258278,
+      "grad_norm": 0.384765625,
+      "learning_rate": 1.1342979173663568e-05,
+      "loss": 1.2256704568862915,
+      "step": 382
+    },
+    {
+      "epoch": 0.6357615894039735,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.1334101859473456e-05,
+      "loss": 1.2296499013900757,
+      "step": 384
+    },
+    {
+      "epoch": 0.6390728476821192,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.1325169441846548e-05,
+      "loss": 1.2478747367858887,
+      "step": 386
+    },
+    {
+      "epoch": 0.6423841059602649,
+      "grad_norm": 0.6875,
+      "learning_rate": 1.1316182039843021e-05,
+      "loss": 1.2903329133987427,
+      "step": 388
+    },
+    {
+      "epoch": 0.6456953642384106,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.1307139773255946e-05,
+      "loss": 1.219477891921997,
+      "step": 390
+    },
+    {
+      "epoch": 0.6490066225165563,
+      "grad_norm": 0.546875,
+      "learning_rate": 1.1298042762609677e-05,
+      "loss": 1.256780743598938,
+      "step": 392
+    },
+    {
+      "epoch": 0.652317880794702,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.1288891129158254e-05,
+      "loss": 1.2532964944839478,
+      "step": 394
+    },
+    {
+      "epoch": 0.6556291390728477,
+      "grad_norm": 0.279296875,
+      "learning_rate": 1.127968499488378e-05,
+      "loss": 1.2641328573226929,
+      "step": 396
+    },
+    {
+      "epoch": 0.6589403973509934,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.1270424482494814e-05,
+      "loss": 1.1905931234359741,
+      "step": 398
+    },
+    {
+      "epoch": 0.6622516556291391,
+      "grad_norm": 0.40234375,
+      "learning_rate": 1.1261109715424701e-05,
+      "loss": 1.2366241216659546,
+      "step": 400
+    },
+    {
+      "epoch": 0.6655629139072847,
+      "grad_norm": 0.29296875,
+      "learning_rate": 1.1251740817829957e-05,
+      "loss": 1.119259238243103,
+      "step": 402
+    },
+    {
+      "epoch": 0.6688741721854304,
+      "grad_norm": 0.5390625,
+      "learning_rate": 1.1242317914588603e-05,
+      "loss": 1.2134146690368652,
+      "step": 404
+    },
+    {
+      "epoch": 0.6721854304635762,
+      "grad_norm": 0.4609375,
+      "learning_rate": 1.12328411312985e-05,
+      "loss": 1.1952847242355347,
+      "step": 406
+    },
+    {
+      "epoch": 0.6754966887417219,
+      "grad_norm": 0.38671875,
+      "learning_rate": 1.1223310594275679e-05,
+      "loss": 1.2045927047729492,
+      "step": 408
+    },
+    {
+      "epoch": 0.6788079470198676,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.1213726430552653e-05,
+      "loss": 1.2554584741592407,
+      "step": 410
+    },
+    {
+      "epoch": 0.6821192052980133,
+      "grad_norm": 0.41015625,
+      "learning_rate": 1.1204088767876721e-05,
+      "loss": 1.201694130897522,
+      "step": 412
+    },
+    {
+      "epoch": 0.6854304635761589,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.1194397734708273e-05,
+      "loss": 1.1945827007293701,
+      "step": 414
+    },
+    {
+      "epoch": 0.6887417218543046,
+      "grad_norm": 1.0390625,
+      "learning_rate": 1.1184653460219077e-05,
+      "loss": 1.2243486642837524,
+      "step": 416
+    },
+    {
+      "epoch": 0.6920529801324503,
+      "grad_norm": 0.306640625,
+      "learning_rate": 1.1174856074290548e-05,
+      "loss": 1.2726621627807617,
+      "step": 418
+    },
+    {
+      "epoch": 0.695364238410596,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.1165005707512026e-05,
+      "loss": 1.205135464668274,
+      "step": 420
+    },
+    {
+      "epoch": 0.6986754966887417,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.1155102491179039e-05,
+      "loss": 1.2586928606033325,
+      "step": 422
+    },
+    {
+      "epoch": 0.7019867549668874,
+      "grad_norm": 0.78125,
+      "learning_rate": 1.1145146557291536e-05,
+      "loss": 1.1896426677703857,
+      "step": 424
+    },
+    {
+      "epoch": 0.7052980132450332,
+      "grad_norm": 0.458984375,
+      "learning_rate": 1.1135138038552144e-05,
+      "loss": 1.2678966522216797,
+      "step": 426
+    },
+    {
+      "epoch": 0.7086092715231788,
+      "grad_norm": 0.58203125,
+      "learning_rate": 1.1125077068364393e-05,
+      "loss": 1.2325910329818726,
+      "step": 428
+    },
+    {
+      "epoch": 0.7119205298013245,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.1114963780830942e-05,
+      "loss": 1.2652318477630615,
+      "step": 430
+    },
+    {
+      "epoch": 0.7152317880794702,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.1104798310751776e-05,
+      "loss": 1.2476012706756592,
+      "step": 432
+    },
+    {
+      "epoch": 0.7185430463576159,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.1094580793622432e-05,
+      "loss": 1.2257249355316162,
+      "step": 434
+    },
+    {
+      "epoch": 0.7218543046357616,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.1084311365632182e-05,
+      "loss": 1.1938482522964478,
+      "step": 436
+    },
+    {
+      "epoch": 0.7251655629139073,
+      "grad_norm": 0.287109375,
+      "learning_rate": 1.1073990163662212e-05,
+      "loss": 1.2081587314605713,
+      "step": 438
+    },
+    {
+      "epoch": 0.7284768211920529,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.1063617325283801e-05,
+      "loss": 1.2109920978546143,
+      "step": 440
+    },
+    {
+      "epoch": 0.7317880794701986,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.1053192988756505e-05,
+      "loss": 1.234221339225769,
+      "step": 442
+    },
+    {
+      "epoch": 0.7350993377483444,
+      "grad_norm": 0.2490234375,
+      "learning_rate": 1.1042717293026281e-05,
+      "loss": 1.2554912567138672,
+      "step": 444
+    },
+    {
+      "epoch": 0.7384105960264901,
+      "grad_norm": 0.67578125,
+      "learning_rate": 1.103219037772366e-05,
+      "loss": 1.2109016180038452,
+      "step": 446
+    },
+    {
+      "epoch": 0.7417218543046358,
+      "grad_norm": 0.474609375,
+      "learning_rate": 1.1021612383161882e-05,
+      "loss": 1.171287178993225,
+      "step": 448
+    },
+    {
+      "epoch": 0.7450331125827815,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.1010983450335014e-05,
+      "loss": 1.2758331298828125,
+      "step": 450
+    },
+    {
+      "epoch": 0.7483443708609272,
+      "grad_norm": 0.65625,
+      "learning_rate": 1.1000303720916088e-05,
+      "loss": 1.2499141693115234,
+      "step": 452
+    },
+    {
+      "epoch": 0.7516556291390728,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.0989573337255199e-05,
+      "loss": 1.2714009284973145,
+      "step": 454
+    },
+    {
+      "epoch": 0.7549668874172185,
+      "grad_norm": 0.32421875,
+      "learning_rate": 1.097879244237761e-05,
+      "loss": 1.3011789321899414,
+      "step": 456
+    },
+    {
+      "epoch": 0.7582781456953642,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0967961179981856e-05,
+      "loss": 1.2752792835235596,
+      "step": 458
+    },
+    {
+      "epoch": 0.7615894039735099,
+      "grad_norm": 2.953125,
+      "learning_rate": 1.095707969443781e-05,
+      "loss": 1.1760289669036865,
+      "step": 460
+    },
+    {
+      "epoch": 0.7649006622516556,
+      "grad_norm": 0.5703125,
+      "learning_rate": 1.0946148130784779e-05,
+      "loss": 1.3046189546585083,
+      "step": 462
+    },
+    {
+      "epoch": 0.7682119205298014,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.0935166634729555e-05,
+      "loss": 1.204204797744751,
+      "step": 464
+    },
+    {
+      "epoch": 0.7715231788079471,
+      "grad_norm": 0.3828125,
+      "learning_rate": 1.092413535264448e-05,
+      "loss": 1.204719066619873,
+      "step": 466
+    },
+    {
+      "epoch": 0.7748344370860927,
+      "grad_norm": 0.28515625,
+      "learning_rate": 1.0913054431565497e-05,
+      "loss": 1.2189865112304688,
+      "step": 468
+    },
+    {
+      "epoch": 0.7781456953642384,
+      "grad_norm": 0.66796875,
+      "learning_rate": 1.090192401919018e-05,
+      "loss": 1.130168080329895,
+      "step": 470
+    },
+    {
+      "epoch": 0.7814569536423841,
+      "grad_norm": 0.330078125,
+      "learning_rate": 1.0890744263875782e-05,
+      "loss": 1.2293215990066528,
+      "step": 472
+    },
+    {
+      "epoch": 0.7847682119205298,
+      "grad_norm": 0.265625,
+      "learning_rate": 1.0879515314637238e-05,
+      "loss": 1.189441442489624,
+      "step": 474
+    },
+    {
+      "epoch": 0.7880794701986755,
+      "grad_norm": 0.345703125,
+      "learning_rate": 1.08682373211452e-05,
+      "loss": 1.2883621454238892,
+      "step": 476
+    },
+    {
+      "epoch": 0.7913907284768212,
+      "grad_norm": 0.83203125,
+      "learning_rate": 1.0856910433724017e-05,
+      "loss": 1.2602189779281616,
+      "step": 478
+    },
+    {
+      "epoch": 0.7947019867549668,
+      "grad_norm": 0.318359375,
+      "learning_rate": 1.0845534803349758e-05,
+      "loss": 1.2539976835250854,
+      "step": 480
+    },
+    {
+      "epoch": 0.7980132450331126,
+      "grad_norm": 0.55078125,
+      "learning_rate": 1.0834110581648181e-05,
+      "loss": 1.27140474319458,
+      "step": 482
+    },
+    {
+      "epoch": 0.8013245033112583,
+      "grad_norm": 0.427734375,
+      "learning_rate": 1.0822637920892714e-05,
+      "loss": 1.2304598093032837,
+      "step": 484
+    },
+    {
+      "epoch": 0.804635761589404,
+      "grad_norm": 0.375,
+      "learning_rate": 1.0811116974002438e-05,
+      "loss": 1.2280398607254028,
+      "step": 486
+    },
+    {
+      "epoch": 0.8079470198675497,
+      "grad_norm": 0.62109375,
+      "learning_rate": 1.0799547894540031e-05,
+      "loss": 1.24711012840271,
+      "step": 488
+    },
+    {
+      "epoch": 0.8112582781456954,
+      "grad_norm": 0.283203125,
+      "learning_rate": 1.0787930836709738e-05,
+      "loss": 1.1827949285507202,
+      "step": 490
+    },
+    {
+      "epoch": 0.8145695364238411,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.07762659553553e-05,
+      "loss": 1.2213504314422607,
+      "step": 492
+    },
+    {
+      "epoch": 0.8178807947019867,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0764553405957902e-05,
+      "loss": 1.2925806045532227,
+      "step": 494
+    },
+    {
+      "epoch": 0.8211920529801324,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.0752793344634095e-05,
+      "loss": 1.318811058998108,
+      "step": 496
+    },
+    {
+      "epoch": 0.8245033112582781,
+      "grad_norm": 0.419921875,
+      "learning_rate": 1.074098592813372e-05,
+      "loss": 1.301307201385498,
+      "step": 498
+    },
+    {
+      "epoch": 0.8278145695364238,
+      "grad_norm": 0.4296875,
+      "learning_rate": 1.0729131313837806e-05,
+      "loss": 1.2577488422393799,
+      "step": 500
+    },
+    {
+      "epoch": 0.8311258278145696,
+      "grad_norm": 0.294921875,
+      "learning_rate": 1.071722965975649e-05,
+      "loss": 1.278800368309021,
+      "step": 502
+    },
+    {
+      "epoch": 0.8344370860927153,
+      "grad_norm": 0.390625,
+      "learning_rate": 1.0705281124526903e-05,
+      "loss": 1.1996169090270996,
+      "step": 504
+    },
+    {
+      "epoch": 0.8377483443708609,
+      "grad_norm": 0.4921875,
+      "learning_rate": 1.069328586741105e-05,
+      "loss": 1.2342064380645752,
+      "step": 506
+    },
+    {
+      "epoch": 0.8410596026490066,
+      "grad_norm": 0.52734375,
+      "learning_rate": 1.0681244048293686e-05,
+      "loss": 1.19930899143219,
+      "step": 508
+    },
+    {
+      "epoch": 0.8443708609271523,
+      "grad_norm": 0.59375,
+      "learning_rate": 1.0669155827680204e-05,
+      "loss": 1.2355972528457642,
+      "step": 510
+    },
+    {
+      "epoch": 0.847682119205298,
+      "grad_norm": 0.35546875,
+      "learning_rate": 1.0657021366694475e-05,
+      "loss": 1.2616443634033203,
+      "step": 512
+    },
+    {
+      "epoch": 0.8509933774834437,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.0644840827076712e-05,
+      "loss": 1.2749875783920288,
+      "step": 514
+    },
+    {
+      "epoch": 0.8543046357615894,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.0632614371181303e-05,
+      "loss": 1.139912486076355,
+      "step": 516
+    },
+    {
+      "epoch": 0.8576158940397351,
+      "grad_norm": 0.404296875,
+      "learning_rate": 1.0620342161974664e-05,
+      "loss": 1.2624098062515259,
+      "step": 518
+    },
+    {
+      "epoch": 0.8609271523178808,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0608024363033044e-05,
+      "loss": 1.2522523403167725,
+      "step": 520
+    },
+    {
+      "epoch": 0.8642384105960265,
+      "grad_norm": 0.46875,
+      "learning_rate": 1.059566113854037e-05,
+      "loss": 1.1696981191635132,
+      "step": 522
+    },
+    {
+      "epoch": 0.8675496688741722,
+      "grad_norm": 0.455078125,
+      "learning_rate": 1.0583252653286037e-05,
+      "loss": 1.283386468887329,
+      "step": 524
+    },
+    {
+      "epoch": 0.8708609271523179,
+      "grad_norm": 0.8671875,
+      "learning_rate": 1.0570799072662724e-05,
+      "loss": 1.2220840454101562,
+      "step": 526
+    },
+    {
+      "epoch": 0.8741721854304636,
+      "grad_norm": 0.435546875,
+      "learning_rate": 1.0558300562664192e-05,
+      "loss": 1.2066237926483154,
+      "step": 528
+    },
+    {
+      "epoch": 0.8774834437086093,
+      "grad_norm": 0.34375,
+      "learning_rate": 1.0545757289883054e-05,
+      "loss": 1.215103268623352,
+      "step": 530
+    },
+    {
+      "epoch": 0.8807947019867549,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.0533169421508575e-05,
+      "loss": 1.152737021446228,
+      "step": 532
+    },
+    {
+      "epoch": 0.8841059602649006,
+      "grad_norm": 0.6328125,
+      "learning_rate": 1.0520537125324432e-05,
+      "loss": 1.2905833721160889,
+      "step": 534
+    },
+    {
+      "epoch": 0.8874172185430463,
+      "grad_norm": 1.0,
+      "learning_rate": 1.0507860569706478e-05,
+      "loss": 1.2769966125488281,
+      "step": 536
+    },
+    {
+      "epoch": 0.890728476821192,
+      "grad_norm": 0.734375,
+      "learning_rate": 1.04951399236205e-05,
+      "loss": 1.217033863067627,
+      "step": 538
+    },
+    {
+      "epoch": 0.8940397350993378,
+      "grad_norm": 0.408203125,
+      "learning_rate": 1.0482375356619973e-05,
+      "loss": 1.2337075471878052,
+      "step": 540
+    },
+    {
+      "epoch": 0.8973509933774835,
+      "grad_norm": 0.396484375,
+      "learning_rate": 1.0469567038843786e-05,
+      "loss": 1.2098217010498047,
+      "step": 542
+    },
+    {
+      "epoch": 0.9006622516556292,
+      "grad_norm": 0.333984375,
+      "learning_rate": 1.0456715141013988e-05,
+      "loss": 1.1911996603012085,
+      "step": 544
+    },
+    {
+      "epoch": 0.9039735099337748,
+      "grad_norm": 0.39453125,
+      "learning_rate": 1.0443819834433498e-05,
+      "loss": 1.2873048782348633,
+      "step": 546
+    },
+    {
+      "epoch": 0.9072847682119205,
+      "grad_norm": 0.271484375,
+      "learning_rate": 1.0430881290983842e-05,
+      "loss": 1.1980925798416138,
+      "step": 548
+    },
+    {
+      "epoch": 0.9105960264900662,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 1.0417899683122844e-05,
+      "loss": 1.1385635137557983,
+      "step": 550
+    },
+    {
+      "epoch": 0.9139072847682119,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.0404875183882331e-05,
+      "loss": 1.2270296812057495,
+      "step": 552
+    },
+    {
+      "epoch": 0.9172185430463576,
+      "grad_norm": 1.1015625,
+      "learning_rate": 1.039180796686583e-05,
+      "loss": 1.1677497625350952,
+      "step": 554
+    },
+    {
+      "epoch": 0.9205298013245033,
+      "grad_norm": 0.3203125,
+      "learning_rate": 1.0378698206246262e-05,
+      "loss": 1.2157952785491943,
+      "step": 556
+    },
+    {
+      "epoch": 0.9238410596026491,
+      "grad_norm": 0.263671875,
+      "learning_rate": 1.0365546076763594e-05,
+      "loss": 1.1570444107055664,
+      "step": 558
+    },
+    {
+      "epoch": 0.9271523178807947,
+      "grad_norm": 0.4375,
+      "learning_rate": 1.0352351753722548e-05,
+      "loss": 1.2186548709869385,
+      "step": 560
+    },
+    {
+      "epoch": 0.9304635761589404,
+      "grad_norm": 0.380859375,
+      "learning_rate": 1.0339115412990228e-05,
+      "loss": 1.2401282787322998,
+      "step": 562
+    },
+    {
+      "epoch": 0.9337748344370861,
+      "grad_norm": 0.365234375,
+      "learning_rate": 1.03258372309938e-05,
+      "loss": 1.2739580869674683,
+      "step": 564
+    },
+    {
+      "epoch": 0.9370860927152318,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.0312517384718124e-05,
+      "loss": 1.2308034896850586,
+      "step": 566
+    },
+    {
+      "epoch": 0.9403973509933775,
+      "grad_norm": 0.359375,
+      "learning_rate": 1.0299156051703417e-05,
+      "loss": 1.1893035173416138,
+      "step": 568
+    },
+    {
+      "epoch": 0.9437086092715232,
+      "grad_norm": 0.341796875,
+      "learning_rate": 1.0285753410042862e-05,
+      "loss": 1.202154517173767,
+      "step": 570
+    },
+    {
+      "epoch": 0.9470198675496688,
+      "grad_norm": 0.33203125,
+      "learning_rate": 1.0272309638380244e-05,
+      "loss": 1.135772943496704,
+      "step": 572
+    },
+    {
+      "epoch": 0.9503311258278145,
+      "grad_norm": 0.51953125,
+      "learning_rate": 1.0258824915907578e-05,
+      "loss": 1.15782630443573,
+      "step": 574
+    },
+    {
+      "epoch": 0.9536423841059603,
+      "grad_norm": 0.357421875,
+      "learning_rate": 1.0245299422362702e-05,
+      "loss": 1.2023342847824097,
+      "step": 576
+    },
+    {
+      "epoch": 0.956953642384106,
+      "grad_norm": 0.3046875,
+      "learning_rate": 1.02317333380269e-05,
+      "loss": 1.2450610399246216,
+      "step": 578
+    },
+    {
+      "epoch": 0.9602649006622517,
+      "grad_norm": 0.298828125,
+      "learning_rate": 1.0218126843722487e-05,
+      "loss": 1.1484456062316895,
+      "step": 580
+    },
+    {
+      "epoch": 0.9635761589403974,
+      "grad_norm": 0.447265625,
+      "learning_rate": 1.0204480120810406e-05,
+      "loss": 1.2417069673538208,
+      "step": 582
+    },
+    {
+      "epoch": 0.9668874172185431,
+      "grad_norm": 0.36328125,
+      "learning_rate": 1.0190793351187798e-05,
+      "loss": 1.257660984992981,
+      "step": 584
+    },
+    {
+      "epoch": 0.9701986754966887,
+      "grad_norm": 0.291015625,
+      "learning_rate": 1.0177066717285597e-05,
+      "loss": 1.1812211275100708,
+      "step": 586
+    },
+    {
+      "epoch": 0.9735099337748344,
+      "grad_norm": 0.400390625,
+      "learning_rate": 1.0163300402066082e-05,
+      "loss": 1.1695009469985962,
+      "step": 588
+    },
+    {
+      "epoch": 0.9768211920529801,
+      "grad_norm": 0.376953125,
+      "learning_rate": 1.0149494589020445e-05,
+      "loss": 1.1857471466064453,
+      "step": 590
+    },
+    {
+      "epoch": 0.9801324503311258,
+      "grad_norm": 0.50390625,
+      "learning_rate": 1.0135649462166347e-05,
+      "loss": 1.3397670984268188,
+      "step": 592
+    },
+    {
+      "epoch": 0.9834437086092715,
+      "grad_norm": 0.44140625,
+      "learning_rate": 1.012176520604546e-05,
+      "loss": 1.1958118677139282,
+      "step": 594
+    },
+    {
+      "epoch": 0.9867549668874173,
+      "grad_norm": 0.431640625,
+      "learning_rate": 1.0107842005721008e-05,
+      "loss": 1.206683874130249,
+      "step": 596
+    },
+    {
+      "epoch": 0.9900662251655629,
+      "grad_norm": 0.392578125,
+      "learning_rate": 1.0093880046775308e-05,
+      "loss": 1.3119841814041138,
+      "step": 598
+    },
+    {
+      "epoch": 0.9933774834437086,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.0079879515307288e-05,
+      "loss": 1.1980386972427368,
+      "step": 600
+    },
+    {
+      "epoch": 0.9966887417218543,
+      "grad_norm": 0.60546875,
+      "learning_rate": 1.0065840597930002e-05,
+      "loss": 1.2894337177276611,
+      "step": 602
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.296875,
+      "learning_rate": 1.0051763481768167e-05,
+      "loss": 1.2193963527679443,
+      "step": 604
+    },
+    {
+      "epoch": 1.0033112582781456,
+      "grad_norm": 0.337890625,
+      "learning_rate": 1.0037648354455634e-05,
+      "loss": 1.1995750665664673,
+      "step": 606
+    },
+    {
+      "epoch": 1.0066225165562914,
+      "grad_norm": 0.322265625,
+      "learning_rate": 1.0023495404132912e-05,
+      "loss": 1.130192756652832,
+      "step": 608
+    },
+    {
+      "epoch": 1.009933774834437,
+      "grad_norm": 0.423828125,
+      "learning_rate": 1.0009304819444656e-05,
+      "loss": 1.068596363067627,
+      "step": 610
+    },
+    {
+      "epoch": 1.0132450331125828,
+      "grad_norm": 0.5390625,
+      "learning_rate": 9.995076789537138e-06,
+      "loss": 1.1101475954055786,
+      "step": 612
+    },
+    {
+      "epoch": 1.0165562913907285,
+      "grad_norm": 0.5625,
+      "learning_rate": 9.980811504055752e-06,
+      "loss": 1.107115626335144,
+      "step": 614
+    },
+    {
+      "epoch": 1.0198675496688743,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.966509153142464e-06,
+      "loss": 1.142488956451416,
+      "step": 616
+    },
+    {
+      "epoch": 1.0231788079470199,
+      "grad_norm": 0.375,
+      "learning_rate": 9.952169927433282e-06,
+      "loss": 1.0813229084014893,
+      "step": 618
+    },
+    {
+      "epoch": 1.0264900662251655,
+      "grad_norm": 0.263671875,
+      "learning_rate": 9.93779401805572e-06,
+      "loss": 1.107801079750061,
+      "step": 620
+    },
+    {
+      "epoch": 1.0298013245033113,
+      "grad_norm": 0.265625,
+      "learning_rate": 9.923381616626257e-06,
+      "loss": 1.1251033544540405,
+      "step": 622
+    },
+    {
+      "epoch": 1.033112582781457,
+      "grad_norm": 0.40625,
+      "learning_rate": 9.908932915247765e-06,
+      "loss": 1.1380716562271118,
+      "step": 624
+    },
+    {
+      "epoch": 1.0364238410596027,
+      "grad_norm": 0.361328125,
+      "learning_rate": 9.894448106506957e-06,
+      "loss": 1.1164017915725708,
+      "step": 626
+    },
+    {
+      "epoch": 1.0397350993377483,
+      "grad_norm": 0.3671875,
+      "learning_rate": 9.879927383471828e-06,
+      "loss": 1.154578447341919,
+      "step": 628
+    },
+    {
+      "epoch": 1.0430463576158941,
+      "grad_norm": 0.31640625,
+      "learning_rate": 9.865370939689072e-06,
+      "loss": 1.2191665172576904,
+      "step": 630
+    },
+    {
+      "epoch": 1.0463576158940397,
+      "grad_norm": 1.15625,
+      "learning_rate": 9.850778969181497e-06,
+      "loss": 1.0986692905426025,
+      "step": 632
+    },
+    {
+      "epoch": 1.0496688741721854,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.836151666445465e-06,
+      "loss": 1.059933066368103,
+      "step": 634
+    },
+    {
+      "epoch": 1.0529801324503312,
+      "grad_norm": 0.40234375,
+      "learning_rate": 9.821489226448259e-06,
+      "loss": 1.0793806314468384,
+      "step": 636
+    },
+    {
+      "epoch": 1.0562913907284768,
+      "grad_norm": 0.271484375,
+      "learning_rate": 9.806791844625525e-06,
+      "loss": 1.1094763278961182,
+      "step": 638
+    },
+    {
+      "epoch": 1.0596026490066226,
+      "grad_norm": 0.4375,
+      "learning_rate": 9.792059716878636e-06,
+      "loss": 1.1059627532958984,
+      "step": 640
+    },
+    {
+      "epoch": 1.0629139072847682,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.7772930395721e-06,
+      "loss": 1.0791089534759521,
+      "step": 642
+    },
+    {
+      "epoch": 1.0662251655629138,
+      "grad_norm": 0.498046875,
+      "learning_rate": 9.762492009530935e-06,
+      "loss": 1.1716126203536987,
+      "step": 644
+    },
+    {
+      "epoch": 1.0695364238410596,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.747656824038044e-06,
+      "loss": 1.0453616380691528,
+      "step": 646
+    },
+    {
+      "epoch": 1.0728476821192052,
+      "grad_norm": 0.396484375,
+      "learning_rate": 9.732787680831589e-06,
+      "loss": 1.0998435020446777,
+      "step": 648
+    },
+    {
+      "epoch": 1.076158940397351,
+      "grad_norm": 0.28125,
+      "learning_rate": 9.71788477810236e-06,
+      "loss": 1.0633697509765625,
+      "step": 650
+    },
+    {
+      "epoch": 1.0794701986754967,
+      "grad_norm": 1.0703125,
+      "learning_rate": 9.702948314491119e-06,
+      "loss": 1.0050365924835205,
+      "step": 652
+    },
+    {
+      "epoch": 1.0827814569536425,
+      "grad_norm": 0.3125,
+      "learning_rate": 9.687978489085959e-06,
+      "loss": 1.1256245374679565,
+      "step": 654
+    },
+    {
+      "epoch": 1.086092715231788,
+      "grad_norm": 0.474609375,
+      "learning_rate": 9.67297550141967e-06,
+      "loss": 1.0635474920272827,
+      "step": 656
+    },
+    {
+      "epoch": 1.0894039735099337,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.657939551467042e-06,
+      "loss": 1.1216498613357544,
+      "step": 658
+    },
+    {
+      "epoch": 1.0927152317880795,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.642870839642226e-06,
+      "loss": 1.0962717533111572,
+      "step": 660
+    },
+    {
+      "epoch": 1.096026490066225,
+      "grad_norm": 0.322265625,
+      "learning_rate": 9.62776956679606e-06,
+      "loss": 1.1562925577163696,
+      "step": 662
+    },
+    {
+      "epoch": 1.099337748344371,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.612635934213384e-06,
+      "loss": 1.1446951627731323,
+      "step": 664
+    },
+    {
+      "epoch": 1.1026490066225165,
+      "grad_norm": 0.431640625,
+      "learning_rate": 9.597470143610362e-06,
+      "loss": 1.1515048742294312,
+      "step": 666
+    },
+    {
+      "epoch": 1.1059602649006623,
+      "grad_norm": 0.60546875,
+      "learning_rate": 9.58227239713179e-06,
+      "loss": 1.097119688987732,
+      "step": 668
+    },
+    {
+      "epoch": 1.109271523178808,
+      "grad_norm": 0.345703125,
+      "learning_rate": 9.567042897348405e-06,
+      "loss": 1.1437350511550903,
+      "step": 670
+    },
+    {
+      "epoch": 1.1125827814569536,
+      "grad_norm": 0.5,
+      "learning_rate": 9.551781847254186e-06,
+      "loss": 1.0003553628921509,
+      "step": 672
+    },
+    {
+      "epoch": 1.1158940397350994,
+      "grad_norm": 0.77734375,
+      "learning_rate": 9.536489450263647e-06,
+      "loss": 1.175012230873108,
+      "step": 674
+    },
+    {
+      "epoch": 1.119205298013245,
+      "grad_norm": 0.5234375,
+      "learning_rate": 9.521165910209115e-06,
+      "loss": 1.0635179281234741,
+      "step": 676
+    },
+    {
+      "epoch": 1.1225165562913908,
+      "grad_norm": 0.46484375,
+      "learning_rate": 9.505811431338033e-06,
+      "loss": 1.119577407836914,
+      "step": 678
+    },
+    {
+      "epoch": 1.1258278145695364,
+      "grad_norm": 0.515625,
+      "learning_rate": 9.490426218310226e-06,
+      "loss": 1.110081672668457,
+      "step": 680
+    },
+    {
+      "epoch": 1.129139072847682,
+      "grad_norm": 0.42578125,
+      "learning_rate": 9.475010476195173e-06,
+      "loss": 1.125766396522522,
+      "step": 682
+    },
+    {
+      "epoch": 1.1324503311258278,
+      "grad_norm": 0.294921875,
+      "learning_rate": 9.459564410469273e-06,
+      "loss": 1.1465976238250732,
+      "step": 684
+    },
+    {
+      "epoch": 1.1357615894039734,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.444088227013111e-06,
+      "loss": 1.0712946653366089,
+      "step": 686
+    },
+    {
+      "epoch": 1.1390728476821192,
+      "grad_norm": 0.578125,
+      "learning_rate": 9.42858213210871e-06,
+      "loss": 1.1463303565979004,
+      "step": 688
+    },
+    {
+      "epoch": 1.1423841059602649,
+      "grad_norm": 0.359375,
+      "learning_rate": 9.413046332436781e-06,
+      "loss": 1.0968739986419678,
+      "step": 690
+    },
+    {
+      "epoch": 1.1456953642384107,
+      "grad_norm": 0.35546875,
+      "learning_rate": 9.39748103507398e-06,
+      "loss": 1.1099607944488525,
+      "step": 692
+    },
+    {
+      "epoch": 1.1490066225165563,
+      "grad_norm": 0.466796875,
+      "learning_rate": 9.38188644749012e-06,
+      "loss": 1.2194066047668457,
+      "step": 694
+    },
+    {
+      "epoch": 1.152317880794702,
+      "grad_norm": 0.53515625,
+      "learning_rate": 9.36626277754544e-06,
+      "loss": 1.0382094383239746,
+      "step": 696
+    },
+    {
+      "epoch": 1.1556291390728477,
+      "grad_norm": 0.302734375,
+      "learning_rate": 9.350610233487807e-06,
+      "loss": 1.1106114387512207,
+      "step": 698
+    },
+    {
+      "epoch": 1.1589403973509933,
+      "grad_norm": 0.388671875,
+      "learning_rate": 9.334929023949954e-06,
+      "loss": 1.104656457901001,
+      "step": 700
+    },
+    {
+      "epoch": 1.1622516556291391,
+      "grad_norm": 0.404296875,
+      "learning_rate": 9.319219357946699e-06,
+      "loss": 1.1264766454696655,
+      "step": 702
+    },
+    {
+      "epoch": 1.1655629139072847,
+      "grad_norm": 0.458984375,
+      "learning_rate": 9.303481444872153e-06,
+      "loss": 1.1377006769180298,
+      "step": 704
+    },
+    {
+      "epoch": 1.1688741721854305,
+      "grad_norm": 0.376953125,
+      "learning_rate": 9.287715494496932e-06,
+      "loss": 1.0277259349822998,
+      "step": 706
+    },
+    {
+      "epoch": 1.1721854304635762,
+      "grad_norm": 0.375,
+      "learning_rate": 9.271921716965366e-06,
+      "loss": 1.1559712886810303,
+      "step": 708
+    },
+    {
+      "epoch": 1.1754966887417218,
+      "grad_norm": 0.328125,
+      "learning_rate": 9.256100322792685e-06,
+      "loss": 1.1573880910873413,
+      "step": 710
+    },
+    {
+      "epoch": 1.1788079470198676,
+      "grad_norm": 0.337890625,
+      "learning_rate": 9.240251522862228e-06,
+      "loss": 1.0647897720336914,
+      "step": 712
+    },
+    {
+      "epoch": 1.1821192052980132,
+      "grad_norm": 0.5,
+      "learning_rate": 9.224375528422625e-06,
+      "loss": 1.1478759050369263,
+      "step": 714
+    },
+    {
+      "epoch": 1.185430463576159,
+      "grad_norm": 0.5,
+      "learning_rate": 9.208472551084974e-06,
+      "loss": 1.171141266822815,
+      "step": 716
+    },
+    {
+      "epoch": 1.1887417218543046,
+      "grad_norm": 0.33984375,
+      "learning_rate": 9.192542802820039e-06,
+      "loss": 1.1633001565933228,
+      "step": 718
+    },
+    {
+      "epoch": 1.1920529801324504,
+      "grad_norm": 0.384765625,
+      "learning_rate": 9.176586495955403e-06,
+      "loss": 1.0764974355697632,
+      "step": 720
+    },
+    {
+      "epoch": 1.195364238410596,
+      "grad_norm": 0.484375,
+      "learning_rate": 9.160603843172655e-06,
+      "loss": 1.094746708869934,
+      "step": 722
+    },
+    {
+      "epoch": 1.1986754966887416,
+      "grad_norm": 0.37109375,
+      "learning_rate": 9.144595057504552e-06,
+      "loss": 1.1030086278915405,
+      "step": 724
+    },
+    {
+      "epoch": 1.2019867549668874,
+      "grad_norm": 0.333984375,
+      "learning_rate": 9.128560352332163e-06,
+      "loss": 1.1092870235443115,
+      "step": 726
+    },
+    {
+      "epoch": 1.205298013245033,
+      "grad_norm": 0.455078125,
+      "learning_rate": 9.112499941382057e-06,
+      "loss": 1.0155194997787476,
+      "step": 728
+    },
+    {
+      "epoch": 1.2086092715231789,
+      "grad_norm": 0.4296875,
+      "learning_rate": 9.096414038723415e-06,
+      "loss": 1.1965831518173218,
+      "step": 730
+    },
+    {
+      "epoch": 1.2119205298013245,
+      "grad_norm": 0.26171875,
+      "learning_rate": 9.080302858765213e-06,
+      "loss": 1.077614665031433,
+      "step": 732
+    },
+    {
+      "epoch": 1.2152317880794703,
+      "grad_norm": 0.64453125,
+      "learning_rate": 9.06416661625334e-06,
+      "loss": 1.1260370016098022,
+      "step": 734
+    },
+    {
+      "epoch": 1.218543046357616,
+      "grad_norm": 0.55859375,
+      "learning_rate": 9.048005526267742e-06,
+      "loss": 1.122714638710022,
+      "step": 736
+    },
+    {
+      "epoch": 1.2218543046357615,
+      "grad_norm": 0.3984375,
+      "learning_rate": 9.031819804219566e-06,
+      "loss": 1.1720261573791504,
+      "step": 738
+    },
+    {
+      "epoch": 1.2251655629139073,
+      "grad_norm": 0.380859375,
+      "learning_rate": 9.01560966584827e-06,
+      "loss": 1.1976429224014282,
+      "step": 740
+    },
+    {
+      "epoch": 1.228476821192053,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.999375327218765e-06,
+      "loss": 1.1546531915664673,
+      "step": 742
+    },
+    {
+      "epoch": 1.2317880794701987,
+      "grad_norm": 0.2890625,
+      "learning_rate": 8.98311700471852e-06,
+      "loss": 1.1697813272476196,
+      "step": 744
+    },
+    {
+      "epoch": 1.2350993377483444,
+      "grad_norm": 0.96875,
+      "learning_rate": 8.966834915054694e-06,
+      "loss": 1.1024788618087769,
+      "step": 746
+    },
+    {
+      "epoch": 1.23841059602649,
+      "grad_norm": 0.291015625,
+      "learning_rate": 8.950529275251226e-06,
+      "loss": 1.1089746952056885,
+      "step": 748
+    },
+    {
+      "epoch": 1.2417218543046358,
+      "grad_norm": 0.88671875,
+      "learning_rate": 8.934200302645968e-06,
+      "loss": 1.0583044290542603,
+      "step": 750
+    },
+    {
+      "epoch": 1.2450331125827814,
+      "grad_norm": 1.6796875,
+      "learning_rate": 8.917848214887765e-06,
+      "loss": 1.0897866487503052,
+      "step": 752
+    },
+    {
+      "epoch": 1.2483443708609272,
+      "grad_norm": 0.267578125,
+      "learning_rate": 8.901473229933567e-06,
+      "loss": 1.0810881853103638,
+      "step": 754
+    },
+    {
+      "epoch": 1.2516556291390728,
+      "grad_norm": 0.8203125,
+      "learning_rate": 8.88507556604552e-06,
+      "loss": 1.0693601369857788,
+      "step": 756
+    },
+    {
+      "epoch": 1.2549668874172186,
+      "grad_norm": 0.375,
+      "learning_rate": 8.868655441788064e-06,
+      "loss": 1.0969278812408447,
+      "step": 758
+    },
+    {
+      "epoch": 1.2582781456953642,
+      "grad_norm": 0.375,
+      "learning_rate": 8.852213076025005e-06,
+      "loss": 1.154745101928711,
+      "step": 760
+    },
+    {
+      "epoch": 1.26158940397351,
+      "grad_norm": 0.30859375,
+      "learning_rate": 8.83574868791661e-06,
+      "loss": 1.1125857830047607,
+      "step": 762
+    },
+    {
+      "epoch": 1.2649006622516556,
+      "grad_norm": 0.39453125,
+      "learning_rate": 8.819262496916676e-06,
+      "loss": 1.1467704772949219,
+      "step": 764
+    },
+    {
+      "epoch": 1.2682119205298013,
+      "grad_norm": 0.486328125,
+      "learning_rate": 8.80275472276962e-06,
+      "loss": 1.1354914903640747,
+      "step": 766
+    },
+    {
+      "epoch": 1.271523178807947,
+      "grad_norm": 0.609375,
+      "learning_rate": 8.786225585507534e-06,
+      "loss": 1.1262799501419067,
+      "step": 768
+    },
+    {
+      "epoch": 1.2748344370860927,
+      "grad_norm": 0.87890625,
+      "learning_rate": 8.769675305447267e-06,
+      "loss": 1.0878307819366455,
+      "step": 770
+    },
+    {
+      "epoch": 1.2781456953642385,
+      "grad_norm": 0.31640625,
+      "learning_rate": 8.753104103187465e-06,
+      "loss": 1.0384912490844727,
+      "step": 772
+    },
+    {
+      "epoch": 1.281456953642384,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.73651219960566e-06,
+      "loss": 1.1220159530639648,
+      "step": 774
+    },
+    {
+      "epoch": 1.2847682119205297,
+      "grad_norm": 0.296875,
+      "learning_rate": 8.719899815855308e-06,
+      "loss": 1.057347059249878,
+      "step": 776
+    },
+    {
+      "epoch": 1.2880794701986755,
+      "grad_norm": 0.53515625,
+      "learning_rate": 8.703267173362838e-06,
+      "loss": 1.1674779653549194,
+      "step": 778
+    },
+    {
+      "epoch": 1.2913907284768211,
+      "grad_norm": 0.4140625,
+      "learning_rate": 8.686614493824725e-06,
+      "loss": 1.1553187370300293,
+      "step": 780
+    },
+    {
+      "epoch": 1.294701986754967,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.6699419992045e-06,
+      "loss": 1.0362199544906616,
+      "step": 782
+    },
+    {
+      "epoch": 1.2980132450331126,
+      "grad_norm": 0.341796875,
+      "learning_rate": 8.653249911729814e-06,
+      "loss": 1.0475202798843384,
+      "step": 784
+    },
+    {
+      "epoch": 1.3013245033112582,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.63653845388948e-06,
+      "loss": 0.9839364886283875,
+      "step": 786
+    },
+    {
+      "epoch": 1.304635761589404,
+      "grad_norm": 0.396484375,
+      "learning_rate": 8.619807848430485e-06,
+      "loss": 1.09470534324646,
+      "step": 788
+    },
+    {
+      "epoch": 1.3079470198675498,
+      "grad_norm": 0.44140625,
+      "learning_rate": 8.603058318355047e-06,
+      "loss": 1.1766548156738281,
+      "step": 790
+    },
+    {
+      "epoch": 1.3112582781456954,
+      "grad_norm": 0.439453125,
+      "learning_rate": 8.586290086917619e-06,
+      "loss": 1.1274049282073975,
+      "step": 792
+    },
+    {
+      "epoch": 1.314569536423841,
+      "grad_norm": 0.345703125,
+      "learning_rate": 8.569503377621931e-06,
+      "loss": 1.1291522979736328,
+      "step": 794
+    },
+    {
+      "epoch": 1.3178807947019868,
+      "grad_norm": 0.419921875,
+      "learning_rate": 8.552698414218007e-06,
+      "loss": 1.100280523300171,
+      "step": 796
+    },
+    {
+      "epoch": 1.3211920529801324,
+      "grad_norm": 0.71484375,
+      "learning_rate": 8.535875420699173e-06,
+      "loss": 1.0891684293746948,
+      "step": 798
+    },
+    {
+      "epoch": 1.3245033112582782,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.519034621299082e-06,
+      "loss": 1.1650758981704712,
+      "step": 800
+    },
+    {
+      "epoch": 1.3278145695364238,
+      "grad_norm": 0.361328125,
+      "learning_rate": 8.502176240488724e-06,
+      "loss": 1.0949400663375854,
+      "step": 802
+    },
+    {
+      "epoch": 1.3311258278145695,
+      "grad_norm": 0.328125,
+      "learning_rate": 8.485300502973424e-06,
+      "loss": 1.0937325954437256,
+      "step": 804
+    },
+    {
+      "epoch": 1.3344370860927153,
+      "grad_norm": 0.447265625,
+      "learning_rate": 8.468407633689864e-06,
+      "loss": 1.0812309980392456,
+      "step": 806
+    },
+    {
+      "epoch": 1.3377483443708609,
+      "grad_norm": 0.42578125,
+      "learning_rate": 8.451497857803066e-06,
+      "loss": 1.0209524631500244,
+      "step": 808
+    },
+    {
+      "epoch": 1.3410596026490067,
+      "grad_norm": 0.3984375,
+      "learning_rate": 8.434571400703408e-06,
+      "loss": 1.1526941061019897,
+      "step": 810
+    },
+    {
+      "epoch": 1.3443708609271523,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.417628488003606e-06,
+      "loss": 1.1507055759429932,
+      "step": 812
+    },
+    {
+      "epoch": 1.347682119205298,
+      "grad_norm": 0.263671875,
+      "learning_rate": 8.400669345535718e-06,
+      "loss": 1.0535410642623901,
+      "step": 814
+    },
+    {
+      "epoch": 1.3509933774834437,
+      "grad_norm": 0.3515625,
+      "learning_rate": 8.383694199348123e-06,
+      "loss": 1.1291109323501587,
+      "step": 816
+    },
+    {
+      "epoch": 1.3543046357615893,
+      "grad_norm": 0.435546875,
+      "learning_rate": 8.366703275702516e-06,
+      "loss": 1.0811841487884521,
+      "step": 818
+    },
+    {
+      "epoch": 1.3576158940397351,
+      "grad_norm": 0.2451171875,
+      "learning_rate": 8.349696801070893e-06,
+      "loss": 1.134947657585144,
+      "step": 820
+    },
+    {
+      "epoch": 1.3609271523178808,
+      "grad_norm": 0.36328125,
+      "learning_rate": 8.332675002132523e-06,
+      "loss": 1.1577849388122559,
+      "step": 822
+    },
+    {
+      "epoch": 1.3642384105960264,
+      "grad_norm": 0.8046875,
+      "learning_rate": 8.315638105770938e-06,
+      "loss": 1.1460561752319336,
+      "step": 824
+    },
+    {
+      "epoch": 1.3675496688741722,
+      "grad_norm": 0.69140625,
+      "learning_rate": 8.298586339070901e-06,
+      "loss": 1.1265171766281128,
+      "step": 826
+    },
+    {
+      "epoch": 1.370860927152318,
+      "grad_norm": 0.31640625,
+      "learning_rate": 8.28151992931538e-06,
+      "loss": 1.2352640628814697,
+      "step": 828
+    },
+    {
+      "epoch": 1.3741721854304636,
+      "grad_norm": 0.36328125,
+      "learning_rate": 8.264439103982522e-06,
+      "loss": 1.1254183053970337,
+      "step": 830
+    },
+    {
+      "epoch": 1.3774834437086092,
+      "grad_norm": 0.53125,
+      "learning_rate": 8.247344090742623e-06,
+      "loss": 1.1543939113616943,
+      "step": 832
+    },
+    {
+      "epoch": 1.380794701986755,
+      "grad_norm": 0.380859375,
+      "learning_rate": 8.230235117455083e-06,
+      "loss": 1.0869652032852173,
+      "step": 834
+    },
+    {
+      "epoch": 1.3841059602649006,
+      "grad_norm": 0.34765625,
+      "learning_rate": 8.213112412165377e-06,
+      "loss": 1.0378072261810303,
+      "step": 836
+    },
+    {
+      "epoch": 1.3874172185430464,
+      "grad_norm": 0.287109375,
+      "learning_rate": 8.195976203102018e-06,
+      "loss": 1.1564513444900513,
+      "step": 838
+    },
+    {
+      "epoch": 1.390728476821192,
+      "grad_norm": 0.287109375,
+      "learning_rate": 8.178826718673508e-06,
+      "loss": 1.117110252380371,
+      "step": 840
+    },
+    {
+      "epoch": 1.3940397350993377,
+      "grad_norm": 0.921875,
+      "learning_rate": 8.161664187465298e-06,
+      "loss": 1.1238434314727783,
+      "step": 842
+    },
+    {
+      "epoch": 1.3973509933774835,
+      "grad_norm": 0.384765625,
+      "learning_rate": 8.144488838236734e-06,
+      "loss": 1.1785955429077148,
+      "step": 844
+    },
+    {
+      "epoch": 1.400662251655629,
+      "grad_norm": 0.40625,
+      "learning_rate": 8.127300899918021e-06,
+      "loss": 1.2380834817886353,
+      "step": 846
+    },
+    {
+      "epoch": 1.403973509933775,
+      "grad_norm": 0.8984375,
+      "learning_rate": 8.110100601607157e-06,
+      "loss": 1.1105948686599731,
+      "step": 848
+    },
+    {
+      "epoch": 1.4072847682119205,
+      "grad_norm": 0.3671875,
+      "learning_rate": 8.092888172566887e-06,
+      "loss": 1.1651510000228882,
+      "step": 850
+    },
+    {
+      "epoch": 1.410596026490066,
+      "grad_norm": 0.302734375,
+      "learning_rate": 8.075663842221656e-06,
+      "loss": 1.0690885782241821,
+      "step": 852
+    },
+    {
+      "epoch": 1.413907284768212,
+      "grad_norm": 1.015625,
+      "learning_rate": 8.058427840154531e-06,
+      "loss": 1.154086947441101,
+      "step": 854
+    },
+    {
+      "epoch": 1.4172185430463577,
+      "grad_norm": 0.41015625,
+      "learning_rate": 8.041180396104154e-06,
+      "loss": 1.0742074251174927,
+      "step": 856
+    },
+    {
+      "epoch": 1.4205298013245033,
+      "grad_norm": 0.90234375,
+      "learning_rate": 8.02392173996168e-06,
+      "loss": 1.0775120258331299,
+      "step": 858
+    },
+    {
+      "epoch": 1.423841059602649,
+      "grad_norm": 0.466796875,
+      "learning_rate": 8.006652101767706e-06,
+      "loss": 1.1411365270614624,
+      "step": 860
+    },
+    {
+      "epoch": 1.4271523178807948,
+      "grad_norm": 0.578125,
+      "learning_rate": 7.989371711709214e-06,
+      "loss": 1.159982442855835,
+      "step": 862
+    },
+    {
+      "epoch": 1.4304635761589404,
+      "grad_norm": 0.462890625,
+      "learning_rate": 7.97208080011649e-06,
+      "loss": 1.1033796072006226,
+      "step": 864
+    },
+    {
+      "epoch": 1.4337748344370862,
+      "grad_norm": 0.419921875,
+      "learning_rate": 7.95477959746007e-06,
+      "loss": 1.1719162464141846,
+      "step": 866
+    },
+    {
+      "epoch": 1.4370860927152318,
+      "grad_norm": 0.4140625,
+      "learning_rate": 7.937468334347649e-06,
+      "loss": 1.202003002166748,
+      "step": 868
+    },
+    {
+      "epoch": 1.4403973509933774,
+      "grad_norm": 0.75390625,
+      "learning_rate": 7.92014724152103e-06,
+      "loss": 1.1364517211914062,
+      "step": 870
+    },
+    {
+      "epoch": 1.4437086092715232,
+      "grad_norm": 0.41796875,
+      "learning_rate": 7.902816549853027e-06,
+      "loss": 1.1248327493667603,
+      "step": 872
+    },
+    {
+      "epoch": 1.4470198675496688,
+      "grad_norm": 0.330078125,
+      "learning_rate": 7.885476490344407e-06,
+      "loss": 1.122451663017273,
+      "step": 874
+    },
+    {
+      "epoch": 1.4503311258278146,
+      "grad_norm": 0.25390625,
+      "learning_rate": 7.868127294120787e-06,
+      "loss": 1.0662963390350342,
+      "step": 876
+    },
+    {
+      "epoch": 1.4536423841059603,
+      "grad_norm": 0.271484375,
+      "learning_rate": 7.850769192429576e-06,
+      "loss": 1.1393651962280273,
+      "step": 878
+    },
+    {
+      "epoch": 1.4569536423841059,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.833402416636885e-06,
+      "loss": 1.051371693611145,
+      "step": 880
+    },
+    {
+      "epoch": 1.4602649006622517,
+      "grad_norm": 0.421875,
+      "learning_rate": 7.816027198224437e-06,
+      "loss": 1.134348750114441,
+      "step": 882
+    },
+    {
+      "epoch": 1.4635761589403973,
+      "grad_norm": 0.4375,
+      "learning_rate": 7.798643768786491e-06,
+      "loss": 1.1689413785934448,
+      "step": 884
+    },
+    {
+      "epoch": 1.466887417218543,
+      "grad_norm": 0.341796875,
+      "learning_rate": 7.781252360026748e-06,
+      "loss": 1.1354305744171143,
+      "step": 886
+    },
+    {
+      "epoch": 1.4701986754966887,
+      "grad_norm": 0.443359375,
+      "learning_rate": 7.763853203755266e-06,
+      "loss": 1.0442321300506592,
+      "step": 888
+    },
+    {
+      "epoch": 1.4735099337748343,
+      "grad_norm": 0.310546875,
+      "learning_rate": 7.746446531885371e-06,
+      "loss": 1.1418088674545288,
+      "step": 890
+    },
+    {
+      "epoch": 1.4768211920529801,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.729032576430563e-06,
+      "loss": 1.1551663875579834,
+      "step": 892
+    },
+    {
+      "epoch": 1.480132450331126,
+      "grad_norm": 0.53125,
+      "learning_rate": 7.711611569501427e-06,
+      "loss": 1.123194694519043,
+      "step": 894
+    },
+    {
+      "epoch": 1.4834437086092715,
+      "grad_norm": 0.494140625,
+      "learning_rate": 7.694183743302533e-06,
+      "loss": 1.1747374534606934,
+      "step": 896
+    },
+    {
+      "epoch": 1.4867549668874172,
+      "grad_norm": 0.546875,
+      "learning_rate": 7.676749330129348e-06,
+      "loss": 1.0527430772781372,
+      "step": 898
+    },
+    {
+      "epoch": 1.490066225165563,
+      "grad_norm": 0.458984375,
+      "learning_rate": 7.659308562365133e-06,
+      "loss": 1.144150733947754,
+      "step": 900
+    },
+    {
+      "epoch": 1.4933774834437086,
+      "grad_norm": 0.46875,
+      "learning_rate": 7.64186167247786e-06,
+      "loss": 1.0819345712661743,
+      "step": 902
+    },
+    {
+      "epoch": 1.4966887417218544,
+      "grad_norm": 1.0703125,
+      "learning_rate": 7.624408893017088e-06,
+      "loss": 1.1327379941940308,
+      "step": 904
+    },
+    {
+      "epoch": 1.5,
+      "grad_norm": 0.31640625,
+      "learning_rate": 7.6069504566108895e-06,
+      "loss": 1.0392861366271973,
+      "step": 906
+    },
+    {
+      "epoch": 1.5033112582781456,
+      "grad_norm": 1.3984375,
+      "learning_rate": 7.589486595962732e-06,
+      "loss": 1.1170624494552612,
+      "step": 908
+    },
+    {
+      "epoch": 1.5066225165562914,
+      "grad_norm": 0.625,
+      "learning_rate": 7.572017543848388e-06,
+      "loss": 1.079361081123352,
+      "step": 910
+    },
+    {
+      "epoch": 1.5099337748344372,
+      "grad_norm": 0.76171875,
+      "learning_rate": 7.5545435331128185e-06,
+      "loss": 1.1161906719207764,
+      "step": 912
+    },
+    {
+      "epoch": 1.5132450331125828,
+      "grad_norm": 0.29296875,
+      "learning_rate": 7.537064796667088e-06,
+      "loss": 1.131893277168274,
+      "step": 914
+    },
+    {
+      "epoch": 1.5165562913907285,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.519581567485246e-06,
+      "loss": 1.1086527109146118,
+      "step": 916
+    },
+    {
+      "epoch": 1.519867549668874,
+      "grad_norm": 0.52734375,
+      "learning_rate": 7.502094078601223e-06,
+      "loss": 1.172413945198059,
+      "step": 918
+    },
+    {
+      "epoch": 1.5231788079470199,
+      "grad_norm": 0.275390625,
+      "learning_rate": 7.48460256310573e-06,
+      "loss": 1.0879758596420288,
+      "step": 920
+    },
+    {
+      "epoch": 1.5264900662251657,
+      "grad_norm": 0.306640625,
+      "learning_rate": 7.4671072541431455e-06,
+      "loss": 1.1638213396072388,
+      "step": 922
+    },
+    {
+      "epoch": 1.5298013245033113,
+      "grad_norm": 0.42578125,
+      "learning_rate": 7.449608384908417e-06,
+      "loss": 1.091423749923706,
+      "step": 924
+    },
+    {
+      "epoch": 1.533112582781457,
+      "grad_norm": 0.435546875,
+      "learning_rate": 7.432106188643942e-06,
+      "loss": 1.1669130325317383,
+      "step": 926
+    },
+    {
+      "epoch": 1.5364238410596025,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.414600898636467e-06,
+      "loss": 1.1157746315002441,
+      "step": 928
+    },
+    {
+      "epoch": 1.5397350993377483,
+      "grad_norm": 0.34375,
+      "learning_rate": 7.39709274821397e-06,
+      "loss": 1.079698085784912,
+      "step": 930
+    },
+    {
+      "epoch": 1.5430463576158941,
+      "grad_norm": 0.333984375,
+      "learning_rate": 7.379581970742558e-06,
+      "loss": 0.983738899230957,
+      "step": 932
+    },
+    {
+      "epoch": 1.5463576158940397,
+      "grad_norm": 0.412109375,
+      "learning_rate": 7.3620687996233606e-06,
+      "loss": 1.1889533996582031,
+      "step": 934
+    },
+    {
+      "epoch": 1.5496688741721854,
+      "grad_norm": 0.353515625,
+      "learning_rate": 7.344553468289401e-06,
+      "loss": 1.1194508075714111,
+      "step": 936
+    },
+    {
+      "epoch": 1.5529801324503312,
+      "grad_norm": 0.37890625,
+      "learning_rate": 7.327036210202505e-06,
+      "loss": 1.1008572578430176,
+      "step": 938
+    },
+    {
+      "epoch": 1.5562913907284768,
+      "grad_norm": 0.337890625,
+      "learning_rate": 7.309517258850176e-06,
+      "loss": 1.0888991355895996,
+      "step": 940
+    },
+    {
+      "epoch": 1.5596026490066226,
+      "grad_norm": 0.51953125,
+      "learning_rate": 7.291996847742485e-06,
+      "loss": 1.1011091470718384,
+      "step": 942
+    },
+    {
+      "epoch": 1.5629139072847682,
+      "grad_norm": 0.5078125,
+      "learning_rate": 7.274475210408963e-06,
+      "loss": 1.0910996198654175,
+      "step": 944
+    },
+    {
+      "epoch": 1.5662251655629138,
+      "grad_norm": 0.380859375,
+      "learning_rate": 7.256952580395487e-06,
+      "loss": 1.086224913597107,
+      "step": 946
+    },
+    {
+      "epoch": 1.5695364238410596,
+      "grad_norm": 0.404296875,
+      "learning_rate": 7.239429191261161e-06,
+      "loss": 1.1097546815872192,
+      "step": 948
+    },
+    {
+      "epoch": 1.5728476821192054,
+      "grad_norm": 0.408203125,
+      "learning_rate": 7.221905276575209e-06,
+      "loss": 1.0408302545547485,
+      "step": 950
+    },
+    {
+      "epoch": 1.576158940397351,
+      "grad_norm": 0.361328125,
+      "learning_rate": 7.204381069913862e-06,
+      "loss": 1.1143653392791748,
+      "step": 952
+    },
+    {
+      "epoch": 1.5794701986754967,
+      "grad_norm": 0.380859375,
+      "learning_rate": 7.186856804857243e-06,
+      "loss": 1.1487960815429688,
+      "step": 954
+    },
+    {
+      "epoch": 1.5827814569536423,
+      "grad_norm": 0.5859375,
+      "learning_rate": 7.1693327149862504e-06,
+      "loss": 1.10153329372406,
+      "step": 956
+    },
+    {
+      "epoch": 1.586092715231788,
+      "grad_norm": 0.384765625,
+      "learning_rate": 7.151809033879449e-06,
+      "loss": 1.0794066190719604,
+      "step": 958
+    },
+    {
+      "epoch": 1.589403973509934,
+      "grad_norm": 0.96484375,
+      "learning_rate": 7.134285995109952e-06,
+      "loss": 1.1781980991363525,
+      "step": 960
+    },
+    {
+      "epoch": 1.5927152317880795,
+      "grad_norm": 0.34765625,
+      "learning_rate": 7.116763832242319e-06,
+      "loss": 1.0936973094940186,
+      "step": 962
+    },
+    {
+      "epoch": 1.596026490066225,
+      "grad_norm": 1.203125,
+      "learning_rate": 7.099242778829429e-06,
+      "loss": 1.1570301055908203,
+      "step": 964
+    },
+    {
+      "epoch": 1.5993377483443707,
+      "grad_norm": 0.32421875,
+      "learning_rate": 7.081723068409368e-06,
+      "loss": 1.1480774879455566,
+      "step": 966
+    },
+    {
+      "epoch": 1.6026490066225165,
+      "grad_norm": 0.416015625,
+      "learning_rate": 7.064204934502333e-06,
+      "loss": 1.1552226543426514,
+      "step": 968
+    },
+    {
+      "epoch": 1.6059602649006623,
+      "grad_norm": 0.5234375,
+      "learning_rate": 7.0466886106075e-06,
+      "loss": 1.2289761304855347,
+      "step": 970
+    },
+    {
+      "epoch": 1.609271523178808,
+      "grad_norm": 0.345703125,
+      "learning_rate": 7.029174330199921e-06,
+      "loss": 1.1841604709625244,
+      "step": 972
+    },
+    {
+      "epoch": 1.6125827814569536,
+      "grad_norm": 0.4765625,
+      "learning_rate": 7.011662326727409e-06,
+      "loss": 1.1432627439498901,
+      "step": 974
+    },
+    {
+      "epoch": 1.6158940397350994,
+      "grad_norm": 0.490234375,
+      "learning_rate": 6.994152833607431e-06,
+      "loss": 1.0701535940170288,
+      "step": 976
+    },
+    {
+      "epoch": 1.6192052980132452,
+      "grad_norm": 0.51171875,
+      "learning_rate": 6.976646084223992e-06,
+      "loss": 1.0911750793457031,
+      "step": 978
+    },
+    {
+      "epoch": 1.6225165562913908,
+      "grad_norm": 0.41796875,
+      "learning_rate": 6.959142311924522e-06,
+      "loss": 1.0632914304733276,
+      "step": 980
+    },
+    {
+      "epoch": 1.6258278145695364,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.9416417500167744e-06,
+      "loss": 1.1161998510360718,
+      "step": 982
+    },
+    {
+      "epoch": 1.629139072847682,
+      "grad_norm": 0.306640625,
+      "learning_rate": 6.924144631765715e-06,
+      "loss": 1.0910724401474,
+      "step": 984
+    },
+    {
+      "epoch": 1.6324503311258278,
+      "grad_norm": 0.26953125,
+      "learning_rate": 6.906651190390396e-06,
+      "loss": 1.0503889322280884,
+      "step": 986
+    },
+    {
+      "epoch": 1.6357615894039736,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.88916165906087e-06,
+      "loss": 1.2001702785491943,
+      "step": 988
+    },
+    {
+      "epoch": 1.6390728476821192,
+      "grad_norm": 0.578125,
+      "learning_rate": 6.8716762708950744e-06,
+      "loss": 1.1387388706207275,
+      "step": 990
+    },
+    {
+      "epoch": 1.6423841059602649,
+      "grad_norm": 0.59765625,
+      "learning_rate": 6.854195258955716e-06,
+      "loss": 1.14443039894104,
+      "step": 992
+    },
+    {
+      "epoch": 1.6456953642384105,
+      "grad_norm": 0.384765625,
+      "learning_rate": 6.836718856247175e-06,
+      "loss": 1.1768776178359985,
+      "step": 994
+    },
+    {
+      "epoch": 1.6490066225165563,
+      "grad_norm": 0.2373046875,
+      "learning_rate": 6.819247295712395e-06,
+      "loss": 1.1593154668807983,
+      "step": 996
+    },
+    {
+      "epoch": 1.652317880794702,
+      "grad_norm": 0.275390625,
+      "learning_rate": 6.801780810229773e-06,
+      "loss": 1.1416690349578857,
+      "step": 998
+    },
+    {
+      "epoch": 1.6556291390728477,
+      "grad_norm": 0.3515625,
+      "learning_rate": 6.784319632610069e-06,
+      "loss": 1.1714385747909546,
+      "step": 1000
+    },
+    {
+      "epoch": 1.6589403973509933,
+      "grad_norm": 0.353515625,
+      "learning_rate": 6.766863995593291e-06,
+      "loss": 1.125061273574829,
+      "step": 1002
+    },
+    {
+      "epoch": 1.6622516556291391,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.749414131845594e-06,
+      "loss": 1.0884270668029785,
+      "step": 1004
+    },
+    {
+      "epoch": 1.6655629139072847,
+      "grad_norm": 0.23046875,
+      "learning_rate": 6.7319702739561835e-06,
+      "loss": 1.0700236558914185,
+      "step": 1006
+    },
+    {
+      "epoch": 1.6688741721854305,
+      "grad_norm": 0.302734375,
+      "learning_rate": 6.714532654434209e-06,
+      "loss": 1.1123465299606323,
+      "step": 1008
+    },
+    {
+      "epoch": 1.6721854304635762,
+      "grad_norm": 0.361328125,
+      "learning_rate": 6.697101505705675e-06,
+      "loss": 1.0608749389648438,
+      "step": 1010
+    },
+    {
+      "epoch": 1.6754966887417218,
+      "grad_norm": 0.349609375,
+      "learning_rate": 6.6796770601103345e-06,
+      "loss": 1.1096681356430054,
+      "step": 1012
+    },
+    {
+      "epoch": 1.6788079470198676,
+      "grad_norm": 0.369140625,
+      "learning_rate": 6.662259549898591e-06,
+      "loss": 1.0963759422302246,
+      "step": 1014
+    },
+    {
+      "epoch": 1.6821192052980134,
+      "grad_norm": 0.5,
+      "learning_rate": 6.644849207228413e-06,
+      "loss": 1.1321171522140503,
+      "step": 1016
+    },
+    {
+      "epoch": 1.685430463576159,
+      "grad_norm": 0.369140625,
+      "learning_rate": 6.627446264162226e-06,
+      "loss": 1.1333951950073242,
+      "step": 1018
+    },
+    {
+      "epoch": 1.6887417218543046,
+      "grad_norm": 0.50390625,
+      "learning_rate": 6.6100509526638295e-06,
+      "loss": 1.0682663917541504,
+      "step": 1020
+    },
+    {
+      "epoch": 1.6920529801324502,
+      "grad_norm": 0.4296875,
+      "learning_rate": 6.5926635045953026e-06,
+      "loss": 1.0695470571517944,
+      "step": 1022
+    },
+    {
+      "epoch": 1.695364238410596,
+      "grad_norm": 0.5625,
+      "learning_rate": 6.57528415171391e-06,
+      "loss": 1.1304333209991455,
+      "step": 1024
+    },
+    {
+      "epoch": 1.6986754966887418,
+      "grad_norm": 0.3671875,
+      "learning_rate": 6.55791312566902e-06,
+      "loss": 1.1207385063171387,
+      "step": 1026
+    },
+    {
+      "epoch": 1.7019867549668874,
+      "grad_norm": 0.64453125,
+      "learning_rate": 6.540550657999007e-06,
+      "loss": 1.0782655477523804,
+      "step": 1028
+    },
+    {
+      "epoch": 1.705298013245033,
+      "grad_norm": 0.314453125,
+      "learning_rate": 6.523196980128175e-06,
+      "loss": 1.0895800590515137,
+      "step": 1030
+    },
+    {
+      "epoch": 1.7086092715231787,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.5058523233636675e-06,
+      "loss": 1.0197545289993286,
+      "step": 1032
+    },
+    {
+      "epoch": 1.7119205298013245,
+      "grad_norm": 0.3984375,
+      "learning_rate": 6.4885169188923835e-06,
+      "loss": 1.1468507051467896,
+      "step": 1034
+    },
+    {
+      "epoch": 1.7152317880794703,
+      "grad_norm": 0.453125,
+      "learning_rate": 6.4711909977779034e-06,
+      "loss": 1.1051777601242065,
+      "step": 1036
+    },
+    {
+      "epoch": 1.718543046357616,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.4538747909574e-06,
+      "loss": 1.0961897373199463,
+      "step": 1038
+    },
+    {
+      "epoch": 1.7218543046357615,
+      "grad_norm": 0.40625,
+      "learning_rate": 6.436568529238562e-06,
+      "loss": 1.1184409856796265,
+      "step": 1040
+    },
+    {
+      "epoch": 1.7251655629139073,
+      "grad_norm": 2.53125,
+      "learning_rate": 6.419272443296529e-06,
+      "loss": 1.058405876159668,
+      "step": 1042
+    },
+    {
+      "epoch": 1.728476821192053,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.401986763670795e-06,
+      "loss": 1.1203678846359253,
+      "step": 1044
+    },
+    {
+      "epoch": 1.7317880794701987,
+      "grad_norm": 0.5,
+      "learning_rate": 6.3847117207621605e-06,
+      "loss": 1.105993390083313,
+      "step": 1046
+    },
+    {
+      "epoch": 1.7350993377483444,
+      "grad_norm": 0.33203125,
+      "learning_rate": 6.367447544829642e-06,
+      "loss": 1.1263104677200317,
+      "step": 1048
+    },
+    {
+      "epoch": 1.73841059602649,
+      "grad_norm": 0.380859375,
+      "learning_rate": 6.35019446598741e-06,
+      "loss": 1.2111080884933472,
+      "step": 1050
+    },
+    {
+      "epoch": 1.7417218543046358,
+      "grad_norm": 0.4921875,
+      "learning_rate": 6.3329527142017255e-06,
+      "loss": 1.1802841424942017,
+      "step": 1052
+    },
+    {
+      "epoch": 1.7450331125827816,
+      "grad_norm": 0.4609375,
+      "learning_rate": 6.31572251928787e-06,
+      "loss": 1.1829028129577637,
+      "step": 1054
+    },
+    {
+      "epoch": 1.7483443708609272,
+      "grad_norm": 0.5078125,
+      "learning_rate": 6.298504110907079e-06,
+      "loss": 1.1439435482025146,
+      "step": 1056
+    },
+    {
+      "epoch": 1.7516556291390728,
+      "grad_norm": 0.4140625,
+      "learning_rate": 6.281297718563491e-06,
+      "loss": 1.0960673093795776,
+      "step": 1058
+    },
+    {
+      "epoch": 1.7549668874172184,
+      "grad_norm": 0.34765625,
+      "learning_rate": 6.264103571601083e-06,
+      "loss": 1.1312483549118042,
+      "step": 1060
+    },
+    {
+      "epoch": 1.7582781456953642,
+      "grad_norm": 0.330078125,
+      "learning_rate": 6.246921899200609e-06,
+      "loss": 1.1018383502960205,
+      "step": 1062
+    },
+    {
+      "epoch": 1.76158940397351,
+      "grad_norm": 0.5859375,
+      "learning_rate": 6.229752930376553e-06,
+      "loss": 1.0974431037902832,
+      "step": 1064
+    },
+    {
+      "epoch": 1.7649006622516556,
+      "grad_norm": 0.8046875,
+      "learning_rate": 6.212596893974069e-06,
+      "loss": 1.0925394296646118,
+      "step": 1066
+    },
+    {
+      "epoch": 1.7682119205298013,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.195454018665944e-06,
+      "loss": 1.0865404605865479,
+      "step": 1068
+    },
+    {
+      "epoch": 1.771523178807947,
+      "grad_norm": 0.5390625,
+      "learning_rate": 6.178324532949529e-06,
+      "loss": 1.0787994861602783,
+      "step": 1070
+    },
+    {
+      "epoch": 1.7748344370860927,
+      "grad_norm": 0.296875,
+      "learning_rate": 6.16120866514371e-06,
+      "loss": 1.0901260375976562,
+      "step": 1072
+    },
+    {
+      "epoch": 1.7781456953642385,
+      "grad_norm": 0.306640625,
+      "learning_rate": 6.1441066433858665e-06,
+      "loss": 1.067402958869934,
+      "step": 1074
+    },
+    {
+      "epoch": 1.781456953642384,
+      "grad_norm": 0.2890625,
+      "learning_rate": 6.127018695628808e-06,
+      "loss": 1.0247200727462769,
+      "step": 1076
+    },
+    {
+      "epoch": 1.7847682119205297,
+      "grad_norm": 0.494140625,
+      "learning_rate": 6.109945049637773e-06,
+      "loss": 1.0284204483032227,
+      "step": 1078
+    },
+    {
+      "epoch": 1.7880794701986755,
+      "grad_norm": 0.3046875,
+      "learning_rate": 6.092885932987351e-06,
+      "loss": 1.0762344598770142,
+      "step": 1080
+    },
+    {
+      "epoch": 1.7913907284768213,
+      "grad_norm": 1.3125,
+      "learning_rate": 6.0758415730584845e-06,
+      "loss": 1.0676106214523315,
+      "step": 1082
+    },
+    {
+      "epoch": 1.794701986754967,
+      "grad_norm": 0.396484375,
+      "learning_rate": 6.058812197035418e-06,
+      "loss": 1.0068953037261963,
+      "step": 1084
+    },
+    {
+      "epoch": 1.7980132450331126,
+      "grad_norm": 0.458984375,
+      "learning_rate": 6.041798031902674e-06,
+      "loss": 1.1694005727767944,
+      "step": 1086
+    },
+    {
+      "epoch": 1.8013245033112582,
+      "grad_norm": 0.373046875,
+      "learning_rate": 6.024799304442037e-06,
+      "loss": 1.1191160678863525,
+      "step": 1088
+    },
+    {
+      "epoch": 1.804635761589404,
+      "grad_norm": 1.546875,
+      "learning_rate": 6.0078162412295115e-06,
+      "loss": 1.0951286554336548,
+      "step": 1090
+    },
+    {
+      "epoch": 1.8079470198675498,
+      "grad_norm": 0.55078125,
+      "learning_rate": 5.990849068632318e-06,
+      "loss": 1.1190029382705688,
+      "step": 1092
+    },
+    {
+      "epoch": 1.8112582781456954,
+      "grad_norm": 0.5,
+      "learning_rate": 5.973898012805875e-06,
+      "loss": 1.0913665294647217,
+      "step": 1094
+    },
+    {
+      "epoch": 1.814569536423841,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.9569632996907755e-06,
+      "loss": 1.069852590560913,
+      "step": 1096
+    },
+    {
+      "epoch": 1.8178807947019866,
+      "grad_norm": 0.31640625,
+      "learning_rate": 5.940045155009784e-06,
+      "loss": 1.0911487340927124,
+      "step": 1098
+    },
+    {
+      "epoch": 1.8211920529801324,
+      "grad_norm": 0.353515625,
+      "learning_rate": 5.923143804264822e-06,
+      "loss": 1.135908603668213,
+      "step": 1100
+    },
+    {
+      "epoch": 1.8245033112582782,
+      "grad_norm": 0.384765625,
+      "learning_rate": 5.906259472733964e-06,
+      "loss": 1.1487995386123657,
+      "step": 1102
+    },
+    {
+      "epoch": 1.8278145695364238,
+      "grad_norm": 0.419921875,
+      "learning_rate": 5.8893923854684386e-06,
+      "loss": 1.1677968502044678,
+      "step": 1104
+    },
+    {
+      "epoch": 1.8311258278145695,
+      "grad_norm": 0.439453125,
+      "learning_rate": 5.872542767289624e-06,
+      "loss": 1.102380633354187,
+      "step": 1106
+    },
+    {
+      "epoch": 1.8344370860927153,
+      "grad_norm": 2.78125,
+      "learning_rate": 5.855710842786053e-06,
+      "loss": 1.1117829084396362,
+      "step": 1108
+    },
+    {
+      "epoch": 1.8377483443708609,
+      "grad_norm": 0.58984375,
+      "learning_rate": 5.838896836310418e-06,
+      "loss": 1.1342206001281738,
+      "step": 1110
+    },
+    {
+      "epoch": 1.8410596026490067,
+      "grad_norm": 0.43359375,
+      "learning_rate": 5.822100971976584e-06,
+      "loss": 1.0774588584899902,
+      "step": 1112
+    },
+    {
+      "epoch": 1.8443708609271523,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.805323473656598e-06,
+      "loss": 1.0986900329589844,
+      "step": 1114
+    },
+    {
+      "epoch": 1.847682119205298,
+      "grad_norm": 0.91015625,
+      "learning_rate": 5.788564564977708e-06,
+      "loss": 1.1522040367126465,
+      "step": 1116
+    },
+    {
+      "epoch": 1.8509933774834437,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.771824469319379e-06,
+      "loss": 1.0625782012939453,
+      "step": 1118
+    },
+    {
+      "epoch": 1.8543046357615895,
+      "grad_norm": 0.56640625,
+      "learning_rate": 5.7551034098103185e-06,
+      "loss": 1.1141998767852783,
+      "step": 1120
+    },
+    {
+      "epoch": 1.8576158940397351,
+      "grad_norm": 0.7109375,
+      "learning_rate": 5.738401609325501e-06,
+      "loss": 1.0555064678192139,
+      "step": 1122
+    },
+    {
+      "epoch": 1.8609271523178808,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.721719290483198e-06,
+      "loss": 1.1195597648620605,
+      "step": 1124
+    },
+    {
+      "epoch": 1.8642384105960264,
+      "grad_norm": 0.41015625,
+      "learning_rate": 5.705056675642008e-06,
+      "loss": 1.1044063568115234,
+      "step": 1126
+    },
+    {
+      "epoch": 1.8675496688741722,
+      "grad_norm": 0.3125,
+      "learning_rate": 5.688413986897899e-06,
+      "loss": 1.2024643421173096,
+      "step": 1128
+    },
+    {
+      "epoch": 1.870860927152318,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.671791446081238e-06,
+      "loss": 1.1156952381134033,
+      "step": 1130
+    },
+    {
+      "epoch": 1.8741721854304636,
+      "grad_norm": 0.6875,
+      "learning_rate": 5.655189274753848e-06,
+      "loss": 1.1415760517120361,
+      "step": 1132
+    },
+    {
+      "epoch": 1.8774834437086092,
+      "grad_norm": 0.55859375,
+      "learning_rate": 5.638607694206041e-06,
+      "loss": 1.1045377254486084,
+      "step": 1134
+    },
+    {
+      "epoch": 1.8807947019867548,
+      "grad_norm": 0.546875,
+      "learning_rate": 5.622046925453673e-06,
+      "loss": 1.0766429901123047,
+      "step": 1136
+    },
+    {
+      "epoch": 1.8841059602649006,
+      "grad_norm": 0.34765625,
+      "learning_rate": 5.605507189235207e-06,
+      "loss": 1.1174238920211792,
+      "step": 1138
+    },
+    {
+      "epoch": 1.8874172185430464,
+      "grad_norm": 0.3203125,
+      "learning_rate": 5.5889887060087605e-06,
+      "loss": 1.1253938674926758,
+      "step": 1140
+    },
+    {
+      "epoch": 1.890728476821192,
+      "grad_norm": 1.203125,
+      "learning_rate": 5.572491695949165e-06,
+      "loss": 1.1572551727294922,
+      "step": 1142
+    },
+    {
+      "epoch": 1.8940397350993377,
+      "grad_norm": 0.65625,
+      "learning_rate": 5.556016378945044e-06,
+      "loss": 1.1192970275878906,
+      "step": 1144
+    },
+    {
+      "epoch": 1.8973509933774835,
+      "grad_norm": 0.333984375,
+      "learning_rate": 5.539562974595868e-06,
+      "loss": 1.1140743494033813,
+      "step": 1146
+    },
+    {
+      "epoch": 1.9006622516556293,
+      "grad_norm": 0.60546875,
+      "learning_rate": 5.523131702209037e-06,
+      "loss": 1.0318939685821533,
+      "step": 1148
+    },
+    {
+      "epoch": 1.903973509933775,
+      "grad_norm": 0.47265625,
+      "learning_rate": 5.506722780796953e-06,
+      "loss": 1.0299806594848633,
+      "step": 1150
+    },
+    {
+      "epoch": 1.9072847682119205,
+      "grad_norm": 0.416015625,
+      "learning_rate": 5.490336429074103e-06,
+      "loss": 1.1678529977798462,
+      "step": 1152
+    },
+    {
+      "epoch": 1.910596026490066,
+      "grad_norm": 0.5234375,
+      "learning_rate": 5.4739728654541415e-06,
+      "loss": 1.0364153385162354,
+      "step": 1154
+    },
+    {
+      "epoch": 1.913907284768212,
+      "grad_norm": 0.330078125,
+      "learning_rate": 5.457632308046978e-06,
+      "loss": 1.1044658422470093,
+      "step": 1156
+    },
+    {
+      "epoch": 1.9172185430463577,
+      "grad_norm": 0.46875,
+      "learning_rate": 5.441314974655879e-06,
+      "loss": 1.1027004718780518,
+      "step": 1158
+    },
+    {
+      "epoch": 1.9205298013245033,
+      "grad_norm": 0.431640625,
+      "learning_rate": 5.425021082774551e-06,
+      "loss": 1.1235613822937012,
+      "step": 1160
+    },
+    {
+      "epoch": 1.923841059602649,
+      "grad_norm": 0.40234375,
+      "learning_rate": 5.408750849584253e-06,
+      "loss": 1.0770944356918335,
+      "step": 1162
+    },
+    {
+      "epoch": 1.9271523178807946,
+      "grad_norm": 0.3671875,
+      "learning_rate": 5.392504491950898e-06,
+      "loss": 1.1669319868087769,
+      "step": 1164
+    },
+    {
+      "epoch": 1.9304635761589404,
+      "grad_norm": 0.341796875,
+      "learning_rate": 5.376282226422157e-06,
+      "loss": 1.064618706703186,
+      "step": 1166
+    },
+    {
+      "epoch": 1.9337748344370862,
+      "grad_norm": 0.404296875,
+      "learning_rate": 5.360084269224581e-06,
+      "loss": 1.0915920734405518,
+      "step": 1168
+    },
+    {
+      "epoch": 1.9370860927152318,
+      "grad_norm": 0.44140625,
+      "learning_rate": 5.343910836260715e-06,
+      "loss": 1.013867974281311,
+      "step": 1170
+    },
+    {
+      "epoch": 1.9403973509933774,
+      "grad_norm": 0.5,
+      "learning_rate": 5.327762143106218e-06,
+      "loss": 1.1231892108917236,
+      "step": 1172
+    },
+    {
+      "epoch": 1.9437086092715232,
+      "grad_norm": 0.357421875,
+      "learning_rate": 5.3116384050069994e-06,
+      "loss": 1.16791832447052,
+      "step": 1174
+    },
+    {
+      "epoch": 1.9470198675496688,
+      "grad_norm": 1.5078125,
+      "learning_rate": 5.295539836876331e-06,
+      "loss": 1.1662765741348267,
+      "step": 1176
+    },
+    {
+      "epoch": 1.9503311258278146,
+      "grad_norm": 1.0625,
+      "learning_rate": 5.279466653292004e-06,
+      "loss": 1.0979350805282593,
+      "step": 1178
+    },
+    {
+      "epoch": 1.9536423841059603,
+      "grad_norm": 0.314453125,
+      "learning_rate": 5.263419068493455e-06,
+      "loss": 1.1612277030944824,
+      "step": 1180
+    },
+    {
+      "epoch": 1.9569536423841059,
+      "grad_norm": 0.48046875,
+      "learning_rate": 5.2473972963789146e-06,
+      "loss": 1.1360162496566772,
+      "step": 1182
+    },
+    {
+      "epoch": 1.9602649006622517,
+      "grad_norm": 0.337890625,
+      "learning_rate": 5.231401550502558e-06,
+      "loss": 1.0562247037887573,
+      "step": 1184
+    },
+    {
+      "epoch": 1.9635761589403975,
+      "grad_norm": 0.400390625,
+      "learning_rate": 5.215432044071655e-06,
+      "loss": 1.0509085655212402,
+      "step": 1186
+    },
+    {
+      "epoch": 1.966887417218543,
+      "grad_norm": 0.28515625,
+      "learning_rate": 5.199488989943729e-06,
+      "loss": 1.0427687168121338,
+      "step": 1188
+    },
+    {
+      "epoch": 1.9701986754966887,
+      "grad_norm": 0.462890625,
+      "learning_rate": 5.183572600623721e-06,
+      "loss": 1.105067491531372,
+      "step": 1190
+    },
+    {
+      "epoch": 1.9735099337748343,
+      "grad_norm": 0.6171875,
+      "learning_rate": 5.167683088261162e-06,
+      "loss": 1.061687707901001,
+      "step": 1192
+    },
+    {
+      "epoch": 1.9768211920529801,
+      "grad_norm": 0.625,
+      "learning_rate": 5.151820664647333e-06,
+      "loss": 1.1441823244094849,
+      "step": 1194
+    },
+    {
+      "epoch": 1.980132450331126,
+      "grad_norm": 0.375,
+      "learning_rate": 5.135985541212451e-06,
+      "loss": 1.0769778490066528,
+      "step": 1196
+    },
+    {
+      "epoch": 1.9834437086092715,
+      "grad_norm": 0.7890625,
+      "learning_rate": 5.1201779290228525e-06,
+      "loss": 1.0631657838821411,
+      "step": 1198
+    },
+    {
+      "epoch": 1.9867549668874172,
+      "grad_norm": 1.234375,
+      "learning_rate": 5.104398038778169e-06,
+      "loss": 1.1530580520629883,
+      "step": 1200
+    },
+    {
+      "epoch": 1.9900662251655628,
+      "grad_norm": 0.26171875,
+      "learning_rate": 5.088646080808537e-06,
+      "loss": 1.1302320957183838,
+      "step": 1202
+    },
+    {
+      "epoch": 1.9933774834437086,
+      "grad_norm": 0.5078125,
+      "learning_rate": 5.072922265071775e-06,
+      "loss": 1.1501787900924683,
+      "step": 1204
+    },
+    {
+      "epoch": 1.9966887417218544,
+      "grad_norm": 0.35546875,
+      "learning_rate": 5.057226801150598e-06,
+      "loss": 1.0549312829971313,
+      "step": 1206
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.451171875,
+      "learning_rate": 5.041559898249817e-06,
+      "loss": 1.0718798637390137,
+      "step": 1208
+    },
+    {
+      "epoch": 2.0033112582781456,
+      "grad_norm": 0.33984375,
+      "learning_rate": 5.025921765193557e-06,
+      "loss": 1.0916215181350708,
+      "step": 1210
+    },
+    {
+      "epoch": 2.006622516556291,
+      "grad_norm": 0.287109375,
+      "learning_rate": 5.010312610422467e-06,
+      "loss": 1.073822259902954,
+      "step": 1212
+    },
+    {
+      "epoch": 2.0099337748344372,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.99473264199094e-06,
+      "loss": 1.042319893836975,
+      "step": 1214
+    },
+    {
+      "epoch": 2.013245033112583,
+      "grad_norm": 0.271484375,
+      "learning_rate": 4.979182067564358e-06,
+      "loss": 1.0481133460998535,
+      "step": 1216
+    },
+    {
+      "epoch": 2.0165562913907285,
+      "grad_norm": 0.31640625,
+      "learning_rate": 4.963661094416295e-06,
+      "loss": 0.9801614284515381,
+      "step": 1218
+    },
+    {
+      "epoch": 2.019867549668874,
+      "grad_norm": 0.91796875,
+      "learning_rate": 4.948169929425778e-06,
+      "loss": 1.0478323698043823,
+      "step": 1220
+    },
+    {
+      "epoch": 2.0231788079470197,
+      "grad_norm": 0.66796875,
+      "learning_rate": 4.9327087790745186e-06,
+      "loss": 1.085882306098938,
+      "step": 1222
+    },
+    {
+      "epoch": 2.0264900662251657,
+      "grad_norm": 0.498046875,
+      "learning_rate": 4.917277849444166e-06,
+      "loss": 1.0048208236694336,
+      "step": 1224
+    },
+    {
+      "epoch": 2.0298013245033113,
+      "grad_norm": 0.5390625,
+      "learning_rate": 4.9018773462135525e-06,
+      "loss": 1.0288286209106445,
+      "step": 1226
+    },
+    {
+      "epoch": 2.033112582781457,
+      "grad_norm": 0.3125,
+      "learning_rate": 4.886507474655965e-06,
+      "loss": 1.102409839630127,
+      "step": 1228
+    },
+    {
+      "epoch": 2.0364238410596025,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.87116843963639e-06,
+      "loss": 1.0566478967666626,
+      "step": 1230
+    },
+    {
+      "epoch": 2.0397350993377485,
+      "grad_norm": 0.30078125,
+      "learning_rate": 4.855860445608805e-06,
+      "loss": 1.0917094945907593,
+      "step": 1232
+    },
+    {
+      "epoch": 2.043046357615894,
+      "grad_norm": 0.431640625,
+      "learning_rate": 4.840583696613433e-06,
+      "loss": 1.0331615209579468,
+      "step": 1234
+    },
+    {
+      "epoch": 2.0463576158940397,
+      "grad_norm": 0.45703125,
+      "learning_rate": 4.825338396274038e-06,
+      "loss": 1.0467015504837036,
+      "step": 1236
+    },
+    {
+      "epoch": 2.0496688741721854,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.810124747795203e-06,
+      "loss": 1.0292361974716187,
+      "step": 1238
+    },
+    {
+      "epoch": 2.052980132450331,
+      "grad_norm": 0.369140625,
+      "learning_rate": 4.7949429539596205e-06,
+      "loss": 0.981902539730072,
+      "step": 1240
+    },
+    {
+      "epoch": 2.056291390728477,
+      "grad_norm": 0.408203125,
+      "learning_rate": 4.779793217125398e-06,
+      "loss": 1.0525181293487549,
+      "step": 1242
+    },
+    {
+      "epoch": 2.0596026490066226,
+      "grad_norm": 0.380859375,
+      "learning_rate": 4.76467573922335e-06,
+      "loss": 1.0744459629058838,
+      "step": 1244
+    },
+    {
+      "epoch": 2.062913907284768,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.749590721754318e-06,
+      "loss": 1.0127346515655518,
+      "step": 1246
+    },
+    {
+      "epoch": 2.066225165562914,
+      "grad_norm": 0.447265625,
+      "learning_rate": 4.734538365786471e-06,
+      "loss": 0.9119741320610046,
+      "step": 1248
+    },
+    {
+      "epoch": 2.0695364238410594,
+      "grad_norm": 0.353515625,
+      "learning_rate": 4.71951887195264e-06,
+      "loss": 1.0265223979949951,
+      "step": 1250
+    },
+    {
+      "epoch": 2.0728476821192054,
+      "grad_norm": 0.69140625,
+      "learning_rate": 4.704532440447632e-06,
+      "loss": 1.0805940628051758,
+      "step": 1252
+    },
+    {
+      "epoch": 2.076158940397351,
+      "grad_norm": 0.462890625,
+      "learning_rate": 4.689579271025565e-06,
+      "loss": 1.0568530559539795,
+      "step": 1254
+    },
+    {
+      "epoch": 2.0794701986754967,
+      "grad_norm": 0.423828125,
+      "learning_rate": 4.674659562997213e-06,
+      "loss": 1.1478904485702515,
+      "step": 1256
+    },
+    {
+      "epoch": 2.0827814569536423,
+      "grad_norm": 0.4375,
+      "learning_rate": 4.6597735152273365e-06,
+      "loss": 0.9818710088729858,
+      "step": 1258
+    },
+    {
+      "epoch": 2.0860927152317883,
+      "grad_norm": 0.515625,
+      "learning_rate": 4.644921326132045e-06,
+      "loss": 1.059829831123352,
+      "step": 1260
+    },
+    {
+      "epoch": 2.089403973509934,
+      "grad_norm": 0.70703125,
+      "learning_rate": 4.630103193676136e-06,
+      "loss": 1.0128612518310547,
+      "step": 1262
+    },
+    {
+      "epoch": 2.0927152317880795,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.61531931537048e-06,
+      "loss": 1.0629007816314697,
+      "step": 1264
+    },
+    {
+      "epoch": 2.096026490066225,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.6005698882693595e-06,
+      "loss": 1.0301984548568726,
+      "step": 1266
+    },
+    {
+      "epoch": 2.0993377483443707,
+      "grad_norm": 0.349609375,
+      "learning_rate": 4.585855108967869e-06,
+      "loss": 1.0227930545806885,
+      "step": 1268
+    },
+    {
+      "epoch": 2.1026490066225167,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.571175173599276e-06,
+      "loss": 1.0368671417236328,
+      "step": 1270
+    },
+    {
+      "epoch": 2.1059602649006623,
+      "grad_norm": 0.396484375,
+      "learning_rate": 4.5565302778324106e-06,
+      "loss": 0.9873349070549011,
+      "step": 1272
+    },
+    {
+      "epoch": 2.109271523178808,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.541920616869069e-06,
+      "loss": 1.007344365119934,
+      "step": 1274
+    },
+    {
+      "epoch": 2.1125827814569536,
+      "grad_norm": 0.38671875,
+      "learning_rate": 4.527346385441395e-06,
+      "loss": 1.0187866687774658,
+      "step": 1276
+    },
+    {
+      "epoch": 2.115894039735099,
+      "grad_norm": 0.41015625,
+      "learning_rate": 4.512807777809299e-06,
+      "loss": 1.082865595817566,
+      "step": 1278
+    },
+    {
+      "epoch": 2.119205298013245,
+      "grad_norm": 0.54296875,
+      "learning_rate": 4.4983049877578516e-06,
+      "loss": 1.0283514261245728,
+      "step": 1280
+    },
+    {
+      "epoch": 2.122516556291391,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.483838208594723e-06,
+      "loss": 1.1053177118301392,
+      "step": 1282
+    },
+    {
+      "epoch": 2.1258278145695364,
+      "grad_norm": 0.419921875,
+      "learning_rate": 4.469407633147583e-06,
+      "loss": 1.0836362838745117,
+      "step": 1284
+    },
+    {
+      "epoch": 2.129139072847682,
+      "grad_norm": 0.4453125,
+      "learning_rate": 4.455013453761553e-06,
+      "loss": 1.0553747415542603,
+      "step": 1286
+    },
+    {
+      "epoch": 2.1324503311258276,
+      "grad_norm": 0.5078125,
+      "learning_rate": 4.440655862296622e-06,
+      "loss": 1.0284078121185303,
+      "step": 1288
+    },
+    {
+      "epoch": 2.1357615894039736,
+      "grad_norm": 0.62109375,
+      "learning_rate": 4.426335050125107e-06,
+      "loss": 0.914071798324585,
+      "step": 1290
+    },
+    {
+      "epoch": 2.1390728476821192,
+      "grad_norm": 0.33203125,
+      "learning_rate": 4.412051208129087e-06,
+      "loss": 1.0339797735214233,
+      "step": 1292
+    },
+    {
+      "epoch": 2.142384105960265,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.3978045266978655e-06,
+      "loss": 1.0838464498519897,
+      "step": 1294
+    },
+    {
+      "epoch": 2.1456953642384105,
+      "grad_norm": 0.734375,
+      "learning_rate": 4.383595195725441e-06,
+      "loss": 1.0508580207824707,
+      "step": 1296
+    },
+    {
+      "epoch": 2.1490066225165565,
+      "grad_norm": 0.7734375,
+      "learning_rate": 4.3694234046079586e-06,
+      "loss": 1.0503878593444824,
+      "step": 1298
+    },
+    {
+      "epoch": 2.152317880794702,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.355289342241201e-06,
+      "loss": 1.066772222518921,
+      "step": 1300
+    },
+    {
+      "epoch": 2.1556291390728477,
+      "grad_norm": 0.390625,
+      "learning_rate": 4.3411931970180586e-06,
+      "loss": 1.0136038064956665,
+      "step": 1302
+    },
+    {
+      "epoch": 2.1589403973509933,
+      "grad_norm": 0.34375,
+      "learning_rate": 4.327135156826031e-06,
+      "loss": 1.0277198553085327,
+      "step": 1304
+    },
+    {
+      "epoch": 2.162251655629139,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.313115409044709e-06,
+      "loss": 1.0602567195892334,
+      "step": 1306
+    },
+    {
+      "epoch": 2.165562913907285,
+      "grad_norm": 0.41796875,
+      "learning_rate": 4.299134140543291e-06,
+      "loss": 1.0680055618286133,
+      "step": 1308
+    },
+    {
+      "epoch": 2.1688741721854305,
+      "grad_norm": 0.60546875,
+      "learning_rate": 4.285191537678081e-06,
+      "loss": 1.0317363739013672,
+      "step": 1310
+    },
+    {
+      "epoch": 2.172185430463576,
+      "grad_norm": 0.56640625,
+      "learning_rate": 4.271287786290006e-06,
+      "loss": 1.0057586431503296,
+      "step": 1312
+    },
+    {
+      "epoch": 2.1754966887417218,
+      "grad_norm": 0.443359375,
+      "learning_rate": 4.257423071702149e-06,
+      "loss": 1.0330718755722046,
+      "step": 1314
+    },
+    {
+      "epoch": 2.1788079470198674,
+      "grad_norm": 0.306640625,
+      "learning_rate": 4.243597578717265e-06,
+      "loss": 1.0199156999588013,
+      "step": 1316
+    },
+    {
+      "epoch": 2.1821192052980134,
+      "grad_norm": 0.361328125,
+      "learning_rate": 4.229811491615329e-06,
+      "loss": 0.9906125664710999,
+      "step": 1318
+    },
+    {
+      "epoch": 2.185430463576159,
+      "grad_norm": 0.302734375,
+      "learning_rate": 4.216064994151071e-06,
+      "loss": 1.1189707517623901,
+      "step": 1320
+    },
+    {
+      "epoch": 2.1887417218543046,
+      "grad_norm": 0.51953125,
+      "learning_rate": 4.202358269551536e-06,
+      "loss": 0.996783971786499,
+      "step": 1322
+    },
+    {
+      "epoch": 2.19205298013245,
+      "grad_norm": 0.31640625,
+      "learning_rate": 4.188691500513628e-06,
+      "loss": 1.0981279611587524,
+      "step": 1324
+    },
+    {
+      "epoch": 2.195364238410596,
+      "grad_norm": 0.33984375,
+      "learning_rate": 4.175064869201694e-06,
+      "loss": 1.0822064876556396,
+      "step": 1326
+    },
+    {
+      "epoch": 2.198675496688742,
+      "grad_norm": 0.58203125,
+      "learning_rate": 4.161478557245076e-06,
+      "loss": 0.9716603755950928,
+      "step": 1328
+    },
+    {
+      "epoch": 2.2019867549668874,
+      "grad_norm": 0.72265625,
+      "learning_rate": 4.147932745735708e-06,
+      "loss": 1.045888900756836,
+      "step": 1330
+    },
+    {
+      "epoch": 2.205298013245033,
+      "grad_norm": 0.359375,
+      "learning_rate": 4.134427615225686e-06,
+      "loss": 1.0230597257614136,
+      "step": 1332
+    },
+    {
+      "epoch": 2.2086092715231787,
+      "grad_norm": 0.53125,
+      "learning_rate": 4.120963345724871e-06,
+      "loss": 1.0504512786865234,
+      "step": 1334
+    },
+    {
+      "epoch": 2.2119205298013247,
+      "grad_norm": 0.39453125,
+      "learning_rate": 4.107540116698491e-06,
+      "loss": 1.048678994178772,
+      "step": 1336
+    },
+    {
+      "epoch": 2.2152317880794703,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.0941581070647416e-06,
+      "loss": 1.0520901679992676,
+      "step": 1338
+    },
+    {
+      "epoch": 2.218543046357616,
+      "grad_norm": 0.466796875,
+      "learning_rate": 4.080817495192409e-06,
+      "loss": 1.0306859016418457,
+      "step": 1340
+    },
+    {
+      "epoch": 2.2218543046357615,
+      "grad_norm": 0.421875,
+      "learning_rate": 4.067518458898486e-06,
+      "loss": 1.067014217376709,
+      "step": 1342
+    },
+    {
+      "epoch": 2.225165562913907,
+      "grad_norm": 0.357421875,
+      "learning_rate": 4.054261175445805e-06,
+      "loss": 0.98519366979599,
+      "step": 1344
+    },
+    {
+      "epoch": 2.228476821192053,
+      "grad_norm": 0.89453125,
+      "learning_rate": 4.041045821540672e-06,
+      "loss": 1.0907796621322632,
+      "step": 1346
+    },
+    {
+      "epoch": 2.2317880794701987,
+      "grad_norm": 0.310546875,
+      "learning_rate": 4.027872573330523e-06,
+      "loss": 0.9753696918487549,
+      "step": 1348
+    },
+    {
+      "epoch": 2.2350993377483444,
+      "grad_norm": 0.40625,
+      "learning_rate": 4.014741606401557e-06,
+      "loss": 1.0596659183502197,
+      "step": 1350
+    },
+    {
+      "epoch": 2.23841059602649,
+      "grad_norm": 0.30859375,
+      "learning_rate": 4.001653095776411e-06,
+      "loss": 1.0401997566223145,
+      "step": 1352
+    },
+    {
+      "epoch": 2.241721854304636,
+      "grad_norm": 0.79296875,
+      "learning_rate": 3.988607215911823e-06,
+      "loss": 1.0707045793533325,
+      "step": 1354
+    },
+    {
+      "epoch": 2.2450331125827816,
+      "grad_norm": 0.310546875,
+      "learning_rate": 3.975604140696301e-06,
+      "loss": 1.0365581512451172,
+      "step": 1356
+    },
+    {
+      "epoch": 2.248344370860927,
+      "grad_norm": 0.384765625,
+      "learning_rate": 3.962644043447818e-06,
+      "loss": 1.0854417085647583,
+      "step": 1358
+    },
+    {
+      "epoch": 2.251655629139073,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.9497270969114825e-06,
+      "loss": 1.0579876899719238,
+      "step": 1360
+    },
+    {
+      "epoch": 2.2549668874172184,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.9368534732572605e-06,
+      "loss": 1.1018257141113281,
+      "step": 1362
+    },
+    {
+      "epoch": 2.258278145695364,
+      "grad_norm": 0.8671875,
+      "learning_rate": 3.924023344077652e-06,
+      "loss": 1.0348024368286133,
+      "step": 1364
+    },
+    {
+      "epoch": 2.26158940397351,
+      "grad_norm": 0.345703125,
+      "learning_rate": 3.911236880385433e-06,
+      "loss": 1.0398342609405518,
+      "step": 1366
+    },
+    {
+      "epoch": 2.2649006622516556,
+      "grad_norm": 0.423828125,
+      "learning_rate": 3.898494252611356e-06,
+      "loss": 1.0437736511230469,
+      "step": 1368
+    },
+    {
+      "epoch": 2.2682119205298013,
+      "grad_norm": 0.609375,
+      "learning_rate": 3.885795630601883e-06,
+      "loss": 1.0380507707595825,
+      "step": 1370
+    },
+    {
+      "epoch": 2.271523178807947,
+      "grad_norm": 0.73046875,
+      "learning_rate": 3.873141183616928e-06,
+      "loss": 1.052344560623169,
+      "step": 1372
+    },
+    {
+      "epoch": 2.274834437086093,
+      "grad_norm": 0.30859375,
+      "learning_rate": 3.860531080327592e-06,
+      "loss": 0.9509193897247314,
+      "step": 1374
+    },
+    {
+      "epoch": 2.2781456953642385,
+      "grad_norm": 0.65625,
+      "learning_rate": 3.847965488813924e-06,
+      "loss": 1.1158931255340576,
+      "step": 1376
+    },
+    {
+      "epoch": 2.281456953642384,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.8354445765626675e-06,
+      "loss": 1.0930873155593872,
+      "step": 1378
+    },
+    {
+      "epoch": 2.2847682119205297,
+      "grad_norm": 0.76171875,
+      "learning_rate": 3.822968510465046e-06,
+      "loss": 1.0825984477996826,
+      "step": 1380
+    },
+    {
+      "epoch": 2.2880794701986753,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.810537456814521e-06,
+      "loss": 1.1007579565048218,
+      "step": 1382
+    },
+    {
+      "epoch": 2.2913907284768213,
+      "grad_norm": 0.74609375,
+      "learning_rate": 3.7981515813045858e-06,
+      "loss": 0.9964556097984314,
+      "step": 1384
+    },
+    {
+      "epoch": 2.294701986754967,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.785811049026554e-06,
+      "loss": 1.0913865566253662,
+      "step": 1386
+    },
+    {
+      "epoch": 2.2980132450331126,
+      "grad_norm": 0.4296875,
+      "learning_rate": 3.773516024467362e-06,
+      "loss": 1.0170255899429321,
+      "step": 1388
+    },
+    {
+      "epoch": 2.301324503311258,
+      "grad_norm": 0.2890625,
+      "learning_rate": 3.76126667150737e-06,
+      "loss": 0.9141421318054199,
+      "step": 1390
+    },
+    {
+      "epoch": 2.304635761589404,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.7490631534181817e-06,
+      "loss": 1.0807045698165894,
+      "step": 1392
+    },
+    {
+      "epoch": 2.30794701986755,
+      "grad_norm": 0.95703125,
+      "learning_rate": 3.7369056328604735e-06,
+      "loss": 1.0416104793548584,
+      "step": 1394
+    },
+    {
+      "epoch": 2.3112582781456954,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.7247942718818125e-06,
+      "loss": 0.938892126083374,
+      "step": 1396
+    },
+    {
+      "epoch": 2.314569536423841,
+      "grad_norm": 0.447265625,
+      "learning_rate": 3.7127292319145126e-06,
+      "loss": 1.0145360231399536,
+      "step": 1398
+    },
+    {
+      "epoch": 2.3178807947019866,
+      "grad_norm": 0.37109375,
+      "learning_rate": 3.7007106737734678e-06,
+      "loss": 1.0101609230041504,
+      "step": 1400
+    },
+    {
+      "epoch": 2.321192052980132,
+      "grad_norm": 0.96484375,
+      "learning_rate": 3.688738757654023e-06,
+      "loss": 1.0995274782180786,
+      "step": 1402
+    },
+    {
+      "epoch": 2.3245033112582782,
+      "grad_norm": 0.4140625,
+      "learning_rate": 3.6768136431298274e-06,
+      "loss": 1.0471631288528442,
+      "step": 1404
+    },
+    {
+      "epoch": 2.327814569536424,
+      "grad_norm": 0.625,
+      "learning_rate": 3.664935489150711e-06,
+      "loss": 1.104356050491333,
+      "step": 1406
+    },
+    {
+      "epoch": 2.3311258278145695,
+      "grad_norm": 0.38671875,
+      "learning_rate": 3.653104454040569e-06,
+      "loss": 1.0226653814315796,
+      "step": 1408
+    },
+    {
+      "epoch": 2.334437086092715,
+      "grad_norm": 0.6875,
+      "learning_rate": 3.6413206954952474e-06,
+      "loss": 1.0596727132797241,
+      "step": 1410
+    },
+    {
+      "epoch": 2.337748344370861,
+      "grad_norm": 0.455078125,
+      "learning_rate": 3.629584370580448e-06,
+      "loss": 1.0580487251281738,
+      "step": 1412
+    },
+    {
+      "epoch": 2.3410596026490067,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.6178956357296224e-06,
+      "loss": 1.0152729749679565,
+      "step": 1414
+    },
+    {
+      "epoch": 2.3443708609271523,
+      "grad_norm": 0.4453125,
+      "learning_rate": 3.6062546467419e-06,
+      "loss": 0.9716925024986267,
+      "step": 1416
+    },
+    {
+      "epoch": 2.347682119205298,
+      "grad_norm": 0.302734375,
+      "learning_rate": 3.5946615587800015e-06,
+      "loss": 1.0557975769042969,
+      "step": 1418
+    },
+    {
+      "epoch": 2.3509933774834435,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.58311652636818e-06,
+      "loss": 0.9758134484291077,
+      "step": 1420
+    },
+    {
+      "epoch": 2.3543046357615895,
+      "grad_norm": 0.390625,
+      "learning_rate": 3.5716197033901525e-06,
+      "loss": 1.0848175287246704,
+      "step": 1422
+    },
+    {
+      "epoch": 2.357615894039735,
+      "grad_norm": 0.6484375,
+      "learning_rate": 3.56017124308705e-06,
+      "loss": 1.0970638990402222,
+      "step": 1424
+    },
+    {
+      "epoch": 2.3609271523178808,
+      "grad_norm": 0.28515625,
+      "learning_rate": 3.5487712980553854e-06,
+      "loss": 1.0622012615203857,
+      "step": 1426
+    },
+    {
+      "epoch": 2.3642384105960264,
+      "grad_norm": 0.63671875,
+      "learning_rate": 3.537420020245004e-06,
+      "loss": 1.0812063217163086,
+      "step": 1428
+    },
+    {
+      "epoch": 2.3675496688741724,
+      "grad_norm": 0.365234375,
+      "learning_rate": 3.526117560957071e-06,
+      "loss": 0.9727545380592346,
+      "step": 1430
+    },
+    {
+      "epoch": 2.370860927152318,
+      "grad_norm": 0.30078125,
+      "learning_rate": 3.5148640708420447e-06,
+      "loss": 1.0257072448730469,
+      "step": 1432
+    },
+    {
+      "epoch": 2.3741721854304636,
+      "grad_norm": 0.388671875,
+      "learning_rate": 3.5036596998976788e-06,
+      "loss": 1.0375983715057373,
+      "step": 1434
+    },
+    {
+      "epoch": 2.377483443708609,
+      "grad_norm": 0.51953125,
+      "learning_rate": 3.492504597467012e-06,
+      "loss": 1.144998550415039,
+      "step": 1436
+    },
+    {
+      "epoch": 2.380794701986755,
+      "grad_norm": 0.44140625,
+      "learning_rate": 3.4813989122363892e-06,
+      "loss": 1.0561128854751587,
+      "step": 1438
+    },
+    {
+      "epoch": 2.384105960264901,
+      "grad_norm": 0.30859375,
+      "learning_rate": 3.4703427922334675e-06,
+      "loss": 1.039366364479065,
+      "step": 1440
+    },
+    {
+      "epoch": 2.3874172185430464,
+      "grad_norm": 1.0078125,
+      "learning_rate": 3.459336384825255e-06,
+      "loss": 1.0634366273880005,
+      "step": 1442
+    },
+    {
+      "epoch": 2.390728476821192,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.4483798367161363e-06,
+      "loss": 1.0396180152893066,
+      "step": 1444
+    },
+    {
+      "epoch": 2.3940397350993377,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.4374732939459216e-06,
+      "loss": 1.0524084568023682,
+      "step": 1446
+    },
+    {
+      "epoch": 2.3973509933774833,
+      "grad_norm": 0.361328125,
+      "learning_rate": 3.4266169018879037e-06,
+      "loss": 1.1443829536437988,
+      "step": 1448
+    },
+    {
+      "epoch": 2.4006622516556293,
+      "grad_norm": 0.380859375,
+      "learning_rate": 3.4158108052469115e-06,
+      "loss": 1.0547349452972412,
+      "step": 1450
+    },
+    {
+      "epoch": 2.403973509933775,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.405055148057391e-06,
+      "loss": 1.09640371799469,
+      "step": 1452
+    },
+    {
+      "epoch": 2.4072847682119205,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.394350073681477e-06,
+      "loss": 1.0381419658660889,
+      "step": 1454
+    },
+    {
+      "epoch": 2.410596026490066,
+      "grad_norm": 0.67578125,
+      "learning_rate": 3.383695724807092e-06,
+      "loss": 1.0096808671951294,
+      "step": 1456
+    },
+    {
+      "epoch": 2.4139072847682117,
+      "grad_norm": 0.3359375,
+      "learning_rate": 3.3730922434460305e-06,
+      "loss": 0.9652847051620483,
+      "step": 1458
+    },
+    {
+      "epoch": 2.4172185430463577,
+      "grad_norm": 0.333984375,
+      "learning_rate": 3.3625397709320806e-06,
+      "loss": 1.0267952680587769,
+      "step": 1460
+    },
+    {
+      "epoch": 2.4205298013245033,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.352038447919129e-06,
+      "loss": 1.1289455890655518,
+      "step": 1462
+    },
+    {
+      "epoch": 2.423841059602649,
+      "grad_norm": 0.421875,
+      "learning_rate": 3.3415884143792916e-06,
+      "loss": 1.0683571100234985,
+      "step": 1464
+    },
+    {
+      "epoch": 2.4271523178807946,
+      "grad_norm": 0.84375,
+      "learning_rate": 3.331189809601052e-06,
+      "loss": 1.0286537408828735,
+      "step": 1466
+    },
+    {
+      "epoch": 2.4304635761589406,
+      "grad_norm": 0.6640625,
+      "learning_rate": 3.320842772187393e-06,
+      "loss": 1.026946783065796,
+      "step": 1468
+    },
+    {
+      "epoch": 2.433774834437086,
+      "grad_norm": 0.58203125,
+      "learning_rate": 3.3105474400539603e-06,
+      "loss": 1.0303407907485962,
+      "step": 1470
+    },
+    {
+      "epoch": 2.437086092715232,
+      "grad_norm": 0.392578125,
+      "learning_rate": 3.3003039504272156e-06,
+      "loss": 0.9855753779411316,
+      "step": 1472
+    },
+    {
+      "epoch": 2.4403973509933774,
+      "grad_norm": 1.4140625,
+      "learning_rate": 3.2901124398426203e-06,
+      "loss": 0.9761983752250671,
+      "step": 1474
+    },
+    {
+      "epoch": 2.443708609271523,
+      "grad_norm": 0.34765625,
+      "learning_rate": 3.2799730441427975e-06,
+      "loss": 1.0473487377166748,
+      "step": 1476
+    },
+    {
+      "epoch": 2.447019867549669,
+      "grad_norm": 0.279296875,
+      "learning_rate": 3.2698858984757387e-06,
+      "loss": 1.0138802528381348,
+      "step": 1478
+    },
+    {
+      "epoch": 2.4503311258278146,
+      "grad_norm": 2.4375,
+      "learning_rate": 3.2598511372929895e-06,
+      "loss": 1.146519660949707,
+      "step": 1480
+    },
+    {
+      "epoch": 2.4536423841059603,
+      "grad_norm": 1.4921875,
+      "learning_rate": 3.249868894347866e-06,
+      "loss": 1.0525798797607422,
+      "step": 1482
+    },
+    {
+      "epoch": 2.456953642384106,
+      "grad_norm": 0.310546875,
+      "learning_rate": 3.239939302693667e-06,
+      "loss": 1.0201683044433594,
+      "step": 1484
+    },
+    {
+      "epoch": 2.460264900662252,
+      "grad_norm": 0.39453125,
+      "learning_rate": 3.230062494681905e-06,
+      "loss": 1.0771936178207397,
+      "step": 1486
+    },
+    {
+      "epoch": 2.4635761589403975,
+      "grad_norm": 0.62890625,
+      "learning_rate": 3.2202386019605365e-06,
+      "loss": 1.1066436767578125,
+      "step": 1488
+    },
+    {
+      "epoch": 2.466887417218543,
+      "grad_norm": 0.35546875,
+      "learning_rate": 3.21046775547221e-06,
+      "loss": 0.9931922554969788,
+      "step": 1490
+    },
+    {
+      "epoch": 2.4701986754966887,
+      "grad_norm": 0.5546875,
+      "learning_rate": 3.2007500854525235e-06,
+      "loss": 1.0755813121795654,
+      "step": 1492
+    },
+    {
+      "epoch": 2.4735099337748343,
+      "grad_norm": 0.44921875,
+      "learning_rate": 3.1910857214282827e-06,
+      "loss": 0.9923776984214783,
+      "step": 1494
+    },
+    {
+      "epoch": 2.47682119205298,
+      "grad_norm": 2.484375,
+      "learning_rate": 3.181474792215783e-06,
+      "loss": 1.1028560400009155,
+      "step": 1496
+    },
+    {
+      "epoch": 2.480132450331126,
+      "grad_norm": 0.46484375,
+      "learning_rate": 3.1719174259190804e-06,
+      "loss": 1.0993995666503906,
+      "step": 1498
+    },
+    {
+      "epoch": 2.4834437086092715,
+      "grad_norm": 0.337890625,
+      "learning_rate": 3.1624137499282962e-06,
+      "loss": 1.0792993307113647,
+      "step": 1500
+    },
+    {
+      "epoch": 2.486754966887417,
+      "grad_norm": 0.4765625,
+      "learning_rate": 3.1529638909179116e-06,
+      "loss": 1.0069727897644043,
+      "step": 1502
+    },
+    {
+      "epoch": 2.4900662251655628,
+      "grad_norm": 0.3203125,
+      "learning_rate": 3.1435679748450805e-06,
+      "loss": 1.0655741691589355,
+      "step": 1504
+    },
+    {
+      "epoch": 2.493377483443709,
+      "grad_norm": 0.40234375,
+      "learning_rate": 3.1342261269479545e-06,
+      "loss": 1.008230209350586,
+      "step": 1506
+    },
+    {
+      "epoch": 2.4966887417218544,
+      "grad_norm": 0.61328125,
+      "learning_rate": 3.124938471744006e-06,
+      "loss": 1.0085946321487427,
+      "step": 1508
+    },
+    {
+      "epoch": 2.5,
+      "grad_norm": 0.376953125,
+      "learning_rate": 3.1157051330283802e-06,
+      "loss": 1.1044775247573853,
+      "step": 1510
+    },
+    {
+      "epoch": 2.5033112582781456,
+      "grad_norm": 0.703125,
+      "learning_rate": 3.1065262338722284e-06,
+      "loss": 1.0872358083724976,
+      "step": 1512
+    },
+    {
+      "epoch": 2.506622516556291,
+      "grad_norm": 0.53515625,
+      "learning_rate": 3.0974018966210855e-06,
+      "loss": 1.109554409980774,
+      "step": 1514
+    },
+    {
+      "epoch": 2.5099337748344372,
+      "grad_norm": 1.8984375,
+      "learning_rate": 3.088332242893225e-06,
+      "loss": 1.1079566478729248,
+      "step": 1516
+    },
+    {
+      "epoch": 2.513245033112583,
+      "grad_norm": 0.70703125,
+      "learning_rate": 3.0793173935780485e-06,
+      "loss": 1.1112803220748901,
+      "step": 1518
+    },
+    {
+      "epoch": 2.5165562913907285,
+      "grad_norm": 0.357421875,
+      "learning_rate": 3.070357468834467e-06,
+      "loss": 1.1184546947479248,
+      "step": 1520
+    },
+    {
+      "epoch": 2.519867549668874,
+      "grad_norm": 0.59375,
+      "learning_rate": 3.061452588089301e-06,
+      "loss": 1.0849051475524902,
+      "step": 1522
+    },
+    {
+      "epoch": 2.52317880794702,
+      "grad_norm": 0.400390625,
+      "learning_rate": 3.0526028700356934e-06,
+      "loss": 1.0355197191238403,
+      "step": 1524
+    },
+    {
+      "epoch": 2.5264900662251657,
+      "grad_norm": 1.3515625,
+      "learning_rate": 3.0438084326315195e-06,
+      "loss": 1.0885202884674072,
+      "step": 1526
+    },
+    {
+      "epoch": 2.5298013245033113,
+      "grad_norm": 0.96875,
+      "learning_rate": 3.0350693930978228e-06,
+      "loss": 1.0239211320877075,
+      "step": 1528
+    },
+    {
+      "epoch": 2.533112582781457,
+      "grad_norm": 0.31640625,
+      "learning_rate": 3.0263858679172452e-06,
+      "loss": 0.9950514435768127,
+      "step": 1530
+    },
+    {
+      "epoch": 2.5364238410596025,
+      "grad_norm": 0.7421875,
+      "learning_rate": 3.0177579728324794e-06,
+      "loss": 1.006035327911377,
+      "step": 1532
+    },
+    {
+      "epoch": 2.539735099337748,
+      "grad_norm": 0.64453125,
+      "learning_rate": 3.0091858228447235e-06,
+      "loss": 1.0517382621765137,
+      "step": 1534
+    },
+    {
+      "epoch": 2.543046357615894,
+      "grad_norm": 0.427734375,
+      "learning_rate": 3.0006695322121533e-06,
+      "loss": 1.0941227674484253,
+      "step": 1536
+    },
+    {
+      "epoch": 2.5463576158940397,
+      "grad_norm": 1.140625,
+      "learning_rate": 2.9922092144483897e-06,
+      "loss": 1.1483227014541626,
+      "step": 1538
+    },
+    {
+      "epoch": 2.5496688741721854,
+      "grad_norm": 1.3828125,
+      "learning_rate": 2.9838049823209965e-06,
+      "loss": 1.0933794975280762,
+      "step": 1540
+    },
+    {
+      "epoch": 2.5529801324503314,
+      "grad_norm": 0.55859375,
+      "learning_rate": 2.9754569478499683e-06,
+      "loss": 1.0416162014007568,
+      "step": 1542
+    },
+    {
+      "epoch": 2.556291390728477,
+      "grad_norm": 0.58984375,
+      "learning_rate": 2.9671652223062422e-06,
+      "loss": 1.0642694234848022,
+      "step": 1544
+    },
+    {
+      "epoch": 2.5596026490066226,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.958929916210215e-06,
+      "loss": 1.0878373384475708,
+      "step": 1546
+    },
+    {
+      "epoch": 2.562913907284768,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.9507511393302675e-06,
+      "loss": 1.1222093105316162,
+      "step": 1548
+    },
+    {
+      "epoch": 2.566225165562914,
+      "grad_norm": 2.359375,
+      "learning_rate": 2.9426290006813043e-06,
+      "loss": 1.0746393203735352,
+      "step": 1550
+    },
+    {
+      "epoch": 2.5695364238410594,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.9345636085232965e-06,
+      "loss": 1.1305041313171387,
+      "step": 1552
+    },
+    {
+      "epoch": 2.5728476821192054,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.9265550703598437e-06,
+      "loss": 1.0597765445709229,
+      "step": 1554
+    },
+    {
+      "epoch": 2.576158940397351,
+      "grad_norm": 0.404296875,
+      "learning_rate": 2.9186034929367357e-06,
+      "loss": 1.111279010772705,
+      "step": 1556
+    },
+    {
+      "epoch": 2.5794701986754967,
+      "grad_norm": 1.09375,
+      "learning_rate": 2.910708982240538e-06,
+      "loss": 1.0620553493499756,
+      "step": 1558
+    },
+    {
+      "epoch": 2.5827814569536423,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.9028716434971685e-06,
+      "loss": 0.9780864715576172,
+      "step": 1560
+    },
+    {
+      "epoch": 2.5860927152317883,
+      "grad_norm": 0.443359375,
+      "learning_rate": 2.895091581170501e-06,
+      "loss": 1.040873408317566,
+      "step": 1562
+    },
+    {
+      "epoch": 2.589403973509934,
+      "grad_norm": 0.484375,
+      "learning_rate": 2.887368898960976e-06,
+      "loss": 1.1218167543411255,
+      "step": 1564
+    },
+    {
+      "epoch": 2.5927152317880795,
+      "grad_norm": 0.30078125,
+      "learning_rate": 2.87970369980421e-06,
+      "loss": 1.0425817966461182,
+      "step": 1566
+    },
+    {
+      "epoch": 2.596026490066225,
+      "grad_norm": 1.03125,
+      "learning_rate": 2.872096085869631e-06,
+      "loss": 1.157275915145874,
+      "step": 1568
+    },
+    {
+      "epoch": 2.5993377483443707,
+      "grad_norm": 0.318359375,
+      "learning_rate": 2.8645461585591106e-06,
+      "loss": 1.0908249616622925,
+      "step": 1570
+    },
+    {
+      "epoch": 2.6026490066225163,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.8570540185056215e-06,
+      "loss": 1.0348210334777832,
+      "step": 1572
+    },
+    {
+      "epoch": 2.6059602649006623,
+      "grad_norm": 0.375,
+      "learning_rate": 2.84961976557188e-06,
+      "loss": 1.02035391330719,
+      "step": 1574
+    },
+    {
+      "epoch": 2.609271523178808,
+      "grad_norm": 0.353515625,
+      "learning_rate": 2.8422434988490363e-06,
+      "loss": 1.078000545501709,
+      "step": 1576
+    },
+    {
+      "epoch": 2.6125827814569536,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.8349253166553342e-06,
+      "loss": 1.1093229055404663,
+      "step": 1578
+    },
+    {
+      "epoch": 2.6158940397350996,
+      "grad_norm": 0.400390625,
+      "learning_rate": 2.8276653165348124e-06,
+      "loss": 1.1029249429702759,
+      "step": 1580
+    },
+    {
+      "epoch": 2.619205298013245,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.820463595256001e-06,
+      "loss": 1.0593039989471436,
+      "step": 1582
+    },
+    {
+      "epoch": 2.622516556291391,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.813320248810631e-06,
+      "loss": 1.1073625087738037,
+      "step": 1584
+    },
+    {
+      "epoch": 2.6258278145695364,
+      "grad_norm": 0.8671875,
+      "learning_rate": 2.8062353724123554e-06,
+      "loss": 0.997706413269043,
+      "step": 1586
+    },
+    {
+      "epoch": 2.629139072847682,
+      "grad_norm": 0.322265625,
+      "learning_rate": 2.799209060495479e-06,
+      "loss": 1.0507164001464844,
+      "step": 1588
+    },
+    {
+      "epoch": 2.6324503311258276,
+      "grad_norm": 0.4296875,
+      "learning_rate": 2.792241406713703e-06,
+      "loss": 1.0913563966751099,
+      "step": 1590
+    },
+    {
+      "epoch": 2.6357615894039736,
+      "grad_norm": 0.279296875,
+      "learning_rate": 2.785332503938872e-06,
+      "loss": 1.0073961019515991,
+      "step": 1592
+    },
+    {
+      "epoch": 2.6390728476821192,
+      "grad_norm": 0.60546875,
+      "learning_rate": 2.7784824442597397e-06,
+      "loss": 0.9370295405387878,
+      "step": 1594
+    },
+    {
+      "epoch": 2.642384105960265,
+      "grad_norm": 0.63671875,
+      "learning_rate": 2.77169131898074e-06,
+      "loss": 1.0220690965652466,
+      "step": 1596
+    },
+    {
+      "epoch": 2.6456953642384105,
+      "grad_norm": 0.42578125,
+      "learning_rate": 2.7649592186207713e-06,
+      "loss": 1.018792748451233,
+      "step": 1598
+    },
+    {
+      "epoch": 2.6490066225165565,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.7582862329119876e-06,
+      "loss": 0.9850583076477051,
+      "step": 1600
+    },
+    {
+      "epoch": 2.652317880794702,
+      "grad_norm": 0.62109375,
+      "learning_rate": 2.751672450798603e-06,
+      "loss": 1.1173804998397827,
+      "step": 1602
+    },
+    {
+      "epoch": 2.6556291390728477,
+      "grad_norm": 0.46875,
+      "learning_rate": 2.745117960435711e-06,
+      "loss": 0.9903439283370972,
+      "step": 1604
+    },
+    {
+      "epoch": 2.6589403973509933,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.7386228491880994e-06,
+      "loss": 1.0194555521011353,
+      "step": 1606
+    },
+    {
+      "epoch": 2.662251655629139,
+      "grad_norm": 0.416015625,
+      "learning_rate": 2.7321872036290987e-06,
+      "loss": 1.010141134262085,
+      "step": 1608
+    },
+    {
+      "epoch": 2.6655629139072845,
+      "grad_norm": 0.31640625,
+      "learning_rate": 2.725811109539415e-06,
+      "loss": 1.0798771381378174,
+      "step": 1610
+    },
+    {
+      "epoch": 2.6688741721854305,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.719494651905997e-06,
+      "loss": 0.9975613355636597,
+      "step": 1612
+    },
+    {
+      "epoch": 2.672185430463576,
+      "grad_norm": 0.37109375,
+      "learning_rate": 2.7132379149208987e-06,
+      "loss": 1.0866070985794067,
+      "step": 1614
+    },
+    {
+      "epoch": 2.6754966887417218,
+      "grad_norm": 0.421875,
+      "learning_rate": 2.707040981980156e-06,
+      "loss": 1.053512692451477,
+      "step": 1616
+    },
+    {
+      "epoch": 2.678807947019868,
+      "grad_norm": 0.3515625,
+      "learning_rate": 2.700903935682677e-06,
+      "loss": 1.0476690530776978,
+      "step": 1618
+    },
+    {
+      "epoch": 2.6821192052980134,
+      "grad_norm": 0.78515625,
+      "learning_rate": 2.6948268578291427e-06,
+      "loss": 1.074185848236084,
+      "step": 1620
+    },
+    {
+      "epoch": 2.685430463576159,
+      "grad_norm": 3.90625,
+      "learning_rate": 2.688809829420914e-06,
+      "loss": 1.0318957567214966,
+      "step": 1622
+    },
+    {
+      "epoch": 2.6887417218543046,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.682852930658951e-06,
+      "loss": 1.0365896224975586,
+      "step": 1624
+    },
+    {
+      "epoch": 2.69205298013245,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.676956240942749e-06,
+      "loss": 0.9852394461631775,
+      "step": 1626
+    },
+    {
+      "epoch": 2.695364238410596,
+      "grad_norm": 0.7265625,
+      "learning_rate": 2.6711198388692742e-06,
+      "loss": 1.0080279111862183,
+      "step": 1628
+    },
+    {
+      "epoch": 2.698675496688742,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.6653438022319214e-06,
+      "loss": 1.0633718967437744,
+      "step": 1630
+    },
+    {
+      "epoch": 2.7019867549668874,
+      "grad_norm": 0.625,
+      "learning_rate": 2.659628208019472e-06,
+      "loss": 1.099438190460205,
+      "step": 1632
+    },
+    {
+      "epoch": 2.705298013245033,
+      "grad_norm": 0.71875,
+      "learning_rate": 2.6539731324150723e-06,
+      "loss": 0.9704924821853638,
+      "step": 1634
+    },
+    {
+      "epoch": 2.7086092715231787,
+      "grad_norm": 0.310546875,
+      "learning_rate": 2.6483786507952167e-06,
+      "loss": 0.9796273708343506,
+      "step": 1636
+    },
+    {
+      "epoch": 2.7119205298013247,
+      "grad_norm": 0.431640625,
+      "learning_rate": 2.642844837728739e-06,
+      "loss": 1.0218433141708374,
+      "step": 1638
+    },
+    {
+      "epoch": 2.7152317880794703,
+      "grad_norm": 0.5234375,
+      "learning_rate": 2.637371766975826e-06,
+      "loss": 0.9966109395027161,
+      "step": 1640
+    },
+    {
+      "epoch": 2.718543046357616,
+      "grad_norm": 0.466796875,
+      "learning_rate": 2.6319595114870276e-06,
+      "loss": 1.0867879390716553,
+      "step": 1642
+    },
+    {
+      "epoch": 2.7218543046357615,
+      "grad_norm": 0.40234375,
+      "learning_rate": 2.62660814340229e-06,
+      "loss": 1.021042823791504,
+      "step": 1644
+    },
+    {
+      "epoch": 2.725165562913907,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.621317734049987e-06,
+      "loss": 1.0033985376358032,
+      "step": 1646
+    },
+    {
+      "epoch": 2.7284768211920527,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.6160883539459785e-06,
+      "loss": 1.008457064628601,
+      "step": 1648
+    },
+    {
+      "epoch": 2.7317880794701987,
+      "grad_norm": 0.5078125,
+      "learning_rate": 2.610920072792662e-06,
+      "loss": 1.0715134143829346,
+      "step": 1650
+    },
+    {
+      "epoch": 2.7350993377483444,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.6058129594780482e-06,
+      "loss": 1.0398625135421753,
+      "step": 1652
+    },
+    {
+      "epoch": 2.73841059602649,
+      "grad_norm": 0.90625,
+      "learning_rate": 2.600767082074842e-06,
+      "loss": 1.050590991973877,
+      "step": 1654
+    },
+    {
+      "epoch": 2.741721854304636,
+      "grad_norm": 0.39453125,
+      "learning_rate": 2.5957825078395354e-06,
+      "loss": 1.068892478942871,
+      "step": 1656
+    },
+    {
+      "epoch": 2.7450331125827816,
+      "grad_norm": 0.3984375,
+      "learning_rate": 2.590859303211511e-06,
+      "loss": 1.0792319774627686,
+      "step": 1658
+    },
+    {
+      "epoch": 2.748344370860927,
+      "grad_norm": 0.41015625,
+      "learning_rate": 2.5859975338121514e-06,
+      "loss": 1.0801936388015747,
+      "step": 1660
+    },
+    {
+      "epoch": 2.751655629139073,
+      "grad_norm": 0.73046875,
+      "learning_rate": 2.581197264443977e-06,
+      "loss": 1.0703206062316895,
+      "step": 1662
+    },
+    {
+      "epoch": 2.7549668874172184,
+      "grad_norm": 1.15625,
+      "learning_rate": 2.5764585590897676e-06,
+      "loss": 1.1012338399887085,
+      "step": 1664
+    },
+    {
+      "epoch": 2.758278145695364,
+      "grad_norm": 0.439453125,
+      "learning_rate": 2.5717814809117207e-06,
+      "loss": 1.0293129682540894,
+      "step": 1666
+    },
+    {
+      "epoch": 2.76158940397351,
+      "grad_norm": 0.47265625,
+      "learning_rate": 2.567166092250602e-06,
+      "loss": 1.0774444341659546,
+      "step": 1668
+    },
+    {
+      "epoch": 2.7649006622516556,
+      "grad_norm": 0.36328125,
+      "learning_rate": 2.5626124546249205e-06,
+      "loss": 1.062659740447998,
+      "step": 1670
+    },
+    {
+      "epoch": 2.7682119205298013,
+      "grad_norm": 0.7109375,
+      "learning_rate": 2.558120628730104e-06,
+      "loss": 0.9909599423408508,
+      "step": 1672
+    },
+    {
+      "epoch": 2.7715231788079473,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.553690674437692e-06,
+      "loss": 1.0109676122665405,
+      "step": 1674
+    },
+    {
+      "epoch": 2.774834437086093,
+      "grad_norm": 1.3671875,
+      "learning_rate": 2.5493226507945386e-06,
+      "loss": 1.0199503898620605,
+      "step": 1676
+    },
+    {
+      "epoch": 2.7781456953642385,
+      "grad_norm": 0.34765625,
+      "learning_rate": 2.545016616022023e-06,
+      "loss": 1.0352284908294678,
+      "step": 1678
+    },
+    {
+      "epoch": 2.781456953642384,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.5407726275152766e-06,
+      "loss": 1.067415714263916,
+      "step": 1680
+    },
+    {
+      "epoch": 2.7847682119205297,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.5365907418424134e-06,
+      "loss": 1.017134428024292,
+      "step": 1682
+    },
+    {
+      "epoch": 2.7880794701986753,
+      "grad_norm": 0.578125,
+      "learning_rate": 2.5324710147437826e-06,
+      "loss": 1.0925308465957642,
+      "step": 1684
+    },
+    {
+      "epoch": 2.7913907284768213,
+      "grad_norm": 0.51953125,
+      "learning_rate": 2.5284135011312176e-06,
+      "loss": 1.0486555099487305,
+      "step": 1686
+    },
+    {
+      "epoch": 2.794701986754967,
+      "grad_norm": 0.38671875,
+      "learning_rate": 2.5244182550873128e-06,
+      "loss": 1.0359236001968384,
+      "step": 1688
+    },
+    {
+      "epoch": 2.7980132450331126,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.5204853298646943e-06,
+      "loss": 1.095540165901184,
+      "step": 1690
+    },
+    {
+      "epoch": 2.801324503311258,
+      "grad_norm": 0.4375,
+      "learning_rate": 2.5166147778853166e-06,
+      "loss": 1.0864918231964111,
+      "step": 1692
+    },
+    {
+      "epoch": 2.804635761589404,
+      "grad_norm": 0.515625,
+      "learning_rate": 2.512806650739761e-06,
+      "loss": 1.026558518409729,
+      "step": 1694
+    },
+    {
+      "epoch": 2.80794701986755,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.5090609991865466e-06,
+      "loss": 1.148924708366394,
+      "step": 1696
+    },
+    {
+      "epoch": 2.8112582781456954,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.5053778731514587e-06,
+      "loss": 0.9244303703308105,
+      "step": 1698
+    },
+    {
+      "epoch": 2.814569536423841,
+      "grad_norm": 1.078125,
+      "learning_rate": 2.5017573217268762e-06,
+      "loss": 1.0666385889053345,
+      "step": 1700
+    },
+    {
+      "epoch": 2.8178807947019866,
+      "grad_norm": 0.412109375,
+      "learning_rate": 2.4981993931711243e-06,
+      "loss": 1.0128355026245117,
+      "step": 1702
+    },
+    {
+      "epoch": 2.821192052980132,
+      "grad_norm": 0.44140625,
+      "learning_rate": 2.494704134907825e-06,
+      "loss": 1.0081136226654053,
+      "step": 1704
+    },
+    {
+      "epoch": 2.8245033112582782,
+      "grad_norm": 0.490234375,
+      "learning_rate": 2.491271593525272e-06,
+      "loss": 1.0477370023727417,
+      "step": 1706
+    },
+    {
+      "epoch": 2.827814569536424,
+      "grad_norm": 0.9453125,
+      "learning_rate": 2.487901814775802e-06,
+      "loss": 1.1048303842544556,
+      "step": 1708
+    },
+    {
+      "epoch": 2.8311258278145695,
+      "grad_norm": 0.546875,
+      "learning_rate": 2.484594843575192e-06,
+      "loss": 1.0124863386154175,
+      "step": 1710
+    },
+    {
+      "epoch": 2.8344370860927155,
+      "grad_norm": 0.474609375,
+      "learning_rate": 2.4813507240020545e-06,
+      "loss": 0.9982864260673523,
+      "step": 1712
+    },
+    {
+      "epoch": 2.837748344370861,
+      "grad_norm": 0.671875,
+      "learning_rate": 2.478169499297254e-06,
+      "loss": 1.0384469032287598,
+      "step": 1714
+    },
+    {
+      "epoch": 2.8410596026490067,
+      "grad_norm": 0.66796875,
+      "learning_rate": 2.475051211863329e-06,
+      "loss": 1.0405954122543335,
+      "step": 1716
+    },
+    {
+      "epoch": 2.8443708609271523,
+      "grad_norm": 0.369140625,
+      "learning_rate": 2.471995903263928e-06,
+      "loss": 1.0376397371292114,
+      "step": 1718
+    },
+    {
+      "epoch": 2.847682119205298,
+      "grad_norm": 0.6640625,
+      "learning_rate": 2.4690036142232548e-06,
+      "loss": 1.0328714847564697,
+      "step": 1720
+    },
+    {
+      "epoch": 2.8509933774834435,
+      "grad_norm": 0.32421875,
+      "learning_rate": 2.4660743846255235e-06,
+      "loss": 1.0731256008148193,
+      "step": 1722
+    },
+    {
+      "epoch": 2.8543046357615895,
+      "grad_norm": 0.330078125,
+      "learning_rate": 2.463208253514432e-06,
+      "loss": 0.9934253692626953,
+      "step": 1724
+    },
+    {
+      "epoch": 2.857615894039735,
+      "grad_norm": 0.2734375,
+      "learning_rate": 2.460405259092636e-06,
+      "loss": 1.0437015295028687,
+      "step": 1726
+    },
+    {
+      "epoch": 2.8609271523178808,
+      "grad_norm": 0.5,
+      "learning_rate": 2.4576654387212443e-06,
+      "loss": 1.0205596685409546,
+      "step": 1728
+    },
+    {
+      "epoch": 2.8642384105960264,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.4549888289193174e-06,
+      "loss": 1.0206191539764404,
+      "step": 1730
+    },
+    {
+      "epoch": 2.8675496688741724,
+      "grad_norm": 0.306640625,
+      "learning_rate": 2.4523754653633833e-06,
+      "loss": 1.005540132522583,
+      "step": 1732
+    },
+    {
+      "epoch": 2.870860927152318,
+      "grad_norm": 0.56640625,
+      "learning_rate": 2.4498253828869606e-06,
+      "loss": 1.1127885580062866,
+      "step": 1734
+    },
+    {
+      "epoch": 2.8741721854304636,
+      "grad_norm": 0.53515625,
+      "learning_rate": 2.447338615480094e-06,
+      "loss": 1.0119414329528809,
+      "step": 1736
+    },
+    {
+      "epoch": 2.877483443708609,
+      "grad_norm": 1.203125,
+      "learning_rate": 2.4449151962889023e-06,
+      "loss": 0.9939915537834167,
+      "step": 1738
+    },
+    {
+      "epoch": 2.880794701986755,
+      "grad_norm": 0.349609375,
+      "learning_rate": 2.442555157615136e-06,
+      "loss": 1.1005805730819702,
+      "step": 1740
+    },
+    {
+      "epoch": 2.8841059602649004,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.440258530915747e-06,
+      "loss": 1.0766074657440186,
+      "step": 1742
+    },
+    {
+      "epoch": 2.8874172185430464,
+      "grad_norm": 0.34375,
+      "learning_rate": 2.438025346802467e-06,
+      "loss": 1.0019570589065552,
+      "step": 1744
+    },
+    {
+      "epoch": 2.890728476821192,
+      "grad_norm": 0.3671875,
+      "learning_rate": 2.4358556350414055e-06,
+      "loss": 1.0643962621688843,
+      "step": 1746
+    },
+    {
+      "epoch": 2.8940397350993377,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.433749424552646e-06,
+      "loss": 0.9804382920265198,
+      "step": 1748
+    },
+    {
+      "epoch": 2.8973509933774837,
+      "grad_norm": 0.76953125,
+      "learning_rate": 2.4317067434098644e-06,
+      "loss": 1.0500622987747192,
+      "step": 1750
+    },
+    {
+      "epoch": 2.9006622516556293,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.4297276188399545e-06,
+      "loss": 1.082712173461914,
+      "step": 1752
+    },
+    {
+      "epoch": 2.903973509933775,
+      "grad_norm": 0.392578125,
+      "learning_rate": 2.4278120772226646e-06,
+      "loss": 0.9876247644424438,
+      "step": 1754
+    },
+    {
+      "epoch": 2.9072847682119205,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.4259601440902454e-06,
+      "loss": 1.0972561836242676,
+      "step": 1756
+    },
+    {
+      "epoch": 2.910596026490066,
+      "grad_norm": 0.341796875,
+      "learning_rate": 2.4241718441271104e-06,
+      "loss": 0.9264360666275024,
+      "step": 1758
+    },
+    {
+      "epoch": 2.9139072847682117,
+      "grad_norm": 0.35546875,
+      "learning_rate": 2.4224472011695073e-06,
+      "loss": 1.1301442384719849,
+      "step": 1760
+    },
+    {
+      "epoch": 2.9172185430463577,
+      "grad_norm": 0.333984375,
+      "learning_rate": 2.420786238205199e-06,
+      "loss": 0.9778857231140137,
+      "step": 1762
+    },
+    {
+      "epoch": 2.9205298013245033,
+      "grad_norm": 0.57421875,
+      "learning_rate": 2.419188977373158e-06,
+      "loss": 1.0964045524597168,
+      "step": 1764
+    },
+    {
+      "epoch": 2.923841059602649,
+      "grad_norm": 0.380859375,
+      "learning_rate": 2.4176554399632713e-06,
+      "loss": 1.0145188570022583,
+      "step": 1766
+    },
+    {
+      "epoch": 2.9271523178807946,
+      "grad_norm": 0.53125,
+      "learning_rate": 2.4161856464160563e-06,
+      "loss": 1.0378607511520386,
+      "step": 1768
+    },
+    {
+      "epoch": 2.9304635761589406,
+      "grad_norm": 0.5,
+      "learning_rate": 2.4147796163223895e-06,
+      "loss": 0.9963395595550537,
+      "step": 1770
+    },
+    {
+      "epoch": 2.933774834437086,
+      "grad_norm": 0.5703125,
+      "learning_rate": 2.413437368423242e-06,
+      "loss": 1.0542939901351929,
+      "step": 1772
+    },
+    {
+      "epoch": 2.937086092715232,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.4121589206094356e-06,
+      "loss": 1.083484411239624,
+      "step": 1774
+    },
+    {
+      "epoch": 2.9403973509933774,
+      "grad_norm": 0.361328125,
+      "learning_rate": 2.4109442899213987e-06,
+      "loss": 0.9940117001533508,
+      "step": 1776
+    },
+    {
+      "epoch": 2.943708609271523,
+      "grad_norm": 0.396484375,
+      "learning_rate": 2.4097934925489416e-06,
+      "loss": 1.0698211193084717,
+      "step": 1778
+    },
+    {
+      "epoch": 2.9470198675496686,
+      "grad_norm": 0.28125,
+      "learning_rate": 2.40870654383104e-06,
+      "loss": 1.026236653327942,
+      "step": 1780
+    },
+    {
+      "epoch": 2.9503311258278146,
+      "grad_norm": 0.296875,
+      "learning_rate": 2.4076834582556318e-06,
+      "loss": 1.0298240184783936,
+      "step": 1782
+    },
+    {
+      "epoch": 2.9536423841059603,
+      "grad_norm": 0.45703125,
+      "learning_rate": 2.4067242494594234e-06,
+      "loss": 0.9907808899879456,
+      "step": 1784
+    },
+    {
+      "epoch": 2.956953642384106,
+      "grad_norm": 0.328125,
+      "learning_rate": 2.4058289302277066e-06,
+      "loss": 0.9502798914909363,
+      "step": 1786
+    },
+    {
+      "epoch": 2.960264900662252,
+      "grad_norm": 0.43359375,
+      "learning_rate": 2.4049975124941905e-06,
+      "loss": 1.0373636484146118,
+      "step": 1788
+    },
+    {
+      "epoch": 2.9635761589403975,
+      "grad_norm": 0.796875,
+      "learning_rate": 2.40423000734084e-06,
+      "loss": 1.010378360748291,
+      "step": 1790
+    },
+    {
+      "epoch": 2.966887417218543,
+      "grad_norm": 0.3359375,
+      "learning_rate": 2.403526424997731e-06,
+      "loss": 1.0332858562469482,
+      "step": 1792
+    },
+    {
+      "epoch": 2.9701986754966887,
+      "grad_norm": 0.75,
+      "learning_rate": 2.4028867748429104e-06,
+      "loss": 1.0457154512405396,
+      "step": 1794
+    },
+    {
+      "epoch": 2.9735099337748343,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.4023110654022747e-06,
+      "loss": 1.130010724067688,
+      "step": 1796
+    },
+    {
+      "epoch": 2.97682119205298,
+      "grad_norm": 0.390625,
+      "learning_rate": 2.401799304349454e-06,
+      "loss": 1.1422545909881592,
+      "step": 1798
+    },
+    {
+      "epoch": 2.980132450331126,
+      "grad_norm": 0.470703125,
+      "learning_rate": 2.40135149850571e-06,
+      "loss": 0.9642534255981445,
+      "step": 1800
+    },
+    {
+      "epoch": 2.9834437086092715,
+      "grad_norm": 0.478515625,
+      "learning_rate": 2.4009676538398466e-06,
+      "loss": 0.9972424507141113,
+      "step": 1802
+    },
+    {
+      "epoch": 2.986754966887417,
+      "grad_norm": 0.83984375,
+      "learning_rate": 2.400647775468129e-06,
+      "loss": 1.1147444248199463,
+      "step": 1804
+    },
+    {
+      "epoch": 2.9900662251655628,
+      "grad_norm": 0.388671875,
+      "learning_rate": 2.4003918676542145e-06,
+      "loss": 0.9947775602340698,
+      "step": 1806
+    },
+    {
+      "epoch": 2.993377483443709,
+      "grad_norm": 0.3828125,
+      "learning_rate": 2.4001999338090986e-06,
+      "loss": 1.022848129272461,
+      "step": 1808
+    },
+    {
+      "epoch": 2.9966887417218544,
+      "grad_norm": 0.375,
+      "learning_rate": 2.400071976491067e-06,
+      "loss": 1.0713096857070923,
+      "step": 1810
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.4140625,
+      "learning_rate": 2.4000079974056627e-06,
+      "loss": 1.0778021812438965,
+      "step": 1812
+    },
+    {
+      "epoch": 3.0,
+      "step": 1812,
+      "total_flos": 4.871467377855824e+18,
+      "train_loss": 1.1715838705861805,
+      "train_runtime": 22979.7614,
+      "train_samples_per_second": 2.523,
+      "train_steps_per_second": 0.079
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1812,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.871467377855824e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}