diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4342 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.3161361141602634,
+  "eval_steps": 50,
+  "global_step": 600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0021953896816684962,
+      "grad_norm": 163.60501272566069,
+      "learning_rate": 0.0,
+      "loss": 1.3923,
+      "step": 1
+    },
+    {
+      "epoch": 0.0043907793633369925,
+      "grad_norm": 170.58640230921233,
+      "learning_rate": 7.2992700729927e-09,
+      "loss": 1.281,
+      "step": 2
+    },
+    {
+      "epoch": 0.006586169045005488,
+      "grad_norm": 131.10623897546836,
+      "learning_rate": 1.45985401459854e-08,
+      "loss": 1.3551,
+      "step": 3
+    },
+    {
+      "epoch": 0.008781558726673985,
+      "grad_norm": 178.96806532547603,
+      "learning_rate": 2.1897810218978102e-08,
+      "loss": 1.3577,
+      "step": 4
+    },
+    {
+      "epoch": 0.010976948408342482,
+      "grad_norm": 132.31431642741137,
+      "learning_rate": 2.91970802919708e-08,
+      "loss": 1.3135,
+      "step": 5
+    },
+    {
+      "epoch": 0.013172338090010977,
+      "grad_norm": 135.62398872133667,
+      "learning_rate": 3.64963503649635e-08,
+      "loss": 1.305,
+      "step": 6
+    },
+    {
+      "epoch": 0.015367727771679473,
+      "grad_norm": 128.75782005700225,
+      "learning_rate": 4.3795620437956203e-08,
+      "loss": 1.4486,
+      "step": 7
+    },
+    {
+      "epoch": 0.01756311745334797,
+      "grad_norm": 138.3771195817014,
+      "learning_rate": 5.10948905109489e-08,
+      "loss": 1.5554,
+      "step": 8
+    },
+    {
+      "epoch": 0.019758507135016465,
+      "grad_norm": 132.69136904564107,
+      "learning_rate": 5.83941605839416e-08,
+      "loss": 1.4041,
+      "step": 9
+    },
+    {
+      "epoch": 0.021953896816684963,
+      "grad_norm": 135.7212626140043,
+      "learning_rate": 6.569343065693431e-08,
+      "loss": 1.2969,
+      "step": 10
+    },
+    {
+      "epoch": 0.024149286498353458,
+      "grad_norm": 150.3565406204981,
+      "learning_rate": 7.2992700729927e-08,
+      "loss": 1.3152,
+      "step": 11
+    },
+    {
+      "epoch": 0.026344676180021953,
+      "grad_norm": 151.3406121199072,
+      "learning_rate": 8.029197080291971e-08,
+      "loss": 1.488,
+      "step": 12
+    },
+    {
+      "epoch": 0.02854006586169045,
+      "grad_norm": 124.37914565637091,
+      "learning_rate": 8.759124087591241e-08,
+      "loss": 1.244,
+      "step": 13
+    },
+    {
+      "epoch": 0.030735455543358946,
+      "grad_norm": 142.5907569544806,
+      "learning_rate": 9.48905109489051e-08,
+      "loss": 1.2863,
+      "step": 14
+    },
+    {
+      "epoch": 0.03293084522502744,
+      "grad_norm": 122.30421005497061,
+      "learning_rate": 1.021897810218978e-07,
+      "loss": 1.3103,
+      "step": 15
+    },
+    {
+      "epoch": 0.03512623490669594,
+      "grad_norm": 200.79885863608268,
+      "learning_rate": 1.0948905109489052e-07,
+      "loss": 1.4004,
+      "step": 16
+    },
+    {
+      "epoch": 0.03732162458836443,
+      "grad_norm": 167.14991660211373,
+      "learning_rate": 1.167883211678832e-07,
+      "loss": 1.2775,
+      "step": 17
+    },
+    {
+      "epoch": 0.03951701427003293,
+      "grad_norm": 131.28993539446094,
+      "learning_rate": 1.240875912408759e-07,
+      "loss": 1.3523,
+      "step": 18
+    },
+    {
+      "epoch": 0.04171240395170143,
+      "grad_norm": 162.3600505958806,
+      "learning_rate": 1.3138686131386862e-07,
+      "loss": 1.3361,
+      "step": 19
+    },
+    {
+      "epoch": 0.043907793633369926,
+      "grad_norm": 133.12292640519846,
+      "learning_rate": 1.386861313868613e-07,
+      "loss": 1.2538,
+      "step": 20
+    },
+    {
+      "epoch": 0.04610318331503842,
+      "grad_norm": 138.78896167890522,
+      "learning_rate": 1.45985401459854e-07,
+      "loss": 1.352,
+      "step": 21
+    },
+    {
+      "epoch": 0.048298572996706916,
+      "grad_norm": 139.69189042882115,
+      "learning_rate": 1.532846715328467e-07,
+      "loss": 1.5154,
+      "step": 22
+    },
+    {
+      "epoch": 0.050493962678375415,
+      "grad_norm": 137.39489076785355,
+      "learning_rate": 1.6058394160583942e-07,
+      "loss": 1.4735,
+      "step": 23
+    },
+    {
+      "epoch": 0.052689352360043906,
+      "grad_norm": 129.9057654211869,
+      "learning_rate": 1.678832116788321e-07,
+      "loss": 1.4751,
+      "step": 24
+    },
+    {
+      "epoch": 0.054884742041712405,
+      "grad_norm": 142.17880454005854,
+      "learning_rate": 1.7518248175182481e-07,
+      "loss": 1.4576,
+      "step": 25
+    },
+    {
+      "epoch": 0.0570801317233809,
+      "grad_norm": 213.94665712809794,
+      "learning_rate": 1.824817518248175e-07,
+      "loss": 1.5012,
+      "step": 26
+    },
+    {
+      "epoch": 0.059275521405049394,
+      "grad_norm": 150.33832122564692,
+      "learning_rate": 1.897810218978102e-07,
+      "loss": 1.4739,
+      "step": 27
+    },
+    {
+      "epoch": 0.06147091108671789,
+      "grad_norm": 135.59843657422903,
+      "learning_rate": 1.9708029197080292e-07,
+      "loss": 1.3503,
+      "step": 28
+    },
+    {
+      "epoch": 0.06366630076838639,
+      "grad_norm": 140.7138855818933,
+      "learning_rate": 2.043795620437956e-07,
+      "loss": 1.2969,
+      "step": 29
+    },
+    {
+      "epoch": 0.06586169045005488,
+      "grad_norm": 203.85681204065455,
+      "learning_rate": 2.116788321167883e-07,
+      "loss": 1.4389,
+      "step": 30
+    },
+    {
+      "epoch": 0.06805708013172337,
+      "grad_norm": 129.70436787634256,
+      "learning_rate": 2.1897810218978103e-07,
+      "loss": 1.352,
+      "step": 31
+    },
+    {
+      "epoch": 0.07025246981339188,
+      "grad_norm": 170.170655877601,
+      "learning_rate": 2.2627737226277372e-07,
+      "loss": 1.4311,
+      "step": 32
+    },
+    {
+      "epoch": 0.07244785949506037,
+      "grad_norm": 167.7752606625615,
+      "learning_rate": 2.335766423357664e-07,
+      "loss": 1.2723,
+      "step": 33
+    },
+    {
+      "epoch": 0.07464324917672886,
+      "grad_norm": 130.878840340516,
+      "learning_rate": 2.408759124087591e-07,
+      "loss": 1.3537,
+      "step": 34
+    },
+    {
+      "epoch": 0.07683863885839737,
+      "grad_norm": 137.56018328091912,
+      "learning_rate": 2.481751824817518e-07,
+      "loss": 1.4013,
+      "step": 35
+    },
+    {
+      "epoch": 0.07903402854006586,
+      "grad_norm": 110.2306130917618,
+      "learning_rate": 2.5547445255474454e-07,
+      "loss": 1.2338,
+      "step": 36
+    },
+    {
+      "epoch": 0.08122941822173436,
+      "grad_norm": 140.88497788079806,
+      "learning_rate": 2.6277372262773725e-07,
+      "loss": 1.2442,
+      "step": 37
+    },
+    {
+      "epoch": 0.08342480790340286,
+      "grad_norm": 131.1603732007104,
+      "learning_rate": 2.700729927007299e-07,
+      "loss": 1.3306,
+      "step": 38
+    },
+    {
+      "epoch": 0.08562019758507135,
+      "grad_norm": 125.8912594433526,
+      "learning_rate": 2.773722627737226e-07,
+      "loss": 1.2538,
+      "step": 39
+    },
+    {
+      "epoch": 0.08781558726673985,
+      "grad_norm": 173.62590699447912,
+      "learning_rate": 2.846715328467153e-07,
+      "loss": 1.282,
+      "step": 40
+    },
+    {
+      "epoch": 0.09001097694840834,
+      "grad_norm": 140.0348709144397,
+      "learning_rate": 2.91970802919708e-07,
+      "loss": 1.2933,
+      "step": 41
+    },
+    {
+      "epoch": 0.09220636663007684,
+      "grad_norm": 138.34191099990304,
+      "learning_rate": 2.9927007299270075e-07,
+      "loss": 1.2246,
+      "step": 42
+    },
+    {
+      "epoch": 0.09440175631174534,
+      "grad_norm": 133.3567358083278,
+      "learning_rate": 3.065693430656934e-07,
+      "loss": 1.2049,
+      "step": 43
+    },
+    {
+      "epoch": 0.09659714599341383,
+      "grad_norm": 149.86829216030878,
+      "learning_rate": 3.138686131386861e-07,
+      "loss": 1.1484,
+      "step": 44
+    },
+    {
+      "epoch": 0.09879253567508232,
+      "grad_norm": 139.54448207721754,
+      "learning_rate": 3.2116788321167883e-07,
+      "loss": 1.1161,
+      "step": 45
+    },
+    {
+      "epoch": 0.10098792535675083,
+      "grad_norm": 144.73507422182286,
+      "learning_rate": 3.284671532846715e-07,
+      "loss": 1.2278,
+      "step": 46
+    },
+    {
+      "epoch": 0.10318331503841932,
+      "grad_norm": 121.6383629486138,
+      "learning_rate": 3.357664233576642e-07,
+      "loss": 1.2288,
+      "step": 47
+    },
+    {
+      "epoch": 0.10537870472008781,
+      "grad_norm": 98.86387010070742,
+      "learning_rate": 3.4306569343065697e-07,
+      "loss": 1.1108,
+      "step": 48
+    },
+    {
+      "epoch": 0.10757409440175632,
+      "grad_norm": 163.06719307818943,
+      "learning_rate": 3.5036496350364963e-07,
+      "loss": 1.1668,
+      "step": 49
+    },
+    {
+      "epoch": 0.10976948408342481,
+      "grad_norm": 115.37025925475122,
+      "learning_rate": 3.5766423357664234e-07,
+      "loss": 1.0684,
+      "step": 50
+    },
+    {
+      "epoch": 0.10976948408342481,
+      "eval_accuracy": 0.506,
+      "eval_loss": 1.019887089729309,
+      "eval_runtime": 62.4289,
+      "eval_samples_per_second": 8.009,
+      "eval_steps_per_second": 1.009,
+      "step": 50
+    },
+    {
+      "epoch": 0.1119648737650933,
+      "grad_norm": 115.00520545818907,
+      "learning_rate": 3.64963503649635e-07,
+      "loss": 1.1397,
+      "step": 51
+    },
+    {
+      "epoch": 0.1141602634467618,
+      "grad_norm": 102.41196522989259,
+      "learning_rate": 3.722627737226277e-07,
+      "loss": 1.0745,
+      "step": 52
+    },
+    {
+      "epoch": 0.1163556531284303,
+      "grad_norm": 117.26316852602802,
+      "learning_rate": 3.795620437956204e-07,
+      "loss": 1.2356,
+      "step": 53
+    },
+    {
+      "epoch": 0.11855104281009879,
+      "grad_norm": 120.91380597609361,
+      "learning_rate": 3.8686131386861313e-07,
+      "loss": 1.2119,
+      "step": 54
+    },
+    {
+      "epoch": 0.1207464324917673,
+      "grad_norm": 92.29300818472724,
+      "learning_rate": 3.9416058394160584e-07,
+      "loss": 1.1304,
+      "step": 55
+    },
+    {
+      "epoch": 0.12294182217343579,
+      "grad_norm": 92.82345907233494,
+      "learning_rate": 4.0145985401459856e-07,
+      "loss": 1.1531,
+      "step": 56
+    },
+    {
+      "epoch": 0.1251372118551043,
+      "grad_norm": 85.5717973777339,
+      "learning_rate": 4.087591240875912e-07,
+      "loss": 0.9308,
+      "step": 57
+    },
+    {
+      "epoch": 0.12733260153677278,
+      "grad_norm": 85.43269620489475,
+      "learning_rate": 4.160583941605839e-07,
+      "loss": 1.1233,
+      "step": 58
+    },
+    {
+      "epoch": 0.12952799121844127,
+      "grad_norm": 115.80515321077907,
+      "learning_rate": 4.233576642335766e-07,
+      "loss": 1.1797,
+      "step": 59
+    },
+    {
+      "epoch": 0.13172338090010977,
+      "grad_norm": 84.2918940049183,
+      "learning_rate": 4.306569343065693e-07,
+      "loss": 1.092,
+      "step": 60
+    },
+    {
+      "epoch": 0.13391877058177826,
+      "grad_norm": 99.39269845007587,
+      "learning_rate": 4.3795620437956206e-07,
+      "loss": 1.1176,
+      "step": 61
+    },
+    {
+      "epoch": 0.13611416026344675,
+      "grad_norm": 70.59454501984223,
+      "learning_rate": 4.452554744525547e-07,
+      "loss": 1.0813,
+      "step": 62
+    },
+    {
+      "epoch": 0.13830954994511527,
+      "grad_norm": 109.5923399556689,
+      "learning_rate": 4.5255474452554743e-07,
+      "loss": 0.9878,
+      "step": 63
+    },
+    {
+      "epoch": 0.14050493962678376,
+      "grad_norm": 63.030131831059876,
+      "learning_rate": 4.5985401459854014e-07,
+      "loss": 0.9483,
+      "step": 64
+    },
+    {
+      "epoch": 0.14270032930845225,
+      "grad_norm": 87.02770783815946,
+      "learning_rate": 4.671532846715328e-07,
+      "loss": 0.9885,
+      "step": 65
+    },
+    {
+      "epoch": 0.14489571899012074,
+      "grad_norm": 70.75751318092878,
+      "learning_rate": 4.744525547445255e-07,
+      "loss": 0.8505,
+      "step": 66
+    },
+    {
+      "epoch": 0.14709110867178923,
+      "grad_norm": 84.18711981429041,
+      "learning_rate": 4.817518248175182e-07,
+      "loss": 0.9274,
+      "step": 67
+    },
+    {
+      "epoch": 0.14928649835345773,
+      "grad_norm": 93.24415466346244,
+      "learning_rate": 4.89051094890511e-07,
+      "loss": 1.0028,
+      "step": 68
+    },
+    {
+      "epoch": 0.15148188803512624,
+      "grad_norm": 58.159526228156615,
+      "learning_rate": 4.963503649635036e-07,
+      "loss": 0.8635,
+      "step": 69
+    },
+    {
+      "epoch": 0.15367727771679474,
+      "grad_norm": 69.94129826644978,
+      "learning_rate": 5.036496350364964e-07,
+      "loss": 0.9774,
+      "step": 70
+    },
+    {
+      "epoch": 0.15587266739846323,
+      "grad_norm": 69.03950330444896,
+      "learning_rate": 5.109489051094891e-07,
+      "loss": 0.9669,
+      "step": 71
+    },
+    {
+      "epoch": 0.15806805708013172,
+      "grad_norm": 76.92567375028155,
+      "learning_rate": 5.182481751824817e-07,
+      "loss": 0.86,
+      "step": 72
+    },
+    {
+      "epoch": 0.1602634467618002,
+      "grad_norm": 74.59715692707641,
+      "learning_rate": 5.255474452554745e-07,
+      "loss": 0.9128,
+      "step": 73
+    },
+    {
+      "epoch": 0.16245883644346873,
+      "grad_norm": 82.61633303553909,
+      "learning_rate": 5.328467153284672e-07,
+      "loss": 0.9546,
+      "step": 74
+    },
+    {
+      "epoch": 0.16465422612513722,
+      "grad_norm": 58.685512250843665,
+      "learning_rate": 5.401459854014598e-07,
+      "loss": 0.875,
+      "step": 75
+    },
+    {
+      "epoch": 0.1668496158068057,
+      "grad_norm": 58.60891793684997,
+      "learning_rate": 5.474452554744526e-07,
+      "loss": 0.8729,
+      "step": 76
+    },
+    {
+      "epoch": 0.1690450054884742,
+      "grad_norm": 68.60986432356273,
+      "learning_rate": 5.547445255474452e-07,
+      "loss": 0.8208,
+      "step": 77
+    },
+    {
+      "epoch": 0.1712403951701427,
+      "grad_norm": 58.35034555605016,
+      "learning_rate": 5.620437956204379e-07,
+      "loss": 0.903,
+      "step": 78
+    },
+    {
+      "epoch": 0.1734357848518112,
+      "grad_norm": 98.61406200927433,
+      "learning_rate": 5.693430656934306e-07,
+      "loss": 0.9472,
+      "step": 79
+    },
+    {
+      "epoch": 0.1756311745334797,
+      "grad_norm": 81.08439956625773,
+      "learning_rate": 5.766423357664233e-07,
+      "loss": 0.9163,
+      "step": 80
+    },
+    {
+      "epoch": 0.1778265642151482,
+      "grad_norm": 53.88557720126228,
+      "learning_rate": 5.83941605839416e-07,
+      "loss": 0.868,
+      "step": 81
+    },
+    {
+      "epoch": 0.1800219538968167,
+      "grad_norm": 53.143301158626926,
+      "learning_rate": 5.912408759124087e-07,
+      "loss": 0.88,
+      "step": 82
+    },
+    {
+      "epoch": 0.18221734357848518,
+      "grad_norm": 49.19404393103411,
+      "learning_rate": 5.985401459854015e-07,
+      "loss": 0.8291,
+      "step": 83
+    },
+    {
+      "epoch": 0.18441273326015367,
+      "grad_norm": 49.9632858603646,
+      "learning_rate": 6.058394160583942e-07,
+      "loss": 0.8155,
+      "step": 84
+    },
+    {
+      "epoch": 0.18660812294182216,
+      "grad_norm": 45.62330278251693,
+      "learning_rate": 6.131386861313868e-07,
+      "loss": 0.7906,
+      "step": 85
+    },
+    {
+      "epoch": 0.18880351262349068,
+      "grad_norm": 41.69982248481481,
+      "learning_rate": 6.204379562043796e-07,
+      "loss": 0.7877,
+      "step": 86
+    },
+    {
+      "epoch": 0.19099890230515917,
+      "grad_norm": 36.534122434280455,
+      "learning_rate": 6.277372262773722e-07,
+      "loss": 0.7527,
+      "step": 87
+    },
+    {
+      "epoch": 0.19319429198682767,
+      "grad_norm": 41.51590795972736,
+      "learning_rate": 6.350364963503649e-07,
+      "loss": 0.7701,
+      "step": 88
+    },
+    {
+      "epoch": 0.19538968166849616,
+      "grad_norm": 50.9447076217186,
+      "learning_rate": 6.423357664233577e-07,
+      "loss": 0.8,
+      "step": 89
+    },
+    {
+      "epoch": 0.19758507135016465,
+      "grad_norm": 39.73817349854984,
+      "learning_rate": 6.496350364963503e-07,
+      "loss": 0.7236,
+      "step": 90
+    },
+    {
+      "epoch": 0.19978046103183314,
+      "grad_norm": 38.43809365198889,
+      "learning_rate": 6.56934306569343e-07,
+      "loss": 0.6534,
+      "step": 91
+    },
+    {
+      "epoch": 0.20197585071350166,
+      "grad_norm": 37.357871240275045,
+      "learning_rate": 6.642335766423358e-07,
+      "loss": 0.7343,
+      "step": 92
+    },
+    {
+      "epoch": 0.20417124039517015,
+      "grad_norm": 44.96345470889994,
+      "learning_rate": 6.715328467153284e-07,
+      "loss": 0.6828,
+      "step": 93
+    },
+    {
+      "epoch": 0.20636663007683864,
+      "grad_norm": 37.388274455357625,
+      "learning_rate": 6.788321167883211e-07,
+      "loss": 0.6308,
+      "step": 94
+    },
+    {
+      "epoch": 0.20856201975850713,
+      "grad_norm": 30.201936308225815,
+      "learning_rate": 6.861313868613139e-07,
+      "loss": 0.6127,
+      "step": 95
+    },
+    {
+      "epoch": 0.21075740944017562,
+      "grad_norm": 35.90933604523665,
+      "learning_rate": 6.934306569343066e-07,
+      "loss": 0.6911,
+      "step": 96
+    },
+    {
+      "epoch": 0.21295279912184412,
+      "grad_norm": 29.700092316262612,
+      "learning_rate": 7.007299270072993e-07,
+      "loss": 0.6556,
+      "step": 97
+    },
+    {
+      "epoch": 0.21514818880351264,
+      "grad_norm": 31.842570539615227,
+      "learning_rate": 7.080291970802919e-07,
+      "loss": 0.6896,
+      "step": 98
+    },
+    {
+      "epoch": 0.21734357848518113,
+      "grad_norm": 31.238683970651707,
+      "learning_rate": 7.153284671532847e-07,
+      "loss": 0.6506,
+      "step": 99
+    },
+    {
+      "epoch": 0.21953896816684962,
+      "grad_norm": 35.48873900248551,
+      "learning_rate": 7.226277372262773e-07,
+      "loss": 0.7495,
+      "step": 100
+    },
+    {
+      "epoch": 0.21953896816684962,
+      "eval_accuracy": 0.682,
+      "eval_loss": 0.562687873840332,
+      "eval_runtime": 62.4373,
+      "eval_samples_per_second": 8.008,
+      "eval_steps_per_second": 1.009,
+      "step": 100
+    },
+    {
+      "epoch": 0.2217343578485181,
+      "grad_norm": 34.153901482286706,
+      "learning_rate": 7.2992700729927e-07,
+      "loss": 0.7136,
+      "step": 101
+    },
+    {
+      "epoch": 0.2239297475301866,
+      "grad_norm": 27.9624148195877,
+      "learning_rate": 7.372262773722628e-07,
+      "loss": 0.7178,
+      "step": 102
+    },
+    {
+      "epoch": 0.2261251372118551,
+      "grad_norm": 22.100202938968764,
+      "learning_rate": 7.445255474452554e-07,
+      "loss": 0.6749,
+      "step": 103
+    },
+    {
+      "epoch": 0.2283205268935236,
+      "grad_norm": 28.18545960285919,
+      "learning_rate": 7.518248175182481e-07,
+      "loss": 0.666,
+      "step": 104
+    },
+    {
+      "epoch": 0.2305159165751921,
+      "grad_norm": 22.264174917923164,
+      "learning_rate": 7.591240875912408e-07,
+      "loss": 0.6312,
+      "step": 105
+    },
+    {
+      "epoch": 0.2327113062568606,
+      "grad_norm": 17.37357275601335,
+      "learning_rate": 7.664233576642335e-07,
+      "loss": 0.6065,
+      "step": 106
+    },
+    {
+      "epoch": 0.2349066959385291,
+      "grad_norm": 19.49426500716658,
+      "learning_rate": 7.737226277372263e-07,
+      "loss": 0.6502,
+      "step": 107
+    },
+    {
+      "epoch": 0.23710208562019758,
+      "grad_norm": 22.022008695087543,
+      "learning_rate": 7.81021897810219e-07,
+      "loss": 0.5885,
+      "step": 108
+    },
+    {
+      "epoch": 0.23929747530186607,
+      "grad_norm": 21.515467941901086,
+      "learning_rate": 7.883211678832117e-07,
+      "loss": 0.5583,
+      "step": 109
+    },
+    {
+      "epoch": 0.2414928649835346,
+      "grad_norm": 25.78728071293923,
+      "learning_rate": 7.956204379562043e-07,
+      "loss": 0.6243,
+      "step": 110
+    },
+    {
+      "epoch": 0.24368825466520308,
+      "grad_norm": 18.18532044027736,
+      "learning_rate": 8.029197080291971e-07,
+      "loss": 0.617,
+      "step": 111
+    },
+    {
+      "epoch": 0.24588364434687157,
+      "grad_norm": 25.226965635992663,
+      "learning_rate": 8.102189781021898e-07,
+      "loss": 0.579,
+      "step": 112
+    },
+    {
+      "epoch": 0.24807903402854006,
+      "grad_norm": 22.512756730293077,
+      "learning_rate": 8.175182481751824e-07,
+      "loss": 0.5939,
+      "step": 113
+    },
+    {
+      "epoch": 0.2502744237102086,
+      "grad_norm": 18.596439780374983,
+      "learning_rate": 8.248175182481751e-07,
+      "loss": 0.6323,
+      "step": 114
+    },
+    {
+      "epoch": 0.2524698133918771,
+      "grad_norm": 17.02105845004187,
+      "learning_rate": 8.321167883211679e-07,
+      "loss": 0.5594,
+      "step": 115
+    },
+    {
+      "epoch": 0.25466520307354557,
+      "grad_norm": 16.250583027072732,
+      "learning_rate": 8.394160583941605e-07,
+      "loss": 0.577,
+      "step": 116
+    },
+    {
+      "epoch": 0.25686059275521406,
+      "grad_norm": 18.5715233786033,
+      "learning_rate": 8.467153284671532e-07,
+      "loss": 0.478,
+      "step": 117
+    },
+    {
+      "epoch": 0.25905598243688255,
+      "grad_norm": 17.614587096288357,
+      "learning_rate": 8.540145985401459e-07,
+      "loss": 0.5567,
+      "step": 118
+    },
+    {
+      "epoch": 0.26125137211855104,
+      "grad_norm": 21.00826003065434,
+      "learning_rate": 8.613138686131386e-07,
+      "loss": 0.5687,
+      "step": 119
+    },
+    {
+      "epoch": 0.26344676180021953,
+      "grad_norm": 15.607731096994556,
+      "learning_rate": 8.686131386861314e-07,
+      "loss": 0.5608,
+      "step": 120
+    },
+    {
+      "epoch": 0.265642151481888,
+      "grad_norm": 15.277425156097365,
+      "learning_rate": 8.759124087591241e-07,
+      "loss": 0.5288,
+      "step": 121
+    },
+    {
+      "epoch": 0.2678375411635565,
+      "grad_norm": 19.651993472199724,
+      "learning_rate": 8.832116788321168e-07,
+      "loss": 0.5026,
+      "step": 122
+    },
+    {
+      "epoch": 0.270032930845225,
+      "grad_norm": 24.891420468239335,
+      "learning_rate": 8.905109489051094e-07,
+      "loss": 0.5651,
+      "step": 123
+    },
+    {
+      "epoch": 0.2722283205268935,
+      "grad_norm": 17.627056937663315,
+      "learning_rate": 8.978102189781022e-07,
+      "loss": 0.5665,
+      "step": 124
+    },
+    {
+      "epoch": 0.27442371020856204,
+      "grad_norm": 16.656331513263602,
+      "learning_rate": 9.051094890510949e-07,
+      "loss": 0.4395,
+      "step": 125
+    },
+    {
+      "epoch": 0.27661909989023054,
+      "grad_norm": 15.567024748783735,
+      "learning_rate": 9.124087591240875e-07,
+      "loss": 0.4625,
+      "step": 126
+    },
+    {
+      "epoch": 0.278814489571899,
+      "grad_norm": 15.002053723893589,
+      "learning_rate": 9.197080291970803e-07,
+      "loss": 0.5267,
+      "step": 127
+    },
+    {
+      "epoch": 0.2810098792535675,
+      "grad_norm": 13.834884625752059,
+      "learning_rate": 9.270072992700729e-07,
+      "loss": 0.5017,
+      "step": 128
+    },
+    {
+      "epoch": 0.283205268935236,
+      "grad_norm": 15.615717349478732,
+      "learning_rate": 9.343065693430656e-07,
+      "loss": 0.4742,
+      "step": 129
+    },
+    {
+      "epoch": 0.2854006586169045,
+      "grad_norm": 13.471813706001054,
+      "learning_rate": 9.416058394160583e-07,
+      "loss": 0.512,
+      "step": 130
+    },
+    {
+      "epoch": 0.287596048298573,
+      "grad_norm": 15.094349701329602,
+      "learning_rate": 9.48905109489051e-07,
+      "loss": 0.5327,
+      "step": 131
+    },
+    {
+      "epoch": 0.2897914379802415,
+      "grad_norm": 13.19342166454616,
+      "learning_rate": 9.562043795620438e-07,
+      "loss": 0.47,
+      "step": 132
+    },
+    {
+      "epoch": 0.29198682766191,
+      "grad_norm": 12.139844355160385,
+      "learning_rate": 9.635036496350364e-07,
+      "loss": 0.5429,
+      "step": 133
+    },
+    {
+      "epoch": 0.29418221734357847,
+      "grad_norm": 13.352063594802008,
+      "learning_rate": 9.708029197080291e-07,
+      "loss": 0.4865,
+      "step": 134
+    },
+    {
+      "epoch": 0.29637760702524696,
+      "grad_norm": 14.225936575772286,
+      "learning_rate": 9.78102189781022e-07,
+      "loss": 0.4784,
+      "step": 135
+    },
+    {
+      "epoch": 0.29857299670691545,
+      "grad_norm": 12.461695683400633,
+      "learning_rate": 9.854014598540146e-07,
+      "loss": 0.4518,
+      "step": 136
+    },
+    {
+      "epoch": 0.300768386388584,
+      "grad_norm": 12.867009884092825,
+      "learning_rate": 9.927007299270073e-07,
+      "loss": 0.4426,
+      "step": 137
+    },
+    {
+      "epoch": 0.3029637760702525,
+      "grad_norm": 20.246601888258336,
+      "learning_rate": 1e-06,
+      "loss": 0.5418,
+      "step": 138
+    },
+    {
+      "epoch": 0.305159165751921,
+      "grad_norm": 19.62917306684865,
+      "learning_rate": 9.999983717412808e-07,
+      "loss": 0.5246,
+      "step": 139
+    },
+    {
+      "epoch": 0.30735455543358947,
+      "grad_norm": 14.70577630004396,
+      "learning_rate": 9.999934869757278e-07,
+      "loss": 0.4996,
+      "step": 140
+    },
+    {
+      "epoch": 0.30954994511525796,
+      "grad_norm": 11.895324662976222,
+      "learning_rate": 9.999853457351558e-07,
+      "loss": 0.4846,
+      "step": 141
+    },
+    {
+      "epoch": 0.31174533479692645,
+      "grad_norm": 11.772224391975989,
+      "learning_rate": 9.999739480725893e-07,
+      "loss": 0.4882,
+      "step": 142
+    },
+    {
+      "epoch": 0.31394072447859495,
+      "grad_norm": 12.342437917454156,
+      "learning_rate": 9.999592940622613e-07,
+      "loss": 0.4973,
+      "step": 143
+    },
+    {
+      "epoch": 0.31613611416026344,
+      "grad_norm": 9.63333593523558,
+      "learning_rate": 9.999413837996137e-07,
+      "loss": 0.4657,
+      "step": 144
+    },
+    {
+      "epoch": 0.31833150384193193,
+      "grad_norm": 12.051946387691274,
+      "learning_rate": 9.999202174012972e-07,
+      "loss": 0.4778,
+      "step": 145
+    },
+    {
+      "epoch": 0.3205268935236004,
+      "grad_norm": 11.959976548769715,
+      "learning_rate": 9.99895795005169e-07,
+      "loss": 0.4527,
+      "step": 146
+    },
+    {
+      "epoch": 0.3227222832052689,
+      "grad_norm": 10.113920097674713,
+      "learning_rate": 9.99868116770293e-07,
+      "loss": 0.4078,
+      "step": 147
+    },
+    {
+      "epoch": 0.32491767288693746,
+      "grad_norm": 12.435033880307502,
+      "learning_rate": 9.998371828769384e-07,
+      "loss": 0.4894,
+      "step": 148
+    },
+    {
+      "epoch": 0.32711306256860595,
+      "grad_norm": 14.939081919754774,
+      "learning_rate": 9.99802993526579e-07,
+      "loss": 0.5065,
+      "step": 149
+    },
+    {
+      "epoch": 0.32930845225027444,
+      "grad_norm": 14.254154244717148,
+      "learning_rate": 9.997655489418912e-07,
+      "loss": 0.494,
+      "step": 150
+    },
+    {
+      "epoch": 0.32930845225027444,
+      "eval_accuracy": 0.786,
+      "eval_loss": 0.435116708278656,
+      "eval_runtime": 62.3466,
+      "eval_samples_per_second": 8.02,
+      "eval_steps_per_second": 1.01,
+      "step": 150
+    },
+    {
+      "epoch": 0.33150384193194293,
+      "grad_norm": 12.039017936319885,
+      "learning_rate": 9.997248493667527e-07,
+      "loss": 0.4839,
+      "step": 151
+    },
+    {
+      "epoch": 0.3336992316136114,
+      "grad_norm": 11.070313457195804,
+      "learning_rate": 9.996808950662413e-07,
+      "loss": 0.4378,
+      "step": 152
+    },
+    {
+      "epoch": 0.3358946212952799,
+      "grad_norm": 9.336009234528,
+      "learning_rate": 9.99633686326633e-07,
+      "loss": 0.4351,
+      "step": 153
+    },
+    {
+      "epoch": 0.3380900109769484,
+      "grad_norm": 11.55439798512532,
+      "learning_rate": 9.995832234554e-07,
+      "loss": 0.4553,
+      "step": 154
+    },
+    {
+      "epoch": 0.3402854006586169,
+      "grad_norm": 12.154145816173985,
+      "learning_rate": 9.995295067812083e-07,
+      "loss": 0.4964,
+      "step": 155
+    },
+    {
+      "epoch": 0.3424807903402854,
+      "grad_norm": 9.628643083965512,
+      "learning_rate": 9.99472536653917e-07,
+      "loss": 0.5015,
+      "step": 156
+    },
+    {
+      "epoch": 0.3446761800219539,
+      "grad_norm": 11.610322067423159,
+      "learning_rate": 9.994123134445746e-07,
+      "loss": 0.4712,
+      "step": 157
+    },
+    {
+      "epoch": 0.3468715697036224,
+      "grad_norm": 12.74943545259812,
+      "learning_rate": 9.993488375454165e-07,
+      "loss": 0.4549,
+      "step": 158
+    },
+    {
+      "epoch": 0.34906695938529086,
+      "grad_norm": 11.741546873031675,
+      "learning_rate": 9.992821093698636e-07,
+      "loss": 0.5129,
+      "step": 159
+    },
+    {
+      "epoch": 0.3512623490669594,
+      "grad_norm": 10.451389621211415,
+      "learning_rate": 9.992121293525188e-07,
+      "loss": 0.4884,
+      "step": 160
+    },
+    {
+      "epoch": 0.3534577387486279,
+      "grad_norm": 9.48305774210746,
+      "learning_rate": 9.991388979491646e-07,
+      "loss": 0.4085,
+      "step": 161
+    },
+    {
+      "epoch": 0.3556531284302964,
+      "grad_norm": 14.783454444270212,
+      "learning_rate": 9.990624156367596e-07,
+      "loss": 0.5289,
+      "step": 162
+    },
+    {
+      "epoch": 0.3578485181119649,
+      "grad_norm": 8.468180083819155,
+      "learning_rate": 9.989826829134356e-07,
+      "loss": 0.449,
+      "step": 163
+    },
+    {
+      "epoch": 0.3600439077936334,
+      "grad_norm": 11.676809759593661,
+      "learning_rate": 9.988997002984949e-07,
+      "loss": 0.5199,
+      "step": 164
+    },
+    {
+      "epoch": 0.36223929747530187,
+      "grad_norm": 11.110035365580961,
+      "learning_rate": 9.988134683324058e-07,
+      "loss": 0.4988,
+      "step": 165
+    },
+    {
+      "epoch": 0.36443468715697036,
+      "grad_norm": 9.194590897066561,
+      "learning_rate": 9.987239875768006e-07,
+      "loss": 0.4447,
+      "step": 166
+    },
+    {
+      "epoch": 0.36663007683863885,
+      "grad_norm": 10.829285977863856,
+      "learning_rate": 9.9863125861447e-07,
+      "loss": 0.4645,
+      "step": 167
+    },
+    {
+      "epoch": 0.36882546652030734,
+      "grad_norm": 11.15768970639428,
+      "learning_rate": 9.985352820493614e-07,
+      "loss": 0.4429,
+      "step": 168
+    },
+    {
+      "epoch": 0.37102085620197583,
+      "grad_norm": 10.615797787300718,
+      "learning_rate": 9.984360585065733e-07,
+      "loss": 0.4236,
+      "step": 169
+    },
+    {
+      "epoch": 0.3732162458836443,
+      "grad_norm": 11.596646960772174,
+      "learning_rate": 9.983335886323524e-07,
+      "loss": 0.4212,
+      "step": 170
+    },
+    {
+      "epoch": 0.3754116355653128,
+      "grad_norm": 11.60172095267533,
+      "learning_rate": 9.98227873094088e-07,
+      "loss": 0.4796,
+      "step": 171
+    },
+    {
+      "epoch": 0.37760702524698136,
+      "grad_norm": 14.055785993956126,
+      "learning_rate": 9.981189125803095e-07,
+      "loss": 0.4841,
+      "step": 172
+    },
+    {
+      "epoch": 0.37980241492864986,
+      "grad_norm": 9.572819362372918,
+      "learning_rate": 9.980067078006804e-07,
+      "loss": 0.3985,
+      "step": 173
+    },
+    {
+      "epoch": 0.38199780461031835,
+      "grad_norm": 11.59813843253133,
+      "learning_rate": 9.978912594859946e-07,
+      "loss": 0.4564,
+      "step": 174
+    },
+    {
+      "epoch": 0.38419319429198684,
+      "grad_norm": 13.737142414943285,
+      "learning_rate": 9.977725683881707e-07,
+      "loss": 0.4522,
+      "step": 175
+    },
+    {
+      "epoch": 0.38638858397365533,
+      "grad_norm": 9.246882570201656,
+      "learning_rate": 9.97650635280248e-07,
+      "loss": 0.4097,
+      "step": 176
+    },
+    {
+      "epoch": 0.3885839736553238,
+      "grad_norm": 9.136180662179372,
+      "learning_rate": 9.97525460956381e-07,
+      "loss": 0.4141,
+      "step": 177
+    },
+    {
+      "epoch": 0.3907793633369923,
+      "grad_norm": 9.921404583485003,
+      "learning_rate": 9.973970462318349e-07,
+      "loss": 0.4357,
+      "step": 178
+    },
+    {
+      "epoch": 0.3929747530186608,
+      "grad_norm": 11.601008865801447,
+      "learning_rate": 9.972653919429788e-07,
+      "loss": 0.4594,
+      "step": 179
+    },
+    {
+      "epoch": 0.3951701427003293,
+      "grad_norm": 10.36094679422873,
+      "learning_rate": 9.971304989472817e-07,
+      "loss": 0.4177,
+      "step": 180
+    },
+    {
+      "epoch": 0.3973655323819978,
+      "grad_norm": 10.948879112431765,
+      "learning_rate": 9.969923681233066e-07,
+      "loss": 0.5064,
+      "step": 181
+    },
+    {
+      "epoch": 0.3995609220636663,
+      "grad_norm": 11.834429314578246,
+      "learning_rate": 9.968510003707042e-07,
+      "loss": 0.4665,
+      "step": 182
+    },
+    {
+      "epoch": 0.40175631174533477,
+      "grad_norm": 11.392606754798651,
+      "learning_rate": 9.967063966102079e-07,
+      "loss": 0.3952,
+      "step": 183
+    },
+    {
+      "epoch": 0.4039517014270033,
+      "grad_norm": 9.253848276057784,
+      "learning_rate": 9.965585577836264e-07,
+      "loss": 0.4712,
+      "step": 184
+    },
+    {
+      "epoch": 0.4061470911086718,
+      "grad_norm": 9.08494168337149,
+      "learning_rate": 9.9640748485384e-07,
+      "loss": 0.4026,
+      "step": 185
+    },
+    {
+      "epoch": 0.4083424807903403,
+      "grad_norm": 9.664013692503126,
+      "learning_rate": 9.962531788047913e-07,
+      "loss": 0.3652,
+      "step": 186
+    },
+    {
+      "epoch": 0.4105378704720088,
+      "grad_norm": 12.118857418627702,
+      "learning_rate": 9.960956406414813e-07,
+      "loss": 0.4189,
+      "step": 187
+    },
+    {
+      "epoch": 0.4127332601536773,
+      "grad_norm": 12.320330458732224,
+      "learning_rate": 9.959348713899613e-07,
+      "loss": 0.4341,
+      "step": 188
+    },
+    {
+      "epoch": 0.4149286498353458,
+      "grad_norm": 12.403712297845825,
+      "learning_rate": 9.957708720973273e-07,
+      "loss": 0.4016,
+      "step": 189
+    },
+    {
+      "epoch": 0.41712403951701427,
+      "grad_norm": 16.26187132736173,
+      "learning_rate": 9.956036438317123e-07,
+      "loss": 0.4644,
+      "step": 190
+    },
+    {
+      "epoch": 0.41931942919868276,
+      "grad_norm": 12.170054665158572,
+      "learning_rate": 9.954331876822798e-07,
+      "loss": 0.4374,
+      "step": 191
+    },
+    {
+      "epoch": 0.42151481888035125,
+      "grad_norm": 14.02551196287224,
+      "learning_rate": 9.952595047592167e-07,
+      "loss": 0.5103,
+      "step": 192
+    },
+    {
+      "epoch": 0.42371020856201974,
+      "grad_norm": 9.533566927751629,
+      "learning_rate": 9.950825961937257e-07,
+      "loss": 0.4551,
+      "step": 193
+    },
+    {
+      "epoch": 0.42590559824368823,
+      "grad_norm": 12.533660392960899,
+      "learning_rate": 9.949024631380189e-07,
+      "loss": 0.4751,
+      "step": 194
+    },
+    {
+      "epoch": 0.4281009879253567,
+      "grad_norm": 11.451058008246223,
+      "learning_rate": 9.94719106765309e-07,
+      "loss": 0.4058,
+      "step": 195
+    },
+    {
+      "epoch": 0.43029637760702527,
+      "grad_norm": 11.13982911192037,
+      "learning_rate": 9.945325282698022e-07,
+      "loss": 0.4395,
+      "step": 196
+    },
+    {
+      "epoch": 0.43249176728869376,
+      "grad_norm": 9.220791965183867,
+      "learning_rate": 9.94342728866691e-07,
+      "loss": 0.4104,
+      "step": 197
+    },
+    {
+      "epoch": 0.43468715697036225,
+      "grad_norm": 10.323390638102808,
+      "learning_rate": 9.941497097921456e-07,
+      "loss": 0.4384,
+      "step": 198
+    },
+    {
+      "epoch": 0.43688254665203075,
+      "grad_norm": 9.058003354224194,
+      "learning_rate": 9.939534723033057e-07,
+      "loss": 0.4268,
+      "step": 199
+    },
+    {
+      "epoch": 0.43907793633369924,
+      "grad_norm": 11.887165620636456,
+      "learning_rate": 9.937540176782731e-07,
+      "loss": 0.4492,
+      "step": 200
+    },
+    {
+      "epoch": 0.43907793633369924,
+      "eval_accuracy": 0.794,
+      "eval_loss": 0.4044578969478607,
+      "eval_runtime": 62.3322,
+      "eval_samples_per_second": 8.022,
+      "eval_steps_per_second": 1.011,
+      "step": 200
+    },
+    {
+      "epoch": 0.44127332601536773,
+      "grad_norm": 12.441278365943637,
+      "learning_rate": 9.935513472161026e-07,
+      "loss": 0.4653,
+      "step": 201
+    },
+    {
+      "epoch": 0.4434687156970362,
+      "grad_norm": 14.201601234233598,
+      "learning_rate": 9.93345462236794e-07,
+      "loss": 0.4865,
+      "step": 202
+    },
+    {
+      "epoch": 0.4456641053787047,
+      "grad_norm": 11.347903109722713,
+      "learning_rate": 9.931363640812837e-07,
+      "loss": 0.4097,
+      "step": 203
+    },
+    {
+      "epoch": 0.4478594950603732,
+      "grad_norm": 11.66358658856107,
+      "learning_rate": 9.929240541114347e-07,
+      "loss": 0.4315,
+      "step": 204
+    },
+    {
+      "epoch": 0.4500548847420417,
+      "grad_norm": 10.618630242743638,
+      "learning_rate": 9.927085337100298e-07,
+      "loss": 0.4678,
+      "step": 205
+    },
+    {
+      "epoch": 0.4522502744237102,
+      "grad_norm": 7.801209288203623,
+      "learning_rate": 9.924898042807604e-07,
+      "loss": 0.4143,
+      "step": 206
+    },
+    {
+      "epoch": 0.4544456641053787,
+      "grad_norm": 7.639533498673639,
+      "learning_rate": 9.922678672482192e-07,
+      "loss": 0.4286,
+      "step": 207
+    },
+    {
+      "epoch": 0.4566410537870472,
+      "grad_norm": 9.76898599627511,
+      "learning_rate": 9.920427240578898e-07,
+      "loss": 0.4785,
+      "step": 208
+    },
+    {
+      "epoch": 0.4588364434687157,
+      "grad_norm": 11.869232979224398,
+      "learning_rate": 9.918143761761376e-07,
+      "loss": 0.4192,
+      "step": 209
+    },
+    {
+      "epoch": 0.4610318331503842,
+      "grad_norm": 7.372984753013456,
+      "learning_rate": 9.915828250902003e-07,
+      "loss": 0.3777,
+      "step": 210
+    },
+    {
+      "epoch": 0.4632272228320527,
+      "grad_norm": 11.121453453807499,
+      "learning_rate": 9.913480723081782e-07,
+      "loss": 0.4398,
+      "step": 211
+    },
+    {
+      "epoch": 0.4654226125137212,
+      "grad_norm": 16.17068331984467,
+      "learning_rate": 9.911101193590243e-07,
+      "loss": 0.4948,
+      "step": 212
+    },
+    {
+      "epoch": 0.4676180021953897,
+      "grad_norm": 7.731820210763722,
+      "learning_rate": 9.908689677925347e-07,
+      "loss": 0.3828,
+      "step": 213
+    },
+    {
+      "epoch": 0.4698133918770582,
+      "grad_norm": 6.955625100083833,
+      "learning_rate": 9.906246191793378e-07,
+      "loss": 0.3839,
+      "step": 214
+    },
+    {
+      "epoch": 0.47200878155872666,
+      "grad_norm": 8.102074000516577,
+      "learning_rate": 9.903770751108845e-07,
+      "loss": 0.4021,
+      "step": 215
+    },
+    {
+      "epoch": 0.47420417124039516,
+      "grad_norm": 9.020918504573277,
+      "learning_rate": 9.901263371994381e-07,
+      "loss": 0.4212,
+      "step": 216
+    },
+    {
+      "epoch": 0.47639956092206365,
+      "grad_norm": 9.885678723860156,
+      "learning_rate": 9.898724070780636e-07,
+      "loss": 0.3961,
+      "step": 217
+    },
+    {
+      "epoch": 0.47859495060373214,
+      "grad_norm": 8.301941197636124,
+      "learning_rate": 9.896152864006163e-07,
+      "loss": 0.4615,
+      "step": 218
+    },
+    {
+      "epoch": 0.4807903402854007,
+      "grad_norm": 6.8025751469318845,
+      "learning_rate": 9.893549768417324e-07,
+      "loss": 0.4108,
+      "step": 219
+    },
+    {
+      "epoch": 0.4829857299670692,
+      "grad_norm": 7.269628718399149,
+      "learning_rate": 9.89091480096817e-07,
+      "loss": 0.4103,
+      "step": 220
+    },
+    {
+      "epoch": 0.48518111964873767,
+      "grad_norm": 7.581903789475218,
+      "learning_rate": 9.888247978820336e-07,
+      "loss": 0.4316,
+      "step": 221
+    },
+    {
+      "epoch": 0.48737650933040616,
+      "grad_norm": 5.990696348928147,
+      "learning_rate": 9.88554931934293e-07,
+      "loss": 0.3884,
+      "step": 222
+    },
+    {
+      "epoch": 0.48957189901207465,
+      "grad_norm": 9.189550396992603,
+      "learning_rate": 9.882818840112412e-07,
+      "loss": 0.4433,
+      "step": 223
+    },
+    {
+      "epoch": 0.49176728869374314,
+      "grad_norm": 7.254234068635307,
+      "learning_rate": 9.88005655891249e-07,
+      "loss": 0.3745,
+      "step": 224
+    },
+    {
+      "epoch": 0.49396267837541163,
+      "grad_norm": 7.202880981381821,
+      "learning_rate": 9.877262493734e-07,
+      "loss": 0.3223,
+      "step": 225
+    },
+    {
+      "epoch": 0.4961580680570801,
+      "grad_norm": 7.173182667515704,
+      "learning_rate": 9.874436662774781e-07,
+      "loss": 0.403,
+      "step": 226
+    },
+    {
+      "epoch": 0.4983534577387486,
+      "grad_norm": 8.612010220051477,
+      "learning_rate": 9.871579084439573e-07,
+      "loss": 0.4335,
+      "step": 227
+    },
+    {
+      "epoch": 0.5005488474204172,
+      "grad_norm": 9.74935576526992,
+      "learning_rate": 9.868689777339882e-07,
+      "loss": 0.3635,
+      "step": 228
+    },
+    {
+      "epoch": 0.5027442371020856,
+      "grad_norm": 8.08908514021729,
+      "learning_rate": 9.865768760293865e-07,
+      "loss": 0.3858,
+      "step": 229
+    },
+    {
+      "epoch": 0.5049396267837541,
+      "grad_norm": 7.235671126052564,
+      "learning_rate": 9.862816052326207e-07,
+      "loss": 0.3883,
+      "step": 230
+    },
+    {
+      "epoch": 0.5071350164654226,
+      "grad_norm": 7.987873641702222,
+      "learning_rate": 9.859831672668001e-07,
+      "loss": 0.4294,
+      "step": 231
+    },
+    {
+      "epoch": 0.5093304061470911,
+      "grad_norm": 8.372151998510336,
+      "learning_rate": 9.856815640756614e-07,
+      "loss": 0.3888,
+      "step": 232
+    },
+    {
+      "epoch": 0.5115257958287596,
+      "grad_norm": 6.161281036256101,
+      "learning_rate": 9.85376797623557e-07,
+      "loss": 0.4193,
+      "step": 233
+    },
+    {
+      "epoch": 0.5137211855104281,
+      "grad_norm": 9.631004836143994,
+      "learning_rate": 9.850688698954408e-07,
+      "loss": 0.4649,
+      "step": 234
+    },
+    {
+      "epoch": 0.5159165751920965,
+      "grad_norm": 6.271647506242595,
+      "learning_rate": 9.847577828968574e-07,
+      "loss": 0.3697,
+      "step": 235
+    },
+    {
+      "epoch": 0.5181119648737651,
+      "grad_norm": 8.270621201274638,
+      "learning_rate": 9.84443538653927e-07,
+      "loss": 0.4409,
+      "step": 236
+    },
+    {
+      "epoch": 0.5203073545554336,
+      "grad_norm": 6.618976245336643,
+      "learning_rate": 9.841261392133334e-07,
+      "loss": 0.4548,
+      "step": 237
+    },
+    {
+      "epoch": 0.5225027442371021,
+      "grad_norm": 5.99945752806192,
+      "learning_rate": 9.838055866423101e-07,
+      "loss": 0.3965,
+      "step": 238
+    },
+    {
+      "epoch": 0.5246981339187706,
+      "grad_norm": 6.170267518062791,
+      "learning_rate": 9.834818830286274e-07,
+      "loss": 0.38,
+      "step": 239
+    },
+    {
+      "epoch": 0.5268935236004391,
+      "grad_norm": 6.8738919806439025,
+      "learning_rate": 9.83155030480578e-07,
+      "loss": 0.4275,
+      "step": 240
+    },
+    {
+      "epoch": 0.5290889132821076,
+      "grad_norm": 6.858809296180418,
+      "learning_rate": 9.82825031126964e-07,
+      "loss": 0.4014,
+      "step": 241
+    },
+    {
+      "epoch": 0.531284302963776,
+      "grad_norm": 8.082830935574739,
+      "learning_rate": 9.82491887117083e-07,
+      "loss": 0.3369,
+      "step": 242
+    },
+    {
+      "epoch": 0.5334796926454446,
+      "grad_norm": 8.12372840208911,
+      "learning_rate": 9.821556006207131e-07,
+      "loss": 0.445,
+      "step": 243
+    },
+    {
+      "epoch": 0.535675082327113,
+      "grad_norm": 7.289328702001413,
+      "learning_rate": 9.818161738281003e-07,
+      "loss": 0.3874,
+      "step": 244
+    },
+    {
+      "epoch": 0.5378704720087816,
+      "grad_norm": 7.954325034186702,
+      "learning_rate": 9.81473608949943e-07,
+      "loss": 0.3915,
+      "step": 245
+    },
+    {
+      "epoch": 0.54006586169045,
+      "grad_norm": 9.55420544614635,
+      "learning_rate": 9.811279082173783e-07,
+      "loss": 0.4023,
+      "step": 246
+    },
+    {
+      "epoch": 0.5422612513721186,
+      "grad_norm": 9.209132221937256,
+      "learning_rate": 9.80779073881967e-07,
+      "loss": 0.4711,
+      "step": 247
+    },
+    {
+      "epoch": 0.544456641053787,
+      "grad_norm": 10.926492351335712,
+      "learning_rate": 9.804271082156792e-07,
+      "loss": 0.4764,
+      "step": 248
+    },
+    {
+      "epoch": 0.5466520307354555,
+      "grad_norm": 10.270432343850993,
+      "learning_rate": 9.800720135108798e-07,
+      "loss": 0.4473,
+      "step": 249
+    },
+    {
+      "epoch": 0.5488474204171241,
+      "grad_norm": 6.965551874562504,
+      "learning_rate": 9.79713792080313e-07,
+      "loss": 0.358,
+      "step": 250
+    },
+    {
+      "epoch": 0.5488474204171241,
+      "eval_accuracy": 0.786,
+      "eval_loss": 0.3991122543811798,
+      "eval_runtime": 62.3075,
+      "eval_samples_per_second": 8.025,
+      "eval_steps_per_second": 1.011,
+      "step": 250
+    },
+    {
+      "epoch": 0.5510428100987925,
+      "grad_norm": 5.5648923276243645,
+      "learning_rate": 9.793524462570874e-07,
+      "loss": 0.4018,
+      "step": 251
+    },
+    {
+      "epoch": 0.5532381997804611,
+      "grad_norm": 6.449792094094442,
+      "learning_rate": 9.78987978394661e-07,
+      "loss": 0.4422,
+      "step": 252
+    },
+    {
+      "epoch": 0.5554335894621295,
+      "grad_norm": 8.297585054184967,
+      "learning_rate": 9.786203908668255e-07,
+      "loss": 0.4425,
+      "step": 253
+    },
+    {
+      "epoch": 0.557628979143798,
+      "grad_norm": 6.0399607888851605,
+      "learning_rate": 9.78249686067691e-07,
+      "loss": 0.3961,
+      "step": 254
+    },
+    {
+      "epoch": 0.5598243688254665,
+      "grad_norm": 5.915525652384803,
+      "learning_rate": 9.778758664116717e-07,
+      "loss": 0.4248,
+      "step": 255
+    },
+    {
+      "epoch": 0.562019758507135,
+      "grad_norm": 7.340368964125289,
+      "learning_rate": 9.774989343334675e-07,
+      "loss": 0.4029,
+      "step": 256
+    },
+    {
+      "epoch": 0.5642151481888035,
+      "grad_norm": 6.723469259580046,
+      "learning_rate": 9.771188922880501e-07,
+      "loss": 0.4067,
+      "step": 257
+    },
+    {
+      "epoch": 0.566410537870472,
+      "grad_norm": 7.335423961287345,
+      "learning_rate": 9.76735742750647e-07,
+      "loss": 0.392,
+      "step": 258
+    },
+    {
+      "epoch": 0.5686059275521405,
+      "grad_norm": 6.368966519220958,
+      "learning_rate": 9.763494882167238e-07,
+      "loss": 0.4085,
+      "step": 259
+    },
+    {
+      "epoch": 0.570801317233809,
+      "grad_norm": 7.236440549041838,
+      "learning_rate": 9.759601312019705e-07,
+      "loss": 0.4253,
+      "step": 260
+    },
+    {
+      "epoch": 0.5729967069154775,
+      "grad_norm": 7.858966758509941,
+      "learning_rate": 9.755676742422824e-07,
+      "loss": 0.3497,
+      "step": 261
+    },
+    {
+      "epoch": 0.575192096597146,
+      "grad_norm": 6.220466022528224,
+      "learning_rate": 9.751721198937457e-07,
+      "loss": 0.3581,
+      "step": 262
+    },
+    {
+      "epoch": 0.5773874862788145,
+      "grad_norm": 7.112804748553098,
+      "learning_rate": 9.747734707326194e-07,
+      "loss": 0.3295,
+      "step": 263
+    },
+    {
+      "epoch": 0.579582875960483,
+      "grad_norm": 6.352762083606537,
+      "learning_rate": 9.743717293553197e-07,
+      "loss": 0.3378,
+      "step": 264
+    },
+    {
+      "epoch": 0.5817782656421515,
+      "grad_norm": 6.836533783139976,
+      "learning_rate": 9.73966898378402e-07,
+      "loss": 0.3405,
+      "step": 265
+    },
+    {
+      "epoch": 0.58397365532382,
+      "grad_norm": 12.555084006443275,
+      "learning_rate": 9.735589804385445e-07,
+      "loss": 0.4567,
+      "step": 266
+    },
+    {
+      "epoch": 0.5861690450054885,
+      "grad_norm": 7.870737080940946,
+      "learning_rate": 9.731479781925308e-07,
+      "loss": 0.344,
+      "step": 267
+    },
+    {
+      "epoch": 0.5883644346871569,
+      "grad_norm": 6.307480660613861,
+      "learning_rate": 9.727338943172335e-07,
+      "loss": 0.3541,
+      "step": 268
+    },
+    {
+      "epoch": 0.5905598243688255,
+      "grad_norm": 7.226896694611319,
+      "learning_rate": 9.723167315095947e-07,
+      "loss": 0.384,
+      "step": 269
+    },
+    {
+      "epoch": 0.5927552140504939,
+      "grad_norm": 11.54903734204424,
+      "learning_rate": 9.718964924866108e-07,
+      "loss": 0.5079,
+      "step": 270
+    },
+    {
+      "epoch": 0.5949506037321625,
+      "grad_norm": 6.907145915491046,
+      "learning_rate": 9.71473179985313e-07,
+      "loss": 0.4009,
+      "step": 271
+    },
+    {
+      "epoch": 0.5971459934138309,
+      "grad_norm": 7.0396191618231425,
+      "learning_rate": 9.710467967627502e-07,
+      "loss": 0.466,
+      "step": 272
+    },
+    {
+      "epoch": 0.5993413830954994,
+      "grad_norm": 6.766494453919865,
+      "learning_rate": 9.706173455959713e-07,
+      "loss": 0.3855,
+      "step": 273
+    },
+    {
+      "epoch": 0.601536772777168,
+      "grad_norm": 6.281456670193716,
+      "learning_rate": 9.701848292820069e-07,
+      "loss": 0.396,
+      "step": 274
+    },
+    {
+      "epoch": 0.6037321624588364,
+      "grad_norm": 8.289418010510577,
+      "learning_rate": 9.697492506378507e-07,
+      "loss": 0.4686,
+      "step": 275
+    },
+    {
+      "epoch": 0.605927552140505,
+      "grad_norm": 5.89839143614937,
+      "learning_rate": 9.693106125004416e-07,
+      "loss": 0.3974,
+      "step": 276
+    },
+    {
+      "epoch": 0.6081229418221734,
+      "grad_norm": 6.733592968899608,
+      "learning_rate": 9.688689177266452e-07,
+      "loss": 0.4243,
+      "step": 277
+    },
+    {
+      "epoch": 0.610318331503842,
+      "grad_norm": 6.051915856327664,
+      "learning_rate": 9.684241691932347e-07,
+      "loss": 0.4216,
+      "step": 278
+    },
+    {
+      "epoch": 0.6125137211855104,
+      "grad_norm": 7.552117598173567,
+      "learning_rate": 9.679763697968732e-07,
+      "loss": 0.4042,
+      "step": 279
+    },
+    {
+      "epoch": 0.6147091108671789,
+      "grad_norm": 5.316578030214541,
+      "learning_rate": 9.675255224540934e-07,
+      "loss": 0.3462,
+      "step": 280
+    },
+    {
+      "epoch": 0.6169045005488474,
+      "grad_norm": 8.421241423628498,
+      "learning_rate": 9.6707163010128e-07,
+      "loss": 0.4272,
+      "step": 281
+    },
+    {
+      "epoch": 0.6190998902305159,
+      "grad_norm": 8.77535805726173,
+      "learning_rate": 9.666146956946496e-07,
+      "loss": 0.4288,
+      "step": 282
+    },
+    {
+      "epoch": 0.6212952799121844,
+      "grad_norm": 6.601900738237074,
+      "learning_rate": 9.661547222102321e-07,
+      "loss": 0.3846,
+      "step": 283
+    },
+    {
+      "epoch": 0.6234906695938529,
+      "grad_norm": 7.4390146186640465,
+      "learning_rate": 9.656917126438508e-07,
+      "loss": 0.4568,
+      "step": 284
+    },
+    {
+      "epoch": 0.6256860592755215,
+      "grad_norm": 5.772511915683897,
+      "learning_rate": 9.65225670011103e-07,
+      "loss": 0.3924,
+      "step": 285
+    },
+    {
+      "epoch": 0.6278814489571899,
+      "grad_norm": 6.955674433545656,
+      "learning_rate": 9.647565973473407e-07,
+      "loss": 0.4033,
+      "step": 286
+    },
+    {
+      "epoch": 0.6300768386388584,
+      "grad_norm": 8.177359289353676,
+      "learning_rate": 9.642844977076507e-07,
+      "loss": 0.4666,
+      "step": 287
+    },
+    {
+      "epoch": 0.6322722283205269,
+      "grad_norm": 6.069507174731214,
+      "learning_rate": 9.63809374166834e-07,
+      "loss": 0.4373,
+      "step": 288
+    },
+    {
+      "epoch": 0.6344676180021954,
+      "grad_norm": 7.849034116126457,
+      "learning_rate": 9.633312298193871e-07,
+      "loss": 0.447,
+      "step": 289
+    },
+    {
+      "epoch": 0.6366630076838639,
+      "grad_norm": 6.926773533221862,
+      "learning_rate": 9.62850067779481e-07,
+      "loss": 0.4244,
+      "step": 290
+    },
+    {
+      "epoch": 0.6388583973655324,
+      "grad_norm": 7.158445163028628,
+      "learning_rate": 9.623658911809404e-07,
+      "loss": 0.3881,
+      "step": 291
+    },
+    {
+      "epoch": 0.6410537870472008,
+      "grad_norm": 5.7467162895599415,
+      "learning_rate": 9.618787031772245e-07,
+      "loss": 0.3568,
+      "step": 292
+    },
+    {
+      "epoch": 0.6432491767288694,
+      "grad_norm": 6.928058648989376,
+      "learning_rate": 9.61388506941406e-07,
+      "loss": 0.4281,
+      "step": 293
+    },
+    {
+      "epoch": 0.6454445664105378,
+      "grad_norm": 6.593161246770192,
+      "learning_rate": 9.6089530566615e-07,
+      "loss": 0.359,
+      "step": 294
+    },
+    {
+      "epoch": 0.6476399560922064,
+      "grad_norm": 7.614859979484859,
+      "learning_rate": 9.603991025636933e-07,
+      "loss": 0.3895,
+      "step": 295
+    },
+    {
+      "epoch": 0.6498353457738749,
+      "grad_norm": 9.246559759408955,
+      "learning_rate": 9.598999008658241e-07,
+      "loss": 0.3959,
+      "step": 296
+    },
+    {
+      "epoch": 0.6520307354555434,
+      "grad_norm": 10.624131439351855,
+      "learning_rate": 9.59397703823861e-07,
+      "loss": 0.4265,
+      "step": 297
+    },
+    {
+      "epoch": 0.6542261251372119,
+      "grad_norm": 7.146670140793182,
+      "learning_rate": 9.588925147086303e-07,
+      "loss": 0.3586,
+      "step": 298
+    },
+    {
+      "epoch": 0.6564215148188803,
+      "grad_norm": 5.708868849662112,
+      "learning_rate": 9.583843368104464e-07,
+      "loss": 0.3444,
+      "step": 299
+    },
+    {
+      "epoch": 0.6586169045005489,
+      "grad_norm": 6.538247134761959,
+      "learning_rate": 9.578731734390898e-07,
+      "loss": 0.3721,
+      "step": 300
+    },
+    {
+      "epoch": 0.6586169045005489,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.3827642798423767,
+      "eval_runtime": 62.403,
+      "eval_samples_per_second": 8.012,
+      "eval_steps_per_second": 1.01,
+      "step": 300
+    },
+    {
+      "epoch": 0.6608122941822173,
+      "grad_norm": 5.4614975487617174,
+      "learning_rate": 9.573590279237854e-07,
+      "loss": 0.3739,
+      "step": 301
+    },
+    {
+      "epoch": 0.6630076838638859,
+      "grad_norm": 5.616403801157149,
+      "learning_rate": 9.568419036131807e-07,
+      "loss": 0.3802,
+      "step": 302
+    },
+    {
+      "epoch": 0.6652030735455543,
+      "grad_norm": 6.54627994415682,
+      "learning_rate": 9.563218038753245e-07,
+      "loss": 0.369,
+      "step": 303
+    },
+    {
+      "epoch": 0.6673984632272228,
+      "grad_norm": 7.562917302782545,
+      "learning_rate": 9.557987320976446e-07,
+      "loss": 0.4048,
+      "step": 304
+    },
+    {
+      "epoch": 0.6695938529088913,
+      "grad_norm": 5.723183604217507,
+      "learning_rate": 9.552726916869254e-07,
+      "loss": 0.3875,
+      "step": 305
+    },
+    {
+      "epoch": 0.6717892425905598,
+      "grad_norm": 5.431663961962029,
+      "learning_rate": 9.547436860692869e-07,
+      "loss": 0.4294,
+      "step": 306
+    },
+    {
+      "epoch": 0.6739846322722283,
+      "grad_norm": 5.515569016162807,
+      "learning_rate": 9.542117186901608e-07,
+      "loss": 0.3534,
+      "step": 307
+    },
+    {
+      "epoch": 0.6761800219538968,
+      "grad_norm": 5.225609276477867,
+      "learning_rate": 9.536767930142692e-07,
+      "loss": 0.3497,
+      "step": 308
+    },
+    {
+      "epoch": 0.6783754116355654,
+      "grad_norm": 5.982849222319877,
+      "learning_rate": 9.53138912525602e-07,
+      "loss": 0.3831,
+      "step": 309
+    },
+    {
+      "epoch": 0.6805708013172338,
+      "grad_norm": 9.316032754601094,
+      "learning_rate": 9.525980807273933e-07,
+      "loss": 0.4859,
+      "step": 310
+    },
+    {
+      "epoch": 0.6827661909989023,
+      "grad_norm": 7.62002827563512,
+      "learning_rate": 9.520543011420994e-07,
+      "loss": 0.4667,
+      "step": 311
+    },
+    {
+      "epoch": 0.6849615806805708,
+      "grad_norm": 7.476256847848758,
+      "learning_rate": 9.515075773113758e-07,
+      "loss": 0.3561,
+      "step": 312
+    },
+    {
+      "epoch": 0.6871569703622393,
+      "grad_norm": 6.0952453407784395,
+      "learning_rate": 9.509579127960541e-07,
+      "loss": 0.3759,
+      "step": 313
+    },
+    {
+      "epoch": 0.6893523600439078,
+      "grad_norm": 5.264194015456075,
+      "learning_rate": 9.504053111761183e-07,
+      "loss": 0.3553,
+      "step": 314
+    },
+    {
+      "epoch": 0.6915477497255763,
+      "grad_norm": 5.722501060128661,
+      "learning_rate": 9.498497760506819e-07,
+      "loss": 0.41,
+      "step": 315
+    },
+    {
+      "epoch": 0.6937431394072447,
+      "grad_norm": 5.385616225560054,
+      "learning_rate": 9.492913110379647e-07,
+      "loss": 0.3811,
+      "step": 316
+    },
+    {
+      "epoch": 0.6959385290889133,
+      "grad_norm": 6.8590494144069565,
+      "learning_rate": 9.487299197752687e-07,
+      "loss": 0.4238,
+      "step": 317
+    },
+    {
+      "epoch": 0.6981339187705817,
+      "grad_norm": 8.374420284758637,
+      "learning_rate": 9.481656059189549e-07,
+      "loss": 0.4383,
+      "step": 318
+    },
+    {
+      "epoch": 0.7003293084522503,
+      "grad_norm": 6.74093160128879,
+      "learning_rate": 9.475983731444191e-07,
+      "loss": 0.4053,
+      "step": 319
+    },
+    {
+      "epoch": 0.7025246981339188,
+      "grad_norm": 6.3600960708427445,
+      "learning_rate": 9.47028225146068e-07,
+      "loss": 0.4135,
+      "step": 320
+    },
+    {
+      "epoch": 0.7047200878155873,
+      "grad_norm": 5.118236300502877,
+      "learning_rate": 9.464551656372955e-07,
+      "loss": 0.3437,
+      "step": 321
+    },
+    {
+      "epoch": 0.7069154774972558,
+      "grad_norm": 6.0512731733617775,
+      "learning_rate": 9.458791983504581e-07,
+      "loss": 0.3719,
+      "step": 322
+    },
+    {
+      "epoch": 0.7091108671789242,
+      "grad_norm": 4.689013131515306,
+      "learning_rate": 9.453003270368509e-07,
+      "loss": 0.4211,
+      "step": 323
+    },
+    {
+      "epoch": 0.7113062568605928,
+      "grad_norm": 5.847735855655551,
+      "learning_rate": 9.44718555466683e-07,
+      "loss": 0.4383,
+      "step": 324
+    },
+    {
+      "epoch": 0.7135016465422612,
+      "grad_norm": 5.665362800637276,
+      "learning_rate": 9.44133887429053e-07,
+      "loss": 0.3934,
+      "step": 325
+    },
+    {
+      "epoch": 0.7156970362239298,
+      "grad_norm": 5.874604417597781,
+      "learning_rate": 9.435463267319239e-07,
+      "loss": 0.3583,
+      "step": 326
+    },
+    {
+      "epoch": 0.7178924259055982,
+      "grad_norm": 5.159270630284724,
+      "learning_rate": 9.429558772020992e-07,
+      "loss": 0.4252,
+      "step": 327
+    },
+    {
+      "epoch": 0.7200878155872668,
+      "grad_norm": 5.706691193740145,
+      "learning_rate": 9.423625426851973e-07,
+      "loss": 0.3727,
+      "step": 328
+    },
+    {
+      "epoch": 0.7222832052689352,
+      "grad_norm": 5.633325508736258,
+      "learning_rate": 9.417663270456267e-07,
+      "loss": 0.3591,
+      "step": 329
+    },
+    {
+      "epoch": 0.7244785949506037,
+      "grad_norm": 6.114919710777612,
+      "learning_rate": 9.411672341665604e-07,
+      "loss": 0.3658,
+      "step": 330
+    },
+    {
+      "epoch": 0.7266739846322722,
+      "grad_norm": 7.068818492445254,
+      "learning_rate": 9.405652679499115e-07,
+      "loss": 0.3954,
+      "step": 331
+    },
+    {
+      "epoch": 0.7288693743139407,
+      "grad_norm": 6.02224410411034,
+      "learning_rate": 9.399604323163068e-07,
+      "loss": 0.346,
+      "step": 332
+    },
+    {
+      "epoch": 0.7310647639956093,
+      "grad_norm": 8.959081519304984,
+      "learning_rate": 9.393527312050617e-07,
+      "loss": 0.4459,
+      "step": 333
+    },
+    {
+      "epoch": 0.7332601536772777,
+      "grad_norm": 7.542390397020765,
+      "learning_rate": 9.387421685741552e-07,
+      "loss": 0.4022,
+      "step": 334
+    },
+    {
+      "epoch": 0.7354555433589463,
+      "grad_norm": 6.8197296180264555,
+      "learning_rate": 9.381287484002027e-07,
+      "loss": 0.3836,
+      "step": 335
+    },
+    {
+      "epoch": 0.7376509330406147,
+      "grad_norm": 5.343272686170338,
+      "learning_rate": 9.375124746784311e-07,
+      "loss": 0.3561,
+      "step": 336
+    },
+    {
+      "epoch": 0.7398463227222832,
+      "grad_norm": 6.461455373048154,
+      "learning_rate": 9.368933514226529e-07,
+      "loss": 0.4096,
+      "step": 337
+    },
+    {
+      "epoch": 0.7420417124039517,
+      "grad_norm": 4.854050051488241,
+      "learning_rate": 9.362713826652392e-07,
+      "loss": 0.3814,
+      "step": 338
+    },
+    {
+      "epoch": 0.7442371020856202,
+      "grad_norm": 5.286657056414475,
+      "learning_rate": 9.356465724570943e-07,
+      "loss": 0.3989,
+      "step": 339
+    },
+    {
+      "epoch": 0.7464324917672887,
+      "grad_norm": 5.696140592098042,
+      "learning_rate": 9.350189248676292e-07,
+      "loss": 0.4114,
+      "step": 340
+    },
+    {
+      "epoch": 0.7486278814489572,
+      "grad_norm": 5.963578851614639,
+      "learning_rate": 9.34388443984734e-07,
+      "loss": 0.4103,
+      "step": 341
+    },
+    {
+      "epoch": 0.7508232711306256,
+      "grad_norm": 5.167078850111144,
+      "learning_rate": 9.33755133914753e-07,
+      "loss": 0.3987,
+      "step": 342
+    },
+    {
+      "epoch": 0.7530186608122942,
+      "grad_norm": 5.365795093462018,
+      "learning_rate": 9.331189987824568e-07,
+      "loss": 0.414,
+      "step": 343
+    },
+    {
+      "epoch": 0.7552140504939627,
+      "grad_norm": 5.155962282558951,
+      "learning_rate": 9.324800427310155e-07,
+      "loss": 0.4107,
+      "step": 344
+    },
+    {
+      "epoch": 0.7574094401756312,
+      "grad_norm": 4.685610227798769,
+      "learning_rate": 9.318382699219722e-07,
+      "loss": 0.383,
+      "step": 345
+    },
+    {
+      "epoch": 0.7596048298572997,
+      "grad_norm": 5.451910902227303,
+      "learning_rate": 9.311936845352157e-07,
+      "loss": 0.3701,
+      "step": 346
+    },
+    {
+      "epoch": 0.7618002195389681,
+      "grad_norm": 4.878532494519716,
+      "learning_rate": 9.305462907689532e-07,
+      "loss": 0.3889,
+      "step": 347
+    },
+    {
+      "epoch": 0.7639956092206367,
+      "grad_norm": 6.488604304889017,
+      "learning_rate": 9.298960928396826e-07,
+      "loss": 0.3769,
+      "step": 348
+    },
+    {
+      "epoch": 0.7661909989023051,
+      "grad_norm": 7.230743847370782,
+      "learning_rate": 9.292430949821659e-07,
+      "loss": 0.3918,
+      "step": 349
+    },
+    {
+      "epoch": 0.7683863885839737,
+      "grad_norm": 8.84856037173542,
+      "learning_rate": 9.285873014494008e-07,
+      "loss": 0.4351,
+      "step": 350
+    },
+    {
+      "epoch": 0.7683863885839737,
+      "eval_accuracy": 0.818,
+      "eval_loss": 0.36839187145233154,
+      "eval_runtime": 62.4628,
+      "eval_samples_per_second": 8.005,
+      "eval_steps_per_second": 1.009,
+      "step": 350
+    },
+    {
+      "epoch": 0.7705817782656421,
+      "grad_norm": 5.617691309816291,
+      "learning_rate": 9.279287165125936e-07,
+      "loss": 0.3329,
+      "step": 351
+    },
+    {
+      "epoch": 0.7727771679473107,
+      "grad_norm": 10.749385714028035,
+      "learning_rate": 9.272673444611308e-07,
+      "loss": 0.4227,
+      "step": 352
+    },
+    {
+      "epoch": 0.7749725576289791,
+      "grad_norm": 6.679314682583184,
+      "learning_rate": 9.266031896025516e-07,
+      "loss": 0.3981,
+      "step": 353
+    },
+    {
+      "epoch": 0.7771679473106476,
+      "grad_norm": 7.906535171339716,
+      "learning_rate": 9.259362562625199e-07,
+      "loss": 0.3457,
+      "step": 354
+    },
+    {
+      "epoch": 0.7793633369923162,
+      "grad_norm": 6.476745284566722,
+      "learning_rate": 9.252665487847957e-07,
+      "loss": 0.351,
+      "step": 355
+    },
+    {
+      "epoch": 0.7815587266739846,
+      "grad_norm": 7.490252907466211,
+      "learning_rate": 9.245940715312074e-07,
+      "loss": 0.3614,
+      "step": 356
+    },
+    {
+      "epoch": 0.7837541163556532,
+      "grad_norm": 10.294959027911519,
+      "learning_rate": 9.239188288816226e-07,
+      "loss": 0.4772,
+      "step": 357
+    },
+    {
+      "epoch": 0.7859495060373216,
+      "grad_norm": 6.749458938354777,
+      "learning_rate": 9.232408252339201e-07,
+      "loss": 0.4137,
+      "step": 358
+    },
+    {
+      "epoch": 0.7881448957189902,
+      "grad_norm": 5.601238558514638,
+      "learning_rate": 9.225600650039615e-07,
+      "loss": 0.4017,
+      "step": 359
+    },
+    {
+      "epoch": 0.7903402854006586,
+      "grad_norm": 5.914207463946434,
+      "learning_rate": 9.218765526255619e-07,
+      "loss": 0.3949,
+      "step": 360
+    },
+    {
+      "epoch": 0.7925356750823271,
+      "grad_norm": 5.60976526230607,
+      "learning_rate": 9.211902925504613e-07,
+      "loss": 0.3947,
+      "step": 361
+    },
+    {
+      "epoch": 0.7947310647639956,
+      "grad_norm": 6.905598884857724,
+      "learning_rate": 9.205012892482952e-07,
+      "loss": 0.3593,
+      "step": 362
+    },
+    {
+      "epoch": 0.7969264544456641,
+      "grad_norm": 5.708597195424006,
+      "learning_rate": 9.198095472065667e-07,
+      "loss": 0.3845,
+      "step": 363
+    },
+    {
+      "epoch": 0.7991218441273326,
+      "grad_norm": 6.3826311537728415,
+      "learning_rate": 9.191150709306155e-07,
+      "loss": 0.426,
+      "step": 364
+    },
+    {
+      "epoch": 0.8013172338090011,
+      "grad_norm": 5.704176371314256,
+      "learning_rate": 9.184178649435896e-07,
+      "loss": 0.3623,
+      "step": 365
+    },
+    {
+      "epoch": 0.8035126234906695,
+      "grad_norm": 6.538927998752732,
+      "learning_rate": 9.177179337864163e-07,
+      "loss": 0.4081,
+      "step": 366
+    },
+    {
+      "epoch": 0.8057080131723381,
+      "grad_norm": 5.417066774084007,
+      "learning_rate": 9.170152820177714e-07,
+      "loss": 0.326,
+      "step": 367
+    },
+    {
+      "epoch": 0.8079034028540066,
+      "grad_norm": 5.7045823769473145,
+      "learning_rate": 9.163099142140505e-07,
+      "loss": 0.3464,
+      "step": 368
+    },
+    {
+      "epoch": 0.8100987925356751,
+      "grad_norm": 6.709508312095118,
+      "learning_rate": 9.156018349693386e-07,
+      "loss": 0.3343,
+      "step": 369
+    },
+    {
+      "epoch": 0.8122941822173436,
+      "grad_norm": 8.175912377684108,
+      "learning_rate": 9.148910488953807e-07,
+      "loss": 0.4461,
+      "step": 370
+    },
+    {
+      "epoch": 0.814489571899012,
+      "grad_norm": 6.172712289986408,
+      "learning_rate": 9.141775606215512e-07,
+      "loss": 0.3778,
+      "step": 371
+    },
+    {
+      "epoch": 0.8166849615806806,
+      "grad_norm": 9.110817197454034,
+      "learning_rate": 9.134613747948238e-07,
+      "loss": 0.4277,
+      "step": 372
+    },
+    {
+      "epoch": 0.818880351262349,
+      "grad_norm": 7.730875175896815,
+      "learning_rate": 9.127424960797423e-07,
+      "loss": 0.3732,
+      "step": 373
+    },
+    {
+      "epoch": 0.8210757409440176,
+      "grad_norm": 6.560594986117142,
+      "learning_rate": 9.120209291583885e-07,
+      "loss": 0.413,
+      "step": 374
+    },
+    {
+      "epoch": 0.823271130625686,
+      "grad_norm": 7.2498236273803,
+      "learning_rate": 9.11296678730353e-07,
+      "loss": 0.4097,
+      "step": 375
+    },
+    {
+      "epoch": 0.8254665203073546,
+      "grad_norm": 5.069648552101249,
+      "learning_rate": 9.10569749512704e-07,
+      "loss": 0.415,
+      "step": 376
+    },
+    {
+      "epoch": 0.827661909989023,
+      "grad_norm": 6.5382746464808355,
+      "learning_rate": 9.098401462399572e-07,
+      "loss": 0.3679,
+      "step": 377
+    },
+    {
+      "epoch": 0.8298572996706916,
+      "grad_norm": 7.063308742163228,
+      "learning_rate": 9.091078736640438e-07,
+      "loss": 0.3834,
+      "step": 378
+    },
+    {
+      "epoch": 0.8320526893523601,
+      "grad_norm": 5.3524000614593445,
+      "learning_rate": 9.083729365542807e-07,
+      "loss": 0.4322,
+      "step": 379
+    },
+    {
+      "epoch": 0.8342480790340285,
+      "grad_norm": 5.251326032637674,
+      "learning_rate": 9.076353396973391e-07,
+      "loss": 0.3737,
+      "step": 380
+    },
+    {
+      "epoch": 0.8364434687156971,
+      "grad_norm": 5.606890380106128,
+      "learning_rate": 9.068950878972128e-07,
+      "loss": 0.3916,
+      "step": 381
+    },
+    {
+      "epoch": 0.8386388583973655,
+      "grad_norm": 5.608911929639908,
+      "learning_rate": 9.06152185975188e-07,
+      "loss": 0.3895,
+      "step": 382
+    },
+    {
+      "epoch": 0.8408342480790341,
+      "grad_norm": 5.326991715289215,
+      "learning_rate": 9.054066387698103e-07,
+      "loss": 0.3505,
+      "step": 383
+    },
+    {
+      "epoch": 0.8430296377607025,
+      "grad_norm": 5.002446774906431,
+      "learning_rate": 9.04658451136855e-07,
+      "loss": 0.3567,
+      "step": 384
+    },
+    {
+      "epoch": 0.845225027442371,
+      "grad_norm": 6.2353029098708355,
+      "learning_rate": 9.039076279492938e-07,
+      "loss": 0.3654,
+      "step": 385
+    },
+    {
+      "epoch": 0.8474204171240395,
+      "grad_norm": 5.383798331488236,
+      "learning_rate": 9.03154174097265e-07,
+      "loss": 0.3528,
+      "step": 386
+    },
+    {
+      "epoch": 0.849615806805708,
+      "grad_norm": 7.522503969731689,
+      "learning_rate": 9.023980944880395e-07,
+      "loss": 0.3802,
+      "step": 387
+    },
+    {
+      "epoch": 0.8518111964873765,
+      "grad_norm": 5.624816423254138,
+      "learning_rate": 9.016393940459901e-07,
+      "loss": 0.3226,
+      "step": 388
+    },
+    {
+      "epoch": 0.854006586169045,
+      "grad_norm": 9.036186780868068,
+      "learning_rate": 9.008780777125592e-07,
+      "loss": 0.3459,
+      "step": 389
+    },
+    {
+      "epoch": 0.8562019758507134,
+      "grad_norm": 7.771864470871327,
+      "learning_rate": 9.001141504462267e-07,
+      "loss": 0.3945,
+      "step": 390
+    },
+    {
+      "epoch": 0.858397365532382,
+      "grad_norm": 10.948074987698849,
+      "learning_rate": 8.993476172224776e-07,
+      "loss": 0.3845,
+      "step": 391
+    },
+    {
+      "epoch": 0.8605927552140505,
+      "grad_norm": 6.279681321387117,
+      "learning_rate": 8.985784830337694e-07,
+      "loss": 0.3346,
+      "step": 392
+    },
+    {
+      "epoch": 0.862788144895719,
+      "grad_norm": 6.22711354144465,
+      "learning_rate": 8.978067528895001e-07,
+      "loss": 0.3589,
+      "step": 393
+    },
+    {
+      "epoch": 0.8649835345773875,
+      "grad_norm": 8.754649020649845,
+      "learning_rate": 8.970324318159747e-07,
+      "loss": 0.4202,
+      "step": 394
+    },
+    {
+      "epoch": 0.867178924259056,
+      "grad_norm": 6.308329054889705,
+      "learning_rate": 8.962555248563737e-07,
+      "loss": 0.3964,
+      "step": 395
+    },
+    {
+      "epoch": 0.8693743139407245,
+      "grad_norm": 5.393641757738077,
+      "learning_rate": 8.95476037070719e-07,
+      "loss": 0.4054,
+      "step": 396
+    },
+    {
+      "epoch": 0.8715697036223929,
+      "grad_norm": 5.6298700405971065,
+      "learning_rate": 8.94693973535842e-07,
+      "loss": 0.3947,
+      "step": 397
+    },
+    {
+      "epoch": 0.8737650933040615,
+      "grad_norm": 4.205803068097396,
+      "learning_rate": 8.939093393453494e-07,
+      "loss": 0.3618,
+      "step": 398
+    },
+    {
+      "epoch": 0.8759604829857299,
+      "grad_norm": 5.473714761916757,
+      "learning_rate": 8.931221396095914e-07,
+      "loss": 0.3509,
+      "step": 399
+    },
+    {
+      "epoch": 0.8781558726673985,
+      "grad_norm": 7.034796071540623,
+      "learning_rate": 8.92332379455627e-07,
+      "loss": 0.3482,
+      "step": 400
+    },
+    {
+      "epoch": 0.8781558726673985,
+      "eval_accuracy": 0.79,
+      "eval_loss": 0.38279932737350464,
+      "eval_runtime": 62.3937,
+      "eval_samples_per_second": 8.014,
+      "eval_steps_per_second": 1.01,
+      "step": 400
+    },
+    {
+      "epoch": 0.8803512623490669,
+      "grad_norm": 5.9849568856182955,
+      "learning_rate": 8.91540064027192e-07,
+      "loss": 0.3945,
+      "step": 401
+    },
+    {
+      "epoch": 0.8825466520307355,
+      "grad_norm": 4.900349286406815,
+      "learning_rate": 8.907451984846642e-07,
+      "loss": 0.3374,
+      "step": 402
+    },
+    {
+      "epoch": 0.884742041712404,
+      "grad_norm": 7.015373419320435,
+      "learning_rate": 8.899477880050305e-07,
+      "loss": 0.4297,
+      "step": 403
+    },
+    {
+      "epoch": 0.8869374313940724,
+      "grad_norm": 5.211500475863296,
+      "learning_rate": 8.891478377818533e-07,
+      "loss": 0.3425,
+      "step": 404
+    },
+    {
+      "epoch": 0.889132821075741,
+      "grad_norm": 4.850386411589254,
+      "learning_rate": 8.883453530252363e-07,
+      "loss": 0.3479,
+      "step": 405
+    },
+    {
+      "epoch": 0.8913282107574094,
+      "grad_norm": 5.755909710211362,
+      "learning_rate": 8.875403389617909e-07,
+      "loss": 0.4106,
+      "step": 406
+    },
+    {
+      "epoch": 0.893523600439078,
+      "grad_norm": 6.753928081610862,
+      "learning_rate": 8.867328008346012e-07,
+      "loss": 0.3862,
+      "step": 407
+    },
+    {
+      "epoch": 0.8957189901207464,
+      "grad_norm": 5.767435648190633,
+      "learning_rate": 8.859227439031917e-07,
+      "loss": 0.389,
+      "step": 408
+    },
+    {
+      "epoch": 0.897914379802415,
+      "grad_norm": 5.3650473270965975,
+      "learning_rate": 8.851101734434916e-07,
+      "loss": 0.3696,
+      "step": 409
+    },
+    {
+      "epoch": 0.9001097694840834,
+      "grad_norm": 6.082270492915926,
+      "learning_rate": 8.842950947478001e-07,
+      "loss": 0.3817,
+      "step": 410
+    },
+    {
+      "epoch": 0.9023051591657519,
+      "grad_norm": 4.627051115673926,
+      "learning_rate": 8.834775131247534e-07,
+      "loss": 0.3615,
+      "step": 411
+    },
+    {
+      "epoch": 0.9045005488474204,
+      "grad_norm": 4.703510731345825,
+      "learning_rate": 8.826574338992893e-07,
+      "loss": 0.3487,
+      "step": 412
+    },
+    {
+      "epoch": 0.9066959385290889,
+      "grad_norm": 4.403248729861,
+      "learning_rate": 8.818348624126122e-07,
+      "loss": 0.3409,
+      "step": 413
+    },
+    {
+      "epoch": 0.9088913282107574,
+      "grad_norm": 7.226848694927689,
+      "learning_rate": 8.810098040221588e-07,
+      "loss": 0.3674,
+      "step": 414
+    },
+    {
+      "epoch": 0.9110867178924259,
+      "grad_norm": 6.0854109691994145,
+      "learning_rate": 8.801822641015635e-07,
+      "loss": 0.3669,
+      "step": 415
+    },
+    {
+      "epoch": 0.9132821075740944,
+      "grad_norm": 5.0811223784714725,
+      "learning_rate": 8.793522480406223e-07,
+      "loss": 0.3236,
+      "step": 416
+    },
+    {
+      "epoch": 0.9154774972557629,
+      "grad_norm": 7.237396760486593,
+      "learning_rate": 8.785197612452591e-07,
+      "loss": 0.4279,
+      "step": 417
+    },
+    {
+      "epoch": 0.9176728869374314,
+      "grad_norm": 5.505858472667137,
+      "learning_rate": 8.776848091374892e-07,
+      "loss": 0.3658,
+      "step": 418
+    },
+    {
+      "epoch": 0.9198682766190999,
+      "grad_norm": 7.040175733229395,
+      "learning_rate": 8.768473971553847e-07,
+      "loss": 0.3852,
+      "step": 419
+    },
+    {
+      "epoch": 0.9220636663007684,
+      "grad_norm": 5.592000526909508,
+      "learning_rate": 8.760075307530392e-07,
+      "loss": 0.3306,
+      "step": 420
+    },
+    {
+      "epoch": 0.9242590559824369,
+      "grad_norm": 6.366945719939024,
+      "learning_rate": 8.75165215400532e-07,
+      "loss": 0.3906,
+      "step": 421
+    },
+    {
+      "epoch": 0.9264544456641054,
+      "grad_norm": 7.370004771820366,
+      "learning_rate": 8.743204565838922e-07,
+      "loss": 0.3853,
+      "step": 422
+    },
+    {
+      "epoch": 0.9286498353457738,
+      "grad_norm": 6.381558651368943,
+      "learning_rate": 8.734732598050636e-07,
+      "loss": 0.4168,
+      "step": 423
+    },
+    {
+      "epoch": 0.9308452250274424,
+      "grad_norm": 5.053966639028662,
+      "learning_rate": 8.726236305818681e-07,
+      "loss": 0.3842,
+      "step": 424
+    },
+    {
+      "epoch": 0.9330406147091108,
+      "grad_norm": 5.398688693219092,
+      "learning_rate": 8.717715744479706e-07,
+      "loss": 0.3838,
+      "step": 425
+    },
+    {
+      "epoch": 0.9352360043907794,
+      "grad_norm": 5.2483617600811625,
+      "learning_rate": 8.709170969528425e-07,
+      "loss": 0.4371,
+      "step": 426
+    },
+    {
+      "epoch": 0.9374313940724479,
+      "grad_norm": 4.227523591223431,
+      "learning_rate": 8.700602036617253e-07,
+      "loss": 0.3653,
+      "step": 427
+    },
+    {
+      "epoch": 0.9396267837541163,
+      "grad_norm": 5.346394544482726,
+      "learning_rate": 8.692009001555951e-07,
+      "loss": 0.4071,
+      "step": 428
+    },
+    {
+      "epoch": 0.9418221734357849,
+      "grad_norm": 5.868340936821356,
+      "learning_rate": 8.683391920311256e-07,
+      "loss": 0.415,
+      "step": 429
+    },
+    {
+      "epoch": 0.9440175631174533,
+      "grad_norm": 3.8517100242733346,
+      "learning_rate": 8.674750849006518e-07,
+      "loss": 0.3518,
+      "step": 430
+    },
+    {
+      "epoch": 0.9462129527991219,
+      "grad_norm": 5.514252217717242,
+      "learning_rate": 8.666085843921337e-07,
+      "loss": 0.3563,
+      "step": 431
+    },
+    {
+      "epoch": 0.9484083424807903,
+      "grad_norm": 4.306700297128568,
+      "learning_rate": 8.65739696149119e-07,
+      "loss": 0.3731,
+      "step": 432
+    },
+    {
+      "epoch": 0.9506037321624589,
+      "grad_norm": 5.318133544247306,
+      "learning_rate": 8.648684258307075e-07,
+      "loss": 0.4034,
+      "step": 433
+    },
+    {
+      "epoch": 0.9527991218441273,
+      "grad_norm": 6.306808553382266,
+      "learning_rate": 8.639947791115131e-07,
+      "loss": 0.3843,
+      "step": 434
+    },
+    {
+      "epoch": 0.9549945115257958,
+      "grad_norm": 4.683006509880547,
+      "learning_rate": 8.631187616816271e-07,
+      "loss": 0.3426,
+      "step": 435
+    },
+    {
+      "epoch": 0.9571899012074643,
+      "grad_norm": 5.272777856944059,
+      "learning_rate": 8.622403792465819e-07,
+      "loss": 0.3919,
+      "step": 436
+    },
+    {
+      "epoch": 0.9593852908891328,
+      "grad_norm": 6.567051623139762,
+      "learning_rate": 8.613596375273127e-07,
+      "loss": 0.3334,
+      "step": 437
+    },
+    {
+      "epoch": 0.9615806805708014,
+      "grad_norm": 4.661181804857022,
+      "learning_rate": 8.604765422601213e-07,
+      "loss": 0.3393,
+      "step": 438
+    },
+    {
+      "epoch": 0.9637760702524698,
+      "grad_norm": 7.723055911423496,
+      "learning_rate": 8.595910991966375e-07,
+      "loss": 0.3987,
+      "step": 439
+    },
+    {
+      "epoch": 0.9659714599341384,
+      "grad_norm": 5.1917875018161705,
+      "learning_rate": 8.587033141037833e-07,
+      "loss": 0.3833,
+      "step": 440
+    },
+    {
+      "epoch": 0.9681668496158068,
+      "grad_norm": 4.93039231657318,
+      "learning_rate": 8.578131927637339e-07,
+      "loss": 0.321,
+      "step": 441
+    },
+    {
+      "epoch": 0.9703622392974753,
+      "grad_norm": 5.662266928236786,
+      "learning_rate": 8.569207409738804e-07,
+      "loss": 0.3442,
+      "step": 442
+    },
+    {
+      "epoch": 0.9725576289791438,
+      "grad_norm": 7.530776092189444,
+      "learning_rate": 8.560259645467927e-07,
+      "loss": 0.3823,
+      "step": 443
+    },
+    {
+      "epoch": 0.9747530186608123,
+      "grad_norm": 5.722097050543397,
+      "learning_rate": 8.551288693101808e-07,
+      "loss": 0.3381,
+      "step": 444
+    },
+    {
+      "epoch": 0.9769484083424808,
+      "grad_norm": 5.2154253758252365,
+      "learning_rate": 8.542294611068573e-07,
+      "loss": 0.355,
+      "step": 445
+    },
+    {
+      "epoch": 0.9791437980241493,
+      "grad_norm": 6.183565635622652,
+      "learning_rate": 8.533277457946988e-07,
+      "loss": 0.3469,
+      "step": 446
+    },
+    {
+      "epoch": 0.9813391877058177,
+      "grad_norm": 6.199123575604012,
+      "learning_rate": 8.524237292466092e-07,
+      "loss": 0.365,
+      "step": 447
+    },
+    {
+      "epoch": 0.9835345773874863,
+      "grad_norm": 5.22531850121621,
+      "learning_rate": 8.515174173504795e-07,
+      "loss": 0.3699,
+      "step": 448
+    },
+    {
+      "epoch": 0.9857299670691547,
+      "grad_norm": 4.777247067342119,
+      "learning_rate": 8.506088160091506e-07,
+      "loss": 0.3778,
+      "step": 449
+    },
+    {
+      "epoch": 0.9879253567508233,
+      "grad_norm": 5.7270931969707854,
+      "learning_rate": 8.49697931140375e-07,
+      "loss": 0.3722,
+      "step": 450
+    },
+    {
+      "epoch": 0.9879253567508233,
+      "eval_accuracy": 0.812,
+      "eval_loss": 0.3796866834163666,
+      "eval_runtime": 62.6515,
+      "eval_samples_per_second": 7.981,
+      "eval_steps_per_second": 1.006,
+      "step": 450
+    },
+    {
+      "epoch": 0.9901207464324918,
+      "grad_norm": 5.722657890993041,
+      "learning_rate": 8.487847686767771e-07,
+      "loss": 0.3671,
+      "step": 451
+    },
+    {
+      "epoch": 0.9923161361141603,
+      "grad_norm": 6.144552426739494,
+      "learning_rate": 8.478693345658165e-07,
+      "loss": 0.3654,
+      "step": 452
+    },
+    {
+      "epoch": 0.9945115257958288,
+      "grad_norm": 6.225509000256566,
+      "learning_rate": 8.469516347697472e-07,
+      "loss": 0.3878,
+      "step": 453
+    },
+    {
+      "epoch": 0.9967069154774972,
+      "grad_norm": 6.652501158992634,
+      "learning_rate": 8.460316752655798e-07,
+      "loss": 0.3277,
+      "step": 454
+    },
+    {
+      "epoch": 0.9989023051591658,
+      "grad_norm": 4.827682071757918,
+      "learning_rate": 8.451094620450431e-07,
+      "loss": 0.3394,
+      "step": 455
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 4.827682071757918,
+      "learning_rate": 8.441850011145435e-07,
+      "loss": 0.3908,
+      "step": 456
+    },
+    {
+      "epoch": 1.0021953896816684,
+      "grad_norm": 8.237448030205368,
+      "learning_rate": 8.432582984951276e-07,
+      "loss": 0.3649,
+      "step": 457
+    },
+    {
+      "epoch": 1.004390779363337,
+      "grad_norm": 5.244969465316055,
+      "learning_rate": 8.423293602224417e-07,
+      "loss": 0.3766,
+      "step": 458
+    },
+    {
+      "epoch": 1.0065861690450055,
+      "grad_norm": 5.735807992043271,
+      "learning_rate": 8.413981923466932e-07,
+      "loss": 0.3734,
+      "step": 459
+    },
+    {
+      "epoch": 1.008781558726674,
+      "grad_norm": 6.279435381847301,
+      "learning_rate": 8.404648009326111e-07,
+      "loss": 0.3716,
+      "step": 460
+    },
+    {
+      "epoch": 1.0109769484083424,
+      "grad_norm": 4.020301331975729,
+      "learning_rate": 8.395291920594061e-07,
+      "loss": 0.3068,
+      "step": 461
+    },
+    {
+      "epoch": 1.013172338090011,
+      "grad_norm": 5.932445501255306,
+      "learning_rate": 8.385913718207313e-07,
+      "loss": 0.3635,
+      "step": 462
+    },
+    {
+      "epoch": 1.0153677277716795,
+      "grad_norm": 6.129385754463756,
+      "learning_rate": 8.376513463246429e-07,
+      "loss": 0.3688,
+      "step": 463
+    },
+    {
+      "epoch": 1.017563117453348,
+      "grad_norm": 5.640316298247342,
+      "learning_rate": 8.367091216935596e-07,
+      "loss": 0.3777,
+      "step": 464
+    },
+    {
+      "epoch": 1.0197585071350164,
+      "grad_norm": 5.53344558624335,
+      "learning_rate": 8.357647040642231e-07,
+      "loss": 0.318,
+      "step": 465
+    },
+    {
+      "epoch": 1.021953896816685,
+      "grad_norm": 5.528084412484368,
+      "learning_rate": 8.348180995876587e-07,
+      "loss": 0.344,
+      "step": 466
+    },
+    {
+      "epoch": 1.0241492864983535,
+      "grad_norm": 4.720679027078674,
+      "learning_rate": 8.338693144291342e-07,
+      "loss": 0.3476,
+      "step": 467
+    },
+    {
+      "epoch": 1.026344676180022,
+      "grad_norm": 4.8918294986912585,
+      "learning_rate": 8.329183547681205e-07,
+      "loss": 0.3584,
+      "step": 468
+    },
+    {
+      "epoch": 1.0285400658616906,
+      "grad_norm": 6.727667695338807,
+      "learning_rate": 8.319652267982508e-07,
+      "loss": 0.3377,
+      "step": 469
+    },
+    {
+      "epoch": 1.030735455543359,
+      "grad_norm": 5.867502190382494,
+      "learning_rate": 8.310099367272812e-07,
+      "loss": 0.3591,
+      "step": 470
+    },
+    {
+      "epoch": 1.0329308452250274,
+      "grad_norm": 6.568389714898798,
+      "learning_rate": 8.30052490777049e-07,
+      "loss": 0.434,
+      "step": 471
+    },
+    {
+      "epoch": 1.0351262349066959,
+      "grad_norm": 5.375130488935029,
+      "learning_rate": 8.29092895183433e-07,
+      "loss": 0.3371,
+      "step": 472
+    },
+    {
+      "epoch": 1.0373216245883645,
+      "grad_norm": 5.0716837196261935,
+      "learning_rate": 8.281311561963129e-07,
+      "loss": 0.3583,
+      "step": 473
+    },
+    {
+      "epoch": 1.039517014270033,
+      "grad_norm": 4.9710453814081035,
+      "learning_rate": 8.271672800795284e-07,
+      "loss": 0.3454,
+      "step": 474
+    },
+    {
+      "epoch": 1.0417124039517014,
+      "grad_norm": 5.372729941376803,
+      "learning_rate": 8.26201273110838e-07,
+      "loss": 0.4327,
+      "step": 475
+    },
+    {
+      "epoch": 1.0439077936333698,
+      "grad_norm": 6.047019069783535,
+      "learning_rate": 8.252331415818788e-07,
+      "loss": 0.3166,
+      "step": 476
+    },
+    {
+      "epoch": 1.0461031833150385,
+      "grad_norm": 4.894109855984722,
+      "learning_rate": 8.242628917981253e-07,
+      "loss": 0.3486,
+      "step": 477
+    },
+    {
+      "epoch": 1.048298572996707,
+      "grad_norm": 4.960287165829203,
+      "learning_rate": 8.232905300788484e-07,
+      "loss": 0.3683,
+      "step": 478
+    },
+    {
+      "epoch": 1.0504939626783754,
+      "grad_norm": 5.9743260148319415,
+      "learning_rate": 8.223160627570736e-07,
+      "loss": 0.2962,
+      "step": 479
+    },
+    {
+      "epoch": 1.0526893523600438,
+      "grad_norm": 4.421002426118984,
+      "learning_rate": 8.213394961795406e-07,
+      "loss": 0.353,
+      "step": 480
+    },
+    {
+      "epoch": 1.0548847420417125,
+      "grad_norm": 5.826816362211787,
+      "learning_rate": 8.203608367066615e-07,
+      "loss": 0.4134,
+      "step": 481
+    },
+    {
+      "epoch": 1.057080131723381,
+      "grad_norm": 6.328500780012271,
+      "learning_rate": 8.193800907124798e-07,
+      "loss": 0.3944,
+      "step": 482
+    },
+    {
+      "epoch": 1.0592755214050493,
+      "grad_norm": 5.761570603930255,
+      "learning_rate": 8.183972645846282e-07,
+      "loss": 0.3999,
+      "step": 483
+    },
+    {
+      "epoch": 1.061470911086718,
+      "grad_norm": 7.269393542940046,
+      "learning_rate": 8.174123647242877e-07,
+      "loss": 0.3465,
+      "step": 484
+    },
+    {
+      "epoch": 1.0636663007683864,
+      "grad_norm": 5.531846580927394,
+      "learning_rate": 8.164253975461453e-07,
+      "loss": 0.3688,
+      "step": 485
+    },
+    {
+      "epoch": 1.0658616904500549,
+      "grad_norm": 4.776075343743403,
+      "learning_rate": 8.154363694783526e-07,
+      "loss": 0.3257,
+      "step": 486
+    },
+    {
+      "epoch": 1.0680570801317233,
+      "grad_norm": 5.378257513897468,
+      "learning_rate": 8.14445286962484e-07,
+      "loss": 0.3389,
+      "step": 487
+    },
+    {
+      "epoch": 1.070252469813392,
+      "grad_norm": 6.510185696868045,
+      "learning_rate": 8.134521564534947e-07,
+      "loss": 0.3559,
+      "step": 488
+    },
+    {
+      "epoch": 1.0724478594950604,
+      "grad_norm": 6.155857107945347,
+      "learning_rate": 8.124569844196779e-07,
+      "loss": 0.3617,
+      "step": 489
+    },
+    {
+      "epoch": 1.0746432491767288,
+      "grad_norm": 6.779049354813868,
+      "learning_rate": 8.11459777342624e-07,
+      "loss": 0.3594,
+      "step": 490
+    },
+    {
+      "epoch": 1.0768386388583973,
+      "grad_norm": 7.620793847029099,
+      "learning_rate": 8.104605417171776e-07,
+      "loss": 0.3984,
+      "step": 491
+    },
+    {
+      "epoch": 1.079034028540066,
+      "grad_norm": 5.064790231482452,
+      "learning_rate": 8.094592840513949e-07,
+      "loss": 0.3915,
+      "step": 492
+    },
+    {
+      "epoch": 1.0812294182217344,
+      "grad_norm": 5.422927842364402,
+      "learning_rate": 8.084560108665023e-07,
+      "loss": 0.3322,
+      "step": 493
+    },
+    {
+      "epoch": 1.0834248079034028,
+      "grad_norm": 4.466736209109948,
+      "learning_rate": 8.074507286968528e-07,
+      "loss": 0.3352,
+      "step": 494
+    },
+    {
+      "epoch": 1.0856201975850714,
+      "grad_norm": 6.338480704389545,
+      "learning_rate": 8.064434440898844e-07,
+      "loss": 0.3471,
+      "step": 495
+    },
+    {
+      "epoch": 1.0878155872667399,
+      "grad_norm": 5.4065627028591425,
+      "learning_rate": 8.054341636060766e-07,
+      "loss": 0.3549,
+      "step": 496
+    },
+    {
+      "epoch": 1.0900109769484083,
+      "grad_norm": 6.850990410700852,
+      "learning_rate": 8.044228938189088e-07,
+      "loss": 0.3903,
+      "step": 497
+    },
+    {
+      "epoch": 1.0922063666300768,
+      "grad_norm": 6.120484367930399,
+      "learning_rate": 8.034096413148161e-07,
+      "loss": 0.3097,
+      "step": 498
+    },
+    {
+      "epoch": 1.0944017563117454,
+      "grad_norm": 4.464128285510355,
+      "learning_rate": 8.023944126931475e-07,
+      "loss": 0.3447,
+      "step": 499
+    },
+    {
+      "epoch": 1.0965971459934138,
+      "grad_norm": 6.600601913970912,
+      "learning_rate": 8.013772145661224e-07,
+      "loss": 0.3887,
+      "step": 500
+    },
+    {
+      "epoch": 1.0965971459934138,
+      "eval_accuracy": 0.818,
+      "eval_loss": 0.36472970247268677,
+      "eval_runtime": 62.4385,
+      "eval_samples_per_second": 8.008,
+      "eval_steps_per_second": 1.009,
+      "step": 500
+    },
+    {
+      "epoch": 1.0987925356750823,
+      "grad_norm": 5.402138515205016,
+      "learning_rate": 8.003580535587877e-07,
+      "loss": 0.3415,
+      "step": 501
+    },
+    {
+      "epoch": 1.1009879253567507,
+      "grad_norm": 7.628594037642485,
+      "learning_rate": 7.993369363089745e-07,
+      "loss": 0.3728,
+      "step": 502
+    },
+    {
+      "epoch": 1.1031833150384194,
+      "grad_norm": 4.697016196200385,
+      "learning_rate": 7.983138694672551e-07,
+      "loss": 0.3638,
+      "step": 503
+    },
+    {
+      "epoch": 1.1053787047200878,
+      "grad_norm": 6.2040879102322215,
+      "learning_rate": 7.972888596968996e-07,
+      "loss": 0.3458,
+      "step": 504
+    },
+    {
+      "epoch": 1.1075740944017562,
+      "grad_norm": 5.283464383567096,
+      "learning_rate": 7.962619136738324e-07,
+      "loss": 0.297,
+      "step": 505
+    },
+    {
+      "epoch": 1.109769484083425,
+      "grad_norm": 5.685282017291933,
+      "learning_rate": 7.952330380865887e-07,
+      "loss": 0.3328,
+      "step": 506
+    },
+    {
+      "epoch": 1.1119648737650933,
+      "grad_norm": 6.504450967378091,
+      "learning_rate": 7.942022396362711e-07,
+      "loss": 0.373,
+      "step": 507
+    },
+    {
+      "epoch": 1.1141602634467618,
+      "grad_norm": 6.224136967697534,
+      "learning_rate": 7.931695250365057e-07,
+      "loss": 0.3587,
+      "step": 508
+    },
+    {
+      "epoch": 1.1163556531284302,
+      "grad_norm": 6.524542381480522,
+      "learning_rate": 7.921349010133989e-07,
+      "loss": 0.3534,
+      "step": 509
+    },
+    {
+      "epoch": 1.1185510428100989,
+      "grad_norm": 6.508769827549269,
+      "learning_rate": 7.910983743054931e-07,
+      "loss": 0.3718,
+      "step": 510
+    },
+    {
+      "epoch": 1.1207464324917673,
+      "grad_norm": 7.148678198675181,
+      "learning_rate": 7.900599516637228e-07,
+      "loss": 0.3331,
+      "step": 511
+    },
+    {
+      "epoch": 1.1229418221734357,
+      "grad_norm": 5.983605133474822,
+      "learning_rate": 7.890196398513707e-07,
+      "loss": 0.353,
+      "step": 512
+    },
+    {
+      "epoch": 1.1251372118551042,
+      "grad_norm": 7.153049742139248,
+      "learning_rate": 7.879774456440242e-07,
+      "loss": 0.348,
+      "step": 513
+    },
+    {
+      "epoch": 1.1273326015367728,
+      "grad_norm": 7.392639449061027,
+      "learning_rate": 7.869333758295304e-07,
+      "loss": 0.3006,
+      "step": 514
+    },
+    {
+      "epoch": 1.1295279912184413,
+      "grad_norm": 6.288847784247055,
+      "learning_rate": 7.858874372079522e-07,
+      "loss": 0.3126,
+      "step": 515
+    },
+    {
+      "epoch": 1.1317233809001097,
+      "grad_norm": 6.530822602581628,
+      "learning_rate": 7.848396365915248e-07,
+      "loss": 0.3757,
+      "step": 516
+    },
+    {
+      "epoch": 1.1339187705817784,
+      "grad_norm": 7.180396451090792,
+      "learning_rate": 7.8378998080461e-07,
+      "loss": 0.3765,
+      "step": 517
+    },
+    {
+      "epoch": 1.1361141602634468,
+      "grad_norm": 7.893013021100936,
+      "learning_rate": 7.827384766836523e-07,
+      "loss": 0.3395,
+      "step": 518
+    },
+    {
+      "epoch": 1.1383095499451152,
+      "grad_norm": 6.419804870104756,
+      "learning_rate": 7.81685131077135e-07,
+      "loss": 0.3196,
+      "step": 519
+    },
+    {
+      "epoch": 1.1405049396267837,
+      "grad_norm": 7.441481132760177,
+      "learning_rate": 7.806299508455345e-07,
+      "loss": 0.4138,
+      "step": 520
+    },
+    {
+      "epoch": 1.1427003293084523,
+      "grad_norm": 5.097261411128716,
+      "learning_rate": 7.795729428612768e-07,
+      "loss": 0.322,
+      "step": 521
+    },
+    {
+      "epoch": 1.1448957189901208,
+      "grad_norm": 5.716684381877557,
+      "learning_rate": 7.785141140086913e-07,
+      "loss": 0.3636,
+      "step": 522
+    },
+    {
+      "epoch": 1.1470911086717892,
+      "grad_norm": 5.867715208154881,
+      "learning_rate": 7.774534711839677e-07,
+      "loss": 0.2981,
+      "step": 523
+    },
+    {
+      "epoch": 1.1492864983534576,
+      "grad_norm": 4.658345922311703,
+      "learning_rate": 7.763910212951095e-07,
+      "loss": 0.3478,
+      "step": 524
+    },
+    {
+      "epoch": 1.1514818880351263,
+      "grad_norm": 4.168344564692843,
+      "learning_rate": 7.753267712618898e-07,
+      "loss": 0.3468,
+      "step": 525
+    },
+    {
+      "epoch": 1.1536772777167947,
+      "grad_norm": 4.7538599995042645,
+      "learning_rate": 7.742607280158064e-07,
+      "loss": 0.3637,
+      "step": 526
+    },
+    {
+      "epoch": 1.1558726673984632,
+      "grad_norm": 6.571603396212342,
+      "learning_rate": 7.731928985000359e-07,
+      "loss": 0.375,
+      "step": 527
+    },
+    {
+      "epoch": 1.1580680570801318,
+      "grad_norm": 5.753219982389443,
+      "learning_rate": 7.721232896693894e-07,
+      "loss": 0.3427,
+      "step": 528
+    },
+    {
+      "epoch": 1.1602634467618003,
+      "grad_norm": 5.225412701161947,
+      "learning_rate": 7.710519084902663e-07,
+      "loss": 0.337,
+      "step": 529
+    },
+    {
+      "epoch": 1.1624588364434687,
+      "grad_norm": 5.484596435215429,
+      "learning_rate": 7.699787619406098e-07,
+      "loss": 0.3408,
+      "step": 530
+    },
+    {
+      "epoch": 1.1646542261251371,
+      "grad_norm": 5.282076505790076,
+      "learning_rate": 7.689038570098608e-07,
+      "loss": 0.3288,
+      "step": 531
+    },
+    {
+      "epoch": 1.1668496158068058,
+      "grad_norm": 4.775504582805927,
+      "learning_rate": 7.678272006989124e-07,
+      "loss": 0.2762,
+      "step": 532
+    },
+    {
+      "epoch": 1.1690450054884742,
+      "grad_norm": 5.78149113535585,
+      "learning_rate": 7.667488000200648e-07,
+      "loss": 0.4126,
+      "step": 533
+    },
+    {
+      "epoch": 1.1712403951701427,
+      "grad_norm": 5.897322988891718,
+      "learning_rate": 7.656686619969794e-07,
+      "loss": 0.3295,
+      "step": 534
+    },
+    {
+      "epoch": 1.173435784851811,
+      "grad_norm": 5.36707442483951,
+      "learning_rate": 7.645867936646327e-07,
+      "loss": 0.3358,
+      "step": 535
+    },
+    {
+      "epoch": 1.1756311745334798,
+      "grad_norm": 6.350922404290376,
+      "learning_rate": 7.635032020692706e-07,
+      "loss": 0.3665,
+      "step": 536
+    },
+    {
+      "epoch": 1.1778265642151482,
+      "grad_norm": 6.672225216206268,
+      "learning_rate": 7.624178942683634e-07,
+      "loss": 0.3237,
+      "step": 537
+    },
+    {
+      "epoch": 1.1800219538968166,
+      "grad_norm": 5.244066732628582,
+      "learning_rate": 7.613308773305584e-07,
+      "loss": 0.3546,
+      "step": 538
+    },
+    {
+      "epoch": 1.1822173435784853,
+      "grad_norm": 5.640267324413999,
+      "learning_rate": 7.60242158335635e-07,
+      "loss": 0.3763,
+      "step": 539
+    },
+    {
+      "epoch": 1.1844127332601537,
+      "grad_norm": 5.735242776747761,
+      "learning_rate": 7.591517443744577e-07,
+      "loss": 0.28,
+      "step": 540
+    },
+    {
+      "epoch": 1.1866081229418222,
+      "grad_norm": 5.370731176540545,
+      "learning_rate": 7.58059642548931e-07,
+      "loss": 0.3005,
+      "step": 541
+    },
+    {
+      "epoch": 1.1888035126234906,
+      "grad_norm": 5.792287386806627,
+      "learning_rate": 7.569658599719519e-07,
+      "loss": 0.3118,
+      "step": 542
+    },
+    {
+      "epoch": 1.1909989023051593,
+      "grad_norm": 6.798904382545207,
+      "learning_rate": 7.558704037673647e-07,
+      "loss": 0.3714,
+      "step": 543
+    },
+    {
+      "epoch": 1.1931942919868277,
+      "grad_norm": 5.46397725868398,
+      "learning_rate": 7.547732810699137e-07,
+      "loss": 0.3581,
+      "step": 544
+    },
+    {
+      "epoch": 1.1953896816684961,
+      "grad_norm": 5.999655519044568,
+      "learning_rate": 7.536744990251973e-07,
+      "loss": 0.3468,
+      "step": 545
+    },
+    {
+      "epoch": 1.1975850713501646,
+      "grad_norm": 5.809032437933095,
+      "learning_rate": 7.525740647896213e-07,
+      "loss": 0.3496,
+      "step": 546
+    },
+    {
+      "epoch": 1.1997804610318332,
+      "grad_norm": 5.6213191897838035,
+      "learning_rate": 7.514719855303524e-07,
+      "loss": 0.2637,
+      "step": 547
+    },
+    {
+      "epoch": 1.2019758507135017,
+      "grad_norm": 6.309518734364315,
+      "learning_rate": 7.503682684252711e-07,
+      "loss": 0.2972,
+      "step": 548
+    },
+    {
+      "epoch": 1.20417124039517,
+      "grad_norm": 5.10655426718776,
+      "learning_rate": 7.492629206629255e-07,
+      "loss": 0.395,
+      "step": 549
+    },
+    {
+      "epoch": 1.2063666300768388,
+      "grad_norm": 6.985855074156219,
+      "learning_rate": 7.481559494424839e-07,
+      "loss": 0.3901,
+      "step": 550
+    },
+    {
+      "epoch": 1.2063666300768388,
+      "eval_accuracy": 0.818,
+      "eval_loss": 0.3550068438053131,
+      "eval_runtime": 62.4228,
+      "eval_samples_per_second": 8.01,
+      "eval_steps_per_second": 1.009,
+      "step": 550
+    },
+    {
+      "epoch": 1.2085620197585072,
+      "grad_norm": 5.452884623196212,
+      "learning_rate": 7.470473619736889e-07,
+      "loss": 0.3635,
+      "step": 551
+    },
+    {
+      "epoch": 1.2107574094401756,
+      "grad_norm": 6.963272384448837,
+      "learning_rate": 7.459371654768088e-07,
+      "loss": 0.3905,
+      "step": 552
+    },
+    {
+      "epoch": 1.212952799121844,
+      "grad_norm": 4.543092820790928,
+      "learning_rate": 7.448253671825926e-07,
+      "loss": 0.3023,
+      "step": 553
+    },
+    {
+      "epoch": 1.2151481888035127,
+      "grad_norm": 5.793455953413061,
+      "learning_rate": 7.43711974332221e-07,
+      "loss": 0.3553,
+      "step": 554
+    },
+    {
+      "epoch": 1.2173435784851812,
+      "grad_norm": 5.5971357032313085,
+      "learning_rate": 7.425969941772606e-07,
+      "loss": 0.3217,
+      "step": 555
+    },
+    {
+      "epoch": 1.2195389681668496,
+      "grad_norm": 7.432739740334462,
+      "learning_rate": 7.414804339796161e-07,
+      "loss": 0.3662,
+      "step": 556
+    },
+    {
+      "epoch": 1.221734357848518,
+      "grad_norm": 5.748364577860177,
+      "learning_rate": 7.403623010114831e-07,
+      "loss": 0.3456,
+      "step": 557
+    },
+    {
+      "epoch": 1.2239297475301867,
+      "grad_norm": 5.661380628505925,
+      "learning_rate": 7.392426025553002e-07,
+      "loss": 0.3484,
+      "step": 558
+    },
+    {
+      "epoch": 1.2261251372118551,
+      "grad_norm": 7.677386392401964,
+      "learning_rate": 7.381213459037031e-07,
+      "loss": 0.3324,
+      "step": 559
+    },
+    {
+      "epoch": 1.2283205268935236,
+      "grad_norm": 5.815978817257181,
+      "learning_rate": 7.36998538359475e-07,
+      "loss": 0.3306,
+      "step": 560
+    },
+    {
+      "epoch": 1.2305159165751922,
+      "grad_norm": 7.157642715894916,
+      "learning_rate": 7.358741872355008e-07,
+      "loss": 0.356,
+      "step": 561
+    },
+    {
+      "epoch": 1.2327113062568607,
+      "grad_norm": 9.217680532850702,
+      "learning_rate": 7.347482998547184e-07,
+      "loss": 0.3601,
+      "step": 562
+    },
+    {
+      "epoch": 1.234906695938529,
+      "grad_norm": 10.045847811800988,
+      "learning_rate": 7.336208835500719e-07,
+      "loss": 0.3745,
+      "step": 563
+    },
+    {
+      "epoch": 1.2371020856201975,
+      "grad_norm": 6.241693811898233,
+      "learning_rate": 7.324919456644627e-07,
+      "loss": 0.3575,
+      "step": 564
+    },
+    {
+      "epoch": 1.239297475301866,
+      "grad_norm": 6.430610540834492,
+      "learning_rate": 7.313614935507028e-07,
+      "loss": 0.3293,
+      "step": 565
+    },
+    {
+      "epoch": 1.2414928649835346,
+      "grad_norm": 5.841827763430915,
+      "learning_rate": 7.302295345714662e-07,
+      "loss": 0.3194,
+      "step": 566
+    },
+    {
+      "epoch": 1.243688254665203,
+      "grad_norm": 5.840825760143234,
+      "learning_rate": 7.290960760992413e-07,
+      "loss": 0.3401,
+      "step": 567
+    },
+    {
+      "epoch": 1.2458836443468715,
+      "grad_norm": 10.441987786035702,
+      "learning_rate": 7.279611255162824e-07,
+      "loss": 0.4262,
+      "step": 568
+    },
+    {
+      "epoch": 1.2480790340285401,
+      "grad_norm": 11.92569950751005,
+      "learning_rate": 7.268246902145625e-07,
+      "loss": 0.3483,
+      "step": 569
+    },
+    {
+      "epoch": 1.2502744237102086,
+      "grad_norm": 6.453124049655114,
+      "learning_rate": 7.256867775957243e-07,
+      "loss": 0.384,
+      "step": 570
+    },
+    {
+      "epoch": 1.252469813391877,
+      "grad_norm": 5.70168194447466,
+      "learning_rate": 7.245473950710322e-07,
+      "loss": 0.3201,
+      "step": 571
+    },
+    {
+      "epoch": 1.2546652030735457,
+      "grad_norm": 6.001337956114435,
+      "learning_rate": 7.234065500613244e-07,
+      "loss": 0.3473,
+      "step": 572
+    },
+    {
+      "epoch": 1.2568605927552141,
+      "grad_norm": 5.271757451994717,
+      "learning_rate": 7.222642499969644e-07,
+      "loss": 0.3644,
+      "step": 573
+    },
+    {
+      "epoch": 1.2590559824368825,
+      "grad_norm": 5.806447016385563,
+      "learning_rate": 7.211205023177923e-07,
+      "loss": 0.3146,
+      "step": 574
+    },
+    {
+      "epoch": 1.261251372118551,
+      "grad_norm": 5.047476182935009,
+      "learning_rate": 7.199753144730765e-07,
+      "loss": 0.3406,
+      "step": 575
+    },
+    {
+      "epoch": 1.2634467618002194,
+      "grad_norm": 5.853247665884473,
+      "learning_rate": 7.188286939214656e-07,
+      "loss": 0.3006,
+      "step": 576
+    },
+    {
+      "epoch": 1.265642151481888,
+      "grad_norm": 6.805598773740254,
+      "learning_rate": 7.176806481309387e-07,
+      "loss": 0.3891,
+      "step": 577
+    },
+    {
+      "epoch": 1.2678375411635565,
+      "grad_norm": 6.946161315608133,
+      "learning_rate": 7.165311845787587e-07,
+      "loss": 0.3392,
+      "step": 578
+    },
+    {
+      "epoch": 1.270032930845225,
+      "grad_norm": 6.46125490152312,
+      "learning_rate": 7.153803107514217e-07,
+      "loss": 0.3478,
+      "step": 579
+    },
+    {
+      "epoch": 1.2722283205268936,
+      "grad_norm": 5.096980092514172,
+      "learning_rate": 7.142280341446089e-07,
+      "loss": 0.3449,
+      "step": 580
+    },
+    {
+      "epoch": 1.274423710208562,
+      "grad_norm": 7.089646471526524,
+      "learning_rate": 7.130743622631378e-07,
+      "loss": 0.3008,
+      "step": 581
+    },
+    {
+      "epoch": 1.2766190998902305,
+      "grad_norm": 5.7888088906336055,
+      "learning_rate": 7.119193026209142e-07,
+      "loss": 0.3902,
+      "step": 582
+    },
+    {
+      "epoch": 1.2788144895718991,
+      "grad_norm": 5.824834458347463,
+      "learning_rate": 7.107628627408813e-07,
+      "loss": 0.3129,
+      "step": 583
+    },
+    {
+      "epoch": 1.2810098792535676,
+      "grad_norm": 8.738486775778084,
+      "learning_rate": 7.096050501549727e-07,
+      "loss": 0.4085,
+      "step": 584
+    },
+    {
+      "epoch": 1.283205268935236,
+      "grad_norm": 6.048998032376583,
+      "learning_rate": 7.084458724040621e-07,
+      "loss": 0.3345,
+      "step": 585
+    },
+    {
+      "epoch": 1.2854006586169044,
+      "grad_norm": 9.605936387748194,
+      "learning_rate": 7.072853370379144e-07,
+      "loss": 0.3393,
+      "step": 586
+    },
+    {
+      "epoch": 1.2875960482985729,
+      "grad_norm": 6.865950862731977,
+      "learning_rate": 7.061234516151371e-07,
+      "loss": 0.3301,
+      "step": 587
+    },
+    {
+      "epoch": 1.2897914379802415,
+      "grad_norm": 6.267095187214198,
+      "learning_rate": 7.049602237031305e-07,
+      "loss": 0.2994,
+      "step": 588
+    },
+    {
+      "epoch": 1.29198682766191,
+      "grad_norm": 5.6232557271177,
+      "learning_rate": 7.037956608780385e-07,
+      "loss": 0.3094,
+      "step": 589
+    },
+    {
+      "epoch": 1.2941822173435784,
+      "grad_norm": 5.801703901127963,
+      "learning_rate": 7.026297707246993e-07,
+      "loss": 0.3549,
+      "step": 590
+    },
+    {
+      "epoch": 1.296377607025247,
+      "grad_norm": 6.414391032155044,
+      "learning_rate": 7.014625608365962e-07,
+      "loss": 0.3183,
+      "step": 591
+    },
+    {
+      "epoch": 1.2985729967069155,
+      "grad_norm": 6.651766488870556,
+      "learning_rate": 7.002940388158083e-07,
+      "loss": 0.3676,
+      "step": 592
+    },
+    {
+      "epoch": 1.300768386388584,
+      "grad_norm": 7.417392797535265,
+      "learning_rate": 6.991242122729596e-07,
+      "loss": 0.4093,
+      "step": 593
+    },
+    {
+      "epoch": 1.3029637760702526,
+      "grad_norm": 6.527075497004787,
+      "learning_rate": 6.979530888271716e-07,
+      "loss": 0.3604,
+      "step": 594
+    },
+    {
+      "epoch": 1.305159165751921,
+      "grad_norm": 6.940696529099327,
+      "learning_rate": 6.96780676106012e-07,
+      "loss": 0.2738,
+      "step": 595
+    },
+    {
+      "epoch": 1.3073545554335895,
+      "grad_norm": 6.972366954094688,
+      "learning_rate": 6.956069817454458e-07,
+      "loss": 0.3495,
+      "step": 596
+    },
+    {
+      "epoch": 1.309549945115258,
+      "grad_norm": 6.427760417094834,
+      "learning_rate": 6.944320133897855e-07,
+      "loss": 0.3368,
+      "step": 597
+    },
+    {
+      "epoch": 1.3117453347969263,
+      "grad_norm": 5.47206109550229,
+      "learning_rate": 6.932557786916406e-07,
+      "loss": 0.3293,
+      "step": 598
+    },
+    {
+      "epoch": 1.313940724478595,
+      "grad_norm": 4.918239274946065,
+      "learning_rate": 6.920782853118689e-07,
+      "loss": 0.3746,
+      "step": 599
+    },
+    {
+      "epoch": 1.3161361141602634,
+      "grad_norm": 6.295802272512743,
+      "learning_rate": 6.908995409195258e-07,
+      "loss": 0.3043,
+      "step": 600
+    },
+    {
+      "epoch": 1.3161361141602634,
+      "eval_accuracy": 0.826,
+      "eval_loss": 0.35173022747039795,
+      "eval_runtime": 62.3581,
+      "eval_samples_per_second": 8.018,
+      "eval_steps_per_second": 1.01,
+      "step": 600
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1368,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}