diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8875 @@
+{
+  "best_metric": 2.457956314086914,
+  "best_model_checkpoint": "bert_base_lda_5_v1_book/checkpoint-580000",
+  "epoch": 25.0,
+  "eval_steps": 10000,
+  "global_step": 593075,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.021076592336551025,
+      "grad_norm": 2.3506898880004883,
+      "learning_rate": 5e-06,
+      "loss": 10.522,
+      "step": 500
+    },
+    {
+      "epoch": 0.04215318467310205,
+      "grad_norm": 1.4383302927017212,
+      "learning_rate": 1e-05,
+      "loss": 8.2998,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06322977700965308,
+      "grad_norm": 1.3909755945205688,
+      "learning_rate": 1.5e-05,
+      "loss": 7.6526,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0843063693462041,
+      "grad_norm": 1.3181076049804688,
+      "learning_rate": 2e-05,
+      "loss": 7.5296,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10538296168275513,
+      "grad_norm": 1.246569275856018,
+      "learning_rate": 2.5e-05,
+      "loss": 7.4694,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12645955401930617,
+      "grad_norm": 1.3694698810577393,
+      "learning_rate": 3e-05,
+      "loss": 7.4298,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14753614635585718,
+      "grad_norm": 1.0647348165512085,
+      "learning_rate": 3.5e-05,
+      "loss": 7.3996,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1686127386924082,
+      "grad_norm": 1.0710080862045288,
+      "learning_rate": 4e-05,
+      "loss": 7.3723,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18968933102895924,
+      "grad_norm": 0.969950258731842,
+      "learning_rate": 4.5e-05,
+      "loss": 7.3473,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21076592336551025,
+      "grad_norm": 1.0767165422439575,
+      "learning_rate": 5e-05,
+      "loss": 7.3313,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2318425157020613,
+      "grad_norm": 1.4764683246612549,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 7.3203,
+      "step": 5500
+    },
+    {
+      "epoch": 0.25291910803861234,
+      "grad_norm": 1.0935750007629395,
+      "learning_rate": 6e-05,
+      "loss": 7.3018,
+      "step": 6000
+    },
+    {
+      "epoch": 0.27399570037516335,
+      "grad_norm": 1.0566405057907104,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 7.2962,
+      "step": 6500
+    },
+    {
+      "epoch": 0.29507229271171437,
+      "grad_norm": 0.8542040586471558,
+      "learning_rate": 7e-05,
+      "loss": 7.2823,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3161488850482654,
+      "grad_norm": 1.0595548152923584,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 7.2746,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3372254773848164,
+      "grad_norm": 1.203078031539917,
+      "learning_rate": 8e-05,
+      "loss": 7.2693,
+      "step": 8000
+    },
+    {
+      "epoch": 0.35830206972136747,
+      "grad_norm": 1.20786452293396,
+      "learning_rate": 8.5e-05,
+      "loss": 7.218,
+      "step": 8500
+    },
+    {
+      "epoch": 0.3793786620579185,
+      "grad_norm": 1.2807844877243042,
+      "learning_rate": 9e-05,
+      "loss": 7.1609,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4004552543944695,
+      "grad_norm": 1.6128579378128052,
+      "learning_rate": 9.5e-05,
+      "loss": 7.1135,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4215318467310205,
+      "grad_norm": 1.4275572299957275,
+      "learning_rate": 0.0001,
+      "loss": 7.0997,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4215318467310205,
+      "eval_accuracy": 0.16385036697608915,
+      "eval_loss": 6.9318108558654785,
+      "eval_runtime": 342.9835,
+      "eval_samples_per_second": 351.929,
+      "eval_steps_per_second": 3.668,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4426084390675716,
+      "grad_norm": 1.165830373764038,
+      "learning_rate": 9.99142477382841e-05,
+      "loss": 7.0864,
+      "step": 10500
+    },
+    {
+      "epoch": 0.4636850314041226,
+      "grad_norm": 0.9837601184844971,
+      "learning_rate": 9.98284954765682e-05,
+      "loss": 7.0772,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4847616237406736,
+      "grad_norm": 1.1474593877792358,
+      "learning_rate": 9.97427432148523e-05,
+      "loss": 7.0766,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5058382160772247,
+      "grad_norm": 1.0772361755371094,
+      "learning_rate": 9.96569909531364e-05,
+      "loss": 7.0632,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5269148084137757,
+      "grad_norm": 1.1484520435333252,
+      "learning_rate": 9.957123869142048e-05,
+      "loss": 7.0567,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5479914007503267,
+      "grad_norm": 1.1547315120697021,
+      "learning_rate": 9.94854864297046e-05,
+      "loss": 7.0538,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5690679930868777,
+      "grad_norm": 1.0673846006393433,
+      "learning_rate": 9.939973416798869e-05,
+      "loss": 7.0431,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5901445854234287,
+      "grad_norm": 0.8327569365501404,
+      "learning_rate": 9.931398190627278e-05,
+      "loss": 7.0384,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6112211777599798,
+      "grad_norm": 0.9184210896492004,
+      "learning_rate": 9.922822964455688e-05,
+      "loss": 7.0311,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6322977700965308,
+      "grad_norm": 0.895837664604187,
+      "learning_rate": 9.914247738284099e-05,
+      "loss": 7.0288,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6533743624330818,
+      "grad_norm": 1.3193985223770142,
+      "learning_rate": 9.905672512112507e-05,
+      "loss": 7.0205,
+      "step": 15500
+    },
+    {
+      "epoch": 0.6744509547696328,
+      "grad_norm": 0.8034365177154541,
+      "learning_rate": 9.897097285940917e-05,
+      "loss": 7.0169,
+      "step": 16000
+    },
+    {
+      "epoch": 0.6955275471061839,
+      "grad_norm": 0.9143784642219543,
+      "learning_rate": 9.888522059769327e-05,
+      "loss": 7.0106,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7166041394427349,
+      "grad_norm": 0.6019152998924255,
+      "learning_rate": 9.879946833597737e-05,
+      "loss": 7.007,
+      "step": 17000
+    },
+    {
+      "epoch": 0.737680731779286,
+      "grad_norm": 0.6809580326080322,
+      "learning_rate": 9.871371607426147e-05,
+      "loss": 6.9992,
+      "step": 17500
+    },
+    {
+      "epoch": 0.758757324115837,
+      "grad_norm": 1.0727837085723877,
+      "learning_rate": 9.862796381254556e-05,
+      "loss": 7.0037,
+      "step": 18000
+    },
+    {
+      "epoch": 0.779833916452388,
+      "grad_norm": 0.908868134021759,
+      "learning_rate": 9.854221155082965e-05,
+      "loss": 6.9948,
+      "step": 18500
+    },
+    {
+      "epoch": 0.800910508788939,
+      "grad_norm": 1.0350536108016968,
+      "learning_rate": 9.845645928911376e-05,
+      "loss": 6.9953,
+      "step": 19000
+    },
+    {
+      "epoch": 0.82198710112549,
+      "grad_norm": 0.7575790286064148,
+      "learning_rate": 9.837070702739786e-05,
+      "loss": 6.9917,
+      "step": 19500
+    },
+    {
+      "epoch": 0.843063693462041,
+      "grad_norm": 1.2296619415283203,
+      "learning_rate": 9.828495476568194e-05,
+      "loss": 6.9855,
+      "step": 20000
+    },
+    {
+      "epoch": 0.843063693462041,
+      "eval_accuracy": 0.16518798283891403,
+      "eval_loss": 6.841960906982422,
+      "eval_runtime": 337.3524,
+      "eval_samples_per_second": 357.804,
+      "eval_steps_per_second": 3.729,
+      "step": 20000
+    },
+    {
+      "epoch": 0.864140285798592,
+      "grad_norm": 0.9227085113525391,
+      "learning_rate": 9.819920250396604e-05,
+      "loss": 6.981,
+      "step": 20500
+    },
+    {
+      "epoch": 0.8852168781351432,
+      "grad_norm": 0.8296341300010681,
+      "learning_rate": 9.811345024225015e-05,
+      "loss": 6.9762,
+      "step": 21000
+    },
+    {
+      "epoch": 0.9062934704716942,
+      "grad_norm": 0.8386650681495667,
+      "learning_rate": 9.802769798053424e-05,
+      "loss": 6.9744,
+      "step": 21500
+    },
+    {
+      "epoch": 0.9273700628082452,
+      "grad_norm": 0.6551395058631897,
+      "learning_rate": 9.794194571881834e-05,
+      "loss": 6.9687,
+      "step": 22000
+    },
+    {
+      "epoch": 0.9484466551447962,
+      "grad_norm": 0.8056783676147461,
+      "learning_rate": 9.785619345710244e-05,
+      "loss": 6.9653,
+      "step": 22500
+    },
+    {
+      "epoch": 0.9695232474813472,
+      "grad_norm": 0.7251622080802917,
+      "learning_rate": 9.777044119538653e-05,
+      "loss": 6.9622,
+      "step": 23000
+    },
+    {
+      "epoch": 0.9905998398178982,
+      "grad_norm": 1.1007863283157349,
+      "learning_rate": 9.768468893367063e-05,
+      "loss": 6.9576,
+      "step": 23500
+    },
+    {
+      "epoch": 1.0116764321544494,
+      "grad_norm": 0.8604030013084412,
+      "learning_rate": 9.759893667195473e-05,
+      "loss": 6.9563,
+      "step": 24000
+    },
+    {
+      "epoch": 1.0327530244910004,
+      "grad_norm": 0.7901830673217773,
+      "learning_rate": 9.751318441023883e-05,
+      "loss": 6.9548,
+      "step": 24500
+    },
+    {
+      "epoch": 1.0538296168275514,
+      "grad_norm": 0.9188992977142334,
+      "learning_rate": 9.742743214852293e-05,
+      "loss": 6.9511,
+      "step": 25000
+    },
+    {
+      "epoch": 1.0749062091641024,
+      "grad_norm": 0.9621573090553284,
+      "learning_rate": 9.734167988680703e-05,
+      "loss": 6.9436,
+      "step": 25500
+    },
+    {
+      "epoch": 1.0959828015006534,
+      "grad_norm": 1.0147000551223755,
+      "learning_rate": 9.725592762509111e-05,
+      "loss": 6.9439,
+      "step": 26000
+    },
+    {
+      "epoch": 1.1170593938372044,
+      "grad_norm": 0.6584479808807373,
+      "learning_rate": 9.717017536337522e-05,
+      "loss": 6.9395,
+      "step": 26500
+    },
+    {
+      "epoch": 1.1381359861737554,
+      "grad_norm": 0.9518819451332092,
+      "learning_rate": 9.708442310165932e-05,
+      "loss": 6.9441,
+      "step": 27000
+    },
+    {
+      "epoch": 1.1592125785103065,
+      "grad_norm": 1.0028880834579468,
+      "learning_rate": 9.69986708399434e-05,
+      "loss": 6.938,
+      "step": 27500
+    },
+    {
+      "epoch": 1.1802891708468575,
+      "grad_norm": 0.7855657935142517,
+      "learning_rate": 9.69129185782275e-05,
+      "loss": 6.941,
+      "step": 28000
+    },
+    {
+      "epoch": 1.2013657631834085,
+      "grad_norm": 0.8446207046508789,
+      "learning_rate": 9.68271663165116e-05,
+      "loss": 6.9306,
+      "step": 28500
+    },
+    {
+      "epoch": 1.2224423555199595,
+      "grad_norm": 0.8305496573448181,
+      "learning_rate": 9.67414140547957e-05,
+      "loss": 6.9281,
+      "step": 29000
+    },
+    {
+      "epoch": 1.2435189478565105,
+      "grad_norm": 0.8320557475090027,
+      "learning_rate": 9.66556617930798e-05,
+      "loss": 6.9248,
+      "step": 29500
+    },
+    {
+      "epoch": 1.2645955401930615,
+      "grad_norm": 1.019494652748108,
+      "learning_rate": 9.656990953136388e-05,
+      "loss": 6.9276,
+      "step": 30000
+    },
+    {
+      "epoch": 1.2645955401930615,
+      "eval_accuracy": 0.16621551482235247,
+      "eval_loss": 6.796502113342285,
+      "eval_runtime": 357.0147,
+      "eval_samples_per_second": 338.098,
+      "eval_steps_per_second": 3.524,
+      "step": 30000
+    },
+    {
+      "epoch": 1.2856721325296125,
+      "grad_norm": 0.7811832427978516,
+      "learning_rate": 9.6484157269648e-05,
+      "loss": 6.9193,
+      "step": 30500
+    },
+    {
+      "epoch": 1.3067487248661636,
+      "grad_norm": 0.6539536118507385,
+      "learning_rate": 9.639840500793209e-05,
+      "loss": 6.9238,
+      "step": 31000
+    },
+    {
+      "epoch": 1.3278253172027146,
+      "grad_norm": 0.7224295735359192,
+      "learning_rate": 9.631265274621618e-05,
+      "loss": 6.9198,
+      "step": 31500
+    },
+    {
+      "epoch": 1.3489019095392658,
+      "grad_norm": 0.837656557559967,
+      "learning_rate": 9.622690048450028e-05,
+      "loss": 6.9227,
+      "step": 32000
+    },
+    {
+      "epoch": 1.3699785018758166,
+      "grad_norm": 1.0153343677520752,
+      "learning_rate": 9.614114822278439e-05,
+      "loss": 6.9161,
+      "step": 32500
+    },
+    {
+      "epoch": 1.3910550942123678,
+      "grad_norm": 0.6399874687194824,
+      "learning_rate": 9.605539596106847e-05,
+      "loss": 6.9188,
+      "step": 33000
+    },
+    {
+      "epoch": 1.4121316865489189,
+      "grad_norm": 0.8211466073989868,
+      "learning_rate": 9.596964369935257e-05,
+      "loss": 6.9143,
+      "step": 33500
+    },
+    {
+      "epoch": 1.4332082788854699,
+      "grad_norm": 0.7502171397209167,
+      "learning_rate": 9.588389143763667e-05,
+      "loss": 6.9109,
+      "step": 34000
+    },
+    {
+      "epoch": 1.4542848712220209,
+      "grad_norm": 1.0810315608978271,
+      "learning_rate": 9.579813917592077e-05,
+      "loss": 6.9125,
+      "step": 34500
+    },
+    {
+      "epoch": 1.475361463558572,
+      "grad_norm": 0.7616167068481445,
+      "learning_rate": 9.571238691420487e-05,
+      "loss": 6.9104,
+      "step": 35000
+    },
+    {
+      "epoch": 1.496438055895123,
+      "grad_norm": 0.7945672273635864,
+      "learning_rate": 9.562663465248896e-05,
+      "loss": 6.9053,
+      "step": 35500
+    },
+    {
+      "epoch": 1.517514648231674,
+      "grad_norm": 0.8658792972564697,
+      "learning_rate": 9.554088239077306e-05,
+      "loss": 6.9009,
+      "step": 36000
+    },
+    {
+      "epoch": 1.538591240568225,
+      "grad_norm": 0.8676465153694153,
+      "learning_rate": 9.545513012905716e-05,
+      "loss": 6.9044,
+      "step": 36500
+    },
+    {
+      "epoch": 1.559667832904776,
+      "grad_norm": 0.7879326939582825,
+      "learning_rate": 9.536937786734126e-05,
+      "loss": 6.9064,
+      "step": 37000
+    },
+    {
+      "epoch": 1.580744425241327,
+      "grad_norm": 0.6512182950973511,
+      "learning_rate": 9.528362560562534e-05,
+      "loss": 6.9021,
+      "step": 37500
+    },
+    {
+      "epoch": 1.601821017577878,
+      "grad_norm": 0.8710217475891113,
+      "learning_rate": 9.519787334390946e-05,
+      "loss": 6.8999,
+      "step": 38000
+    },
+    {
+      "epoch": 1.622897609914429,
+      "grad_norm": 0.6441351175308228,
+      "learning_rate": 9.511212108219355e-05,
+      "loss": 6.8972,
+      "step": 38500
+    },
+    {
+      "epoch": 1.64397420225098,
+      "grad_norm": 0.7855520844459534,
+      "learning_rate": 9.502636882047764e-05,
+      "loss": 6.8989,
+      "step": 39000
+    },
+    {
+      "epoch": 1.6650507945875312,
+      "grad_norm": 1.0443763732910156,
+      "learning_rate": 9.494061655876174e-05,
+      "loss": 6.8978,
+      "step": 39500
+    },
+    {
+      "epoch": 1.686127386924082,
+      "grad_norm": 0.8631075024604797,
+      "learning_rate": 9.485486429704585e-05,
+      "loss": 6.8979,
+      "step": 40000
+    },
+    {
+      "epoch": 1.686127386924082,
+      "eval_accuracy": 0.16639960576724372,
+      "eval_loss": 6.770650386810303,
+      "eval_runtime": 372.7441,
+      "eval_samples_per_second": 323.831,
+      "eval_steps_per_second": 3.375,
+      "step": 40000
+    },
+    {
+      "epoch": 1.7072039792606333,
+      "grad_norm": 1.1212265491485596,
+      "learning_rate": 9.476911203532993e-05,
+      "loss": 6.8898,
+      "step": 40500
+    },
+    {
+      "epoch": 1.728280571597184,
+      "grad_norm": 0.8171544075012207,
+      "learning_rate": 9.468335977361403e-05,
+      "loss": 6.8916,
+      "step": 41000
+    },
+    {
+      "epoch": 1.7493571639337353,
+      "grad_norm": 0.6501728296279907,
+      "learning_rate": 9.459760751189813e-05,
+      "loss": 6.8936,
+      "step": 41500
+    },
+    {
+      "epoch": 1.770433756270286,
+      "grad_norm": 0.8541089296340942,
+      "learning_rate": 9.451185525018223e-05,
+      "loss": 6.8929,
+      "step": 42000
+    },
+    {
+      "epoch": 1.7915103486068373,
+      "grad_norm": 1.0465905666351318,
+      "learning_rate": 9.442610298846633e-05,
+      "loss": 6.8893,
+      "step": 42500
+    },
+    {
+      "epoch": 1.8125869409433881,
+      "grad_norm": 0.982352614402771,
+      "learning_rate": 9.434035072675043e-05,
+      "loss": 6.8868,
+      "step": 43000
+    },
+    {
+      "epoch": 1.8336635332799394,
+      "grad_norm": 0.6767280697822571,
+      "learning_rate": 9.425459846503451e-05,
+      "loss": 6.8854,
+      "step": 43500
+    },
+    {
+      "epoch": 1.8547401256164904,
+      "grad_norm": 0.5541186928749084,
+      "learning_rate": 9.416884620331862e-05,
+      "loss": 6.8844,
+      "step": 44000
+    },
+    {
+      "epoch": 1.8758167179530414,
+      "grad_norm": 0.8303530812263489,
+      "learning_rate": 9.408309394160272e-05,
+      "loss": 6.8872,
+      "step": 44500
+    },
+    {
+      "epoch": 1.8968933102895924,
+      "grad_norm": 0.9653485417366028,
+      "learning_rate": 9.39973416798868e-05,
+      "loss": 6.8813,
+      "step": 45000
+    },
+    {
+      "epoch": 1.9179699026261434,
+      "grad_norm": 0.5843878984451294,
+      "learning_rate": 9.39115894181709e-05,
+      "loss": 6.8796,
+      "step": 45500
+    },
+    {
+      "epoch": 1.9390464949626944,
+      "grad_norm": 0.667246401309967,
+      "learning_rate": 9.382583715645502e-05,
+      "loss": 6.8796,
+      "step": 46000
+    },
+    {
+      "epoch": 1.9601230872992454,
+      "grad_norm": 0.9902203679084778,
+      "learning_rate": 9.37400848947391e-05,
+      "loss": 6.8759,
+      "step": 46500
+    },
+    {
+      "epoch": 1.9811996796357965,
+      "grad_norm": 0.7747452855110168,
+      "learning_rate": 9.36543326330232e-05,
+      "loss": 6.8856,
+      "step": 47000
+    },
+    {
+      "epoch": 2.0022762719723475,
+      "grad_norm": 1.011152744293213,
+      "learning_rate": 9.35685803713073e-05,
+      "loss": 6.8834,
+      "step": 47500
+    },
+    {
+      "epoch": 2.0233528643088987,
+      "grad_norm": 0.6567111611366272,
+      "learning_rate": 9.34828281095914e-05,
+      "loss": 6.8748,
+      "step": 48000
+    },
+    {
+      "epoch": 2.0444294566454495,
+      "grad_norm": 0.7529711127281189,
+      "learning_rate": 9.339707584787549e-05,
+      "loss": 6.8739,
+      "step": 48500
+    },
+    {
+      "epoch": 2.0655060489820007,
+      "grad_norm": 0.9052157402038574,
+      "learning_rate": 9.331132358615959e-05,
+      "loss": 6.8717,
+      "step": 49000
+    },
+    {
+      "epoch": 2.0865826413185515,
+      "grad_norm": 0.5829957127571106,
+      "learning_rate": 9.322557132444369e-05,
+      "loss": 6.8697,
+      "step": 49500
+    },
+    {
+      "epoch": 2.1076592336551028,
+      "grad_norm": 0.7997853755950928,
+      "learning_rate": 9.313981906272779e-05,
+      "loss": 6.8708,
+      "step": 50000
+    },
+    {
+      "epoch": 2.1076592336551028,
+      "eval_accuracy": 0.16733511131416295,
+      "eval_loss": 6.750965595245361,
+      "eval_runtime": 490.5955,
+      "eval_samples_per_second": 246.04,
+      "eval_steps_per_second": 2.564,
+      "step": 50000
+    },
+    {
+      "epoch": 2.1287358259916536,
+      "grad_norm": 0.8557528853416443,
+      "learning_rate": 9.305406680101189e-05,
+      "loss": 6.8741,
+      "step": 50500
+    },
+    {
+      "epoch": 2.149812418328205,
+      "grad_norm": 0.810947060585022,
+      "learning_rate": 9.296831453929597e-05,
+      "loss": 6.8725,
+      "step": 51000
+    },
+    {
+      "epoch": 2.1708890106647556,
+      "grad_norm": 0.7566709518432617,
+      "learning_rate": 9.288256227758008e-05,
+      "loss": 6.8662,
+      "step": 51500
+    },
+    {
+      "epoch": 2.191965603001307,
+      "grad_norm": 0.9078630805015564,
+      "learning_rate": 9.279681001586418e-05,
+      "loss": 6.8664,
+      "step": 52000
+    },
+    {
+      "epoch": 2.2130421953378576,
+      "grad_norm": 0.7223352789878845,
+      "learning_rate": 9.271105775414827e-05,
+      "loss": 6.8642,
+      "step": 52500
+    },
+    {
+      "epoch": 2.234118787674409,
+      "grad_norm": 0.6951504349708557,
+      "learning_rate": 9.262530549243236e-05,
+      "loss": 6.8654,
+      "step": 53000
+    },
+    {
+      "epoch": 2.2551953800109596,
+      "grad_norm": 0.5903010368347168,
+      "learning_rate": 9.253955323071646e-05,
+      "loss": 6.8668,
+      "step": 53500
+    },
+    {
+      "epoch": 2.276271972347511,
+      "grad_norm": 0.6211978793144226,
+      "learning_rate": 9.245380096900056e-05,
+      "loss": 6.8709,
+      "step": 54000
+    },
+    {
+      "epoch": 2.2973485646840617,
+      "grad_norm": 0.8644723892211914,
+      "learning_rate": 9.236804870728466e-05,
+      "loss": 6.8615,
+      "step": 54500
+    },
+    {
+      "epoch": 2.318425157020613,
+      "grad_norm": 0.7224984765052795,
+      "learning_rate": 9.228229644556876e-05,
+      "loss": 6.8632,
+      "step": 55000
+    },
+    {
+      "epoch": 2.339501749357164,
+      "grad_norm": 0.6594499945640564,
+      "learning_rate": 9.219654418385286e-05,
+      "loss": 6.8591,
+      "step": 55500
+    },
+    {
+      "epoch": 2.360578341693715,
+      "grad_norm": 0.7944409251213074,
+      "learning_rate": 9.211079192213695e-05,
+      "loss": 6.8646,
+      "step": 56000
+    },
+    {
+      "epoch": 2.381654934030266,
+      "grad_norm": 1.1470361948013306,
+      "learning_rate": 9.202503966042105e-05,
+      "loss": 6.86,
+      "step": 56500
+    },
+    {
+      "epoch": 2.402731526366817,
+      "grad_norm": 0.7851831316947937,
+      "learning_rate": 9.193928739870514e-05,
+      "loss": 6.8619,
+      "step": 57000
+    },
+    {
+      "epoch": 2.423808118703368,
+      "grad_norm": 0.5950602293014526,
+      "learning_rate": 9.185353513698925e-05,
+      "loss": 6.857,
+      "step": 57500
+    },
+    {
+      "epoch": 2.444884711039919,
+      "grad_norm": 0.6637733578681946,
+      "learning_rate": 9.176778287527335e-05,
+      "loss": 6.8568,
+      "step": 58000
+    },
+    {
+      "epoch": 2.4659613033764702,
+      "grad_norm": 0.6513834595680237,
+      "learning_rate": 9.168203061355743e-05,
+      "loss": 6.8577,
+      "step": 58500
+    },
+    {
+      "epoch": 2.487037895713021,
+      "grad_norm": 0.7247253656387329,
+      "learning_rate": 9.159627835184153e-05,
+      "loss": 6.8578,
+      "step": 59000
+    },
+    {
+      "epoch": 2.5081144880495723,
+      "grad_norm": 0.5381003022193909,
+      "learning_rate": 9.151052609012564e-05,
+      "loss": 6.8565,
+      "step": 59500
+    },
+    {
+      "epoch": 2.529191080386123,
+      "grad_norm": 0.7330372333526611,
+      "learning_rate": 9.142477382840973e-05,
+      "loss": 6.8545,
+      "step": 60000
+    },
+    {
+      "epoch": 2.529191080386123,
+      "eval_accuracy": 0.16820468541804126,
+      "eval_loss": 6.7322492599487305,
+      "eval_runtime": 346.5404,
+      "eval_samples_per_second": 348.317,
+      "eval_steps_per_second": 3.63,
+      "step": 60000
+    },
+    {
+      "epoch": 2.5502676727226743,
+      "grad_norm": 0.572146475315094,
+      "learning_rate": 9.133902156669383e-05,
+      "loss": 6.8491,
+      "step": 60500
+    },
+    {
+      "epoch": 2.571344265059225,
+      "grad_norm": 0.7418913245201111,
+      "learning_rate": 9.125326930497792e-05,
+      "loss": 6.8517,
+      "step": 61000
+    },
+    {
+      "epoch": 2.5924208573957763,
+      "grad_norm": 0.639075517654419,
+      "learning_rate": 9.116751704326202e-05,
+      "loss": 6.8469,
+      "step": 61500
+    },
+    {
+      "epoch": 2.613497449732327,
+      "grad_norm": 0.8623602986335754,
+      "learning_rate": 9.108176478154612e-05,
+      "loss": 6.846,
+      "step": 62000
+    },
+    {
+      "epoch": 2.6345740420688784,
+      "grad_norm": 0.7321968078613281,
+      "learning_rate": 9.09960125198302e-05,
+      "loss": 6.8527,
+      "step": 62500
+    },
+    {
+      "epoch": 2.655650634405429,
+      "grad_norm": 0.9571184515953064,
+      "learning_rate": 9.091026025811432e-05,
+      "loss": 6.8468,
+      "step": 63000
+    },
+    {
+      "epoch": 2.6767272267419804,
+      "grad_norm": 0.8065069317817688,
+      "learning_rate": 9.082450799639842e-05,
+      "loss": 6.8442,
+      "step": 63500
+    },
+    {
+      "epoch": 2.6978038190785316,
+      "grad_norm": 0.6704916954040527,
+      "learning_rate": 9.07387557346825e-05,
+      "loss": 6.8409,
+      "step": 64000
+    },
+    {
+      "epoch": 2.7188804114150824,
+      "grad_norm": 0.71282559633255,
+      "learning_rate": 9.06530034729666e-05,
+      "loss": 6.8368,
+      "step": 64500
+    },
+    {
+      "epoch": 2.739957003751633,
+      "grad_norm": 0.7310541272163391,
+      "learning_rate": 9.05672512112507e-05,
+      "loss": 6.8348,
+      "step": 65000
+    },
+    {
+      "epoch": 2.7610335960881844,
+      "grad_norm": 0.7924425005912781,
+      "learning_rate": 9.04814989495348e-05,
+      "loss": 6.8288,
+      "step": 65500
+    },
+    {
+      "epoch": 2.7821101884247357,
+      "grad_norm": 0.6368079781532288,
+      "learning_rate": 9.03957466878189e-05,
+      "loss": 6.8363,
+      "step": 66000
+    },
+    {
+      "epoch": 2.8031867807612865,
+      "grad_norm": 0.9884883165359497,
+      "learning_rate": 9.030999442610299e-05,
+      "loss": 6.7893,
+      "step": 66500
+    },
+    {
+      "epoch": 2.8242633730978377,
+      "grad_norm": 1.618870735168457,
+      "learning_rate": 9.022424216438709e-05,
+      "loss": 5.9981,
+      "step": 67000
+    },
+    {
+      "epoch": 2.8453399654343885,
+      "grad_norm": 1.4446355104446411,
+      "learning_rate": 9.013848990267119e-05,
+      "loss": 4.9438,
+      "step": 67500
+    },
+    {
+      "epoch": 2.8664165577709397,
+      "grad_norm": 1.1582340002059937,
+      "learning_rate": 9.005273764095529e-05,
+      "loss": 4.6468,
+      "step": 68000
+    },
+    {
+      "epoch": 2.8874931501074905,
+      "grad_norm": 1.2521469593048096,
+      "learning_rate": 8.996698537923937e-05,
+      "loss": 4.4884,
+      "step": 68500
+    },
+    {
+      "epoch": 2.9085697424440418,
+      "grad_norm": 1.2781131267547607,
+      "learning_rate": 8.988123311752348e-05,
+      "loss": 4.3502,
+      "step": 69000
+    },
+    {
+      "epoch": 2.9296463347805926,
+      "grad_norm": 1.1442620754241943,
+      "learning_rate": 8.979548085580758e-05,
+      "loss": 4.2435,
+      "step": 69500
+    },
+    {
+      "epoch": 2.950722927117144,
+      "grad_norm": 1.2043843269348145,
+      "learning_rate": 8.970972859409167e-05,
+      "loss": 4.1528,
+      "step": 70000
+    },
+    {
+      "epoch": 2.950722927117144,
+      "eval_accuracy": 0.5410019024051829,
+      "eval_loss": 3.808379888534546,
+      "eval_runtime": 368.6387,
+      "eval_samples_per_second": 327.437,
+      "eval_steps_per_second": 3.413,
+      "step": 70000
+    },
+    {
+      "epoch": 2.9717995194536946,
+      "grad_norm": 1.0711034536361694,
+      "learning_rate": 8.962397633237576e-05,
+      "loss": 4.0909,
+      "step": 70500
+    },
+    {
+      "epoch": 2.992876111790246,
+      "grad_norm": 1.041051983833313,
+      "learning_rate": 8.953822407065988e-05,
+      "loss": 4.0286,
+      "step": 71000
+    },
+    {
+      "epoch": 3.0139527041267966,
+      "grad_norm": 1.0935906171798706,
+      "learning_rate": 8.945247180894396e-05,
+      "loss": 3.9756,
+      "step": 71500
+    },
+    {
+      "epoch": 3.035029296463348,
+      "grad_norm": 1.1561946868896484,
+      "learning_rate": 8.936671954722806e-05,
+      "loss": 3.9307,
+      "step": 72000
+    },
+    {
+      "epoch": 3.0561058887998986,
+      "grad_norm": 1.1335831880569458,
+      "learning_rate": 8.928096728551216e-05,
+      "loss": 3.8953,
+      "step": 72500
+    },
+    {
+      "epoch": 3.07718248113645,
+      "grad_norm": 1.0603992938995361,
+      "learning_rate": 8.919521502379626e-05,
+      "loss": 3.8574,
+      "step": 73000
+    },
+    {
+      "epoch": 3.0982590734730007,
+      "grad_norm": 1.2654424905776978,
+      "learning_rate": 8.910946276208035e-05,
+      "loss": 3.8298,
+      "step": 73500
+    },
+    {
+      "epoch": 3.119335665809552,
+      "grad_norm": 1.0848084688186646,
+      "learning_rate": 8.902371050036445e-05,
+      "loss": 3.7919,
+      "step": 74000
+    },
+    {
+      "epoch": 3.140412258146103,
+      "grad_norm": 1.0994441509246826,
+      "learning_rate": 8.893795823864855e-05,
+      "loss": 3.7729,
+      "step": 74500
+    },
+    {
+      "epoch": 3.161488850482654,
+      "grad_norm": 1.1664998531341553,
+      "learning_rate": 8.885220597693265e-05,
+      "loss": 3.7483,
+      "step": 75000
+    },
+    {
+      "epoch": 3.182565442819205,
+      "grad_norm": 1.1818132400512695,
+      "learning_rate": 8.876645371521675e-05,
+      "loss": 3.7235,
+      "step": 75500
+    },
+    {
+      "epoch": 3.203642035155756,
+      "grad_norm": 1.058892846107483,
+      "learning_rate": 8.868070145350083e-05,
+      "loss": 3.704,
+      "step": 76000
+    },
+    {
+      "epoch": 3.224718627492307,
+      "grad_norm": 1.1381815671920776,
+      "learning_rate": 8.859494919178493e-05,
+      "loss": 3.6856,
+      "step": 76500
+    },
+    {
+      "epoch": 3.245795219828858,
+      "grad_norm": 1.247073769569397,
+      "learning_rate": 8.850919693006904e-05,
+      "loss": 3.6627,
+      "step": 77000
+    },
+    {
+      "epoch": 3.2668718121654092,
+      "grad_norm": 1.0255544185638428,
+      "learning_rate": 8.842344466835313e-05,
+      "loss": 3.6448,
+      "step": 77500
+    },
+    {
+      "epoch": 3.28794840450196,
+      "grad_norm": 1.219139814376831,
+      "learning_rate": 8.833769240663723e-05,
+      "loss": 3.6279,
+      "step": 78000
+    },
+    {
+      "epoch": 3.3090249968385113,
+      "grad_norm": 1.1450467109680176,
+      "learning_rate": 8.825194014492132e-05,
+      "loss": 3.614,
+      "step": 78500
+    },
+    {
+      "epoch": 3.330101589175062,
+      "grad_norm": 1.247581958770752,
+      "learning_rate": 8.816618788320542e-05,
+      "loss": 3.5992,
+      "step": 79000
+    },
+    {
+      "epoch": 3.3511781815116133,
+      "grad_norm": 1.0956048965454102,
+      "learning_rate": 8.808043562148952e-05,
+      "loss": 3.5852,
+      "step": 79500
+    },
+    {
+      "epoch": 3.372254773848164,
+      "grad_norm": 1.0880578756332397,
+      "learning_rate": 8.799468335977362e-05,
+      "loss": 3.5714,
+      "step": 80000
+    },
+    {
+      "epoch": 3.372254773848164,
+      "eval_accuracy": 0.6195616570769593,
+      "eval_loss": 3.283751964569092,
+      "eval_runtime": 400.1483,
+      "eval_samples_per_second": 301.653,
+      "eval_steps_per_second": 3.144,
+      "step": 80000
+    },
+    {
+      "epoch": 3.3933313661847153,
+      "grad_norm": 1.0292654037475586,
+      "learning_rate": 8.790893109805772e-05,
+      "loss": 3.5591,
+      "step": 80500
+    },
+    {
+      "epoch": 3.414407958521266,
+      "grad_norm": 1.0292627811431885,
+      "learning_rate": 8.782317883634182e-05,
+      "loss": 3.5453,
+      "step": 81000
+    },
+    {
+      "epoch": 3.4354845508578173,
+      "grad_norm": 1.0922260284423828,
+      "learning_rate": 8.773742657462591e-05,
+      "loss": 3.5327,
+      "step": 81500
+    },
+    {
+      "epoch": 3.456561143194368,
+      "grad_norm": 1.156275749206543,
+      "learning_rate": 8.765167431291e-05,
+      "loss": 3.5219,
+      "step": 82000
+    },
+    {
+      "epoch": 3.4776377355309194,
+      "grad_norm": 1.0010485649108887,
+      "learning_rate": 8.756592205119411e-05,
+      "loss": 3.5113,
+      "step": 82500
+    },
+    {
+      "epoch": 3.4987143278674706,
+      "grad_norm": 1.117132306098938,
+      "learning_rate": 8.748016978947821e-05,
+      "loss": 3.5014,
+      "step": 83000
+    },
+    {
+      "epoch": 3.5197909202040214,
+      "grad_norm": 1.1347253322601318,
+      "learning_rate": 8.73944175277623e-05,
+      "loss": 3.4909,
+      "step": 83500
+    },
+    {
+      "epoch": 3.540867512540572,
+      "grad_norm": 1.2612993717193604,
+      "learning_rate": 8.730866526604639e-05,
+      "loss": 3.4833,
+      "step": 84000
+    },
+    {
+      "epoch": 3.5619441048771234,
+      "grad_norm": 1.0251561403274536,
+      "learning_rate": 8.72229130043305e-05,
+      "loss": 3.4734,
+      "step": 84500
+    },
+    {
+      "epoch": 3.5830206972136747,
+      "grad_norm": 1.2212690114974976,
+      "learning_rate": 8.713716074261459e-05,
+      "loss": 3.4651,
+      "step": 85000
+    },
+    {
+      "epoch": 3.6040972895502255,
+      "grad_norm": 1.1672661304473877,
+      "learning_rate": 8.705140848089869e-05,
+      "loss": 3.4552,
+      "step": 85500
+    },
+    {
+      "epoch": 3.6251738818867767,
+      "grad_norm": 1.1889482736587524,
+      "learning_rate": 8.696565621918278e-05,
+      "loss": 3.4433,
+      "step": 86000
+    },
+    {
+      "epoch": 3.6462504742233275,
+      "grad_norm": 1.035056233406067,
+      "learning_rate": 8.687990395746688e-05,
+      "loss": 3.4372,
+      "step": 86500
+    },
+    {
+      "epoch": 3.6673270665598787,
+      "grad_norm": 1.01767098903656,
+      "learning_rate": 8.679415169575098e-05,
+      "loss": 3.43,
+      "step": 87000
+    },
+    {
+      "epoch": 3.6884036588964295,
+      "grad_norm": 1.149052619934082,
+      "learning_rate": 8.670839943403508e-05,
+      "loss": 3.4225,
+      "step": 87500
+    },
+    {
+      "epoch": 3.7094802512329808,
+      "grad_norm": 1.263179063796997,
+      "learning_rate": 8.662264717231916e-05,
+      "loss": 3.4147,
+      "step": 88000
+    },
+    {
+      "epoch": 3.7305568435695315,
+      "grad_norm": 1.1375000476837158,
+      "learning_rate": 8.653689491060328e-05,
+      "loss": 3.4076,
+      "step": 88500
+    },
+    {
+      "epoch": 3.751633435906083,
+      "grad_norm": 1.1006150245666504,
+      "learning_rate": 8.645114264888737e-05,
+      "loss": 3.4015,
+      "step": 89000
+    },
+    {
+      "epoch": 3.7727100282426336,
+      "grad_norm": 1.1122090816497803,
+      "learning_rate": 8.636539038717146e-05,
+      "loss": 3.3897,
+      "step": 89500
+    },
+    {
+      "epoch": 3.793786620579185,
+      "grad_norm": 1.2452670335769653,
+      "learning_rate": 8.627963812545556e-05,
+      "loss": 3.3831,
+      "step": 90000
+    },
+    {
+      "epoch": 3.793786620579185,
+      "eval_accuracy": 0.6469675251855653,
+      "eval_loss": 3.110896587371826,
+      "eval_runtime": 384.2416,
+      "eval_samples_per_second": 314.141,
+      "eval_steps_per_second": 3.274,
+      "step": 90000
+    },
+    {
+      "epoch": 3.8148632129157356,
+      "grad_norm": 1.1392483711242676,
+      "learning_rate": 8.619388586373967e-05,
+      "loss": 3.3811,
+      "step": 90500
+    },
+    {
+      "epoch": 3.835939805252287,
+      "grad_norm": 1.1481903791427612,
+      "learning_rate": 8.610813360202375e-05,
+      "loss": 3.3686,
+      "step": 91000
+    },
+    {
+      "epoch": 3.857016397588838,
+      "grad_norm": 1.103860855102539,
+      "learning_rate": 8.602238134030785e-05,
+      "loss": 3.3662,
+      "step": 91500
+    },
+    {
+      "epoch": 3.878092989925389,
+      "grad_norm": 1.1099293231964111,
+      "learning_rate": 8.593662907859195e-05,
+      "loss": 3.3536,
+      "step": 92000
+    },
+    {
+      "epoch": 3.8991695822619397,
+      "grad_norm": 0.9830219745635986,
+      "learning_rate": 8.585087681687605e-05,
+      "loss": 3.352,
+      "step": 92500
+    },
+    {
+      "epoch": 3.920246174598491,
+      "grad_norm": 1.107896089553833,
+      "learning_rate": 8.576512455516015e-05,
+      "loss": 3.3485,
+      "step": 93000
+    },
+    {
+      "epoch": 3.941322766935042,
+      "grad_norm": 1.2672383785247803,
+      "learning_rate": 8.567937229344425e-05,
+      "loss": 3.3396,
+      "step": 93500
+    },
+    {
+      "epoch": 3.962399359271593,
+      "grad_norm": 1.0287253856658936,
+      "learning_rate": 8.559362003172834e-05,
+      "loss": 3.3355,
+      "step": 94000
+    },
+    {
+      "epoch": 3.983475951608144,
+      "grad_norm": 1.1357077360153198,
+      "learning_rate": 8.550786777001244e-05,
+      "loss": 3.3302,
+      "step": 94500
+    },
+    {
+      "epoch": 4.004552543944695,
+      "grad_norm": 1.084702491760254,
+      "learning_rate": 8.542211550829654e-05,
+      "loss": 3.3232,
+      "step": 95000
+    },
+    {
+      "epoch": 4.025629136281246,
+      "grad_norm": 1.0517680644989014,
+      "learning_rate": 8.533636324658063e-05,
+      "loss": 3.3111,
+      "step": 95500
+    },
+    {
+      "epoch": 4.046705728617797,
+      "grad_norm": 1.0653400421142578,
+      "learning_rate": 8.525061098486474e-05,
+      "loss": 3.3122,
+      "step": 96000
+    },
+    {
+      "epoch": 4.067782320954348,
+      "grad_norm": 1.031496524810791,
+      "learning_rate": 8.516485872314884e-05,
+      "loss": 3.3015,
+      "step": 96500
+    },
+    {
+      "epoch": 4.088858913290899,
+      "grad_norm": 1.0832912921905518,
+      "learning_rate": 8.507910646143292e-05,
+      "loss": 3.299,
+      "step": 97000
+    },
+    {
+      "epoch": 4.10993550562745,
+      "grad_norm": 1.1457117795944214,
+      "learning_rate": 8.499335419971702e-05,
+      "loss": 3.2933,
+      "step": 97500
+    },
+    {
+      "epoch": 4.1310120979640015,
+      "grad_norm": 1.0359746217727661,
+      "learning_rate": 8.490760193800112e-05,
+      "loss": 3.2884,
+      "step": 98000
+    },
+    {
+      "epoch": 4.152088690300552,
+      "grad_norm": 1.2416901588439941,
+      "learning_rate": 8.482184967628522e-05,
+      "loss": 3.283,
+      "step": 98500
+    },
+    {
+      "epoch": 4.173165282637103,
+      "grad_norm": 1.068037509918213,
+      "learning_rate": 8.473609741456931e-05,
+      "loss": 3.2854,
+      "step": 99000
+    },
+    {
+      "epoch": 4.194241874973654,
+      "grad_norm": 1.2899117469787598,
+      "learning_rate": 8.46503451528534e-05,
+      "loss": 3.2816,
+      "step": 99500
+    },
+    {
+      "epoch": 4.2153184673102055,
+      "grad_norm": 1.0810317993164062,
+      "learning_rate": 8.456459289113751e-05,
+      "loss": 3.2724,
+      "step": 100000
+    },
+    {
+      "epoch": 4.2153184673102055,
+      "eval_accuracy": 0.6627946292165982,
+      "eval_loss": 3.0118486881256104,
+      "eval_runtime": 380.9852,
+      "eval_samples_per_second": 316.826,
+      "eval_steps_per_second": 3.302,
+      "step": 100000
+    },
+    {
+      "epoch": 4.236395059646756,
+      "grad_norm": 1.1014373302459717,
+      "learning_rate": 8.447884062942161e-05,
+      "loss": 3.2728,
+      "step": 100500
+    },
+    {
+      "epoch": 4.257471651983307,
+      "grad_norm": 1.189949631690979,
+      "learning_rate": 8.43930883677057e-05,
+      "loss": 3.2689,
+      "step": 101000
+    },
+    {
+      "epoch": 4.278548244319858,
+      "grad_norm": 1.0565866231918335,
+      "learning_rate": 8.430733610598979e-05,
+      "loss": 3.26,
+      "step": 101500
+    },
+    {
+      "epoch": 4.29962483665641,
+      "grad_norm": 1.0556780099868774,
+      "learning_rate": 8.42215838442739e-05,
+      "loss": 3.2559,
+      "step": 102000
+    },
+    {
+      "epoch": 4.32070142899296,
+      "grad_norm": 0.9717570543289185,
+      "learning_rate": 8.413583158255799e-05,
+      "loss": 3.2532,
+      "step": 102500
+    },
+    {
+      "epoch": 4.341778021329511,
+      "grad_norm": 0.9797452092170715,
+      "learning_rate": 8.405007932084209e-05,
+      "loss": 3.247,
+      "step": 103000
+    },
+    {
+      "epoch": 4.362854613666062,
+      "grad_norm": 1.0778639316558838,
+      "learning_rate": 8.396432705912619e-05,
+      "loss": 3.2461,
+      "step": 103500
+    },
+    {
+      "epoch": 4.383931206002614,
+      "grad_norm": 1.0978378057479858,
+      "learning_rate": 8.387857479741028e-05,
+      "loss": 3.2453,
+      "step": 104000
+    },
+    {
+      "epoch": 4.405007798339165,
+      "grad_norm": 1.0536212921142578,
+      "learning_rate": 8.379282253569438e-05,
+      "loss": 3.2392,
+      "step": 104500
+    },
+    {
+      "epoch": 4.426084390675715,
+      "grad_norm": 1.0425158739089966,
+      "learning_rate": 8.370707027397848e-05,
+      "loss": 3.2341,
+      "step": 105000
+    },
+    {
+      "epoch": 4.4471609830122665,
+      "grad_norm": 1.002151608467102,
+      "learning_rate": 8.362131801226258e-05,
+      "loss": 3.2303,
+      "step": 105500
+    },
+    {
+      "epoch": 4.468237575348818,
+      "grad_norm": 1.0568463802337646,
+      "learning_rate": 8.353556575054668e-05,
+      "loss": 3.2247,
+      "step": 106000
+    },
+    {
+      "epoch": 4.489314167685369,
+      "grad_norm": 1.0124139785766602,
+      "learning_rate": 8.344981348883077e-05,
+      "loss": 3.2264,
+      "step": 106500
+    },
+    {
+      "epoch": 4.510390760021919,
+      "grad_norm": 1.0119850635528564,
+      "learning_rate": 8.336406122711486e-05,
+      "loss": 3.2188,
+      "step": 107000
+    },
+    {
+      "epoch": 4.5314673523584705,
+      "grad_norm": 1.0425482988357544,
+      "learning_rate": 8.327830896539897e-05,
+      "loss": 3.2216,
+      "step": 107500
+    },
+    {
+      "epoch": 4.552543944695022,
+      "grad_norm": 1.2428576946258545,
+      "learning_rate": 8.319255670368307e-05,
+      "loss": 3.2149,
+      "step": 108000
+    },
+    {
+      "epoch": 4.573620537031573,
+      "grad_norm": 1.178147792816162,
+      "learning_rate": 8.310680444196715e-05,
+      "loss": 3.2113,
+      "step": 108500
+    },
+    {
+      "epoch": 4.594697129368123,
+      "grad_norm": 1.0125010013580322,
+      "learning_rate": 8.302105218025125e-05,
+      "loss": 3.2021,
+      "step": 109000
+    },
+    {
+      "epoch": 4.615773721704675,
+      "grad_norm": 1.0107455253601074,
+      "learning_rate": 8.293529991853536e-05,
+      "loss": 3.2088,
+      "step": 109500
+    },
+    {
+      "epoch": 4.636850314041226,
+      "grad_norm": 1.0441641807556152,
+      "learning_rate": 8.284954765681945e-05,
+      "loss": 3.2036,
+      "step": 110000
+    },
+    {
+      "epoch": 4.636850314041226,
+      "eval_accuracy": 0.673201946120713,
+      "eval_loss": 2.9436707496643066,
+      "eval_runtime": 417.9807,
+      "eval_samples_per_second": 288.784,
+      "eval_steps_per_second": 3.01,
+      "step": 110000
+    },
+    {
+      "epoch": 4.657926906377777,
+      "grad_norm": 0.9846625328063965,
+      "learning_rate": 8.276379539510355e-05,
+      "loss": 3.2004,
+      "step": 110500
+    },
+    {
+      "epoch": 4.679003498714328,
+      "grad_norm": 1.1222068071365356,
+      "learning_rate": 8.267804313338765e-05,
+      "loss": 3.1973,
+      "step": 111000
+    },
+    {
+      "epoch": 4.700080091050879,
+      "grad_norm": 1.1272120475769043,
+      "learning_rate": 8.259229087167174e-05,
+      "loss": 3.1933,
+      "step": 111500
+    },
+    {
+      "epoch": 4.72115668338743,
+      "grad_norm": 1.0700374841690063,
+      "learning_rate": 8.250653860995584e-05,
+      "loss": 3.1877,
+      "step": 112000
+    },
+    {
+      "epoch": 4.742233275723981,
+      "grad_norm": 1.091374397277832,
+      "learning_rate": 8.242078634823994e-05,
+      "loss": 3.1877,
+      "step": 112500
+    },
+    {
+      "epoch": 4.763309868060532,
+      "grad_norm": 1.103630781173706,
+      "learning_rate": 8.233503408652403e-05,
+      "loss": 3.1848,
+      "step": 113000
+    },
+    {
+      "epoch": 4.784386460397083,
+      "grad_norm": 1.080336570739746,
+      "learning_rate": 8.224928182480814e-05,
+      "loss": 3.1857,
+      "step": 113500
+    },
+    {
+      "epoch": 4.805463052733634,
+      "grad_norm": 1.239959716796875,
+      "learning_rate": 8.216352956309224e-05,
+      "loss": 3.1746,
+      "step": 114000
+    },
+    {
+      "epoch": 4.826539645070185,
+      "grad_norm": 1.0558053255081177,
+      "learning_rate": 8.207777730137632e-05,
+      "loss": 3.1738,
+      "step": 114500
+    },
+    {
+      "epoch": 4.847616237406736,
+      "grad_norm": 1.1350589990615845,
+      "learning_rate": 8.199202503966042e-05,
+      "loss": 3.1728,
+      "step": 115000
+    },
+    {
+      "epoch": 4.868692829743287,
+      "grad_norm": 1.1721267700195312,
+      "learning_rate": 8.190627277794453e-05,
+      "loss": 3.1696,
+      "step": 115500
+    },
+    {
+      "epoch": 4.889769422079838,
+      "grad_norm": 1.0804462432861328,
+      "learning_rate": 8.182052051622862e-05,
+      "loss": 3.1647,
+      "step": 116000
+    },
+    {
+      "epoch": 4.910846014416389,
+      "grad_norm": 1.2322784662246704,
+      "learning_rate": 8.173476825451271e-05,
+      "loss": 3.1646,
+      "step": 116500
+    },
+    {
+      "epoch": 4.9319226067529405,
+      "grad_norm": 1.2686874866485596,
+      "learning_rate": 8.164901599279681e-05,
+      "loss": 3.1626,
+      "step": 117000
+    },
+    {
+      "epoch": 4.952999199089491,
+      "grad_norm": 1.169320821762085,
+      "learning_rate": 8.156326373108091e-05,
+      "loss": 3.1627,
+      "step": 117500
+    },
+    {
+      "epoch": 4.974075791426042,
+      "grad_norm": 1.0978491306304932,
+      "learning_rate": 8.147751146936501e-05,
+      "loss": 3.1585,
+      "step": 118000
+    },
+    {
+      "epoch": 4.995152383762593,
+      "grad_norm": 1.2362549304962158,
+      "learning_rate": 8.139175920764911e-05,
+      "loss": 3.1517,
+      "step": 118500
+    },
+    {
+      "epoch": 5.0162289760991445,
+      "grad_norm": 1.1215428113937378,
+      "learning_rate": 8.13060069459332e-05,
+      "loss": 3.1513,
+      "step": 119000
+    },
+    {
+      "epoch": 5.037305568435695,
+      "grad_norm": 1.0707147121429443,
+      "learning_rate": 8.12202546842173e-05,
+      "loss": 3.1443,
+      "step": 119500
+    },
+    {
+      "epoch": 5.058382160772246,
+      "grad_norm": 1.1055184602737427,
+      "learning_rate": 8.11345024225014e-05,
+      "loss": 3.1442,
+      "step": 120000
+    },
+    {
+      "epoch": 5.058382160772246,
+      "eval_accuracy": 0.6821309195553876,
+      "eval_loss": 2.8916923999786377,
+      "eval_runtime": 511.3627,
+      "eval_samples_per_second": 236.048,
+      "eval_steps_per_second": 2.46,
+      "step": 120000
+    },
+    {
+      "epoch": 5.079458753108797,
+      "grad_norm": 1.12617027759552,
+      "learning_rate": 8.104875016078549e-05,
+      "loss": 3.1361,
+      "step": 120500
+    },
+    {
+      "epoch": 5.100535345445349,
+      "grad_norm": 1.0911645889282227,
+      "learning_rate": 8.09629978990696e-05,
+      "loss": 3.1348,
+      "step": 121000
+    },
+    {
+      "epoch": 5.1216119377819,
+      "grad_norm": 1.1265887022018433,
+      "learning_rate": 8.08772456373537e-05,
+      "loss": 3.1361,
+      "step": 121500
+    },
+    {
+      "epoch": 5.14268853011845,
+      "grad_norm": 1.064839482307434,
+      "learning_rate": 8.079149337563778e-05,
+      "loss": 3.1341,
+      "step": 122000
+    },
+    {
+      "epoch": 5.163765122455001,
+      "grad_norm": 1.020843505859375,
+      "learning_rate": 8.070574111392188e-05,
+      "loss": 3.1287,
+      "step": 122500
+    },
+    {
+      "epoch": 5.184841714791553,
+      "grad_norm": 1.2647985219955444,
+      "learning_rate": 8.061998885220599e-05,
+      "loss": 3.1272,
+      "step": 123000
+    },
+    {
+      "epoch": 5.205918307128104,
+      "grad_norm": 1.0691004991531372,
+      "learning_rate": 8.053423659049008e-05,
+      "loss": 3.126,
+      "step": 123500
+    },
+    {
+      "epoch": 5.226994899464654,
+      "grad_norm": 1.0166749954223633,
+      "learning_rate": 8.044848432877418e-05,
+      "loss": 3.1257,
+      "step": 124000
+    },
+    {
+      "epoch": 5.2480714918012055,
+      "grad_norm": 1.0583964586257935,
+      "learning_rate": 8.036273206705827e-05,
+      "loss": 3.1262,
+      "step": 124500
+    },
+    {
+      "epoch": 5.269148084137757,
+      "grad_norm": 1.0643447637557983,
+      "learning_rate": 8.027697980534237e-05,
+      "loss": 3.1191,
+      "step": 125000
+    },
+    {
+      "epoch": 5.290224676474308,
+      "grad_norm": 1.1366677284240723,
+      "learning_rate": 8.019122754362647e-05,
+      "loss": 3.1202,
+      "step": 125500
+    },
+    {
+      "epoch": 5.311301268810858,
+      "grad_norm": 1.0518467426300049,
+      "learning_rate": 8.010547528191057e-05,
+      "loss": 3.1182,
+      "step": 126000
+    },
+    {
+      "epoch": 5.3323778611474095,
+      "grad_norm": 1.1233019828796387,
+      "learning_rate": 8.001972302019465e-05,
+      "loss": 3.1118,
+      "step": 126500
+    },
+    {
+      "epoch": 5.353454453483961,
+      "grad_norm": 1.1249240636825562,
+      "learning_rate": 7.993397075847876e-05,
+      "loss": 3.1102,
+      "step": 127000
+    },
+    {
+      "epoch": 5.374531045820512,
+      "grad_norm": 1.2353320121765137,
+      "learning_rate": 7.984821849676286e-05,
+      "loss": 3.1129,
+      "step": 127500
+    },
+    {
+      "epoch": 5.395607638157063,
+      "grad_norm": 1.2399853467941284,
+      "learning_rate": 7.976246623504695e-05,
+      "loss": 3.1089,
+      "step": 128000
+    },
+    {
+      "epoch": 5.416684230493614,
+      "grad_norm": 1.0391329526901245,
+      "learning_rate": 7.967671397333105e-05,
+      "loss": 3.1028,
+      "step": 128500
+    },
+    {
+      "epoch": 5.437760822830165,
+      "grad_norm": 1.0521605014801025,
+      "learning_rate": 7.959096171161516e-05,
+      "loss": 3.1023,
+      "step": 129000
+    },
+    {
+      "epoch": 5.458837415166716,
+      "grad_norm": 1.021182894706726,
+      "learning_rate": 7.950520944989924e-05,
+      "loss": 3.0997,
+      "step": 129500
+    },
+    {
+      "epoch": 5.479914007503266,
+      "grad_norm": 1.0592447519302368,
+      "learning_rate": 7.941945718818334e-05,
+      "loss": 3.0973,
+      "step": 130000
+    },
+    {
+      "epoch": 5.479914007503266,
+      "eval_accuracy": 0.6892488590492393,
+      "eval_loss": 2.8480405807495117,
+      "eval_runtime": 381.1504,
+      "eval_samples_per_second": 316.689,
+      "eval_steps_per_second": 3.301,
+      "step": 130000
+    },
+    {
+      "epoch": 5.500990599839818,
+      "grad_norm": 1.0475915670394897,
+      "learning_rate": 7.933370492646744e-05,
+      "loss": 3.1024,
+      "step": 130500
+    },
+    {
+      "epoch": 5.522067192176369,
+      "grad_norm": 1.6009825468063354,
+      "learning_rate": 7.924795266475154e-05,
+      "loss": 3.0932,
+      "step": 131000
+    },
+    {
+      "epoch": 5.54314378451292,
+      "grad_norm": 1.0657505989074707,
+      "learning_rate": 7.916220040303564e-05,
+      "loss": 3.0885,
+      "step": 131500
+    },
+    {
+      "epoch": 5.564220376849471,
+      "grad_norm": 1.13173508644104,
+      "learning_rate": 7.907644814131972e-05,
+      "loss": 3.092,
+      "step": 132000
+    },
+    {
+      "epoch": 5.585296969186022,
+      "grad_norm": 1.1277837753295898,
+      "learning_rate": 7.899069587960383e-05,
+      "loss": 3.0864,
+      "step": 132500
+    },
+    {
+      "epoch": 5.606373561522573,
+      "grad_norm": 0.9967442154884338,
+      "learning_rate": 7.890494361788793e-05,
+      "loss": 3.0877,
+      "step": 133000
+    },
+    {
+      "epoch": 5.627450153859124,
+      "grad_norm": 1.1431701183319092,
+      "learning_rate": 7.881919135617202e-05,
+      "loss": 3.0822,
+      "step": 133500
+    },
+    {
+      "epoch": 5.648526746195675,
+      "grad_norm": 1.0610195398330688,
+      "learning_rate": 7.873343909445611e-05,
+      "loss": 3.0796,
+      "step": 134000
+    },
+    {
+      "epoch": 5.669603338532226,
+      "grad_norm": 1.0659259557724,
+      "learning_rate": 7.864768683274023e-05,
+      "loss": 3.0851,
+      "step": 134500
+    },
+    {
+      "epoch": 5.690679930868777,
+      "grad_norm": 1.0729259252548218,
+      "learning_rate": 7.856193457102431e-05,
+      "loss": 3.0833,
+      "step": 135000
+    },
+    {
+      "epoch": 5.711756523205328,
+      "grad_norm": 1.2815239429473877,
+      "learning_rate": 7.847618230930841e-05,
+      "loss": 3.0811,
+      "step": 135500
+    },
+    {
+      "epoch": 5.7328331155418795,
+      "grad_norm": 1.1783567667007446,
+      "learning_rate": 7.839043004759251e-05,
+      "loss": 3.0789,
+      "step": 136000
+    },
+    {
+      "epoch": 5.75390970787843,
+      "grad_norm": 1.0411161184310913,
+      "learning_rate": 7.83046777858766e-05,
+      "loss": 3.0754,
+      "step": 136500
+    },
+    {
+      "epoch": 5.774986300214981,
+      "grad_norm": 1.092773675918579,
+      "learning_rate": 7.82189255241607e-05,
+      "loss": 3.0738,
+      "step": 137000
+    },
+    {
+      "epoch": 5.796062892551532,
+      "grad_norm": 1.392815113067627,
+      "learning_rate": 7.81331732624448e-05,
+      "loss": 3.0718,
+      "step": 137500
+    },
+    {
+      "epoch": 5.8171394848880835,
+      "grad_norm": 1.016334891319275,
+      "learning_rate": 7.804742100072889e-05,
+      "loss": 3.068,
+      "step": 138000
+    },
+    {
+      "epoch": 5.838216077224635,
+      "grad_norm": 1.1654419898986816,
+      "learning_rate": 7.7961668739013e-05,
+      "loss": 3.0706,
+      "step": 138500
+    },
+    {
+      "epoch": 5.859292669561185,
+      "grad_norm": 1.040810465812683,
+      "learning_rate": 7.78759164772971e-05,
+      "loss": 3.068,
+      "step": 139000
+    },
+    {
+      "epoch": 5.880369261897736,
+      "grad_norm": 1.0924264192581177,
+      "learning_rate": 7.779016421558118e-05,
+      "loss": 3.0651,
+      "step": 139500
+    },
+    {
+      "epoch": 5.901445854234288,
+      "grad_norm": 1.1327614784240723,
+      "learning_rate": 7.770441195386528e-05,
+      "loss": 3.0658,
+      "step": 140000
+    },
+    {
+      "epoch": 5.901445854234288,
+      "eval_accuracy": 0.6945600434897065,
+      "eval_loss": 2.8129515647888184,
+      "eval_runtime": 362.0103,
+      "eval_samples_per_second": 333.432,
+      "eval_steps_per_second": 3.475,
+      "step": 140000
+    },
+    {
+      "epoch": 5.922522446570838,
+      "grad_norm": 1.0177291631698608,
+      "learning_rate": 7.761865969214939e-05,
+      "loss": 3.0576,
+      "step": 140500
+    },
+    {
+      "epoch": 5.943599038907389,
+      "grad_norm": 1.0612154006958008,
+      "learning_rate": 7.753290743043348e-05,
+      "loss": 3.0623,
+      "step": 141000
+    },
+    {
+      "epoch": 5.96467563124394,
+      "grad_norm": 1.1396235227584839,
+      "learning_rate": 7.744715516871758e-05,
+      "loss": 3.0614,
+      "step": 141500
+    },
+    {
+      "epoch": 5.985752223580492,
+      "grad_norm": 1.1105926036834717,
+      "learning_rate": 7.736140290700167e-05,
+      "loss": 3.0552,
+      "step": 142000
+    },
+    {
+      "epoch": 6.006828815917043,
+      "grad_norm": 1.2960937023162842,
+      "learning_rate": 7.727565064528577e-05,
+      "loss": 3.0545,
+      "step": 142500
+    },
+    {
+      "epoch": 6.027905408253593,
+      "grad_norm": 1.1689718961715698,
+      "learning_rate": 7.718989838356987e-05,
+      "loss": 3.0503,
+      "step": 143000
+    },
+    {
+      "epoch": 6.0489820005901445,
+      "grad_norm": 1.1816215515136719,
+      "learning_rate": 7.710414612185397e-05,
+      "loss": 3.0519,
+      "step": 143500
+    },
+    {
+      "epoch": 6.070058592926696,
+      "grad_norm": 1.0606178045272827,
+      "learning_rate": 7.701839386013807e-05,
+      "loss": 3.0472,
+      "step": 144000
+    },
+    {
+      "epoch": 6.091135185263247,
+      "grad_norm": 1.033261775970459,
+      "learning_rate": 7.693264159842217e-05,
+      "loss": 3.0438,
+      "step": 144500
+    },
+    {
+      "epoch": 6.112211777599797,
+      "grad_norm": 1.05693781375885,
+      "learning_rate": 7.684688933670626e-05,
+      "loss": 3.0437,
+      "step": 145000
+    },
+    {
+      "epoch": 6.1332883699363485,
+      "grad_norm": 1.1272575855255127,
+      "learning_rate": 7.676113707499035e-05,
+      "loss": 3.0422,
+      "step": 145500
+    },
+    {
+      "epoch": 6.1543649622729,
+      "grad_norm": 1.0037453174591064,
+      "learning_rate": 7.667538481327446e-05,
+      "loss": 3.0413,
+      "step": 146000
+    },
+    {
+      "epoch": 6.175441554609451,
+      "grad_norm": 1.147699236869812,
+      "learning_rate": 7.658963255155856e-05,
+      "loss": 3.0448,
+      "step": 146500
+    },
+    {
+      "epoch": 6.196518146946001,
+      "grad_norm": 1.07542884349823,
+      "learning_rate": 7.650388028984264e-05,
+      "loss": 3.0404,
+      "step": 147000
+    },
+    {
+      "epoch": 6.217594739282553,
+      "grad_norm": 1.3079478740692139,
+      "learning_rate": 7.641812802812674e-05,
+      "loss": 3.0379,
+      "step": 147500
+    },
+    {
+      "epoch": 6.238671331619104,
+      "grad_norm": 1.059332013130188,
+      "learning_rate": 7.633237576641084e-05,
+      "loss": 3.0361,
+      "step": 148000
+    },
+    {
+      "epoch": 6.259747923955655,
+      "grad_norm": 1.1732531785964966,
+      "learning_rate": 7.624662350469494e-05,
+      "loss": 3.0353,
+      "step": 148500
+    },
+    {
+      "epoch": 6.280824516292206,
+      "grad_norm": 1.1434768438339233,
+      "learning_rate": 7.616087124297904e-05,
+      "loss": 3.0335,
+      "step": 149000
+    },
+    {
+      "epoch": 6.301901108628757,
+      "grad_norm": 1.1131255626678467,
+      "learning_rate": 7.607511898126313e-05,
+      "loss": 3.0363,
+      "step": 149500
+    },
+    {
+      "epoch": 6.322977700965308,
+      "grad_norm": 1.0412867069244385,
+      "learning_rate": 7.598936671954723e-05,
+      "loss": 3.03,
+      "step": 150000
+    },
+    {
+      "epoch": 6.322977700965308,
+      "eval_accuracy": 0.6992791348738577,
+      "eval_loss": 2.7858409881591797,
+      "eval_runtime": 389.7146,
+      "eval_samples_per_second": 309.729,
+      "eval_steps_per_second": 3.228,
+      "step": 150000
+    },
+    {
+      "epoch": 6.344054293301859,
+      "grad_norm": 1.0903054475784302,
+      "learning_rate": 7.590361445783133e-05,
+      "loss": 3.0312,
+      "step": 150500
+    },
+    {
+      "epoch": 6.36513088563841,
+      "grad_norm": 1.0271650552749634,
+      "learning_rate": 7.581786219611543e-05,
+      "loss": 3.0289,
+      "step": 151000
+    },
+    {
+      "epoch": 6.386207477974961,
+      "grad_norm": 1.0638962984085083,
+      "learning_rate": 7.573210993439951e-05,
+      "loss": 3.0288,
+      "step": 151500
+    },
+    {
+      "epoch": 6.407284070311512,
+      "grad_norm": 1.1337742805480957,
+      "learning_rate": 7.564635767268363e-05,
+      "loss": 3.0307,
+      "step": 152000
+    },
+    {
+      "epoch": 6.428360662648063,
+      "grad_norm": 1.0439515113830566,
+      "learning_rate": 7.556060541096772e-05,
+      "loss": 3.0248,
+      "step": 152500
+    },
+    {
+      "epoch": 6.449437254984614,
+      "grad_norm": 1.0447839498519897,
+      "learning_rate": 7.547485314925181e-05,
+      "loss": 3.0249,
+      "step": 153000
+    },
+    {
+      "epoch": 6.470513847321165,
+      "grad_norm": 1.1477222442626953,
+      "learning_rate": 7.538910088753591e-05,
+      "loss": 3.0234,
+      "step": 153500
+    },
+    {
+      "epoch": 6.491590439657716,
+      "grad_norm": 1.0677307844161987,
+      "learning_rate": 7.530334862582002e-05,
+      "loss": 3.0199,
+      "step": 154000
+    },
+    {
+      "epoch": 6.512667031994267,
+      "grad_norm": 1.3771947622299194,
+      "learning_rate": 7.52175963641041e-05,
+      "loss": 3.0218,
+      "step": 154500
+    },
+    {
+      "epoch": 6.5337436243308185,
+      "grad_norm": 1.0529382228851318,
+      "learning_rate": 7.51318441023882e-05,
+      "loss": 3.022,
+      "step": 155000
+    },
+    {
+      "epoch": 6.55482021666737,
+      "grad_norm": 1.0381561517715454,
+      "learning_rate": 7.50460918406723e-05,
+      "loss": 3.0175,
+      "step": 155500
+    },
+    {
+      "epoch": 6.57589680900392,
+      "grad_norm": 1.0798434019088745,
+      "learning_rate": 7.49603395789564e-05,
+      "loss": 3.0167,
+      "step": 156000
+    },
+    {
+      "epoch": 6.596973401340471,
+      "grad_norm": 1.1701490879058838,
+      "learning_rate": 7.48745873172405e-05,
+      "loss": 3.0109,
+      "step": 156500
+    },
+    {
+      "epoch": 6.6180499936770225,
+      "grad_norm": 1.2867263555526733,
+      "learning_rate": 7.47888350555246e-05,
+      "loss": 3.0096,
+      "step": 157000
+    },
+    {
+      "epoch": 6.639126586013573,
+      "grad_norm": 1.0917506217956543,
+      "learning_rate": 7.47030827938087e-05,
+      "loss": 3.0108,
+      "step": 157500
+    },
+    {
+      "epoch": 6.660203178350124,
+      "grad_norm": 1.139657974243164,
+      "learning_rate": 7.461733053209279e-05,
+      "loss": 3.0124,
+      "step": 158000
+    },
+    {
+      "epoch": 6.681279770686675,
+      "grad_norm": 1.0662206411361694,
+      "learning_rate": 7.453157827037689e-05,
+      "loss": 3.0149,
+      "step": 158500
+    },
+    {
+      "epoch": 6.702356363023227,
+      "grad_norm": 1.140265941619873,
+      "learning_rate": 7.444582600866098e-05,
+      "loss": 3.0067,
+      "step": 159000
+    },
+    {
+      "epoch": 6.723432955359778,
+      "grad_norm": 1.1930358409881592,
+      "learning_rate": 7.436007374694507e-05,
+      "loss": 3.0094,
+      "step": 159500
+    },
+    {
+      "epoch": 6.744509547696328,
+      "grad_norm": 1.1119061708450317,
+      "learning_rate": 7.427432148522919e-05,
+      "loss": 3.0096,
+      "step": 160000
+    },
+    {
+      "epoch": 6.744509547696328,
+      "eval_accuracy": 0.7035452983707882,
+      "eval_loss": 2.761711359024048,
+      "eval_runtime": 377.3227,
+      "eval_samples_per_second": 319.901,
+      "eval_steps_per_second": 3.334,
+      "step": 160000
+    },
+    {
+      "epoch": 6.765586140032879,
+      "grad_norm": 1.0554249286651611,
+      "learning_rate": 7.418856922351327e-05,
+      "loss": 3.0058,
+      "step": 160500
+    },
+    {
+      "epoch": 6.786662732369431,
+      "grad_norm": 1.1619874238967896,
+      "learning_rate": 7.410281696179737e-05,
+      "loss": 3.0026,
+      "step": 161000
+    },
+    {
+      "epoch": 6.807739324705982,
+      "grad_norm": 1.1459603309631348,
+      "learning_rate": 7.401706470008147e-05,
+      "loss": 3.0001,
+      "step": 161500
+    },
+    {
+      "epoch": 6.828815917042532,
+      "grad_norm": 1.139062523841858,
+      "learning_rate": 7.393131243836557e-05,
+      "loss": 3.0004,
+      "step": 162000
+    },
+    {
+      "epoch": 6.8498925093790834,
+      "grad_norm": 1.1050411462783813,
+      "learning_rate": 7.384556017664966e-05,
+      "loss": 2.998,
+      "step": 162500
+    },
+    {
+      "epoch": 6.870969101715635,
+      "grad_norm": 1.0926766395568848,
+      "learning_rate": 7.375980791493376e-05,
+      "loss": 2.9947,
+      "step": 163000
+    },
+    {
+      "epoch": 6.892045694052186,
+      "grad_norm": 1.1599478721618652,
+      "learning_rate": 7.367405565321786e-05,
+      "loss": 2.9937,
+      "step": 163500
+    },
+    {
+      "epoch": 6.913122286388736,
+      "grad_norm": 1.0994690656661987,
+      "learning_rate": 7.358830339150196e-05,
+      "loss": 2.9959,
+      "step": 164000
+    },
+    {
+      "epoch": 6.9341988787252875,
+      "grad_norm": 1.0609668493270874,
+      "learning_rate": 7.350255112978606e-05,
+      "loss": 2.9991,
+      "step": 164500
+    },
+    {
+      "epoch": 6.955275471061839,
+      "grad_norm": 1.0970399379730225,
+      "learning_rate": 7.341679886807014e-05,
+      "loss": 2.994,
+      "step": 165000
+    },
+    {
+      "epoch": 6.97635206339839,
+      "grad_norm": 1.144362449645996,
+      "learning_rate": 7.333104660635425e-05,
+      "loss": 2.9914,
+      "step": 165500
+    },
+    {
+      "epoch": 6.997428655734941,
+      "grad_norm": 1.0675320625305176,
+      "learning_rate": 7.324529434463835e-05,
+      "loss": 2.9901,
+      "step": 166000
+    },
+    {
+      "epoch": 7.018505248071492,
+      "grad_norm": 1.1204904317855835,
+      "learning_rate": 7.315954208292244e-05,
+      "loss": 2.9891,
+      "step": 166500
+    },
+    {
+      "epoch": 7.039581840408043,
+      "grad_norm": 1.0776515007019043,
+      "learning_rate": 7.307378982120653e-05,
+      "loss": 2.9822,
+      "step": 167000
+    },
+    {
+      "epoch": 7.060658432744594,
+      "grad_norm": 1.1404248476028442,
+      "learning_rate": 7.298803755949063e-05,
+      "loss": 2.9831,
+      "step": 167500
+    },
+    {
+      "epoch": 7.081735025081145,
+      "grad_norm": 1.115059733390808,
+      "learning_rate": 7.290228529777473e-05,
+      "loss": 2.9813,
+      "step": 168000
+    },
+    {
+      "epoch": 7.102811617417696,
+      "grad_norm": 1.1070072650909424,
+      "learning_rate": 7.281653303605883e-05,
+      "loss": 2.9834,
+      "step": 168500
+    },
+    {
+      "epoch": 7.123888209754247,
+      "grad_norm": 1.1591709852218628,
+      "learning_rate": 7.273078077434293e-05,
+      "loss": 2.9868,
+      "step": 169000
+    },
+    {
+      "epoch": 7.144964802090798,
+      "grad_norm": 1.0755780935287476,
+      "learning_rate": 7.264502851262703e-05,
+      "loss": 2.9783,
+      "step": 169500
+    },
+    {
+      "epoch": 7.166041394427349,
+      "grad_norm": 1.1395283937454224,
+      "learning_rate": 7.255927625091112e-05,
+      "loss": 2.9758,
+      "step": 170000
+    },
+    {
+      "epoch": 7.166041394427349,
+      "eval_accuracy": 0.7076852270529109,
+      "eval_loss": 2.738133668899536,
+      "eval_runtime": 475.5178,
+      "eval_samples_per_second": 253.841,
+      "eval_steps_per_second": 2.646,
+      "step": 170000
+    },
+    {
+      "epoch": 7.1871179867639,
+      "grad_norm": 1.0432250499725342,
+      "learning_rate": 7.247352398919521e-05,
+      "loss": 2.9808,
+      "step": 170500
+    },
+    {
+      "epoch": 7.208194579100451,
+      "grad_norm": 1.0722516775131226,
+      "learning_rate": 7.238777172747931e-05,
+      "loss": 2.976,
+      "step": 171000
+    },
+    {
+      "epoch": 7.229271171437002,
+      "grad_norm": 1.0883610248565674,
+      "learning_rate": 7.230201946576342e-05,
+      "loss": 2.9784,
+      "step": 171500
+    },
+    {
+      "epoch": 7.250347763773553,
+      "grad_norm": 1.1319477558135986,
+      "learning_rate": 7.22162672040475e-05,
+      "loss": 2.976,
+      "step": 172000
+    },
+    {
+      "epoch": 7.271424356110104,
+      "grad_norm": 1.1574044227600098,
+      "learning_rate": 7.21305149423316e-05,
+      "loss": 2.9727,
+      "step": 172500
+    },
+    {
+      "epoch": 7.292500948446655,
+      "grad_norm": 1.096825122833252,
+      "learning_rate": 7.20447626806157e-05,
+      "loss": 2.9679,
+      "step": 173000
+    },
+    {
+      "epoch": 7.313577540783206,
+      "grad_norm": 1.1164765357971191,
+      "learning_rate": 7.19590104188998e-05,
+      "loss": 2.9695,
+      "step": 173500
+    },
+    {
+      "epoch": 7.3346541331197574,
+      "grad_norm": 1.1495282649993896,
+      "learning_rate": 7.18732581571839e-05,
+      "loss": 2.9712,
+      "step": 174000
+    },
+    {
+      "epoch": 7.355730725456308,
+      "grad_norm": 1.1710656881332397,
+      "learning_rate": 7.1787505895468e-05,
+      "loss": 2.9694,
+      "step": 174500
+    },
+    {
+      "epoch": 7.376807317792859,
+      "grad_norm": 1.053619146347046,
+      "learning_rate": 7.17017536337521e-05,
+      "loss": 2.9691,
+      "step": 175000
+    },
+    {
+      "epoch": 7.39788391012941,
+      "grad_norm": 1.0789093971252441,
+      "learning_rate": 7.161600137203619e-05,
+      "loss": 2.9708,
+      "step": 175500
+    },
+    {
+      "epoch": 7.4189605024659615,
+      "grad_norm": 1.0652003288269043,
+      "learning_rate": 7.153024911032029e-05,
+      "loss": 2.9638,
+      "step": 176000
+    },
+    {
+      "epoch": 7.440037094802513,
+      "grad_norm": 1.1143277883529663,
+      "learning_rate": 7.144449684860438e-05,
+      "loss": 2.9663,
+      "step": 176500
+    },
+    {
+      "epoch": 7.461113687139063,
+      "grad_norm": 1.1678093671798706,
+      "learning_rate": 7.135874458688849e-05,
+      "loss": 2.9682,
+      "step": 177000
+    },
+    {
+      "epoch": 7.482190279475614,
+      "grad_norm": 1.103148102760315,
+      "learning_rate": 7.127299232517259e-05,
+      "loss": 2.9678,
+      "step": 177500
+    },
+    {
+      "epoch": 7.503266871812166,
+      "grad_norm": 1.0633866786956787,
+      "learning_rate": 7.118724006345667e-05,
+      "loss": 2.964,
+      "step": 178000
+    },
+    {
+      "epoch": 7.524343464148717,
+      "grad_norm": 1.1731432676315308,
+      "learning_rate": 7.110148780174077e-05,
+      "loss": 2.96,
+      "step": 178500
+    },
+    {
+      "epoch": 7.545420056485267,
+      "grad_norm": 1.1338860988616943,
+      "learning_rate": 7.101573554002488e-05,
+      "loss": 2.9595,
+      "step": 179000
+    },
+    {
+      "epoch": 7.566496648821818,
+      "grad_norm": 1.1755574941635132,
+      "learning_rate": 7.092998327830897e-05,
+      "loss": 2.9617,
+      "step": 179500
+    },
+    {
+      "epoch": 7.58757324115837,
+      "grad_norm": 1.2211264371871948,
+      "learning_rate": 7.084423101659306e-05,
+      "loss": 2.9593,
+      "step": 180000
+    },
+    {
+      "epoch": 7.58757324115837,
+      "eval_accuracy": 0.7100974479512268,
+      "eval_loss": 2.724198579788208,
+      "eval_runtime": 383.1464,
+      "eval_samples_per_second": 315.039,
+      "eval_steps_per_second": 3.283,
+      "step": 180000
+    },
+    {
+      "epoch": 7.608649833494921,
+      "grad_norm": 1.054659366607666,
+      "learning_rate": 7.075847875487716e-05,
+      "loss": 2.9576,
+      "step": 180500
+    },
+    {
+      "epoch": 7.629726425831471,
+      "grad_norm": 1.052811861038208,
+      "learning_rate": 7.067272649316126e-05,
+      "loss": 2.9591,
+      "step": 181000
+    },
+    {
+      "epoch": 7.650803018168022,
+      "grad_norm": 1.1816107034683228,
+      "learning_rate": 7.058697423144536e-05,
+      "loss": 2.9525,
+      "step": 181500
+    },
+    {
+      "epoch": 7.671879610504574,
+      "grad_norm": 1.1687514781951904,
+      "learning_rate": 7.050122196972946e-05,
+      "loss": 2.9528,
+      "step": 182000
+    },
+    {
+      "epoch": 7.692956202841125,
+      "grad_norm": 1.0765714645385742,
+      "learning_rate": 7.041546970801354e-05,
+      "loss": 2.9559,
+      "step": 182500
+    },
+    {
+      "epoch": 7.714032795177676,
+      "grad_norm": 1.1206670999526978,
+      "learning_rate": 7.032971744629765e-05,
+      "loss": 2.9567,
+      "step": 183000
+    },
+    {
+      "epoch": 7.7351093875142265,
+      "grad_norm": 1.0698055028915405,
+      "learning_rate": 7.024396518458175e-05,
+      "loss": 2.9565,
+      "step": 183500
+    },
+    {
+      "epoch": 7.756185979850778,
+      "grad_norm": 1.0336474180221558,
+      "learning_rate": 7.015821292286584e-05,
+      "loss": 2.9528,
+      "step": 184000
+    },
+    {
+      "epoch": 7.777262572187329,
+      "grad_norm": 1.1118927001953125,
+      "learning_rate": 7.007246066114993e-05,
+      "loss": 2.9521,
+      "step": 184500
+    },
+    {
+      "epoch": 7.798339164523879,
+      "grad_norm": 1.1753329038619995,
+      "learning_rate": 6.998670839943405e-05,
+      "loss": 2.9512,
+      "step": 185000
+    },
+    {
+      "epoch": 7.8194157568604306,
+      "grad_norm": 1.0788778066635132,
+      "learning_rate": 6.990095613771813e-05,
+      "loss": 2.9482,
+      "step": 185500
+    },
+    {
+      "epoch": 7.840492349196982,
+      "grad_norm": 1.1203080415725708,
+      "learning_rate": 6.981520387600223e-05,
+      "loss": 2.9475,
+      "step": 186000
+    },
+    {
+      "epoch": 7.861568941533533,
+      "grad_norm": 1.0537283420562744,
+      "learning_rate": 6.972945161428633e-05,
+      "loss": 2.9487,
+      "step": 186500
+    },
+    {
+      "epoch": 7.882645533870084,
+      "grad_norm": 1.0483893156051636,
+      "learning_rate": 6.964369935257043e-05,
+      "loss": 2.9476,
+      "step": 187000
+    },
+    {
+      "epoch": 7.903722126206635,
+      "grad_norm": 1.0506216287612915,
+      "learning_rate": 6.955794709085452e-05,
+      "loss": 2.9475,
+      "step": 187500
+    },
+    {
+      "epoch": 7.924798718543186,
+      "grad_norm": 1.1045602560043335,
+      "learning_rate": 6.947219482913862e-05,
+      "loss": 2.9366,
+      "step": 188000
+    },
+    {
+      "epoch": 7.945875310879737,
+      "grad_norm": 1.2198230028152466,
+      "learning_rate": 6.938644256742272e-05,
+      "loss": 2.9457,
+      "step": 188500
+    },
+    {
+      "epoch": 7.966951903216288,
+      "grad_norm": 1.1009573936462402,
+      "learning_rate": 6.930069030570682e-05,
+      "loss": 2.943,
+      "step": 189000
+    },
+    {
+      "epoch": 7.988028495552839,
+      "grad_norm": 1.1872934103012085,
+      "learning_rate": 6.921493804399092e-05,
+      "loss": 2.9399,
+      "step": 189500
+    },
+    {
+      "epoch": 8.00910508788939,
+      "grad_norm": 1.1348000764846802,
+      "learning_rate": 6.9129185782275e-05,
+      "loss": 2.9411,
+      "step": 190000
+    },
+    {
+      "epoch": 8.00910508788939,
+      "eval_accuracy": 0.7130135736773657,
+      "eval_loss": 2.7033603191375732,
+      "eval_runtime": 360.2947,
+      "eval_samples_per_second": 335.02,
+      "eval_steps_per_second": 3.492,
+      "step": 190000
+    },
+    {
+      "epoch": 8.03018168022594,
+      "grad_norm": 1.0649981498718262,
+      "learning_rate": 6.904343352055911e-05,
+      "loss": 2.9354,
+      "step": 190500
+    },
+    {
+      "epoch": 8.051258272562492,
+      "grad_norm": 1.091935157775879,
+      "learning_rate": 6.895768125884321e-05,
+      "loss": 2.9368,
+      "step": 191000
+    },
+    {
+      "epoch": 8.072334864899043,
+      "grad_norm": 1.2174708843231201,
+      "learning_rate": 6.88719289971273e-05,
+      "loss": 2.9356,
+      "step": 191500
+    },
+    {
+      "epoch": 8.093411457235595,
+      "grad_norm": 1.0804139375686646,
+      "learning_rate": 6.87861767354114e-05,
+      "loss": 2.9297,
+      "step": 192000
+    },
+    {
+      "epoch": 8.114488049572145,
+      "grad_norm": 1.1897525787353516,
+      "learning_rate": 6.870042447369551e-05,
+      "loss": 2.9356,
+      "step": 192500
+    },
+    {
+      "epoch": 8.135564641908696,
+      "grad_norm": 1.0706359148025513,
+      "learning_rate": 6.861467221197959e-05,
+      "loss": 2.9354,
+      "step": 193000
+    },
+    {
+      "epoch": 8.156641234245248,
+      "grad_norm": 1.2781833410263062,
+      "learning_rate": 6.852891995026369e-05,
+      "loss": 2.9353,
+      "step": 193500
+    },
+    {
+      "epoch": 8.177717826581798,
+      "grad_norm": 1.0859172344207764,
+      "learning_rate": 6.844316768854779e-05,
+      "loss": 2.9347,
+      "step": 194000
+    },
+    {
+      "epoch": 8.19879441891835,
+      "grad_norm": 1.0777952671051025,
+      "learning_rate": 6.835741542683189e-05,
+      "loss": 2.9322,
+      "step": 194500
+    },
+    {
+      "epoch": 8.2198710112549,
+      "grad_norm": 1.1690711975097656,
+      "learning_rate": 6.827166316511599e-05,
+      "loss": 2.9273,
+      "step": 195000
+    },
+    {
+      "epoch": 8.24094760359145,
+      "grad_norm": 1.1113920211791992,
+      "learning_rate": 6.818591090340008e-05,
+      "loss": 2.931,
+      "step": 195500
+    },
+    {
+      "epoch": 8.262024195928003,
+      "grad_norm": 1.1240168809890747,
+      "learning_rate": 6.810015864168417e-05,
+      "loss": 2.9259,
+      "step": 196000
+    },
+    {
+      "epoch": 8.283100788264553,
+      "grad_norm": 1.092856764793396,
+      "learning_rate": 6.801440637996828e-05,
+      "loss": 2.9267,
+      "step": 196500
+    },
+    {
+      "epoch": 8.304177380601104,
+      "grad_norm": 1.1313588619232178,
+      "learning_rate": 6.792865411825238e-05,
+      "loss": 2.9293,
+      "step": 197000
+    },
+    {
+      "epoch": 8.325253972937656,
+      "grad_norm": 1.0551503896713257,
+      "learning_rate": 6.784290185653646e-05,
+      "loss": 2.9257,
+      "step": 197500
+    },
+    {
+      "epoch": 8.346330565274206,
+      "grad_norm": 1.3216464519500732,
+      "learning_rate": 6.775714959482056e-05,
+      "loss": 2.924,
+      "step": 198000
+    },
+    {
+      "epoch": 8.367407157610758,
+      "grad_norm": 1.1672749519348145,
+      "learning_rate": 6.767139733310467e-05,
+      "loss": 2.9266,
+      "step": 198500
+    },
+    {
+      "epoch": 8.388483749947309,
+      "grad_norm": 1.090099811553955,
+      "learning_rate": 6.758564507138876e-05,
+      "loss": 2.9282,
+      "step": 199000
+    },
+    {
+      "epoch": 8.409560342283859,
+      "grad_norm": 1.0704500675201416,
+      "learning_rate": 6.749989280967286e-05,
+      "loss": 2.9269,
+      "step": 199500
+    },
+    {
+      "epoch": 8.430636934620411,
+      "grad_norm": 1.1324026584625244,
+      "learning_rate": 6.741414054795696e-05,
+      "loss": 2.923,
+      "step": 200000
+    },
+    {
+      "epoch": 8.430636934620411,
+      "eval_accuracy": 0.7158247161749128,
+      "eval_loss": 2.6897358894348145,
+      "eval_runtime": 381.2392,
+      "eval_samples_per_second": 316.615,
+      "eval_steps_per_second": 3.3,
+      "step": 200000
+    },
+    {
+      "epoch": 8.451713526956961,
+      "grad_norm": 1.1459347009658813,
+      "learning_rate": 6.732838828624105e-05,
+      "loss": 2.9219,
+      "step": 200500
+    },
+    {
+      "epoch": 8.472790119293512,
+      "grad_norm": 1.1052474975585938,
+      "learning_rate": 6.724263602452515e-05,
+      "loss": 2.9237,
+      "step": 201000
+    },
+    {
+      "epoch": 8.493866711630064,
+      "grad_norm": 1.1957297325134277,
+      "learning_rate": 6.715688376280924e-05,
+      "loss": 2.9207,
+      "step": 201500
+    },
+    {
+      "epoch": 8.514943303966614,
+      "grad_norm": 1.1772258281707764,
+      "learning_rate": 6.707113150109335e-05,
+      "loss": 2.9193,
+      "step": 202000
+    },
+    {
+      "epoch": 8.536019896303166,
+      "grad_norm": 1.1550202369689941,
+      "learning_rate": 6.698537923937745e-05,
+      "loss": 2.9223,
+      "step": 202500
+    },
+    {
+      "epoch": 8.557096488639717,
+      "grad_norm": 1.0653135776519775,
+      "learning_rate": 6.689962697766153e-05,
+      "loss": 2.919,
+      "step": 203000
+    },
+    {
+      "epoch": 8.578173080976267,
+      "grad_norm": 1.1351792812347412,
+      "learning_rate": 6.681387471594563e-05,
+      "loss": 2.9181,
+      "step": 203500
+    },
+    {
+      "epoch": 8.59924967331282,
+      "grad_norm": 1.2342908382415771,
+      "learning_rate": 6.672812245422974e-05,
+      "loss": 2.9161,
+      "step": 204000
+    },
+    {
+      "epoch": 8.62032626564937,
+      "grad_norm": 1.1976215839385986,
+      "learning_rate": 6.664237019251383e-05,
+      "loss": 2.9179,
+      "step": 204500
+    },
+    {
+      "epoch": 8.64140285798592,
+      "grad_norm": 1.1582359075546265,
+      "learning_rate": 6.655661793079792e-05,
+      "loss": 2.9157,
+      "step": 205000
+    },
+    {
+      "epoch": 8.662479450322472,
+      "grad_norm": 1.1149996519088745,
+      "learning_rate": 6.647086566908202e-05,
+      "loss": 2.9112,
+      "step": 205500
+    },
+    {
+      "epoch": 8.683556042659022,
+      "grad_norm": 1.098393440246582,
+      "learning_rate": 6.638511340736612e-05,
+      "loss": 2.9126,
+      "step": 206000
+    },
+    {
+      "epoch": 8.704632634995574,
+      "grad_norm": 1.0889263153076172,
+      "learning_rate": 6.629936114565022e-05,
+      "loss": 2.9103,
+      "step": 206500
+    },
+    {
+      "epoch": 8.725709227332125,
+      "grad_norm": 1.0951812267303467,
+      "learning_rate": 6.621360888393432e-05,
+      "loss": 2.9127,
+      "step": 207000
+    },
+    {
+      "epoch": 8.746785819668675,
+      "grad_norm": 1.0896576642990112,
+      "learning_rate": 6.61278566222184e-05,
+      "loss": 2.9041,
+      "step": 207500
+    },
+    {
+      "epoch": 8.767862412005227,
+      "grad_norm": 1.1914458274841309,
+      "learning_rate": 6.604210436050251e-05,
+      "loss": 2.9108,
+      "step": 208000
+    },
+    {
+      "epoch": 8.788939004341778,
+      "grad_norm": 1.3060169219970703,
+      "learning_rate": 6.595635209878661e-05,
+      "loss": 2.91,
+      "step": 208500
+    },
+    {
+      "epoch": 8.81001559667833,
+      "grad_norm": 1.1106061935424805,
+      "learning_rate": 6.58705998370707e-05,
+      "loss": 2.9108,
+      "step": 209000
+    },
+    {
+      "epoch": 8.83109218901488,
+      "grad_norm": 1.074009895324707,
+      "learning_rate": 6.57848475753548e-05,
+      "loss": 2.911,
+      "step": 209500
+    },
+    {
+      "epoch": 8.85216878135143,
+      "grad_norm": 1.122781753540039,
+      "learning_rate": 6.569909531363891e-05,
+      "loss": 2.9048,
+      "step": 210000
+    },
+    {
+      "epoch": 8.85216878135143,
+      "eval_accuracy": 0.7181487245498546,
+      "eval_loss": 2.67653489112854,
+      "eval_runtime": 354.4662,
+      "eval_samples_per_second": 340.529,
+      "eval_steps_per_second": 3.549,
+      "step": 210000
+    },
+    {
+      "epoch": 8.873245373687983,
+      "grad_norm": 1.0612046718597412,
+      "learning_rate": 6.561334305192299e-05,
+      "loss": 2.9079,
+      "step": 210500
+    },
+    {
+      "epoch": 8.894321966024533,
+      "grad_norm": 1.0794860124588013,
+      "learning_rate": 6.552759079020709e-05,
+      "loss": 2.9091,
+      "step": 211000
+    },
+    {
+      "epoch": 8.915398558361083,
+      "grad_norm": 1.1518737077713013,
+      "learning_rate": 6.544183852849119e-05,
+      "loss": 2.9053,
+      "step": 211500
+    },
+    {
+      "epoch": 8.936475150697635,
+      "grad_norm": 1.119808316230774,
+      "learning_rate": 6.535608626677529e-05,
+      "loss": 2.9095,
+      "step": 212000
+    },
+    {
+      "epoch": 8.957551743034186,
+      "grad_norm": 1.1451191902160645,
+      "learning_rate": 6.527033400505939e-05,
+      "loss": 2.9081,
+      "step": 212500
+    },
+    {
+      "epoch": 8.978628335370738,
+      "grad_norm": 1.1871023178100586,
+      "learning_rate": 6.518458174334348e-05,
+      "loss": 2.9086,
+      "step": 213000
+    },
+    {
+      "epoch": 8.999704927707288,
+      "grad_norm": 1.179527759552002,
+      "learning_rate": 6.509882948162758e-05,
+      "loss": 2.9057,
+      "step": 213500
+    },
+    {
+      "epoch": 9.020781520043839,
+      "grad_norm": 1.1570216417312622,
+      "learning_rate": 6.501307721991168e-05,
+      "loss": 2.8998,
+      "step": 214000
+    },
+    {
+      "epoch": 9.04185811238039,
+      "grad_norm": 1.1221469640731812,
+      "learning_rate": 6.492732495819578e-05,
+      "loss": 2.8984,
+      "step": 214500
+    },
+    {
+      "epoch": 9.062934704716941,
+      "grad_norm": 1.2023974657058716,
+      "learning_rate": 6.484157269647986e-05,
+      "loss": 2.8989,
+      "step": 215000
+    },
+    {
+      "epoch": 9.084011297053493,
+      "grad_norm": 1.0986675024032593,
+      "learning_rate": 6.475582043476398e-05,
+      "loss": 2.8989,
+      "step": 215500
+    },
+    {
+      "epoch": 9.105087889390044,
+      "grad_norm": 1.0840119123458862,
+      "learning_rate": 6.467006817304807e-05,
+      "loss": 2.8967,
+      "step": 216000
+    },
+    {
+      "epoch": 9.126164481726594,
+      "grad_norm": 1.1448246240615845,
+      "learning_rate": 6.458431591133216e-05,
+      "loss": 2.8928,
+      "step": 216500
+    },
+    {
+      "epoch": 9.147241074063146,
+      "grad_norm": 1.1093271970748901,
+      "learning_rate": 6.449856364961626e-05,
+      "loss": 2.8957,
+      "step": 217000
+    },
+    {
+      "epoch": 9.168317666399696,
+      "grad_norm": 1.096163034439087,
+      "learning_rate": 6.441281138790037e-05,
+      "loss": 2.893,
+      "step": 217500
+    },
+    {
+      "epoch": 9.189394258736247,
+      "grad_norm": 1.085171103477478,
+      "learning_rate": 6.432705912618445e-05,
+      "loss": 2.8946,
+      "step": 218000
+    },
+    {
+      "epoch": 9.210470851072799,
+      "grad_norm": 1.0827836990356445,
+      "learning_rate": 6.424130686446855e-05,
+      "loss": 2.895,
+      "step": 218500
+    },
+    {
+      "epoch": 9.23154744340935,
+      "grad_norm": 1.1749422550201416,
+      "learning_rate": 6.415555460275265e-05,
+      "loss": 2.8942,
+      "step": 219000
+    },
+    {
+      "epoch": 9.252624035745901,
+      "grad_norm": 1.1791439056396484,
+      "learning_rate": 6.406980234103675e-05,
+      "loss": 2.895,
+      "step": 219500
+    },
+    {
+      "epoch": 9.273700628082452,
+      "grad_norm": 1.1061257123947144,
+      "learning_rate": 6.398405007932085e-05,
+      "loss": 2.8966,
+      "step": 220000
+    },
+    {
+      "epoch": 9.273700628082452,
+      "eval_accuracy": 0.7204496954231441,
+      "eval_loss": 2.662158250808716,
+      "eval_runtime": 572.3077,
+      "eval_samples_per_second": 210.911,
+      "eval_steps_per_second": 2.198,
+      "step": 220000
+    },
+    {
+      "epoch": 9.294777220419002,
+      "grad_norm": 1.0660297870635986,
+      "learning_rate": 6.389829781760495e-05,
+      "loss": 2.8904,
+      "step": 220500
+    },
+    {
+      "epoch": 9.315853812755554,
+      "grad_norm": 1.1441725492477417,
+      "learning_rate": 6.381254555588903e-05,
+      "loss": 2.8942,
+      "step": 221000
+    },
+    {
+      "epoch": 9.336930405092104,
+      "grad_norm": 1.0496788024902344,
+      "learning_rate": 6.372679329417314e-05,
+      "loss": 2.8886,
+      "step": 221500
+    },
+    {
+      "epoch": 9.358006997428657,
+      "grad_norm": 1.1793948411941528,
+      "learning_rate": 6.364104103245724e-05,
+      "loss": 2.8876,
+      "step": 222000
+    },
+    {
+      "epoch": 9.379083589765207,
+      "grad_norm": 1.124091625213623,
+      "learning_rate": 6.355528877074132e-05,
+      "loss": 2.8916,
+      "step": 222500
+    },
+    {
+      "epoch": 9.400160182101757,
+      "grad_norm": 1.1525397300720215,
+      "learning_rate": 6.346953650902542e-05,
+      "loss": 2.8879,
+      "step": 223000
+    },
+    {
+      "epoch": 9.42123677443831,
+      "grad_norm": 1.131447434425354,
+      "learning_rate": 6.338378424730954e-05,
+      "loss": 2.8896,
+      "step": 223500
+    },
+    {
+      "epoch": 9.44231336677486,
+      "grad_norm": 1.1651755571365356,
+      "learning_rate": 6.329803198559362e-05,
+      "loss": 2.8862,
+      "step": 224000
+    },
+    {
+      "epoch": 9.46338995911141,
+      "grad_norm": 1.1790530681610107,
+      "learning_rate": 6.321227972387772e-05,
+      "loss": 2.8872,
+      "step": 224500
+    },
+    {
+      "epoch": 9.484466551447962,
+      "grad_norm": 1.1216908693313599,
+      "learning_rate": 6.312652746216182e-05,
+      "loss": 2.8829,
+      "step": 225000
+    },
+    {
+      "epoch": 9.505543143784513,
+      "grad_norm": 1.0767239332199097,
+      "learning_rate": 6.304077520044591e-05,
+      "loss": 2.89,
+      "step": 225500
+    },
+    {
+      "epoch": 9.526619736121065,
+      "grad_norm": 1.1002498865127563,
+      "learning_rate": 6.295502293873001e-05,
+      "loss": 2.8837,
+      "step": 226000
+    },
+    {
+      "epoch": 9.547696328457615,
+      "grad_norm": 1.1905837059020996,
+      "learning_rate": 6.286927067701411e-05,
+      "loss": 2.8823,
+      "step": 226500
+    },
+    {
+      "epoch": 9.568772920794165,
+      "grad_norm": 1.2649986743927002,
+      "learning_rate": 6.278351841529821e-05,
+      "loss": 2.8864,
+      "step": 227000
+    },
+    {
+      "epoch": 9.589849513130718,
+      "grad_norm": 1.0912470817565918,
+      "learning_rate": 6.269776615358231e-05,
+      "loss": 2.8834,
+      "step": 227500
+    },
+    {
+      "epoch": 9.610926105467268,
+      "grad_norm": 1.155846118927002,
+      "learning_rate": 6.26120138918664e-05,
+      "loss": 2.8824,
+      "step": 228000
+    },
+    {
+      "epoch": 9.63200269780382,
+      "grad_norm": 1.0676798820495605,
+      "learning_rate": 6.252626163015049e-05,
+      "loss": 2.8807,
+      "step": 228500
+    },
+    {
+      "epoch": 9.65307929014037,
+      "grad_norm": 1.2211025953292847,
+      "learning_rate": 6.24405093684346e-05,
+      "loss": 2.8796,
+      "step": 229000
+    },
+    {
+      "epoch": 9.67415588247692,
+      "grad_norm": 1.1760472059249878,
+      "learning_rate": 6.23547571067187e-05,
+      "loss": 2.879,
+      "step": 229500
+    },
+    {
+      "epoch": 9.695232474813473,
+      "grad_norm": 1.1819926500320435,
+      "learning_rate": 6.226900484500279e-05,
+      "loss": 2.8762,
+      "step": 230000
+    },
+    {
+      "epoch": 9.695232474813473,
+      "eval_accuracy": 0.7223368868496577,
+      "eval_loss": 2.6504571437835693,
+      "eval_runtime": 408.4744,
+      "eval_samples_per_second": 295.504,
+      "eval_steps_per_second": 3.08,
+      "step": 230000
+    },
+    {
+      "epoch": 9.716309067150023,
+      "grad_norm": 1.1507211923599243,
+      "learning_rate": 6.218325258328688e-05,
+      "loss": 2.8835,
+      "step": 230500
+    },
+    {
+      "epoch": 9.737385659486574,
+      "grad_norm": 1.0716516971588135,
+      "learning_rate": 6.209750032157098e-05,
+      "loss": 2.8803,
+      "step": 231000
+    },
+    {
+      "epoch": 9.758462251823126,
+      "grad_norm": 1.1526685953140259,
+      "learning_rate": 6.201174805985508e-05,
+      "loss": 2.8765,
+      "step": 231500
+    },
+    {
+      "epoch": 9.779538844159676,
+      "grad_norm": 1.1408942937850952,
+      "learning_rate": 6.192599579813918e-05,
+      "loss": 2.877,
+      "step": 232000
+    },
+    {
+      "epoch": 9.800615436496226,
+      "grad_norm": 1.0803561210632324,
+      "learning_rate": 6.184024353642328e-05,
+      "loss": 2.8739,
+      "step": 232500
+    },
+    {
+      "epoch": 9.821692028832778,
+      "grad_norm": 1.1361820697784424,
+      "learning_rate": 6.175449127470738e-05,
+      "loss": 2.8733,
+      "step": 233000
+    },
+    {
+      "epoch": 9.842768621169329,
+      "grad_norm": 1.1026684045791626,
+      "learning_rate": 6.166873901299147e-05,
+      "loss": 2.879,
+      "step": 233500
+    },
+    {
+      "epoch": 9.863845213505881,
+      "grad_norm": 1.0774463415145874,
+      "learning_rate": 6.158298675127557e-05,
+      "loss": 2.8733,
+      "step": 234000
+    },
+    {
+      "epoch": 9.884921805842431,
+      "grad_norm": 1.2058387994766235,
+      "learning_rate": 6.149723448955966e-05,
+      "loss": 2.8757,
+      "step": 234500
+    },
+    {
+      "epoch": 9.905998398178982,
+      "grad_norm": 1.1163556575775146,
+      "learning_rate": 6.141148222784377e-05,
+      "loss": 2.8701,
+      "step": 235000
+    },
+    {
+      "epoch": 9.927074990515534,
+      "grad_norm": 1.202458381652832,
+      "learning_rate": 6.132572996612787e-05,
+      "loss": 2.8708,
+      "step": 235500
+    },
+    {
+      "epoch": 9.948151582852084,
+      "grad_norm": 1.1671936511993408,
+      "learning_rate": 6.123997770441195e-05,
+      "loss": 2.8757,
+      "step": 236000
+    },
+    {
+      "epoch": 9.969228175188636,
+      "grad_norm": 1.215464472770691,
+      "learning_rate": 6.115422544269605e-05,
+      "loss": 2.8755,
+      "step": 236500
+    },
+    {
+      "epoch": 9.990304767525187,
+      "grad_norm": 1.214748740196228,
+      "learning_rate": 6.106847318098015e-05,
+      "loss": 2.8717,
+      "step": 237000
+    },
+    {
+      "epoch": 10.011381359861737,
+      "grad_norm": 1.0600272417068481,
+      "learning_rate": 6.098272091926425e-05,
+      "loss": 2.871,
+      "step": 237500
+    },
+    {
+      "epoch": 10.032457952198289,
+      "grad_norm": 1.0806080102920532,
+      "learning_rate": 6.0896968657548345e-05,
+      "loss": 2.8685,
+      "step": 238000
+    },
+    {
+      "epoch": 10.05353454453484,
+      "grad_norm": 1.1183685064315796,
+      "learning_rate": 6.081121639583244e-05,
+      "loss": 2.8662,
+      "step": 238500
+    },
+    {
+      "epoch": 10.07461113687139,
+      "grad_norm": 1.1194103956222534,
+      "learning_rate": 6.072546413411654e-05,
+      "loss": 2.867,
+      "step": 239000
+    },
+    {
+      "epoch": 10.095687729207942,
+      "grad_norm": 1.1299004554748535,
+      "learning_rate": 6.063971187240064e-05,
+      "loss": 2.8655,
+      "step": 239500
+    },
+    {
+      "epoch": 10.116764321544492,
+      "grad_norm": 1.1387616395950317,
+      "learning_rate": 6.055395961068473e-05,
+      "loss": 2.8607,
+      "step": 240000
+    },
+    {
+      "epoch": 10.116764321544492,
+      "eval_accuracy": 0.724530309311634,
+      "eval_loss": 2.639788866043091,
+      "eval_runtime": 503.3995,
+      "eval_samples_per_second": 239.782,
+      "eval_steps_per_second": 2.499,
+      "step": 240000
+    },
+    {
+      "epoch": 10.137840913881044,
+      "grad_norm": 1.1641356945037842,
+      "learning_rate": 6.046820734896883e-05,
+      "loss": 2.8638,
+      "step": 240500
+    },
+    {
+      "epoch": 10.158917506217595,
+      "grad_norm": 1.1925791501998901,
+      "learning_rate": 6.038245508725293e-05,
+      "loss": 2.865,
+      "step": 241000
+    },
+    {
+      "epoch": 10.179994098554145,
+      "grad_norm": 1.1233131885528564,
+      "learning_rate": 6.029670282553702e-05,
+      "loss": 2.863,
+      "step": 241500
+    },
+    {
+      "epoch": 10.201070690890697,
+      "grad_norm": 1.1123764514923096,
+      "learning_rate": 6.0210950563821125e-05,
+      "loss": 2.8664,
+      "step": 242000
+    },
+    {
+      "epoch": 10.222147283227248,
+      "grad_norm": 1.3262293338775635,
+      "learning_rate": 6.0125198302105223e-05,
+      "loss": 2.8635,
+      "step": 242500
+    },
+    {
+      "epoch": 10.2432238755638,
+      "grad_norm": 1.1125178337097168,
+      "learning_rate": 6.0039446040389315e-05,
+      "loss": 2.8619,
+      "step": 243000
+    },
+    {
+      "epoch": 10.26430046790035,
+      "grad_norm": 1.196913719177246,
+      "learning_rate": 5.995369377867341e-05,
+      "loss": 2.8664,
+      "step": 243500
+    },
+    {
+      "epoch": 10.2853770602369,
+      "grad_norm": 1.0928704738616943,
+      "learning_rate": 5.986794151695752e-05,
+      "loss": 2.857,
+      "step": 244000
+    },
+    {
+      "epoch": 10.306453652573452,
+      "grad_norm": 1.2177590131759644,
+      "learning_rate": 5.97821892552416e-05,
+      "loss": 2.8645,
+      "step": 244500
+    },
+    {
+      "epoch": 10.327530244910003,
+      "grad_norm": 1.1060130596160889,
+      "learning_rate": 5.969643699352571e-05,
+      "loss": 2.863,
+      "step": 245000
+    },
+    {
+      "epoch": 10.348606837246553,
+      "grad_norm": 1.109161138534546,
+      "learning_rate": 5.9610684731809806e-05,
+      "loss": 2.8618,
+      "step": 245500
+    },
+    {
+      "epoch": 10.369683429583105,
+      "grad_norm": 1.0750991106033325,
+      "learning_rate": 5.95249324700939e-05,
+      "loss": 2.86,
+      "step": 246000
+    },
+    {
+      "epoch": 10.390760021919656,
+      "grad_norm": 1.1675221920013428,
+      "learning_rate": 5.9439180208377996e-05,
+      "loss": 2.8613,
+      "step": 246500
+    },
+    {
+      "epoch": 10.411836614256208,
+      "grad_norm": 1.1404515504837036,
+      "learning_rate": 5.93534279466621e-05,
+      "loss": 2.8578,
+      "step": 247000
+    },
+    {
+      "epoch": 10.432913206592758,
+      "grad_norm": 1.090852975845337,
+      "learning_rate": 5.9267675684946186e-05,
+      "loss": 2.8615,
+      "step": 247500
+    },
+    {
+      "epoch": 10.453989798929308,
+      "grad_norm": 1.180174469947815,
+      "learning_rate": 5.918192342323029e-05,
+      "loss": 2.8539,
+      "step": 248000
+    },
+    {
+      "epoch": 10.47506639126586,
+      "grad_norm": 1.0894999504089355,
+      "learning_rate": 5.909617116151439e-05,
+      "loss": 2.8567,
+      "step": 248500
+    },
+    {
+      "epoch": 10.496142983602411,
+      "grad_norm": 1.143641710281372,
+      "learning_rate": 5.901041889979848e-05,
+      "loss": 2.8534,
+      "step": 249000
+    },
+    {
+      "epoch": 10.517219575938963,
+      "grad_norm": 1.0824109315872192,
+      "learning_rate": 5.892466663808258e-05,
+      "loss": 2.8586,
+      "step": 249500
+    },
+    {
+      "epoch": 10.538296168275513,
+      "grad_norm": 1.0723791122436523,
+      "learning_rate": 5.8838914376366684e-05,
+      "loss": 2.8571,
+      "step": 250000
+    },
+    {
+      "epoch": 10.538296168275513,
+      "eval_accuracy": 0.7262084430068972,
+      "eval_loss": 2.6271183490753174,
+      "eval_runtime": 386.5161,
+      "eval_samples_per_second": 312.292,
+      "eval_steps_per_second": 3.255,
+      "step": 250000
+    },
+    {
+      "epoch": 10.559372760612064,
+      "grad_norm": 1.1291708946228027,
+      "learning_rate": 5.8753162114650776e-05,
+      "loss": 2.8594,
+      "step": 250500
+    },
+    {
+      "epoch": 10.580449352948616,
+      "grad_norm": 1.204417109489441,
+      "learning_rate": 5.8667409852934874e-05,
+      "loss": 2.854,
+      "step": 251000
+    },
+    {
+      "epoch": 10.601525945285166,
+      "grad_norm": 1.1110678911209106,
+      "learning_rate": 5.858165759121897e-05,
+      "loss": 2.8495,
+      "step": 251500
+    },
+    {
+      "epoch": 10.622602537621717,
+      "grad_norm": 1.077514410018921,
+      "learning_rate": 5.8495905329503064e-05,
+      "loss": 2.8538,
+      "step": 252000
+    },
+    {
+      "epoch": 10.643679129958269,
+      "grad_norm": 1.087298035621643,
+      "learning_rate": 5.841015306778716e-05,
+      "loss": 2.8532,
+      "step": 252500
+    },
+    {
+      "epoch": 10.664755722294819,
+      "grad_norm": 1.135702133178711,
+      "learning_rate": 5.832440080607127e-05,
+      "loss": 2.8538,
+      "step": 253000
+    },
+    {
+      "epoch": 10.685832314631371,
+      "grad_norm": 1.0863332748413086,
+      "learning_rate": 5.823864854435536e-05,
+      "loss": 2.8479,
+      "step": 253500
+    },
+    {
+      "epoch": 10.706908906967922,
+      "grad_norm": 1.1756316423416138,
+      "learning_rate": 5.815289628263946e-05,
+      "loss": 2.8537,
+      "step": 254000
+    },
+    {
+      "epoch": 10.727985499304472,
+      "grad_norm": 1.1488475799560547,
+      "learning_rate": 5.8067144020923556e-05,
+      "loss": 2.8553,
+      "step": 254500
+    },
+    {
+      "epoch": 10.749062091641024,
+      "grad_norm": 1.1579923629760742,
+      "learning_rate": 5.798139175920765e-05,
+      "loss": 2.8499,
+      "step": 255000
+    },
+    {
+      "epoch": 10.770138683977574,
+      "grad_norm": 1.1076788902282715,
+      "learning_rate": 5.789563949749175e-05,
+      "loss": 2.8515,
+      "step": 255500
+    },
+    {
+      "epoch": 10.791215276314126,
+      "grad_norm": 1.2158606052398682,
+      "learning_rate": 5.780988723577585e-05,
+      "loss": 2.8496,
+      "step": 256000
+    },
+    {
+      "epoch": 10.812291868650677,
+      "grad_norm": 1.126074194908142,
+      "learning_rate": 5.772413497405994e-05,
+      "loss": 2.8505,
+      "step": 256500
+    },
+    {
+      "epoch": 10.833368460987227,
+      "grad_norm": 1.1246585845947266,
+      "learning_rate": 5.763838271234404e-05,
+      "loss": 2.8457,
+      "step": 257000
+    },
+    {
+      "epoch": 10.85444505332378,
+      "grad_norm": 1.0921372175216675,
+      "learning_rate": 5.7552630450628146e-05,
+      "loss": 2.8491,
+      "step": 257500
+    },
+    {
+      "epoch": 10.87552164566033,
+      "grad_norm": 1.1925691366195679,
+      "learning_rate": 5.746687818891223e-05,
+      "loss": 2.8468,
+      "step": 258000
+    },
+    {
+      "epoch": 10.89659823799688,
+      "grad_norm": 1.1455470323562622,
+      "learning_rate": 5.7381125927196335e-05,
+      "loss": 2.8472,
+      "step": 258500
+    },
+    {
+      "epoch": 10.917674830333432,
+      "grad_norm": 1.1722052097320557,
+      "learning_rate": 5.7295373665480434e-05,
+      "loss": 2.8472,
+      "step": 259000
+    },
+    {
+      "epoch": 10.938751422669982,
+      "grad_norm": 1.0971215963363647,
+      "learning_rate": 5.7209621403764525e-05,
+      "loss": 2.8467,
+      "step": 259500
+    },
+    {
+      "epoch": 10.959828015006533,
+      "grad_norm": 1.177807331085205,
+      "learning_rate": 5.7123869142048624e-05,
+      "loss": 2.8425,
+      "step": 260000
+    },
+    {
+      "epoch": 10.959828015006533,
+      "eval_accuracy": 0.727975746170522,
+      "eval_loss": 2.617532968521118,
+      "eval_runtime": 366.6846,
+      "eval_samples_per_second": 329.182,
+      "eval_steps_per_second": 3.431,
+      "step": 260000
+    },
+    {
+      "epoch": 10.980904607343085,
+      "grad_norm": 1.1334214210510254,
+      "learning_rate": 5.703811688033273e-05,
+      "loss": 2.8416,
+      "step": 260500
+    },
+    {
+      "epoch": 11.001981199679635,
+      "grad_norm": 1.140517234802246,
+      "learning_rate": 5.695236461861681e-05,
+      "loss": 2.8437,
+      "step": 261000
+    },
+    {
+      "epoch": 11.023057792016187,
+      "grad_norm": 1.1567606925964355,
+      "learning_rate": 5.686661235690092e-05,
+      "loss": 2.8367,
+      "step": 261500
+    },
+    {
+      "epoch": 11.044134384352738,
+      "grad_norm": 1.107995867729187,
+      "learning_rate": 5.678086009518502e-05,
+      "loss": 2.8443,
+      "step": 262000
+    },
+    {
+      "epoch": 11.065210976689288,
+      "grad_norm": 1.2605202198028564,
+      "learning_rate": 5.669510783346911e-05,
+      "loss": 2.8449,
+      "step": 262500
+    },
+    {
+      "epoch": 11.08628756902584,
+      "grad_norm": 1.1701496839523315,
+      "learning_rate": 5.6609355571753207e-05,
+      "loss": 2.8391,
+      "step": 263000
+    },
+    {
+      "epoch": 11.10736416136239,
+      "grad_norm": 1.17499840259552,
+      "learning_rate": 5.652360331003731e-05,
+      "loss": 2.8412,
+      "step": 263500
+    },
+    {
+      "epoch": 11.128440753698943,
+      "grad_norm": 1.272225260734558,
+      "learning_rate": 5.6437851048321396e-05,
+      "loss": 2.8366,
+      "step": 264000
+    },
+    {
+      "epoch": 11.149517346035493,
+      "grad_norm": 1.1453450918197632,
+      "learning_rate": 5.63520987866055e-05,
+      "loss": 2.8414,
+      "step": 264500
+    },
+    {
+      "epoch": 11.170593938372043,
+      "grad_norm": 1.1126829385757446,
+      "learning_rate": 5.62663465248896e-05,
+      "loss": 2.8391,
+      "step": 265000
+    },
+    {
+      "epoch": 11.191670530708596,
+      "grad_norm": 1.138759732246399,
+      "learning_rate": 5.618059426317369e-05,
+      "loss": 2.8347,
+      "step": 265500
+    },
+    {
+      "epoch": 11.212747123045146,
+      "grad_norm": 1.170114278793335,
+      "learning_rate": 5.609484200145779e-05,
+      "loss": 2.8394,
+      "step": 266000
+    },
+    {
+      "epoch": 11.233823715381696,
+      "grad_norm": 1.1583267450332642,
+      "learning_rate": 5.6009089739741895e-05,
+      "loss": 2.8392,
+      "step": 266500
+    },
+    {
+      "epoch": 11.254900307718248,
+      "grad_norm": 1.1162408590316772,
+      "learning_rate": 5.5923337478025986e-05,
+      "loss": 2.8347,
+      "step": 267000
+    },
+    {
+      "epoch": 11.275976900054799,
+      "grad_norm": 1.1335992813110352,
+      "learning_rate": 5.5837585216310085e-05,
+      "loss": 2.8395,
+      "step": 267500
+    },
+    {
+      "epoch": 11.29705349239135,
+      "grad_norm": 1.2482997179031372,
+      "learning_rate": 5.575183295459418e-05,
+      "loss": 2.8331,
+      "step": 268000
+    },
+    {
+      "epoch": 11.318130084727901,
+      "grad_norm": 1.143761157989502,
+      "learning_rate": 5.5666080692878274e-05,
+      "loss": 2.8376,
+      "step": 268500
+    },
+    {
+      "epoch": 11.339206677064452,
+      "grad_norm": 1.12701416015625,
+      "learning_rate": 5.558032843116238e-05,
+      "loss": 2.8364,
+      "step": 269000
+    },
+    {
+      "epoch": 11.360283269401004,
+      "grad_norm": 1.1339936256408691,
+      "learning_rate": 5.549457616944648e-05,
+      "loss": 2.8372,
+      "step": 269500
+    },
+    {
+      "epoch": 11.381359861737554,
+      "grad_norm": 1.1112282276153564,
+      "learning_rate": 5.540882390773057e-05,
+      "loss": 2.8318,
+      "step": 270000
+    },
+    {
+      "epoch": 11.381359861737554,
+      "eval_accuracy": 0.7292257531459444,
+      "eval_loss": 2.61082124710083,
+      "eval_runtime": 761.9787,
+      "eval_samples_per_second": 158.411,
+      "eval_steps_per_second": 1.651,
+      "step": 270000
+    },
+    {
+      "epoch": 11.402436454074106,
+      "grad_norm": 1.11770761013031,
+      "learning_rate": 5.532307164601467e-05,
+      "loss": 2.8344,
+      "step": 270500
+    },
+    {
+      "epoch": 11.423513046410656,
+      "grad_norm": 1.1740740537643433,
+      "learning_rate": 5.5237319384298766e-05,
+      "loss": 2.8329,
+      "step": 271000
+    },
+    {
+      "epoch": 11.444589638747207,
+      "grad_norm": 1.0773067474365234,
+      "learning_rate": 5.515156712258286e-05,
+      "loss": 2.8292,
+      "step": 271500
+    },
+    {
+      "epoch": 11.465666231083759,
+      "grad_norm": 1.1690468788146973,
+      "learning_rate": 5.506581486086696e-05,
+      "loss": 2.8333,
+      "step": 272000
+    },
+    {
+      "epoch": 11.48674282342031,
+      "grad_norm": 1.2153018712997437,
+      "learning_rate": 5.498006259915105e-05,
+      "loss": 2.8312,
+      "step": 272500
+    },
+    {
+      "epoch": 11.50781941575686,
+      "grad_norm": 1.0754259824752808,
+      "learning_rate": 5.489431033743515e-05,
+      "loss": 2.8309,
+      "step": 273000
+    },
+    {
+      "epoch": 11.528896008093412,
+      "grad_norm": 1.1420719623565674,
+      "learning_rate": 5.480855807571925e-05,
+      "loss": 2.8334,
+      "step": 273500
+    },
+    {
+      "epoch": 11.549972600429962,
+      "grad_norm": 1.065546989440918,
+      "learning_rate": 5.472280581400334e-05,
+      "loss": 2.8294,
+      "step": 274000
+    },
+    {
+      "epoch": 11.571049192766514,
+      "grad_norm": 1.209484338760376,
+      "learning_rate": 5.463705355228744e-05,
+      "loss": 2.8293,
+      "step": 274500
+    },
+    {
+      "epoch": 11.592125785103065,
+      "grad_norm": 1.101347804069519,
+      "learning_rate": 5.4551301290571546e-05,
+      "loss": 2.833,
+      "step": 275000
+    },
+    {
+      "epoch": 11.613202377439615,
+      "grad_norm": 1.1171296834945679,
+      "learning_rate": 5.446554902885563e-05,
+      "loss": 2.8302,
+      "step": 275500
+    },
+    {
+      "epoch": 11.634278969776167,
+      "grad_norm": 1.0985829830169678,
+      "learning_rate": 5.4379796767139736e-05,
+      "loss": 2.8267,
+      "step": 276000
+    },
+    {
+      "epoch": 11.655355562112717,
+      "grad_norm": 1.1095987558364868,
+      "learning_rate": 5.4294044505423834e-05,
+      "loss": 2.8273,
+      "step": 276500
+    },
+    {
+      "epoch": 11.67643215444927,
+      "grad_norm": 1.0924415588378906,
+      "learning_rate": 5.4208292243707925e-05,
+      "loss": 2.8286,
+      "step": 277000
+    },
+    {
+      "epoch": 11.69750874678582,
+      "grad_norm": 1.0707364082336426,
+      "learning_rate": 5.4122539981992024e-05,
+      "loss": 2.8279,
+      "step": 277500
+    },
+    {
+      "epoch": 11.71858533912237,
+      "grad_norm": 1.0952256917953491,
+      "learning_rate": 5.403678772027613e-05,
+      "loss": 2.8231,
+      "step": 278000
+    },
+    {
+      "epoch": 11.739661931458922,
+      "grad_norm": 1.115564227104187,
+      "learning_rate": 5.395103545856022e-05,
+      "loss": 2.8266,
+      "step": 278500
+    },
+    {
+      "epoch": 11.760738523795473,
+      "grad_norm": 1.2027214765548706,
+      "learning_rate": 5.386528319684432e-05,
+      "loss": 2.8238,
+      "step": 279000
+    },
+    {
+      "epoch": 11.781815116132023,
+      "grad_norm": 1.214417576789856,
+      "learning_rate": 5.377953093512842e-05,
+      "loss": 2.8216,
+      "step": 279500
+    },
+    {
+      "epoch": 11.802891708468575,
+      "grad_norm": 1.1173959970474243,
+      "learning_rate": 5.369377867341251e-05,
+      "loss": 2.8289,
+      "step": 280000
+    },
+    {
+      "epoch": 11.802891708468575,
+      "eval_accuracy": 0.7310976309084538,
+      "eval_loss": 2.60066294670105,
+      "eval_runtime": 397.4611,
+      "eval_samples_per_second": 303.693,
+      "eval_steps_per_second": 3.165,
+      "step": 280000
+    },
+    {
+      "epoch": 11.823968300805126,
+      "grad_norm": 1.1264653205871582,
+      "learning_rate": 5.3608026411696614e-05,
+      "loss": 2.8299,
+      "step": 280500
+    },
+    {
+      "epoch": 11.845044893141678,
+      "grad_norm": 1.2253224849700928,
+      "learning_rate": 5.352227414998071e-05,
+      "loss": 2.8197,
+      "step": 281000
+    },
+    {
+      "epoch": 11.866121485478228,
+      "grad_norm": 1.2499597072601318,
+      "learning_rate": 5.34365218882648e-05,
+      "loss": 2.8232,
+      "step": 281500
+    },
+    {
+      "epoch": 11.887198077814778,
+      "grad_norm": 1.0643383264541626,
+      "learning_rate": 5.33507696265489e-05,
+      "loss": 2.8168,
+      "step": 282000
+    },
+    {
+      "epoch": 11.90827467015133,
+      "grad_norm": 1.137691617012024,
+      "learning_rate": 5.3265017364833e-05,
+      "loss": 2.8257,
+      "step": 282500
+    },
+    {
+      "epoch": 11.92935126248788,
+      "grad_norm": 1.195610761642456,
+      "learning_rate": 5.317926510311709e-05,
+      "loss": 2.8205,
+      "step": 283000
+    },
+    {
+      "epoch": 11.950427854824433,
+      "grad_norm": 1.1147677898406982,
+      "learning_rate": 5.3093512841401197e-05,
+      "loss": 2.8206,
+      "step": 283500
+    },
+    {
+      "epoch": 11.971504447160983,
+      "grad_norm": 1.128434419631958,
+      "learning_rate": 5.3007760579685295e-05,
+      "loss": 2.8235,
+      "step": 284000
+    },
+    {
+      "epoch": 11.992581039497534,
+      "grad_norm": 1.1825289726257324,
+      "learning_rate": 5.2922008317969386e-05,
+      "loss": 2.8244,
+      "step": 284500
+    },
+    {
+      "epoch": 12.013657631834086,
+      "grad_norm": 1.109700322151184,
+      "learning_rate": 5.2836256056253485e-05,
+      "loss": 2.8196,
+      "step": 285000
+    },
+    {
+      "epoch": 12.034734224170636,
+      "grad_norm": 1.17766535282135,
+      "learning_rate": 5.275050379453759e-05,
+      "loss": 2.8202,
+      "step": 285500
+    },
+    {
+      "epoch": 12.055810816507186,
+      "grad_norm": 1.1091578006744385,
+      "learning_rate": 5.2664751532821675e-05,
+      "loss": 2.816,
+      "step": 286000
+    },
+    {
+      "epoch": 12.076887408843739,
+      "grad_norm": 1.132075309753418,
+      "learning_rate": 5.257899927110578e-05,
+      "loss": 2.8166,
+      "step": 286500
+    },
+    {
+      "epoch": 12.097964001180289,
+      "grad_norm": 1.0818796157836914,
+      "learning_rate": 5.249324700938988e-05,
+      "loss": 2.8142,
+      "step": 287000
+    },
+    {
+      "epoch": 12.119040593516841,
+      "grad_norm": 1.1289353370666504,
+      "learning_rate": 5.240749474767397e-05,
+      "loss": 2.8145,
+      "step": 287500
+    },
+    {
+      "epoch": 12.140117185853391,
+      "grad_norm": 1.2059327363967896,
+      "learning_rate": 5.232174248595807e-05,
+      "loss": 2.8198,
+      "step": 288000
+    },
+    {
+      "epoch": 12.161193778189942,
+      "grad_norm": 1.0750534534454346,
+      "learning_rate": 5.223599022424217e-05,
+      "loss": 2.8169,
+      "step": 288500
+    },
+    {
+      "epoch": 12.182270370526494,
+      "grad_norm": 1.1586952209472656,
+      "learning_rate": 5.215023796252626e-05,
+      "loss": 2.822,
+      "step": 289000
+    },
+    {
+      "epoch": 12.203346962863044,
+      "grad_norm": 1.1892727613449097,
+      "learning_rate": 5.206448570081036e-05,
+      "loss": 2.8126,
+      "step": 289500
+    },
+    {
+      "epoch": 12.224423555199595,
+      "grad_norm": 1.1751580238342285,
+      "learning_rate": 5.197873343909446e-05,
+      "loss": 2.8124,
+      "step": 290000
+    },
+    {
+      "epoch": 12.224423555199595,
+      "eval_accuracy": 0.7323904312306266,
+      "eval_loss": 2.592902183532715,
+      "eval_runtime": 353.2013,
+      "eval_samples_per_second": 341.748,
+      "eval_steps_per_second": 3.562,
+      "step": 290000
+    },
+    {
+      "epoch": 12.245500147536147,
+      "grad_norm": 1.1863564252853394,
+      "learning_rate": 5.189298117737855e-05,
+      "loss": 2.8143,
+      "step": 290500
+    },
+    {
+      "epoch": 12.266576739872697,
+      "grad_norm": 1.122219443321228,
+      "learning_rate": 5.180722891566265e-05,
+      "loss": 2.8127,
+      "step": 291000
+    },
+    {
+      "epoch": 12.28765333220925,
+      "grad_norm": 1.1502673625946045,
+      "learning_rate": 5.1721476653946756e-05,
+      "loss": 2.81,
+      "step": 291500
+    },
+    {
+      "epoch": 12.3087299245458,
+      "grad_norm": 1.1467839479446411,
+      "learning_rate": 5.163572439223085e-05,
+      "loss": 2.8128,
+      "step": 292000
+    },
+    {
+      "epoch": 12.32980651688235,
+      "grad_norm": 1.1073864698410034,
+      "learning_rate": 5.1549972130514946e-05,
+      "loss": 2.811,
+      "step": 292500
+    },
+    {
+      "epoch": 12.350883109218902,
+      "grad_norm": 1.179865837097168,
+      "learning_rate": 5.1464219868799044e-05,
+      "loss": 2.8116,
+      "step": 293000
+    },
+    {
+      "epoch": 12.371959701555452,
+      "grad_norm": 1.2482935190200806,
+      "learning_rate": 5.1378467607083136e-05,
+      "loss": 2.8109,
+      "step": 293500
+    },
+    {
+      "epoch": 12.393036293892003,
+      "grad_norm": 1.1366822719573975,
+      "learning_rate": 5.1292715345367234e-05,
+      "loss": 2.812,
+      "step": 294000
+    },
+    {
+      "epoch": 12.414112886228555,
+      "grad_norm": 1.1328504085540771,
+      "learning_rate": 5.120696308365134e-05,
+      "loss": 2.8086,
+      "step": 294500
+    },
+    {
+      "epoch": 12.435189478565105,
+      "grad_norm": 1.1318508386611938,
+      "learning_rate": 5.112121082193543e-05,
+      "loss": 2.8107,
+      "step": 295000
+    },
+    {
+      "epoch": 12.456266070901657,
+      "grad_norm": 1.2119477987289429,
+      "learning_rate": 5.103545856021953e-05,
+      "loss": 2.8101,
+      "step": 295500
+    },
+    {
+      "epoch": 12.477342663238208,
+      "grad_norm": 1.1600226163864136,
+      "learning_rate": 5.094970629850363e-05,
+      "loss": 2.8088,
+      "step": 296000
+    },
+    {
+      "epoch": 12.498419255574758,
+      "grad_norm": 1.0812925100326538,
+      "learning_rate": 5.086395403678772e-05,
+      "loss": 2.8101,
+      "step": 296500
+    },
+    {
+      "epoch": 12.51949584791131,
+      "grad_norm": 1.1103565692901611,
+      "learning_rate": 5.0778201775071824e-05,
+      "loss": 2.8131,
+      "step": 297000
+    },
+    {
+      "epoch": 12.54057244024786,
+      "grad_norm": 1.217098593711853,
+      "learning_rate": 5.069244951335592e-05,
+      "loss": 2.8103,
+      "step": 297500
+    },
+    {
+      "epoch": 12.561649032584413,
+      "grad_norm": 1.1655622720718384,
+      "learning_rate": 5.0606697251640014e-05,
+      "loss": 2.8055,
+      "step": 298000
+    },
+    {
+      "epoch": 12.582725624920963,
+      "grad_norm": 1.1700767278671265,
+      "learning_rate": 5.052094498992411e-05,
+      "loss": 2.8118,
+      "step": 298500
+    },
+    {
+      "epoch": 12.603802217257513,
+      "grad_norm": 1.1530096530914307,
+      "learning_rate": 5.043519272820822e-05,
+      "loss": 2.8116,
+      "step": 299000
+    },
+    {
+      "epoch": 12.624878809594065,
+      "grad_norm": 1.10191011428833,
+      "learning_rate": 5.03494404664923e-05,
+      "loss": 2.805,
+      "step": 299500
+    },
+    {
+      "epoch": 12.645955401930616,
+      "grad_norm": 1.1382973194122314,
+      "learning_rate": 5.026368820477641e-05,
+      "loss": 2.8057,
+      "step": 300000
+    },
+    {
+      "epoch": 12.645955401930616,
+      "eval_accuracy": 0.7342881530475287,
+      "eval_loss": 2.5820624828338623,
+      "eval_runtime": 364.1062,
+      "eval_samples_per_second": 331.513,
+      "eval_steps_per_second": 3.455,
+      "step": 300000
+    },
+    {
+      "epoch": 12.667031994267166,
+      "grad_norm": 1.1284699440002441,
+      "learning_rate": 5.0177935943060505e-05,
+      "loss": 2.8015,
+      "step": 300500
+    },
+    {
+      "epoch": 12.688108586603718,
+      "grad_norm": 1.1500294208526611,
+      "learning_rate": 5.00921836813446e-05,
+      "loss": 2.808,
+      "step": 301000
+    },
+    {
+      "epoch": 12.709185178940269,
+      "grad_norm": 1.1069673299789429,
+      "learning_rate": 5.0006431419628695e-05,
+      "loss": 2.8061,
+      "step": 301500
+    },
+    {
+      "epoch": 12.73026177127682,
+      "grad_norm": 1.142751693725586,
+      "learning_rate": 4.992067915791279e-05,
+      "loss": 2.8049,
+      "step": 302000
+    },
+    {
+      "epoch": 12.751338363613371,
+      "grad_norm": 1.1469497680664062,
+      "learning_rate": 4.9834926896196885e-05,
+      "loss": 2.8029,
+      "step": 302500
+    },
+    {
+      "epoch": 12.772414955949921,
+      "grad_norm": 1.1647357940673828,
+      "learning_rate": 4.974917463448099e-05,
+      "loss": 2.8065,
+      "step": 303000
+    },
+    {
+      "epoch": 12.793491548286474,
+      "grad_norm": 1.1697022914886475,
+      "learning_rate": 4.966342237276508e-05,
+      "loss": 2.8083,
+      "step": 303500
+    },
+    {
+      "epoch": 12.814568140623024,
+      "grad_norm": 1.119096279144287,
+      "learning_rate": 4.957767011104918e-05,
+      "loss": 2.8052,
+      "step": 304000
+    },
+    {
+      "epoch": 12.835644732959576,
+      "grad_norm": 1.1486164331436157,
+      "learning_rate": 4.949191784933328e-05,
+      "loss": 2.8042,
+      "step": 304500
+    },
+    {
+      "epoch": 12.856721325296126,
+      "grad_norm": 1.1053950786590576,
+      "learning_rate": 4.9406165587617376e-05,
+      "loss": 2.8014,
+      "step": 305000
+    },
+    {
+      "epoch": 12.877797917632677,
+      "grad_norm": 1.1498428583145142,
+      "learning_rate": 4.932041332590147e-05,
+      "loss": 2.8062,
+      "step": 305500
+    },
+    {
+      "epoch": 12.898874509969229,
+      "grad_norm": 1.1528254747390747,
+      "learning_rate": 4.923466106418557e-05,
+      "loss": 2.8011,
+      "step": 306000
+    },
+    {
+      "epoch": 12.91995110230578,
+      "grad_norm": 1.1447033882141113,
+      "learning_rate": 4.9148908802469665e-05,
+      "loss": 2.7996,
+      "step": 306500
+    },
+    {
+      "epoch": 12.94102769464233,
+      "grad_norm": 1.131354570388794,
+      "learning_rate": 4.906315654075376e-05,
+      "loss": 2.8021,
+      "step": 307000
+    },
+    {
+      "epoch": 12.962104286978882,
+      "grad_norm": 1.2118381261825562,
+      "learning_rate": 4.897740427903786e-05,
+      "loss": 2.8044,
+      "step": 307500
+    },
+    {
+      "epoch": 12.983180879315432,
+      "grad_norm": 1.2008392810821533,
+      "learning_rate": 4.889165201732196e-05,
+      "loss": 2.8043,
+      "step": 308000
+    },
+    {
+      "epoch": 13.004257471651984,
+      "grad_norm": 1.1118990182876587,
+      "learning_rate": 4.880589975560606e-05,
+      "loss": 2.7994,
+      "step": 308500
+    },
+    {
+      "epoch": 13.025334063988534,
+      "grad_norm": 1.1810768842697144,
+      "learning_rate": 4.8720147493890156e-05,
+      "loss": 2.7934,
+      "step": 309000
+    },
+    {
+      "epoch": 13.046410656325085,
+      "grad_norm": 1.1686005592346191,
+      "learning_rate": 4.8634395232174254e-05,
+      "loss": 2.7984,
+      "step": 309500
+    },
+    {
+      "epoch": 13.067487248661637,
+      "grad_norm": 1.2280247211456299,
+      "learning_rate": 4.8548642970458346e-05,
+      "loss": 2.7945,
+      "step": 310000
+    },
+    {
+      "epoch": 13.067487248661637,
+      "eval_accuracy": 0.7354074569007104,
+      "eval_loss": 2.5764663219451904,
+      "eval_runtime": 387.9335,
+      "eval_samples_per_second": 311.151,
+      "eval_steps_per_second": 3.243,
+      "step": 310000
+    },
+    {
+      "epoch": 13.088563840998187,
+      "grad_norm": 1.138336420059204,
+      "learning_rate": 4.846289070874245e-05,
+      "loss": 2.797,
+      "step": 310500
+    },
+    {
+      "epoch": 13.109640433334738,
+      "grad_norm": 1.1514745950698853,
+      "learning_rate": 4.837713844702654e-05,
+      "loss": 2.7979,
+      "step": 311000
+    },
+    {
+      "epoch": 13.13071702567129,
+      "grad_norm": 1.2146300077438354,
+      "learning_rate": 4.829138618531064e-05,
+      "loss": 2.7927,
+      "step": 311500
+    },
+    {
+      "epoch": 13.15179361800784,
+      "grad_norm": 1.1501473188400269,
+      "learning_rate": 4.820563392359474e-05,
+      "loss": 2.7967,
+      "step": 312000
+    },
+    {
+      "epoch": 13.172870210344392,
+      "grad_norm": 1.1547178030014038,
+      "learning_rate": 4.811988166187884e-05,
+      "loss": 2.7925,
+      "step": 312500
+    },
+    {
+      "epoch": 13.193946802680943,
+      "grad_norm": 1.1092034578323364,
+      "learning_rate": 4.803412940016293e-05,
+      "loss": 2.7931,
+      "step": 313000
+    },
+    {
+      "epoch": 13.215023395017493,
+      "grad_norm": 1.1844452619552612,
+      "learning_rate": 4.7948377138447034e-05,
+      "loss": 2.7949,
+      "step": 313500
+    },
+    {
+      "epoch": 13.236099987354045,
+      "grad_norm": 1.1107574701309204,
+      "learning_rate": 4.7862624876731126e-05,
+      "loss": 2.7908,
+      "step": 314000
+    },
+    {
+      "epoch": 13.257176579690595,
+      "grad_norm": 1.1270263195037842,
+      "learning_rate": 4.7776872615015224e-05,
+      "loss": 2.7929,
+      "step": 314500
+    },
+    {
+      "epoch": 13.278253172027147,
+      "grad_norm": 1.2822831869125366,
+      "learning_rate": 4.7691120353299315e-05,
+      "loss": 2.7899,
+      "step": 315000
+    },
+    {
+      "epoch": 13.299329764363698,
+      "grad_norm": 1.0917562246322632,
+      "learning_rate": 4.760536809158342e-05,
+      "loss": 2.7958,
+      "step": 315500
+    },
+    {
+      "epoch": 13.320406356700248,
+      "grad_norm": 1.100634217262268,
+      "learning_rate": 4.751961582986751e-05,
+      "loss": 2.7947,
+      "step": 316000
+    },
+    {
+      "epoch": 13.3414829490368,
+      "grad_norm": 1.1184725761413574,
+      "learning_rate": 4.743386356815161e-05,
+      "loss": 2.7936,
+      "step": 316500
+    },
+    {
+      "epoch": 13.36255954137335,
+      "grad_norm": 1.1428985595703125,
+      "learning_rate": 4.734811130643571e-05,
+      "loss": 2.7922,
+      "step": 317000
+    },
+    {
+      "epoch": 13.383636133709901,
+      "grad_norm": 1.1732453107833862,
+      "learning_rate": 4.726235904471981e-05,
+      "loss": 2.7889,
+      "step": 317500
+    },
+    {
+      "epoch": 13.404712726046453,
+      "grad_norm": 1.1908643245697021,
+      "learning_rate": 4.71766067830039e-05,
+      "loss": 2.7912,
+      "step": 318000
+    },
+    {
+      "epoch": 13.425789318383003,
+      "grad_norm": 1.202691674232483,
+      "learning_rate": 4.7090854521288004e-05,
+      "loss": 2.794,
+      "step": 318500
+    },
+    {
+      "epoch": 13.446865910719556,
+      "grad_norm": 1.172513723373413,
+      "learning_rate": 4.7005102259572095e-05,
+      "loss": 2.7894,
+      "step": 319000
+    },
+    {
+      "epoch": 13.467942503056106,
+      "grad_norm": 1.2765130996704102,
+      "learning_rate": 4.6919349997856193e-05,
+      "loss": 2.7934,
+      "step": 319500
+    },
+    {
+      "epoch": 13.489019095392656,
+      "grad_norm": 1.1488903760910034,
+      "learning_rate": 4.683359773614029e-05,
+      "loss": 2.7875,
+      "step": 320000
+    },
+    {
+      "epoch": 13.489019095392656,
+      "eval_accuracy": 0.7365860316177575,
+      "eval_loss": 2.5696897506713867,
+      "eval_runtime": 472.6451,
+      "eval_samples_per_second": 255.384,
+      "eval_steps_per_second": 2.662,
+      "step": 320000
+    },
+    {
+      "epoch": 13.510095687729208,
+      "grad_norm": 1.2637044191360474,
+      "learning_rate": 4.674784547442439e-05,
+      "loss": 2.7948,
+      "step": 320500
+    },
+    {
+      "epoch": 13.531172280065759,
+      "grad_norm": 1.2421412467956543,
+      "learning_rate": 4.666209321270849e-05,
+      "loss": 2.79,
+      "step": 321000
+    },
+    {
+      "epoch": 13.55224887240231,
+      "grad_norm": 1.1114425659179688,
+      "learning_rate": 4.657634095099259e-05,
+      "loss": 2.7894,
+      "step": 321500
+    },
+    {
+      "epoch": 13.573325464738861,
+      "grad_norm": 1.1476738452911377,
+      "learning_rate": 4.6490588689276685e-05,
+      "loss": 2.7906,
+      "step": 322000
+    },
+    {
+      "epoch": 13.594402057075412,
+      "grad_norm": 1.1564421653747559,
+      "learning_rate": 4.6404836427560777e-05,
+      "loss": 2.7901,
+      "step": 322500
+    },
+    {
+      "epoch": 13.615478649411964,
+      "grad_norm": 1.126970887184143,
+      "learning_rate": 4.6319084165844875e-05,
+      "loss": 2.7857,
+      "step": 323000
+    },
+    {
+      "epoch": 13.636555241748514,
+      "grad_norm": 1.1418492794036865,
+      "learning_rate": 4.623333190412897e-05,
+      "loss": 2.7903,
+      "step": 323500
+    },
+    {
+      "epoch": 13.657631834085064,
+      "grad_norm": 1.1753501892089844,
+      "learning_rate": 4.614757964241307e-05,
+      "loss": 2.786,
+      "step": 324000
+    },
+    {
+      "epoch": 13.678708426421617,
+      "grad_norm": 1.1679949760437012,
+      "learning_rate": 4.606182738069717e-05,
+      "loss": 2.7883,
+      "step": 324500
+    },
+    {
+      "epoch": 13.699785018758167,
+      "grad_norm": 1.1923571825027466,
+      "learning_rate": 4.597607511898127e-05,
+      "loss": 2.7865,
+      "step": 325000
+    },
+    {
+      "epoch": 13.720861611094719,
+      "grad_norm": 1.1143501996994019,
+      "learning_rate": 4.589032285726536e-05,
+      "loss": 2.7866,
+      "step": 325500
+    },
+    {
+      "epoch": 13.74193820343127,
+      "grad_norm": 1.1569753885269165,
+      "learning_rate": 4.5804570595549465e-05,
+      "loss": 2.7857,
+      "step": 326000
+    },
+    {
+      "epoch": 13.76301479576782,
+      "grad_norm": 1.1929373741149902,
+      "learning_rate": 4.5718818333833556e-05,
+      "loss": 2.7884,
+      "step": 326500
+    },
+    {
+      "epoch": 13.784091388104372,
+      "grad_norm": 1.1675968170166016,
+      "learning_rate": 4.5633066072117655e-05,
+      "loss": 2.7874,
+      "step": 327000
+    },
+    {
+      "epoch": 13.805167980440922,
+      "grad_norm": 1.2171367406845093,
+      "learning_rate": 4.554731381040175e-05,
+      "loss": 2.7823,
+      "step": 327500
+    },
+    {
+      "epoch": 13.826244572777473,
+      "grad_norm": 1.1480249166488647,
+      "learning_rate": 4.546156154868585e-05,
+      "loss": 2.7861,
+      "step": 328000
+    },
+    {
+      "epoch": 13.847321165114025,
+      "grad_norm": 1.1949799060821533,
+      "learning_rate": 4.537580928696994e-05,
+      "loss": 2.7851,
+      "step": 328500
+    },
+    {
+      "epoch": 13.868397757450575,
+      "grad_norm": 1.2694050073623657,
+      "learning_rate": 4.529005702525405e-05,
+      "loss": 2.7849,
+      "step": 329000
+    },
+    {
+      "epoch": 13.889474349787127,
+      "grad_norm": 1.1850781440734863,
+      "learning_rate": 4.520430476353814e-05,
+      "loss": 2.7825,
+      "step": 329500
+    },
+    {
+      "epoch": 13.910550942123677,
+      "grad_norm": 1.1561588048934937,
+      "learning_rate": 4.511855250182224e-05,
+      "loss": 2.782,
+      "step": 330000
+    },
+    {
+      "epoch": 13.910550942123677,
+      "eval_accuracy": 0.737297420861579,
+      "eval_loss": 2.563389301300049,
+      "eval_runtime": 370.2722,
+      "eval_samples_per_second": 325.993,
+      "eval_steps_per_second": 3.398,
+      "step": 330000
+    },
+    {
+      "epoch": 13.931627534460228,
+      "grad_norm": 1.115807056427002,
+      "learning_rate": 4.5032800240106336e-05,
+      "loss": 2.7846,
+      "step": 330500
+    },
+    {
+      "epoch": 13.95270412679678,
+      "grad_norm": 1.192780613899231,
+      "learning_rate": 4.4947047978390434e-05,
+      "loss": 2.7874,
+      "step": 331000
+    },
+    {
+      "epoch": 13.97378071913333,
+      "grad_norm": 1.238399624824524,
+      "learning_rate": 4.4861295716674526e-05,
+      "loss": 2.7851,
+      "step": 331500
+    },
+    {
+      "epoch": 13.994857311469882,
+      "grad_norm": 1.1732228994369507,
+      "learning_rate": 4.4775543454958624e-05,
+      "loss": 2.7847,
+      "step": 332000
+    },
+    {
+      "epoch": 14.015933903806433,
+      "grad_norm": 1.2015936374664307,
+      "learning_rate": 4.468979119324272e-05,
+      "loss": 2.7761,
+      "step": 332500
+    },
+    {
+      "epoch": 14.037010496142983,
+      "grad_norm": 1.199040174484253,
+      "learning_rate": 4.460403893152682e-05,
+      "loss": 2.7781,
+      "step": 333000
+    },
+    {
+      "epoch": 14.058087088479535,
+      "grad_norm": 1.1536718606948853,
+      "learning_rate": 4.451828666981092e-05,
+      "loss": 2.7774,
+      "step": 333500
+    },
+    {
+      "epoch": 14.079163680816086,
+      "grad_norm": 1.2549256086349487,
+      "learning_rate": 4.443253440809502e-05,
+      "loss": 2.7773,
+      "step": 334000
+    },
+    {
+      "epoch": 14.100240273152636,
+      "grad_norm": 1.1009291410446167,
+      "learning_rate": 4.4346782146379116e-05,
+      "loss": 2.7771,
+      "step": 334500
+    },
+    {
+      "epoch": 14.121316865489188,
+      "grad_norm": 1.2212251424789429,
+      "learning_rate": 4.426102988466321e-05,
+      "loss": 2.7747,
+      "step": 335000
+    },
+    {
+      "epoch": 14.142393457825738,
+      "grad_norm": 1.1055370569229126,
+      "learning_rate": 4.4175277622947305e-05,
+      "loss": 2.7715,
+      "step": 335500
+    },
+    {
+      "epoch": 14.16347005016229,
+      "grad_norm": 1.2312239408493042,
+      "learning_rate": 4.4089525361231404e-05,
+      "loss": 2.7763,
+      "step": 336000
+    },
+    {
+      "epoch": 14.184546642498841,
+      "grad_norm": 1.126004934310913,
+      "learning_rate": 4.40037730995155e-05,
+      "loss": 2.7741,
+      "step": 336500
+    },
+    {
+      "epoch": 14.205623234835391,
+      "grad_norm": 1.1435521841049194,
+      "learning_rate": 4.39180208377996e-05,
+      "loss": 2.7772,
+      "step": 337000
+    },
+    {
+      "epoch": 14.226699827171943,
+      "grad_norm": 1.1560245752334595,
+      "learning_rate": 4.38322685760837e-05,
+      "loss": 2.7786,
+      "step": 337500
+    },
+    {
+      "epoch": 14.247776419508494,
+      "grad_norm": 1.1577421426773071,
+      "learning_rate": 4.374651631436779e-05,
+      "loss": 2.7753,
+      "step": 338000
+    },
+    {
+      "epoch": 14.268853011845044,
+      "grad_norm": 1.139666199684143,
+      "learning_rate": 4.3660764052651895e-05,
+      "loss": 2.7774,
+      "step": 338500
+    },
+    {
+      "epoch": 14.289929604181596,
+      "grad_norm": 1.1352485418319702,
+      "learning_rate": 4.357501179093599e-05,
+      "loss": 2.7768,
+      "step": 339000
+    },
+    {
+      "epoch": 14.311006196518147,
+      "grad_norm": 1.157981514930725,
+      "learning_rate": 4.3489259529220085e-05,
+      "loss": 2.7731,
+      "step": 339500
+    },
+    {
+      "epoch": 14.332082788854699,
+      "grad_norm": 1.2241332530975342,
+      "learning_rate": 4.3403507267504183e-05,
+      "loss": 2.7765,
+      "step": 340000
+    },
+    {
+      "epoch": 14.332082788854699,
+      "eval_accuracy": 0.7390404702836486,
+      "eval_loss": 2.5551633834838867,
+      "eval_runtime": 409.3333,
+      "eval_samples_per_second": 294.884,
+      "eval_steps_per_second": 3.073,
+      "step": 340000
+    },
+    {
+      "epoch": 14.353159381191249,
+      "grad_norm": 1.1707687377929688,
+      "learning_rate": 4.331775500578828e-05,
+      "loss": 2.7731,
+      "step": 340500
+    },
+    {
+      "epoch": 14.3742359735278,
+      "grad_norm": 1.1384748220443726,
+      "learning_rate": 4.323200274407237e-05,
+      "loss": 2.7741,
+      "step": 341000
+    },
+    {
+      "epoch": 14.395312565864351,
+      "grad_norm": 1.212461233139038,
+      "learning_rate": 4.314625048235648e-05,
+      "loss": 2.7729,
+      "step": 341500
+    },
+    {
+      "epoch": 14.416389158200902,
+      "grad_norm": 1.134296178817749,
+      "learning_rate": 4.306049822064057e-05,
+      "loss": 2.7756,
+      "step": 342000
+    },
+    {
+      "epoch": 14.437465750537454,
+      "grad_norm": 1.1625412702560425,
+      "learning_rate": 4.297474595892467e-05,
+      "loss": 2.7743,
+      "step": 342500
+    },
+    {
+      "epoch": 14.458542342874004,
+      "grad_norm": 1.129514455795288,
+      "learning_rate": 4.2888993697208767e-05,
+      "loss": 2.7711,
+      "step": 343000
+    },
+    {
+      "epoch": 14.479618935210555,
+      "grad_norm": 1.0871129035949707,
+      "learning_rate": 4.2803241435492865e-05,
+      "loss": 2.7741,
+      "step": 343500
+    },
+    {
+      "epoch": 14.500695527547107,
+      "grad_norm": 1.1068123579025269,
+      "learning_rate": 4.2717489173776956e-05,
+      "loss": 2.7717,
+      "step": 344000
+    },
+    {
+      "epoch": 14.521772119883657,
+      "grad_norm": 1.1828094720840454,
+      "learning_rate": 4.263173691206106e-05,
+      "loss": 2.7705,
+      "step": 344500
+    },
+    {
+      "epoch": 14.542848712220207,
+      "grad_norm": 1.144195795059204,
+      "learning_rate": 4.254598465034515e-05,
+      "loss": 2.7736,
+      "step": 345000
+    },
+    {
+      "epoch": 14.56392530455676,
+      "grad_norm": 1.115132451057434,
+      "learning_rate": 4.246023238862925e-05,
+      "loss": 2.7702,
+      "step": 345500
+    },
+    {
+      "epoch": 14.58500189689331,
+      "grad_norm": 1.218706727027893,
+      "learning_rate": 4.237448012691335e-05,
+      "loss": 2.7714,
+      "step": 346000
+    },
+    {
+      "epoch": 14.606078489229862,
+      "grad_norm": 1.1598670482635498,
+      "learning_rate": 4.228872786519745e-05,
+      "loss": 2.7691,
+      "step": 346500
+    },
+    {
+      "epoch": 14.627155081566412,
+      "grad_norm": 1.2239512205123901,
+      "learning_rate": 4.220297560348154e-05,
+      "loss": 2.7717,
+      "step": 347000
+    },
+    {
+      "epoch": 14.648231673902963,
+      "grad_norm": 1.1809344291687012,
+      "learning_rate": 4.2117223341765645e-05,
+      "loss": 2.7689,
+      "step": 347500
+    },
+    {
+      "epoch": 14.669308266239515,
+      "grad_norm": 1.1817526817321777,
+      "learning_rate": 4.2031471080049736e-05,
+      "loss": 2.7695,
+      "step": 348000
+    },
+    {
+      "epoch": 14.690384858576065,
+      "grad_norm": 1.126387596130371,
+      "learning_rate": 4.1945718818333834e-05,
+      "loss": 2.7696,
+      "step": 348500
+    },
+    {
+      "epoch": 14.711461450912616,
+      "grad_norm": 1.0974743366241455,
+      "learning_rate": 4.185996655661793e-05,
+      "loss": 2.7702,
+      "step": 349000
+    },
+    {
+      "epoch": 14.732538043249168,
+      "grad_norm": 1.1268750429153442,
+      "learning_rate": 4.177421429490203e-05,
+      "loss": 2.771,
+      "step": 349500
+    },
+    {
+      "epoch": 14.753614635585718,
+      "grad_norm": 1.1845661401748657,
+      "learning_rate": 4.168846203318613e-05,
+      "loss": 2.7667,
+      "step": 350000
+    },
+    {
+      "epoch": 14.753614635585718,
+      "eval_accuracy": 0.7398303808951591,
+      "eval_loss": 2.5492777824401855,
+      "eval_runtime": 381.8703,
+      "eval_samples_per_second": 316.092,
+      "eval_steps_per_second": 3.294,
+      "step": 350000
+    },
+    {
+      "epoch": 14.77469122792227,
+      "grad_norm": 1.1551570892333984,
+      "learning_rate": 4.160270977147022e-05,
+      "loss": 2.7659,
+      "step": 350500
+    },
+    {
+      "epoch": 14.79576782025882,
+      "grad_norm": 1.135872721672058,
+      "learning_rate": 4.1516957509754326e-05,
+      "loss": 2.77,
+      "step": 351000
+    },
+    {
+      "epoch": 14.81684441259537,
+      "grad_norm": 1.1674211025238037,
+      "learning_rate": 4.143120524803842e-05,
+      "loss": 2.7655,
+      "step": 351500
+    },
+    {
+      "epoch": 14.837921004931923,
+      "grad_norm": 1.1730097532272339,
+      "learning_rate": 4.1345452986322516e-05,
+      "loss": 2.7687,
+      "step": 352000
+    },
+    {
+      "epoch": 14.858997597268473,
+      "grad_norm": 1.1170876026153564,
+      "learning_rate": 4.1259700724606614e-05,
+      "loss": 2.7667,
+      "step": 352500
+    },
+    {
+      "epoch": 14.880074189605025,
+      "grad_norm": 1.1467403173446655,
+      "learning_rate": 4.117394846289071e-05,
+      "loss": 2.7671,
+      "step": 353000
+    },
+    {
+      "epoch": 14.901150781941576,
+      "grad_norm": 1.152738332748413,
+      "learning_rate": 4.1088196201174804e-05,
+      "loss": 2.7665,
+      "step": 353500
+    },
+    {
+      "epoch": 14.922227374278126,
+      "grad_norm": 1.1858787536621094,
+      "learning_rate": 4.100244393945891e-05,
+      "loss": 2.7641,
+      "step": 354000
+    },
+    {
+      "epoch": 14.943303966614678,
+      "grad_norm": 1.1796566247940063,
+      "learning_rate": 4.0916691677743e-05,
+      "loss": 2.7703,
+      "step": 354500
+    },
+    {
+      "epoch": 14.964380558951229,
+      "grad_norm": 1.1750434637069702,
+      "learning_rate": 4.08309394160271e-05,
+      "loss": 2.7669,
+      "step": 355000
+    },
+    {
+      "epoch": 14.985457151287779,
+      "grad_norm": 1.1137551069259644,
+      "learning_rate": 4.07451871543112e-05,
+      "loss": 2.7662,
+      "step": 355500
+    },
+    {
+      "epoch": 15.006533743624331,
+      "grad_norm": 1.1712113618850708,
+      "learning_rate": 4.0659434892595295e-05,
+      "loss": 2.7609,
+      "step": 356000
+    },
+    {
+      "epoch": 15.027610335960881,
+      "grad_norm": 1.1130077838897705,
+      "learning_rate": 4.057368263087939e-05,
+      "loss": 2.7565,
+      "step": 356500
+    },
+    {
+      "epoch": 15.048686928297434,
+      "grad_norm": 1.1662648916244507,
+      "learning_rate": 4.048793036916349e-05,
+      "loss": 2.7621,
+      "step": 357000
+    },
+    {
+      "epoch": 15.069763520633984,
+      "grad_norm": 1.2039918899536133,
+      "learning_rate": 4.0402178107447584e-05,
+      "loss": 2.7632,
+      "step": 357500
+    },
+    {
+      "epoch": 15.090840112970534,
+      "grad_norm": 1.1620346307754517,
+      "learning_rate": 4.031642584573168e-05,
+      "loss": 2.7627,
+      "step": 358000
+    },
+    {
+      "epoch": 15.111916705307086,
+      "grad_norm": 1.2013987302780151,
+      "learning_rate": 4.023067358401578e-05,
+      "loss": 2.7595,
+      "step": 358500
+    },
+    {
+      "epoch": 15.132993297643637,
+      "grad_norm": 1.1608084440231323,
+      "learning_rate": 4.014492132229988e-05,
+      "loss": 2.7626,
+      "step": 359000
+    },
+    {
+      "epoch": 15.154069889980189,
+      "grad_norm": 1.1765363216400146,
+      "learning_rate": 4.005916906058397e-05,
+      "loss": 2.7596,
+      "step": 359500
+    },
+    {
+      "epoch": 15.17514648231674,
+      "grad_norm": 1.2068012952804565,
+      "learning_rate": 3.9973416798868075e-05,
+      "loss": 2.7611,
+      "step": 360000
+    },
+    {
+      "epoch": 15.17514648231674,
+      "eval_accuracy": 0.7407492198785559,
+      "eval_loss": 2.54384708404541,
+      "eval_runtime": 407.4139,
+      "eval_samples_per_second": 296.274,
+      "eval_steps_per_second": 3.088,
+      "step": 360000
+    },
+    {
+      "epoch": 15.19622307465329,
+      "grad_norm": 1.2233223915100098,
+      "learning_rate": 3.988766453715217e-05,
+      "loss": 2.7618,
+      "step": 360500
+    },
+    {
+      "epoch": 15.217299666989842,
+      "grad_norm": 1.1961357593536377,
+      "learning_rate": 3.9801912275436265e-05,
+      "loss": 2.761,
+      "step": 361000
+    },
+    {
+      "epoch": 15.238376259326392,
+      "grad_norm": 1.192737340927124,
+      "learning_rate": 3.971616001372036e-05,
+      "loss": 2.7631,
+      "step": 361500
+    },
+    {
+      "epoch": 15.259452851662942,
+      "grad_norm": 1.1609729528427124,
+      "learning_rate": 3.963040775200446e-05,
+      "loss": 2.7603,
+      "step": 362000
+    },
+    {
+      "epoch": 15.280529443999495,
+      "grad_norm": 1.223203182220459,
+      "learning_rate": 3.954465549028856e-05,
+      "loss": 2.7586,
+      "step": 362500
+    },
+    {
+      "epoch": 15.301606036336045,
+      "grad_norm": 1.1583315134048462,
+      "learning_rate": 3.945890322857266e-05,
+      "loss": 2.7598,
+      "step": 363000
+    },
+    {
+      "epoch": 15.322682628672597,
+      "grad_norm": 1.1016000509262085,
+      "learning_rate": 3.9373150966856756e-05,
+      "loss": 2.7568,
+      "step": 363500
+    },
+    {
+      "epoch": 15.343759221009147,
+      "grad_norm": 1.1786805391311646,
+      "learning_rate": 3.928739870514085e-05,
+      "loss": 2.7592,
+      "step": 364000
+    },
+    {
+      "epoch": 15.364835813345698,
+      "grad_norm": 1.1580727100372314,
+      "learning_rate": 3.920164644342495e-05,
+      "loss": 2.7562,
+      "step": 364500
+    },
+    {
+      "epoch": 15.38591240568225,
+      "grad_norm": 1.1919409036636353,
+      "learning_rate": 3.9115894181709045e-05,
+      "loss": 2.7564,
+      "step": 365000
+    },
+    {
+      "epoch": 15.4069889980188,
+      "grad_norm": 1.1301671266555786,
+      "learning_rate": 3.903014191999314e-05,
+      "loss": 2.7539,
+      "step": 365500
+    },
+    {
+      "epoch": 15.42806559035535,
+      "grad_norm": 1.1871968507766724,
+      "learning_rate": 3.894438965827724e-05,
+      "loss": 2.7582,
+      "step": 366000
+    },
+    {
+      "epoch": 15.449142182691903,
+      "grad_norm": 1.1255131959915161,
+      "learning_rate": 3.885863739656134e-05,
+      "loss": 2.757,
+      "step": 366500
+    },
+    {
+      "epoch": 15.470218775028453,
+      "grad_norm": 1.1734973192214966,
+      "learning_rate": 3.877288513484543e-05,
+      "loss": 2.7545,
+      "step": 367000
+    },
+    {
+      "epoch": 15.491295367365005,
+      "grad_norm": 1.17353093624115,
+      "learning_rate": 3.868713287312953e-05,
+      "loss": 2.7572,
+      "step": 367500
+    },
+    {
+      "epoch": 15.512371959701555,
+      "grad_norm": 1.1776635646820068,
+      "learning_rate": 3.860138061141363e-05,
+      "loss": 2.7556,
+      "step": 368000
+    },
+    {
+      "epoch": 15.533448552038106,
+      "grad_norm": 1.254614233970642,
+      "learning_rate": 3.8515628349697726e-05,
+      "loss": 2.757,
+      "step": 368500
+    },
+    {
+      "epoch": 15.554525144374658,
+      "grad_norm": 1.1643548011779785,
+      "learning_rate": 3.842987608798182e-05,
+      "loss": 2.7562,
+      "step": 369000
+    },
+    {
+      "epoch": 15.575601736711208,
+      "grad_norm": 1.2005078792572021,
+      "learning_rate": 3.834412382626592e-05,
+      "loss": 2.7521,
+      "step": 369500
+    },
+    {
+      "epoch": 15.596678329047759,
+      "grad_norm": 1.1525814533233643,
+      "learning_rate": 3.8258371564550014e-05,
+      "loss": 2.7551,
+      "step": 370000
+    },
+    {
+      "epoch": 15.596678329047759,
+      "eval_accuracy": 0.7419281885177323,
+      "eval_loss": 2.5371034145355225,
+      "eval_runtime": 381.4384,
+      "eval_samples_per_second": 316.45,
+      "eval_steps_per_second": 3.298,
+      "step": 370000
+    },
+    {
+      "epoch": 15.61775492138431,
+      "grad_norm": 1.1777580976486206,
+      "learning_rate": 3.817261930283411e-05,
+      "loss": 2.7528,
+      "step": 370500
+    },
+    {
+      "epoch": 15.638831513720861,
+      "grad_norm": 1.222572922706604,
+      "learning_rate": 3.808686704111821e-05,
+      "loss": 2.7545,
+      "step": 371000
+    },
+    {
+      "epoch": 15.659908106057413,
+      "grad_norm": 1.1559070348739624,
+      "learning_rate": 3.800111477940231e-05,
+      "loss": 2.7548,
+      "step": 371500
+    },
+    {
+      "epoch": 15.680984698393964,
+      "grad_norm": 1.1902825832366943,
+      "learning_rate": 3.79153625176864e-05,
+      "loss": 2.7569,
+      "step": 372000
+    },
+    {
+      "epoch": 15.702061290730514,
+      "grad_norm": 1.132583737373352,
+      "learning_rate": 3.7829610255970506e-05,
+      "loss": 2.7569,
+      "step": 372500
+    },
+    {
+      "epoch": 15.723137883067066,
+      "grad_norm": 1.1792211532592773,
+      "learning_rate": 3.77438579942546e-05,
+      "loss": 2.7522,
+      "step": 373000
+    },
+    {
+      "epoch": 15.744214475403616,
+      "grad_norm": 1.1509644985198975,
+      "learning_rate": 3.7658105732538696e-05,
+      "loss": 2.7501,
+      "step": 373500
+    },
+    {
+      "epoch": 15.765291067740169,
+      "grad_norm": 1.1753661632537842,
+      "learning_rate": 3.7572353470822794e-05,
+      "loss": 2.7533,
+      "step": 374000
+    },
+    {
+      "epoch": 15.786367660076719,
+      "grad_norm": 1.1464297771453857,
+      "learning_rate": 3.748660120910689e-05,
+      "loss": 2.7503,
+      "step": 374500
+    },
+    {
+      "epoch": 15.80744425241327,
+      "grad_norm": 1.1978617906570435,
+      "learning_rate": 3.740084894739099e-05,
+      "loss": 2.747,
+      "step": 375000
+    },
+    {
+      "epoch": 15.828520844749821,
+      "grad_norm": 1.1943421363830566,
+      "learning_rate": 3.731509668567509e-05,
+      "loss": 2.7493,
+      "step": 375500
+    },
+    {
+      "epoch": 15.849597437086372,
+      "grad_norm": 1.1577798128128052,
+      "learning_rate": 3.722934442395919e-05,
+      "loss": 2.7534,
+      "step": 376000
+    },
+    {
+      "epoch": 15.870674029422922,
+      "grad_norm": 1.1518703699111938,
+      "learning_rate": 3.714359216224328e-05,
+      "loss": 2.7514,
+      "step": 376500
+    },
+    {
+      "epoch": 15.891750621759474,
+      "grad_norm": 1.173925518989563,
+      "learning_rate": 3.705783990052738e-05,
+      "loss": 2.7536,
+      "step": 377000
+    },
+    {
+      "epoch": 15.912827214096025,
+      "grad_norm": 1.1717406511306763,
+      "learning_rate": 3.6972087638811475e-05,
+      "loss": 2.748,
+      "step": 377500
+    },
+    {
+      "epoch": 15.933903806432577,
+      "grad_norm": 1.1404868364334106,
+      "learning_rate": 3.6886335377095574e-05,
+      "loss": 2.7524,
+      "step": 378000
+    },
+    {
+      "epoch": 15.954980398769127,
+      "grad_norm": 1.1270850896835327,
+      "learning_rate": 3.680058311537967e-05,
+      "loss": 2.7491,
+      "step": 378500
+    },
+    {
+      "epoch": 15.976056991105677,
+      "grad_norm": 1.1875402927398682,
+      "learning_rate": 3.671483085366377e-05,
+      "loss": 2.7516,
+      "step": 379000
+    },
+    {
+      "epoch": 15.99713358344223,
+      "grad_norm": 1.1255375146865845,
+      "learning_rate": 3.662907859194786e-05,
+      "loss": 2.7464,
+      "step": 379500
+    },
+    {
+      "epoch": 16.01821017577878,
+      "grad_norm": 1.1832513809204102,
+      "learning_rate": 3.654332633023197e-05,
+      "loss": 2.7481,
+      "step": 380000
+    },
+    {
+      "epoch": 16.01821017577878,
+      "eval_accuracy": 0.7429914458716615,
+      "eval_loss": 2.5312561988830566,
+      "eval_runtime": 371.7369,
+      "eval_samples_per_second": 324.708,
+      "eval_steps_per_second": 3.384,
+      "step": 380000
+    },
+    {
+      "epoch": 16.039286768115332,
+      "grad_norm": 1.1279016733169556,
+      "learning_rate": 3.645757406851606e-05,
+      "loss": 2.7449,
+      "step": 380500
+    },
+    {
+      "epoch": 16.06036336045188,
+      "grad_norm": 1.1797829866409302,
+      "learning_rate": 3.6371821806800157e-05,
+      "loss": 2.7433,
+      "step": 381000
+    },
+    {
+      "epoch": 16.081439952788433,
+      "grad_norm": 1.120765209197998,
+      "learning_rate": 3.6286069545084255e-05,
+      "loss": 2.7487,
+      "step": 381500
+    },
+    {
+      "epoch": 16.102516545124985,
+      "grad_norm": 1.1624466180801392,
+      "learning_rate": 3.620031728336835e-05,
+      "loss": 2.7452,
+      "step": 382000
+    },
+    {
+      "epoch": 16.123593137461537,
+      "grad_norm": 1.1526817083358765,
+      "learning_rate": 3.6114565021652445e-05,
+      "loss": 2.7404,
+      "step": 382500
+    },
+    {
+      "epoch": 16.144669729798085,
+      "grad_norm": 1.1349482536315918,
+      "learning_rate": 3.602881275993655e-05,
+      "loss": 2.7426,
+      "step": 383000
+    },
+    {
+      "epoch": 16.165746322134638,
+      "grad_norm": 1.2012494802474976,
+      "learning_rate": 3.594306049822064e-05,
+      "loss": 2.748,
+      "step": 383500
+    },
+    {
+      "epoch": 16.18682291447119,
+      "grad_norm": 1.1900912523269653,
+      "learning_rate": 3.585730823650474e-05,
+      "loss": 2.7446,
+      "step": 384000
+    },
+    {
+      "epoch": 16.20789950680774,
+      "grad_norm": 1.1075726747512817,
+      "learning_rate": 3.577155597478883e-05,
+      "loss": 2.7447,
+      "step": 384500
+    },
+    {
+      "epoch": 16.22897609914429,
+      "grad_norm": 1.1566611528396606,
+      "learning_rate": 3.5685803713072936e-05,
+      "loss": 2.7396,
+      "step": 385000
+    },
+    {
+      "epoch": 16.250052691480843,
+      "grad_norm": 1.181900978088379,
+      "learning_rate": 3.560005145135703e-05,
+      "loss": 2.7438,
+      "step": 385500
+    },
+    {
+      "epoch": 16.27112928381739,
+      "grad_norm": 1.210671305656433,
+      "learning_rate": 3.5514299189641126e-05,
+      "loss": 2.7409,
+      "step": 386000
+    },
+    {
+      "epoch": 16.292205876153943,
+      "grad_norm": 1.1298484802246094,
+      "learning_rate": 3.5428546927925224e-05,
+      "loss": 2.7449,
+      "step": 386500
+    },
+    {
+      "epoch": 16.313282468490495,
+      "grad_norm": 1.1701059341430664,
+      "learning_rate": 3.534279466620932e-05,
+      "loss": 2.7411,
+      "step": 387000
+    },
+    {
+      "epoch": 16.334359060827044,
+      "grad_norm": 1.1417850255966187,
+      "learning_rate": 3.525704240449342e-05,
+      "loss": 2.7427,
+      "step": 387500
+    },
+    {
+      "epoch": 16.355435653163596,
+      "grad_norm": 1.1800585985183716,
+      "learning_rate": 3.517129014277752e-05,
+      "loss": 2.7419,
+      "step": 388000
+    },
+    {
+      "epoch": 16.376512245500148,
+      "grad_norm": 1.218755841255188,
+      "learning_rate": 3.508553788106161e-05,
+      "loss": 2.7407,
+      "step": 388500
+    },
+    {
+      "epoch": 16.3975888378367,
+      "grad_norm": 1.2374519109725952,
+      "learning_rate": 3.499978561934571e-05,
+      "loss": 2.7398,
+      "step": 389000
+    },
+    {
+      "epoch": 16.41866543017325,
+      "grad_norm": 1.1342873573303223,
+      "learning_rate": 3.491403335762981e-05,
+      "loss": 2.741,
+      "step": 389500
+    },
+    {
+      "epoch": 16.4397420225098,
+      "grad_norm": 1.2349729537963867,
+      "learning_rate": 3.4828281095913906e-05,
+      "loss": 2.7426,
+      "step": 390000
+    },
+    {
+      "epoch": 16.4397420225098,
+      "eval_accuracy": 0.743898797295235,
+      "eval_loss": 2.5264172554016113,
+      "eval_runtime": 660.3587,
+      "eval_samples_per_second": 182.789,
+      "eval_steps_per_second": 1.905,
+      "step": 390000
+    },
+    {
+      "epoch": 16.460818614846353,
+      "grad_norm": 1.213935375213623,
+      "learning_rate": 3.4742528834198004e-05,
+      "loss": 2.7391,
+      "step": 390500
+    },
+    {
+      "epoch": 16.4818952071829,
+      "grad_norm": 1.2012463808059692,
+      "learning_rate": 3.46567765724821e-05,
+      "loss": 2.7423,
+      "step": 391000
+    },
+    {
+      "epoch": 16.502971799519454,
+      "grad_norm": 1.1893736124038696,
+      "learning_rate": 3.45710243107662e-05,
+      "loss": 2.7417,
+      "step": 391500
+    },
+    {
+      "epoch": 16.524048391856006,
+      "grad_norm": 1.180709958076477,
+      "learning_rate": 3.448527204905029e-05,
+      "loss": 2.7448,
+      "step": 392000
+    },
+    {
+      "epoch": 16.545124984192555,
+      "grad_norm": 1.155045986175537,
+      "learning_rate": 3.43995197873344e-05,
+      "loss": 2.7371,
+      "step": 392500
+    },
+    {
+      "epoch": 16.566201576529107,
+      "grad_norm": 1.135329246520996,
+      "learning_rate": 3.431376752561849e-05,
+      "loss": 2.7408,
+      "step": 393000
+    },
+    {
+      "epoch": 16.58727816886566,
+      "grad_norm": 1.1811561584472656,
+      "learning_rate": 3.422801526390259e-05,
+      "loss": 2.7416,
+      "step": 393500
+    },
+    {
+      "epoch": 16.608354761202207,
+      "grad_norm": 1.204284906387329,
+      "learning_rate": 3.4142263002186686e-05,
+      "loss": 2.7409,
+      "step": 394000
+    },
+    {
+      "epoch": 16.62943135353876,
+      "grad_norm": 1.1798804998397827,
+      "learning_rate": 3.4056510740470784e-05,
+      "loss": 2.7449,
+      "step": 394500
+    },
+    {
+      "epoch": 16.65050794587531,
+      "grad_norm": 1.1905360221862793,
+      "learning_rate": 3.3970758478754875e-05,
+      "loss": 2.7412,
+      "step": 395000
+    },
+    {
+      "epoch": 16.67158453821186,
+      "grad_norm": 1.1624568700790405,
+      "learning_rate": 3.388500621703898e-05,
+      "loss": 2.7414,
+      "step": 395500
+    },
+    {
+      "epoch": 16.692661130548412,
+      "grad_norm": 1.2100841999053955,
+      "learning_rate": 3.379925395532307e-05,
+      "loss": 2.7403,
+      "step": 396000
+    },
+    {
+      "epoch": 16.713737722884964,
+      "grad_norm": 1.1385600566864014,
+      "learning_rate": 3.371350169360717e-05,
+      "loss": 2.7371,
+      "step": 396500
+    },
+    {
+      "epoch": 16.734814315221517,
+      "grad_norm": 1.1348662376403809,
+      "learning_rate": 3.362774943189127e-05,
+      "loss": 2.7411,
+      "step": 397000
+    },
+    {
+      "epoch": 16.755890907558065,
+      "grad_norm": 1.1313577890396118,
+      "learning_rate": 3.354199717017537e-05,
+      "loss": 2.7373,
+      "step": 397500
+    },
+    {
+      "epoch": 16.776967499894617,
+      "grad_norm": 1.1876378059387207,
+      "learning_rate": 3.345624490845946e-05,
+      "loss": 2.7377,
+      "step": 398000
+    },
+    {
+      "epoch": 16.79804409223117,
+      "grad_norm": 1.1396428346633911,
+      "learning_rate": 3.3370492646743564e-05,
+      "loss": 2.7348,
+      "step": 398500
+    },
+    {
+      "epoch": 16.819120684567718,
+      "grad_norm": 1.182817816734314,
+      "learning_rate": 3.3284740385027655e-05,
+      "loss": 2.7418,
+      "step": 399000
+    },
+    {
+      "epoch": 16.84019727690427,
+      "grad_norm": 1.1320631504058838,
+      "learning_rate": 3.319898812331175e-05,
+      "loss": 2.7343,
+      "step": 399500
+    },
+    {
+      "epoch": 16.861273869240822,
+      "grad_norm": 1.1981252431869507,
+      "learning_rate": 3.311323586159585e-05,
+      "loss": 2.7361,
+      "step": 400000
+    },
+    {
+      "epoch": 16.861273869240822,
+      "eval_accuracy": 0.7447062381847003,
+      "eval_loss": 2.522871971130371,
+      "eval_runtime": 362.0759,
+      "eval_samples_per_second": 333.372,
+      "eval_steps_per_second": 3.474,
+      "step": 400000
+    },
+    {
+      "epoch": 16.88235046157737,
+      "grad_norm": 1.279401421546936,
+      "learning_rate": 3.302748359987995e-05,
+      "loss": 2.7334,
+      "step": 400500
+    },
+    {
+      "epoch": 16.903427053913923,
+      "grad_norm": 1.1670730113983154,
+      "learning_rate": 3.294173133816404e-05,
+      "loss": 2.7359,
+      "step": 401000
+    },
+    {
+      "epoch": 16.924503646250475,
+      "grad_norm": 1.2154498100280762,
+      "learning_rate": 3.2855979076448147e-05,
+      "loss": 2.7396,
+      "step": 401500
+    },
+    {
+      "epoch": 16.945580238587024,
+      "grad_norm": 1.1619014739990234,
+      "learning_rate": 3.277022681473224e-05,
+      "loss": 2.7367,
+      "step": 402000
+    },
+    {
+      "epoch": 16.966656830923576,
+      "grad_norm": 1.1775766611099243,
+      "learning_rate": 3.2684474553016336e-05,
+      "loss": 2.7353,
+      "step": 402500
+    },
+    {
+      "epoch": 16.987733423260128,
+      "grad_norm": 1.1922035217285156,
+      "learning_rate": 3.2598722291300435e-05,
+      "loss": 2.7385,
+      "step": 403000
+    },
+    {
+      "epoch": 17.00881001559668,
+      "grad_norm": 1.1692768335342407,
+      "learning_rate": 3.251297002958453e-05,
+      "loss": 2.7359,
+      "step": 403500
+    },
+    {
+      "epoch": 17.02988660793323,
+      "grad_norm": 1.2390981912612915,
+      "learning_rate": 3.242721776786863e-05,
+      "loss": 2.7329,
+      "step": 404000
+    },
+    {
+      "epoch": 17.05096320026978,
+      "grad_norm": 1.135617971420288,
+      "learning_rate": 3.234146550615272e-05,
+      "loss": 2.7304,
+      "step": 404500
+    },
+    {
+      "epoch": 17.072039792606333,
+      "grad_norm": 1.165669560432434,
+      "learning_rate": 3.225571324443683e-05,
+      "loss": 2.7318,
+      "step": 405000
+    },
+    {
+      "epoch": 17.09311638494288,
+      "grad_norm": 1.2223178148269653,
+      "learning_rate": 3.216996098272092e-05,
+      "loss": 2.7309,
+      "step": 405500
+    },
+    {
+      "epoch": 17.114192977279433,
+      "grad_norm": 1.1757044792175293,
+      "learning_rate": 3.208420872100502e-05,
+      "loss": 2.7326,
+      "step": 406000
+    },
+    {
+      "epoch": 17.135269569615986,
+      "grad_norm": 1.1698079109191895,
+      "learning_rate": 3.1998456459289116e-05,
+      "loss": 2.7279,
+      "step": 406500
+    },
+    {
+      "epoch": 17.156346161952534,
+      "grad_norm": 1.1588577032089233,
+      "learning_rate": 3.1912704197573214e-05,
+      "loss": 2.7283,
+      "step": 407000
+    },
+    {
+      "epoch": 17.177422754289086,
+      "grad_norm": 1.1593503952026367,
+      "learning_rate": 3.1826951935857306e-05,
+      "loss": 2.7277,
+      "step": 407500
+    },
+    {
+      "epoch": 17.19849934662564,
+      "grad_norm": 1.1717785596847534,
+      "learning_rate": 3.174119967414141e-05,
+      "loss": 2.7323,
+      "step": 408000
+    },
+    {
+      "epoch": 17.219575938962187,
+      "grad_norm": 1.181837558746338,
+      "learning_rate": 3.16554474124255e-05,
+      "loss": 2.7315,
+      "step": 408500
+    },
+    {
+      "epoch": 17.24065253129874,
+      "grad_norm": 1.2145661115646362,
+      "learning_rate": 3.15696951507096e-05,
+      "loss": 2.7316,
+      "step": 409000
+    },
+    {
+      "epoch": 17.26172912363529,
+      "grad_norm": 1.162132740020752,
+      "learning_rate": 3.14839428889937e-05,
+      "loss": 2.7285,
+      "step": 409500
+    },
+    {
+      "epoch": 17.282805715971843,
+      "grad_norm": 1.1986910104751587,
+      "learning_rate": 3.13981906272778e-05,
+      "loss": 2.7309,
+      "step": 410000
+    },
+    {
+      "epoch": 17.282805715971843,
+      "eval_accuracy": 0.7457944101195303,
+      "eval_loss": 2.5151779651641846,
+      "eval_runtime": 367.2618,
+      "eval_samples_per_second": 328.665,
+      "eval_steps_per_second": 3.425,
+      "step": 410000
+    },
+    {
+      "epoch": 17.303882308308392,
+      "grad_norm": 1.2232367992401123,
+      "learning_rate": 3.131243836556189e-05,
+      "loss": 2.7328,
+      "step": 410500
+    },
+    {
+      "epoch": 17.324958900644944,
+      "grad_norm": 1.2635061740875244,
+      "learning_rate": 3.1226686103845994e-05,
+      "loss": 2.7256,
+      "step": 411000
+    },
+    {
+      "epoch": 17.346035492981496,
+      "grad_norm": 1.2222546339035034,
+      "learning_rate": 3.1140933842130086e-05,
+      "loss": 2.7281,
+      "step": 411500
+    },
+    {
+      "epoch": 17.367112085318045,
+      "grad_norm": 1.2480653524398804,
+      "learning_rate": 3.1055181580414184e-05,
+      "loss": 2.7313,
+      "step": 412000
+    },
+    {
+      "epoch": 17.388188677654597,
+      "grad_norm": 1.1441736221313477,
+      "learning_rate": 3.096942931869828e-05,
+      "loss": 2.7328,
+      "step": 412500
+    },
+    {
+      "epoch": 17.40926526999115,
+      "grad_norm": 1.156179428100586,
+      "learning_rate": 3.088367705698238e-05,
+      "loss": 2.7261,
+      "step": 413000
+    },
+    {
+      "epoch": 17.430341862327698,
+      "grad_norm": 1.2253141403198242,
+      "learning_rate": 3.079792479526647e-05,
+      "loss": 2.7288,
+      "step": 413500
+    },
+    {
+      "epoch": 17.45141845466425,
+      "grad_norm": 1.1211682558059692,
+      "learning_rate": 3.071217253355058e-05,
+      "loss": 2.7272,
+      "step": 414000
+    },
+    {
+      "epoch": 17.472495047000802,
+      "grad_norm": 1.1524442434310913,
+      "learning_rate": 3.062642027183467e-05,
+      "loss": 2.7269,
+      "step": 414500
+    },
+    {
+      "epoch": 17.49357163933735,
+      "grad_norm": 1.26809823513031,
+      "learning_rate": 3.054066801011877e-05,
+      "loss": 2.7269,
+      "step": 415000
+    },
+    {
+      "epoch": 17.514648231673903,
+      "grad_norm": 1.2086609601974487,
+      "learning_rate": 3.045491574840287e-05,
+      "loss": 2.7277,
+      "step": 415500
+    },
+    {
+      "epoch": 17.535724824010455,
+      "grad_norm": 1.2233141660690308,
+      "learning_rate": 3.0369163486686964e-05,
+      "loss": 2.7263,
+      "step": 416000
+    },
+    {
+      "epoch": 17.556801416347007,
+      "grad_norm": 1.2027802467346191,
+      "learning_rate": 3.028341122497106e-05,
+      "loss": 2.7251,
+      "step": 416500
+    },
+    {
+      "epoch": 17.577878008683555,
+      "grad_norm": 1.2254323959350586,
+      "learning_rate": 3.019765896325516e-05,
+      "loss": 2.7298,
+      "step": 417000
+    },
+    {
+      "epoch": 17.598954601020107,
+      "grad_norm": 1.1907049417495728,
+      "learning_rate": 3.0111906701539255e-05,
+      "loss": 2.7284,
+      "step": 417500
+    },
+    {
+      "epoch": 17.62003119335666,
+      "grad_norm": 1.1579463481903076,
+      "learning_rate": 3.002615443982335e-05,
+      "loss": 2.7249,
+      "step": 418000
+    },
+    {
+      "epoch": 17.641107785693208,
+      "grad_norm": 1.1593762636184692,
+      "learning_rate": 2.9940402178107452e-05,
+      "loss": 2.7256,
+      "step": 418500
+    },
+    {
+      "epoch": 17.66218437802976,
+      "grad_norm": 1.2478811740875244,
+      "learning_rate": 2.9854649916391547e-05,
+      "loss": 2.7234,
+      "step": 419000
+    },
+    {
+      "epoch": 17.683260970366312,
+      "grad_norm": 1.2037094831466675,
+      "learning_rate": 2.976889765467564e-05,
+      "loss": 2.7302,
+      "step": 419500
+    },
+    {
+      "epoch": 17.70433756270286,
+      "grad_norm": 1.2048442363739014,
+      "learning_rate": 2.9683145392959737e-05,
+      "loss": 2.7245,
+      "step": 420000
+    },
+    {
+      "epoch": 17.70433756270286,
+      "eval_accuracy": 0.7467075483187184,
+      "eval_loss": 2.512053966522217,
+      "eval_runtime": 272.2716,
+      "eval_samples_per_second": 443.329,
+      "eval_steps_per_second": 4.62,
+      "step": 420000
+    },
+    {
+      "epoch": 17.725414155039413,
+      "grad_norm": 1.2275172472000122,
+      "learning_rate": 2.9597393131243838e-05,
+      "loss": 2.7272,
+      "step": 420500
+    },
+    {
+      "epoch": 17.746490747375965,
+      "grad_norm": 1.154929518699646,
+      "learning_rate": 2.9511640869527933e-05,
+      "loss": 2.7239,
+      "step": 421000
+    },
+    {
+      "epoch": 17.767567339712514,
+      "grad_norm": 1.259440302848816,
+      "learning_rate": 2.942588860781203e-05,
+      "loss": 2.7228,
+      "step": 421500
+    },
+    {
+      "epoch": 17.788643932049066,
+      "grad_norm": 1.1905819177627563,
+      "learning_rate": 2.934013634609613e-05,
+      "loss": 2.725,
+      "step": 422000
+    },
+    {
+      "epoch": 17.809720524385618,
+      "grad_norm": 1.193031668663025,
+      "learning_rate": 2.9254384084380225e-05,
+      "loss": 2.726,
+      "step": 422500
+    },
+    {
+      "epoch": 17.830797116722167,
+      "grad_norm": 1.1710149049758911,
+      "learning_rate": 2.9168631822664323e-05,
+      "loss": 2.726,
+      "step": 423000
+    },
+    {
+      "epoch": 17.85187370905872,
+      "grad_norm": 1.1775387525558472,
+      "learning_rate": 2.908287956094842e-05,
+      "loss": 2.7254,
+      "step": 423500
+    },
+    {
+      "epoch": 17.87295030139527,
+      "grad_norm": 1.1622763872146606,
+      "learning_rate": 2.899712729923252e-05,
+      "loss": 2.7257,
+      "step": 424000
+    },
+    {
+      "epoch": 17.894026893731823,
+      "grad_norm": 1.170509934425354,
+      "learning_rate": 2.8911375037516615e-05,
+      "loss": 2.7236,
+      "step": 424500
+    },
+    {
+      "epoch": 17.91510348606837,
+      "grad_norm": 1.1378254890441895,
+      "learning_rate": 2.8825622775800716e-05,
+      "loss": 2.7257,
+      "step": 425000
+    },
+    {
+      "epoch": 17.936180078404924,
+      "grad_norm": 1.2029753923416138,
+      "learning_rate": 2.873987051408481e-05,
+      "loss": 2.7232,
+      "step": 425500
+    },
+    {
+      "epoch": 17.957256670741476,
+      "grad_norm": 1.1676582098007202,
+      "learning_rate": 2.8654118252368906e-05,
+      "loss": 2.7216,
+      "step": 426000
+    },
+    {
+      "epoch": 17.978333263078024,
+      "grad_norm": 1.1963940858840942,
+      "learning_rate": 2.8568365990653008e-05,
+      "loss": 2.7212,
+      "step": 426500
+    },
+    {
+      "epoch": 17.999409855414576,
+      "grad_norm": 1.1567802429199219,
+      "learning_rate": 2.8482613728937103e-05,
+      "loss": 2.7246,
+      "step": 427000
+    },
+    {
+      "epoch": 18.02048644775113,
+      "grad_norm": 1.1727653741836548,
+      "learning_rate": 2.8396861467221198e-05,
+      "loss": 2.7208,
+      "step": 427500
+    },
+    {
+      "epoch": 18.041563040087677,
+      "grad_norm": 1.1888779401779175,
+      "learning_rate": 2.83111092055053e-05,
+      "loss": 2.7178,
+      "step": 428000
+    },
+    {
+      "epoch": 18.06263963242423,
+      "grad_norm": 1.1389132738113403,
+      "learning_rate": 2.8225356943789394e-05,
+      "loss": 2.7195,
+      "step": 428500
+    },
+    {
+      "epoch": 18.08371622476078,
+      "grad_norm": 1.1756205558776855,
+      "learning_rate": 2.813960468207349e-05,
+      "loss": 2.7186,
+      "step": 429000
+    },
+    {
+      "epoch": 18.10479281709733,
+      "grad_norm": 1.143972635269165,
+      "learning_rate": 2.805385242035759e-05,
+      "loss": 2.7169,
+      "step": 429500
+    },
+    {
+      "epoch": 18.125869409433882,
+      "grad_norm": 1.1781307458877563,
+      "learning_rate": 2.7968100158641686e-05,
+      "loss": 2.7188,
+      "step": 430000
+    },
+    {
+      "epoch": 18.125869409433882,
+      "eval_accuracy": 0.747145791443299,
+      "eval_loss": 2.508552074432373,
+      "eval_runtime": 256.3448,
+      "eval_samples_per_second": 470.874,
+      "eval_steps_per_second": 4.907,
+      "step": 430000
+    },
+    {
+      "epoch": 18.146946001770434,
+      "grad_norm": 1.2986667156219482,
+      "learning_rate": 2.788234789692578e-05,
+      "loss": 2.7189,
+      "step": 430500
+    },
+    {
+      "epoch": 18.168022594106986,
+      "grad_norm": 1.2140634059906006,
+      "learning_rate": 2.7796595635209882e-05,
+      "loss": 2.7185,
+      "step": 431000
+    },
+    {
+      "epoch": 18.189099186443535,
+      "grad_norm": 1.1596591472625732,
+      "learning_rate": 2.7710843373493977e-05,
+      "loss": 2.7196,
+      "step": 431500
+    },
+    {
+      "epoch": 18.210175778780087,
+      "grad_norm": 1.1550323963165283,
+      "learning_rate": 2.7625091111778072e-05,
+      "loss": 2.7217,
+      "step": 432000
+    },
+    {
+      "epoch": 18.23125237111664,
+      "grad_norm": 1.2120144367218018,
+      "learning_rate": 2.7539338850062174e-05,
+      "loss": 2.7177,
+      "step": 432500
+    },
+    {
+      "epoch": 18.252328963453188,
+      "grad_norm": 1.1742066144943237,
+      "learning_rate": 2.745358658834627e-05,
+      "loss": 2.7193,
+      "step": 433000
+    },
+    {
+      "epoch": 18.27340555578974,
+      "grad_norm": 1.1574739217758179,
+      "learning_rate": 2.7367834326630364e-05,
+      "loss": 2.7154,
+      "step": 433500
+    },
+    {
+      "epoch": 18.294482148126292,
+      "grad_norm": 1.1957488059997559,
+      "learning_rate": 2.7282082064914465e-05,
+      "loss": 2.717,
+      "step": 434000
+    },
+    {
+      "epoch": 18.31555874046284,
+      "grad_norm": 1.1930450201034546,
+      "learning_rate": 2.719632980319856e-05,
+      "loss": 2.7173,
+      "step": 434500
+    },
+    {
+      "epoch": 18.336635332799393,
+      "grad_norm": 1.1961368322372437,
+      "learning_rate": 2.7110577541482655e-05,
+      "loss": 2.7165,
+      "step": 435000
+    },
+    {
+      "epoch": 18.357711925135945,
+      "grad_norm": 1.1572502851486206,
+      "learning_rate": 2.7024825279766757e-05,
+      "loss": 2.719,
+      "step": 435500
+    },
+    {
+      "epoch": 18.378788517472493,
+      "grad_norm": 1.2107810974121094,
+      "learning_rate": 2.6939073018050852e-05,
+      "loss": 2.7161,
+      "step": 436000
+    },
+    {
+      "epoch": 18.399865109809046,
+      "grad_norm": 1.1581470966339111,
+      "learning_rate": 2.685332075633495e-05,
+      "loss": 2.718,
+      "step": 436500
+    },
+    {
+      "epoch": 18.420941702145598,
+      "grad_norm": 1.1284382343292236,
+      "learning_rate": 2.6767568494619045e-05,
+      "loss": 2.7161,
+      "step": 437000
+    },
+    {
+      "epoch": 18.44201829448215,
+      "grad_norm": 1.1889055967330933,
+      "learning_rate": 2.6681816232903143e-05,
+      "loss": 2.7148,
+      "step": 437500
+    },
+    {
+      "epoch": 18.4630948868187,
+      "grad_norm": 1.2170681953430176,
+      "learning_rate": 2.6596063971187242e-05,
+      "loss": 2.7115,
+      "step": 438000
+    },
+    {
+      "epoch": 18.48417147915525,
+      "grad_norm": 1.2624510526657104,
+      "learning_rate": 2.6510311709471337e-05,
+      "loss": 2.714,
+      "step": 438500
+    },
+    {
+      "epoch": 18.505248071491803,
+      "grad_norm": 1.2359105348587036,
+      "learning_rate": 2.642455944775544e-05,
+      "loss": 2.7148,
+      "step": 439000
+    },
+    {
+      "epoch": 18.52632466382835,
+      "grad_norm": 1.1381685733795166,
+      "learning_rate": 2.6338807186039533e-05,
+      "loss": 2.7139,
+      "step": 439500
+    },
+    {
+      "epoch": 18.547401256164903,
+      "grad_norm": 1.155768871307373,
+      "learning_rate": 2.6253054924323628e-05,
+      "loss": 2.7113,
+      "step": 440000
+    },
+    {
+      "epoch": 18.547401256164903,
+      "eval_accuracy": 0.7477654218386418,
+      "eval_loss": 2.5051496028900146,
+      "eval_runtime": 257.7992,
+      "eval_samples_per_second": 468.217,
+      "eval_steps_per_second": 4.88,
+      "step": 440000
+    },
+    {
+      "epoch": 18.568477848501455,
+      "grad_norm": 1.144463062286377,
+      "learning_rate": 2.616730266260773e-05,
+      "loss": 2.7113,
+      "step": 440500
+    },
+    {
+      "epoch": 18.589554440838004,
+      "grad_norm": 1.233823299407959,
+      "learning_rate": 2.6081550400891825e-05,
+      "loss": 2.7152,
+      "step": 441000
+    },
+    {
+      "epoch": 18.610631033174556,
+      "grad_norm": 1.2617125511169434,
+      "learning_rate": 2.599579813917592e-05,
+      "loss": 2.7174,
+      "step": 441500
+    },
+    {
+      "epoch": 18.63170762551111,
+      "grad_norm": 1.1806477308273315,
+      "learning_rate": 2.591004587746002e-05,
+      "loss": 2.7147,
+      "step": 442000
+    },
+    {
+      "epoch": 18.652784217847657,
+      "grad_norm": 1.1930525302886963,
+      "learning_rate": 2.5824293615744116e-05,
+      "loss": 2.7141,
+      "step": 442500
+    },
+    {
+      "epoch": 18.67386081018421,
+      "grad_norm": 1.1764447689056396,
+      "learning_rate": 2.573854135402821e-05,
+      "loss": 2.7141,
+      "step": 443000
+    },
+    {
+      "epoch": 18.69493740252076,
+      "grad_norm": 1.2201206684112549,
+      "learning_rate": 2.5652789092312313e-05,
+      "loss": 2.7136,
+      "step": 443500
+    },
+    {
+      "epoch": 18.716013994857313,
+      "grad_norm": 1.212592601776123,
+      "learning_rate": 2.5567036830596408e-05,
+      "loss": 2.7121,
+      "step": 444000
+    },
+    {
+      "epoch": 18.737090587193862,
+      "grad_norm": 1.2113837003707886,
+      "learning_rate": 2.5481284568880503e-05,
+      "loss": 2.7143,
+      "step": 444500
+    },
+    {
+      "epoch": 18.758167179530414,
+      "grad_norm": 1.187363624572754,
+      "learning_rate": 2.5395532307164605e-05,
+      "loss": 2.7097,
+      "step": 445000
+    },
+    {
+      "epoch": 18.779243771866966,
+      "grad_norm": 1.2141813039779663,
+      "learning_rate": 2.53097800454487e-05,
+      "loss": 2.7179,
+      "step": 445500
+    },
+    {
+      "epoch": 18.800320364203515,
+      "grad_norm": 1.137945294380188,
+      "learning_rate": 2.5224027783732794e-05,
+      "loss": 2.7144,
+      "step": 446000
+    },
+    {
+      "epoch": 18.821396956540067,
+      "grad_norm": 1.2426265478134155,
+      "learning_rate": 2.5138275522016896e-05,
+      "loss": 2.7047,
+      "step": 446500
+    },
+    {
+      "epoch": 18.84247354887662,
+      "grad_norm": 1.2007626295089722,
+      "learning_rate": 2.505252326030099e-05,
+      "loss": 2.7095,
+      "step": 447000
+    },
+    {
+      "epoch": 18.863550141213167,
+      "grad_norm": 1.2146817445755005,
+      "learning_rate": 2.496677099858509e-05,
+      "loss": 2.7102,
+      "step": 447500
+    },
+    {
+      "epoch": 18.88462673354972,
+      "grad_norm": 1.1678818464279175,
+      "learning_rate": 2.4881018736869184e-05,
+      "loss": 2.7126,
+      "step": 448000
+    },
+    {
+      "epoch": 18.90570332588627,
+      "grad_norm": 1.21388840675354,
+      "learning_rate": 2.4795266475153283e-05,
+      "loss": 2.7098,
+      "step": 448500
+    },
+    {
+      "epoch": 18.92677991822282,
+      "grad_norm": 1.253048062324524,
+      "learning_rate": 2.470951421343738e-05,
+      "loss": 2.7146,
+      "step": 449000
+    },
+    {
+      "epoch": 18.947856510559372,
+      "grad_norm": 1.1792420148849487,
+      "learning_rate": 2.4623761951721476e-05,
+      "loss": 2.7091,
+      "step": 449500
+    },
+    {
+      "epoch": 18.968933102895924,
+      "grad_norm": 1.177870750427246,
+      "learning_rate": 2.4538009690005574e-05,
+      "loss": 2.7108,
+      "step": 450000
+    },
+    {
+      "epoch": 18.968933102895924,
+      "eval_accuracy": 0.7488836256454823,
+      "eval_loss": 2.4989075660705566,
+      "eval_runtime": 257.6231,
+      "eval_samples_per_second": 468.537,
+      "eval_steps_per_second": 4.883,
+      "step": 450000
+    },
+    {
+      "epoch": 18.990009695232473,
+      "grad_norm": 1.1918195486068726,
+      "learning_rate": 2.4452257428289672e-05,
+      "loss": 2.7074,
+      "step": 450500
+    },
+    {
+      "epoch": 19.011086287569025,
+      "grad_norm": 1.2545666694641113,
+      "learning_rate": 2.436650516657377e-05,
+      "loss": 2.7074,
+      "step": 451000
+    },
+    {
+      "epoch": 19.032162879905577,
+      "grad_norm": 1.2269465923309326,
+      "learning_rate": 2.428075290485787e-05,
+      "loss": 2.7061,
+      "step": 451500
+    },
+    {
+      "epoch": 19.05323947224213,
+      "grad_norm": 1.167538046836853,
+      "learning_rate": 2.4195000643141964e-05,
+      "loss": 2.7081,
+      "step": 452000
+    },
+    {
+      "epoch": 19.074316064578678,
+      "grad_norm": 1.2128009796142578,
+      "learning_rate": 2.4109248381426062e-05,
+      "loss": 2.7092,
+      "step": 452500
+    },
+    {
+      "epoch": 19.09539265691523,
+      "grad_norm": 1.1592979431152344,
+      "learning_rate": 2.402349611971016e-05,
+      "loss": 2.7085,
+      "step": 453000
+    },
+    {
+      "epoch": 19.116469249251782,
+      "grad_norm": 1.2084112167358398,
+      "learning_rate": 2.3937743857994255e-05,
+      "loss": 2.7049,
+      "step": 453500
+    },
+    {
+      "epoch": 19.13754584158833,
+      "grad_norm": 1.2066115140914917,
+      "learning_rate": 2.3851991596278354e-05,
+      "loss": 2.7085,
+      "step": 454000
+    },
+    {
+      "epoch": 19.158622433924883,
+      "grad_norm": 1.2209522724151611,
+      "learning_rate": 2.3766239334562452e-05,
+      "loss": 2.7038,
+      "step": 454500
+    },
+    {
+      "epoch": 19.179699026261435,
+      "grad_norm": 1.2040867805480957,
+      "learning_rate": 2.3680487072846547e-05,
+      "loss": 2.7029,
+      "step": 455000
+    },
+    {
+      "epoch": 19.200775618597984,
+      "grad_norm": 1.2055507898330688,
+      "learning_rate": 2.3594734811130645e-05,
+      "loss": 2.703,
+      "step": 455500
+    },
+    {
+      "epoch": 19.221852210934536,
+      "grad_norm": 1.1465665102005005,
+      "learning_rate": 2.3508982549414744e-05,
+      "loss": 2.7035,
+      "step": 456000
+    },
+    {
+      "epoch": 19.242928803271088,
+      "grad_norm": 1.1993789672851562,
+      "learning_rate": 2.342323028769884e-05,
+      "loss": 2.7069,
+      "step": 456500
+    },
+    {
+      "epoch": 19.264005395607636,
+      "grad_norm": 1.219387173652649,
+      "learning_rate": 2.3337478025982937e-05,
+      "loss": 2.7041,
+      "step": 457000
+    },
+    {
+      "epoch": 19.28508198794419,
+      "grad_norm": 1.245322346687317,
+      "learning_rate": 2.3251725764267035e-05,
+      "loss": 2.7054,
+      "step": 457500
+    },
+    {
+      "epoch": 19.30615858028074,
+      "grad_norm": 1.1792999505996704,
+      "learning_rate": 2.316597350255113e-05,
+      "loss": 2.7071,
+      "step": 458000
+    },
+    {
+      "epoch": 19.327235172617293,
+      "grad_norm": 1.1668041944503784,
+      "learning_rate": 2.308022124083523e-05,
+      "loss": 2.7046,
+      "step": 458500
+    },
+    {
+      "epoch": 19.34831176495384,
+      "grad_norm": 1.2187529802322388,
+      "learning_rate": 2.2994468979119323e-05,
+      "loss": 2.7073,
+      "step": 459000
+    },
+    {
+      "epoch": 19.369388357290394,
+      "grad_norm": 1.2755134105682373,
+      "learning_rate": 2.290871671740342e-05,
+      "loss": 2.7039,
+      "step": 459500
+    },
+    {
+      "epoch": 19.390464949626946,
+      "grad_norm": 1.2453250885009766,
+      "learning_rate": 2.282296445568752e-05,
+      "loss": 2.7047,
+      "step": 460000
+    },
+    {
+      "epoch": 19.390464949626946,
+      "eval_accuracy": 0.7496133995082144,
+      "eval_loss": 2.4949228763580322,
+      "eval_runtime": 256.6075,
+      "eval_samples_per_second": 470.392,
+      "eval_steps_per_second": 4.902,
+      "step": 460000
+    },
+    {
+      "epoch": 19.411541541963494,
+      "grad_norm": 1.1905615329742432,
+      "learning_rate": 2.2737212193971615e-05,
+      "loss": 2.7028,
+      "step": 460500
+    },
+    {
+      "epoch": 19.432618134300046,
+      "grad_norm": 1.2662273645401,
+      "learning_rate": 2.2651459932255713e-05,
+      "loss": 2.7074,
+      "step": 461000
+    },
+    {
+      "epoch": 19.4536947266366,
+      "grad_norm": 1.2058972120285034,
+      "learning_rate": 2.256570767053981e-05,
+      "loss": 2.7011,
+      "step": 461500
+    },
+    {
+      "epoch": 19.474771318973147,
+      "grad_norm": 1.1653393507003784,
+      "learning_rate": 2.2479955408823906e-05,
+      "loss": 2.7021,
+      "step": 462000
+    },
+    {
+      "epoch": 19.4958479113097,
+      "grad_norm": 1.2116100788116455,
+      "learning_rate": 2.2394203147108005e-05,
+      "loss": 2.7036,
+      "step": 462500
+    },
+    {
+      "epoch": 19.51692450364625,
+      "grad_norm": 1.198673129081726,
+      "learning_rate": 2.2308450885392103e-05,
+      "loss": 2.7027,
+      "step": 463000
+    },
+    {
+      "epoch": 19.5380010959828,
+      "grad_norm": 1.1892800331115723,
+      "learning_rate": 2.22226986236762e-05,
+      "loss": 2.7008,
+      "step": 463500
+    },
+    {
+      "epoch": 19.559077688319352,
+      "grad_norm": 1.1824392080307007,
+      "learning_rate": 2.21369463619603e-05,
+      "loss": 2.702,
+      "step": 464000
+    },
+    {
+      "epoch": 19.580154280655904,
+      "grad_norm": 1.2906430959701538,
+      "learning_rate": 2.2051194100244394e-05,
+      "loss": 2.7013,
+      "step": 464500
+    },
+    {
+      "epoch": 19.601230872992456,
+      "grad_norm": 1.2470998764038086,
+      "learning_rate": 2.1965441838528493e-05,
+      "loss": 2.7029,
+      "step": 465000
+    },
+    {
+      "epoch": 19.622307465329005,
+      "grad_norm": 1.1859519481658936,
+      "learning_rate": 2.187968957681259e-05,
+      "loss": 2.7012,
+      "step": 465500
+    },
+    {
+      "epoch": 19.643384057665557,
+      "grad_norm": 1.2869160175323486,
+      "learning_rate": 2.179393731509669e-05,
+      "loss": 2.7033,
+      "step": 466000
+    },
+    {
+      "epoch": 19.66446065000211,
+      "grad_norm": 1.201958417892456,
+      "learning_rate": 2.1708185053380784e-05,
+      "loss": 2.7011,
+      "step": 466500
+    },
+    {
+      "epoch": 19.685537242338658,
+      "grad_norm": 1.153571367263794,
+      "learning_rate": 2.1622432791664883e-05,
+      "loss": 2.7032,
+      "step": 467000
+    },
+    {
+      "epoch": 19.70661383467521,
+      "grad_norm": 1.2097396850585938,
+      "learning_rate": 2.1536680529948978e-05,
+      "loss": 2.7021,
+      "step": 467500
+    },
+    {
+      "epoch": 19.727690427011762,
+      "grad_norm": 1.1822502613067627,
+      "learning_rate": 2.1450928268233076e-05,
+      "loss": 2.7015,
+      "step": 468000
+    },
+    {
+      "epoch": 19.74876701934831,
+      "grad_norm": 1.1519300937652588,
+      "learning_rate": 2.1365176006517174e-05,
+      "loss": 2.7015,
+      "step": 468500
+    },
+    {
+      "epoch": 19.769843611684863,
+      "grad_norm": 1.2372256517410278,
+      "learning_rate": 2.127942374480127e-05,
+      "loss": 2.7017,
+      "step": 469000
+    },
+    {
+      "epoch": 19.790920204021415,
+      "grad_norm": 1.2088857889175415,
+      "learning_rate": 2.1193671483085367e-05,
+      "loss": 2.7048,
+      "step": 469500
+    },
+    {
+      "epoch": 19.811996796357963,
+      "grad_norm": 1.176517128944397,
+      "learning_rate": 2.1107919221369466e-05,
+      "loss": 2.7021,
+      "step": 470000
+    },
+    {
+      "epoch": 19.811996796357963,
+      "eval_accuracy": 0.7501600209815477,
+      "eval_loss": 2.490896701812744,
+      "eval_runtime": 256.3218,
+      "eval_samples_per_second": 470.916,
+      "eval_steps_per_second": 4.908,
+      "step": 470000
+    },
+    {
+      "epoch": 19.833073388694515,
+      "grad_norm": 1.1569997072219849,
+      "learning_rate": 2.102216695965356e-05,
+      "loss": 2.701,
+      "step": 470500
+    },
+    {
+      "epoch": 19.854149981031068,
+      "grad_norm": 1.1770853996276855,
+      "learning_rate": 2.093641469793766e-05,
+      "loss": 2.7003,
+      "step": 471000
+    },
+    {
+      "epoch": 19.87522657336762,
+      "grad_norm": 1.1452209949493408,
+      "learning_rate": 2.0850662436221757e-05,
+      "loss": 2.6995,
+      "step": 471500
+    },
+    {
+      "epoch": 19.89630316570417,
+      "grad_norm": 1.1811836957931519,
+      "learning_rate": 2.0764910174505852e-05,
+      "loss": 2.6996,
+      "step": 472000
+    },
+    {
+      "epoch": 19.91737975804072,
+      "grad_norm": 1.157072901725769,
+      "learning_rate": 2.067915791278995e-05,
+      "loss": 2.7012,
+      "step": 472500
+    },
+    {
+      "epoch": 19.938456350377272,
+      "grad_norm": 1.227238416671753,
+      "learning_rate": 2.059340565107405e-05,
+      "loss": 2.6975,
+      "step": 473000
+    },
+    {
+      "epoch": 19.95953294271382,
+      "grad_norm": 1.1845409870147705,
+      "learning_rate": 2.0507653389358144e-05,
+      "loss": 2.6978,
+      "step": 473500
+    },
+    {
+      "epoch": 19.980609535050373,
+      "grad_norm": 1.3223782777786255,
+      "learning_rate": 2.0421901127642242e-05,
+      "loss": 2.697,
+      "step": 474000
+    },
+    {
+      "epoch": 20.001686127386925,
+      "grad_norm": 1.2172801494598389,
+      "learning_rate": 2.033614886592634e-05,
+      "loss": 2.7044,
+      "step": 474500
+    },
+    {
+      "epoch": 20.022762719723474,
+      "grad_norm": 1.177558422088623,
+      "learning_rate": 2.0250396604210435e-05,
+      "loss": 2.6987,
+      "step": 475000
+    },
+    {
+      "epoch": 20.043839312060026,
+      "grad_norm": 1.2198539972305298,
+      "learning_rate": 2.0164644342494534e-05,
+      "loss": 2.6978,
+      "step": 475500
+    },
+    {
+      "epoch": 20.064915904396578,
+      "grad_norm": 1.1658097505569458,
+      "learning_rate": 2.007889208077863e-05,
+      "loss": 2.6926,
+      "step": 476000
+    },
+    {
+      "epoch": 20.085992496733127,
+      "grad_norm": 1.1734035015106201,
+      "learning_rate": 1.9993139819062727e-05,
+      "loss": 2.694,
+      "step": 476500
+    },
+    {
+      "epoch": 20.10706908906968,
+      "grad_norm": 1.2007174491882324,
+      "learning_rate": 1.9907387557346825e-05,
+      "loss": 2.6958,
+      "step": 477000
+    },
+    {
+      "epoch": 20.12814568140623,
+      "grad_norm": 1.239338994026184,
+      "learning_rate": 1.9821635295630923e-05,
+      "loss": 2.6939,
+      "step": 477500
+    },
+    {
+      "epoch": 20.14922227374278,
+      "grad_norm": 1.245455265045166,
+      "learning_rate": 1.9735883033915022e-05,
+      "loss": 2.6934,
+      "step": 478000
+    },
+    {
+      "epoch": 20.17029886607933,
+      "grad_norm": 1.206384301185608,
+      "learning_rate": 1.965013077219912e-05,
+      "loss": 2.6976,
+      "step": 478500
+    },
+    {
+      "epoch": 20.191375458415884,
+      "grad_norm": 1.1990792751312256,
+      "learning_rate": 1.9564378510483215e-05,
+      "loss": 2.6937,
+      "step": 479000
+    },
+    {
+      "epoch": 20.212452050752436,
+      "grad_norm": 1.1855634450912476,
+      "learning_rate": 1.9478626248767313e-05,
+      "loss": 2.6977,
+      "step": 479500
+    },
+    {
+      "epoch": 20.233528643088984,
+      "grad_norm": 1.1927183866500854,
+      "learning_rate": 1.939287398705141e-05,
+      "loss": 2.6941,
+      "step": 480000
+    },
+    {
+      "epoch": 20.233528643088984,
+      "eval_accuracy": 0.7508731287716771,
+      "eval_loss": 2.486924409866333,
+      "eval_runtime": 257.2243,
+      "eval_samples_per_second": 469.264,
+      "eval_steps_per_second": 4.891,
+      "step": 480000
+    },
+    {
+      "epoch": 20.254605235425537,
+      "grad_norm": 1.1990382671356201,
+      "learning_rate": 1.9307121725335506e-05,
+      "loss": 2.6906,
+      "step": 480500
+    },
+    {
+      "epoch": 20.27568182776209,
+      "grad_norm": 1.2779431343078613,
+      "learning_rate": 1.9221369463619605e-05,
+      "loss": 2.6907,
+      "step": 481000
+    },
+    {
+      "epoch": 20.296758420098637,
+      "grad_norm": 1.1959766149520874,
+      "learning_rate": 1.9135617201903703e-05,
+      "loss": 2.6918,
+      "step": 481500
+    },
+    {
+      "epoch": 20.31783501243519,
+      "grad_norm": 1.2122710943222046,
+      "learning_rate": 1.9049864940187798e-05,
+      "loss": 2.6927,
+      "step": 482000
+    },
+    {
+      "epoch": 20.33891160477174,
+      "grad_norm": 1.2448711395263672,
+      "learning_rate": 1.8964112678471896e-05,
+      "loss": 2.6949,
+      "step": 482500
+    },
+    {
+      "epoch": 20.35998819710829,
+      "grad_norm": 1.171743392944336,
+      "learning_rate": 1.8878360416755995e-05,
+      "loss": 2.6959,
+      "step": 483000
+    },
+    {
+      "epoch": 20.381064789444842,
+      "grad_norm": 1.162561297416687,
+      "learning_rate": 1.879260815504009e-05,
+      "loss": 2.6921,
+      "step": 483500
+    },
+    {
+      "epoch": 20.402141381781394,
+      "grad_norm": 1.225953459739685,
+      "learning_rate": 1.8706855893324188e-05,
+      "loss": 2.6908,
+      "step": 484000
+    },
+    {
+      "epoch": 20.423217974117943,
+      "grad_norm": 1.1718711853027344,
+      "learning_rate": 1.8621103631608283e-05,
+      "loss": 2.6927,
+      "step": 484500
+    },
+    {
+      "epoch": 20.444294566454495,
+      "grad_norm": 1.2143754959106445,
+      "learning_rate": 1.853535136989238e-05,
+      "loss": 2.6923,
+      "step": 485000
+    },
+    {
+      "epoch": 20.465371158791047,
+      "grad_norm": 1.215700387954712,
+      "learning_rate": 1.844959910817648e-05,
+      "loss": 2.6898,
+      "step": 485500
+    },
+    {
+      "epoch": 20.4864477511276,
+      "grad_norm": 1.2255021333694458,
+      "learning_rate": 1.8363846846460574e-05,
+      "loss": 2.6951,
+      "step": 486000
+    },
+    {
+      "epoch": 20.507524343464148,
+      "grad_norm": 1.2252496480941772,
+      "learning_rate": 1.8278094584744673e-05,
+      "loss": 2.6893,
+      "step": 486500
+    },
+    {
+      "epoch": 20.5286009358007,
+      "grad_norm": 1.2190343141555786,
+      "learning_rate": 1.819234232302877e-05,
+      "loss": 2.6933,
+      "step": 487000
+    },
+    {
+      "epoch": 20.549677528137252,
+      "grad_norm": 1.1811158657073975,
+      "learning_rate": 1.8106590061312866e-05,
+      "loss": 2.6934,
+      "step": 487500
+    },
+    {
+      "epoch": 20.5707541204738,
+      "grad_norm": 1.1878994703292847,
+      "learning_rate": 1.8020837799596964e-05,
+      "loss": 2.6932,
+      "step": 488000
+    },
+    {
+      "epoch": 20.591830712810353,
+      "grad_norm": 1.1952636241912842,
+      "learning_rate": 1.7935085537881062e-05,
+      "loss": 2.6932,
+      "step": 488500
+    },
+    {
+      "epoch": 20.612907305146905,
+      "grad_norm": 1.238442063331604,
+      "learning_rate": 1.7849333276165157e-05,
+      "loss": 2.6915,
+      "step": 489000
+    },
+    {
+      "epoch": 20.633983897483454,
+      "grad_norm": 1.256340742111206,
+      "learning_rate": 1.7763581014449256e-05,
+      "loss": 2.6908,
+      "step": 489500
+    },
+    {
+      "epoch": 20.655060489820006,
+      "grad_norm": 1.1913901567459106,
+      "learning_rate": 1.7677828752733354e-05,
+      "loss": 2.6883,
+      "step": 490000
+    },
+    {
+      "epoch": 20.655060489820006,
+      "eval_accuracy": 0.7515946310389225,
+      "eval_loss": 2.482805013656616,
+      "eval_runtime": 256.7469,
+      "eval_samples_per_second": 470.136,
+      "eval_steps_per_second": 4.9,
+      "step": 490000
+    },
+    {
+      "epoch": 20.676137082156558,
+      "grad_norm": 1.2993087768554688,
+      "learning_rate": 1.7592076491017452e-05,
+      "loss": 2.6917,
+      "step": 490500
+    },
+    {
+      "epoch": 20.697213674493106,
+      "grad_norm": 1.2293691635131836,
+      "learning_rate": 1.7506324229301547e-05,
+      "loss": 2.6936,
+      "step": 491000
+    },
+    {
+      "epoch": 20.71829026682966,
+      "grad_norm": 1.2329825162887573,
+      "learning_rate": 1.7420571967585646e-05,
+      "loss": 2.6918,
+      "step": 491500
+    },
+    {
+      "epoch": 20.73936685916621,
+      "grad_norm": 1.2400989532470703,
+      "learning_rate": 1.7334819705869744e-05,
+      "loss": 2.6896,
+      "step": 492000
+    },
+    {
+      "epoch": 20.760443451502763,
+      "grad_norm": 1.1925376653671265,
+      "learning_rate": 1.7249067444153842e-05,
+      "loss": 2.6939,
+      "step": 492500
+    },
+    {
+      "epoch": 20.78152004383931,
+      "grad_norm": 1.1843823194503784,
+      "learning_rate": 1.7163315182437937e-05,
+      "loss": 2.689,
+      "step": 493000
+    },
+    {
+      "epoch": 20.802596636175863,
+      "grad_norm": 1.1863384246826172,
+      "learning_rate": 1.7077562920722035e-05,
+      "loss": 2.6881,
+      "step": 493500
+    },
+    {
+      "epoch": 20.823673228512416,
+      "grad_norm": 1.2092801332473755,
+      "learning_rate": 1.6991810659006134e-05,
+      "loss": 2.6898,
+      "step": 494000
+    },
+    {
+      "epoch": 20.844749820848964,
+      "grad_norm": 1.1610995531082153,
+      "learning_rate": 1.690605839729023e-05,
+      "loss": 2.6906,
+      "step": 494500
+    },
+    {
+      "epoch": 20.865826413185516,
+      "grad_norm": 1.1718530654907227,
+      "learning_rate": 1.6820306135574327e-05,
+      "loss": 2.689,
+      "step": 495000
+    },
+    {
+      "epoch": 20.88690300552207,
+      "grad_norm": 1.2568415403366089,
+      "learning_rate": 1.6734553873858425e-05,
+      "loss": 2.6913,
+      "step": 495500
+    },
+    {
+      "epoch": 20.907979597858617,
+      "grad_norm": 1.2411779165267944,
+      "learning_rate": 1.664880161214252e-05,
+      "loss": 2.6886,
+      "step": 496000
+    },
+    {
+      "epoch": 20.92905619019517,
+      "grad_norm": 1.1798317432403564,
+      "learning_rate": 1.656304935042662e-05,
+      "loss": 2.691,
+      "step": 496500
+    },
+    {
+      "epoch": 20.95013278253172,
+      "grad_norm": 1.1697680950164795,
+      "learning_rate": 1.6477297088710717e-05,
+      "loss": 2.6878,
+      "step": 497000
+    },
+    {
+      "epoch": 20.97120937486827,
+      "grad_norm": 1.2008106708526611,
+      "learning_rate": 1.639154482699481e-05,
+      "loss": 2.6899,
+      "step": 497500
+    },
+    {
+      "epoch": 20.992285967204822,
+      "grad_norm": 1.1727839708328247,
+      "learning_rate": 1.630579256527891e-05,
+      "loss": 2.6871,
+      "step": 498000
+    },
+    {
+      "epoch": 21.013362559541374,
+      "grad_norm": 1.2058137655258179,
+      "learning_rate": 1.6220040303563008e-05,
+      "loss": 2.6871,
+      "step": 498500
+    },
+    {
+      "epoch": 21.034439151877926,
+      "grad_norm": 1.2576847076416016,
+      "learning_rate": 1.6134288041847103e-05,
+      "loss": 2.688,
+      "step": 499000
+    },
+    {
+      "epoch": 21.055515744214475,
+      "grad_norm": 1.1873375177383423,
+      "learning_rate": 1.60485357801312e-05,
+      "loss": 2.6836,
+      "step": 499500
+    },
+    {
+      "epoch": 21.076592336551027,
+      "grad_norm": 1.2221812009811401,
+      "learning_rate": 1.59627835184153e-05,
+      "loss": 2.6843,
+      "step": 500000
+    },
+    {
+      "epoch": 21.076592336551027,
+      "eval_accuracy": 0.7522070062872485,
+      "eval_loss": 2.4785408973693848,
+      "eval_runtime": 256.1298,
+      "eval_samples_per_second": 471.269,
+      "eval_steps_per_second": 4.912,
+      "step": 500000
+    },
+    {
+      "epoch": 21.09766892888758,
+      "grad_norm": 1.2302268743515015,
+      "learning_rate": 1.5877031256699395e-05,
+      "loss": 2.6836,
+      "step": 500500
+    },
+    {
+      "epoch": 21.118745521224128,
+      "grad_norm": 1.2112387418746948,
+      "learning_rate": 1.5791278994983493e-05,
+      "loss": 2.6878,
+      "step": 501000
+    },
+    {
+      "epoch": 21.13982211356068,
+      "grad_norm": 1.199427843093872,
+      "learning_rate": 1.570552673326759e-05,
+      "loss": 2.683,
+      "step": 501500
+    },
+    {
+      "epoch": 21.16089870589723,
+      "grad_norm": 1.2058426141738892,
+      "learning_rate": 1.5619774471551686e-05,
+      "loss": 2.6874,
+      "step": 502000
+    },
+    {
+      "epoch": 21.18197529823378,
+      "grad_norm": 1.1829116344451904,
+      "learning_rate": 1.5534022209835785e-05,
+      "loss": 2.6867,
+      "step": 502500
+    },
+    {
+      "epoch": 21.203051890570332,
+      "grad_norm": 1.1598479747772217,
+      "learning_rate": 1.544826994811988e-05,
+      "loss": 2.6846,
+      "step": 503000
+    },
+    {
+      "epoch": 21.224128482906885,
+      "grad_norm": 1.1928588151931763,
+      "learning_rate": 1.5362517686403978e-05,
+      "loss": 2.6851,
+      "step": 503500
+    },
+    {
+      "epoch": 21.245205075243433,
+      "grad_norm": 1.18240487575531,
+      "learning_rate": 1.5276765424688076e-05,
+      "loss": 2.6864,
+      "step": 504000
+    },
+    {
+      "epoch": 21.266281667579985,
+      "grad_norm": 1.2900537252426147,
+      "learning_rate": 1.5191013162972173e-05,
+      "loss": 2.6838,
+      "step": 504500
+    },
+    {
+      "epoch": 21.287358259916537,
+      "grad_norm": 1.2119444608688354,
+      "learning_rate": 1.5105260901256271e-05,
+      "loss": 2.6857,
+      "step": 505000
+    },
+    {
+      "epoch": 21.308434852253086,
+      "grad_norm": 1.1992489099502563,
+      "learning_rate": 1.501950863954037e-05,
+      "loss": 2.6857,
+      "step": 505500
+    },
+    {
+      "epoch": 21.329511444589638,
+      "grad_norm": 1.2674964666366577,
+      "learning_rate": 1.4933756377824464e-05,
+      "loss": 2.6824,
+      "step": 506000
+    },
+    {
+      "epoch": 21.35058803692619,
+      "grad_norm": 1.202696681022644,
+      "learning_rate": 1.4848004116108563e-05,
+      "loss": 2.68,
+      "step": 506500
+    },
+    {
+      "epoch": 21.371664629262742,
+      "grad_norm": 1.2270926237106323,
+      "learning_rate": 1.4762251854392661e-05,
+      "loss": 2.6861,
+      "step": 507000
+    },
+    {
+      "epoch": 21.39274122159929,
+      "grad_norm": 1.2181349992752075,
+      "learning_rate": 1.4676499592676758e-05,
+      "loss": 2.6856,
+      "step": 507500
+    },
+    {
+      "epoch": 21.413817813935843,
+      "grad_norm": 1.1818658113479614,
+      "learning_rate": 1.4590747330960856e-05,
+      "loss": 2.6836,
+      "step": 508000
+    },
+    {
+      "epoch": 21.434894406272395,
+      "grad_norm": 1.1808966398239136,
+      "learning_rate": 1.4504995069244952e-05,
+      "loss": 2.6801,
+      "step": 508500
+    },
+    {
+      "epoch": 21.455970998608944,
+      "grad_norm": 1.1550387144088745,
+      "learning_rate": 1.4419242807529049e-05,
+      "loss": 2.6839,
+      "step": 509000
+    },
+    {
+      "epoch": 21.477047590945496,
+      "grad_norm": 1.1675546169281006,
+      "learning_rate": 1.4333490545813147e-05,
+      "loss": 2.6845,
+      "step": 509500
+    },
+    {
+      "epoch": 21.498124183282048,
+      "grad_norm": 1.1917377710342407,
+      "learning_rate": 1.4247738284097246e-05,
+      "loss": 2.681,
+      "step": 510000
+    },
+    {
+      "epoch": 21.498124183282048,
+      "eval_accuracy": 0.7529528884201871,
+      "eval_loss": 2.4754700660705566,
+      "eval_runtime": 256.7167,
+      "eval_samples_per_second": 470.191,
+      "eval_steps_per_second": 4.9,
+      "step": 510000
+    },
+    {
+      "epoch": 21.519200775618597,
+      "grad_norm": 1.2591363191604614,
+      "learning_rate": 1.416198602238134e-05,
+      "loss": 2.6846,
+      "step": 510500
+    },
+    {
+      "epoch": 21.54027736795515,
+      "grad_norm": 1.1656168699264526,
+      "learning_rate": 1.4076233760665439e-05,
+      "loss": 2.6824,
+      "step": 511000
+    },
+    {
+      "epoch": 21.5613539602917,
+      "grad_norm": 1.2221752405166626,
+      "learning_rate": 1.3990481498949534e-05,
+      "loss": 2.6809,
+      "step": 511500
+    },
+    {
+      "epoch": 21.58243055262825,
+      "grad_norm": 1.2157338857650757,
+      "learning_rate": 1.3904729237233632e-05,
+      "loss": 2.6843,
+      "step": 512000
+    },
+    {
+      "epoch": 21.6035071449648,
+      "grad_norm": 1.175045132637024,
+      "learning_rate": 1.381897697551773e-05,
+      "loss": 2.68,
+      "step": 512500
+    },
+    {
+      "epoch": 21.624583737301354,
+      "grad_norm": 1.1866666078567505,
+      "learning_rate": 1.3733224713801825e-05,
+      "loss": 2.6847,
+      "step": 513000
+    },
+    {
+      "epoch": 21.645660329637906,
+      "grad_norm": 1.2531830072402954,
+      "learning_rate": 1.3647472452085924e-05,
+      "loss": 2.6851,
+      "step": 513500
+    },
+    {
+      "epoch": 21.666736921974454,
+      "grad_norm": 1.2429271936416626,
+      "learning_rate": 1.3561720190370022e-05,
+      "loss": 2.6831,
+      "step": 514000
+    },
+    {
+      "epoch": 21.687813514311006,
+      "grad_norm": 1.18976891040802,
+      "learning_rate": 1.3475967928654119e-05,
+      "loss": 2.6817,
+      "step": 514500
+    },
+    {
+      "epoch": 21.70889010664756,
+      "grad_norm": 1.2261024713516235,
+      "learning_rate": 1.3390215666938217e-05,
+      "loss": 2.6833,
+      "step": 515000
+    },
+    {
+      "epoch": 21.729966698984107,
+      "grad_norm": 1.1794220209121704,
+      "learning_rate": 1.3304463405222315e-05,
+      "loss": 2.6829,
+      "step": 515500
+    },
+    {
+      "epoch": 21.75104329132066,
+      "grad_norm": 1.2634303569793701,
+      "learning_rate": 1.321871114350641e-05,
+      "loss": 2.6851,
+      "step": 516000
+    },
+    {
+      "epoch": 21.77211988365721,
+      "grad_norm": 1.2092864513397217,
+      "learning_rate": 1.3132958881790508e-05,
+      "loss": 2.6797,
+      "step": 516500
+    },
+    {
+      "epoch": 21.79319647599376,
+      "grad_norm": 1.1771397590637207,
+      "learning_rate": 1.3047206620074607e-05,
+      "loss": 2.6787,
+      "step": 517000
+    },
+    {
+      "epoch": 21.814273068330312,
+      "grad_norm": 1.207892656326294,
+      "learning_rate": 1.2961454358358702e-05,
+      "loss": 2.6803,
+      "step": 517500
+    },
+    {
+      "epoch": 21.835349660666864,
+      "grad_norm": 1.2433807849884033,
+      "learning_rate": 1.28757020966428e-05,
+      "loss": 2.6792,
+      "step": 518000
+    },
+    {
+      "epoch": 21.856426253003413,
+      "grad_norm": 1.2182440757751465,
+      "learning_rate": 1.2789949834926898e-05,
+      "loss": 2.6775,
+      "step": 518500
+    },
+    {
+      "epoch": 21.877502845339965,
+      "grad_norm": 1.171600580215454,
+      "learning_rate": 1.2704197573210993e-05,
+      "loss": 2.6841,
+      "step": 519000
+    },
+    {
+      "epoch": 21.898579437676517,
+      "grad_norm": 1.206408977508545,
+      "learning_rate": 1.2618445311495092e-05,
+      "loss": 2.6768,
+      "step": 519500
+    },
+    {
+      "epoch": 21.91965603001307,
+      "grad_norm": 1.256736159324646,
+      "learning_rate": 1.2532693049779186e-05,
+      "loss": 2.6756,
+      "step": 520000
+    },
+    {
+      "epoch": 21.91965603001307,
+      "eval_accuracy": 0.7534510827131166,
+      "eval_loss": 2.4729316234588623,
+      "eval_runtime": 254.958,
+      "eval_samples_per_second": 473.435,
+      "eval_steps_per_second": 4.934,
+      "step": 520000
+    },
+    {
+      "epoch": 21.940732622349618,
+      "grad_norm": 1.2883257865905762,
+      "learning_rate": 1.2446940788063285e-05,
+      "loss": 2.6803,
+      "step": 520500
+    },
+    {
+      "epoch": 21.96180921468617,
+      "grad_norm": 1.2392154932022095,
+      "learning_rate": 1.2361188526347383e-05,
+      "loss": 2.6801,
+      "step": 521000
+    },
+    {
+      "epoch": 21.982885807022722,
+      "grad_norm": 1.2188835144042969,
+      "learning_rate": 1.227543626463148e-05,
+      "loss": 2.6804,
+      "step": 521500
+    },
+    {
+      "epoch": 22.00396239935927,
+      "grad_norm": 1.2353577613830566,
+      "learning_rate": 1.2189684002915578e-05,
+      "loss": 2.675,
+      "step": 522000
+    },
+    {
+      "epoch": 22.025038991695823,
+      "grad_norm": 1.22325599193573,
+      "learning_rate": 1.2103931741199675e-05,
+      "loss": 2.6751,
+      "step": 522500
+    },
+    {
+      "epoch": 22.046115584032375,
+      "grad_norm": 1.186126470565796,
+      "learning_rate": 1.2018179479483773e-05,
+      "loss": 2.6756,
+      "step": 523000
+    },
+    {
+      "epoch": 22.067192176368923,
+      "grad_norm": 1.2205013036727905,
+      "learning_rate": 1.193242721776787e-05,
+      "loss": 2.6782,
+      "step": 523500
+    },
+    {
+      "epoch": 22.088268768705476,
+      "grad_norm": 1.2109812498092651,
+      "learning_rate": 1.1846674956051966e-05,
+      "loss": 2.6794,
+      "step": 524000
+    },
+    {
+      "epoch": 22.109345361042028,
+      "grad_norm": 1.2474455833435059,
+      "learning_rate": 1.1760922694336064e-05,
+      "loss": 2.6739,
+      "step": 524500
+    },
+    {
+      "epoch": 22.130421953378576,
+      "grad_norm": 1.2150230407714844,
+      "learning_rate": 1.1675170432620161e-05,
+      "loss": 2.6803,
+      "step": 525000
+    },
+    {
+      "epoch": 22.15149854571513,
+      "grad_norm": 1.1918593645095825,
+      "learning_rate": 1.1589418170904258e-05,
+      "loss": 2.6719,
+      "step": 525500
+    },
+    {
+      "epoch": 22.17257513805168,
+      "grad_norm": 1.1611078977584839,
+      "learning_rate": 1.1503665909188354e-05,
+      "loss": 2.6747,
+      "step": 526000
+    },
+    {
+      "epoch": 22.193651730388233,
+      "grad_norm": 1.2288941144943237,
+      "learning_rate": 1.1417913647472453e-05,
+      "loss": 2.6754,
+      "step": 526500
+    },
+    {
+      "epoch": 22.21472832272478,
+      "grad_norm": 1.192989468574524,
+      "learning_rate": 1.133216138575655e-05,
+      "loss": 2.6755,
+      "step": 527000
+    },
+    {
+      "epoch": 22.235804915061333,
+      "grad_norm": 1.2392932176589966,
+      "learning_rate": 1.1246409124040646e-05,
+      "loss": 2.6739,
+      "step": 527500
+    },
+    {
+      "epoch": 22.256881507397885,
+      "grad_norm": 1.2431585788726807,
+      "learning_rate": 1.1160656862324744e-05,
+      "loss": 2.6765,
+      "step": 528000
+    },
+    {
+      "epoch": 22.277958099734434,
+      "grad_norm": 1.198577880859375,
+      "learning_rate": 1.1074904600608842e-05,
+      "loss": 2.6757,
+      "step": 528500
+    },
+    {
+      "epoch": 22.299034692070986,
+      "grad_norm": 1.2593759298324585,
+      "learning_rate": 1.0989152338892939e-05,
+      "loss": 2.6723,
+      "step": 529000
+    },
+    {
+      "epoch": 22.320111284407538,
+      "grad_norm": 1.2454047203063965,
+      "learning_rate": 1.0903400077177037e-05,
+      "loss": 2.677,
+      "step": 529500
+    },
+    {
+      "epoch": 22.341187876744087,
+      "grad_norm": 1.176487922668457,
+      "learning_rate": 1.0817647815461134e-05,
+      "loss": 2.6757,
+      "step": 530000
+    },
+    {
+      "epoch": 22.341187876744087,
+      "eval_accuracy": 0.7538600036371459,
+      "eval_loss": 2.4706943035125732,
+      "eval_runtime": 257.0305,
+      "eval_samples_per_second": 469.617,
+      "eval_steps_per_second": 4.894,
+      "step": 530000
+    },
+    {
+      "epoch": 22.36226446908064,
+      "grad_norm": 1.1798120737075806,
+      "learning_rate": 1.073189555374523e-05,
+      "loss": 2.6775,
+      "step": 530500
+    },
+    {
+      "epoch": 22.38334106141719,
+      "grad_norm": 1.294094204902649,
+      "learning_rate": 1.0646143292029327e-05,
+      "loss": 2.6782,
+      "step": 531000
+    },
+    {
+      "epoch": 22.40441765375374,
+      "grad_norm": 1.1631076335906982,
+      "learning_rate": 1.0560391030313425e-05,
+      "loss": 2.6726,
+      "step": 531500
+    },
+    {
+      "epoch": 22.42549424609029,
+      "grad_norm": 1.1891592741012573,
+      "learning_rate": 1.0474638768597522e-05,
+      "loss": 2.6725,
+      "step": 532000
+    },
+    {
+      "epoch": 22.446570838426844,
+      "grad_norm": 1.2378301620483398,
+      "learning_rate": 1.0388886506881619e-05,
+      "loss": 2.6737,
+      "step": 532500
+    },
+    {
+      "epoch": 22.467647430763392,
+      "grad_norm": 1.2077827453613281,
+      "learning_rate": 1.0303134245165717e-05,
+      "loss": 2.6756,
+      "step": 533000
+    },
+    {
+      "epoch": 22.488724023099945,
+      "grad_norm": 1.2310006618499756,
+      "learning_rate": 1.0217381983449814e-05,
+      "loss": 2.673,
+      "step": 533500
+    },
+    {
+      "epoch": 22.509800615436497,
+      "grad_norm": 1.1543177366256714,
+      "learning_rate": 1.013162972173391e-05,
+      "loss": 2.6735,
+      "step": 534000
+    },
+    {
+      "epoch": 22.53087720777305,
+      "grad_norm": 1.2122540473937988,
+      "learning_rate": 1.0045877460018009e-05,
+      "loss": 2.6755,
+      "step": 534500
+    },
+    {
+      "epoch": 22.551953800109597,
+      "grad_norm": 1.2248764038085938,
+      "learning_rate": 9.960125198302105e-06,
+      "loss": 2.677,
+      "step": 535000
+    },
+    {
+      "epoch": 22.57303039244615,
+      "grad_norm": 1.1913903951644897,
+      "learning_rate": 9.874372936586203e-06,
+      "loss": 2.674,
+      "step": 535500
+    },
+    {
+      "epoch": 22.5941069847827,
+      "grad_norm": 1.1807990074157715,
+      "learning_rate": 9.7886206748703e-06,
+      "loss": 2.6731,
+      "step": 536000
+    },
+    {
+      "epoch": 22.61518357711925,
+      "grad_norm": 1.2151296138763428,
+      "learning_rate": 9.702868413154398e-06,
+      "loss": 2.6714,
+      "step": 536500
+    },
+    {
+      "epoch": 22.636260169455802,
+      "grad_norm": 1.2244514226913452,
+      "learning_rate": 9.617116151438495e-06,
+      "loss": 2.6731,
+      "step": 537000
+    },
+    {
+      "epoch": 22.657336761792354,
+      "grad_norm": 1.28788423538208,
+      "learning_rate": 9.531363889722592e-06,
+      "loss": 2.674,
+      "step": 537500
+    },
+    {
+      "epoch": 22.678413354128903,
+      "grad_norm": 1.193604826927185,
+      "learning_rate": 9.44561162800669e-06,
+      "loss": 2.6717,
+      "step": 538000
+    },
+    {
+      "epoch": 22.699489946465455,
+      "grad_norm": 1.249402642250061,
+      "learning_rate": 9.359859366290787e-06,
+      "loss": 2.6725,
+      "step": 538500
+    },
+    {
+      "epoch": 22.720566538802007,
+      "grad_norm": 1.208331823348999,
+      "learning_rate": 9.274107104574883e-06,
+      "loss": 2.6755,
+      "step": 539000
+    },
+    {
+      "epoch": 22.741643131138556,
+      "grad_norm": 1.172566533088684,
+      "learning_rate": 9.18835484285898e-06,
+      "loss": 2.6707,
+      "step": 539500
+    },
+    {
+      "epoch": 22.762719723475108,
+      "grad_norm": 1.2443684339523315,
+      "learning_rate": 9.102602581143078e-06,
+      "loss": 2.6705,
+      "step": 540000
+    },
+    {
+      "epoch": 22.762719723475108,
+      "eval_accuracy": 0.7540602933304733,
+      "eval_loss": 2.469379186630249,
+      "eval_runtime": 255.7871,
+      "eval_samples_per_second": 471.9,
+      "eval_steps_per_second": 4.918,
+      "step": 540000
+    },
+    {
+      "epoch": 22.78379631581166,
+      "grad_norm": 1.2495062351226807,
+      "learning_rate": 9.016850319427175e-06,
+      "loss": 2.6746,
+      "step": 540500
+    },
+    {
+      "epoch": 22.804872908148212,
+      "grad_norm": 1.217437982559204,
+      "learning_rate": 8.931098057711271e-06,
+      "loss": 2.6732,
+      "step": 541000
+    },
+    {
+      "epoch": 22.82594950048476,
+      "grad_norm": 1.1984294652938843,
+      "learning_rate": 8.84534579599537e-06,
+      "loss": 2.673,
+      "step": 541500
+    },
+    {
+      "epoch": 22.847026092821313,
+      "grad_norm": 1.2583712339401245,
+      "learning_rate": 8.759593534279468e-06,
+      "loss": 2.6732,
+      "step": 542000
+    },
+    {
+      "epoch": 22.868102685157865,
+      "grad_norm": 1.1942957639694214,
+      "learning_rate": 8.673841272563565e-06,
+      "loss": 2.6698,
+      "step": 542500
+    },
+    {
+      "epoch": 22.889179277494414,
+      "grad_norm": 1.248986840248108,
+      "learning_rate": 8.588089010847663e-06,
+      "loss": 2.6725,
+      "step": 543000
+    },
+    {
+      "epoch": 22.910255869830966,
+      "grad_norm": 1.2051935195922852,
+      "learning_rate": 8.50233674913176e-06,
+      "loss": 2.6723,
+      "step": 543500
+    },
+    {
+      "epoch": 22.931332462167518,
+      "grad_norm": 1.2573319673538208,
+      "learning_rate": 8.416584487415856e-06,
+      "loss": 2.6679,
+      "step": 544000
+    },
+    {
+      "epoch": 22.952409054504066,
+      "grad_norm": 1.2693605422973633,
+      "learning_rate": 8.330832225699953e-06,
+      "loss": 2.6684,
+      "step": 544500
+    },
+    {
+      "epoch": 22.97348564684062,
+      "grad_norm": 1.2628155946731567,
+      "learning_rate": 8.245079963984051e-06,
+      "loss": 2.6682,
+      "step": 545000
+    },
+    {
+      "epoch": 22.99456223917717,
+      "grad_norm": 1.2379945516586304,
+      "learning_rate": 8.159327702268148e-06,
+      "loss": 2.673,
+      "step": 545500
+    },
+    {
+      "epoch": 23.01563883151372,
+      "grad_norm": 1.2032510042190552,
+      "learning_rate": 8.073575440552244e-06,
+      "loss": 2.6696,
+      "step": 546000
+    },
+    {
+      "epoch": 23.03671542385027,
+      "grad_norm": 1.206868290901184,
+      "learning_rate": 7.987823178836343e-06,
+      "loss": 2.6707,
+      "step": 546500
+    },
+    {
+      "epoch": 23.057792016186824,
+      "grad_norm": 1.2096521854400635,
+      "learning_rate": 7.90207091712044e-06,
+      "loss": 2.668,
+      "step": 547000
+    },
+    {
+      "epoch": 23.078868608523376,
+      "grad_norm": 1.2348042726516724,
+      "learning_rate": 7.816318655404536e-06,
+      "loss": 2.6668,
+      "step": 547500
+    },
+    {
+      "epoch": 23.099945200859924,
+      "grad_norm": 1.236573576927185,
+      "learning_rate": 7.730566393688634e-06,
+      "loss": 2.6697,
+      "step": 548000
+    },
+    {
+      "epoch": 23.121021793196476,
+      "grad_norm": 1.2323729991912842,
+      "learning_rate": 7.64481413197273e-06,
+      "loss": 2.6688,
+      "step": 548500
+    },
+    {
+      "epoch": 23.14209838553303,
+      "grad_norm": 1.2238707542419434,
+      "learning_rate": 7.559061870256828e-06,
+      "loss": 2.6683,
+      "step": 549000
+    },
+    {
+      "epoch": 23.163174977869577,
+      "grad_norm": 1.2081378698349,
+      "learning_rate": 7.473309608540925e-06,
+      "loss": 2.6678,
+      "step": 549500
+    },
+    {
+      "epoch": 23.18425157020613,
+      "grad_norm": 1.2485332489013672,
+      "learning_rate": 7.387557346825023e-06,
+      "loss": 2.6708,
+      "step": 550000
+    },
+    {
+      "epoch": 23.18425157020613,
+      "eval_accuracy": 0.7545105547672611,
+      "eval_loss": 2.4654266834259033,
+      "eval_runtime": 256.6246,
+      "eval_samples_per_second": 470.36,
+      "eval_steps_per_second": 4.902,
+      "step": 550000
+    },
+    {
+      "epoch": 23.20532816254268,
+      "grad_norm": 1.2178773880004883,
+      "learning_rate": 7.3018050851091205e-06,
+      "loss": 2.6648,
+      "step": 550500
+    },
+    {
+      "epoch": 23.22640475487923,
+      "grad_norm": 1.2079859972000122,
+      "learning_rate": 7.216052823393217e-06,
+      "loss": 2.6686,
+      "step": 551000
+    },
+    {
+      "epoch": 23.247481347215782,
+      "grad_norm": 1.272722840309143,
+      "learning_rate": 7.1303005616773155e-06,
+      "loss": 2.6673,
+      "step": 551500
+    },
+    {
+      "epoch": 23.268557939552334,
+      "grad_norm": 1.2521681785583496,
+      "learning_rate": 7.044548299961412e-06,
+      "loss": 2.6718,
+      "step": 552000
+    },
+    {
+      "epoch": 23.289634531888883,
+      "grad_norm": 1.2028534412384033,
+      "learning_rate": 6.958796038245509e-06,
+      "loss": 2.6623,
+      "step": 552500
+    },
+    {
+      "epoch": 23.310711124225435,
+      "grad_norm": 1.3018420934677124,
+      "learning_rate": 6.873043776529606e-06,
+      "loss": 2.6702,
+      "step": 553000
+    },
+    {
+      "epoch": 23.331787716561987,
+      "grad_norm": 1.2654086351394653,
+      "learning_rate": 6.787291514813704e-06,
+      "loss": 2.6649,
+      "step": 553500
+    },
+    {
+      "epoch": 23.35286430889854,
+      "grad_norm": 1.253478765487671,
+      "learning_rate": 6.701539253097801e-06,
+      "loss": 2.6675,
+      "step": 554000
+    },
+    {
+      "epoch": 23.373940901235088,
+      "grad_norm": 1.281093716621399,
+      "learning_rate": 6.615786991381898e-06,
+      "loss": 2.6668,
+      "step": 554500
+    },
+    {
+      "epoch": 23.39501749357164,
+      "grad_norm": 1.2161346673965454,
+      "learning_rate": 6.530034729665996e-06,
+      "loss": 2.6676,
+      "step": 555000
+    },
+    {
+      "epoch": 23.416094085908192,
+      "grad_norm": 1.2410410642623901,
+      "learning_rate": 6.444282467950093e-06,
+      "loss": 2.6666,
+      "step": 555500
+    },
+    {
+      "epoch": 23.43717067824474,
+      "grad_norm": 1.1940752267837524,
+      "learning_rate": 6.358530206234189e-06,
+      "loss": 2.6662,
+      "step": 556000
+    },
+    {
+      "epoch": 23.458247270581293,
+      "grad_norm": 1.2339050769805908,
+      "learning_rate": 6.272777944518287e-06,
+      "loss": 2.6691,
+      "step": 556500
+    },
+    {
+      "epoch": 23.479323862917845,
+      "grad_norm": 1.2406237125396729,
+      "learning_rate": 6.187025682802384e-06,
+      "loss": 2.6684,
+      "step": 557000
+    },
+    {
+      "epoch": 23.500400455254393,
+      "grad_norm": 1.2184898853302002,
+      "learning_rate": 6.101273421086482e-06,
+      "loss": 2.6667,
+      "step": 557500
+    },
+    {
+      "epoch": 23.521477047590945,
+      "grad_norm": 1.2442213296890259,
+      "learning_rate": 6.015521159370579e-06,
+      "loss": 2.6667,
+      "step": 558000
+    },
+    {
+      "epoch": 23.542553639927497,
+      "grad_norm": 1.2631967067718506,
+      "learning_rate": 5.929768897654676e-06,
+      "loss": 2.6638,
+      "step": 558500
+    },
+    {
+      "epoch": 23.563630232264046,
+      "grad_norm": 1.2186017036437988,
+      "learning_rate": 5.844016635938773e-06,
+      "loss": 2.6661,
+      "step": 559000
+    },
+    {
+      "epoch": 23.584706824600598,
+      "grad_norm": 1.1857177019119263,
+      "learning_rate": 5.75826437422287e-06,
+      "loss": 2.6663,
+      "step": 559500
+    },
+    {
+      "epoch": 23.60578341693715,
+      "grad_norm": 1.2447664737701416,
+      "learning_rate": 5.672512112506967e-06,
+      "loss": 2.6665,
+      "step": 560000
+    },
+    {
+      "epoch": 23.60578341693715,
+      "eval_accuracy": 0.7552870707086258,
+      "eval_loss": 2.462599039077759,
+      "eval_runtime": 390.0508,
+      "eval_samples_per_second": 309.462,
+      "eval_steps_per_second": 3.225,
+      "step": 560000
+    },
+    {
+      "epoch": 23.6268600092737,
+      "grad_norm": 1.2108244895935059,
+      "learning_rate": 5.5867598507910655e-06,
+      "loss": 2.6655,
+      "step": 560500
+    },
+    {
+      "epoch": 23.64793660161025,
+      "grad_norm": 1.217085599899292,
+      "learning_rate": 5.501007589075162e-06,
+      "loss": 2.6653,
+      "step": 561000
+    },
+    {
+      "epoch": 23.669013193946803,
+      "grad_norm": 1.2652370929718018,
+      "learning_rate": 5.41525532735926e-06,
+      "loss": 2.6666,
+      "step": 561500
+    },
+    {
+      "epoch": 23.690089786283355,
+      "grad_norm": 1.2941522598266602,
+      "learning_rate": 5.329503065643356e-06,
+      "loss": 2.6624,
+      "step": 562000
+    },
+    {
+      "epoch": 23.711166378619904,
+      "grad_norm": 1.2568550109863281,
+      "learning_rate": 5.243750803927454e-06,
+      "loss": 2.6642,
+      "step": 562500
+    },
+    {
+      "epoch": 23.732242970956456,
+      "grad_norm": 1.2126754522323608,
+      "learning_rate": 5.157998542211551e-06,
+      "loss": 2.6666,
+      "step": 563000
+    },
+    {
+      "epoch": 23.753319563293008,
+      "grad_norm": 1.2100588083267212,
+      "learning_rate": 5.072246280495649e-06,
+      "loss": 2.6638,
+      "step": 563500
+    },
+    {
+      "epoch": 23.774396155629557,
+      "grad_norm": 1.204321265220642,
+      "learning_rate": 4.986494018779746e-06,
+      "loss": 2.6673,
+      "step": 564000
+    },
+    {
+      "epoch": 23.79547274796611,
+      "grad_norm": 1.2687700986862183,
+      "learning_rate": 4.900741757063843e-06,
+      "loss": 2.6632,
+      "step": 564500
+    },
+    {
+      "epoch": 23.81654934030266,
+      "grad_norm": 1.2125352621078491,
+      "learning_rate": 4.81498949534794e-06,
+      "loss": 2.6635,
+      "step": 565000
+    },
+    {
+      "epoch": 23.83762593263921,
+      "grad_norm": 1.1975284814834595,
+      "learning_rate": 4.729237233632038e-06,
+      "loss": 2.6627,
+      "step": 565500
+    },
+    {
+      "epoch": 23.85870252497576,
+      "grad_norm": 1.2041300535202026,
+      "learning_rate": 4.643484971916134e-06,
+      "loss": 2.6665,
+      "step": 566000
+    },
+    {
+      "epoch": 23.879779117312314,
+      "grad_norm": 1.2488267421722412,
+      "learning_rate": 4.557732710200232e-06,
+      "loss": 2.6627,
+      "step": 566500
+    },
+    {
+      "epoch": 23.900855709648862,
+      "grad_norm": 1.2528446912765503,
+      "learning_rate": 4.471980448484329e-06,
+      "loss": 2.6622,
+      "step": 567000
+    },
+    {
+      "epoch": 23.921932301985414,
+      "grad_norm": 1.2793368101119995,
+      "learning_rate": 4.386228186768427e-06,
+      "loss": 2.663,
+      "step": 567500
+    },
+    {
+      "epoch": 23.943008894321967,
+      "grad_norm": 1.2312911748886108,
+      "learning_rate": 4.300475925052524e-06,
+      "loss": 2.6632,
+      "step": 568000
+    },
+    {
+      "epoch": 23.96408548665852,
+      "grad_norm": 1.2059766054153442,
+      "learning_rate": 4.214723663336621e-06,
+      "loss": 2.6671,
+      "step": 568500
+    },
+    {
+      "epoch": 23.985162078995067,
+      "grad_norm": 1.2856242656707764,
+      "learning_rate": 4.128971401620718e-06,
+      "loss": 2.661,
+      "step": 569000
+    },
+    {
+      "epoch": 24.00623867133162,
+      "grad_norm": 1.2255595922470093,
+      "learning_rate": 4.043219139904815e-06,
+      "loss": 2.6638,
+      "step": 569500
+    },
+    {
+      "epoch": 24.02731526366817,
+      "grad_norm": 1.2126749753952026,
+      "learning_rate": 3.957466878188912e-06,
+      "loss": 2.6622,
+      "step": 570000
+    },
+    {
+      "epoch": 24.02731526366817,
+      "eval_accuracy": 0.755508287382015,
+      "eval_loss": 2.4618396759033203,
+      "eval_runtime": 428.4095,
+      "eval_samples_per_second": 281.754,
+      "eval_steps_per_second": 2.936,
+      "step": 570000
+    },
+    {
+      "epoch": 24.04839185600472,
+      "grad_norm": 1.2358497381210327,
+      "learning_rate": 3.87171461647301e-06,
+      "loss": 2.6651,
+      "step": 570500
+    },
+    {
+      "epoch": 24.069468448341272,
+      "grad_norm": 1.204716444015503,
+      "learning_rate": 3.7859623547571067e-06,
+      "loss": 2.6599,
+      "step": 571000
+    },
+    {
+      "epoch": 24.090545040677824,
+      "grad_norm": 1.2521451711654663,
+      "learning_rate": 3.7002100930412046e-06,
+      "loss": 2.6583,
+      "step": 571500
+    },
+    {
+      "epoch": 24.111621633014373,
+      "grad_norm": 1.2290995121002197,
+      "learning_rate": 3.614457831325301e-06,
+      "loss": 2.6671,
+      "step": 572000
+    },
+    {
+      "epoch": 24.132698225350925,
+      "grad_norm": 1.2650110721588135,
+      "learning_rate": 3.5287055696093987e-06,
+      "loss": 2.6625,
+      "step": 572500
+    },
+    {
+      "epoch": 24.153774817687477,
+      "grad_norm": 1.2165331840515137,
+      "learning_rate": 3.4429533078934957e-06,
+      "loss": 2.6618,
+      "step": 573000
+    },
+    {
+      "epoch": 24.174851410024026,
+      "grad_norm": 1.234910249710083,
+      "learning_rate": 3.357201046177593e-06,
+      "loss": 2.6659,
+      "step": 573500
+    },
+    {
+      "epoch": 24.195928002360578,
+      "grad_norm": 1.2204464673995972,
+      "learning_rate": 3.2714487844616906e-06,
+      "loss": 2.6633,
+      "step": 574000
+    },
+    {
+      "epoch": 24.21700459469713,
+      "grad_norm": 1.2501003742218018,
+      "learning_rate": 3.1856965227457872e-06,
+      "loss": 2.6644,
+      "step": 574500
+    },
+    {
+      "epoch": 24.238081187033682,
+      "grad_norm": 1.2189737558364868,
+      "learning_rate": 3.099944261029885e-06,
+      "loss": 2.6656,
+      "step": 575000
+    },
+    {
+      "epoch": 24.25915777937023,
+      "grad_norm": 1.272934913635254,
+      "learning_rate": 3.014191999313982e-06,
+      "loss": 2.6624,
+      "step": 575500
+    },
+    {
+      "epoch": 24.280234371706783,
+      "grad_norm": 1.215233564376831,
+      "learning_rate": 2.928439737598079e-06,
+      "loss": 2.6588,
+      "step": 576000
+    },
+    {
+      "epoch": 24.301310964043335,
+      "grad_norm": 1.2652108669281006,
+      "learning_rate": 2.8426874758821767e-06,
+      "loss": 2.6629,
+      "step": 576500
+    },
+    {
+      "epoch": 24.322387556379883,
+      "grad_norm": 1.266037106513977,
+      "learning_rate": 2.7569352141662737e-06,
+      "loss": 2.6578,
+      "step": 577000
+    },
+    {
+      "epoch": 24.343464148716436,
+      "grad_norm": 1.2390358448028564,
+      "learning_rate": 2.6711829524503707e-06,
+      "loss": 2.6595,
+      "step": 577500
+    },
+    {
+      "epoch": 24.364540741052988,
+      "grad_norm": 1.2625863552093506,
+      "learning_rate": 2.585430690734468e-06,
+      "loss": 2.6589,
+      "step": 578000
+    },
+    {
+      "epoch": 24.385617333389536,
+      "grad_norm": 1.1894887685775757,
+      "learning_rate": 2.4996784290185657e-06,
+      "loss": 2.662,
+      "step": 578500
+    },
+    {
+      "epoch": 24.40669392572609,
+      "grad_norm": 1.2772036790847778,
+      "learning_rate": 2.4139261673026627e-06,
+      "loss": 2.6581,
+      "step": 579000
+    },
+    {
+      "epoch": 24.42777051806264,
+      "grad_norm": 1.214478611946106,
+      "learning_rate": 2.3281739055867597e-06,
+      "loss": 2.6604,
+      "step": 579500
+    },
+    {
+      "epoch": 24.44884711039919,
+      "grad_norm": 1.2057948112487793,
+      "learning_rate": 2.242421643870857e-06,
+      "loss": 2.6612,
+      "step": 580000
+    },
+    {
+      "epoch": 24.44884711039919,
+      "eval_accuracy": 0.7561029266191164,
+      "eval_loss": 2.457956314086914,
+      "eval_runtime": 395.579,
+      "eval_samples_per_second": 305.138,
+      "eval_steps_per_second": 3.18,
+      "step": 580000
+    },
+    {
+      "epoch": 24.46992370273574,
+      "grad_norm": 1.2924978733062744,
+      "learning_rate": 2.1566693821549547e-06,
+      "loss": 2.659,
+      "step": 580500
+    },
+    {
+      "epoch": 24.491000295072293,
+      "grad_norm": 1.2107150554656982,
+      "learning_rate": 2.0709171204390517e-06,
+      "loss": 2.6631,
+      "step": 581000
+    },
+    {
+      "epoch": 24.512076887408845,
+      "grad_norm": 1.2180763483047485,
+      "learning_rate": 1.985164858723149e-06,
+      "loss": 2.6583,
+      "step": 581500
+    },
+    {
+      "epoch": 24.533153479745394,
+      "grad_norm": 1.283178687095642,
+      "learning_rate": 1.8994125970072462e-06,
+      "loss": 2.6641,
+      "step": 582000
+    },
+    {
+      "epoch": 24.554230072081946,
+      "grad_norm": 1.21387779712677,
+      "learning_rate": 1.8136603352913432e-06,
+      "loss": 2.6584,
+      "step": 582500
+    },
+    {
+      "epoch": 24.5753066644185,
+      "grad_norm": 1.1845364570617676,
+      "learning_rate": 1.7279080735754405e-06,
+      "loss": 2.6616,
+      "step": 583000
+    },
+    {
+      "epoch": 24.596383256755047,
+      "grad_norm": 1.2499839067459106,
+      "learning_rate": 1.642155811859538e-06,
+      "loss": 2.6625,
+      "step": 583500
+    },
+    {
+      "epoch": 24.6174598490916,
+      "grad_norm": 1.218228816986084,
+      "learning_rate": 1.5564035501436352e-06,
+      "loss": 2.6608,
+      "step": 584000
+    },
+    {
+      "epoch": 24.63853644142815,
+      "grad_norm": 1.2125355005264282,
+      "learning_rate": 1.4706512884277324e-06,
+      "loss": 2.6619,
+      "step": 584500
+    },
+    {
+      "epoch": 24.6596130337647,
+      "grad_norm": 1.1831626892089844,
+      "learning_rate": 1.3848990267118295e-06,
+      "loss": 2.6578,
+      "step": 585000
+    },
+    {
+      "epoch": 24.680689626101252,
+      "grad_norm": 1.1932088136672974,
+      "learning_rate": 1.299146764995927e-06,
+      "loss": 2.66,
+      "step": 585500
+    },
+    {
+      "epoch": 24.701766218437804,
+      "grad_norm": 1.2371885776519775,
+      "learning_rate": 1.213394503280024e-06,
+      "loss": 2.6583,
+      "step": 586000
+    },
+    {
+      "epoch": 24.722842810774353,
+      "grad_norm": 1.2574782371520996,
+      "learning_rate": 1.1276422415641212e-06,
+      "loss": 2.6596,
+      "step": 586500
+    },
+    {
+      "epoch": 24.743919403110905,
+      "grad_norm": 1.2201411724090576,
+      "learning_rate": 1.0418899798482187e-06,
+      "loss": 2.661,
+      "step": 587000
+    },
+    {
+      "epoch": 24.764995995447457,
+      "grad_norm": 1.2513254880905151,
+      "learning_rate": 9.561377181323157e-07,
+      "loss": 2.6572,
+      "step": 587500
+    },
+    {
+      "epoch": 24.786072587784005,
+      "grad_norm": 1.2408982515335083,
+      "learning_rate": 8.70385456416413e-07,
+      "loss": 2.6583,
+      "step": 588000
+    },
+    {
+      "epoch": 24.807149180120557,
+      "grad_norm": 1.1847333908081055,
+      "learning_rate": 7.846331947005103e-07,
+      "loss": 2.6597,
+      "step": 588500
+    },
+    {
+      "epoch": 24.82822577245711,
+      "grad_norm": 1.2239362001419067,
+      "learning_rate": 6.988809329846075e-07,
+      "loss": 2.6568,
+      "step": 589000
+    },
+    {
+      "epoch": 24.84930236479366,
+      "grad_norm": 1.2334727048873901,
+      "learning_rate": 6.131286712687047e-07,
+      "loss": 2.6615,
+      "step": 589500
+    },
+    {
+      "epoch": 24.87037895713021,
+      "grad_norm": 1.3068723678588867,
+      "learning_rate": 5.27376409552802e-07,
+      "loss": 2.6573,
+      "step": 590000
+    },
+    {
+      "epoch": 24.87037895713021,
+      "eval_accuracy": 0.7559469170014751,
+      "eval_loss": 2.4581093788146973,
+      "eval_runtime": 391.1938,
+      "eval_samples_per_second": 308.558,
+      "eval_steps_per_second": 3.216,
+      "step": 590000
+    },
+    {
+      "epoch": 24.891455549466762,
+      "grad_norm": 1.2096562385559082,
+      "learning_rate": 4.4162414783689917e-07,
+      "loss": 2.6592,
+      "step": 590500
+    },
+    {
+      "epoch": 24.912532141803315,
+      "grad_norm": 1.2075059413909912,
+      "learning_rate": 3.558718861209964e-07,
+      "loss": 2.6589,
+      "step": 591000
+    },
+    {
+      "epoch": 24.933608734139863,
+      "grad_norm": 1.254931092262268,
+      "learning_rate": 2.701196244050937e-07,
+      "loss": 2.6591,
+      "step": 591500
+    },
+    {
+      "epoch": 24.954685326476415,
+      "grad_norm": 1.2780890464782715,
+      "learning_rate": 1.8436736268919094e-07,
+      "loss": 2.6589,
+      "step": 592000
+    },
+    {
+      "epoch": 24.975761918812967,
+      "grad_norm": 1.2007554769515991,
+      "learning_rate": 9.861510097328816e-08,
+      "loss": 2.6574,
+      "step": 592500
+    },
+    {
+      "epoch": 24.996838511149516,
+      "grad_norm": 1.2141892910003662,
+      "learning_rate": 1.2862839257385414e-08,
+      "loss": 2.6586,
+      "step": 593000
+    },
+    {
+      "epoch": 25.0,
+      "step": 593075,
+      "total_flos": 1.5089151321197107e+19,
+      "train_loss": 3.3307897637649497,
+      "train_runtime": 296764.7539,
+      "train_samples_per_second": 191.847,
+      "train_steps_per_second": 1.998
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 593075,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 25,
+  "save_steps": 10000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5089151321197107e+19,
+  "train_batch_size": 96,
+  "trial_name": null,
+  "trial_params": null
+}