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": 3.5705835819244385,
+  "best_model_checkpoint": "distilbert_lda_20_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": 2.1125710010528564,
+      "learning_rate": 5e-06,
+      "loss": 11.9854,
+      "step": 500
+    },
+    {
+      "epoch": 0.04215318467310205,
+      "grad_norm": 0.957831084728241,
+      "learning_rate": 1e-05,
+      "loss": 9.7005,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06322977700965308,
+      "grad_norm": 0.8790013790130615,
+      "learning_rate": 1.5e-05,
+      "loss": 9.0664,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0843063693462041,
+      "grad_norm": 1.1625961065292358,
+      "learning_rate": 2e-05,
+      "loss": 8.9307,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10538296168275513,
+      "grad_norm": 0.9050658345222473,
+      "learning_rate": 2.5e-05,
+      "loss": 8.8632,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12645955401930617,
+      "grad_norm": 0.9540181756019592,
+      "learning_rate": 3e-05,
+      "loss": 8.8209,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14753614635585718,
+      "grad_norm": 0.9363327622413635,
+      "learning_rate": 3.5e-05,
+      "loss": 8.7916,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1686127386924082,
+      "grad_norm": 0.9414223432540894,
+      "learning_rate": 4e-05,
+      "loss": 8.7647,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18968933102895924,
+      "grad_norm": 1.0075184106826782,
+      "learning_rate": 4.5e-05,
+      "loss": 8.7389,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21076592336551025,
+      "grad_norm": 0.998175323009491,
+      "learning_rate": 5e-05,
+      "loss": 8.7221,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2318425157020613,
+      "grad_norm": 0.8386073112487793,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 8.7119,
+      "step": 5500
+    },
+    {
+      "epoch": 0.25291910803861234,
+      "grad_norm": 0.9923796057701111,
+      "learning_rate": 6e-05,
+      "loss": 8.6896,
+      "step": 6000
+    },
+    {
+      "epoch": 0.27399570037516335,
+      "grad_norm": 0.9337378144264221,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 8.6652,
+      "step": 6500
+    },
+    {
+      "epoch": 0.29507229271171437,
+      "grad_norm": 1.0165833234786987,
+      "learning_rate": 7e-05,
+      "loss": 8.6353,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3161488850482654,
+      "grad_norm": 1.2013641595840454,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 8.5853,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3372254773848164,
+      "grad_norm": 1.4257361888885498,
+      "learning_rate": 8e-05,
+      "loss": 8.4962,
+      "step": 8000
+    },
+    {
+      "epoch": 0.35830206972136747,
+      "grad_norm": 1.541422963142395,
+      "learning_rate": 8.5e-05,
+      "loss": 8.363,
+      "step": 8500
+    },
+    {
+      "epoch": 0.3793786620579185,
+      "grad_norm": 1.7964773178100586,
+      "learning_rate": 9e-05,
+      "loss": 8.1965,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4004552543944695,
+      "grad_norm": 2.672678232192993,
+      "learning_rate": 9.5e-05,
+      "loss": 7.8403,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4215318467310205,
+      "grad_norm": 2.47609806060791,
+      "learning_rate": 0.0001,
+      "loss": 7.2789,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4215318467310205,
+      "eval_accuracy": 0.36781781219989773,
+      "eval_loss": 6.573305130004883,
+      "eval_runtime": 504.8686,
+      "eval_samples_per_second": 239.084,
+      "eval_steps_per_second": 2.492,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4426084390675716,
+      "grad_norm": 1.7028785943984985,
+      "learning_rate": 9.99142477382841e-05,
+      "loss": 6.683,
+      "step": 10500
+    },
+    {
+      "epoch": 0.4636850314041226,
+      "grad_norm": 1.6939502954483032,
+      "learning_rate": 9.98284954765682e-05,
+      "loss": 6.2549,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4847616237406736,
+      "grad_norm": 1.6161152124404907,
+      "learning_rate": 9.97427432148523e-05,
+      "loss": 6.0382,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5058382160772247,
+      "grad_norm": 1.4230468273162842,
+      "learning_rate": 9.96569909531364e-05,
+      "loss": 5.8807,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5269148084137757,
+      "grad_norm": 1.282158374786377,
+      "learning_rate": 9.957123869142048e-05,
+      "loss": 5.763,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5479914007503267,
+      "grad_norm": 1.4138014316558838,
+      "learning_rate": 9.94854864297046e-05,
+      "loss": 5.6695,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5690679930868777,
+      "grad_norm": 1.3267381191253662,
+      "learning_rate": 9.939973416798869e-05,
+      "loss": 5.582,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5901445854234287,
+      "grad_norm": 1.2328766584396362,
+      "learning_rate": 9.931398190627278e-05,
+      "loss": 5.5158,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6112211777599798,
+      "grad_norm": 1.3452967405319214,
+      "learning_rate": 9.922822964455688e-05,
+      "loss": 5.4422,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6322977700965308,
+      "grad_norm": 1.3835461139678955,
+      "learning_rate": 9.914247738284099e-05,
+      "loss": 5.3845,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6533743624330818,
+      "grad_norm": 1.2751452922821045,
+      "learning_rate": 9.905672512112507e-05,
+      "loss": 5.3282,
+      "step": 15500
+    },
+    {
+      "epoch": 0.6744509547696328,
+      "grad_norm": 1.3650610446929932,
+      "learning_rate": 9.897097285940917e-05,
+      "loss": 5.2803,
+      "step": 16000
+    },
+    {
+      "epoch": 0.6955275471061839,
+      "grad_norm": 1.3306357860565186,
+      "learning_rate": 9.888522059769327e-05,
+      "loss": 5.2247,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7166041394427349,
+      "grad_norm": 1.2730858325958252,
+      "learning_rate": 9.879946833597737e-05,
+      "loss": 5.1737,
+      "step": 17000
+    },
+    {
+      "epoch": 0.737680731779286,
+      "grad_norm": 1.3529385328292847,
+      "learning_rate": 9.871371607426147e-05,
+      "loss": 5.1267,
+      "step": 17500
+    },
+    {
+      "epoch": 0.758757324115837,
+      "grad_norm": 1.4120761156082153,
+      "learning_rate": 9.862796381254556e-05,
+      "loss": 5.091,
+      "step": 18000
+    },
+    {
+      "epoch": 0.779833916452388,
+      "grad_norm": 1.3667072057724,
+      "learning_rate": 9.854221155082965e-05,
+      "loss": 5.0476,
+      "step": 18500
+    },
+    {
+      "epoch": 0.800910508788939,
+      "grad_norm": 1.3867383003234863,
+      "learning_rate": 9.845645928911376e-05,
+      "loss": 5.0215,
+      "step": 19000
+    },
+    {
+      "epoch": 0.82198710112549,
+      "grad_norm": 1.3622369766235352,
+      "learning_rate": 9.837070702739786e-05,
+      "loss": 4.986,
+      "step": 19500
+    },
+    {
+      "epoch": 0.843063693462041,
+      "grad_norm": 1.3549550771713257,
+      "learning_rate": 9.828495476568194e-05,
+      "loss": 4.955,
+      "step": 20000
+    },
+    {
+      "epoch": 0.843063693462041,
+      "eval_accuracy": 0.5844104039847753,
+      "eval_loss": 4.592887878417969,
+      "eval_runtime": 431.9906,
+      "eval_samples_per_second": 279.418,
+      "eval_steps_per_second": 2.912,
+      "step": 20000
+    },
+    {
+      "epoch": 0.864140285798592,
+      "grad_norm": 1.4442979097366333,
+      "learning_rate": 9.819920250396604e-05,
+      "loss": 4.9252,
+      "step": 20500
+    },
+    {
+      "epoch": 0.8852168781351432,
+      "grad_norm": 1.3139088153839111,
+      "learning_rate": 9.811345024225015e-05,
+      "loss": 4.8993,
+      "step": 21000
+    },
+    {
+      "epoch": 0.9062934704716942,
+      "grad_norm": 1.3067762851715088,
+      "learning_rate": 9.802769798053424e-05,
+      "loss": 4.88,
+      "step": 21500
+    },
+    {
+      "epoch": 0.9273700628082452,
+      "grad_norm": 1.3485420942306519,
+      "learning_rate": 9.794194571881834e-05,
+      "loss": 4.8515,
+      "step": 22000
+    },
+    {
+      "epoch": 0.9484466551447962,
+      "grad_norm": 1.4817858934402466,
+      "learning_rate": 9.785619345710244e-05,
+      "loss": 4.8298,
+      "step": 22500
+    },
+    {
+      "epoch": 0.9695232474813472,
+      "grad_norm": 1.4857697486877441,
+      "learning_rate": 9.777044119538653e-05,
+      "loss": 4.8071,
+      "step": 23000
+    },
+    {
+      "epoch": 0.9905998398178982,
+      "grad_norm": 1.4700037240982056,
+      "learning_rate": 9.768468893367063e-05,
+      "loss": 4.7904,
+      "step": 23500
+    },
+    {
+      "epoch": 1.0116764321544494,
+      "grad_norm": 1.3063980340957642,
+      "learning_rate": 9.759893667195473e-05,
+      "loss": 4.7751,
+      "step": 24000
+    },
+    {
+      "epoch": 1.0327530244910004,
+      "grad_norm": 1.3363605737686157,
+      "learning_rate": 9.751318441023883e-05,
+      "loss": 4.7581,
+      "step": 24500
+    },
+    {
+      "epoch": 1.0538296168275514,
+      "grad_norm": 1.4756698608398438,
+      "learning_rate": 9.742743214852293e-05,
+      "loss": 4.7401,
+      "step": 25000
+    },
+    {
+      "epoch": 1.0749062091641024,
+      "grad_norm": 1.4594119787216187,
+      "learning_rate": 9.734167988680703e-05,
+      "loss": 4.7254,
+      "step": 25500
+    },
+    {
+      "epoch": 1.0959828015006534,
+      "grad_norm": 1.321815013885498,
+      "learning_rate": 9.725592762509111e-05,
+      "loss": 4.7091,
+      "step": 26000
+    },
+    {
+      "epoch": 1.1170593938372044,
+      "grad_norm": 1.299394130706787,
+      "learning_rate": 9.717017536337522e-05,
+      "loss": 4.6962,
+      "step": 26500
+    },
+    {
+      "epoch": 1.1381359861737554,
+      "grad_norm": 1.3042415380477905,
+      "learning_rate": 9.708442310165932e-05,
+      "loss": 4.6871,
+      "step": 27000
+    },
+    {
+      "epoch": 1.1592125785103065,
+      "grad_norm": 1.2775287628173828,
+      "learning_rate": 9.69986708399434e-05,
+      "loss": 4.6723,
+      "step": 27500
+    },
+    {
+      "epoch": 1.1802891708468575,
+      "grad_norm": 1.3007296323776245,
+      "learning_rate": 9.69129185782275e-05,
+      "loss": 4.6658,
+      "step": 28000
+    },
+    {
+      "epoch": 1.2013657631834085,
+      "grad_norm": 1.2816871404647827,
+      "learning_rate": 9.68271663165116e-05,
+      "loss": 4.6479,
+      "step": 28500
+    },
+    {
+      "epoch": 1.2224423555199595,
+      "grad_norm": 1.3169293403625488,
+      "learning_rate": 9.67414140547957e-05,
+      "loss": 4.6384,
+      "step": 29000
+    },
+    {
+      "epoch": 1.2435189478565105,
+      "grad_norm": 1.4479376077651978,
+      "learning_rate": 9.66556617930798e-05,
+      "loss": 4.6203,
+      "step": 29500
+    },
+    {
+      "epoch": 1.2645955401930615,
+      "grad_norm": 1.2178654670715332,
+      "learning_rate": 9.656990953136388e-05,
+      "loss": 4.6182,
+      "step": 30000
+    },
+    {
+      "epoch": 1.2645955401930615,
+      "eval_accuracy": 0.6199481702154195,
+      "eval_loss": 4.301407337188721,
+      "eval_runtime": 496.4338,
+      "eval_samples_per_second": 243.146,
+      "eval_steps_per_second": 2.534,
+      "step": 30000
+    },
+    {
+      "epoch": 1.2856721325296125,
+      "grad_norm": 1.3004441261291504,
+      "learning_rate": 9.6484157269648e-05,
+      "loss": 4.6028,
+      "step": 30500
+    },
+    {
+      "epoch": 1.3067487248661636,
+      "grad_norm": 1.3483167886734009,
+      "learning_rate": 9.639840500793209e-05,
+      "loss": 4.6008,
+      "step": 31000
+    },
+    {
+      "epoch": 1.3278253172027146,
+      "grad_norm": 1.3567254543304443,
+      "learning_rate": 9.631265274621618e-05,
+      "loss": 4.588,
+      "step": 31500
+    },
+    {
+      "epoch": 1.3489019095392658,
+      "grad_norm": 1.379468321800232,
+      "learning_rate": 9.622690048450028e-05,
+      "loss": 4.5851,
+      "step": 32000
+    },
+    {
+      "epoch": 1.3699785018758166,
+      "grad_norm": 1.2407667636871338,
+      "learning_rate": 9.614114822278439e-05,
+      "loss": 4.5734,
+      "step": 32500
+    },
+    {
+      "epoch": 1.3910550942123678,
+      "grad_norm": 1.2499284744262695,
+      "learning_rate": 9.605539596106847e-05,
+      "loss": 4.5701,
+      "step": 33000
+    },
+    {
+      "epoch": 1.4121316865489189,
+      "grad_norm": 1.2258241176605225,
+      "learning_rate": 9.596964369935257e-05,
+      "loss": 4.557,
+      "step": 33500
+    },
+    {
+      "epoch": 1.4332082788854699,
+      "grad_norm": 1.2973231077194214,
+      "learning_rate": 9.588389143763667e-05,
+      "loss": 4.553,
+      "step": 34000
+    },
+    {
+      "epoch": 1.4542848712220209,
+      "grad_norm": 1.200750708580017,
+      "learning_rate": 9.579813917592077e-05,
+      "loss": 4.5429,
+      "step": 34500
+    },
+    {
+      "epoch": 1.475361463558572,
+      "grad_norm": 1.297882318496704,
+      "learning_rate": 9.571238691420487e-05,
+      "loss": 4.5385,
+      "step": 35000
+    },
+    {
+      "epoch": 1.496438055895123,
+      "grad_norm": 1.2399722337722778,
+      "learning_rate": 9.562663465248896e-05,
+      "loss": 4.5286,
+      "step": 35500
+    },
+    {
+      "epoch": 1.517514648231674,
+      "grad_norm": 1.4111851453781128,
+      "learning_rate": 9.554088239077306e-05,
+      "loss": 4.5162,
+      "step": 36000
+    },
+    {
+      "epoch": 1.538591240568225,
+      "grad_norm": 1.223677158355713,
+      "learning_rate": 9.545513012905716e-05,
+      "loss": 4.5131,
+      "step": 36500
+    },
+    {
+      "epoch": 1.559667832904776,
+      "grad_norm": 1.206773042678833,
+      "learning_rate": 9.536937786734126e-05,
+      "loss": 4.5147,
+      "step": 37000
+    },
+    {
+      "epoch": 1.580744425241327,
+      "grad_norm": 1.269779086112976,
+      "learning_rate": 9.528362560562534e-05,
+      "loss": 4.5031,
+      "step": 37500
+    },
+    {
+      "epoch": 1.601821017577878,
+      "grad_norm": 1.4207309484481812,
+      "learning_rate": 9.519787334390946e-05,
+      "loss": 4.4955,
+      "step": 38000
+    },
+    {
+      "epoch": 1.622897609914429,
+      "grad_norm": 1.35732102394104,
+      "learning_rate": 9.511212108219355e-05,
+      "loss": 4.4885,
+      "step": 38500
+    },
+    {
+      "epoch": 1.64397420225098,
+      "grad_norm": 1.3769789934158325,
+      "learning_rate": 9.502636882047764e-05,
+      "loss": 4.4816,
+      "step": 39000
+    },
+    {
+      "epoch": 1.6650507945875312,
+      "grad_norm": 1.2985422611236572,
+      "learning_rate": 9.494061655876174e-05,
+      "loss": 4.477,
+      "step": 39500
+    },
+    {
+      "epoch": 1.686127386924082,
+      "grad_norm": 1.2566149234771729,
+      "learning_rate": 9.485486429704585e-05,
+      "loss": 4.4751,
+      "step": 40000
+    },
+    {
+      "epoch": 1.686127386924082,
+      "eval_accuracy": 0.6386551087013674,
+      "eval_loss": 4.169163703918457,
+      "eval_runtime": 462.3255,
+      "eval_samples_per_second": 261.084,
+      "eval_steps_per_second": 2.721,
+      "step": 40000
+    },
+    {
+      "epoch": 1.7072039792606333,
+      "grad_norm": 1.2702150344848633,
+      "learning_rate": 9.476911203532993e-05,
+      "loss": 4.4677,
+      "step": 40500
+    },
+    {
+      "epoch": 1.728280571597184,
+      "grad_norm": 1.278765320777893,
+      "learning_rate": 9.468335977361403e-05,
+      "loss": 4.4644,
+      "step": 41000
+    },
+    {
+      "epoch": 1.7493571639337353,
+      "grad_norm": 1.2696791887283325,
+      "learning_rate": 9.459760751189813e-05,
+      "loss": 4.4601,
+      "step": 41500
+    },
+    {
+      "epoch": 1.770433756270286,
+      "grad_norm": 1.3403836488723755,
+      "learning_rate": 9.451185525018223e-05,
+      "loss": 4.4539,
+      "step": 42000
+    },
+    {
+      "epoch": 1.7915103486068373,
+      "grad_norm": 1.2827568054199219,
+      "learning_rate": 9.442610298846633e-05,
+      "loss": 4.4494,
+      "step": 42500
+    },
+    {
+      "epoch": 1.8125869409433881,
+      "grad_norm": 1.334587812423706,
+      "learning_rate": 9.434035072675043e-05,
+      "loss": 4.4411,
+      "step": 43000
+    },
+    {
+      "epoch": 1.8336635332799394,
+      "grad_norm": 1.2485697269439697,
+      "learning_rate": 9.425459846503451e-05,
+      "loss": 4.4357,
+      "step": 43500
+    },
+    {
+      "epoch": 1.8547401256164904,
+      "grad_norm": 1.2948163747787476,
+      "learning_rate": 9.416884620331862e-05,
+      "loss": 4.431,
+      "step": 44000
+    },
+    {
+      "epoch": 1.8758167179530414,
+      "grad_norm": 1.3448460102081299,
+      "learning_rate": 9.408309394160272e-05,
+      "loss": 4.4316,
+      "step": 44500
+    },
+    {
+      "epoch": 1.8968933102895924,
+      "grad_norm": 1.306273341178894,
+      "learning_rate": 9.39973416798868e-05,
+      "loss": 4.4207,
+      "step": 45000
+    },
+    {
+      "epoch": 1.9179699026261434,
+      "grad_norm": 1.218953013420105,
+      "learning_rate": 9.39115894181709e-05,
+      "loss": 4.4209,
+      "step": 45500
+    },
+    {
+      "epoch": 1.9390464949626944,
+      "grad_norm": 1.326711893081665,
+      "learning_rate": 9.382583715645502e-05,
+      "loss": 4.4111,
+      "step": 46000
+    },
+    {
+      "epoch": 1.9601230872992454,
+      "grad_norm": 1.3056893348693848,
+      "learning_rate": 9.37400848947391e-05,
+      "loss": 4.4076,
+      "step": 46500
+    },
+    {
+      "epoch": 1.9811996796357965,
+      "grad_norm": 1.2805821895599365,
+      "learning_rate": 9.36543326330232e-05,
+      "loss": 4.4068,
+      "step": 47000
+    },
+    {
+      "epoch": 2.0022762719723475,
+      "grad_norm": 1.3365784883499146,
+      "learning_rate": 9.35685803713073e-05,
+      "loss": 4.4052,
+      "step": 47500
+    },
+    {
+      "epoch": 2.0233528643088987,
+      "grad_norm": 1.2664375305175781,
+      "learning_rate": 9.34828281095914e-05,
+      "loss": 4.3915,
+      "step": 48000
+    },
+    {
+      "epoch": 2.0444294566454495,
+      "grad_norm": 1.2173092365264893,
+      "learning_rate": 9.339707584787549e-05,
+      "loss": 4.3898,
+      "step": 48500
+    },
+    {
+      "epoch": 2.0655060489820007,
+      "grad_norm": 1.3102564811706543,
+      "learning_rate": 9.331132358615959e-05,
+      "loss": 4.3821,
+      "step": 49000
+    },
+    {
+      "epoch": 2.0865826413185515,
+      "grad_norm": 1.2926936149597168,
+      "learning_rate": 9.322557132444369e-05,
+      "loss": 4.3782,
+      "step": 49500
+    },
+    {
+      "epoch": 2.1076592336551028,
+      "grad_norm": 1.3074634075164795,
+      "learning_rate": 9.313981906272779e-05,
+      "loss": 4.3802,
+      "step": 50000
+    },
+    {
+      "epoch": 2.1076592336551028,
+      "eval_accuracy": 0.6498215028393357,
+      "eval_loss": 4.0856475830078125,
+      "eval_runtime": 493.7892,
+      "eval_samples_per_second": 244.448,
+      "eval_steps_per_second": 2.548,
+      "step": 50000
+    },
+    {
+      "epoch": 2.1287358259916536,
+      "grad_norm": 1.257688283920288,
+      "learning_rate": 9.305406680101189e-05,
+      "loss": 4.3805,
+      "step": 50500
+    },
+    {
+      "epoch": 2.149812418328205,
+      "grad_norm": 1.2499240636825562,
+      "learning_rate": 9.296831453929597e-05,
+      "loss": 4.3722,
+      "step": 51000
+    },
+    {
+      "epoch": 2.1708890106647556,
+      "grad_norm": 1.2979816198349,
+      "learning_rate": 9.288256227758008e-05,
+      "loss": 4.3653,
+      "step": 51500
+    },
+    {
+      "epoch": 2.191965603001307,
+      "grad_norm": 1.2507808208465576,
+      "learning_rate": 9.279681001586418e-05,
+      "loss": 4.3653,
+      "step": 52000
+    },
+    {
+      "epoch": 2.2130421953378576,
+      "grad_norm": 1.2701256275177002,
+      "learning_rate": 9.271105775414827e-05,
+      "loss": 4.3576,
+      "step": 52500
+    },
+    {
+      "epoch": 2.234118787674409,
+      "grad_norm": 1.3122810125350952,
+      "learning_rate": 9.262530549243236e-05,
+      "loss": 4.3593,
+      "step": 53000
+    },
+    {
+      "epoch": 2.2551953800109596,
+      "grad_norm": 1.2580366134643555,
+      "learning_rate": 9.253955323071646e-05,
+      "loss": 4.3561,
+      "step": 53500
+    },
+    {
+      "epoch": 2.276271972347511,
+      "grad_norm": 1.337925672531128,
+      "learning_rate": 9.245380096900056e-05,
+      "loss": 4.3565,
+      "step": 54000
+    },
+    {
+      "epoch": 2.2973485646840617,
+      "grad_norm": 1.3932888507843018,
+      "learning_rate": 9.236804870728466e-05,
+      "loss": 4.3448,
+      "step": 54500
+    },
+    {
+      "epoch": 2.318425157020613,
+      "grad_norm": 1.2486608028411865,
+      "learning_rate": 9.228229644556876e-05,
+      "loss": 4.3458,
+      "step": 55000
+    },
+    {
+      "epoch": 2.339501749357164,
+      "grad_norm": 1.296960711479187,
+      "learning_rate": 9.219654418385286e-05,
+      "loss": 4.3379,
+      "step": 55500
+    },
+    {
+      "epoch": 2.360578341693715,
+      "grad_norm": 1.2127395868301392,
+      "learning_rate": 9.211079192213695e-05,
+      "loss": 4.344,
+      "step": 56000
+    },
+    {
+      "epoch": 2.381654934030266,
+      "grad_norm": 1.218783974647522,
+      "learning_rate": 9.202503966042105e-05,
+      "loss": 4.3373,
+      "step": 56500
+    },
+    {
+      "epoch": 2.402731526366817,
+      "grad_norm": 1.3169888257980347,
+      "learning_rate": 9.193928739870514e-05,
+      "loss": 4.3319,
+      "step": 57000
+    },
+    {
+      "epoch": 2.423808118703368,
+      "grad_norm": 1.2963619232177734,
+      "learning_rate": 9.185353513698925e-05,
+      "loss": 4.33,
+      "step": 57500
+    },
+    {
+      "epoch": 2.444884711039919,
+      "grad_norm": 1.2948452234268188,
+      "learning_rate": 9.176778287527335e-05,
+      "loss": 4.3282,
+      "step": 58000
+    },
+    {
+      "epoch": 2.4659613033764702,
+      "grad_norm": 1.2105095386505127,
+      "learning_rate": 9.168203061355743e-05,
+      "loss": 4.326,
+      "step": 58500
+    },
+    {
+      "epoch": 2.487037895713021,
+      "grad_norm": 1.261460781097412,
+      "learning_rate": 9.159627835184153e-05,
+      "loss": 4.3224,
+      "step": 59000
+    },
+    {
+      "epoch": 2.5081144880495723,
+      "grad_norm": 1.2296700477600098,
+      "learning_rate": 9.151052609012564e-05,
+      "loss": 4.3232,
+      "step": 59500
+    },
+    {
+      "epoch": 2.529191080386123,
+      "grad_norm": 1.2361059188842773,
+      "learning_rate": 9.142477382840973e-05,
+      "loss": 4.3162,
+      "step": 60000
+    },
+    {
+      "epoch": 2.529191080386123,
+      "eval_accuracy": 0.6598053358070456,
+      "eval_loss": 4.020321369171143,
+      "eval_runtime": 458.6416,
+      "eval_samples_per_second": 263.182,
+      "eval_steps_per_second": 2.743,
+      "step": 60000
+    },
+    {
+      "epoch": 2.5502676727226743,
+      "grad_norm": 1.259150505065918,
+      "learning_rate": 9.133902156669383e-05,
+      "loss": 4.3129,
+      "step": 60500
+    },
+    {
+      "epoch": 2.571344265059225,
+      "grad_norm": 1.305823564529419,
+      "learning_rate": 9.125326930497792e-05,
+      "loss": 4.3116,
+      "step": 61000
+    },
+    {
+      "epoch": 2.5924208573957763,
+      "grad_norm": 1.2827887535095215,
+      "learning_rate": 9.116751704326202e-05,
+      "loss": 4.3059,
+      "step": 61500
+    },
+    {
+      "epoch": 2.613497449732327,
+      "grad_norm": 1.2761354446411133,
+      "learning_rate": 9.108176478154612e-05,
+      "loss": 4.3,
+      "step": 62000
+    },
+    {
+      "epoch": 2.6345740420688784,
+      "grad_norm": 1.2681866884231567,
+      "learning_rate": 9.09960125198302e-05,
+      "loss": 4.308,
+      "step": 62500
+    },
+    {
+      "epoch": 2.655650634405429,
+      "grad_norm": 1.269339919090271,
+      "learning_rate": 9.091026025811432e-05,
+      "loss": 4.2999,
+      "step": 63000
+    },
+    {
+      "epoch": 2.6767272267419804,
+      "grad_norm": 1.2736406326293945,
+      "learning_rate": 9.082450799639842e-05,
+      "loss": 4.2969,
+      "step": 63500
+    },
+    {
+      "epoch": 2.6978038190785316,
+      "grad_norm": 1.2559821605682373,
+      "learning_rate": 9.07387557346825e-05,
+      "loss": 4.2982,
+      "step": 64000
+    },
+    {
+      "epoch": 2.7188804114150824,
+      "grad_norm": 1.3215770721435547,
+      "learning_rate": 9.06530034729666e-05,
+      "loss": 4.2935,
+      "step": 64500
+    },
+    {
+      "epoch": 2.739957003751633,
+      "grad_norm": 1.340651273727417,
+      "learning_rate": 9.05672512112507e-05,
+      "loss": 4.2904,
+      "step": 65000
+    },
+    {
+      "epoch": 2.7610335960881844,
+      "grad_norm": 1.4504210948944092,
+      "learning_rate": 9.04814989495348e-05,
+      "loss": 4.2876,
+      "step": 65500
+    },
+    {
+      "epoch": 2.7821101884247357,
+      "grad_norm": 1.2401155233383179,
+      "learning_rate": 9.03957466878189e-05,
+      "loss": 4.2908,
+      "step": 66000
+    },
+    {
+      "epoch": 2.8031867807612865,
+      "grad_norm": 1.412730097770691,
+      "learning_rate": 9.030999442610299e-05,
+      "loss": 4.2822,
+      "step": 66500
+    },
+    {
+      "epoch": 2.8242633730978377,
+      "grad_norm": 1.2546416521072388,
+      "learning_rate": 9.022424216438709e-05,
+      "loss": 4.2781,
+      "step": 67000
+    },
+    {
+      "epoch": 2.8453399654343885,
+      "grad_norm": 1.2268093824386597,
+      "learning_rate": 9.013848990267119e-05,
+      "loss": 4.2762,
+      "step": 67500
+    },
+    {
+      "epoch": 2.8664165577709397,
+      "grad_norm": 1.2974580526351929,
+      "learning_rate": 9.005273764095529e-05,
+      "loss": 4.2735,
+      "step": 68000
+    },
+    {
+      "epoch": 2.8874931501074905,
+      "grad_norm": 1.3605877161026,
+      "learning_rate": 8.996698537923937e-05,
+      "loss": 4.2767,
+      "step": 68500
+    },
+    {
+      "epoch": 2.9085697424440418,
+      "grad_norm": 1.247268557548523,
+      "learning_rate": 8.988123311752348e-05,
+      "loss": 4.2698,
+      "step": 69000
+    },
+    {
+      "epoch": 2.9296463347805926,
+      "grad_norm": 1.3056007623672485,
+      "learning_rate": 8.979548085580758e-05,
+      "loss": 4.2698,
+      "step": 69500
+    },
+    {
+      "epoch": 2.950722927117144,
+      "grad_norm": 1.215798258781433,
+      "learning_rate": 8.970972859409167e-05,
+      "loss": 4.2637,
+      "step": 70000
+    },
+    {
+      "epoch": 2.950722927117144,
+      "eval_accuracy": 0.665788232001193,
+      "eval_loss": 3.9797568321228027,
+      "eval_runtime": 545.814,
+      "eval_samples_per_second": 221.149,
+      "eval_steps_per_second": 2.305,
+      "step": 70000
+    },
+    {
+      "epoch": 2.9717995194536946,
+      "grad_norm": 1.241657018661499,
+      "learning_rate": 8.962397633237576e-05,
+      "loss": 4.2655,
+      "step": 70500
+    },
+    {
+      "epoch": 2.992876111790246,
+      "grad_norm": 1.2538546323776245,
+      "learning_rate": 8.953822407065988e-05,
+      "loss": 4.2611,
+      "step": 71000
+    },
+    {
+      "epoch": 3.0139527041267966,
+      "grad_norm": 1.2865099906921387,
+      "learning_rate": 8.945247180894396e-05,
+      "loss": 4.2599,
+      "step": 71500
+    },
+    {
+      "epoch": 3.035029296463348,
+      "grad_norm": 1.2967705726623535,
+      "learning_rate": 8.936671954722806e-05,
+      "loss": 4.2547,
+      "step": 72000
+    },
+    {
+      "epoch": 3.0561058887998986,
+      "grad_norm": 1.2914801836013794,
+      "learning_rate": 8.928096728551216e-05,
+      "loss": 4.2574,
+      "step": 72500
+    },
+    {
+      "epoch": 3.07718248113645,
+      "grad_norm": 1.3269233703613281,
+      "learning_rate": 8.919521502379626e-05,
+      "loss": 4.2501,
+      "step": 73000
+    },
+    {
+      "epoch": 3.0982590734730007,
+      "grad_norm": 1.2776845693588257,
+      "learning_rate": 8.910946276208035e-05,
+      "loss": 4.2527,
+      "step": 73500
+    },
+    {
+      "epoch": 3.119335665809552,
+      "grad_norm": 1.2184683084487915,
+      "learning_rate": 8.902371050036445e-05,
+      "loss": 4.2427,
+      "step": 74000
+    },
+    {
+      "epoch": 3.140412258146103,
+      "grad_norm": 1.2680565118789673,
+      "learning_rate": 8.893795823864855e-05,
+      "loss": 4.2493,
+      "step": 74500
+    },
+    {
+      "epoch": 3.161488850482654,
+      "grad_norm": 1.4478607177734375,
+      "learning_rate": 8.885220597693265e-05,
+      "loss": 4.2428,
+      "step": 75000
+    },
+    {
+      "epoch": 3.182565442819205,
+      "grad_norm": 1.300204873085022,
+      "learning_rate": 8.876645371521675e-05,
+      "loss": 4.2418,
+      "step": 75500
+    },
+    {
+      "epoch": 3.203642035155756,
+      "grad_norm": 1.27613365650177,
+      "learning_rate": 8.868070145350083e-05,
+      "loss": 4.2409,
+      "step": 76000
+    },
+    {
+      "epoch": 3.224718627492307,
+      "grad_norm": 1.271437644958496,
+      "learning_rate": 8.859494919178493e-05,
+      "loss": 4.2415,
+      "step": 76500
+    },
+    {
+      "epoch": 3.245795219828858,
+      "grad_norm": 1.3246439695358276,
+      "learning_rate": 8.850919693006904e-05,
+      "loss": 4.2361,
+      "step": 77000
+    },
+    {
+      "epoch": 3.2668718121654092,
+      "grad_norm": 1.202181100845337,
+      "learning_rate": 8.842344466835313e-05,
+      "loss": 4.2336,
+      "step": 77500
+    },
+    {
+      "epoch": 3.28794840450196,
+      "grad_norm": 1.2833359241485596,
+      "learning_rate": 8.833769240663723e-05,
+      "loss": 4.2294,
+      "step": 78000
+    },
+    {
+      "epoch": 3.3090249968385113,
+      "grad_norm": 1.2585039138793945,
+      "learning_rate": 8.825194014492132e-05,
+      "loss": 4.2299,
+      "step": 78500
+    },
+    {
+      "epoch": 3.330101589175062,
+      "grad_norm": 1.312504768371582,
+      "learning_rate": 8.816618788320542e-05,
+      "loss": 4.2292,
+      "step": 79000
+    },
+    {
+      "epoch": 3.3511781815116133,
+      "grad_norm": 1.4075509309768677,
+      "learning_rate": 8.808043562148952e-05,
+      "loss": 4.2279,
+      "step": 79500
+    },
+    {
+      "epoch": 3.372254773848164,
+      "grad_norm": 1.250618577003479,
+      "learning_rate": 8.799468335977362e-05,
+      "loss": 4.225,
+      "step": 80000
+    },
+    {
+      "epoch": 3.372254773848164,
+      "eval_accuracy": 0.6715542055531594,
+      "eval_loss": 3.941906213760376,
+      "eval_runtime": 476.772,
+      "eval_samples_per_second": 253.173,
+      "eval_steps_per_second": 2.639,
+      "step": 80000
+    },
+    {
+      "epoch": 3.3933313661847153,
+      "grad_norm": 1.2815290689468384,
+      "learning_rate": 8.790893109805772e-05,
+      "loss": 4.2255,
+      "step": 80500
+    },
+    {
+      "epoch": 3.414407958521266,
+      "grad_norm": 1.256801962852478,
+      "learning_rate": 8.782317883634182e-05,
+      "loss": 4.2219,
+      "step": 81000
+    },
+    {
+      "epoch": 3.4354845508578173,
+      "grad_norm": 1.3086819648742676,
+      "learning_rate": 8.773742657462591e-05,
+      "loss": 4.2193,
+      "step": 81500
+    },
+    {
+      "epoch": 3.456561143194368,
+      "grad_norm": 1.308670163154602,
+      "learning_rate": 8.765167431291e-05,
+      "loss": 4.2187,
+      "step": 82000
+    },
+    {
+      "epoch": 3.4776377355309194,
+      "grad_norm": 1.338402271270752,
+      "learning_rate": 8.756592205119411e-05,
+      "loss": 4.2154,
+      "step": 82500
+    },
+    {
+      "epoch": 3.4987143278674706,
+      "grad_norm": 1.2115265130996704,
+      "learning_rate": 8.748016978947821e-05,
+      "loss": 4.2167,
+      "step": 83000
+    },
+    {
+      "epoch": 3.5197909202040214,
+      "grad_norm": 1.2838630676269531,
+      "learning_rate": 8.73944175277623e-05,
+      "loss": 4.2145,
+      "step": 83500
+    },
+    {
+      "epoch": 3.540867512540572,
+      "grad_norm": 1.3645919561386108,
+      "learning_rate": 8.730866526604639e-05,
+      "loss": 4.2162,
+      "step": 84000
+    },
+    {
+      "epoch": 3.5619441048771234,
+      "grad_norm": 1.2744603157043457,
+      "learning_rate": 8.72229130043305e-05,
+      "loss": 4.2139,
+      "step": 84500
+    },
+    {
+      "epoch": 3.5830206972136747,
+      "grad_norm": 1.2527600526809692,
+      "learning_rate": 8.713716074261459e-05,
+      "loss": 4.2094,
+      "step": 85000
+    },
+    {
+      "epoch": 3.6040972895502255,
+      "grad_norm": 1.2738324403762817,
+      "learning_rate": 8.705140848089869e-05,
+      "loss": 4.2096,
+      "step": 85500
+    },
+    {
+      "epoch": 3.6251738818867767,
+      "grad_norm": 1.3594094514846802,
+      "learning_rate": 8.696565621918278e-05,
+      "loss": 4.2021,
+      "step": 86000
+    },
+    {
+      "epoch": 3.6462504742233275,
+      "grad_norm": 1.2782235145568848,
+      "learning_rate": 8.687990395746688e-05,
+      "loss": 4.2061,
+      "step": 86500
+    },
+    {
+      "epoch": 3.6673270665598787,
+      "grad_norm": 1.2291390895843506,
+      "learning_rate": 8.679415169575098e-05,
+      "loss": 4.2048,
+      "step": 87000
+    },
+    {
+      "epoch": 3.6884036588964295,
+      "grad_norm": 1.2707864046096802,
+      "learning_rate": 8.670839943403508e-05,
+      "loss": 4.202,
+      "step": 87500
+    },
+    {
+      "epoch": 3.7094802512329808,
+      "grad_norm": 1.3397955894470215,
+      "learning_rate": 8.662264717231916e-05,
+      "loss": 4.2023,
+      "step": 88000
+    },
+    {
+      "epoch": 3.7305568435695315,
+      "grad_norm": 1.2625789642333984,
+      "learning_rate": 8.653689491060328e-05,
+      "loss": 4.2022,
+      "step": 88500
+    },
+    {
+      "epoch": 3.751633435906083,
+      "grad_norm": 1.2575842142105103,
+      "learning_rate": 8.645114264888737e-05,
+      "loss": 4.2012,
+      "step": 89000
+    },
+    {
+      "epoch": 3.7727100282426336,
+      "grad_norm": 1.2701308727264404,
+      "learning_rate": 8.636539038717146e-05,
+      "loss": 4.1938,
+      "step": 89500
+    },
+    {
+      "epoch": 3.793786620579185,
+      "grad_norm": 1.2316595315933228,
+      "learning_rate": 8.627963812545556e-05,
+      "loss": 4.1939,
+      "step": 90000
+    },
+    {
+      "epoch": 3.793786620579185,
+      "eval_accuracy": 0.6760549675711476,
+      "eval_loss": 3.9100935459136963,
+      "eval_runtime": 453.7255,
+      "eval_samples_per_second": 266.033,
+      "eval_steps_per_second": 2.773,
+      "step": 90000
+    },
+    {
+      "epoch": 3.8148632129157356,
+      "grad_norm": 1.3288357257843018,
+      "learning_rate": 8.619388586373967e-05,
+      "loss": 4.1976,
+      "step": 90500
+    },
+    {
+      "epoch": 3.835939805252287,
+      "grad_norm": 1.303652048110962,
+      "learning_rate": 8.610813360202375e-05,
+      "loss": 4.1895,
+      "step": 91000
+    },
+    {
+      "epoch": 3.857016397588838,
+      "grad_norm": 1.289547324180603,
+      "learning_rate": 8.602238134030785e-05,
+      "loss": 4.1922,
+      "step": 91500
+    },
+    {
+      "epoch": 3.878092989925389,
+      "grad_norm": 1.2502143383026123,
+      "learning_rate": 8.593662907859195e-05,
+      "loss": 4.1834,
+      "step": 92000
+    },
+    {
+      "epoch": 3.8991695822619397,
+      "grad_norm": 1.1820487976074219,
+      "learning_rate": 8.585087681687605e-05,
+      "loss": 4.1883,
+      "step": 92500
+    },
+    {
+      "epoch": 3.920246174598491,
+      "grad_norm": 1.2446610927581787,
+      "learning_rate": 8.576512455516015e-05,
+      "loss": 4.1874,
+      "step": 93000
+    },
+    {
+      "epoch": 3.941322766935042,
+      "grad_norm": 1.2163784503936768,
+      "learning_rate": 8.567937229344425e-05,
+      "loss": 4.1843,
+      "step": 93500
+    },
+    {
+      "epoch": 3.962399359271593,
+      "grad_norm": 1.2012404203414917,
+      "learning_rate": 8.559362003172834e-05,
+      "loss": 4.183,
+      "step": 94000
+    },
+    {
+      "epoch": 3.983475951608144,
+      "grad_norm": 1.2290505170822144,
+      "learning_rate": 8.550786777001244e-05,
+      "loss": 4.1827,
+      "step": 94500
+    },
+    {
+      "epoch": 4.004552543944695,
+      "grad_norm": 1.3255467414855957,
+      "learning_rate": 8.542211550829654e-05,
+      "loss": 4.1779,
+      "step": 95000
+    },
+    {
+      "epoch": 4.025629136281246,
+      "grad_norm": 1.305121898651123,
+      "learning_rate": 8.533636324658063e-05,
+      "loss": 4.1705,
+      "step": 95500
+    },
+    {
+      "epoch": 4.046705728617797,
+      "grad_norm": 1.3128089904785156,
+      "learning_rate": 8.525061098486474e-05,
+      "loss": 4.1733,
+      "step": 96000
+    },
+    {
+      "epoch": 4.067782320954348,
+      "grad_norm": 1.2922817468643188,
+      "learning_rate": 8.516485872314884e-05,
+      "loss": 4.17,
+      "step": 96500
+    },
+    {
+      "epoch": 4.088858913290899,
+      "grad_norm": 1.1880995035171509,
+      "learning_rate": 8.507910646143292e-05,
+      "loss": 4.1697,
+      "step": 97000
+    },
+    {
+      "epoch": 4.10993550562745,
+      "grad_norm": 1.4019254446029663,
+      "learning_rate": 8.499335419971702e-05,
+      "loss": 4.167,
+      "step": 97500
+    },
+    {
+      "epoch": 4.1310120979640015,
+      "grad_norm": 1.213587999343872,
+      "learning_rate": 8.490760193800112e-05,
+      "loss": 4.1666,
+      "step": 98000
+    },
+    {
+      "epoch": 4.152088690300552,
+      "grad_norm": 1.2260299921035767,
+      "learning_rate": 8.482184967628522e-05,
+      "loss": 4.165,
+      "step": 98500
+    },
+    {
+      "epoch": 4.173165282637103,
+      "grad_norm": 1.2662341594696045,
+      "learning_rate": 8.473609741456931e-05,
+      "loss": 4.1692,
+      "step": 99000
+    },
+    {
+      "epoch": 4.194241874973654,
+      "grad_norm": 1.2768611907958984,
+      "learning_rate": 8.46503451528534e-05,
+      "loss": 4.169,
+      "step": 99500
+    },
+    {
+      "epoch": 4.2153184673102055,
+      "grad_norm": 1.2082992792129517,
+      "learning_rate": 8.456459289113751e-05,
+      "loss": 4.1637,
+      "step": 100000
+    },
+    {
+      "epoch": 4.2153184673102055,
+      "eval_accuracy": 0.6795155577036994,
+      "eval_loss": 3.889165163040161,
+      "eval_runtime": 431.7036,
+      "eval_samples_per_second": 279.604,
+      "eval_steps_per_second": 2.914,
+      "step": 100000
+    },
+    {
+      "epoch": 4.236395059646756,
+      "grad_norm": 1.2176048755645752,
+      "learning_rate": 8.447884062942161e-05,
+      "loss": 4.1663,
+      "step": 100500
+    },
+    {
+      "epoch": 4.257471651983307,
+      "grad_norm": 1.2972294092178345,
+      "learning_rate": 8.43930883677057e-05,
+      "loss": 4.1649,
+      "step": 101000
+    },
+    {
+      "epoch": 4.278548244319858,
+      "grad_norm": 1.3176345825195312,
+      "learning_rate": 8.430733610598979e-05,
+      "loss": 4.1599,
+      "step": 101500
+    },
+    {
+      "epoch": 4.29962483665641,
+      "grad_norm": 1.2824589014053345,
+      "learning_rate": 8.42215838442739e-05,
+      "loss": 4.1586,
+      "step": 102000
+    },
+    {
+      "epoch": 4.32070142899296,
+      "grad_norm": 1.2740110158920288,
+      "learning_rate": 8.413583158255799e-05,
+      "loss": 4.1595,
+      "step": 102500
+    },
+    {
+      "epoch": 4.341778021329511,
+      "grad_norm": 1.213599681854248,
+      "learning_rate": 8.405007932084209e-05,
+      "loss": 4.156,
+      "step": 103000
+    },
+    {
+      "epoch": 4.362854613666062,
+      "grad_norm": 1.3011438846588135,
+      "learning_rate": 8.396432705912619e-05,
+      "loss": 4.1582,
+      "step": 103500
+    },
+    {
+      "epoch": 4.383931206002614,
+      "grad_norm": 1.2148698568344116,
+      "learning_rate": 8.387857479741028e-05,
+      "loss": 4.1594,
+      "step": 104000
+    },
+    {
+      "epoch": 4.405007798339165,
+      "grad_norm": 1.2764326333999634,
+      "learning_rate": 8.379282253569438e-05,
+      "loss": 4.1572,
+      "step": 104500
+    },
+    {
+      "epoch": 4.426084390675715,
+      "grad_norm": 1.2693339586257935,
+      "learning_rate": 8.370707027397848e-05,
+      "loss": 4.1548,
+      "step": 105000
+    },
+    {
+      "epoch": 4.4471609830122665,
+      "grad_norm": 1.33685302734375,
+      "learning_rate": 8.362131801226258e-05,
+      "loss": 4.1525,
+      "step": 105500
+    },
+    {
+      "epoch": 4.468237575348818,
+      "grad_norm": 1.2214077711105347,
+      "learning_rate": 8.353556575054668e-05,
+      "loss": 4.1501,
+      "step": 106000
+    },
+    {
+      "epoch": 4.489314167685369,
+      "grad_norm": 1.271178960800171,
+      "learning_rate": 8.344981348883077e-05,
+      "loss": 4.1524,
+      "step": 106500
+    },
+    {
+      "epoch": 4.510390760021919,
+      "grad_norm": 1.323753833770752,
+      "learning_rate": 8.336406122711486e-05,
+      "loss": 4.1482,
+      "step": 107000
+    },
+    {
+      "epoch": 4.5314673523584705,
+      "grad_norm": 1.3438806533813477,
+      "learning_rate": 8.327830896539897e-05,
+      "loss": 4.1526,
+      "step": 107500
+    },
+    {
+      "epoch": 4.552543944695022,
+      "grad_norm": 1.2887499332427979,
+      "learning_rate": 8.319255670368307e-05,
+      "loss": 4.1461,
+      "step": 108000
+    },
+    {
+      "epoch": 4.573620537031573,
+      "grad_norm": 1.2705481052398682,
+      "learning_rate": 8.310680444196715e-05,
+      "loss": 4.1457,
+      "step": 108500
+    },
+    {
+      "epoch": 4.594697129368123,
+      "grad_norm": 1.1926896572113037,
+      "learning_rate": 8.302105218025125e-05,
+      "loss": 4.1399,
+      "step": 109000
+    },
+    {
+      "epoch": 4.615773721704675,
+      "grad_norm": 1.320517897605896,
+      "learning_rate": 8.293529991853536e-05,
+      "loss": 4.1472,
+      "step": 109500
+    },
+    {
+      "epoch": 4.636850314041226,
+      "grad_norm": 1.31378173828125,
+      "learning_rate": 8.284954765681945e-05,
+      "loss": 4.1428,
+      "step": 110000
+    },
+    {
+      "epoch": 4.636850314041226,
+      "eval_accuracy": 0.6829516058281011,
+      "eval_loss": 3.8674232959747314,
+      "eval_runtime": 452.1403,
+      "eval_samples_per_second": 266.966,
+      "eval_steps_per_second": 2.782,
+      "step": 110000
+    },
+    {
+      "epoch": 4.657926906377777,
+      "grad_norm": 1.3386913537979126,
+      "learning_rate": 8.276379539510355e-05,
+      "loss": 4.1427,
+      "step": 110500
+    },
+    {
+      "epoch": 4.679003498714328,
+      "grad_norm": 1.2905458211898804,
+      "learning_rate": 8.267804313338765e-05,
+      "loss": 4.143,
+      "step": 111000
+    },
+    {
+      "epoch": 4.700080091050879,
+      "grad_norm": 1.2975270748138428,
+      "learning_rate": 8.259229087167174e-05,
+      "loss": 4.1394,
+      "step": 111500
+    },
+    {
+      "epoch": 4.72115668338743,
+      "grad_norm": 1.2702622413635254,
+      "learning_rate": 8.250653860995584e-05,
+      "loss": 4.1373,
+      "step": 112000
+    },
+    {
+      "epoch": 4.742233275723981,
+      "grad_norm": 1.2720346450805664,
+      "learning_rate": 8.242078634823994e-05,
+      "loss": 4.1381,
+      "step": 112500
+    },
+    {
+      "epoch": 4.763309868060532,
+      "grad_norm": 1.2733843326568604,
+      "learning_rate": 8.233503408652403e-05,
+      "loss": 4.138,
+      "step": 113000
+    },
+    {
+      "epoch": 4.784386460397083,
+      "grad_norm": 1.3934731483459473,
+      "learning_rate": 8.224928182480814e-05,
+      "loss": 4.1405,
+      "step": 113500
+    },
+    {
+      "epoch": 4.805463052733634,
+      "grad_norm": 1.2896430492401123,
+      "learning_rate": 8.216352956309224e-05,
+      "loss": 4.1321,
+      "step": 114000
+    },
+    {
+      "epoch": 4.826539645070185,
+      "grad_norm": 1.3406416177749634,
+      "learning_rate": 8.207777730137632e-05,
+      "loss": 4.1339,
+      "step": 114500
+    },
+    {
+      "epoch": 4.847616237406736,
+      "grad_norm": 1.288157343864441,
+      "learning_rate": 8.199202503966042e-05,
+      "loss": 4.1319,
+      "step": 115000
+    },
+    {
+      "epoch": 4.868692829743287,
+      "grad_norm": 1.2445929050445557,
+      "learning_rate": 8.190627277794453e-05,
+      "loss": 4.133,
+      "step": 115500
+    },
+    {
+      "epoch": 4.889769422079838,
+      "grad_norm": 1.2539482116699219,
+      "learning_rate": 8.182052051622862e-05,
+      "loss": 4.1295,
+      "step": 116000
+    },
+    {
+      "epoch": 4.910846014416389,
+      "grad_norm": 1.3659460544586182,
+      "learning_rate": 8.173476825451271e-05,
+      "loss": 4.1301,
+      "step": 116500
+    },
+    {
+      "epoch": 4.9319226067529405,
+      "grad_norm": 1.2577362060546875,
+      "learning_rate": 8.164901599279681e-05,
+      "loss": 4.1325,
+      "step": 117000
+    },
+    {
+      "epoch": 4.952999199089491,
+      "grad_norm": 1.2471140623092651,
+      "learning_rate": 8.156326373108091e-05,
+      "loss": 4.1314,
+      "step": 117500
+    },
+    {
+      "epoch": 4.974075791426042,
+      "grad_norm": 1.2989256381988525,
+      "learning_rate": 8.147751146936501e-05,
+      "loss": 4.1281,
+      "step": 118000
+    },
+    {
+      "epoch": 4.995152383762593,
+      "grad_norm": 1.3426097631454468,
+      "learning_rate": 8.139175920764911e-05,
+      "loss": 4.122,
+      "step": 118500
+    },
+    {
+      "epoch": 5.0162289760991445,
+      "grad_norm": 1.3568274974822998,
+      "learning_rate": 8.13060069459332e-05,
+      "loss": 4.1236,
+      "step": 119000
+    },
+    {
+      "epoch": 5.037305568435695,
+      "grad_norm": 1.2204523086547852,
+      "learning_rate": 8.12202546842173e-05,
+      "loss": 4.1209,
+      "step": 119500
+    },
+    {
+      "epoch": 5.058382160772246,
+      "grad_norm": 1.405532956123352,
+      "learning_rate": 8.11345024225014e-05,
+      "loss": 4.1204,
+      "step": 120000
+    },
+    {
+      "epoch": 5.058382160772246,
+      "eval_accuracy": 0.6860231780467186,
+      "eval_loss": 3.843247890472412,
+      "eval_runtime": 432.9677,
+      "eval_samples_per_second": 278.788,
+      "eval_steps_per_second": 2.906,
+      "step": 120000
+    },
+    {
+      "epoch": 5.079458753108797,
+      "grad_norm": 1.3123078346252441,
+      "learning_rate": 8.104875016078549e-05,
+      "loss": 4.1137,
+      "step": 120500
+    },
+    {
+      "epoch": 5.100535345445349,
+      "grad_norm": 1.2861294746398926,
+      "learning_rate": 8.09629978990696e-05,
+      "loss": 4.1159,
+      "step": 121000
+    },
+    {
+      "epoch": 5.1216119377819,
+      "grad_norm": 1.3627774715423584,
+      "learning_rate": 8.08772456373537e-05,
+      "loss": 4.1179,
+      "step": 121500
+    },
+    {
+      "epoch": 5.14268853011845,
+      "grad_norm": 1.3723270893096924,
+      "learning_rate": 8.079149337563778e-05,
+      "loss": 4.1163,
+      "step": 122000
+    },
+    {
+      "epoch": 5.163765122455001,
+      "grad_norm": 1.2253586053848267,
+      "learning_rate": 8.070574111392188e-05,
+      "loss": 4.1136,
+      "step": 122500
+    },
+    {
+      "epoch": 5.184841714791553,
+      "grad_norm": 1.3064696788787842,
+      "learning_rate": 8.061998885220599e-05,
+      "loss": 4.1147,
+      "step": 123000
+    },
+    {
+      "epoch": 5.205918307128104,
+      "grad_norm": 1.3062758445739746,
+      "learning_rate": 8.053423659049008e-05,
+      "loss": 4.1143,
+      "step": 123500
+    },
+    {
+      "epoch": 5.226994899464654,
+      "grad_norm": 1.2994937896728516,
+      "learning_rate": 8.044848432877418e-05,
+      "loss": 4.1145,
+      "step": 124000
+    },
+    {
+      "epoch": 5.2480714918012055,
+      "grad_norm": 1.335684895515442,
+      "learning_rate": 8.036273206705827e-05,
+      "loss": 4.1175,
+      "step": 124500
+    },
+    {
+      "epoch": 5.269148084137757,
+      "grad_norm": 1.3654730319976807,
+      "learning_rate": 8.027697980534237e-05,
+      "loss": 4.1082,
+      "step": 125000
+    },
+    {
+      "epoch": 5.290224676474308,
+      "grad_norm": 1.3133584260940552,
+      "learning_rate": 8.019122754362647e-05,
+      "loss": 4.1134,
+      "step": 125500
+    },
+    {
+      "epoch": 5.311301268810858,
+      "grad_norm": 1.318641185760498,
+      "learning_rate": 8.010547528191057e-05,
+      "loss": 4.1108,
+      "step": 126000
+    },
+    {
+      "epoch": 5.3323778611474095,
+      "grad_norm": 1.2622636556625366,
+      "learning_rate": 8.001972302019465e-05,
+      "loss": 4.1067,
+      "step": 126500
+    },
+    {
+      "epoch": 5.353454453483961,
+      "grad_norm": 1.3190937042236328,
+      "learning_rate": 7.993397075847876e-05,
+      "loss": 4.1073,
+      "step": 127000
+    },
+    {
+      "epoch": 5.374531045820512,
+      "grad_norm": 1.2658008337020874,
+      "learning_rate": 7.984821849676286e-05,
+      "loss": 4.1083,
+      "step": 127500
+    },
+    {
+      "epoch": 5.395607638157063,
+      "grad_norm": 1.2833892107009888,
+      "learning_rate": 7.976246623504695e-05,
+      "loss": 4.1065,
+      "step": 128000
+    },
+    {
+      "epoch": 5.416684230493614,
+      "grad_norm": 1.375807523727417,
+      "learning_rate": 7.967671397333105e-05,
+      "loss": 4.1032,
+      "step": 128500
+    },
+    {
+      "epoch": 5.437760822830165,
+      "grad_norm": 1.2585948705673218,
+      "learning_rate": 7.959096171161516e-05,
+      "loss": 4.1032,
+      "step": 129000
+    },
+    {
+      "epoch": 5.458837415166716,
+      "grad_norm": 1.2781621217727661,
+      "learning_rate": 7.950520944989924e-05,
+      "loss": 4.1043,
+      "step": 129500
+    },
+    {
+      "epoch": 5.479914007503266,
+      "grad_norm": 1.278773546218872,
+      "learning_rate": 7.941945718818334e-05,
+      "loss": 4.1028,
+      "step": 130000
+    },
+    {
+      "epoch": 5.479914007503266,
+      "eval_accuracy": 0.688711696434632,
+      "eval_loss": 3.8268468379974365,
+      "eval_runtime": 472.1878,
+      "eval_samples_per_second": 255.631,
+      "eval_steps_per_second": 2.664,
+      "step": 130000
+    },
+    {
+      "epoch": 5.500990599839818,
+      "grad_norm": 1.2571120262145996,
+      "learning_rate": 7.933370492646744e-05,
+      "loss": 4.108,
+      "step": 130500
+    },
+    {
+      "epoch": 5.522067192176369,
+      "grad_norm": 1.2858752012252808,
+      "learning_rate": 7.924795266475154e-05,
+      "loss": 4.0996,
+      "step": 131000
+    },
+    {
+      "epoch": 5.54314378451292,
+      "grad_norm": 1.3101019859313965,
+      "learning_rate": 7.916220040303564e-05,
+      "loss": 4.0951,
+      "step": 131500
+    },
+    {
+      "epoch": 5.564220376849471,
+      "grad_norm": 1.3078222274780273,
+      "learning_rate": 7.907644814131972e-05,
+      "loss": 4.1029,
+      "step": 132000
+    },
+    {
+      "epoch": 5.585296969186022,
+      "grad_norm": 1.2949804067611694,
+      "learning_rate": 7.899069587960383e-05,
+      "loss": 4.0949,
+      "step": 132500
+    },
+    {
+      "epoch": 5.606373561522573,
+      "grad_norm": 1.2311137914657593,
+      "learning_rate": 7.890494361788793e-05,
+      "loss": 4.0969,
+      "step": 133000
+    },
+    {
+      "epoch": 5.627450153859124,
+      "grad_norm": 1.305440902709961,
+      "learning_rate": 7.881919135617202e-05,
+      "loss": 4.094,
+      "step": 133500
+    },
+    {
+      "epoch": 5.648526746195675,
+      "grad_norm": 1.3193600177764893,
+      "learning_rate": 7.873343909445611e-05,
+      "loss": 4.0912,
+      "step": 134000
+    },
+    {
+      "epoch": 5.669603338532226,
+      "grad_norm": 1.3457558155059814,
+      "learning_rate": 7.864768683274023e-05,
+      "loss": 4.1002,
+      "step": 134500
+    },
+    {
+      "epoch": 5.690679930868777,
+      "grad_norm": 1.3072319030761719,
+      "learning_rate": 7.856193457102431e-05,
+      "loss": 4.0981,
+      "step": 135000
+    },
+    {
+      "epoch": 5.711756523205328,
+      "grad_norm": 1.3082062005996704,
+      "learning_rate": 7.847618230930841e-05,
+      "loss": 4.0948,
+      "step": 135500
+    },
+    {
+      "epoch": 5.7328331155418795,
+      "grad_norm": 1.3051823377609253,
+      "learning_rate": 7.839043004759251e-05,
+      "loss": 4.0947,
+      "step": 136000
+    },
+    {
+      "epoch": 5.75390970787843,
+      "grad_norm": 1.3111071586608887,
+      "learning_rate": 7.83046777858766e-05,
+      "loss": 4.0922,
+      "step": 136500
+    },
+    {
+      "epoch": 5.774986300214981,
+      "grad_norm": 1.2713321447372437,
+      "learning_rate": 7.82189255241607e-05,
+      "loss": 4.0914,
+      "step": 137000
+    },
+    {
+      "epoch": 5.796062892551532,
+      "grad_norm": 1.3324121236801147,
+      "learning_rate": 7.81331732624448e-05,
+      "loss": 4.0916,
+      "step": 137500
+    },
+    {
+      "epoch": 5.8171394848880835,
+      "grad_norm": 1.2724429368972778,
+      "learning_rate": 7.804742100072889e-05,
+      "loss": 4.0878,
+      "step": 138000
+    },
+    {
+      "epoch": 5.838216077224635,
+      "grad_norm": 1.3340167999267578,
+      "learning_rate": 7.7961668739013e-05,
+      "loss": 4.093,
+      "step": 138500
+    },
+    {
+      "epoch": 5.859292669561185,
+      "grad_norm": 1.3792376518249512,
+      "learning_rate": 7.78759164772971e-05,
+      "loss": 4.088,
+      "step": 139000
+    },
+    {
+      "epoch": 5.880369261897736,
+      "grad_norm": 1.3087481260299683,
+      "learning_rate": 7.779016421558118e-05,
+      "loss": 4.0863,
+      "step": 139500
+    },
+    {
+      "epoch": 5.901445854234288,
+      "grad_norm": 1.3278608322143555,
+      "learning_rate": 7.770441195386528e-05,
+      "loss": 4.091,
+      "step": 140000
+    },
+    {
+      "epoch": 5.901445854234288,
+      "eval_accuracy": 0.6911929919332662,
+      "eval_loss": 3.8095269203186035,
+      "eval_runtime": 483.3698,
+      "eval_samples_per_second": 249.718,
+      "eval_steps_per_second": 2.603,
+      "step": 140000
+    },
+    {
+      "epoch": 5.922522446570838,
+      "grad_norm": 1.3651552200317383,
+      "learning_rate": 7.761865969214939e-05,
+      "loss": 4.0818,
+      "step": 140500
+    },
+    {
+      "epoch": 5.943599038907389,
+      "grad_norm": 1.276672601699829,
+      "learning_rate": 7.753290743043348e-05,
+      "loss": 4.0875,
+      "step": 141000
+    },
+    {
+      "epoch": 5.96467563124394,
+      "grad_norm": 1.3164788484573364,
+      "learning_rate": 7.744715516871758e-05,
+      "loss": 4.0868,
+      "step": 141500
+    },
+    {
+      "epoch": 5.985752223580492,
+      "grad_norm": 1.3242648839950562,
+      "learning_rate": 7.736140290700167e-05,
+      "loss": 4.081,
+      "step": 142000
+    },
+    {
+      "epoch": 6.006828815917043,
+      "grad_norm": 1.3866955041885376,
+      "learning_rate": 7.727565064528577e-05,
+      "loss": 4.0818,
+      "step": 142500
+    },
+    {
+      "epoch": 6.027905408253593,
+      "grad_norm": 1.2436034679412842,
+      "learning_rate": 7.718989838356987e-05,
+      "loss": 4.0782,
+      "step": 143000
+    },
+    {
+      "epoch": 6.0489820005901445,
+      "grad_norm": 1.362056016921997,
+      "learning_rate": 7.710414612185397e-05,
+      "loss": 4.0811,
+      "step": 143500
+    },
+    {
+      "epoch": 6.070058592926696,
+      "grad_norm": 1.280554175376892,
+      "learning_rate": 7.701839386013807e-05,
+      "loss": 4.0773,
+      "step": 144000
+    },
+    {
+      "epoch": 6.091135185263247,
+      "grad_norm": 1.3033353090286255,
+      "learning_rate": 7.693264159842217e-05,
+      "loss": 4.0772,
+      "step": 144500
+    },
+    {
+      "epoch": 6.112211777599797,
+      "grad_norm": 1.2917585372924805,
+      "learning_rate": 7.684688933670626e-05,
+      "loss": 4.0771,
+      "step": 145000
+    },
+    {
+      "epoch": 6.1332883699363485,
+      "grad_norm": 1.2532538175582886,
+      "learning_rate": 7.676113707499035e-05,
+      "loss": 4.0776,
+      "step": 145500
+    },
+    {
+      "epoch": 6.1543649622729,
+      "grad_norm": 1.2743442058563232,
+      "learning_rate": 7.667538481327446e-05,
+      "loss": 4.0764,
+      "step": 146000
+    },
+    {
+      "epoch": 6.175441554609451,
+      "grad_norm": 1.3272305727005005,
+      "learning_rate": 7.658963255155856e-05,
+      "loss": 4.077,
+      "step": 146500
+    },
+    {
+      "epoch": 6.196518146946001,
+      "grad_norm": 1.3062189817428589,
+      "learning_rate": 7.650388028984264e-05,
+      "loss": 4.0746,
+      "step": 147000
+    },
+    {
+      "epoch": 6.217594739282553,
+      "grad_norm": 1.3125274181365967,
+      "learning_rate": 7.641812802812674e-05,
+      "loss": 4.0729,
+      "step": 147500
+    },
+    {
+      "epoch": 6.238671331619104,
+      "grad_norm": 1.2309629917144775,
+      "learning_rate": 7.633237576641084e-05,
+      "loss": 4.0715,
+      "step": 148000
+    },
+    {
+      "epoch": 6.259747923955655,
+      "grad_norm": 1.4368162155151367,
+      "learning_rate": 7.624662350469494e-05,
+      "loss": 4.0739,
+      "step": 148500
+    },
+    {
+      "epoch": 6.280824516292206,
+      "grad_norm": 1.2945464849472046,
+      "learning_rate": 7.616087124297904e-05,
+      "loss": 4.0736,
+      "step": 149000
+    },
+    {
+      "epoch": 6.301901108628757,
+      "grad_norm": 1.3388108015060425,
+      "learning_rate": 7.607511898126313e-05,
+      "loss": 4.0723,
+      "step": 149500
+    },
+    {
+      "epoch": 6.322977700965308,
+      "grad_norm": 1.32869553565979,
+      "learning_rate": 7.598936671954723e-05,
+      "loss": 4.0686,
+      "step": 150000
+    },
+    {
+      "epoch": 6.322977700965308,
+      "eval_accuracy": 0.6935591178818449,
+      "eval_loss": 3.7973339557647705,
+      "eval_runtime": 485.5417,
+      "eval_samples_per_second": 248.601,
+      "eval_steps_per_second": 2.591,
+      "step": 150000
+    },
+    {
+      "epoch": 6.344054293301859,
+      "grad_norm": 1.3697847127914429,
+      "learning_rate": 7.590361445783133e-05,
+      "loss": 4.0706,
+      "step": 150500
+    },
+    {
+      "epoch": 6.36513088563841,
+      "grad_norm": 1.240012526512146,
+      "learning_rate": 7.581786219611543e-05,
+      "loss": 4.0694,
+      "step": 151000
+    },
+    {
+      "epoch": 6.386207477974961,
+      "grad_norm": 1.3116835355758667,
+      "learning_rate": 7.573210993439951e-05,
+      "loss": 4.0698,
+      "step": 151500
+    },
+    {
+      "epoch": 6.407284070311512,
+      "grad_norm": 1.2528948783874512,
+      "learning_rate": 7.564635767268363e-05,
+      "loss": 4.0685,
+      "step": 152000
+    },
+    {
+      "epoch": 6.428360662648063,
+      "grad_norm": 1.2877901792526245,
+      "learning_rate": 7.556060541096772e-05,
+      "loss": 4.0645,
+      "step": 152500
+    },
+    {
+      "epoch": 6.449437254984614,
+      "grad_norm": 1.2688461542129517,
+      "learning_rate": 7.547485314925181e-05,
+      "loss": 4.0682,
+      "step": 153000
+    },
+    {
+      "epoch": 6.470513847321165,
+      "grad_norm": 1.2985889911651611,
+      "learning_rate": 7.538910088753591e-05,
+      "loss": 4.0664,
+      "step": 153500
+    },
+    {
+      "epoch": 6.491590439657716,
+      "grad_norm": 1.263596773147583,
+      "learning_rate": 7.530334862582002e-05,
+      "loss": 4.066,
+      "step": 154000
+    },
+    {
+      "epoch": 6.512667031994267,
+      "grad_norm": 1.3664573431015015,
+      "learning_rate": 7.52175963641041e-05,
+      "loss": 4.0681,
+      "step": 154500
+    },
+    {
+      "epoch": 6.5337436243308185,
+      "grad_norm": 1.322748064994812,
+      "learning_rate": 7.51318441023882e-05,
+      "loss": 4.0682,
+      "step": 155000
+    },
+    {
+      "epoch": 6.55482021666737,
+      "grad_norm": 1.2959364652633667,
+      "learning_rate": 7.50460918406723e-05,
+      "loss": 4.0653,
+      "step": 155500
+    },
+    {
+      "epoch": 6.57589680900392,
+      "grad_norm": 1.285718560218811,
+      "learning_rate": 7.49603395789564e-05,
+      "loss": 4.0625,
+      "step": 156000
+    },
+    {
+      "epoch": 6.596973401340471,
+      "grad_norm": 1.2567390203475952,
+      "learning_rate": 7.48745873172405e-05,
+      "loss": 4.0578,
+      "step": 156500
+    },
+    {
+      "epoch": 6.6180499936770225,
+      "grad_norm": 1.3125557899475098,
+      "learning_rate": 7.47888350555246e-05,
+      "loss": 4.0566,
+      "step": 157000
+    },
+    {
+      "epoch": 6.639126586013573,
+      "grad_norm": 1.3314908742904663,
+      "learning_rate": 7.47030827938087e-05,
+      "loss": 4.0603,
+      "step": 157500
+    },
+    {
+      "epoch": 6.660203178350124,
+      "grad_norm": 1.3004220724105835,
+      "learning_rate": 7.461733053209279e-05,
+      "loss": 4.0605,
+      "step": 158000
+    },
+    {
+      "epoch": 6.681279770686675,
+      "grad_norm": 1.3171597719192505,
+      "learning_rate": 7.453157827037689e-05,
+      "loss": 4.0642,
+      "step": 158500
+    },
+    {
+      "epoch": 6.702356363023227,
+      "grad_norm": 1.277429223060608,
+      "learning_rate": 7.444582600866098e-05,
+      "loss": 4.0566,
+      "step": 159000
+    },
+    {
+      "epoch": 6.723432955359778,
+      "grad_norm": 1.345645546913147,
+      "learning_rate": 7.436007374694507e-05,
+      "loss": 4.0618,
+      "step": 159500
+    },
+    {
+      "epoch": 6.744509547696328,
+      "grad_norm": 1.3768807649612427,
+      "learning_rate": 7.427432148522919e-05,
+      "loss": 4.0613,
+      "step": 160000
+    },
+    {
+      "epoch": 6.744509547696328,
+      "eval_accuracy": 0.6951553248466421,
+      "eval_loss": 3.785357713699341,
+      "eval_runtime": 495.1066,
+      "eval_samples_per_second": 243.798,
+      "eval_steps_per_second": 2.541,
+      "step": 160000
+    },
+    {
+      "epoch": 6.765586140032879,
+      "grad_norm": 1.3072136640548706,
+      "learning_rate": 7.418856922351327e-05,
+      "loss": 4.0581,
+      "step": 160500
+    },
+    {
+      "epoch": 6.786662732369431,
+      "grad_norm": 1.3120360374450684,
+      "learning_rate": 7.410281696179737e-05,
+      "loss": 4.0554,
+      "step": 161000
+    },
+    {
+      "epoch": 6.807739324705982,
+      "grad_norm": 1.2523508071899414,
+      "learning_rate": 7.401706470008147e-05,
+      "loss": 4.0529,
+      "step": 161500
+    },
+    {
+      "epoch": 6.828815917042532,
+      "grad_norm": 1.2762928009033203,
+      "learning_rate": 7.393131243836557e-05,
+      "loss": 4.0531,
+      "step": 162000
+    },
+    {
+      "epoch": 6.8498925093790834,
+      "grad_norm": 1.2954990863800049,
+      "learning_rate": 7.384556017664966e-05,
+      "loss": 4.051,
+      "step": 162500
+    },
+    {
+      "epoch": 6.870969101715635,
+      "grad_norm": 1.29587721824646,
+      "learning_rate": 7.375980791493376e-05,
+      "loss": 4.0513,
+      "step": 163000
+    },
+    {
+      "epoch": 6.892045694052186,
+      "grad_norm": 1.3240844011306763,
+      "learning_rate": 7.367405565321786e-05,
+      "loss": 4.0502,
+      "step": 163500
+    },
+    {
+      "epoch": 6.913122286388736,
+      "grad_norm": 1.3163014650344849,
+      "learning_rate": 7.358830339150196e-05,
+      "loss": 4.0501,
+      "step": 164000
+    },
+    {
+      "epoch": 6.9341988787252875,
+      "grad_norm": 1.3085330724716187,
+      "learning_rate": 7.350255112978606e-05,
+      "loss": 4.0563,
+      "step": 164500
+    },
+    {
+      "epoch": 6.955275471061839,
+      "grad_norm": 1.2987468242645264,
+      "learning_rate": 7.341679886807014e-05,
+      "loss": 4.0516,
+      "step": 165000
+    },
+    {
+      "epoch": 6.97635206339839,
+      "grad_norm": 1.3119937181472778,
+      "learning_rate": 7.333104660635425e-05,
+      "loss": 4.0482,
+      "step": 165500
+    },
+    {
+      "epoch": 6.997428655734941,
+      "grad_norm": 1.2574567794799805,
+      "learning_rate": 7.324529434463835e-05,
+      "loss": 4.049,
+      "step": 166000
+    },
+    {
+      "epoch": 7.018505248071492,
+      "grad_norm": 1.3008226156234741,
+      "learning_rate": 7.315954208292244e-05,
+      "loss": 4.0473,
+      "step": 166500
+    },
+    {
+      "epoch": 7.039581840408043,
+      "grad_norm": 1.3551065921783447,
+      "learning_rate": 7.307378982120653e-05,
+      "loss": 4.0434,
+      "step": 167000
+    },
+    {
+      "epoch": 7.060658432744594,
+      "grad_norm": 1.303310513496399,
+      "learning_rate": 7.298803755949063e-05,
+      "loss": 4.0442,
+      "step": 167500
+    },
+    {
+      "epoch": 7.081735025081145,
+      "grad_norm": 1.2999267578125,
+      "learning_rate": 7.290228529777473e-05,
+      "loss": 4.0428,
+      "step": 168000
+    },
+    {
+      "epoch": 7.102811617417696,
+      "grad_norm": 1.24193274974823,
+      "learning_rate": 7.281653303605883e-05,
+      "loss": 4.044,
+      "step": 168500
+    },
+    {
+      "epoch": 7.123888209754247,
+      "grad_norm": 1.299760341644287,
+      "learning_rate": 7.273078077434293e-05,
+      "loss": 4.0485,
+      "step": 169000
+    },
+    {
+      "epoch": 7.144964802090798,
+      "grad_norm": 1.3078458309173584,
+      "learning_rate": 7.264502851262703e-05,
+      "loss": 4.0414,
+      "step": 169500
+    },
+    {
+      "epoch": 7.166041394427349,
+      "grad_norm": 1.3176155090332031,
+      "learning_rate": 7.255927625091112e-05,
+      "loss": 4.0375,
+      "step": 170000
+    },
+    {
+      "epoch": 7.166041394427349,
+      "eval_accuracy": 0.6975991628592397,
+      "eval_loss": 3.771228551864624,
+      "eval_runtime": 441.9875,
+      "eval_samples_per_second": 273.098,
+      "eval_steps_per_second": 2.846,
+      "step": 170000
+    },
+    {
+      "epoch": 7.1871179867639,
+      "grad_norm": 1.3206108808517456,
+      "learning_rate": 7.247352398919521e-05,
+      "loss": 4.0436,
+      "step": 170500
+    },
+    {
+      "epoch": 7.208194579100451,
+      "grad_norm": 1.2868762016296387,
+      "learning_rate": 7.238777172747931e-05,
+      "loss": 4.0417,
+      "step": 171000
+    },
+    {
+      "epoch": 7.229271171437002,
+      "grad_norm": 1.2857002019882202,
+      "learning_rate": 7.230201946576342e-05,
+      "loss": 4.0427,
+      "step": 171500
+    },
+    {
+      "epoch": 7.250347763773553,
+      "grad_norm": 1.3073184490203857,
+      "learning_rate": 7.22162672040475e-05,
+      "loss": 4.0415,
+      "step": 172000
+    },
+    {
+      "epoch": 7.271424356110104,
+      "grad_norm": 1.304558515548706,
+      "learning_rate": 7.21305149423316e-05,
+      "loss": 4.0398,
+      "step": 172500
+    },
+    {
+      "epoch": 7.292500948446655,
+      "grad_norm": 1.2772414684295654,
+      "learning_rate": 7.20447626806157e-05,
+      "loss": 4.0358,
+      "step": 173000
+    },
+    {
+      "epoch": 7.313577540783206,
+      "grad_norm": 1.3549686670303345,
+      "learning_rate": 7.19590104188998e-05,
+      "loss": 4.0372,
+      "step": 173500
+    },
+    {
+      "epoch": 7.3346541331197574,
+      "grad_norm": 1.2849866151809692,
+      "learning_rate": 7.18732581571839e-05,
+      "loss": 4.0379,
+      "step": 174000
+    },
+    {
+      "epoch": 7.355730725456308,
+      "grad_norm": 1.270983099937439,
+      "learning_rate": 7.1787505895468e-05,
+      "loss": 4.0366,
+      "step": 174500
+    },
+    {
+      "epoch": 7.376807317792859,
+      "grad_norm": 1.3086708784103394,
+      "learning_rate": 7.17017536337521e-05,
+      "loss": 4.038,
+      "step": 175000
+    },
+    {
+      "epoch": 7.39788391012941,
+      "grad_norm": 1.2888997793197632,
+      "learning_rate": 7.161600137203619e-05,
+      "loss": 4.0382,
+      "step": 175500
+    },
+    {
+      "epoch": 7.4189605024659615,
+      "grad_norm": 1.2552061080932617,
+      "learning_rate": 7.153024911032029e-05,
+      "loss": 4.0324,
+      "step": 176000
+    },
+    {
+      "epoch": 7.440037094802513,
+      "grad_norm": 1.316170573234558,
+      "learning_rate": 7.144449684860438e-05,
+      "loss": 4.0366,
+      "step": 176500
+    },
+    {
+      "epoch": 7.461113687139063,
+      "grad_norm": 1.2737340927124023,
+      "learning_rate": 7.135874458688849e-05,
+      "loss": 4.0376,
+      "step": 177000
+    },
+    {
+      "epoch": 7.482190279475614,
+      "grad_norm": 1.3429104089736938,
+      "learning_rate": 7.127299232517259e-05,
+      "loss": 4.0384,
+      "step": 177500
+    },
+    {
+      "epoch": 7.503266871812166,
+      "grad_norm": 1.3066461086273193,
+      "learning_rate": 7.118724006345667e-05,
+      "loss": 4.0358,
+      "step": 178000
+    },
+    {
+      "epoch": 7.524343464148717,
+      "grad_norm": 1.3475944995880127,
+      "learning_rate": 7.110148780174077e-05,
+      "loss": 4.0321,
+      "step": 178500
+    },
+    {
+      "epoch": 7.545420056485267,
+      "grad_norm": 1.257920742034912,
+      "learning_rate": 7.101573554002488e-05,
+      "loss": 4.0318,
+      "step": 179000
+    },
+    {
+      "epoch": 7.566496648821818,
+      "grad_norm": 1.2701014280319214,
+      "learning_rate": 7.092998327830897e-05,
+      "loss": 4.0347,
+      "step": 179500
+    },
+    {
+      "epoch": 7.58757324115837,
+      "grad_norm": 1.2982786893844604,
+      "learning_rate": 7.084423101659306e-05,
+      "loss": 4.0301,
+      "step": 180000
+    },
+    {
+      "epoch": 7.58757324115837,
+      "eval_accuracy": 0.6984443699421429,
+      "eval_loss": 3.763975143432617,
+      "eval_runtime": 453.2313,
+      "eval_samples_per_second": 266.323,
+      "eval_steps_per_second": 2.776,
+      "step": 180000
+    },
+    {
+      "epoch": 7.608649833494921,
+      "grad_norm": 1.384023904800415,
+      "learning_rate": 7.075847875487716e-05,
+      "loss": 4.0301,
+      "step": 180500
+    },
+    {
+      "epoch": 7.629726425831471,
+      "grad_norm": 1.3524622917175293,
+      "learning_rate": 7.067272649316126e-05,
+      "loss": 4.0323,
+      "step": 181000
+    },
+    {
+      "epoch": 7.650803018168022,
+      "grad_norm": 1.325691819190979,
+      "learning_rate": 7.058697423144536e-05,
+      "loss": 4.0259,
+      "step": 181500
+    },
+    {
+      "epoch": 7.671879610504574,
+      "grad_norm": 1.2585123777389526,
+      "learning_rate": 7.050122196972946e-05,
+      "loss": 4.0269,
+      "step": 182000
+    },
+    {
+      "epoch": 7.692956202841125,
+      "grad_norm": 1.2671507596969604,
+      "learning_rate": 7.041546970801354e-05,
+      "loss": 4.0317,
+      "step": 182500
+    },
+    {
+      "epoch": 7.714032795177676,
+      "grad_norm": 1.3290849924087524,
+      "learning_rate": 7.032971744629765e-05,
+      "loss": 4.0309,
+      "step": 183000
+    },
+    {
+      "epoch": 7.7351093875142265,
+      "grad_norm": 1.3145209550857544,
+      "learning_rate": 7.024396518458175e-05,
+      "loss": 4.0319,
+      "step": 183500
+    },
+    {
+      "epoch": 7.756185979850778,
+      "grad_norm": 1.2444292306900024,
+      "learning_rate": 7.015821292286584e-05,
+      "loss": 4.0294,
+      "step": 184000
+    },
+    {
+      "epoch": 7.777262572187329,
+      "grad_norm": 1.3552958965301514,
+      "learning_rate": 7.007246066114993e-05,
+      "loss": 4.0263,
+      "step": 184500
+    },
+    {
+      "epoch": 7.798339164523879,
+      "grad_norm": 1.3454278707504272,
+      "learning_rate": 6.998670839943405e-05,
+      "loss": 4.0281,
+      "step": 185000
+    },
+    {
+      "epoch": 7.8194157568604306,
+      "grad_norm": 1.3203349113464355,
+      "learning_rate": 6.990095613771813e-05,
+      "loss": 4.0259,
+      "step": 185500
+    },
+    {
+      "epoch": 7.840492349196982,
+      "grad_norm": 1.4024029970169067,
+      "learning_rate": 6.981520387600223e-05,
+      "loss": 4.0251,
+      "step": 186000
+    },
+    {
+      "epoch": 7.861568941533533,
+      "grad_norm": 1.287111520767212,
+      "learning_rate": 6.972945161428633e-05,
+      "loss": 4.027,
+      "step": 186500
+    },
+    {
+      "epoch": 7.882645533870084,
+      "grad_norm": 1.3077434301376343,
+      "learning_rate": 6.964369935257043e-05,
+      "loss": 4.0263,
+      "step": 187000
+    },
+    {
+      "epoch": 7.903722126206635,
+      "grad_norm": 1.3586114645004272,
+      "learning_rate": 6.955794709085452e-05,
+      "loss": 4.0279,
+      "step": 187500
+    },
+    {
+      "epoch": 7.924798718543186,
+      "grad_norm": 1.3378574848175049,
+      "learning_rate": 6.947219482913862e-05,
+      "loss": 4.0157,
+      "step": 188000
+    },
+    {
+      "epoch": 7.945875310879737,
+      "grad_norm": 1.2944865226745605,
+      "learning_rate": 6.938644256742272e-05,
+      "loss": 4.0263,
+      "step": 188500
+    },
+    {
+      "epoch": 7.966951903216288,
+      "grad_norm": 1.3497295379638672,
+      "learning_rate": 6.930069030570682e-05,
+      "loss": 4.0231,
+      "step": 189000
+    },
+    {
+      "epoch": 7.988028495552839,
+      "grad_norm": 1.2943888902664185,
+      "learning_rate": 6.921493804399092e-05,
+      "loss": 4.019,
+      "step": 189500
+    },
+    {
+      "epoch": 8.00910508788939,
+      "grad_norm": 1.3155953884124756,
+      "learning_rate": 6.9129185782275e-05,
+      "loss": 4.0202,
+      "step": 190000
+    },
+    {
+      "epoch": 8.00910508788939,
+      "eval_accuracy": 0.7001893196227409,
+      "eval_loss": 3.751873254776001,
+      "eval_runtime": 420.6972,
+      "eval_samples_per_second": 286.919,
+      "eval_steps_per_second": 2.99,
+      "step": 190000
+    },
+    {
+      "epoch": 8.03018168022594,
+      "grad_norm": 1.3342705965042114,
+      "learning_rate": 6.904343352055911e-05,
+      "loss": 4.0167,
+      "step": 190500
+    },
+    {
+      "epoch": 8.051258272562492,
+      "grad_norm": 1.3504923582077026,
+      "learning_rate": 6.895768125884321e-05,
+      "loss": 4.0185,
+      "step": 191000
+    },
+    {
+      "epoch": 8.072334864899043,
+      "grad_norm": 1.3414698839187622,
+      "learning_rate": 6.88719289971273e-05,
+      "loss": 4.0179,
+      "step": 191500
+    },
+    {
+      "epoch": 8.093411457235595,
+      "grad_norm": 1.2660714387893677,
+      "learning_rate": 6.87861767354114e-05,
+      "loss": 4.0123,
+      "step": 192000
+    },
+    {
+      "epoch": 8.114488049572145,
+      "grad_norm": 1.3551139831542969,
+      "learning_rate": 6.870042447369551e-05,
+      "loss": 4.0191,
+      "step": 192500
+    },
+    {
+      "epoch": 8.135564641908696,
+      "grad_norm": 1.2952131032943726,
+      "learning_rate": 6.861467221197959e-05,
+      "loss": 4.0179,
+      "step": 193000
+    },
+    {
+      "epoch": 8.156641234245248,
+      "grad_norm": 1.3086978197097778,
+      "learning_rate": 6.852891995026369e-05,
+      "loss": 4.0193,
+      "step": 193500
+    },
+    {
+      "epoch": 8.177717826581798,
+      "grad_norm": 1.2983644008636475,
+      "learning_rate": 6.844316768854779e-05,
+      "loss": 4.0188,
+      "step": 194000
+    },
+    {
+      "epoch": 8.19879441891835,
+      "grad_norm": 1.2912441492080688,
+      "learning_rate": 6.835741542683189e-05,
+      "loss": 4.0169,
+      "step": 194500
+    },
+    {
+      "epoch": 8.2198710112549,
+      "grad_norm": 1.306266188621521,
+      "learning_rate": 6.827166316511599e-05,
+      "loss": 4.014,
+      "step": 195000
+    },
+    {
+      "epoch": 8.24094760359145,
+      "grad_norm": 1.2981733083724976,
+      "learning_rate": 6.818591090340008e-05,
+      "loss": 4.0152,
+      "step": 195500
+    },
+    {
+      "epoch": 8.262024195928003,
+      "grad_norm": 1.2391489744186401,
+      "learning_rate": 6.810015864168417e-05,
+      "loss": 4.0132,
+      "step": 196000
+    },
+    {
+      "epoch": 8.283100788264553,
+      "grad_norm": 1.2956194877624512,
+      "learning_rate": 6.801440637996828e-05,
+      "loss": 4.0108,
+      "step": 196500
+    },
+    {
+      "epoch": 8.304177380601104,
+      "grad_norm": 1.3567521572113037,
+      "learning_rate": 6.792865411825238e-05,
+      "loss": 4.0141,
+      "step": 197000
+    },
+    {
+      "epoch": 8.325253972937656,
+      "grad_norm": 1.3292361497879028,
+      "learning_rate": 6.784290185653646e-05,
+      "loss": 4.0109,
+      "step": 197500
+    },
+    {
+      "epoch": 8.346330565274206,
+      "grad_norm": 1.3758251667022705,
+      "learning_rate": 6.775714959482056e-05,
+      "loss": 4.0108,
+      "step": 198000
+    },
+    {
+      "epoch": 8.367407157610758,
+      "grad_norm": 1.3197747468948364,
+      "learning_rate": 6.767139733310467e-05,
+      "loss": 4.0124,
+      "step": 198500
+    },
+    {
+      "epoch": 8.388483749947309,
+      "grad_norm": 1.3496253490447998,
+      "learning_rate": 6.758564507138876e-05,
+      "loss": 4.0156,
+      "step": 199000
+    },
+    {
+      "epoch": 8.409560342283859,
+      "grad_norm": 1.2221095561981201,
+      "learning_rate": 6.749989280967286e-05,
+      "loss": 4.0152,
+      "step": 199500
+    },
+    {
+      "epoch": 8.430636934620411,
+      "grad_norm": 1.3079551458358765,
+      "learning_rate": 6.741414054795696e-05,
+      "loss": 4.0102,
+      "step": 200000
+    },
+    {
+      "epoch": 8.430636934620411,
+      "eval_accuracy": 0.7018057152534135,
+      "eval_loss": 3.7429323196411133,
+      "eval_runtime": 452.9195,
+      "eval_samples_per_second": 266.507,
+      "eval_steps_per_second": 2.778,
+      "step": 200000
+    },
+    {
+      "epoch": 8.451713526956961,
+      "grad_norm": 1.3180047273635864,
+      "learning_rate": 6.732838828624105e-05,
+      "loss": 4.0097,
+      "step": 200500
+    },
+    {
+      "epoch": 8.472790119293512,
+      "grad_norm": 1.326328992843628,
+      "learning_rate": 6.724263602452515e-05,
+      "loss": 4.0121,
+      "step": 201000
+    },
+    {
+      "epoch": 8.493866711630064,
+      "grad_norm": 1.3245339393615723,
+      "learning_rate": 6.715688376280924e-05,
+      "loss": 4.0083,
+      "step": 201500
+    },
+    {
+      "epoch": 8.514943303966614,
+      "grad_norm": 1.2803126573562622,
+      "learning_rate": 6.707113150109335e-05,
+      "loss": 4.0088,
+      "step": 202000
+    },
+    {
+      "epoch": 8.536019896303166,
+      "grad_norm": 1.3394393920898438,
+      "learning_rate": 6.698537923937745e-05,
+      "loss": 4.0114,
+      "step": 202500
+    },
+    {
+      "epoch": 8.557096488639717,
+      "grad_norm": 1.2923822402954102,
+      "learning_rate": 6.689962697766153e-05,
+      "loss": 4.0089,
+      "step": 203000
+    },
+    {
+      "epoch": 8.578173080976267,
+      "grad_norm": 1.3765919208526611,
+      "learning_rate": 6.681387471594563e-05,
+      "loss": 4.0074,
+      "step": 203500
+    },
+    {
+      "epoch": 8.59924967331282,
+      "grad_norm": 1.2911906242370605,
+      "learning_rate": 6.672812245422974e-05,
+      "loss": 4.0064,
+      "step": 204000
+    },
+    {
+      "epoch": 8.62032626564937,
+      "grad_norm": 1.302582859992981,
+      "learning_rate": 6.664237019251383e-05,
+      "loss": 4.0077,
+      "step": 204500
+    },
+    {
+      "epoch": 8.64140285798592,
+      "grad_norm": 1.3165802955627441,
+      "learning_rate": 6.655661793079792e-05,
+      "loss": 4.0076,
+      "step": 205000
+    },
+    {
+      "epoch": 8.662479450322472,
+      "grad_norm": 1.3761391639709473,
+      "learning_rate": 6.647086566908202e-05,
+      "loss": 4.0031,
+      "step": 205500
+    },
+    {
+      "epoch": 8.683556042659022,
+      "grad_norm": 1.3036824464797974,
+      "learning_rate": 6.638511340736612e-05,
+      "loss": 4.0039,
+      "step": 206000
+    },
+    {
+      "epoch": 8.704632634995574,
+      "grad_norm": 1.3353699445724487,
+      "learning_rate": 6.629936114565022e-05,
+      "loss": 4.0009,
+      "step": 206500
+    },
+    {
+      "epoch": 8.725709227332125,
+      "grad_norm": 1.348020076751709,
+      "learning_rate": 6.621360888393432e-05,
+      "loss": 4.0036,
+      "step": 207000
+    },
+    {
+      "epoch": 8.746785819668675,
+      "grad_norm": 1.2659090757369995,
+      "learning_rate": 6.61278566222184e-05,
+      "loss": 3.9964,
+      "step": 207500
+    },
+    {
+      "epoch": 8.767862412005227,
+      "grad_norm": 1.3252865076065063,
+      "learning_rate": 6.604210436050251e-05,
+      "loss": 4.003,
+      "step": 208000
+    },
+    {
+      "epoch": 8.788939004341778,
+      "grad_norm": 1.344516634941101,
+      "learning_rate": 6.595635209878661e-05,
+      "loss": 4.0042,
+      "step": 208500
+    },
+    {
+      "epoch": 8.81001559667833,
+      "grad_norm": 1.2576165199279785,
+      "learning_rate": 6.58705998370707e-05,
+      "loss": 4.0038,
+      "step": 209000
+    },
+    {
+      "epoch": 8.83109218901488,
+      "grad_norm": 1.2966135740280151,
+      "learning_rate": 6.57848475753548e-05,
+      "loss": 4.0014,
+      "step": 209500
+    },
+    {
+      "epoch": 8.85216878135143,
+      "grad_norm": 1.3721510171890259,
+      "learning_rate": 6.569909531363891e-05,
+      "loss": 3.9985,
+      "step": 210000
+    },
+    {
+      "epoch": 8.85216878135143,
+      "eval_accuracy": 0.7030542926243116,
+      "eval_loss": 3.735699415206909,
+      "eval_runtime": 449.25,
+      "eval_samples_per_second": 268.683,
+      "eval_steps_per_second": 2.8,
+      "step": 210000
+    },
+    {
+      "epoch": 8.873245373687983,
+      "grad_norm": 1.2964622974395752,
+      "learning_rate": 6.561334305192299e-05,
+      "loss": 4.0,
+      "step": 210500
+    },
+    {
+      "epoch": 8.894321966024533,
+      "grad_norm": 1.2438499927520752,
+      "learning_rate": 6.552759079020709e-05,
+      "loss": 4.0019,
+      "step": 211000
+    },
+    {
+      "epoch": 8.915398558361083,
+      "grad_norm": 1.2948734760284424,
+      "learning_rate": 6.544183852849119e-05,
+      "loss": 3.9984,
+      "step": 211500
+    },
+    {
+      "epoch": 8.936475150697635,
+      "grad_norm": 1.3451166152954102,
+      "learning_rate": 6.535608626677529e-05,
+      "loss": 4.0057,
+      "step": 212000
+    },
+    {
+      "epoch": 8.957551743034186,
+      "grad_norm": 1.278096079826355,
+      "learning_rate": 6.527033400505939e-05,
+      "loss": 4.003,
+      "step": 212500
+    },
+    {
+      "epoch": 8.978628335370738,
+      "grad_norm": 1.3273265361785889,
+      "learning_rate": 6.518458174334348e-05,
+      "loss": 4.0023,
+      "step": 213000
+    },
+    {
+      "epoch": 8.999704927707288,
+      "grad_norm": 1.34186851978302,
+      "learning_rate": 6.509882948162758e-05,
+      "loss": 4.0011,
+      "step": 213500
+    },
+    {
+      "epoch": 9.020781520043839,
+      "grad_norm": 1.3863059282302856,
+      "learning_rate": 6.501307721991168e-05,
+      "loss": 3.9961,
+      "step": 214000
+    },
+    {
+      "epoch": 9.04185811238039,
+      "grad_norm": 1.2717092037200928,
+      "learning_rate": 6.492732495819578e-05,
+      "loss": 3.9935,
+      "step": 214500
+    },
+    {
+      "epoch": 9.062934704716941,
+      "grad_norm": 1.3696802854537964,
+      "learning_rate": 6.484157269647986e-05,
+      "loss": 3.997,
+      "step": 215000
+    },
+    {
+      "epoch": 9.084011297053493,
+      "grad_norm": 1.357700228691101,
+      "learning_rate": 6.475582043476398e-05,
+      "loss": 3.9964,
+      "step": 215500
+    },
+    {
+      "epoch": 9.105087889390044,
+      "grad_norm": 1.3730473518371582,
+      "learning_rate": 6.467006817304807e-05,
+      "loss": 3.9933,
+      "step": 216000
+    },
+    {
+      "epoch": 9.126164481726594,
+      "grad_norm": 1.2956339120864868,
+      "learning_rate": 6.458431591133216e-05,
+      "loss": 3.9865,
+      "step": 216500
+    },
+    {
+      "epoch": 9.147241074063146,
+      "grad_norm": 1.3330892324447632,
+      "learning_rate": 6.449856364961626e-05,
+      "loss": 3.9941,
+      "step": 217000
+    },
+    {
+      "epoch": 9.168317666399696,
+      "grad_norm": 1.3341469764709473,
+      "learning_rate": 6.441281138790037e-05,
+      "loss": 3.9909,
+      "step": 217500
+    },
+    {
+      "epoch": 9.189394258736247,
+      "grad_norm": 1.3247278928756714,
+      "learning_rate": 6.432705912618445e-05,
+      "loss": 3.9922,
+      "step": 218000
+    },
+    {
+      "epoch": 9.210470851072799,
+      "grad_norm": 1.288197636604309,
+      "learning_rate": 6.424130686446855e-05,
+      "loss": 3.9955,
+      "step": 218500
+    },
+    {
+      "epoch": 9.23154744340935,
+      "grad_norm": 1.3217947483062744,
+      "learning_rate": 6.415555460275265e-05,
+      "loss": 3.9912,
+      "step": 219000
+    },
+    {
+      "epoch": 9.252624035745901,
+      "grad_norm": 1.3281906843185425,
+      "learning_rate": 6.406980234103675e-05,
+      "loss": 3.9925,
+      "step": 219500
+    },
+    {
+      "epoch": 9.273700628082452,
+      "grad_norm": 1.400024175643921,
+      "learning_rate": 6.398405007932085e-05,
+      "loss": 3.9959,
+      "step": 220000
+    },
+    {
+      "epoch": 9.273700628082452,
+      "eval_accuracy": 0.7043909447736604,
+      "eval_loss": 3.726987361907959,
+      "eval_runtime": 446.4463,
+      "eval_samples_per_second": 270.371,
+      "eval_steps_per_second": 2.818,
+      "step": 220000
+    },
+    {
+      "epoch": 9.294777220419002,
+      "grad_norm": 1.2846523523330688,
+      "learning_rate": 6.389829781760495e-05,
+      "loss": 3.99,
+      "step": 220500
+    },
+    {
+      "epoch": 9.315853812755554,
+      "grad_norm": 1.3969154357910156,
+      "learning_rate": 6.381254555588903e-05,
+      "loss": 3.9925,
+      "step": 221000
+    },
+    {
+      "epoch": 9.336930405092104,
+      "grad_norm": 1.3198400735855103,
+      "learning_rate": 6.372679329417314e-05,
+      "loss": 3.9887,
+      "step": 221500
+    },
+    {
+      "epoch": 9.358006997428657,
+      "grad_norm": 1.3707762956619263,
+      "learning_rate": 6.364104103245724e-05,
+      "loss": 3.9838,
+      "step": 222000
+    },
+    {
+      "epoch": 9.379083589765207,
+      "grad_norm": 1.2777632474899292,
+      "learning_rate": 6.355528877074132e-05,
+      "loss": 3.9913,
+      "step": 222500
+    },
+    {
+      "epoch": 9.400160182101757,
+      "grad_norm": 1.3324333429336548,
+      "learning_rate": 6.346953650902542e-05,
+      "loss": 3.9891,
+      "step": 223000
+    },
+    {
+      "epoch": 9.42123677443831,
+      "grad_norm": 1.3184070587158203,
+      "learning_rate": 6.338378424730954e-05,
+      "loss": 3.9883,
+      "step": 223500
+    },
+    {
+      "epoch": 9.44231336677486,
+      "grad_norm": 1.3994964361190796,
+      "learning_rate": 6.329803198559362e-05,
+      "loss": 3.9867,
+      "step": 224000
+    },
+    {
+      "epoch": 9.46338995911141,
+      "grad_norm": 1.330715298652649,
+      "learning_rate": 6.321227972387772e-05,
+      "loss": 3.9869,
+      "step": 224500
+    },
+    {
+      "epoch": 9.484466551447962,
+      "grad_norm": 1.2989274263381958,
+      "learning_rate": 6.312652746216182e-05,
+      "loss": 3.9828,
+      "step": 225000
+    },
+    {
+      "epoch": 9.505543143784513,
+      "grad_norm": 1.3056893348693848,
+      "learning_rate": 6.304077520044591e-05,
+      "loss": 3.9904,
+      "step": 225500
+    },
+    {
+      "epoch": 9.526619736121065,
+      "grad_norm": 1.3067275285720825,
+      "learning_rate": 6.295502293873001e-05,
+      "loss": 3.9858,
+      "step": 226000
+    },
+    {
+      "epoch": 9.547696328457615,
+      "grad_norm": 1.3893433809280396,
+      "learning_rate": 6.286927067701411e-05,
+      "loss": 3.9824,
+      "step": 226500
+    },
+    {
+      "epoch": 9.568772920794165,
+      "grad_norm": 1.414292335510254,
+      "learning_rate": 6.278351841529821e-05,
+      "loss": 3.986,
+      "step": 227000
+    },
+    {
+      "epoch": 9.589849513130718,
+      "grad_norm": 1.3334447145462036,
+      "learning_rate": 6.269776615358231e-05,
+      "loss": 3.9858,
+      "step": 227500
+    },
+    {
+      "epoch": 9.610926105467268,
+      "grad_norm": 1.300183653831482,
+      "learning_rate": 6.26120138918664e-05,
+      "loss": 3.9835,
+      "step": 228000
+    },
+    {
+      "epoch": 9.63200269780382,
+      "grad_norm": 1.346247673034668,
+      "learning_rate": 6.252626163015049e-05,
+      "loss": 3.9845,
+      "step": 228500
+    },
+    {
+      "epoch": 9.65307929014037,
+      "grad_norm": 1.3147140741348267,
+      "learning_rate": 6.24405093684346e-05,
+      "loss": 3.9811,
+      "step": 229000
+    },
+    {
+      "epoch": 9.67415588247692,
+      "grad_norm": 1.3234723806381226,
+      "learning_rate": 6.23547571067187e-05,
+      "loss": 3.9809,
+      "step": 229500
+    },
+    {
+      "epoch": 9.695232474813473,
+      "grad_norm": 1.285512089729309,
+      "learning_rate": 6.226900484500279e-05,
+      "loss": 3.9781,
+      "step": 230000
+    },
+    {
+      "epoch": 9.695232474813473,
+      "eval_accuracy": 0.7057613390175896,
+      "eval_loss": 3.7160940170288086,
+      "eval_runtime": 450.0882,
+      "eval_samples_per_second": 268.183,
+      "eval_steps_per_second": 2.795,
+      "step": 230000
+    },
+    {
+      "epoch": 9.716309067150023,
+      "grad_norm": 1.300737738609314,
+      "learning_rate": 6.218325258328688e-05,
+      "loss": 3.9867,
+      "step": 230500
+    },
+    {
+      "epoch": 9.737385659486574,
+      "grad_norm": 1.320038080215454,
+      "learning_rate": 6.209750032157098e-05,
+      "loss": 3.9847,
+      "step": 231000
+    },
+    {
+      "epoch": 9.758462251823126,
+      "grad_norm": 1.3389012813568115,
+      "learning_rate": 6.201174805985508e-05,
+      "loss": 3.9794,
+      "step": 231500
+    },
+    {
+      "epoch": 9.779538844159676,
+      "grad_norm": 1.341302752494812,
+      "learning_rate": 6.192599579813918e-05,
+      "loss": 3.9808,
+      "step": 232000
+    },
+    {
+      "epoch": 9.800615436496226,
+      "grad_norm": 1.3164373636245728,
+      "learning_rate": 6.184024353642328e-05,
+      "loss": 3.9797,
+      "step": 232500
+    },
+    {
+      "epoch": 9.821692028832778,
+      "grad_norm": 1.3226326704025269,
+      "learning_rate": 6.175449127470738e-05,
+      "loss": 3.9781,
+      "step": 233000
+    },
+    {
+      "epoch": 9.842768621169329,
+      "grad_norm": 1.3816205263137817,
+      "learning_rate": 6.166873901299147e-05,
+      "loss": 3.9844,
+      "step": 233500
+    },
+    {
+      "epoch": 9.863845213505881,
+      "grad_norm": 1.3084101676940918,
+      "learning_rate": 6.158298675127557e-05,
+      "loss": 3.9774,
+      "step": 234000
+    },
+    {
+      "epoch": 9.884921805842431,
+      "grad_norm": 1.3305227756500244,
+      "learning_rate": 6.149723448955966e-05,
+      "loss": 3.9819,
+      "step": 234500
+    },
+    {
+      "epoch": 9.905998398178982,
+      "grad_norm": 1.2952533960342407,
+      "learning_rate": 6.141148222784377e-05,
+      "loss": 3.9739,
+      "step": 235000
+    },
+    {
+      "epoch": 9.927074990515534,
+      "grad_norm": 1.2237471342086792,
+      "learning_rate": 6.132572996612787e-05,
+      "loss": 3.9766,
+      "step": 235500
+    },
+    {
+      "epoch": 9.948151582852084,
+      "grad_norm": 1.3733131885528564,
+      "learning_rate": 6.123997770441195e-05,
+      "loss": 3.979,
+      "step": 236000
+    },
+    {
+      "epoch": 9.969228175188636,
+      "grad_norm": 1.3964906930923462,
+      "learning_rate": 6.115422544269605e-05,
+      "loss": 3.9825,
+      "step": 236500
+    },
+    {
+      "epoch": 9.990304767525187,
+      "grad_norm": 1.3560963869094849,
+      "learning_rate": 6.106847318098015e-05,
+      "loss": 3.9793,
+      "step": 237000
+    },
+    {
+      "epoch": 10.011381359861737,
+      "grad_norm": 1.277612328529358,
+      "learning_rate": 6.098272091926425e-05,
+      "loss": 3.9777,
+      "step": 237500
+    },
+    {
+      "epoch": 10.032457952198289,
+      "grad_norm": 1.255475640296936,
+      "learning_rate": 6.0896968657548345e-05,
+      "loss": 3.977,
+      "step": 238000
+    },
+    {
+      "epoch": 10.05353454453484,
+      "grad_norm": 1.3513543605804443,
+      "learning_rate": 6.081121639583244e-05,
+      "loss": 3.9712,
+      "step": 238500
+    },
+    {
+      "epoch": 10.07461113687139,
+      "grad_norm": 1.2540451288223267,
+      "learning_rate": 6.072546413411654e-05,
+      "loss": 3.9742,
+      "step": 239000
+    },
+    {
+      "epoch": 10.095687729207942,
+      "grad_norm": 1.3529155254364014,
+      "learning_rate": 6.063971187240064e-05,
+      "loss": 3.9736,
+      "step": 239500
+    },
+    {
+      "epoch": 10.116764321544492,
+      "grad_norm": 1.3082735538482666,
+      "learning_rate": 6.055395961068473e-05,
+      "loss": 3.9676,
+      "step": 240000
+    },
+    {
+      "epoch": 10.116764321544492,
+      "eval_accuracy": 0.706924505851032,
+      "eval_loss": 3.7122199535369873,
+      "eval_runtime": 467.1574,
+      "eval_samples_per_second": 258.384,
+      "eval_steps_per_second": 2.693,
+      "step": 240000
+    },
+    {
+      "epoch": 10.137840913881044,
+      "grad_norm": 1.3115047216415405,
+      "learning_rate": 6.046820734896883e-05,
+      "loss": 3.9703,
+      "step": 240500
+    },
+    {
+      "epoch": 10.158917506217595,
+      "grad_norm": 1.3535202741622925,
+      "learning_rate": 6.038245508725293e-05,
+      "loss": 3.972,
+      "step": 241000
+    },
+    {
+      "epoch": 10.179994098554145,
+      "grad_norm": 1.3090511560440063,
+      "learning_rate": 6.029670282553702e-05,
+      "loss": 3.9707,
+      "step": 241500
+    },
+    {
+      "epoch": 10.201070690890697,
+      "grad_norm": 1.311006784439087,
+      "learning_rate": 6.0210950563821125e-05,
+      "loss": 3.9753,
+      "step": 242000
+    },
+    {
+      "epoch": 10.222147283227248,
+      "grad_norm": 1.3313339948654175,
+      "learning_rate": 6.0125198302105223e-05,
+      "loss": 3.9717,
+      "step": 242500
+    },
+    {
+      "epoch": 10.2432238755638,
+      "grad_norm": 1.4123648405075073,
+      "learning_rate": 6.0039446040389315e-05,
+      "loss": 3.9706,
+      "step": 243000
+    },
+    {
+      "epoch": 10.26430046790035,
+      "grad_norm": 1.3286292552947998,
+      "learning_rate": 5.995369377867341e-05,
+      "loss": 3.9749,
+      "step": 243500
+    },
+    {
+      "epoch": 10.2853770602369,
+      "grad_norm": 1.2810206413269043,
+      "learning_rate": 5.986794151695752e-05,
+      "loss": 3.9658,
+      "step": 244000
+    },
+    {
+      "epoch": 10.306453652573452,
+      "grad_norm": 1.2968727350234985,
+      "learning_rate": 5.97821892552416e-05,
+      "loss": 3.9732,
+      "step": 244500
+    },
+    {
+      "epoch": 10.327530244910003,
+      "grad_norm": 1.2787100076675415,
+      "learning_rate": 5.969643699352571e-05,
+      "loss": 3.9706,
+      "step": 245000
+    },
+    {
+      "epoch": 10.348606837246553,
+      "grad_norm": 1.3248896598815918,
+      "learning_rate": 5.9610684731809806e-05,
+      "loss": 3.97,
+      "step": 245500
+    },
+    {
+      "epoch": 10.369683429583105,
+      "grad_norm": 1.3055840730667114,
+      "learning_rate": 5.95249324700939e-05,
+      "loss": 3.9693,
+      "step": 246000
+    },
+    {
+      "epoch": 10.390760021919656,
+      "grad_norm": 1.29989492893219,
+      "learning_rate": 5.9439180208377996e-05,
+      "loss": 3.9681,
+      "step": 246500
+    },
+    {
+      "epoch": 10.411836614256208,
+      "grad_norm": 1.3006956577301025,
+      "learning_rate": 5.93534279466621e-05,
+      "loss": 3.9684,
+      "step": 247000
+    },
+    {
+      "epoch": 10.432913206592758,
+      "grad_norm": 1.3117653131484985,
+      "learning_rate": 5.9267675684946186e-05,
+      "loss": 3.9706,
+      "step": 247500
+    },
+    {
+      "epoch": 10.453989798929308,
+      "grad_norm": 1.3730003833770752,
+      "learning_rate": 5.918192342323029e-05,
+      "loss": 3.9651,
+      "step": 248000
+    },
+    {
+      "epoch": 10.47506639126586,
+      "grad_norm": 1.3368933200836182,
+      "learning_rate": 5.909617116151439e-05,
+      "loss": 3.9666,
+      "step": 248500
+    },
+    {
+      "epoch": 10.496142983602411,
+      "grad_norm": 1.4073560237884521,
+      "learning_rate": 5.901041889979848e-05,
+      "loss": 3.9647,
+      "step": 249000
+    },
+    {
+      "epoch": 10.517219575938963,
+      "grad_norm": 1.3097727298736572,
+      "learning_rate": 5.892466663808258e-05,
+      "loss": 3.9669,
+      "step": 249500
+    },
+    {
+      "epoch": 10.538296168275513,
+      "grad_norm": 1.315977931022644,
+      "learning_rate": 5.8838914376366684e-05,
+      "loss": 3.9672,
+      "step": 250000
+    },
+    {
+      "epoch": 10.538296168275513,
+      "eval_accuracy": 0.7080399347699322,
+      "eval_loss": 3.7028486728668213,
+      "eval_runtime": 473.9259,
+      "eval_samples_per_second": 254.694,
+      "eval_steps_per_second": 2.654,
+      "step": 250000
+    },
+    {
+      "epoch": 10.559372760612064,
+      "grad_norm": 1.3029711246490479,
+      "learning_rate": 5.8753162114650776e-05,
+      "loss": 3.9713,
+      "step": 250500
+    },
+    {
+      "epoch": 10.580449352948616,
+      "grad_norm": 1.3542011976242065,
+      "learning_rate": 5.8667409852934874e-05,
+      "loss": 3.9647,
+      "step": 251000
+    },
+    {
+      "epoch": 10.601525945285166,
+      "grad_norm": 1.2729461193084717,
+      "learning_rate": 5.858165759121897e-05,
+      "loss": 3.9616,
+      "step": 251500
+    },
+    {
+      "epoch": 10.622602537621717,
+      "grad_norm": 1.331905722618103,
+      "learning_rate": 5.8495905329503064e-05,
+      "loss": 3.9664,
+      "step": 252000
+    },
+    {
+      "epoch": 10.643679129958269,
+      "grad_norm": 1.2787367105484009,
+      "learning_rate": 5.841015306778716e-05,
+      "loss": 3.9655,
+      "step": 252500
+    },
+    {
+      "epoch": 10.664755722294819,
+      "grad_norm": 1.3172341585159302,
+      "learning_rate": 5.832440080607127e-05,
+      "loss": 3.9672,
+      "step": 253000
+    },
+    {
+      "epoch": 10.685832314631371,
+      "grad_norm": 1.41899573802948,
+      "learning_rate": 5.823864854435536e-05,
+      "loss": 3.9604,
+      "step": 253500
+    },
+    {
+      "epoch": 10.706908906967922,
+      "grad_norm": 1.3492988348007202,
+      "learning_rate": 5.815289628263946e-05,
+      "loss": 3.9625,
+      "step": 254000
+    },
+    {
+      "epoch": 10.727985499304472,
+      "grad_norm": 1.3632858991622925,
+      "learning_rate": 5.8067144020923556e-05,
+      "loss": 3.9685,
+      "step": 254500
+    },
+    {
+      "epoch": 10.749062091641024,
+      "grad_norm": 1.3340235948562622,
+      "learning_rate": 5.798139175920765e-05,
+      "loss": 3.964,
+      "step": 255000
+    },
+    {
+      "epoch": 10.770138683977574,
+      "grad_norm": 1.3169883489608765,
+      "learning_rate": 5.789563949749175e-05,
+      "loss": 3.9657,
+      "step": 255500
+    },
+    {
+      "epoch": 10.791215276314126,
+      "grad_norm": 1.350654125213623,
+      "learning_rate": 5.780988723577585e-05,
+      "loss": 3.9641,
+      "step": 256000
+    },
+    {
+      "epoch": 10.812291868650677,
+      "grad_norm": 1.33375883102417,
+      "learning_rate": 5.772413497405994e-05,
+      "loss": 3.9646,
+      "step": 256500
+    },
+    {
+      "epoch": 10.833368460987227,
+      "grad_norm": 1.341024398803711,
+      "learning_rate": 5.763838271234404e-05,
+      "loss": 3.9607,
+      "step": 257000
+    },
+    {
+      "epoch": 10.85444505332378,
+      "grad_norm": 1.3053178787231445,
+      "learning_rate": 5.7552630450628146e-05,
+      "loss": 3.9626,
+      "step": 257500
+    },
+    {
+      "epoch": 10.87552164566033,
+      "grad_norm": 1.3103811740875244,
+      "learning_rate": 5.746687818891223e-05,
+      "loss": 3.9599,
+      "step": 258000
+    },
+    {
+      "epoch": 10.89659823799688,
+      "grad_norm": 1.3415056467056274,
+      "learning_rate": 5.7381125927196335e-05,
+      "loss": 3.9623,
+      "step": 258500
+    },
+    {
+      "epoch": 10.917674830333432,
+      "grad_norm": 1.3429609537124634,
+      "learning_rate": 5.7295373665480434e-05,
+      "loss": 3.962,
+      "step": 259000
+    },
+    {
+      "epoch": 10.938751422669982,
+      "grad_norm": 1.324076533317566,
+      "learning_rate": 5.7209621403764525e-05,
+      "loss": 3.9603,
+      "step": 259500
+    },
+    {
+      "epoch": 10.959828015006533,
+      "grad_norm": 1.293860912322998,
+      "learning_rate": 5.7123869142048624e-05,
+      "loss": 3.9573,
+      "step": 260000
+    },
+    {
+      "epoch": 10.959828015006533,
+      "eval_accuracy": 0.7092515432042134,
+      "eval_loss": 3.696906328201294,
+      "eval_runtime": 457.5027,
+      "eval_samples_per_second": 263.837,
+      "eval_steps_per_second": 2.75,
+      "step": 260000
+    },
+    {
+      "epoch": 10.980904607343085,
+      "grad_norm": 1.317211389541626,
+      "learning_rate": 5.703811688033273e-05,
+      "loss": 3.9552,
+      "step": 260500
+    },
+    {
+      "epoch": 11.001981199679635,
+      "grad_norm": 1.318996787071228,
+      "learning_rate": 5.695236461861681e-05,
+      "loss": 3.957,
+      "step": 261000
+    },
+    {
+      "epoch": 11.023057792016187,
+      "grad_norm": 1.3456875085830688,
+      "learning_rate": 5.686661235690092e-05,
+      "loss": 3.9526,
+      "step": 261500
+    },
+    {
+      "epoch": 11.044134384352738,
+      "grad_norm": 1.3531146049499512,
+      "learning_rate": 5.678086009518502e-05,
+      "loss": 3.9595,
+      "step": 262000
+    },
+    {
+      "epoch": 11.065210976689288,
+      "grad_norm": 1.4028359651565552,
+      "learning_rate": 5.669510783346911e-05,
+      "loss": 3.9587,
+      "step": 262500
+    },
+    {
+      "epoch": 11.08628756902584,
+      "grad_norm": 1.3248058557510376,
+      "learning_rate": 5.6609355571753207e-05,
+      "loss": 3.9549,
+      "step": 263000
+    },
+    {
+      "epoch": 11.10736416136239,
+      "grad_norm": 1.421217918395996,
+      "learning_rate": 5.652360331003731e-05,
+      "loss": 3.9567,
+      "step": 263500
+    },
+    {
+      "epoch": 11.128440753698943,
+      "grad_norm": 1.374590277671814,
+      "learning_rate": 5.6437851048321396e-05,
+      "loss": 3.9534,
+      "step": 264000
+    },
+    {
+      "epoch": 11.149517346035493,
+      "grad_norm": 1.3743417263031006,
+      "learning_rate": 5.63520987866055e-05,
+      "loss": 3.9561,
+      "step": 264500
+    },
+    {
+      "epoch": 11.170593938372043,
+      "grad_norm": 1.3148473501205444,
+      "learning_rate": 5.62663465248896e-05,
+      "loss": 3.9533,
+      "step": 265000
+    },
+    {
+      "epoch": 11.191670530708596,
+      "grad_norm": 1.3165109157562256,
+      "learning_rate": 5.618059426317369e-05,
+      "loss": 3.9532,
+      "step": 265500
+    },
+    {
+      "epoch": 11.212747123045146,
+      "grad_norm": 1.3530795574188232,
+      "learning_rate": 5.609484200145779e-05,
+      "loss": 3.9568,
+      "step": 266000
+    },
+    {
+      "epoch": 11.233823715381696,
+      "grad_norm": 1.3105723857879639,
+      "learning_rate": 5.6009089739741895e-05,
+      "loss": 3.956,
+      "step": 266500
+    },
+    {
+      "epoch": 11.254900307718248,
+      "grad_norm": 1.4123492240905762,
+      "learning_rate": 5.5923337478025986e-05,
+      "loss": 3.9518,
+      "step": 267000
+    },
+    {
+      "epoch": 11.275976900054799,
+      "grad_norm": 1.3657774925231934,
+      "learning_rate": 5.5837585216310085e-05,
+      "loss": 3.955,
+      "step": 267500
+    },
+    {
+      "epoch": 11.29705349239135,
+      "grad_norm": 1.3964136838912964,
+      "learning_rate": 5.575183295459418e-05,
+      "loss": 3.9503,
+      "step": 268000
+    },
+    {
+      "epoch": 11.318130084727901,
+      "grad_norm": 1.4228318929672241,
+      "learning_rate": 5.5666080692878274e-05,
+      "loss": 3.9539,
+      "step": 268500
+    },
+    {
+      "epoch": 11.339206677064452,
+      "grad_norm": 1.3295122385025024,
+      "learning_rate": 5.558032843116238e-05,
+      "loss": 3.9543,
+      "step": 269000
+    },
+    {
+      "epoch": 11.360283269401004,
+      "grad_norm": 1.3555238246917725,
+      "learning_rate": 5.549457616944648e-05,
+      "loss": 3.9531,
+      "step": 269500
+    },
+    {
+      "epoch": 11.381359861737554,
+      "grad_norm": 1.3527286052703857,
+      "learning_rate": 5.540882390773057e-05,
+      "loss": 3.9497,
+      "step": 270000
+    },
+    {
+      "epoch": 11.381359861737554,
+      "eval_accuracy": 0.7097630167913684,
+      "eval_loss": 3.6923487186431885,
+      "eval_runtime": 452.852,
+      "eval_samples_per_second": 266.546,
+      "eval_steps_per_second": 2.778,
+      "step": 270000
+    },
+    {
+      "epoch": 11.402436454074106,
+      "grad_norm": 1.3435646295547485,
+      "learning_rate": 5.532307164601467e-05,
+      "loss": 3.953,
+      "step": 270500
+    },
+    {
+      "epoch": 11.423513046410656,
+      "grad_norm": 1.3339743614196777,
+      "learning_rate": 5.5237319384298766e-05,
+      "loss": 3.9502,
+      "step": 271000
+    },
+    {
+      "epoch": 11.444589638747207,
+      "grad_norm": 1.281876564025879,
+      "learning_rate": 5.515156712258286e-05,
+      "loss": 3.9475,
+      "step": 271500
+    },
+    {
+      "epoch": 11.465666231083759,
+      "grad_norm": 1.3370720148086548,
+      "learning_rate": 5.506581486086696e-05,
+      "loss": 3.9497,
+      "step": 272000
+    },
+    {
+      "epoch": 11.48674282342031,
+      "grad_norm": 1.3272731304168701,
+      "learning_rate": 5.498006259915105e-05,
+      "loss": 3.9494,
+      "step": 272500
+    },
+    {
+      "epoch": 11.50781941575686,
+      "grad_norm": 1.3450390100479126,
+      "learning_rate": 5.489431033743515e-05,
+      "loss": 3.9489,
+      "step": 273000
+    },
+    {
+      "epoch": 11.528896008093412,
+      "grad_norm": 1.3018174171447754,
+      "learning_rate": 5.480855807571925e-05,
+      "loss": 3.953,
+      "step": 273500
+    },
+    {
+      "epoch": 11.549972600429962,
+      "grad_norm": 1.3982748985290527,
+      "learning_rate": 5.472280581400334e-05,
+      "loss": 3.9476,
+      "step": 274000
+    },
+    {
+      "epoch": 11.571049192766514,
+      "grad_norm": 1.3610074520111084,
+      "learning_rate": 5.463705355228744e-05,
+      "loss": 3.9496,
+      "step": 274500
+    },
+    {
+      "epoch": 11.592125785103065,
+      "grad_norm": 1.321820855140686,
+      "learning_rate": 5.4551301290571546e-05,
+      "loss": 3.9516,
+      "step": 275000
+    },
+    {
+      "epoch": 11.613202377439615,
+      "grad_norm": 1.3456435203552246,
+      "learning_rate": 5.446554902885563e-05,
+      "loss": 3.9489,
+      "step": 275500
+    },
+    {
+      "epoch": 11.634278969776167,
+      "grad_norm": 1.329667568206787,
+      "learning_rate": 5.4379796767139736e-05,
+      "loss": 3.9465,
+      "step": 276000
+    },
+    {
+      "epoch": 11.655355562112717,
+      "grad_norm": 1.337854266166687,
+      "learning_rate": 5.4294044505423834e-05,
+      "loss": 3.9472,
+      "step": 276500
+    },
+    {
+      "epoch": 11.67643215444927,
+      "grad_norm": 1.3235496282577515,
+      "learning_rate": 5.4208292243707925e-05,
+      "loss": 3.9467,
+      "step": 277000
+    },
+    {
+      "epoch": 11.69750874678582,
+      "grad_norm": 1.33723783493042,
+      "learning_rate": 5.4122539981992024e-05,
+      "loss": 3.9479,
+      "step": 277500
+    },
+    {
+      "epoch": 11.71858533912237,
+      "grad_norm": 1.3138431310653687,
+      "learning_rate": 5.403678772027613e-05,
+      "loss": 3.9431,
+      "step": 278000
+    },
+    {
+      "epoch": 11.739661931458922,
+      "grad_norm": 1.3257819414138794,
+      "learning_rate": 5.395103545856022e-05,
+      "loss": 3.948,
+      "step": 278500
+    },
+    {
+      "epoch": 11.760738523795473,
+      "grad_norm": 1.3107185363769531,
+      "learning_rate": 5.386528319684432e-05,
+      "loss": 3.9454,
+      "step": 279000
+    },
+    {
+      "epoch": 11.781815116132023,
+      "grad_norm": 1.311624526977539,
+      "learning_rate": 5.377953093512842e-05,
+      "loss": 3.9425,
+      "step": 279500
+    },
+    {
+      "epoch": 11.802891708468575,
+      "grad_norm": 1.3920516967773438,
+      "learning_rate": 5.369377867341251e-05,
+      "loss": 3.9483,
+      "step": 280000
+    },
+    {
+      "epoch": 11.802891708468575,
+      "eval_accuracy": 0.7110541710307098,
+      "eval_loss": 3.6840808391571045,
+      "eval_runtime": 451.4635,
+      "eval_samples_per_second": 267.366,
+      "eval_steps_per_second": 2.786,
+      "step": 280000
+    },
+    {
+      "epoch": 11.823968300805126,
+      "grad_norm": 1.38111412525177,
+      "learning_rate": 5.3608026411696614e-05,
+      "loss": 3.9491,
+      "step": 280500
+    },
+    {
+      "epoch": 11.845044893141678,
+      "grad_norm": 1.3376792669296265,
+      "learning_rate": 5.352227414998071e-05,
+      "loss": 3.9403,
+      "step": 281000
+    },
+    {
+      "epoch": 11.866121485478228,
+      "grad_norm": 1.4167546033859253,
+      "learning_rate": 5.34365218882648e-05,
+      "loss": 3.9427,
+      "step": 281500
+    },
+    {
+      "epoch": 11.887198077814778,
+      "grad_norm": 1.2788422107696533,
+      "learning_rate": 5.33507696265489e-05,
+      "loss": 3.9388,
+      "step": 282000
+    },
+    {
+      "epoch": 11.90827467015133,
+      "grad_norm": 1.4020700454711914,
+      "learning_rate": 5.3265017364833e-05,
+      "loss": 3.9473,
+      "step": 282500
+    },
+    {
+      "epoch": 11.92935126248788,
+      "grad_norm": 1.4099026918411255,
+      "learning_rate": 5.317926510311709e-05,
+      "loss": 3.9402,
+      "step": 283000
+    },
+    {
+      "epoch": 11.950427854824433,
+      "grad_norm": 1.411189317703247,
+      "learning_rate": 5.3093512841401197e-05,
+      "loss": 3.9394,
+      "step": 283500
+    },
+    {
+      "epoch": 11.971504447160983,
+      "grad_norm": 1.3513870239257812,
+      "learning_rate": 5.3007760579685295e-05,
+      "loss": 3.9447,
+      "step": 284000
+    },
+    {
+      "epoch": 11.992581039497534,
+      "grad_norm": 1.3571823835372925,
+      "learning_rate": 5.2922008317969386e-05,
+      "loss": 3.9465,
+      "step": 284500
+    },
+    {
+      "epoch": 12.013657631834086,
+      "grad_norm": 1.3905842304229736,
+      "learning_rate": 5.2836256056253485e-05,
+      "loss": 3.9404,
+      "step": 285000
+    },
+    {
+      "epoch": 12.034734224170636,
+      "grad_norm": 1.3413423299789429,
+      "learning_rate": 5.275050379453759e-05,
+      "loss": 3.9416,
+      "step": 285500
+    },
+    {
+      "epoch": 12.055810816507186,
+      "grad_norm": 1.2919282913208008,
+      "learning_rate": 5.2664751532821675e-05,
+      "loss": 3.9393,
+      "step": 286000
+    },
+    {
+      "epoch": 12.076887408843739,
+      "grad_norm": 1.3663338422775269,
+      "learning_rate": 5.257899927110578e-05,
+      "loss": 3.9379,
+      "step": 286500
+    },
+    {
+      "epoch": 12.097964001180289,
+      "grad_norm": 1.3264641761779785,
+      "learning_rate": 5.249324700938988e-05,
+      "loss": 3.937,
+      "step": 287000
+    },
+    {
+      "epoch": 12.119040593516841,
+      "grad_norm": 1.3299224376678467,
+      "learning_rate": 5.240749474767397e-05,
+      "loss": 3.9347,
+      "step": 287500
+    },
+    {
+      "epoch": 12.140117185853391,
+      "grad_norm": 1.3693588972091675,
+      "learning_rate": 5.232174248595807e-05,
+      "loss": 3.94,
+      "step": 288000
+    },
+    {
+      "epoch": 12.161193778189942,
+      "grad_norm": 1.3290989398956299,
+      "learning_rate": 5.223599022424217e-05,
+      "loss": 3.94,
+      "step": 288500
+    },
+    {
+      "epoch": 12.182270370526494,
+      "grad_norm": 1.3596065044403076,
+      "learning_rate": 5.215023796252626e-05,
+      "loss": 3.9449,
+      "step": 289000
+    },
+    {
+      "epoch": 12.203346962863044,
+      "grad_norm": 1.3313069343566895,
+      "learning_rate": 5.206448570081036e-05,
+      "loss": 3.934,
+      "step": 289500
+    },
+    {
+      "epoch": 12.224423555199595,
+      "grad_norm": 1.3465008735656738,
+      "learning_rate": 5.197873343909446e-05,
+      "loss": 3.9348,
+      "step": 290000
+    },
+    {
+      "epoch": 12.224423555199595,
+      "eval_accuracy": 0.7118759736686153,
+      "eval_loss": 3.6785988807678223,
+      "eval_runtime": 428.4669,
+      "eval_samples_per_second": 281.716,
+      "eval_steps_per_second": 2.936,
+      "step": 290000
+    },
+    {
+      "epoch": 12.245500147536147,
+      "grad_norm": 1.3097912073135376,
+      "learning_rate": 5.189298117737855e-05,
+      "loss": 3.9373,
+      "step": 290500
+    },
+    {
+      "epoch": 12.266576739872697,
+      "grad_norm": 1.356719732284546,
+      "learning_rate": 5.180722891566265e-05,
+      "loss": 3.9355,
+      "step": 291000
+    },
+    {
+      "epoch": 12.28765333220925,
+      "grad_norm": 1.3698605298995972,
+      "learning_rate": 5.1721476653946756e-05,
+      "loss": 3.9332,
+      "step": 291500
+    },
+    {
+      "epoch": 12.3087299245458,
+      "grad_norm": 1.329801082611084,
+      "learning_rate": 5.163572439223085e-05,
+      "loss": 3.9367,
+      "step": 292000
+    },
+    {
+      "epoch": 12.32980651688235,
+      "grad_norm": 1.275471568107605,
+      "learning_rate": 5.1549972130514946e-05,
+      "loss": 3.937,
+      "step": 292500
+    },
+    {
+      "epoch": 12.350883109218902,
+      "grad_norm": 1.392946481704712,
+      "learning_rate": 5.1464219868799044e-05,
+      "loss": 3.9354,
+      "step": 293000
+    },
+    {
+      "epoch": 12.371959701555452,
+      "grad_norm": 1.3525758981704712,
+      "learning_rate": 5.1378467607083136e-05,
+      "loss": 3.9325,
+      "step": 293500
+    },
+    {
+      "epoch": 12.393036293892003,
+      "grad_norm": 1.366074562072754,
+      "learning_rate": 5.1292715345367234e-05,
+      "loss": 3.9338,
+      "step": 294000
+    },
+    {
+      "epoch": 12.414112886228555,
+      "grad_norm": 1.337479591369629,
+      "learning_rate": 5.120696308365134e-05,
+      "loss": 3.9321,
+      "step": 294500
+    },
+    {
+      "epoch": 12.435189478565105,
+      "grad_norm": 1.3327606916427612,
+      "learning_rate": 5.112121082193543e-05,
+      "loss": 3.9355,
+      "step": 295000
+    },
+    {
+      "epoch": 12.456266070901657,
+      "grad_norm": 1.3819180727005005,
+      "learning_rate": 5.103545856021953e-05,
+      "loss": 3.9328,
+      "step": 295500
+    },
+    {
+      "epoch": 12.477342663238208,
+      "grad_norm": 1.3422201871871948,
+      "learning_rate": 5.094970629850363e-05,
+      "loss": 3.9344,
+      "step": 296000
+    },
+    {
+      "epoch": 12.498419255574758,
+      "grad_norm": 1.3337793350219727,
+      "learning_rate": 5.086395403678772e-05,
+      "loss": 3.9328,
+      "step": 296500
+    },
+    {
+      "epoch": 12.51949584791131,
+      "grad_norm": 1.3290514945983887,
+      "learning_rate": 5.0778201775071824e-05,
+      "loss": 3.9391,
+      "step": 297000
+    },
+    {
+      "epoch": 12.54057244024786,
+      "grad_norm": 1.3194193840026855,
+      "learning_rate": 5.069244951335592e-05,
+      "loss": 3.9354,
+      "step": 297500
+    },
+    {
+      "epoch": 12.561649032584413,
+      "grad_norm": 1.3082661628723145,
+      "learning_rate": 5.0606697251640014e-05,
+      "loss": 3.9315,
+      "step": 298000
+    },
+    {
+      "epoch": 12.582725624920963,
+      "grad_norm": 1.3915637731552124,
+      "learning_rate": 5.052094498992411e-05,
+      "loss": 3.9351,
+      "step": 298500
+    },
+    {
+      "epoch": 12.603802217257513,
+      "grad_norm": 1.3420003652572632,
+      "learning_rate": 5.043519272820822e-05,
+      "loss": 3.9356,
+      "step": 299000
+    },
+    {
+      "epoch": 12.624878809594065,
+      "grad_norm": 1.3456547260284424,
+      "learning_rate": 5.03494404664923e-05,
+      "loss": 3.9291,
+      "step": 299500
+    },
+    {
+      "epoch": 12.645955401930616,
+      "grad_norm": 1.3348097801208496,
+      "learning_rate": 5.026368820477641e-05,
+      "loss": 3.9304,
+      "step": 300000
+    },
+    {
+      "epoch": 12.645955401930616,
+      "eval_accuracy": 0.7131752116622819,
+      "eval_loss": 3.6702513694763184,
+      "eval_runtime": 440.9939,
+      "eval_samples_per_second": 273.714,
+      "eval_steps_per_second": 2.853,
+      "step": 300000
+    },
+    {
+      "epoch": 12.667031994267166,
+      "grad_norm": 1.3779054880142212,
+      "learning_rate": 5.0177935943060505e-05,
+      "loss": 3.9266,
+      "step": 300500
+    },
+    {
+      "epoch": 12.688108586603718,
+      "grad_norm": 1.347821831703186,
+      "learning_rate": 5.00921836813446e-05,
+      "loss": 3.9351,
+      "step": 301000
+    },
+    {
+      "epoch": 12.709185178940269,
+      "grad_norm": 1.3372251987457275,
+      "learning_rate": 5.0006431419628695e-05,
+      "loss": 3.9313,
+      "step": 301500
+    },
+    {
+      "epoch": 12.73026177127682,
+      "grad_norm": 1.3927217721939087,
+      "learning_rate": 4.992067915791279e-05,
+      "loss": 3.9326,
+      "step": 302000
+    },
+    {
+      "epoch": 12.751338363613371,
+      "grad_norm": 1.355270504951477,
+      "learning_rate": 4.9834926896196885e-05,
+      "loss": 3.9261,
+      "step": 302500
+    },
+    {
+      "epoch": 12.772414955949921,
+      "grad_norm": 1.3721328973770142,
+      "learning_rate": 4.974917463448099e-05,
+      "loss": 3.9328,
+      "step": 303000
+    },
+    {
+      "epoch": 12.793491548286474,
+      "grad_norm": 1.3368809223175049,
+      "learning_rate": 4.966342237276508e-05,
+      "loss": 3.9328,
+      "step": 303500
+    },
+    {
+      "epoch": 12.814568140623024,
+      "grad_norm": 1.318307876586914,
+      "learning_rate": 4.957767011104918e-05,
+      "loss": 3.9304,
+      "step": 304000
+    },
+    {
+      "epoch": 12.835644732959576,
+      "grad_norm": 1.3181530237197876,
+      "learning_rate": 4.949191784933328e-05,
+      "loss": 3.9297,
+      "step": 304500
+    },
+    {
+      "epoch": 12.856721325296126,
+      "grad_norm": 1.3040449619293213,
+      "learning_rate": 4.9406165587617376e-05,
+      "loss": 3.9276,
+      "step": 305000
+    },
+    {
+      "epoch": 12.877797917632677,
+      "grad_norm": 1.3593695163726807,
+      "learning_rate": 4.932041332590147e-05,
+      "loss": 3.9319,
+      "step": 305500
+    },
+    {
+      "epoch": 12.898874509969229,
+      "grad_norm": 1.4277350902557373,
+      "learning_rate": 4.923466106418557e-05,
+      "loss": 3.9283,
+      "step": 306000
+    },
+    {
+      "epoch": 12.91995110230578,
+      "grad_norm": 1.343863606452942,
+      "learning_rate": 4.9148908802469665e-05,
+      "loss": 3.9251,
+      "step": 306500
+    },
+    {
+      "epoch": 12.94102769464233,
+      "grad_norm": 1.3200805187225342,
+      "learning_rate": 4.906315654075376e-05,
+      "loss": 3.9282,
+      "step": 307000
+    },
+    {
+      "epoch": 12.962104286978882,
+      "grad_norm": 1.4030075073242188,
+      "learning_rate": 4.897740427903786e-05,
+      "loss": 3.9329,
+      "step": 307500
+    },
+    {
+      "epoch": 12.983180879315432,
+      "grad_norm": 1.368229866027832,
+      "learning_rate": 4.889165201732196e-05,
+      "loss": 3.9312,
+      "step": 308000
+    },
+    {
+      "epoch": 13.004257471651984,
+      "grad_norm": 1.3602405786514282,
+      "learning_rate": 4.880589975560606e-05,
+      "loss": 3.9281,
+      "step": 308500
+    },
+    {
+      "epoch": 13.025334063988534,
+      "grad_norm": 1.3407719135284424,
+      "learning_rate": 4.8720147493890156e-05,
+      "loss": 3.9201,
+      "step": 309000
+    },
+    {
+      "epoch": 13.046410656325085,
+      "grad_norm": 1.3522615432739258,
+      "learning_rate": 4.8634395232174254e-05,
+      "loss": 3.9246,
+      "step": 309500
+    },
+    {
+      "epoch": 13.067487248661637,
+      "grad_norm": 1.3443254232406616,
+      "learning_rate": 4.8548642970458346e-05,
+      "loss": 3.9205,
+      "step": 310000
+    },
+    {
+      "epoch": 13.067487248661637,
+      "eval_accuracy": 0.7139006420764209,
+      "eval_loss": 3.6665778160095215,
+      "eval_runtime": 549.3651,
+      "eval_samples_per_second": 219.719,
+      "eval_steps_per_second": 2.29,
+      "step": 310000
+    },
+    {
+      "epoch": 13.088563840998187,
+      "grad_norm": 1.35757577419281,
+      "learning_rate": 4.846289070874245e-05,
+      "loss": 3.9248,
+      "step": 310500
+    },
+    {
+      "epoch": 13.109640433334738,
+      "grad_norm": 1.40268075466156,
+      "learning_rate": 4.837713844702654e-05,
+      "loss": 3.9248,
+      "step": 311000
+    },
+    {
+      "epoch": 13.13071702567129,
+      "grad_norm": 1.3761773109436035,
+      "learning_rate": 4.829138618531064e-05,
+      "loss": 3.9206,
+      "step": 311500
+    },
+    {
+      "epoch": 13.15179361800784,
+      "grad_norm": 1.3423832654953003,
+      "learning_rate": 4.820563392359474e-05,
+      "loss": 3.9254,
+      "step": 312000
+    },
+    {
+      "epoch": 13.172870210344392,
+      "grad_norm": 1.3482491970062256,
+      "learning_rate": 4.811988166187884e-05,
+      "loss": 3.9211,
+      "step": 312500
+    },
+    {
+      "epoch": 13.193946802680943,
+      "grad_norm": 1.3268697261810303,
+      "learning_rate": 4.803412940016293e-05,
+      "loss": 3.9206,
+      "step": 313000
+    },
+    {
+      "epoch": 13.215023395017493,
+      "grad_norm": 1.3990137577056885,
+      "learning_rate": 4.7948377138447034e-05,
+      "loss": 3.9235,
+      "step": 313500
+    },
+    {
+      "epoch": 13.236099987354045,
+      "grad_norm": 1.3782047033309937,
+      "learning_rate": 4.7862624876731126e-05,
+      "loss": 3.9203,
+      "step": 314000
+    },
+    {
+      "epoch": 13.257176579690595,
+      "grad_norm": 1.3955236673355103,
+      "learning_rate": 4.7776872615015224e-05,
+      "loss": 3.9217,
+      "step": 314500
+    },
+    {
+      "epoch": 13.278253172027147,
+      "grad_norm": 1.3820618391036987,
+      "learning_rate": 4.7691120353299315e-05,
+      "loss": 3.9177,
+      "step": 315000
+    },
+    {
+      "epoch": 13.299329764363698,
+      "grad_norm": 1.3180475234985352,
+      "learning_rate": 4.760536809158342e-05,
+      "loss": 3.925,
+      "step": 315500
+    },
+    {
+      "epoch": 13.320406356700248,
+      "grad_norm": 1.3343690633773804,
+      "learning_rate": 4.751961582986751e-05,
+      "loss": 3.9217,
+      "step": 316000
+    },
+    {
+      "epoch": 13.3414829490368,
+      "grad_norm": 1.341372013092041,
+      "learning_rate": 4.743386356815161e-05,
+      "loss": 3.9228,
+      "step": 316500
+    },
+    {
+      "epoch": 13.36255954137335,
+      "grad_norm": 1.2916312217712402,
+      "learning_rate": 4.734811130643571e-05,
+      "loss": 3.9201,
+      "step": 317000
+    },
+    {
+      "epoch": 13.383636133709901,
+      "grad_norm": 1.3684160709381104,
+      "learning_rate": 4.726235904471981e-05,
+      "loss": 3.9192,
+      "step": 317500
+    },
+    {
+      "epoch": 13.404712726046453,
+      "grad_norm": 1.4369447231292725,
+      "learning_rate": 4.71766067830039e-05,
+      "loss": 3.9218,
+      "step": 318000
+    },
+    {
+      "epoch": 13.425789318383003,
+      "grad_norm": 1.3311858177185059,
+      "learning_rate": 4.7090854521288004e-05,
+      "loss": 3.9232,
+      "step": 318500
+    },
+    {
+      "epoch": 13.446865910719556,
+      "grad_norm": 1.3725388050079346,
+      "learning_rate": 4.7005102259572095e-05,
+      "loss": 3.9196,
+      "step": 319000
+    },
+    {
+      "epoch": 13.467942503056106,
+      "grad_norm": 1.4294073581695557,
+      "learning_rate": 4.6919349997856193e-05,
+      "loss": 3.9216,
+      "step": 319500
+    },
+    {
+      "epoch": 13.489019095392656,
+      "grad_norm": 1.3710520267486572,
+      "learning_rate": 4.683359773614029e-05,
+      "loss": 3.9202,
+      "step": 320000
+    },
+    {
+      "epoch": 13.489019095392656,
+      "eval_accuracy": 0.7146433059668893,
+      "eval_loss": 3.6619887351989746,
+      "eval_runtime": 453.0866,
+      "eval_samples_per_second": 266.408,
+      "eval_steps_per_second": 2.777,
+      "step": 320000
+    },
+    {
+      "epoch": 13.510095687729208,
+      "grad_norm": 1.4209895133972168,
+      "learning_rate": 4.674784547442439e-05,
+      "loss": 3.9236,
+      "step": 320500
+    },
+    {
+      "epoch": 13.531172280065759,
+      "grad_norm": 1.3584743738174438,
+      "learning_rate": 4.666209321270849e-05,
+      "loss": 3.9198,
+      "step": 321000
+    },
+    {
+      "epoch": 13.55224887240231,
+      "grad_norm": 1.3449934720993042,
+      "learning_rate": 4.657634095099259e-05,
+      "loss": 3.9208,
+      "step": 321500
+    },
+    {
+      "epoch": 13.573325464738861,
+      "grad_norm": 1.4054151773452759,
+      "learning_rate": 4.6490588689276685e-05,
+      "loss": 3.9199,
+      "step": 322000
+    },
+    {
+      "epoch": 13.594402057075412,
+      "grad_norm": 1.357309103012085,
+      "learning_rate": 4.6404836427560777e-05,
+      "loss": 3.9219,
+      "step": 322500
+    },
+    {
+      "epoch": 13.615478649411964,
+      "grad_norm": 1.386186957359314,
+      "learning_rate": 4.6319084165844875e-05,
+      "loss": 3.9143,
+      "step": 323000
+    },
+    {
+      "epoch": 13.636555241748514,
+      "grad_norm": 1.3696508407592773,
+      "learning_rate": 4.623333190412897e-05,
+      "loss": 3.9209,
+      "step": 323500
+    },
+    {
+      "epoch": 13.657631834085064,
+      "grad_norm": 1.3496469259262085,
+      "learning_rate": 4.614757964241307e-05,
+      "loss": 3.916,
+      "step": 324000
+    },
+    {
+      "epoch": 13.678708426421617,
+      "grad_norm": 1.3469682931900024,
+      "learning_rate": 4.606182738069717e-05,
+      "loss": 3.9182,
+      "step": 324500
+    },
+    {
+      "epoch": 13.699785018758167,
+      "grad_norm": 1.3763799667358398,
+      "learning_rate": 4.597607511898127e-05,
+      "loss": 3.9168,
+      "step": 325000
+    },
+    {
+      "epoch": 13.720861611094719,
+      "grad_norm": 1.3688344955444336,
+      "learning_rate": 4.589032285726536e-05,
+      "loss": 3.918,
+      "step": 325500
+    },
+    {
+      "epoch": 13.74193820343127,
+      "grad_norm": 1.3626452684402466,
+      "learning_rate": 4.5804570595549465e-05,
+      "loss": 3.9171,
+      "step": 326000
+    },
+    {
+      "epoch": 13.76301479576782,
+      "grad_norm": 1.310539722442627,
+      "learning_rate": 4.5718818333833556e-05,
+      "loss": 3.919,
+      "step": 326500
+    },
+    {
+      "epoch": 13.784091388104372,
+      "grad_norm": 1.3512582778930664,
+      "learning_rate": 4.5633066072117655e-05,
+      "loss": 3.9195,
+      "step": 327000
+    },
+    {
+      "epoch": 13.805167980440922,
+      "grad_norm": 1.3566906452178955,
+      "learning_rate": 4.554731381040175e-05,
+      "loss": 3.9132,
+      "step": 327500
+    },
+    {
+      "epoch": 13.826244572777473,
+      "grad_norm": 1.3821688890457153,
+      "learning_rate": 4.546156154868585e-05,
+      "loss": 3.9165,
+      "step": 328000
+    },
+    {
+      "epoch": 13.847321165114025,
+      "grad_norm": 1.4025585651397705,
+      "learning_rate": 4.537580928696994e-05,
+      "loss": 3.9167,
+      "step": 328500
+    },
+    {
+      "epoch": 13.868397757450575,
+      "grad_norm": 1.3905564546585083,
+      "learning_rate": 4.529005702525405e-05,
+      "loss": 3.9164,
+      "step": 329000
+    },
+    {
+      "epoch": 13.889474349787127,
+      "grad_norm": 1.3904305696487427,
+      "learning_rate": 4.520430476353814e-05,
+      "loss": 3.9137,
+      "step": 329500
+    },
+    {
+      "epoch": 13.910550942123677,
+      "grad_norm": 1.4082283973693848,
+      "learning_rate": 4.511855250182224e-05,
+      "loss": 3.9152,
+      "step": 330000
+    },
+    {
+      "epoch": 13.910550942123677,
+      "eval_accuracy": 0.7152312618690059,
+      "eval_loss": 3.6580519676208496,
+      "eval_runtime": 420.4923,
+      "eval_samples_per_second": 287.059,
+      "eval_steps_per_second": 2.992,
+      "step": 330000
+    },
+    {
+      "epoch": 13.931627534460228,
+      "grad_norm": 1.3082201480865479,
+      "learning_rate": 4.5032800240106336e-05,
+      "loss": 3.9164,
+      "step": 330500
+    },
+    {
+      "epoch": 13.95270412679678,
+      "grad_norm": 1.3675941228866577,
+      "learning_rate": 4.4947047978390434e-05,
+      "loss": 3.9189,
+      "step": 331000
+    },
+    {
+      "epoch": 13.97378071913333,
+      "grad_norm": 1.3766281604766846,
+      "learning_rate": 4.4861295716674526e-05,
+      "loss": 3.9148,
+      "step": 331500
+    },
+    {
+      "epoch": 13.994857311469882,
+      "grad_norm": 1.3512485027313232,
+      "learning_rate": 4.4775543454958624e-05,
+      "loss": 3.9167,
+      "step": 332000
+    },
+    {
+      "epoch": 14.015933903806433,
+      "grad_norm": 1.3324275016784668,
+      "learning_rate": 4.468979119324272e-05,
+      "loss": 3.9094,
+      "step": 332500
+    },
+    {
+      "epoch": 14.037010496142983,
+      "grad_norm": 1.3503332138061523,
+      "learning_rate": 4.460403893152682e-05,
+      "loss": 3.9105,
+      "step": 333000
+    },
+    {
+      "epoch": 14.058087088479535,
+      "grad_norm": 1.397404432296753,
+      "learning_rate": 4.451828666981092e-05,
+      "loss": 3.9096,
+      "step": 333500
+    },
+    {
+      "epoch": 14.079163680816086,
+      "grad_norm": 1.3128241300582886,
+      "learning_rate": 4.443253440809502e-05,
+      "loss": 3.9103,
+      "step": 334000
+    },
+    {
+      "epoch": 14.100240273152636,
+      "grad_norm": 1.3401113748550415,
+      "learning_rate": 4.4346782146379116e-05,
+      "loss": 3.9101,
+      "step": 334500
+    },
+    {
+      "epoch": 14.121316865489188,
+      "grad_norm": 1.3705230951309204,
+      "learning_rate": 4.426102988466321e-05,
+      "loss": 3.9086,
+      "step": 335000
+    },
+    {
+      "epoch": 14.142393457825738,
+      "grad_norm": 1.3618402481079102,
+      "learning_rate": 4.4175277622947305e-05,
+      "loss": 3.9031,
+      "step": 335500
+    },
+    {
+      "epoch": 14.16347005016229,
+      "grad_norm": 1.3938554525375366,
+      "learning_rate": 4.4089525361231404e-05,
+      "loss": 3.9103,
+      "step": 336000
+    },
+    {
+      "epoch": 14.184546642498841,
+      "grad_norm": 1.3375216722488403,
+      "learning_rate": 4.40037730995155e-05,
+      "loss": 3.9039,
+      "step": 336500
+    },
+    {
+      "epoch": 14.205623234835391,
+      "grad_norm": 1.3672751188278198,
+      "learning_rate": 4.39180208377996e-05,
+      "loss": 3.9103,
+      "step": 337000
+    },
+    {
+      "epoch": 14.226699827171943,
+      "grad_norm": 1.4030122756958008,
+      "learning_rate": 4.38322685760837e-05,
+      "loss": 3.9095,
+      "step": 337500
+    },
+    {
+      "epoch": 14.247776419508494,
+      "grad_norm": 1.4050158262252808,
+      "learning_rate": 4.374651631436779e-05,
+      "loss": 3.9088,
+      "step": 338000
+    },
+    {
+      "epoch": 14.268853011845044,
+      "grad_norm": 1.3781143426895142,
+      "learning_rate": 4.3660764052651895e-05,
+      "loss": 3.9109,
+      "step": 338500
+    },
+    {
+      "epoch": 14.289929604181596,
+      "grad_norm": 1.3735519647598267,
+      "learning_rate": 4.357501179093599e-05,
+      "loss": 3.9085,
+      "step": 339000
+    },
+    {
+      "epoch": 14.311006196518147,
+      "grad_norm": 1.3722783327102661,
+      "learning_rate": 4.3489259529220085e-05,
+      "loss": 3.9084,
+      "step": 339500
+    },
+    {
+      "epoch": 14.332082788854699,
+      "grad_norm": 1.4383596181869507,
+      "learning_rate": 4.3403507267504183e-05,
+      "loss": 3.9105,
+      "step": 340000
+    },
+    {
+      "epoch": 14.332082788854699,
+      "eval_accuracy": 0.7167108275119457,
+      "eval_loss": 3.650089979171753,
+      "eval_runtime": 434.7007,
+      "eval_samples_per_second": 277.676,
+      "eval_steps_per_second": 2.894,
+      "step": 340000
+    },
+    {
+      "epoch": 14.353159381191249,
+      "grad_norm": 1.3863632678985596,
+      "learning_rate": 4.331775500578828e-05,
+      "loss": 3.9062,
+      "step": 340500
+    },
+    {
+      "epoch": 14.3742359735278,
+      "grad_norm": 1.346764087677002,
+      "learning_rate": 4.323200274407237e-05,
+      "loss": 3.905,
+      "step": 341000
+    },
+    {
+      "epoch": 14.395312565864351,
+      "grad_norm": 1.3845573663711548,
+      "learning_rate": 4.314625048235648e-05,
+      "loss": 3.9058,
+      "step": 341500
+    },
+    {
+      "epoch": 14.416389158200902,
+      "grad_norm": 1.3589117527008057,
+      "learning_rate": 4.306049822064057e-05,
+      "loss": 3.91,
+      "step": 342000
+    },
+    {
+      "epoch": 14.437465750537454,
+      "grad_norm": 1.352136254310608,
+      "learning_rate": 4.297474595892467e-05,
+      "loss": 3.9092,
+      "step": 342500
+    },
+    {
+      "epoch": 14.458542342874004,
+      "grad_norm": 1.3610610961914062,
+      "learning_rate": 4.2888993697208767e-05,
+      "loss": 3.9045,
+      "step": 343000
+    },
+    {
+      "epoch": 14.479618935210555,
+      "grad_norm": 1.3463726043701172,
+      "learning_rate": 4.2803241435492865e-05,
+      "loss": 3.9075,
+      "step": 343500
+    },
+    {
+      "epoch": 14.500695527547107,
+      "grad_norm": 1.3395646810531616,
+      "learning_rate": 4.2717489173776956e-05,
+      "loss": 3.9073,
+      "step": 344000
+    },
+    {
+      "epoch": 14.521772119883657,
+      "grad_norm": 1.3587359189987183,
+      "learning_rate": 4.263173691206106e-05,
+      "loss": 3.9035,
+      "step": 344500
+    },
+    {
+      "epoch": 14.542848712220207,
+      "grad_norm": 1.4076464176177979,
+      "learning_rate": 4.254598465034515e-05,
+      "loss": 3.9063,
+      "step": 345000
+    },
+    {
+      "epoch": 14.56392530455676,
+      "grad_norm": 1.3363561630249023,
+      "learning_rate": 4.246023238862925e-05,
+      "loss": 3.9051,
+      "step": 345500
+    },
+    {
+      "epoch": 14.58500189689331,
+      "grad_norm": 1.3723201751708984,
+      "learning_rate": 4.237448012691335e-05,
+      "loss": 3.9055,
+      "step": 346000
+    },
+    {
+      "epoch": 14.606078489229862,
+      "grad_norm": 1.357417345046997,
+      "learning_rate": 4.228872786519745e-05,
+      "loss": 3.9049,
+      "step": 346500
+    },
+    {
+      "epoch": 14.627155081566412,
+      "grad_norm": 1.3918380737304688,
+      "learning_rate": 4.220297560348154e-05,
+      "loss": 3.9071,
+      "step": 347000
+    },
+    {
+      "epoch": 14.648231673902963,
+      "grad_norm": 1.3359774351119995,
+      "learning_rate": 4.2117223341765645e-05,
+      "loss": 3.9044,
+      "step": 347500
+    },
+    {
+      "epoch": 14.669308266239515,
+      "grad_norm": 1.3652353286743164,
+      "learning_rate": 4.2031471080049736e-05,
+      "loss": 3.9021,
+      "step": 348000
+    },
+    {
+      "epoch": 14.690384858576065,
+      "grad_norm": 1.398751139640808,
+      "learning_rate": 4.1945718818333834e-05,
+      "loss": 3.9044,
+      "step": 348500
+    },
+    {
+      "epoch": 14.711461450912616,
+      "grad_norm": 1.347855567932129,
+      "learning_rate": 4.185996655661793e-05,
+      "loss": 3.9067,
+      "step": 349000
+    },
+    {
+      "epoch": 14.732538043249168,
+      "grad_norm": 1.3462238311767578,
+      "learning_rate": 4.177421429490203e-05,
+      "loss": 3.9067,
+      "step": 349500
+    },
+    {
+      "epoch": 14.753614635585718,
+      "grad_norm": 1.4078598022460938,
+      "learning_rate": 4.168846203318613e-05,
+      "loss": 3.9024,
+      "step": 350000
+    },
+    {
+      "epoch": 14.753614635585718,
+      "eval_accuracy": 0.7167539008982116,
+      "eval_loss": 3.6466331481933594,
+      "eval_runtime": 460.7019,
+      "eval_samples_per_second": 262.005,
+      "eval_steps_per_second": 2.731,
+      "step": 350000
+    },
+    {
+      "epoch": 14.77469122792227,
+      "grad_norm": 1.355332374572754,
+      "learning_rate": 4.160270977147022e-05,
+      "loss": 3.9027,
+      "step": 350500
+    },
+    {
+      "epoch": 14.79576782025882,
+      "grad_norm": 1.3641587495803833,
+      "learning_rate": 4.1516957509754326e-05,
+      "loss": 3.9049,
+      "step": 351000
+    },
+    {
+      "epoch": 14.81684441259537,
+      "grad_norm": 1.3629987239837646,
+      "learning_rate": 4.143120524803842e-05,
+      "loss": 3.9008,
+      "step": 351500
+    },
+    {
+      "epoch": 14.837921004931923,
+      "grad_norm": 1.3509442806243896,
+      "learning_rate": 4.1345452986322516e-05,
+      "loss": 3.9022,
+      "step": 352000
+    },
+    {
+      "epoch": 14.858997597268473,
+      "grad_norm": 1.3462828397750854,
+      "learning_rate": 4.1259700724606614e-05,
+      "loss": 3.9024,
+      "step": 352500
+    },
+    {
+      "epoch": 14.880074189605025,
+      "grad_norm": 1.358713984489441,
+      "learning_rate": 4.117394846289071e-05,
+      "loss": 3.9019,
+      "step": 353000
+    },
+    {
+      "epoch": 14.901150781941576,
+      "grad_norm": 1.3813903331756592,
+      "learning_rate": 4.1088196201174804e-05,
+      "loss": 3.9041,
+      "step": 353500
+    },
+    {
+      "epoch": 14.922227374278126,
+      "grad_norm": 1.3608967065811157,
+      "learning_rate": 4.100244393945891e-05,
+      "loss": 3.9004,
+      "step": 354000
+    },
+    {
+      "epoch": 14.943303966614678,
+      "grad_norm": 1.3387380838394165,
+      "learning_rate": 4.0916691677743e-05,
+      "loss": 3.9055,
+      "step": 354500
+    },
+    {
+      "epoch": 14.964380558951229,
+      "grad_norm": 1.3064637184143066,
+      "learning_rate": 4.08309394160271e-05,
+      "loss": 3.9041,
+      "step": 355000
+    },
+    {
+      "epoch": 14.985457151287779,
+      "grad_norm": 1.3509303331375122,
+      "learning_rate": 4.07451871543112e-05,
+      "loss": 3.9019,
+      "step": 355500
+    },
+    {
+      "epoch": 15.006533743624331,
+      "grad_norm": 1.3530592918395996,
+      "learning_rate": 4.0659434892595295e-05,
+      "loss": 3.8972,
+      "step": 356000
+    },
+    {
+      "epoch": 15.027610335960881,
+      "grad_norm": 1.4001927375793457,
+      "learning_rate": 4.057368263087939e-05,
+      "loss": 3.8924,
+      "step": 356500
+    },
+    {
+      "epoch": 15.048686928297434,
+      "grad_norm": 1.3543769121170044,
+      "learning_rate": 4.048793036916349e-05,
+      "loss": 3.8978,
+      "step": 357000
+    },
+    {
+      "epoch": 15.069763520633984,
+      "grad_norm": 1.4348479509353638,
+      "learning_rate": 4.0402178107447584e-05,
+      "loss": 3.9002,
+      "step": 357500
+    },
+    {
+      "epoch": 15.090840112970534,
+      "grad_norm": 1.4076569080352783,
+      "learning_rate": 4.031642584573168e-05,
+      "loss": 3.8966,
+      "step": 358000
+    },
+    {
+      "epoch": 15.111916705307086,
+      "grad_norm": 1.4077661037445068,
+      "learning_rate": 4.023067358401578e-05,
+      "loss": 3.8964,
+      "step": 358500
+    },
+    {
+      "epoch": 15.132993297643637,
+      "grad_norm": 1.39332115650177,
+      "learning_rate": 4.014492132229988e-05,
+      "loss": 3.8982,
+      "step": 359000
+    },
+    {
+      "epoch": 15.154069889980189,
+      "grad_norm": 1.3874651193618774,
+      "learning_rate": 4.005916906058397e-05,
+      "loss": 3.8973,
+      "step": 359500
+    },
+    {
+      "epoch": 15.17514648231674,
+      "grad_norm": 1.3772706985473633,
+      "learning_rate": 3.9973416798868075e-05,
+      "loss": 3.8976,
+      "step": 360000
+    },
+    {
+      "epoch": 15.17514648231674,
+      "eval_accuracy": 0.717306808959644,
+      "eval_loss": 3.644414186477661,
+      "eval_runtime": 419.553,
+      "eval_samples_per_second": 287.701,
+      "eval_steps_per_second": 2.998,
+      "step": 360000
+    },
+    {
+      "epoch": 15.19622307465329,
+      "grad_norm": 1.414156436920166,
+      "learning_rate": 3.988766453715217e-05,
+      "loss": 3.9002,
+      "step": 360500
+    },
+    {
+      "epoch": 15.217299666989842,
+      "grad_norm": 1.3461275100708008,
+      "learning_rate": 3.9801912275436265e-05,
+      "loss": 3.8984,
+      "step": 361000
+    },
+    {
+      "epoch": 15.238376259326392,
+      "grad_norm": 1.3717570304870605,
+      "learning_rate": 3.971616001372036e-05,
+      "loss": 3.8985,
+      "step": 361500
+    },
+    {
+      "epoch": 15.259452851662942,
+      "grad_norm": 1.417664885520935,
+      "learning_rate": 3.963040775200446e-05,
+      "loss": 3.8964,
+      "step": 362000
+    },
+    {
+      "epoch": 15.280529443999495,
+      "grad_norm": 1.4374829530715942,
+      "learning_rate": 3.954465549028856e-05,
+      "loss": 3.8952,
+      "step": 362500
+    },
+    {
+      "epoch": 15.301606036336045,
+      "grad_norm": 1.3791372776031494,
+      "learning_rate": 3.945890322857266e-05,
+      "loss": 3.8985,
+      "step": 363000
+    },
+    {
+      "epoch": 15.322682628672597,
+      "grad_norm": 1.4383161067962646,
+      "learning_rate": 3.9373150966856756e-05,
+      "loss": 3.8944,
+      "step": 363500
+    },
+    {
+      "epoch": 15.343759221009147,
+      "grad_norm": 1.3761835098266602,
+      "learning_rate": 3.928739870514085e-05,
+      "loss": 3.8958,
+      "step": 364000
+    },
+    {
+      "epoch": 15.364835813345698,
+      "grad_norm": 1.4183019399642944,
+      "learning_rate": 3.920164644342495e-05,
+      "loss": 3.8937,
+      "step": 364500
+    },
+    {
+      "epoch": 15.38591240568225,
+      "grad_norm": 1.392530918121338,
+      "learning_rate": 3.9115894181709045e-05,
+      "loss": 3.8955,
+      "step": 365000
+    },
+    {
+      "epoch": 15.4069889980188,
+      "grad_norm": 1.3752367496490479,
+      "learning_rate": 3.903014191999314e-05,
+      "loss": 3.8917,
+      "step": 365500
+    },
+    {
+      "epoch": 15.42806559035535,
+      "grad_norm": 1.4111636877059937,
+      "learning_rate": 3.894438965827724e-05,
+      "loss": 3.8981,
+      "step": 366000
+    },
+    {
+      "epoch": 15.449142182691903,
+      "grad_norm": 1.3426021337509155,
+      "learning_rate": 3.885863739656134e-05,
+      "loss": 3.8941,
+      "step": 366500
+    },
+    {
+      "epoch": 15.470218775028453,
+      "grad_norm": 1.406751275062561,
+      "learning_rate": 3.877288513484543e-05,
+      "loss": 3.8937,
+      "step": 367000
+    },
+    {
+      "epoch": 15.491295367365005,
+      "grad_norm": 1.3726617097854614,
+      "learning_rate": 3.868713287312953e-05,
+      "loss": 3.8949,
+      "step": 367500
+    },
+    {
+      "epoch": 15.512371959701555,
+      "grad_norm": 1.4048482179641724,
+      "learning_rate": 3.860138061141363e-05,
+      "loss": 3.8944,
+      "step": 368000
+    },
+    {
+      "epoch": 15.533448552038106,
+      "grad_norm": 1.3621785640716553,
+      "learning_rate": 3.8515628349697726e-05,
+      "loss": 3.8956,
+      "step": 368500
+    },
+    {
+      "epoch": 15.554525144374658,
+      "grad_norm": 1.3970487117767334,
+      "learning_rate": 3.842987608798182e-05,
+      "loss": 3.8943,
+      "step": 369000
+    },
+    {
+      "epoch": 15.575601736711208,
+      "grad_norm": 1.4143363237380981,
+      "learning_rate": 3.834412382626592e-05,
+      "loss": 3.8902,
+      "step": 369500
+    },
+    {
+      "epoch": 15.596678329047759,
+      "grad_norm": 1.351871371269226,
+      "learning_rate": 3.8258371564550014e-05,
+      "loss": 3.8948,
+      "step": 370000
+    },
+    {
+      "epoch": 15.596678329047759,
+      "eval_accuracy": 0.7185628559511288,
+      "eval_loss": 3.6362369060516357,
+      "eval_runtime": 382.986,
+      "eval_samples_per_second": 315.171,
+      "eval_steps_per_second": 3.285,
+      "step": 370000
+    },
+    {
+      "epoch": 15.61775492138431,
+      "grad_norm": 1.3705445528030396,
+      "learning_rate": 3.817261930283411e-05,
+      "loss": 3.8904,
+      "step": 370500
+    },
+    {
+      "epoch": 15.638831513720861,
+      "grad_norm": 1.443784236907959,
+      "learning_rate": 3.808686704111821e-05,
+      "loss": 3.8929,
+      "step": 371000
+    },
+    {
+      "epoch": 15.659908106057413,
+      "grad_norm": 1.423122763633728,
+      "learning_rate": 3.800111477940231e-05,
+      "loss": 3.8929,
+      "step": 371500
+    },
+    {
+      "epoch": 15.680984698393964,
+      "grad_norm": 1.3741027116775513,
+      "learning_rate": 3.79153625176864e-05,
+      "loss": 3.8941,
+      "step": 372000
+    },
+    {
+      "epoch": 15.702061290730514,
+      "grad_norm": 1.3473587036132812,
+      "learning_rate": 3.7829610255970506e-05,
+      "loss": 3.8963,
+      "step": 372500
+    },
+    {
+      "epoch": 15.723137883067066,
+      "grad_norm": 1.4233283996582031,
+      "learning_rate": 3.77438579942546e-05,
+      "loss": 3.8903,
+      "step": 373000
+    },
+    {
+      "epoch": 15.744214475403616,
+      "grad_norm": 1.3788650035858154,
+      "learning_rate": 3.7658105732538696e-05,
+      "loss": 3.887,
+      "step": 373500
+    },
+    {
+      "epoch": 15.765291067740169,
+      "grad_norm": 1.3347578048706055,
+      "learning_rate": 3.7572353470822794e-05,
+      "loss": 3.8916,
+      "step": 374000
+    },
+    {
+      "epoch": 15.786367660076719,
+      "grad_norm": 1.3971338272094727,
+      "learning_rate": 3.748660120910689e-05,
+      "loss": 3.8879,
+      "step": 374500
+    },
+    {
+      "epoch": 15.80744425241327,
+      "grad_norm": 1.4271931648254395,
+      "learning_rate": 3.740084894739099e-05,
+      "loss": 3.8859,
+      "step": 375000
+    },
+    {
+      "epoch": 15.828520844749821,
+      "grad_norm": 1.3356118202209473,
+      "learning_rate": 3.731509668567509e-05,
+      "loss": 3.8885,
+      "step": 375500
+    },
+    {
+      "epoch": 15.849597437086372,
+      "grad_norm": 1.3640503883361816,
+      "learning_rate": 3.722934442395919e-05,
+      "loss": 3.891,
+      "step": 376000
+    },
+    {
+      "epoch": 15.870674029422922,
+      "grad_norm": 1.3343900442123413,
+      "learning_rate": 3.714359216224328e-05,
+      "loss": 3.8893,
+      "step": 376500
+    },
+    {
+      "epoch": 15.891750621759474,
+      "grad_norm": 1.4103041887283325,
+      "learning_rate": 3.705783990052738e-05,
+      "loss": 3.8954,
+      "step": 377000
+    },
+    {
+      "epoch": 15.912827214096025,
+      "grad_norm": 1.3846709728240967,
+      "learning_rate": 3.6972087638811475e-05,
+      "loss": 3.8882,
+      "step": 377500
+    },
+    {
+      "epoch": 15.933903806432577,
+      "grad_norm": 1.3725395202636719,
+      "learning_rate": 3.6886335377095574e-05,
+      "loss": 3.8919,
+      "step": 378000
+    },
+    {
+      "epoch": 15.954980398769127,
+      "grad_norm": 1.3512264490127563,
+      "learning_rate": 3.680058311537967e-05,
+      "loss": 3.8898,
+      "step": 378500
+    },
+    {
+      "epoch": 15.976056991105677,
+      "grad_norm": 1.3892171382904053,
+      "learning_rate": 3.671483085366377e-05,
+      "loss": 3.891,
+      "step": 379000
+    },
+    {
+      "epoch": 15.99713358344223,
+      "grad_norm": 1.3391282558441162,
+      "learning_rate": 3.662907859194786e-05,
+      "loss": 3.8863,
+      "step": 379500
+    },
+    {
+      "epoch": 16.01821017577878,
+      "grad_norm": 1.420074462890625,
+      "learning_rate": 3.654332633023197e-05,
+      "loss": 3.8875,
+      "step": 380000
+    },
+    {
+      "epoch": 16.01821017577878,
+      "eval_accuracy": 0.7192764335972195,
+      "eval_loss": 3.631619453430176,
+      "eval_runtime": 703.6318,
+      "eval_samples_per_second": 171.547,
+      "eval_steps_per_second": 1.788,
+      "step": 380000
+    },
+    {
+      "epoch": 16.039286768115332,
+      "grad_norm": 1.3218660354614258,
+      "learning_rate": 3.645757406851606e-05,
+      "loss": 3.8845,
+      "step": 380500
+    },
+    {
+      "epoch": 16.06036336045188,
+      "grad_norm": 1.4044084548950195,
+      "learning_rate": 3.6371821806800157e-05,
+      "loss": 3.8852,
+      "step": 381000
+    },
+    {
+      "epoch": 16.081439952788433,
+      "grad_norm": 1.3800437450408936,
+      "learning_rate": 3.6286069545084255e-05,
+      "loss": 3.8904,
+      "step": 381500
+    },
+    {
+      "epoch": 16.102516545124985,
+      "grad_norm": 1.3692717552185059,
+      "learning_rate": 3.620031728336835e-05,
+      "loss": 3.8864,
+      "step": 382000
+    },
+    {
+      "epoch": 16.123593137461537,
+      "grad_norm": 1.394648790359497,
+      "learning_rate": 3.6114565021652445e-05,
+      "loss": 3.8798,
+      "step": 382500
+    },
+    {
+      "epoch": 16.144669729798085,
+      "grad_norm": 1.3787832260131836,
+      "learning_rate": 3.602881275993655e-05,
+      "loss": 3.8832,
+      "step": 383000
+    },
+    {
+      "epoch": 16.165746322134638,
+      "grad_norm": 1.3669034242630005,
+      "learning_rate": 3.594306049822064e-05,
+      "loss": 3.8858,
+      "step": 383500
+    },
+    {
+      "epoch": 16.18682291447119,
+      "grad_norm": 1.4652361869812012,
+      "learning_rate": 3.585730823650474e-05,
+      "loss": 3.8837,
+      "step": 384000
+    },
+    {
+      "epoch": 16.20789950680774,
+      "grad_norm": 1.3588998317718506,
+      "learning_rate": 3.577155597478883e-05,
+      "loss": 3.8855,
+      "step": 384500
+    },
+    {
+      "epoch": 16.22897609914429,
+      "grad_norm": 1.3947973251342773,
+      "learning_rate": 3.5685803713072936e-05,
+      "loss": 3.8814,
+      "step": 385000
+    },
+    {
+      "epoch": 16.250052691480843,
+      "grad_norm": 1.3708503246307373,
+      "learning_rate": 3.560005145135703e-05,
+      "loss": 3.884,
+      "step": 385500
+    },
+    {
+      "epoch": 16.27112928381739,
+      "grad_norm": 1.351488709449768,
+      "learning_rate": 3.5514299189641126e-05,
+      "loss": 3.8799,
+      "step": 386000
+    },
+    {
+      "epoch": 16.292205876153943,
+      "grad_norm": 1.4085997343063354,
+      "learning_rate": 3.5428546927925224e-05,
+      "loss": 3.8869,
+      "step": 386500
+    },
+    {
+      "epoch": 16.313282468490495,
+      "grad_norm": 1.372379183769226,
+      "learning_rate": 3.534279466620932e-05,
+      "loss": 3.8815,
+      "step": 387000
+    },
+    {
+      "epoch": 16.334359060827044,
+      "grad_norm": 1.3764355182647705,
+      "learning_rate": 3.525704240449342e-05,
+      "loss": 3.8828,
+      "step": 387500
+    },
+    {
+      "epoch": 16.355435653163596,
+      "grad_norm": 1.4133485555648804,
+      "learning_rate": 3.517129014277752e-05,
+      "loss": 3.8805,
+      "step": 388000
+    },
+    {
+      "epoch": 16.376512245500148,
+      "grad_norm": 1.4221103191375732,
+      "learning_rate": 3.508553788106161e-05,
+      "loss": 3.883,
+      "step": 388500
+    },
+    {
+      "epoch": 16.3975888378367,
+      "grad_norm": 1.3595232963562012,
+      "learning_rate": 3.499978561934571e-05,
+      "loss": 3.8809,
+      "step": 389000
+    },
+    {
+      "epoch": 16.41866543017325,
+      "grad_norm": 1.366932988166809,
+      "learning_rate": 3.491403335762981e-05,
+      "loss": 3.8812,
+      "step": 389500
+    },
+    {
+      "epoch": 16.4397420225098,
+      "grad_norm": 1.3680415153503418,
+      "learning_rate": 3.4828281095913906e-05,
+      "loss": 3.8842,
+      "step": 390000
+    },
+    {
+      "epoch": 16.4397420225098,
+      "eval_accuracy": 0.7198632894278593,
+      "eval_loss": 3.62803316116333,
+      "eval_runtime": 411.8597,
+      "eval_samples_per_second": 293.076,
+      "eval_steps_per_second": 3.054,
+      "step": 390000
+    },
+    {
+      "epoch": 16.460818614846353,
+      "grad_norm": 1.4084402322769165,
+      "learning_rate": 3.4742528834198004e-05,
+      "loss": 3.8797,
+      "step": 390500
+    },
+    {
+      "epoch": 16.4818952071829,
+      "grad_norm": 1.373098373413086,
+      "learning_rate": 3.46567765724821e-05,
+      "loss": 3.8831,
+      "step": 391000
+    },
+    {
+      "epoch": 16.502971799519454,
+      "grad_norm": 1.4028816223144531,
+      "learning_rate": 3.45710243107662e-05,
+      "loss": 3.8826,
+      "step": 391500
+    },
+    {
+      "epoch": 16.524048391856006,
+      "grad_norm": 1.4195351600646973,
+      "learning_rate": 3.448527204905029e-05,
+      "loss": 3.8844,
+      "step": 392000
+    },
+    {
+      "epoch": 16.545124984192555,
+      "grad_norm": 1.4147838354110718,
+      "learning_rate": 3.43995197873344e-05,
+      "loss": 3.8786,
+      "step": 392500
+    },
+    {
+      "epoch": 16.566201576529107,
+      "grad_norm": 1.4225722551345825,
+      "learning_rate": 3.431376752561849e-05,
+      "loss": 3.8817,
+      "step": 393000
+    },
+    {
+      "epoch": 16.58727816886566,
+      "grad_norm": 1.3380063772201538,
+      "learning_rate": 3.422801526390259e-05,
+      "loss": 3.8853,
+      "step": 393500
+    },
+    {
+      "epoch": 16.608354761202207,
+      "grad_norm": 1.3951219320297241,
+      "learning_rate": 3.4142263002186686e-05,
+      "loss": 3.8825,
+      "step": 394000
+    },
+    {
+      "epoch": 16.62943135353876,
+      "grad_norm": 1.4150776863098145,
+      "learning_rate": 3.4056510740470784e-05,
+      "loss": 3.8877,
+      "step": 394500
+    },
+    {
+      "epoch": 16.65050794587531,
+      "grad_norm": 1.3852022886276245,
+      "learning_rate": 3.3970758478754875e-05,
+      "loss": 3.8833,
+      "step": 395000
+    },
+    {
+      "epoch": 16.67158453821186,
+      "grad_norm": 1.3775135278701782,
+      "learning_rate": 3.388500621703898e-05,
+      "loss": 3.8829,
+      "step": 395500
+    },
+    {
+      "epoch": 16.692661130548412,
+      "grad_norm": 1.3926010131835938,
+      "learning_rate": 3.379925395532307e-05,
+      "loss": 3.8825,
+      "step": 396000
+    },
+    {
+      "epoch": 16.713737722884964,
+      "grad_norm": 1.366296410560608,
+      "learning_rate": 3.371350169360717e-05,
+      "loss": 3.8797,
+      "step": 396500
+    },
+    {
+      "epoch": 16.734814315221517,
+      "grad_norm": 1.4791797399520874,
+      "learning_rate": 3.362774943189127e-05,
+      "loss": 3.8806,
+      "step": 397000
+    },
+    {
+      "epoch": 16.755890907558065,
+      "grad_norm": 1.3370732069015503,
+      "learning_rate": 3.354199717017537e-05,
+      "loss": 3.881,
+      "step": 397500
+    },
+    {
+      "epoch": 16.776967499894617,
+      "grad_norm": 1.3912434577941895,
+      "learning_rate": 3.345624490845946e-05,
+      "loss": 3.8798,
+      "step": 398000
+    },
+    {
+      "epoch": 16.79804409223117,
+      "grad_norm": 1.4337376356124878,
+      "learning_rate": 3.3370492646743564e-05,
+      "loss": 3.8778,
+      "step": 398500
+    },
+    {
+      "epoch": 16.819120684567718,
+      "grad_norm": 1.3442845344543457,
+      "learning_rate": 3.3284740385027655e-05,
+      "loss": 3.8837,
+      "step": 399000
+    },
+    {
+      "epoch": 16.84019727690427,
+      "grad_norm": 1.3388097286224365,
+      "learning_rate": 3.319898812331175e-05,
+      "loss": 3.8764,
+      "step": 399500
+    },
+    {
+      "epoch": 16.861273869240822,
+      "grad_norm": 1.398321270942688,
+      "learning_rate": 3.311323586159585e-05,
+      "loss": 3.8773,
+      "step": 400000
+    },
+    {
+      "epoch": 16.861273869240822,
+      "eval_accuracy": 0.7202512508914409,
+      "eval_loss": 3.6256191730499268,
+      "eval_runtime": 432.4867,
+      "eval_samples_per_second": 279.098,
+      "eval_steps_per_second": 2.909,
+      "step": 400000
+    },
+    {
+      "epoch": 16.88235046157737,
+      "grad_norm": 1.4053997993469238,
+      "learning_rate": 3.302748359987995e-05,
+      "loss": 3.876,
+      "step": 400500
+    },
+    {
+      "epoch": 16.903427053913923,
+      "grad_norm": 1.398554801940918,
+      "learning_rate": 3.294173133816404e-05,
+      "loss": 3.8782,
+      "step": 401000
+    },
+    {
+      "epoch": 16.924503646250475,
+      "grad_norm": 1.4151557683944702,
+      "learning_rate": 3.2855979076448147e-05,
+      "loss": 3.8826,
+      "step": 401500
+    },
+    {
+      "epoch": 16.945580238587024,
+      "grad_norm": 1.4280130863189697,
+      "learning_rate": 3.277022681473224e-05,
+      "loss": 3.8772,
+      "step": 402000
+    },
+    {
+      "epoch": 16.966656830923576,
+      "grad_norm": 1.4069057703018188,
+      "learning_rate": 3.2684474553016336e-05,
+      "loss": 3.8801,
+      "step": 402500
+    },
+    {
+      "epoch": 16.987733423260128,
+      "grad_norm": 1.4176504611968994,
+      "learning_rate": 3.2598722291300435e-05,
+      "loss": 3.8793,
+      "step": 403000
+    },
+    {
+      "epoch": 17.00881001559668,
+      "grad_norm": 1.39607834815979,
+      "learning_rate": 3.251297002958453e-05,
+      "loss": 3.8769,
+      "step": 403500
+    },
+    {
+      "epoch": 17.02988660793323,
+      "grad_norm": 1.405268907546997,
+      "learning_rate": 3.242721776786863e-05,
+      "loss": 3.8764,
+      "step": 404000
+    },
+    {
+      "epoch": 17.05096320026978,
+      "grad_norm": 1.349086880683899,
+      "learning_rate": 3.234146550615272e-05,
+      "loss": 3.8735,
+      "step": 404500
+    },
+    {
+      "epoch": 17.072039792606333,
+      "grad_norm": 1.4098995923995972,
+      "learning_rate": 3.225571324443683e-05,
+      "loss": 3.8746,
+      "step": 405000
+    },
+    {
+      "epoch": 17.09311638494288,
+      "grad_norm": 1.4462028741836548,
+      "learning_rate": 3.216996098272092e-05,
+      "loss": 3.875,
+      "step": 405500
+    },
+    {
+      "epoch": 17.114192977279433,
+      "grad_norm": 1.4027295112609863,
+      "learning_rate": 3.208420872100502e-05,
+      "loss": 3.8754,
+      "step": 406000
+    },
+    {
+      "epoch": 17.135269569615986,
+      "grad_norm": 1.4020326137542725,
+      "learning_rate": 3.1998456459289116e-05,
+      "loss": 3.8716,
+      "step": 406500
+    },
+    {
+      "epoch": 17.156346161952534,
+      "grad_norm": 1.4437731504440308,
+      "learning_rate": 3.1912704197573214e-05,
+      "loss": 3.8712,
+      "step": 407000
+    },
+    {
+      "epoch": 17.177422754289086,
+      "grad_norm": 1.3855340480804443,
+      "learning_rate": 3.1826951935857306e-05,
+      "loss": 3.8708,
+      "step": 407500
+    },
+    {
+      "epoch": 17.19849934662564,
+      "grad_norm": 1.3807625770568848,
+      "learning_rate": 3.174119967414141e-05,
+      "loss": 3.8763,
+      "step": 408000
+    },
+    {
+      "epoch": 17.219575938962187,
+      "grad_norm": 1.4479700326919556,
+      "learning_rate": 3.16554474124255e-05,
+      "loss": 3.8733,
+      "step": 408500
+    },
+    {
+      "epoch": 17.24065253129874,
+      "grad_norm": 1.3731745481491089,
+      "learning_rate": 3.15696951507096e-05,
+      "loss": 3.8762,
+      "step": 409000
+    },
+    {
+      "epoch": 17.26172912363529,
+      "grad_norm": 1.4186406135559082,
+      "learning_rate": 3.14839428889937e-05,
+      "loss": 3.8733,
+      "step": 409500
+    },
+    {
+      "epoch": 17.282805715971843,
+      "grad_norm": 1.5018037557601929,
+      "learning_rate": 3.13981906272778e-05,
+      "loss": 3.8766,
+      "step": 410000
+    },
+    {
+      "epoch": 17.282805715971843,
+      "eval_accuracy": 0.7214251202295732,
+      "eval_loss": 3.619047164916992,
+      "eval_runtime": 443.6704,
+      "eval_samples_per_second": 272.062,
+      "eval_steps_per_second": 2.835,
+      "step": 410000
+    },
+    {
+      "epoch": 17.303882308308392,
+      "grad_norm": 1.4225496053695679,
+      "learning_rate": 3.131243836556189e-05,
+      "loss": 3.8759,
+      "step": 410500
+    },
+    {
+      "epoch": 17.324958900644944,
+      "grad_norm": 1.3987301588058472,
+      "learning_rate": 3.1226686103845994e-05,
+      "loss": 3.8694,
+      "step": 411000
+    },
+    {
+      "epoch": 17.346035492981496,
+      "grad_norm": 1.4396098852157593,
+      "learning_rate": 3.1140933842130086e-05,
+      "loss": 3.8708,
+      "step": 411500
+    },
+    {
+      "epoch": 17.367112085318045,
+      "grad_norm": 1.4017897844314575,
+      "learning_rate": 3.1055181580414184e-05,
+      "loss": 3.8742,
+      "step": 412000
+    },
+    {
+      "epoch": 17.388188677654597,
+      "grad_norm": 1.3844493627548218,
+      "learning_rate": 3.096942931869828e-05,
+      "loss": 3.8761,
+      "step": 412500
+    },
+    {
+      "epoch": 17.40926526999115,
+      "grad_norm": 1.3915551900863647,
+      "learning_rate": 3.088367705698238e-05,
+      "loss": 3.8695,
+      "step": 413000
+    },
+    {
+      "epoch": 17.430341862327698,
+      "grad_norm": 1.3475651741027832,
+      "learning_rate": 3.079792479526647e-05,
+      "loss": 3.8737,
+      "step": 413500
+    },
+    {
+      "epoch": 17.45141845466425,
+      "grad_norm": 1.3939110040664673,
+      "learning_rate": 3.071217253355058e-05,
+      "loss": 3.8713,
+      "step": 414000
+    },
+    {
+      "epoch": 17.472495047000802,
+      "grad_norm": 1.3998979330062866,
+      "learning_rate": 3.062642027183467e-05,
+      "loss": 3.8693,
+      "step": 414500
+    },
+    {
+      "epoch": 17.49357163933735,
+      "grad_norm": 1.3880306482315063,
+      "learning_rate": 3.054066801011877e-05,
+      "loss": 3.8718,
+      "step": 415000
+    },
+    {
+      "epoch": 17.514648231673903,
+      "grad_norm": 1.4597318172454834,
+      "learning_rate": 3.045491574840287e-05,
+      "loss": 3.8725,
+      "step": 415500
+    },
+    {
+      "epoch": 17.535724824010455,
+      "grad_norm": 1.4306273460388184,
+      "learning_rate": 3.0369163486686964e-05,
+      "loss": 3.8711,
+      "step": 416000
+    },
+    {
+      "epoch": 17.556801416347007,
+      "grad_norm": 1.4294953346252441,
+      "learning_rate": 3.028341122497106e-05,
+      "loss": 3.8683,
+      "step": 416500
+    },
+    {
+      "epoch": 17.577878008683555,
+      "grad_norm": 1.459107756614685,
+      "learning_rate": 3.019765896325516e-05,
+      "loss": 3.8769,
+      "step": 417000
+    },
+    {
+      "epoch": 17.598954601020107,
+      "grad_norm": 1.4384026527404785,
+      "learning_rate": 3.0111906701539255e-05,
+      "loss": 3.8724,
+      "step": 417500
+    },
+    {
+      "epoch": 17.62003119335666,
+      "grad_norm": 1.4187747240066528,
+      "learning_rate": 3.002615443982335e-05,
+      "loss": 3.8703,
+      "step": 418000
+    },
+    {
+      "epoch": 17.641107785693208,
+      "grad_norm": 1.4249036312103271,
+      "learning_rate": 2.9940402178107452e-05,
+      "loss": 3.8689,
+      "step": 418500
+    },
+    {
+      "epoch": 17.66218437802976,
+      "grad_norm": 1.4037829637527466,
+      "learning_rate": 2.9854649916391547e-05,
+      "loss": 3.8685,
+      "step": 419000
+    },
+    {
+      "epoch": 17.683260970366312,
+      "grad_norm": 1.4106301069259644,
+      "learning_rate": 2.976889765467564e-05,
+      "loss": 3.8743,
+      "step": 419500
+    },
+    {
+      "epoch": 17.70433756270286,
+      "grad_norm": 1.4151989221572876,
+      "learning_rate": 2.9683145392959737e-05,
+      "loss": 3.8712,
+      "step": 420000
+    },
+    {
+      "epoch": 17.70433756270286,
+      "eval_accuracy": 0.7220104313751191,
+      "eval_loss": 3.61545991897583,
+      "eval_runtime": 439.0096,
+      "eval_samples_per_second": 274.951,
+      "eval_steps_per_second": 2.866,
+      "step": 420000
+    },
+    {
+      "epoch": 17.725414155039413,
+      "grad_norm": 1.4545838832855225,
+      "learning_rate": 2.9597393131243838e-05,
+      "loss": 3.8711,
+      "step": 420500
+    },
+    {
+      "epoch": 17.746490747375965,
+      "grad_norm": 1.410272479057312,
+      "learning_rate": 2.9511640869527933e-05,
+      "loss": 3.8672,
+      "step": 421000
+    },
+    {
+      "epoch": 17.767567339712514,
+      "grad_norm": 1.4058136940002441,
+      "learning_rate": 2.942588860781203e-05,
+      "loss": 3.8682,
+      "step": 421500
+    },
+    {
+      "epoch": 17.788643932049066,
+      "grad_norm": 1.4541923999786377,
+      "learning_rate": 2.934013634609613e-05,
+      "loss": 3.8708,
+      "step": 422000
+    },
+    {
+      "epoch": 17.809720524385618,
+      "grad_norm": 1.3767180442810059,
+      "learning_rate": 2.9254384084380225e-05,
+      "loss": 3.8719,
+      "step": 422500
+    },
+    {
+      "epoch": 17.830797116722167,
+      "grad_norm": 1.4262168407440186,
+      "learning_rate": 2.9168631822664323e-05,
+      "loss": 3.8696,
+      "step": 423000
+    },
+    {
+      "epoch": 17.85187370905872,
+      "grad_norm": 1.3709843158721924,
+      "learning_rate": 2.908287956094842e-05,
+      "loss": 3.8706,
+      "step": 423500
+    },
+    {
+      "epoch": 17.87295030139527,
+      "grad_norm": 1.3741497993469238,
+      "learning_rate": 2.899712729923252e-05,
+      "loss": 3.8704,
+      "step": 424000
+    },
+    {
+      "epoch": 17.894026893731823,
+      "grad_norm": 1.383151650428772,
+      "learning_rate": 2.8911375037516615e-05,
+      "loss": 3.8689,
+      "step": 424500
+    },
+    {
+      "epoch": 17.91510348606837,
+      "grad_norm": 1.399422287940979,
+      "learning_rate": 2.8825622775800716e-05,
+      "loss": 3.8719,
+      "step": 425000
+    },
+    {
+      "epoch": 17.936180078404924,
+      "grad_norm": 1.4091109037399292,
+      "learning_rate": 2.873987051408481e-05,
+      "loss": 3.8687,
+      "step": 425500
+    },
+    {
+      "epoch": 17.957256670741476,
+      "grad_norm": 1.4434553384780884,
+      "learning_rate": 2.8654118252368906e-05,
+      "loss": 3.8666,
+      "step": 426000
+    },
+    {
+      "epoch": 17.978333263078024,
+      "grad_norm": 1.4144442081451416,
+      "learning_rate": 2.8568365990653008e-05,
+      "loss": 3.866,
+      "step": 426500
+    },
+    {
+      "epoch": 17.999409855414576,
+      "grad_norm": 1.4095628261566162,
+      "learning_rate": 2.8482613728937103e-05,
+      "loss": 3.8716,
+      "step": 427000
+    },
+    {
+      "epoch": 18.02048644775113,
+      "grad_norm": 1.398498296737671,
+      "learning_rate": 2.8396861467221198e-05,
+      "loss": 3.8655,
+      "step": 427500
+    },
+    {
+      "epoch": 18.041563040087677,
+      "grad_norm": 1.4346047639846802,
+      "learning_rate": 2.83111092055053e-05,
+      "loss": 3.8653,
+      "step": 428000
+    },
+    {
+      "epoch": 18.06263963242423,
+      "grad_norm": 1.3989639282226562,
+      "learning_rate": 2.8225356943789394e-05,
+      "loss": 3.866,
+      "step": 428500
+    },
+    {
+      "epoch": 18.08371622476078,
+      "grad_norm": 1.4516538381576538,
+      "learning_rate": 2.813960468207349e-05,
+      "loss": 3.8635,
+      "step": 429000
+    },
+    {
+      "epoch": 18.10479281709733,
+      "grad_norm": 1.410131812095642,
+      "learning_rate": 2.805385242035759e-05,
+      "loss": 3.8634,
+      "step": 429500
+    },
+    {
+      "epoch": 18.125869409433882,
+      "grad_norm": 1.4575915336608887,
+      "learning_rate": 2.7968100158641686e-05,
+      "loss": 3.8638,
+      "step": 430000
+    },
+    {
+      "epoch": 18.125869409433882,
+      "eval_accuracy": 0.7225158881005217,
+      "eval_loss": 3.6134326457977295,
+      "eval_runtime": 440.0803,
+      "eval_samples_per_second": 274.282,
+      "eval_steps_per_second": 2.859,
+      "step": 430000
+    },
+    {
+      "epoch": 18.146946001770434,
+      "grad_norm": 1.401827335357666,
+      "learning_rate": 2.788234789692578e-05,
+      "loss": 3.8645,
+      "step": 430500
+    },
+    {
+      "epoch": 18.168022594106986,
+      "grad_norm": 1.426401972770691,
+      "learning_rate": 2.7796595635209882e-05,
+      "loss": 3.865,
+      "step": 431000
+    },
+    {
+      "epoch": 18.189099186443535,
+      "grad_norm": 1.43930184841156,
+      "learning_rate": 2.7710843373493977e-05,
+      "loss": 3.8662,
+      "step": 431500
+    },
+    {
+      "epoch": 18.210175778780087,
+      "grad_norm": 1.4230502843856812,
+      "learning_rate": 2.7625091111778072e-05,
+      "loss": 3.8671,
+      "step": 432000
+    },
+    {
+      "epoch": 18.23125237111664,
+      "grad_norm": 1.4341344833374023,
+      "learning_rate": 2.7539338850062174e-05,
+      "loss": 3.8647,
+      "step": 432500
+    },
+    {
+      "epoch": 18.252328963453188,
+      "grad_norm": 1.3954546451568604,
+      "learning_rate": 2.745358658834627e-05,
+      "loss": 3.8644,
+      "step": 433000
+    },
+    {
+      "epoch": 18.27340555578974,
+      "grad_norm": 1.4276723861694336,
+      "learning_rate": 2.7367834326630364e-05,
+      "loss": 3.8618,
+      "step": 433500
+    },
+    {
+      "epoch": 18.294482148126292,
+      "grad_norm": 1.433361530303955,
+      "learning_rate": 2.7282082064914465e-05,
+      "loss": 3.8625,
+      "step": 434000
+    },
+    {
+      "epoch": 18.31555874046284,
+      "grad_norm": 1.4398785829544067,
+      "learning_rate": 2.719632980319856e-05,
+      "loss": 3.8618,
+      "step": 434500
+    },
+    {
+      "epoch": 18.336635332799393,
+      "grad_norm": 1.415863037109375,
+      "learning_rate": 2.7110577541482655e-05,
+      "loss": 3.8612,
+      "step": 435000
+    },
+    {
+      "epoch": 18.357711925135945,
+      "grad_norm": 1.4131532907485962,
+      "learning_rate": 2.7024825279766757e-05,
+      "loss": 3.8632,
+      "step": 435500
+    },
+    {
+      "epoch": 18.378788517472493,
+      "grad_norm": 1.458092212677002,
+      "learning_rate": 2.6939073018050852e-05,
+      "loss": 3.8651,
+      "step": 436000
+    },
+    {
+      "epoch": 18.399865109809046,
+      "grad_norm": 1.378008246421814,
+      "learning_rate": 2.685332075633495e-05,
+      "loss": 3.8632,
+      "step": 436500
+    },
+    {
+      "epoch": 18.420941702145598,
+      "grad_norm": 1.3960801362991333,
+      "learning_rate": 2.6767568494619045e-05,
+      "loss": 3.8624,
+      "step": 437000
+    },
+    {
+      "epoch": 18.44201829448215,
+      "grad_norm": 1.4601411819458008,
+      "learning_rate": 2.6681816232903143e-05,
+      "loss": 3.8624,
+      "step": 437500
+    },
+    {
+      "epoch": 18.4630948868187,
+      "grad_norm": 1.4669278860092163,
+      "learning_rate": 2.6596063971187242e-05,
+      "loss": 3.8591,
+      "step": 438000
+    },
+    {
+      "epoch": 18.48417147915525,
+      "grad_norm": 1.4654794931411743,
+      "learning_rate": 2.6510311709471337e-05,
+      "loss": 3.8609,
+      "step": 438500
+    },
+    {
+      "epoch": 18.505248071491803,
+      "grad_norm": 1.378740906715393,
+      "learning_rate": 2.642455944775544e-05,
+      "loss": 3.8625,
+      "step": 439000
+    },
+    {
+      "epoch": 18.52632466382835,
+      "grad_norm": 1.416479468345642,
+      "learning_rate": 2.6338807186039533e-05,
+      "loss": 3.8592,
+      "step": 439500
+    },
+    {
+      "epoch": 18.547401256164903,
+      "grad_norm": 1.4315968751907349,
+      "learning_rate": 2.6253054924323628e-05,
+      "loss": 3.8588,
+      "step": 440000
+    },
+    {
+      "epoch": 18.547401256164903,
+      "eval_accuracy": 0.7228014089919648,
+      "eval_loss": 3.610879898071289,
+      "eval_runtime": 437.2081,
+      "eval_samples_per_second": 276.084,
+      "eval_steps_per_second": 2.877,
+      "step": 440000
+    },
+    {
+      "epoch": 18.568477848501455,
+      "grad_norm": 1.3782013654708862,
+      "learning_rate": 2.616730266260773e-05,
+      "loss": 3.859,
+      "step": 440500
+    },
+    {
+      "epoch": 18.589554440838004,
+      "grad_norm": 1.4390465021133423,
+      "learning_rate": 2.6081550400891825e-05,
+      "loss": 3.8621,
+      "step": 441000
+    },
+    {
+      "epoch": 18.610631033174556,
+      "grad_norm": 1.402696132659912,
+      "learning_rate": 2.599579813917592e-05,
+      "loss": 3.8617,
+      "step": 441500
+    },
+    {
+      "epoch": 18.63170762551111,
+      "grad_norm": 1.3902013301849365,
+      "learning_rate": 2.591004587746002e-05,
+      "loss": 3.8612,
+      "step": 442000
+    },
+    {
+      "epoch": 18.652784217847657,
+      "grad_norm": 1.489107608795166,
+      "learning_rate": 2.5824293615744116e-05,
+      "loss": 3.8606,
+      "step": 442500
+    },
+    {
+      "epoch": 18.67386081018421,
+      "grad_norm": 1.474097728729248,
+      "learning_rate": 2.573854135402821e-05,
+      "loss": 3.8628,
+      "step": 443000
+    },
+    {
+      "epoch": 18.69493740252076,
+      "grad_norm": 1.404527187347412,
+      "learning_rate": 2.5652789092312313e-05,
+      "loss": 3.8596,
+      "step": 443500
+    },
+    {
+      "epoch": 18.716013994857313,
+      "grad_norm": 1.424486756324768,
+      "learning_rate": 2.5567036830596408e-05,
+      "loss": 3.8593,
+      "step": 444000
+    },
+    {
+      "epoch": 18.737090587193862,
+      "grad_norm": 1.4863967895507812,
+      "learning_rate": 2.5481284568880503e-05,
+      "loss": 3.8609,
+      "step": 444500
+    },
+    {
+      "epoch": 18.758167179530414,
+      "grad_norm": 1.437961220741272,
+      "learning_rate": 2.5395532307164605e-05,
+      "loss": 3.8551,
+      "step": 445000
+    },
+    {
+      "epoch": 18.779243771866966,
+      "grad_norm": 1.5082110166549683,
+      "learning_rate": 2.53097800454487e-05,
+      "loss": 3.8641,
+      "step": 445500
+    },
+    {
+      "epoch": 18.800320364203515,
+      "grad_norm": 1.3742308616638184,
+      "learning_rate": 2.5224027783732794e-05,
+      "loss": 3.8621,
+      "step": 446000
+    },
+    {
+      "epoch": 18.821396956540067,
+      "grad_norm": 1.4945086240768433,
+      "learning_rate": 2.5138275522016896e-05,
+      "loss": 3.8539,
+      "step": 446500
+    },
+    {
+      "epoch": 18.84247354887662,
+      "grad_norm": 1.4091224670410156,
+      "learning_rate": 2.505252326030099e-05,
+      "loss": 3.8577,
+      "step": 447000
+    },
+    {
+      "epoch": 18.863550141213167,
+      "grad_norm": 1.435324788093567,
+      "learning_rate": 2.496677099858509e-05,
+      "loss": 3.8576,
+      "step": 447500
+    },
+    {
+      "epoch": 18.88462673354972,
+      "grad_norm": 1.3894648551940918,
+      "learning_rate": 2.4881018736869184e-05,
+      "loss": 3.8616,
+      "step": 448000
+    },
+    {
+      "epoch": 18.90570332588627,
+      "grad_norm": 1.462371826171875,
+      "learning_rate": 2.4795266475153283e-05,
+      "loss": 3.8565,
+      "step": 448500
+    },
+    {
+      "epoch": 18.92677991822282,
+      "grad_norm": 1.4791711568832397,
+      "learning_rate": 2.470951421343738e-05,
+      "loss": 3.8635,
+      "step": 449000
+    },
+    {
+      "epoch": 18.947856510559372,
+      "grad_norm": 1.3700664043426514,
+      "learning_rate": 2.4623761951721476e-05,
+      "loss": 3.856,
+      "step": 449500
+    },
+    {
+      "epoch": 18.968933102895924,
+      "grad_norm": 1.4212837219238281,
+      "learning_rate": 2.4538009690005574e-05,
+      "loss": 3.8586,
+      "step": 450000
+    },
+    {
+      "epoch": 18.968933102895924,
+      "eval_accuracy": 0.7236755234892206,
+      "eval_loss": 3.605743169784546,
+      "eval_runtime": 505.2787,
+      "eval_samples_per_second": 238.89,
+      "eval_steps_per_second": 2.49,
+      "step": 450000
+    },
+    {
+      "epoch": 18.990009695232473,
+      "grad_norm": 1.4671857357025146,
+      "learning_rate": 2.4452257428289672e-05,
+      "loss": 3.8541,
+      "step": 450500
+    },
+    {
+      "epoch": 19.011086287569025,
+      "grad_norm": 1.45199716091156,
+      "learning_rate": 2.436650516657377e-05,
+      "loss": 3.8549,
+      "step": 451000
+    },
+    {
+      "epoch": 19.032162879905577,
+      "grad_norm": 1.4702253341674805,
+      "learning_rate": 2.428075290485787e-05,
+      "loss": 3.8542,
+      "step": 451500
+    },
+    {
+      "epoch": 19.05323947224213,
+      "grad_norm": 1.4331493377685547,
+      "learning_rate": 2.4195000643141964e-05,
+      "loss": 3.8567,
+      "step": 452000
+    },
+    {
+      "epoch": 19.074316064578678,
+      "grad_norm": 1.4340629577636719,
+      "learning_rate": 2.4109248381426062e-05,
+      "loss": 3.8572,
+      "step": 452500
+    },
+    {
+      "epoch": 19.09539265691523,
+      "grad_norm": 1.3608511686325073,
+      "learning_rate": 2.402349611971016e-05,
+      "loss": 3.8569,
+      "step": 453000
+    },
+    {
+      "epoch": 19.116469249251782,
+      "grad_norm": 1.3855551481246948,
+      "learning_rate": 2.3937743857994255e-05,
+      "loss": 3.8536,
+      "step": 453500
+    },
+    {
+      "epoch": 19.13754584158833,
+      "grad_norm": 1.4599756002426147,
+      "learning_rate": 2.3851991596278354e-05,
+      "loss": 3.8569,
+      "step": 454000
+    },
+    {
+      "epoch": 19.158622433924883,
+      "grad_norm": 1.4378066062927246,
+      "learning_rate": 2.3766239334562452e-05,
+      "loss": 3.8537,
+      "step": 454500
+    },
+    {
+      "epoch": 19.179699026261435,
+      "grad_norm": 1.4464505910873413,
+      "learning_rate": 2.3680487072846547e-05,
+      "loss": 3.8503,
+      "step": 455000
+    },
+    {
+      "epoch": 19.200775618597984,
+      "grad_norm": 1.4673713445663452,
+      "learning_rate": 2.3594734811130645e-05,
+      "loss": 3.8509,
+      "step": 455500
+    },
+    {
+      "epoch": 19.221852210934536,
+      "grad_norm": 1.4127763509750366,
+      "learning_rate": 2.3508982549414744e-05,
+      "loss": 3.852,
+      "step": 456000
+    },
+    {
+      "epoch": 19.242928803271088,
+      "grad_norm": 1.4590644836425781,
+      "learning_rate": 2.342323028769884e-05,
+      "loss": 3.8539,
+      "step": 456500
+    },
+    {
+      "epoch": 19.264005395607636,
+      "grad_norm": 1.4234756231307983,
+      "learning_rate": 2.3337478025982937e-05,
+      "loss": 3.8504,
+      "step": 457000
+    },
+    {
+      "epoch": 19.28508198794419,
+      "grad_norm": 1.484137773513794,
+      "learning_rate": 2.3251725764267035e-05,
+      "loss": 3.8527,
+      "step": 457500
+    },
+    {
+      "epoch": 19.30615858028074,
+      "grad_norm": 1.468321681022644,
+      "learning_rate": 2.316597350255113e-05,
+      "loss": 3.8576,
+      "step": 458000
+    },
+    {
+      "epoch": 19.327235172617293,
+      "grad_norm": 1.4544819593429565,
+      "learning_rate": 2.308022124083523e-05,
+      "loss": 3.854,
+      "step": 458500
+    },
+    {
+      "epoch": 19.34831176495384,
+      "grad_norm": 1.4457937479019165,
+      "learning_rate": 2.2994468979119323e-05,
+      "loss": 3.8546,
+      "step": 459000
+    },
+    {
+      "epoch": 19.369388357290394,
+      "grad_norm": 1.4557251930236816,
+      "learning_rate": 2.290871671740342e-05,
+      "loss": 3.8534,
+      "step": 459500
+    },
+    {
+      "epoch": 19.390464949626946,
+      "grad_norm": 1.4009417295455933,
+      "learning_rate": 2.282296445568752e-05,
+      "loss": 3.8534,
+      "step": 460000
+    },
+    {
+      "epoch": 19.390464949626946,
+      "eval_accuracy": 0.7240966661472035,
+      "eval_loss": 3.601783037185669,
+      "eval_runtime": 361.006,
+      "eval_samples_per_second": 334.36,
+      "eval_steps_per_second": 3.485,
+      "step": 460000
+    },
+    {
+      "epoch": 19.411541541963494,
+      "grad_norm": 1.4687793254852295,
+      "learning_rate": 2.2737212193971615e-05,
+      "loss": 3.8509,
+      "step": 460500
+    },
+    {
+      "epoch": 19.432618134300046,
+      "grad_norm": 1.4314830303192139,
+      "learning_rate": 2.2651459932255713e-05,
+      "loss": 3.8562,
+      "step": 461000
+    },
+    {
+      "epoch": 19.4536947266366,
+      "grad_norm": 1.435370683670044,
+      "learning_rate": 2.256570767053981e-05,
+      "loss": 3.8515,
+      "step": 461500
+    },
+    {
+      "epoch": 19.474771318973147,
+      "grad_norm": 1.4463789463043213,
+      "learning_rate": 2.2479955408823906e-05,
+      "loss": 3.8508,
+      "step": 462000
+    },
+    {
+      "epoch": 19.4958479113097,
+      "grad_norm": 1.4501289129257202,
+      "learning_rate": 2.2394203147108005e-05,
+      "loss": 3.8522,
+      "step": 462500
+    },
+    {
+      "epoch": 19.51692450364625,
+      "grad_norm": 1.4594707489013672,
+      "learning_rate": 2.2308450885392103e-05,
+      "loss": 3.8495,
+      "step": 463000
+    },
+    {
+      "epoch": 19.5380010959828,
+      "grad_norm": 1.4631088972091675,
+      "learning_rate": 2.22226986236762e-05,
+      "loss": 3.8499,
+      "step": 463500
+    },
+    {
+      "epoch": 19.559077688319352,
+      "grad_norm": 1.4246662855148315,
+      "learning_rate": 2.21369463619603e-05,
+      "loss": 3.8519,
+      "step": 464000
+    },
+    {
+      "epoch": 19.580154280655904,
+      "grad_norm": 1.4750233888626099,
+      "learning_rate": 2.2051194100244394e-05,
+      "loss": 3.851,
+      "step": 464500
+    },
+    {
+      "epoch": 19.601230872992456,
+      "grad_norm": 1.4828217029571533,
+      "learning_rate": 2.1965441838528493e-05,
+      "loss": 3.8519,
+      "step": 465000
+    },
+    {
+      "epoch": 19.622307465329005,
+      "grad_norm": 1.432034969329834,
+      "learning_rate": 2.187968957681259e-05,
+      "loss": 3.85,
+      "step": 465500
+    },
+    {
+      "epoch": 19.643384057665557,
+      "grad_norm": 1.406863808631897,
+      "learning_rate": 2.179393731509669e-05,
+      "loss": 3.852,
+      "step": 466000
+    },
+    {
+      "epoch": 19.66446065000211,
+      "grad_norm": 1.4497827291488647,
+      "learning_rate": 2.1708185053380784e-05,
+      "loss": 3.8504,
+      "step": 466500
+    },
+    {
+      "epoch": 19.685537242338658,
+      "grad_norm": 1.4448015689849854,
+      "learning_rate": 2.1622432791664883e-05,
+      "loss": 3.8532,
+      "step": 467000
+    },
+    {
+      "epoch": 19.70661383467521,
+      "grad_norm": 1.4377310276031494,
+      "learning_rate": 2.1536680529948978e-05,
+      "loss": 3.8498,
+      "step": 467500
+    },
+    {
+      "epoch": 19.727690427011762,
+      "grad_norm": 1.4505053758621216,
+      "learning_rate": 2.1450928268233076e-05,
+      "loss": 3.8522,
+      "step": 468000
+    },
+    {
+      "epoch": 19.74876701934831,
+      "grad_norm": 1.3930619955062866,
+      "learning_rate": 2.1365176006517174e-05,
+      "loss": 3.8499,
+      "step": 468500
+    },
+    {
+      "epoch": 19.769843611684863,
+      "grad_norm": 1.4799753427505493,
+      "learning_rate": 2.127942374480127e-05,
+      "loss": 3.8503,
+      "step": 469000
+    },
+    {
+      "epoch": 19.790920204021415,
+      "grad_norm": 1.4870669841766357,
+      "learning_rate": 2.1193671483085367e-05,
+      "loss": 3.8541,
+      "step": 469500
+    },
+    {
+      "epoch": 19.811996796357963,
+      "grad_norm": 1.4387937784194946,
+      "learning_rate": 2.1107919221369466e-05,
+      "loss": 3.8508,
+      "step": 470000
+    },
+    {
+      "epoch": 19.811996796357963,
+      "eval_accuracy": 0.7246566808683377,
+      "eval_loss": 3.5987603664398193,
+      "eval_runtime": 413.3331,
+      "eval_samples_per_second": 292.031,
+      "eval_steps_per_second": 3.044,
+      "step": 470000
+    },
+    {
+      "epoch": 19.833073388694515,
+      "grad_norm": 1.4134844541549683,
+      "learning_rate": 2.102216695965356e-05,
+      "loss": 3.8524,
+      "step": 470500
+    },
+    {
+      "epoch": 19.854149981031068,
+      "grad_norm": 1.3924450874328613,
+      "learning_rate": 2.093641469793766e-05,
+      "loss": 3.8497,
+      "step": 471000
+    },
+    {
+      "epoch": 19.87522657336762,
+      "grad_norm": 1.431532859802246,
+      "learning_rate": 2.0850662436221757e-05,
+      "loss": 3.8496,
+      "step": 471500
+    },
+    {
+      "epoch": 19.89630316570417,
+      "grad_norm": 1.4352811574935913,
+      "learning_rate": 2.0764910174505852e-05,
+      "loss": 3.8486,
+      "step": 472000
+    },
+    {
+      "epoch": 19.91737975804072,
+      "grad_norm": 1.498489260673523,
+      "learning_rate": 2.067915791278995e-05,
+      "loss": 3.8523,
+      "step": 472500
+    },
+    {
+      "epoch": 19.938456350377272,
+      "grad_norm": 1.5420033931732178,
+      "learning_rate": 2.059340565107405e-05,
+      "loss": 3.8443,
+      "step": 473000
+    },
+    {
+      "epoch": 19.95953294271382,
+      "grad_norm": 1.3824357986450195,
+      "learning_rate": 2.0507653389358144e-05,
+      "loss": 3.8493,
+      "step": 473500
+    },
+    {
+      "epoch": 19.980609535050373,
+      "grad_norm": 1.4685912132263184,
+      "learning_rate": 2.0421901127642242e-05,
+      "loss": 3.8466,
+      "step": 474000
+    },
+    {
+      "epoch": 20.001686127386925,
+      "grad_norm": 1.4587396383285522,
+      "learning_rate": 2.033614886592634e-05,
+      "loss": 3.8538,
+      "step": 474500
+    },
+    {
+      "epoch": 20.022762719723474,
+      "grad_norm": 1.4618656635284424,
+      "learning_rate": 2.0250396604210435e-05,
+      "loss": 3.8483,
+      "step": 475000
+    },
+    {
+      "epoch": 20.043839312060026,
+      "grad_norm": 1.531733512878418,
+      "learning_rate": 2.0164644342494534e-05,
+      "loss": 3.8482,
+      "step": 475500
+    },
+    {
+      "epoch": 20.064915904396578,
+      "grad_norm": 1.4332327842712402,
+      "learning_rate": 2.007889208077863e-05,
+      "loss": 3.8424,
+      "step": 476000
+    },
+    {
+      "epoch": 20.085992496733127,
+      "grad_norm": 1.4737993478775024,
+      "learning_rate": 1.9993139819062727e-05,
+      "loss": 3.846,
+      "step": 476500
+    },
+    {
+      "epoch": 20.10706908906968,
+      "grad_norm": 1.4912726879119873,
+      "learning_rate": 1.9907387557346825e-05,
+      "loss": 3.8481,
+      "step": 477000
+    },
+    {
+      "epoch": 20.12814568140623,
+      "grad_norm": 1.4285738468170166,
+      "learning_rate": 1.9821635295630923e-05,
+      "loss": 3.8463,
+      "step": 477500
+    },
+    {
+      "epoch": 20.14922227374278,
+      "grad_norm": 1.4813228845596313,
+      "learning_rate": 1.9735883033915022e-05,
+      "loss": 3.844,
+      "step": 478000
+    },
+    {
+      "epoch": 20.17029886607933,
+      "grad_norm": 1.4908586740493774,
+      "learning_rate": 1.965013077219912e-05,
+      "loss": 3.8489,
+      "step": 478500
+    },
+    {
+      "epoch": 20.191375458415884,
+      "grad_norm": 1.4351298809051514,
+      "learning_rate": 1.9564378510483215e-05,
+      "loss": 3.8427,
+      "step": 479000
+    },
+    {
+      "epoch": 20.212452050752436,
+      "grad_norm": 1.4278494119644165,
+      "learning_rate": 1.9478626248767313e-05,
+      "loss": 3.8477,
+      "step": 479500
+    },
+    {
+      "epoch": 20.233528643088984,
+      "grad_norm": 1.4952630996704102,
+      "learning_rate": 1.939287398705141e-05,
+      "loss": 3.8443,
+      "step": 480000
+    },
+    {
+      "epoch": 20.233528643088984,
+      "eval_accuracy": 0.7251729101393275,
+      "eval_loss": 3.5953001976013184,
+      "eval_runtime": 389.9341,
+      "eval_samples_per_second": 309.555,
+      "eval_steps_per_second": 3.226,
+      "step": 480000
+    },
+    {
+      "epoch": 20.254605235425537,
+      "grad_norm": 1.4034298658370972,
+      "learning_rate": 1.9307121725335506e-05,
+      "loss": 3.8415,
+      "step": 480500
+    },
+    {
+      "epoch": 20.27568182776209,
+      "grad_norm": 1.4524632692337036,
+      "learning_rate": 1.9221369463619605e-05,
+      "loss": 3.8422,
+      "step": 481000
+    },
+    {
+      "epoch": 20.296758420098637,
+      "grad_norm": 1.4348094463348389,
+      "learning_rate": 1.9135617201903703e-05,
+      "loss": 3.8424,
+      "step": 481500
+    },
+    {
+      "epoch": 20.31783501243519,
+      "grad_norm": 1.4646574258804321,
+      "learning_rate": 1.9049864940187798e-05,
+      "loss": 3.843,
+      "step": 482000
+    },
+    {
+      "epoch": 20.33891160477174,
+      "grad_norm": 1.4683212041854858,
+      "learning_rate": 1.8964112678471896e-05,
+      "loss": 3.8445,
+      "step": 482500
+    },
+    {
+      "epoch": 20.35998819710829,
+      "grad_norm": 1.422366738319397,
+      "learning_rate": 1.8878360416755995e-05,
+      "loss": 3.8451,
+      "step": 483000
+    },
+    {
+      "epoch": 20.381064789444842,
+      "grad_norm": 1.43455970287323,
+      "learning_rate": 1.879260815504009e-05,
+      "loss": 3.8438,
+      "step": 483500
+    },
+    {
+      "epoch": 20.402141381781394,
+      "grad_norm": 1.426229476928711,
+      "learning_rate": 1.8706855893324188e-05,
+      "loss": 3.8403,
+      "step": 484000
+    },
+    {
+      "epoch": 20.423217974117943,
+      "grad_norm": 1.4760392904281616,
+      "learning_rate": 1.8621103631608283e-05,
+      "loss": 3.8433,
+      "step": 484500
+    },
+    {
+      "epoch": 20.444294566454495,
+      "grad_norm": 1.4479515552520752,
+      "learning_rate": 1.853535136989238e-05,
+      "loss": 3.8443,
+      "step": 485000
+    },
+    {
+      "epoch": 20.465371158791047,
+      "grad_norm": 1.4691495895385742,
+      "learning_rate": 1.844959910817648e-05,
+      "loss": 3.8421,
+      "step": 485500
+    },
+    {
+      "epoch": 20.4864477511276,
+      "grad_norm": 1.4617035388946533,
+      "learning_rate": 1.8363846846460574e-05,
+      "loss": 3.8446,
+      "step": 486000
+    },
+    {
+      "epoch": 20.507524343464148,
+      "grad_norm": 1.50462007522583,
+      "learning_rate": 1.8278094584744673e-05,
+      "loss": 3.8396,
+      "step": 486500
+    },
+    {
+      "epoch": 20.5286009358007,
+      "grad_norm": 1.4610159397125244,
+      "learning_rate": 1.819234232302877e-05,
+      "loss": 3.8448,
+      "step": 487000
+    },
+    {
+      "epoch": 20.549677528137252,
+      "grad_norm": 1.4093517065048218,
+      "learning_rate": 1.8106590061312866e-05,
+      "loss": 3.8437,
+      "step": 487500
+    },
+    {
+      "epoch": 20.5707541204738,
+      "grad_norm": 1.4394655227661133,
+      "learning_rate": 1.8020837799596964e-05,
+      "loss": 3.8453,
+      "step": 488000
+    },
+    {
+      "epoch": 20.591830712810353,
+      "grad_norm": 1.4095388650894165,
+      "learning_rate": 1.7935085537881062e-05,
+      "loss": 3.8459,
+      "step": 488500
+    },
+    {
+      "epoch": 20.612907305146905,
+      "grad_norm": 1.4793789386749268,
+      "learning_rate": 1.7849333276165157e-05,
+      "loss": 3.8416,
+      "step": 489000
+    },
+    {
+      "epoch": 20.633983897483454,
+      "grad_norm": 1.4835070371627808,
+      "learning_rate": 1.7763581014449256e-05,
+      "loss": 3.8412,
+      "step": 489500
+    },
+    {
+      "epoch": 20.655060489820006,
+      "grad_norm": 1.4489439725875854,
+      "learning_rate": 1.7677828752733354e-05,
+      "loss": 3.8393,
+      "step": 490000
+    },
+    {
+      "epoch": 20.655060489820006,
+      "eval_accuracy": 0.7257299424268523,
+      "eval_loss": 3.5920610427856445,
+      "eval_runtime": 376.942,
+      "eval_samples_per_second": 320.224,
+      "eval_steps_per_second": 3.337,
+      "step": 490000
+    },
+    {
+      "epoch": 20.676137082156558,
+      "grad_norm": 1.4870858192443848,
+      "learning_rate": 1.7592076491017452e-05,
+      "loss": 3.8431,
+      "step": 490500
+    },
+    {
+      "epoch": 20.697213674493106,
+      "grad_norm": 1.45559561252594,
+      "learning_rate": 1.7506324229301547e-05,
+      "loss": 3.8444,
+      "step": 491000
+    },
+    {
+      "epoch": 20.71829026682966,
+      "grad_norm": 1.4609463214874268,
+      "learning_rate": 1.7420571967585646e-05,
+      "loss": 3.8443,
+      "step": 491500
+    },
+    {
+      "epoch": 20.73936685916621,
+      "grad_norm": 1.5050770044326782,
+      "learning_rate": 1.7334819705869744e-05,
+      "loss": 3.8415,
+      "step": 492000
+    },
+    {
+      "epoch": 20.760443451502763,
+      "grad_norm": 1.422317385673523,
+      "learning_rate": 1.7249067444153842e-05,
+      "loss": 3.8447,
+      "step": 492500
+    },
+    {
+      "epoch": 20.78152004383931,
+      "grad_norm": 1.4750112295150757,
+      "learning_rate": 1.7163315182437937e-05,
+      "loss": 3.8385,
+      "step": 493000
+    },
+    {
+      "epoch": 20.802596636175863,
+      "grad_norm": 1.4566717147827148,
+      "learning_rate": 1.7077562920722035e-05,
+      "loss": 3.8411,
+      "step": 493500
+    },
+    {
+      "epoch": 20.823673228512416,
+      "grad_norm": 1.4421502351760864,
+      "learning_rate": 1.6991810659006134e-05,
+      "loss": 3.8423,
+      "step": 494000
+    },
+    {
+      "epoch": 20.844749820848964,
+      "grad_norm": 1.4118074178695679,
+      "learning_rate": 1.690605839729023e-05,
+      "loss": 3.8409,
+      "step": 494500
+    },
+    {
+      "epoch": 20.865826413185516,
+      "grad_norm": 1.4179068803787231,
+      "learning_rate": 1.6820306135574327e-05,
+      "loss": 3.8408,
+      "step": 495000
+    },
+    {
+      "epoch": 20.88690300552207,
+      "grad_norm": 1.4959323406219482,
+      "learning_rate": 1.6734553873858425e-05,
+      "loss": 3.841,
+      "step": 495500
+    },
+    {
+      "epoch": 20.907979597858617,
+      "grad_norm": 1.4930814504623413,
+      "learning_rate": 1.664880161214252e-05,
+      "loss": 3.8406,
+      "step": 496000
+    },
+    {
+      "epoch": 20.92905619019517,
+      "grad_norm": 1.4512773752212524,
+      "learning_rate": 1.656304935042662e-05,
+      "loss": 3.8436,
+      "step": 496500
+    },
+    {
+      "epoch": 20.95013278253172,
+      "grad_norm": 1.446810007095337,
+      "learning_rate": 1.6477297088710717e-05,
+      "loss": 3.8393,
+      "step": 497000
+    },
+    {
+      "epoch": 20.97120937486827,
+      "grad_norm": 1.4742056131362915,
+      "learning_rate": 1.639154482699481e-05,
+      "loss": 3.8389,
+      "step": 497500
+    },
+    {
+      "epoch": 20.992285967204822,
+      "grad_norm": 1.4186142683029175,
+      "learning_rate": 1.630579256527891e-05,
+      "loss": 3.8407,
+      "step": 498000
+    },
+    {
+      "epoch": 21.013362559541374,
+      "grad_norm": 1.4623297452926636,
+      "learning_rate": 1.6220040303563008e-05,
+      "loss": 3.8398,
+      "step": 498500
+    },
+    {
+      "epoch": 21.034439151877926,
+      "grad_norm": 1.4685580730438232,
+      "learning_rate": 1.6134288041847103e-05,
+      "loss": 3.8405,
+      "step": 499000
+    },
+    {
+      "epoch": 21.055515744214475,
+      "grad_norm": 1.4637525081634521,
+      "learning_rate": 1.60485357801312e-05,
+      "loss": 3.8361,
+      "step": 499500
+    },
+    {
+      "epoch": 21.076592336551027,
+      "grad_norm": 1.4745171070098877,
+      "learning_rate": 1.59627835184153e-05,
+      "loss": 3.8366,
+      "step": 500000
+    },
+    {
+      "epoch": 21.076592336551027,
+      "eval_accuracy": 0.7259951569195014,
+      "eval_loss": 3.5885629653930664,
+      "eval_runtime": 401.792,
+      "eval_samples_per_second": 300.419,
+      "eval_steps_per_second": 3.131,
+      "step": 500000
+    },
+    {
+      "epoch": 21.09766892888758,
+      "grad_norm": 1.4456353187561035,
+      "learning_rate": 1.5877031256699395e-05,
+      "loss": 3.8366,
+      "step": 500500
+    },
+    {
+      "epoch": 21.118745521224128,
+      "grad_norm": 1.447122573852539,
+      "learning_rate": 1.5791278994983493e-05,
+      "loss": 3.8383,
+      "step": 501000
+    },
+    {
+      "epoch": 21.13982211356068,
+      "grad_norm": 1.4379470348358154,
+      "learning_rate": 1.570552673326759e-05,
+      "loss": 3.835,
+      "step": 501500
+    },
+    {
+      "epoch": 21.16089870589723,
+      "grad_norm": 1.439568281173706,
+      "learning_rate": 1.5619774471551686e-05,
+      "loss": 3.8389,
+      "step": 502000
+    },
+    {
+      "epoch": 21.18197529823378,
+      "grad_norm": 1.4287375211715698,
+      "learning_rate": 1.5534022209835785e-05,
+      "loss": 3.8385,
+      "step": 502500
+    },
+    {
+      "epoch": 21.203051890570332,
+      "grad_norm": 1.417061448097229,
+      "learning_rate": 1.544826994811988e-05,
+      "loss": 3.837,
+      "step": 503000
+    },
+    {
+      "epoch": 21.224128482906885,
+      "grad_norm": 1.430117130279541,
+      "learning_rate": 1.5362517686403978e-05,
+      "loss": 3.8371,
+      "step": 503500
+    },
+    {
+      "epoch": 21.245205075243433,
+      "grad_norm": 1.422243356704712,
+      "learning_rate": 1.5276765424688076e-05,
+      "loss": 3.8357,
+      "step": 504000
+    },
+    {
+      "epoch": 21.266281667579985,
+      "grad_norm": 1.4663926362991333,
+      "learning_rate": 1.5191013162972173e-05,
+      "loss": 3.835,
+      "step": 504500
+    },
+    {
+      "epoch": 21.287358259916537,
+      "grad_norm": 1.5152007341384888,
+      "learning_rate": 1.5105260901256271e-05,
+      "loss": 3.8367,
+      "step": 505000
+    },
+    {
+      "epoch": 21.308434852253086,
+      "grad_norm": 1.5031744241714478,
+      "learning_rate": 1.501950863954037e-05,
+      "loss": 3.8377,
+      "step": 505500
+    },
+    {
+      "epoch": 21.329511444589638,
+      "grad_norm": 1.466470718383789,
+      "learning_rate": 1.4933756377824464e-05,
+      "loss": 3.8328,
+      "step": 506000
+    },
+    {
+      "epoch": 21.35058803692619,
+      "grad_norm": 1.457750916481018,
+      "learning_rate": 1.4848004116108563e-05,
+      "loss": 3.8312,
+      "step": 506500
+    },
+    {
+      "epoch": 21.371664629262742,
+      "grad_norm": 1.4462709426879883,
+      "learning_rate": 1.4762251854392661e-05,
+      "loss": 3.8394,
+      "step": 507000
+    },
+    {
+      "epoch": 21.39274122159929,
+      "grad_norm": 1.4526947736740112,
+      "learning_rate": 1.4676499592676758e-05,
+      "loss": 3.838,
+      "step": 507500
+    },
+    {
+      "epoch": 21.413817813935843,
+      "grad_norm": 1.4195399284362793,
+      "learning_rate": 1.4590747330960856e-05,
+      "loss": 3.836,
+      "step": 508000
+    },
+    {
+      "epoch": 21.434894406272395,
+      "grad_norm": 1.4341137409210205,
+      "learning_rate": 1.4504995069244952e-05,
+      "loss": 3.8346,
+      "step": 508500
+    },
+    {
+      "epoch": 21.455970998608944,
+      "grad_norm": 1.392270565032959,
+      "learning_rate": 1.4419242807529049e-05,
+      "loss": 3.8373,
+      "step": 509000
+    },
+    {
+      "epoch": 21.477047590945496,
+      "grad_norm": 1.3599987030029297,
+      "learning_rate": 1.4333490545813147e-05,
+      "loss": 3.836,
+      "step": 509500
+    },
+    {
+      "epoch": 21.498124183282048,
+      "grad_norm": 1.4515693187713623,
+      "learning_rate": 1.4247738284097246e-05,
+      "loss": 3.8324,
+      "step": 510000
+    },
+    {
+      "epoch": 21.498124183282048,
+      "eval_accuracy": 0.7266227756044478,
+      "eval_loss": 3.585749864578247,
+      "eval_runtime": 385.9238,
+      "eval_samples_per_second": 312.772,
+      "eval_steps_per_second": 3.26,
+      "step": 510000
+    },
+    {
+      "epoch": 21.519200775618597,
+      "grad_norm": 1.445045828819275,
+      "learning_rate": 1.416198602238134e-05,
+      "loss": 3.8365,
+      "step": 510500
+    },
+    {
+      "epoch": 21.54027736795515,
+      "grad_norm": 1.3891557455062866,
+      "learning_rate": 1.4076233760665439e-05,
+      "loss": 3.8333,
+      "step": 511000
+    },
+    {
+      "epoch": 21.5613539602917,
+      "grad_norm": 1.4166592359542847,
+      "learning_rate": 1.3990481498949534e-05,
+      "loss": 3.8315,
+      "step": 511500
+    },
+    {
+      "epoch": 21.58243055262825,
+      "grad_norm": 1.4153056144714355,
+      "learning_rate": 1.3904729237233632e-05,
+      "loss": 3.8375,
+      "step": 512000
+    },
+    {
+      "epoch": 21.6035071449648,
+      "grad_norm": 1.4367507696151733,
+      "learning_rate": 1.381897697551773e-05,
+      "loss": 3.8339,
+      "step": 512500
+    },
+    {
+      "epoch": 21.624583737301354,
+      "grad_norm": 1.4607244729995728,
+      "learning_rate": 1.3733224713801825e-05,
+      "loss": 3.8375,
+      "step": 513000
+    },
+    {
+      "epoch": 21.645660329637906,
+      "grad_norm": 1.5248762369155884,
+      "learning_rate": 1.3647472452085924e-05,
+      "loss": 3.8388,
+      "step": 513500
+    },
+    {
+      "epoch": 21.666736921974454,
+      "grad_norm": 1.4983315467834473,
+      "learning_rate": 1.3561720190370022e-05,
+      "loss": 3.8364,
+      "step": 514000
+    },
+    {
+      "epoch": 21.687813514311006,
+      "grad_norm": 1.4084669351577759,
+      "learning_rate": 1.3475967928654119e-05,
+      "loss": 3.834,
+      "step": 514500
+    },
+    {
+      "epoch": 21.70889010664756,
+      "grad_norm": 1.4565075635910034,
+      "learning_rate": 1.3390215666938217e-05,
+      "loss": 3.8367,
+      "step": 515000
+    },
+    {
+      "epoch": 21.729966698984107,
+      "grad_norm": 1.4589897394180298,
+      "learning_rate": 1.3304463405222315e-05,
+      "loss": 3.8357,
+      "step": 515500
+    },
+    {
+      "epoch": 21.75104329132066,
+      "grad_norm": 1.4460359811782837,
+      "learning_rate": 1.321871114350641e-05,
+      "loss": 3.8374,
+      "step": 516000
+    },
+    {
+      "epoch": 21.77211988365721,
+      "grad_norm": 1.4368832111358643,
+      "learning_rate": 1.3132958881790508e-05,
+      "loss": 3.8312,
+      "step": 516500
+    },
+    {
+      "epoch": 21.79319647599376,
+      "grad_norm": 1.4360949993133545,
+      "learning_rate": 1.3047206620074607e-05,
+      "loss": 3.8326,
+      "step": 517000
+    },
+    {
+      "epoch": 21.814273068330312,
+      "grad_norm": 1.488476037979126,
+      "learning_rate": 1.2961454358358702e-05,
+      "loss": 3.8344,
+      "step": 517500
+    },
+    {
+      "epoch": 21.835349660666864,
+      "grad_norm": 1.4997751712799072,
+      "learning_rate": 1.28757020966428e-05,
+      "loss": 3.8311,
+      "step": 518000
+    },
+    {
+      "epoch": 21.856426253003413,
+      "grad_norm": 1.4730007648468018,
+      "learning_rate": 1.2789949834926898e-05,
+      "loss": 3.8303,
+      "step": 518500
+    },
+    {
+      "epoch": 21.877502845339965,
+      "grad_norm": 1.402915358543396,
+      "learning_rate": 1.2704197573210993e-05,
+      "loss": 3.8386,
+      "step": 519000
+    },
+    {
+      "epoch": 21.898579437676517,
+      "grad_norm": 1.438438057899475,
+      "learning_rate": 1.2618445311495092e-05,
+      "loss": 3.8317,
+      "step": 519500
+    },
+    {
+      "epoch": 21.91965603001307,
+      "grad_norm": 1.4627808332443237,
+      "learning_rate": 1.2532693049779186e-05,
+      "loss": 3.8286,
+      "step": 520000
+    },
+    {
+      "epoch": 21.91965603001307,
+      "eval_accuracy": 0.7268745022256391,
+      "eval_loss": 3.5842621326446533,
+      "eval_runtime": 373.2341,
+      "eval_samples_per_second": 323.406,
+      "eval_steps_per_second": 3.371,
+      "step": 520000
+    },
+    {
+      "epoch": 21.940732622349618,
+      "grad_norm": 1.5239932537078857,
+      "learning_rate": 1.2446940788063285e-05,
+      "loss": 3.831,
+      "step": 520500
+    },
+    {
+      "epoch": 21.96180921468617,
+      "grad_norm": 1.4567456245422363,
+      "learning_rate": 1.2361188526347383e-05,
+      "loss": 3.8326,
+      "step": 521000
+    },
+    {
+      "epoch": 21.982885807022722,
+      "grad_norm": 1.4611061811447144,
+      "learning_rate": 1.227543626463148e-05,
+      "loss": 3.8332,
+      "step": 521500
+    },
+    {
+      "epoch": 22.00396239935927,
+      "grad_norm": 1.4981279373168945,
+      "learning_rate": 1.2189684002915578e-05,
+      "loss": 3.8298,
+      "step": 522000
+    },
+    {
+      "epoch": 22.025038991695823,
+      "grad_norm": 1.4818317890167236,
+      "learning_rate": 1.2103931741199675e-05,
+      "loss": 3.8277,
+      "step": 522500
+    },
+    {
+      "epoch": 22.046115584032375,
+      "grad_norm": 1.4652643203735352,
+      "learning_rate": 1.2018179479483773e-05,
+      "loss": 3.8301,
+      "step": 523000
+    },
+    {
+      "epoch": 22.067192176368923,
+      "grad_norm": 1.4457072019577026,
+      "learning_rate": 1.193242721776787e-05,
+      "loss": 3.8312,
+      "step": 523500
+    },
+    {
+      "epoch": 22.088268768705476,
+      "grad_norm": 1.4907163381576538,
+      "learning_rate": 1.1846674956051966e-05,
+      "loss": 3.8337,
+      "step": 524000
+    },
+    {
+      "epoch": 22.109345361042028,
+      "grad_norm": 1.4775421619415283,
+      "learning_rate": 1.1760922694336064e-05,
+      "loss": 3.8271,
+      "step": 524500
+    },
+    {
+      "epoch": 22.130421953378576,
+      "grad_norm": 1.4771358966827393,
+      "learning_rate": 1.1675170432620161e-05,
+      "loss": 3.834,
+      "step": 525000
+    },
+    {
+      "epoch": 22.15149854571513,
+      "grad_norm": 1.4330247640609741,
+      "learning_rate": 1.1589418170904258e-05,
+      "loss": 3.8284,
+      "step": 525500
+    },
+    {
+      "epoch": 22.17257513805168,
+      "grad_norm": 1.4383000135421753,
+      "learning_rate": 1.1503665909188354e-05,
+      "loss": 3.8275,
+      "step": 526000
+    },
+    {
+      "epoch": 22.193651730388233,
+      "grad_norm": 1.5176277160644531,
+      "learning_rate": 1.1417913647472453e-05,
+      "loss": 3.8298,
+      "step": 526500
+    },
+    {
+      "epoch": 22.21472832272478,
+      "grad_norm": 1.4661767482757568,
+      "learning_rate": 1.133216138575655e-05,
+      "loss": 3.8298,
+      "step": 527000
+    },
+    {
+      "epoch": 22.235804915061333,
+      "grad_norm": 1.501128911972046,
+      "learning_rate": 1.1246409124040646e-05,
+      "loss": 3.8275,
+      "step": 527500
+    },
+    {
+      "epoch": 22.256881507397885,
+      "grad_norm": 1.4709721803665161,
+      "learning_rate": 1.1160656862324744e-05,
+      "loss": 3.831,
+      "step": 528000
+    },
+    {
+      "epoch": 22.277958099734434,
+      "grad_norm": 1.4800392389297485,
+      "learning_rate": 1.1074904600608842e-05,
+      "loss": 3.8279,
+      "step": 528500
+    },
+    {
+      "epoch": 22.299034692070986,
+      "grad_norm": 1.4496607780456543,
+      "learning_rate": 1.0989152338892939e-05,
+      "loss": 3.8263,
+      "step": 529000
+    },
+    {
+      "epoch": 22.320111284407538,
+      "grad_norm": 1.5543688535690308,
+      "learning_rate": 1.0903400077177037e-05,
+      "loss": 3.8305,
+      "step": 529500
+    },
+    {
+      "epoch": 22.341187876744087,
+      "grad_norm": 1.409175992012024,
+      "learning_rate": 1.0817647815461134e-05,
+      "loss": 3.8317,
+      "step": 530000
+    },
+    {
+      "epoch": 22.341187876744087,
+      "eval_accuracy": 0.727347362476077,
+      "eval_loss": 3.5817906856536865,
+      "eval_runtime": 350.7509,
+      "eval_samples_per_second": 344.136,
+      "eval_steps_per_second": 3.587,
+      "step": 530000
+    },
+    {
+      "epoch": 22.36226446908064,
+      "grad_norm": 1.4431580305099487,
+      "learning_rate": 1.073189555374523e-05,
+      "loss": 3.8301,
+      "step": 530500
+    },
+    {
+      "epoch": 22.38334106141719,
+      "grad_norm": 1.5079700946807861,
+      "learning_rate": 1.0646143292029327e-05,
+      "loss": 3.8314,
+      "step": 531000
+    },
+    {
+      "epoch": 22.40441765375374,
+      "grad_norm": 1.4759889841079712,
+      "learning_rate": 1.0560391030313425e-05,
+      "loss": 3.8245,
+      "step": 531500
+    },
+    {
+      "epoch": 22.42549424609029,
+      "grad_norm": 1.4321341514587402,
+      "learning_rate": 1.0474638768597522e-05,
+      "loss": 3.8268,
+      "step": 532000
+    },
+    {
+      "epoch": 22.446570838426844,
+      "grad_norm": 1.4706180095672607,
+      "learning_rate": 1.0388886506881619e-05,
+      "loss": 3.828,
+      "step": 532500
+    },
+    {
+      "epoch": 22.467647430763392,
+      "grad_norm": 1.484107255935669,
+      "learning_rate": 1.0303134245165717e-05,
+      "loss": 3.8299,
+      "step": 533000
+    },
+    {
+      "epoch": 22.488724023099945,
+      "grad_norm": 1.4722779989242554,
+      "learning_rate": 1.0217381983449814e-05,
+      "loss": 3.826,
+      "step": 533500
+    },
+    {
+      "epoch": 22.509800615436497,
+      "grad_norm": 1.4284369945526123,
+      "learning_rate": 1.013162972173391e-05,
+      "loss": 3.827,
+      "step": 534000
+    },
+    {
+      "epoch": 22.53087720777305,
+      "grad_norm": 1.4722604751586914,
+      "learning_rate": 1.0045877460018009e-05,
+      "loss": 3.8306,
+      "step": 534500
+    },
+    {
+      "epoch": 22.551953800109597,
+      "grad_norm": 1.4664009809494019,
+      "learning_rate": 9.960125198302105e-06,
+      "loss": 3.8323,
+      "step": 535000
+    },
+    {
+      "epoch": 22.57303039244615,
+      "grad_norm": 1.4398150444030762,
+      "learning_rate": 9.874372936586203e-06,
+      "loss": 3.829,
+      "step": 535500
+    },
+    {
+      "epoch": 22.5941069847827,
+      "grad_norm": 1.4772063493728638,
+      "learning_rate": 9.7886206748703e-06,
+      "loss": 3.8241,
+      "step": 536000
+    },
+    {
+      "epoch": 22.61518357711925,
+      "grad_norm": 1.4745800495147705,
+      "learning_rate": 9.702868413154398e-06,
+      "loss": 3.8251,
+      "step": 536500
+    },
+    {
+      "epoch": 22.636260169455802,
+      "grad_norm": 1.4967702627182007,
+      "learning_rate": 9.617116151438495e-06,
+      "loss": 3.8265,
+      "step": 537000
+    },
+    {
+      "epoch": 22.657336761792354,
+      "grad_norm": 1.473240613937378,
+      "learning_rate": 9.531363889722592e-06,
+      "loss": 3.8274,
+      "step": 537500
+    },
+    {
+      "epoch": 22.678413354128903,
+      "grad_norm": 1.4615875482559204,
+      "learning_rate": 9.44561162800669e-06,
+      "loss": 3.8254,
+      "step": 538000
+    },
+    {
+      "epoch": 22.699489946465455,
+      "grad_norm": 1.4585999250411987,
+      "learning_rate": 9.359859366290787e-06,
+      "loss": 3.8278,
+      "step": 538500
+    },
+    {
+      "epoch": 22.720566538802007,
+      "grad_norm": 1.4159053564071655,
+      "learning_rate": 9.274107104574883e-06,
+      "loss": 3.8284,
+      "step": 539000
+    },
+    {
+      "epoch": 22.741643131138556,
+      "grad_norm": 1.4406895637512207,
+      "learning_rate": 9.18835484285898e-06,
+      "loss": 3.8238,
+      "step": 539500
+    },
+    {
+      "epoch": 22.762719723475108,
+      "grad_norm": 1.4831504821777344,
+      "learning_rate": 9.102602581143078e-06,
+      "loss": 3.8243,
+      "step": 540000
+    },
+    {
+      "epoch": 22.762719723475108,
+      "eval_accuracy": 0.7275684882715138,
+      "eval_loss": 3.5811142921447754,
+      "eval_runtime": 351.8646,
+      "eval_samples_per_second": 343.047,
+      "eval_steps_per_second": 3.575,
+      "step": 540000
+    },
+    {
+      "epoch": 22.78379631581166,
+      "grad_norm": 1.505347490310669,
+      "learning_rate": 9.016850319427175e-06,
+      "loss": 3.8285,
+      "step": 540500
+    },
+    {
+      "epoch": 22.804872908148212,
+      "grad_norm": 1.4828722476959229,
+      "learning_rate": 8.931098057711271e-06,
+      "loss": 3.8275,
+      "step": 541000
+    },
+    {
+      "epoch": 22.82594950048476,
+      "grad_norm": 1.5085564851760864,
+      "learning_rate": 8.84534579599537e-06,
+      "loss": 3.8277,
+      "step": 541500
+    },
+    {
+      "epoch": 22.847026092821313,
+      "grad_norm": 1.5291873216629028,
+      "learning_rate": 8.759593534279468e-06,
+      "loss": 3.8278,
+      "step": 542000
+    },
+    {
+      "epoch": 22.868102685157865,
+      "grad_norm": 1.4287058115005493,
+      "learning_rate": 8.673841272563565e-06,
+      "loss": 3.8254,
+      "step": 542500
+    },
+    {
+      "epoch": 22.889179277494414,
+      "grad_norm": 1.4588969945907593,
+      "learning_rate": 8.588089010847663e-06,
+      "loss": 3.8295,
+      "step": 543000
+    },
+    {
+      "epoch": 22.910255869830966,
+      "grad_norm": 1.4794321060180664,
+      "learning_rate": 8.50233674913176e-06,
+      "loss": 3.8255,
+      "step": 543500
+    },
+    {
+      "epoch": 22.931332462167518,
+      "grad_norm": 1.537391185760498,
+      "learning_rate": 8.416584487415856e-06,
+      "loss": 3.8235,
+      "step": 544000
+    },
+    {
+      "epoch": 22.952409054504066,
+      "grad_norm": 1.5013211965560913,
+      "learning_rate": 8.330832225699953e-06,
+      "loss": 3.8209,
+      "step": 544500
+    },
+    {
+      "epoch": 22.97348564684062,
+      "grad_norm": 1.4525611400604248,
+      "learning_rate": 8.245079963984051e-06,
+      "loss": 3.8232,
+      "step": 545000
+    },
+    {
+      "epoch": 22.99456223917717,
+      "grad_norm": 1.5167020559310913,
+      "learning_rate": 8.159327702268148e-06,
+      "loss": 3.8272,
+      "step": 545500
+    },
+    {
+      "epoch": 23.01563883151372,
+      "grad_norm": 1.4540268182754517,
+      "learning_rate": 8.073575440552244e-06,
+      "loss": 3.8241,
+      "step": 546000
+    },
+    {
+      "epoch": 23.03671542385027,
+      "grad_norm": 1.4601263999938965,
+      "learning_rate": 7.987823178836343e-06,
+      "loss": 3.8236,
+      "step": 546500
+    },
+    {
+      "epoch": 23.057792016186824,
+      "grad_norm": 1.4744279384613037,
+      "learning_rate": 7.90207091712044e-06,
+      "loss": 3.823,
+      "step": 547000
+    },
+    {
+      "epoch": 23.078868608523376,
+      "grad_norm": 1.531900405883789,
+      "learning_rate": 7.816318655404536e-06,
+      "loss": 3.8238,
+      "step": 547500
+    },
+    {
+      "epoch": 23.099945200859924,
+      "grad_norm": 1.46993088722229,
+      "learning_rate": 7.730566393688634e-06,
+      "loss": 3.8244,
+      "step": 548000
+    },
+    {
+      "epoch": 23.121021793196476,
+      "grad_norm": 1.464343547821045,
+      "learning_rate": 7.64481413197273e-06,
+      "loss": 3.8239,
+      "step": 548500
+    },
+    {
+      "epoch": 23.14209838553303,
+      "grad_norm": 1.493773102760315,
+      "learning_rate": 7.559061870256828e-06,
+      "loss": 3.8252,
+      "step": 549000
+    },
+    {
+      "epoch": 23.163174977869577,
+      "grad_norm": 1.4178105592727661,
+      "learning_rate": 7.473309608540925e-06,
+      "loss": 3.8226,
+      "step": 549500
+    },
+    {
+      "epoch": 23.18425157020613,
+      "grad_norm": 1.4994392395019531,
+      "learning_rate": 7.387557346825023e-06,
+      "loss": 3.825,
+      "step": 550000
+    },
+    {
+      "epoch": 23.18425157020613,
+      "eval_accuracy": 0.7279059535983837,
+      "eval_loss": 3.5771191120147705,
+      "eval_runtime": 454.7952,
+      "eval_samples_per_second": 265.407,
+      "eval_steps_per_second": 2.766,
+      "step": 550000
+    },
+    {
+      "epoch": 23.20532816254268,
+      "grad_norm": 1.5079748630523682,
+      "learning_rate": 7.3018050851091205e-06,
+      "loss": 3.8203,
+      "step": 550500
+    },
+    {
+      "epoch": 23.22640475487923,
+      "grad_norm": 1.458299994468689,
+      "learning_rate": 7.216052823393217e-06,
+      "loss": 3.823,
+      "step": 551000
+    },
+    {
+      "epoch": 23.247481347215782,
+      "grad_norm": 1.4897270202636719,
+      "learning_rate": 7.1303005616773155e-06,
+      "loss": 3.823,
+      "step": 551500
+    },
+    {
+      "epoch": 23.268557939552334,
+      "grad_norm": 1.513429880142212,
+      "learning_rate": 7.044548299961412e-06,
+      "loss": 3.8277,
+      "step": 552000
+    },
+    {
+      "epoch": 23.289634531888883,
+      "grad_norm": 1.5009933710098267,
+      "learning_rate": 6.958796038245509e-06,
+      "loss": 3.8164,
+      "step": 552500
+    },
+    {
+      "epoch": 23.310711124225435,
+      "grad_norm": 1.553181767463684,
+      "learning_rate": 6.873043776529606e-06,
+      "loss": 3.8275,
+      "step": 553000
+    },
+    {
+      "epoch": 23.331787716561987,
+      "grad_norm": 1.4983340501785278,
+      "learning_rate": 6.787291514813704e-06,
+      "loss": 3.8197,
+      "step": 553500
+    },
+    {
+      "epoch": 23.35286430889854,
+      "grad_norm": 1.4900944232940674,
+      "learning_rate": 6.701539253097801e-06,
+      "loss": 3.8219,
+      "step": 554000
+    },
+    {
+      "epoch": 23.373940901235088,
+      "grad_norm": 1.5108531713485718,
+      "learning_rate": 6.615786991381898e-06,
+      "loss": 3.8225,
+      "step": 554500
+    },
+    {
+      "epoch": 23.39501749357164,
+      "grad_norm": 1.425534963607788,
+      "learning_rate": 6.530034729665996e-06,
+      "loss": 3.8223,
+      "step": 555000
+    },
+    {
+      "epoch": 23.416094085908192,
+      "grad_norm": 1.4707149267196655,
+      "learning_rate": 6.444282467950093e-06,
+      "loss": 3.8231,
+      "step": 555500
+    },
+    {
+      "epoch": 23.43717067824474,
+      "grad_norm": 1.464470624923706,
+      "learning_rate": 6.358530206234189e-06,
+      "loss": 3.8192,
+      "step": 556000
+    },
+    {
+      "epoch": 23.458247270581293,
+      "grad_norm": 1.4853909015655518,
+      "learning_rate": 6.272777944518287e-06,
+      "loss": 3.8233,
+      "step": 556500
+    },
+    {
+      "epoch": 23.479323862917845,
+      "grad_norm": 1.475545883178711,
+      "learning_rate": 6.187025682802384e-06,
+      "loss": 3.8236,
+      "step": 557000
+    },
+    {
+      "epoch": 23.500400455254393,
+      "grad_norm": 1.5471032857894897,
+      "learning_rate": 6.101273421086482e-06,
+      "loss": 3.8211,
+      "step": 557500
+    },
+    {
+      "epoch": 23.521477047590945,
+      "grad_norm": 1.451910138130188,
+      "learning_rate": 6.015521159370579e-06,
+      "loss": 3.8205,
+      "step": 558000
+    },
+    {
+      "epoch": 23.542553639927497,
+      "grad_norm": 1.4798431396484375,
+      "learning_rate": 5.929768897654676e-06,
+      "loss": 3.8202,
+      "step": 558500
+    },
+    {
+      "epoch": 23.563630232264046,
+      "grad_norm": 1.4835596084594727,
+      "learning_rate": 5.844016635938773e-06,
+      "loss": 3.8188,
+      "step": 559000
+    },
+    {
+      "epoch": 23.584706824600598,
+      "grad_norm": 1.4456976652145386,
+      "learning_rate": 5.75826437422287e-06,
+      "loss": 3.8213,
+      "step": 559500
+    },
+    {
+      "epoch": 23.60578341693715,
+      "grad_norm": 1.528678297996521,
+      "learning_rate": 5.672512112506967e-06,
+      "loss": 3.8227,
+      "step": 560000
+    },
+    {
+      "epoch": 23.60578341693715,
+      "eval_accuracy": 0.7284422515045915,
+      "eval_loss": 3.5752360820770264,
+      "eval_runtime": 390.9964,
+      "eval_samples_per_second": 308.714,
+      "eval_steps_per_second": 3.217,
+      "step": 560000
+    },
+    {
+      "epoch": 23.6268600092737,
+      "grad_norm": 1.512743353843689,
+      "learning_rate": 5.5867598507910655e-06,
+      "loss": 3.8214,
+      "step": 560500
+    },
+    {
+      "epoch": 23.64793660161025,
+      "grad_norm": 1.4717857837677002,
+      "learning_rate": 5.501007589075162e-06,
+      "loss": 3.8211,
+      "step": 561000
+    },
+    {
+      "epoch": 23.669013193946803,
+      "grad_norm": 1.5611804723739624,
+      "learning_rate": 5.41525532735926e-06,
+      "loss": 3.8222,
+      "step": 561500
+    },
+    {
+      "epoch": 23.690089786283355,
+      "grad_norm": 1.5186790227890015,
+      "learning_rate": 5.329503065643356e-06,
+      "loss": 3.8172,
+      "step": 562000
+    },
+    {
+      "epoch": 23.711166378619904,
+      "grad_norm": 1.5380473136901855,
+      "learning_rate": 5.243750803927454e-06,
+      "loss": 3.8181,
+      "step": 562500
+    },
+    {
+      "epoch": 23.732242970956456,
+      "grad_norm": 1.455582618713379,
+      "learning_rate": 5.157998542211551e-06,
+      "loss": 3.8214,
+      "step": 563000
+    },
+    {
+      "epoch": 23.753319563293008,
+      "grad_norm": 1.436585545539856,
+      "learning_rate": 5.072246280495649e-06,
+      "loss": 3.8165,
+      "step": 563500
+    },
+    {
+      "epoch": 23.774396155629557,
+      "grad_norm": 1.4792672395706177,
+      "learning_rate": 4.986494018779746e-06,
+      "loss": 3.8196,
+      "step": 564000
+    },
+    {
+      "epoch": 23.79547274796611,
+      "grad_norm": 1.5203779935836792,
+      "learning_rate": 4.900741757063843e-06,
+      "loss": 3.8186,
+      "step": 564500
+    },
+    {
+      "epoch": 23.81654934030266,
+      "grad_norm": 1.5053966045379639,
+      "learning_rate": 4.81498949534794e-06,
+      "loss": 3.8185,
+      "step": 565000
+    },
+    {
+      "epoch": 23.83762593263921,
+      "grad_norm": 1.4720145463943481,
+      "learning_rate": 4.729237233632038e-06,
+      "loss": 3.8198,
+      "step": 565500
+    },
+    {
+      "epoch": 23.85870252497576,
+      "grad_norm": 1.449937105178833,
+      "learning_rate": 4.643484971916134e-06,
+      "loss": 3.8203,
+      "step": 566000
+    },
+    {
+      "epoch": 23.879779117312314,
+      "grad_norm": 1.486830711364746,
+      "learning_rate": 4.557732710200232e-06,
+      "loss": 3.8182,
+      "step": 566500
+    },
+    {
+      "epoch": 23.900855709648862,
+      "grad_norm": 1.4971548318862915,
+      "learning_rate": 4.471980448484329e-06,
+      "loss": 3.8185,
+      "step": 567000
+    },
+    {
+      "epoch": 23.921932301985414,
+      "grad_norm": 1.5345858335494995,
+      "learning_rate": 4.386228186768427e-06,
+      "loss": 3.8172,
+      "step": 567500
+    },
+    {
+      "epoch": 23.943008894321967,
+      "grad_norm": 1.4782081842422485,
+      "learning_rate": 4.300475925052524e-06,
+      "loss": 3.8192,
+      "step": 568000
+    },
+    {
+      "epoch": 23.96408548665852,
+      "grad_norm": 1.4870619773864746,
+      "learning_rate": 4.214723663336621e-06,
+      "loss": 3.823,
+      "step": 568500
+    },
+    {
+      "epoch": 23.985162078995067,
+      "grad_norm": 1.5145857334136963,
+      "learning_rate": 4.128971401620718e-06,
+      "loss": 3.816,
+      "step": 569000
+    },
+    {
+      "epoch": 24.00623867133162,
+      "grad_norm": 1.4795106649398804,
+      "learning_rate": 4.043219139904815e-06,
+      "loss": 3.8187,
+      "step": 569500
+    },
+    {
+      "epoch": 24.02731526366817,
+      "grad_norm": 1.4708019495010376,
+      "learning_rate": 3.957466878188912e-06,
+      "loss": 3.8173,
+      "step": 570000
+    },
+    {
+      "epoch": 24.02731526366817,
+      "eval_accuracy": 0.7286428497693995,
+      "eval_loss": 3.5740878582000732,
+      "eval_runtime": 356.1729,
+      "eval_samples_per_second": 338.897,
+      "eval_steps_per_second": 3.532,
+      "step": 570000
+    },
+    {
+      "epoch": 24.04839185600472,
+      "grad_norm": 1.517836332321167,
+      "learning_rate": 3.87171461647301e-06,
+      "loss": 3.8228,
+      "step": 570500
+    },
+    {
+      "epoch": 24.069468448341272,
+      "grad_norm": 1.475856900215149,
+      "learning_rate": 3.7859623547571067e-06,
+      "loss": 3.8162,
+      "step": 571000
+    },
+    {
+      "epoch": 24.090545040677824,
+      "grad_norm": 1.499794363975525,
+      "learning_rate": 3.7002100930412046e-06,
+      "loss": 3.8143,
+      "step": 571500
+    },
+    {
+      "epoch": 24.111621633014373,
+      "grad_norm": 1.4421933889389038,
+      "learning_rate": 3.614457831325301e-06,
+      "loss": 3.8222,
+      "step": 572000
+    },
+    {
+      "epoch": 24.132698225350925,
+      "grad_norm": 1.5154500007629395,
+      "learning_rate": 3.5287055696093987e-06,
+      "loss": 3.8178,
+      "step": 572500
+    },
+    {
+      "epoch": 24.153774817687477,
+      "grad_norm": 1.4712227582931519,
+      "learning_rate": 3.4429533078934957e-06,
+      "loss": 3.818,
+      "step": 573000
+    },
+    {
+      "epoch": 24.174851410024026,
+      "grad_norm": 1.4797468185424805,
+      "learning_rate": 3.357201046177593e-06,
+      "loss": 3.8211,
+      "step": 573500
+    },
+    {
+      "epoch": 24.195928002360578,
+      "grad_norm": 1.4831593036651611,
+      "learning_rate": 3.2714487844616906e-06,
+      "loss": 3.8201,
+      "step": 574000
+    },
+    {
+      "epoch": 24.21700459469713,
+      "grad_norm": 1.4542059898376465,
+      "learning_rate": 3.1856965227457872e-06,
+      "loss": 3.8197,
+      "step": 574500
+    },
+    {
+      "epoch": 24.238081187033682,
+      "grad_norm": 1.450979232788086,
+      "learning_rate": 3.099944261029885e-06,
+      "loss": 3.8231,
+      "step": 575000
+    },
+    {
+      "epoch": 24.25915777937023,
+      "grad_norm": 1.517579197883606,
+      "learning_rate": 3.014191999313982e-06,
+      "loss": 3.8192,
+      "step": 575500
+    },
+    {
+      "epoch": 24.280234371706783,
+      "grad_norm": 1.517797589302063,
+      "learning_rate": 2.928439737598079e-06,
+      "loss": 3.8154,
+      "step": 576000
+    },
+    {
+      "epoch": 24.301310964043335,
+      "grad_norm": 1.4999089241027832,
+      "learning_rate": 2.8426874758821767e-06,
+      "loss": 3.8181,
+      "step": 576500
+    },
+    {
+      "epoch": 24.322387556379883,
+      "grad_norm": 1.45829176902771,
+      "learning_rate": 2.7569352141662737e-06,
+      "loss": 3.8127,
+      "step": 577000
+    },
+    {
+      "epoch": 24.343464148716436,
+      "grad_norm": 1.5128980875015259,
+      "learning_rate": 2.6711829524503707e-06,
+      "loss": 3.8141,
+      "step": 577500
+    },
+    {
+      "epoch": 24.364540741052988,
+      "grad_norm": 1.5487525463104248,
+      "learning_rate": 2.585430690734468e-06,
+      "loss": 3.8139,
+      "step": 578000
+    },
+    {
+      "epoch": 24.385617333389536,
+      "grad_norm": 1.4674789905548096,
+      "learning_rate": 2.4996784290185657e-06,
+      "loss": 3.8169,
+      "step": 578500
+    },
+    {
+      "epoch": 24.40669392572609,
+      "grad_norm": 1.5204170942306519,
+      "learning_rate": 2.4139261673026627e-06,
+      "loss": 3.8125,
+      "step": 579000
+    },
+    {
+      "epoch": 24.42777051806264,
+      "grad_norm": 1.4631026983261108,
+      "learning_rate": 2.3281739055867597e-06,
+      "loss": 3.8156,
+      "step": 579500
+    },
+    {
+      "epoch": 24.44884711039919,
+      "grad_norm": 1.4859623908996582,
+      "learning_rate": 2.242421643870857e-06,
+      "loss": 3.8174,
+      "step": 580000
+    },
+    {
+      "epoch": 24.44884711039919,
+      "eval_accuracy": 0.7290958354126942,
+      "eval_loss": 3.5705835819244385,
+      "eval_runtime": 336.1839,
+      "eval_samples_per_second": 359.047,
+      "eval_steps_per_second": 3.742,
+      "step": 580000
+    },
+    {
+      "epoch": 24.46992370273574,
+      "grad_norm": 1.5068862438201904,
+      "learning_rate": 2.1566693821549547e-06,
+      "loss": 3.8151,
+      "step": 580500
+    },
+    {
+      "epoch": 24.491000295072293,
+      "grad_norm": 1.5207760334014893,
+      "learning_rate": 2.0709171204390517e-06,
+      "loss": 3.8181,
+      "step": 581000
+    },
+    {
+      "epoch": 24.512076887408845,
+      "grad_norm": 1.509458065032959,
+      "learning_rate": 1.985164858723149e-06,
+      "loss": 3.813,
+      "step": 581500
+    },
+    {
+      "epoch": 24.533153479745394,
+      "grad_norm": 1.506588339805603,
+      "learning_rate": 1.8994125970072462e-06,
+      "loss": 3.8192,
+      "step": 582000
+    },
+    {
+      "epoch": 24.554230072081946,
+      "grad_norm": 1.507524013519287,
+      "learning_rate": 1.8136603352913432e-06,
+      "loss": 3.8147,
+      "step": 582500
+    },
+    {
+      "epoch": 24.5753066644185,
+      "grad_norm": 1.4904346466064453,
+      "learning_rate": 1.7279080735754405e-06,
+      "loss": 3.8185,
+      "step": 583000
+    },
+    {
+      "epoch": 24.596383256755047,
+      "grad_norm": 1.5087367296218872,
+      "learning_rate": 1.642155811859538e-06,
+      "loss": 3.8176,
+      "step": 583500
+    },
+    {
+      "epoch": 24.6174598490916,
+      "grad_norm": 1.496445655822754,
+      "learning_rate": 1.5564035501436352e-06,
+      "loss": 3.8152,
+      "step": 584000
+    },
+    {
+      "epoch": 24.63853644142815,
+      "grad_norm": 1.4876972436904907,
+      "learning_rate": 1.4706512884277324e-06,
+      "loss": 3.8169,
+      "step": 584500
+    },
+    {
+      "epoch": 24.6596130337647,
+      "grad_norm": 1.4217205047607422,
+      "learning_rate": 1.3848990267118295e-06,
+      "loss": 3.8131,
+      "step": 585000
+    },
+    {
+      "epoch": 24.680689626101252,
+      "grad_norm": 1.5284245014190674,
+      "learning_rate": 1.299146764995927e-06,
+      "loss": 3.8157,
+      "step": 585500
+    },
+    {
+      "epoch": 24.701766218437804,
+      "grad_norm": 1.5545495748519897,
+      "learning_rate": 1.213394503280024e-06,
+      "loss": 3.8128,
+      "step": 586000
+    },
+    {
+      "epoch": 24.722842810774353,
+      "grad_norm": 1.490014910697937,
+      "learning_rate": 1.1276422415641212e-06,
+      "loss": 3.8141,
+      "step": 586500
+    },
+    {
+      "epoch": 24.743919403110905,
+      "grad_norm": 1.5050127506256104,
+      "learning_rate": 1.0418899798482187e-06,
+      "loss": 3.8179,
+      "step": 587000
+    },
+    {
+      "epoch": 24.764995995447457,
+      "grad_norm": 1.5044207572937012,
+      "learning_rate": 9.561377181323157e-07,
+      "loss": 3.8126,
+      "step": 587500
+    },
+    {
+      "epoch": 24.786072587784005,
+      "grad_norm": 1.484796166419983,
+      "learning_rate": 8.70385456416413e-07,
+      "loss": 3.8148,
+      "step": 588000
+    },
+    {
+      "epoch": 24.807149180120557,
+      "grad_norm": 1.4478814601898193,
+      "learning_rate": 7.846331947005103e-07,
+      "loss": 3.8156,
+      "step": 588500
+    },
+    {
+      "epoch": 24.82822577245711,
+      "grad_norm": 1.472593069076538,
+      "learning_rate": 6.988809329846075e-07,
+      "loss": 3.8112,
+      "step": 589000
+    },
+    {
+      "epoch": 24.84930236479366,
+      "grad_norm": 1.5285813808441162,
+      "learning_rate": 6.131286712687047e-07,
+      "loss": 3.8159,
+      "step": 589500
+    },
+    {
+      "epoch": 24.87037895713021,
+      "grad_norm": 1.48330557346344,
+      "learning_rate": 5.27376409552802e-07,
+      "loss": 3.8141,
+      "step": 590000
+    },
+    {
+      "epoch": 24.87037895713021,
+      "eval_accuracy": 0.728923763224559,
+      "eval_loss": 3.5708119869232178,
+      "eval_runtime": 304.2836,
+      "eval_samples_per_second": 396.689,
+      "eval_steps_per_second": 4.134,
+      "step": 590000
+    },
+    {
+      "epoch": 24.891455549466762,
+      "grad_norm": 1.4938112497329712,
+      "learning_rate": 4.4162414783689917e-07,
+      "loss": 3.8144,
+      "step": 590500
+    },
+    {
+      "epoch": 24.912532141803315,
+      "grad_norm": 1.4734259843826294,
+      "learning_rate": 3.558718861209964e-07,
+      "loss": 3.8135,
+      "step": 591000
+    },
+    {
+      "epoch": 24.933608734139863,
+      "grad_norm": 1.5110267400741577,
+      "learning_rate": 2.701196244050937e-07,
+      "loss": 3.8148,
+      "step": 591500
+    },
+    {
+      "epoch": 24.954685326476415,
+      "grad_norm": 1.5400500297546387,
+      "learning_rate": 1.8436736268919094e-07,
+      "loss": 3.8152,
+      "step": 592000
+    },
+    {
+      "epoch": 24.975761918812967,
+      "grad_norm": 1.4537055492401123,
+      "learning_rate": 9.861510097328816e-08,
+      "loss": 3.8128,
+      "step": 592500
+    },
+    {
+      "epoch": 24.996838511149516,
+      "grad_norm": 1.4816259145736694,
+      "learning_rate": 1.2862839257385414e-08,
+      "loss": 3.8132,
+      "step": 593000
+    },
+    {
+      "epoch": 25.0,
+      "step": 593075,
+      "total_flos": 7.653161630006784e+18,
+      "train_loss": 4.095287137130649,
+      "train_runtime": 218339.8454,
+      "train_samples_per_second": 260.757,
+      "train_steps_per_second": 2.716
+    }
+  ],
+  "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": 7.653161630006784e+18,
+  "train_batch_size": 96,
+  "trial_name": null,
+  "trial_params": null
+}