diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8863 @@
+{
+  "best_metric": 3.1032865047454834,
+  "best_model_checkpoint": "tiny_bert_bc_rand_5_v1/checkpoint-340000",
+  "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": 3.145738124847412,
+      "learning_rate": 5e-06,
+      "loss": 11.2754,
+      "step": 500
+    },
+    {
+      "epoch": 0.04215318467310205,
+      "grad_norm": 1.5895618200302124,
+      "learning_rate": 1e-05,
+      "loss": 9.2694,
+      "step": 1000
+    },
+    {
+      "epoch": 0.06322977700965308,
+      "grad_norm": 1.838132381439209,
+      "learning_rate": 1.5e-05,
+      "loss": 7.9814,
+      "step": 1500
+    },
+    {
+      "epoch": 0.0843063693462041,
+      "grad_norm": 1.5314830541610718,
+      "learning_rate": 2e-05,
+      "loss": 7.7335,
+      "step": 2000
+    },
+    {
+      "epoch": 0.10538296168275513,
+      "grad_norm": 1.6739658117294312,
+      "learning_rate": 2.5e-05,
+      "loss": 7.6209,
+      "step": 2500
+    },
+    {
+      "epoch": 0.12645955401930617,
+      "grad_norm": 1.3434497117996216,
+      "learning_rate": 3e-05,
+      "loss": 7.5511,
+      "step": 3000
+    },
+    {
+      "epoch": 0.14753614635585718,
+      "grad_norm": 1.6727228164672852,
+      "learning_rate": 3.5e-05,
+      "loss": 7.5028,
+      "step": 3500
+    },
+    {
+      "epoch": 0.1686127386924082,
+      "grad_norm": 0.9008095860481262,
+      "learning_rate": 4e-05,
+      "loss": 7.4633,
+      "step": 4000
+    },
+    {
+      "epoch": 0.18968933102895924,
+      "grad_norm": 2.350503444671631,
+      "learning_rate": 4.5e-05,
+      "loss": 7.4294,
+      "step": 4500
+    },
+    {
+      "epoch": 0.21076592336551025,
+      "grad_norm": 1.2562755346298218,
+      "learning_rate": 5e-05,
+      "loss": 7.4059,
+      "step": 5000
+    },
+    {
+      "epoch": 0.2318425157020613,
+      "grad_norm": 2.4663517475128174,
+      "learning_rate": 5.500000000000001e-05,
+      "loss": 7.3895,
+      "step": 5500
+    },
+    {
+      "epoch": 0.25291910803861234,
+      "grad_norm": 1.3434195518493652,
+      "learning_rate": 6e-05,
+      "loss": 7.3648,
+      "step": 6000
+    },
+    {
+      "epoch": 0.27399570037516335,
+      "grad_norm": 1.0526429414749146,
+      "learning_rate": 6.500000000000001e-05,
+      "loss": 7.3559,
+      "step": 6500
+    },
+    {
+      "epoch": 0.29507229271171437,
+      "grad_norm": 1.7030152082443237,
+      "learning_rate": 7e-05,
+      "loss": 7.3382,
+      "step": 7000
+    },
+    {
+      "epoch": 0.3161488850482654,
+      "grad_norm": 1.8441736698150635,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 7.3277,
+      "step": 7500
+    },
+    {
+      "epoch": 0.3372254773848164,
+      "grad_norm": 2.034295082092285,
+      "learning_rate": 8e-05,
+      "loss": 7.3228,
+      "step": 8000
+    },
+    {
+      "epoch": 0.35830206972136747,
+      "grad_norm": 1.9254813194274902,
+      "learning_rate": 8.5e-05,
+      "loss": 7.3081,
+      "step": 8500
+    },
+    {
+      "epoch": 0.3793786620579185,
+      "grad_norm": 1.9745136499404907,
+      "learning_rate": 9e-05,
+      "loss": 7.2967,
+      "step": 9000
+    },
+    {
+      "epoch": 0.4004552543944695,
+      "grad_norm": 2.017653465270996,
+      "learning_rate": 9.5e-05,
+      "loss": 7.2885,
+      "step": 9500
+    },
+    {
+      "epoch": 0.4215318467310205,
+      "grad_norm": 1.2156219482421875,
+      "learning_rate": 0.0001,
+      "loss": 7.2824,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4215318467310205,
+      "eval_accuracy": 0.15827706969843475,
+      "eval_loss": 7.1206817626953125,
+      "eval_runtime": 171.532,
+      "eval_samples_per_second": 703.694,
+      "eval_steps_per_second": 7.334,
+      "step": 10000
+    },
+    {
+      "epoch": 0.4426084390675716,
+      "grad_norm": 1.5811258554458618,
+      "learning_rate": 9.99142477382841e-05,
+      "loss": 7.2728,
+      "step": 10500
+    },
+    {
+      "epoch": 0.4636850314041226,
+      "grad_norm": 1.505975365638733,
+      "learning_rate": 9.98284954765682e-05,
+      "loss": 7.2682,
+      "step": 11000
+    },
+    {
+      "epoch": 0.4847616237406736,
+      "grad_norm": 1.8331820964813232,
+      "learning_rate": 9.97427432148523e-05,
+      "loss": 7.2676,
+      "step": 11500
+    },
+    {
+      "epoch": 0.5058382160772247,
+      "grad_norm": 1.3567615747451782,
+      "learning_rate": 9.96569909531364e-05,
+      "loss": 7.2568,
+      "step": 12000
+    },
+    {
+      "epoch": 0.5269148084137757,
+      "grad_norm": 1.5720763206481934,
+      "learning_rate": 9.957123869142048e-05,
+      "loss": 7.2516,
+      "step": 12500
+    },
+    {
+      "epoch": 0.5479914007503267,
+      "grad_norm": 2.517540216445923,
+      "learning_rate": 9.94854864297046e-05,
+      "loss": 7.2492,
+      "step": 13000
+    },
+    {
+      "epoch": 0.5690679930868777,
+      "grad_norm": 0.9260111451148987,
+      "learning_rate": 9.939973416798869e-05,
+      "loss": 7.2389,
+      "step": 13500
+    },
+    {
+      "epoch": 0.5901445854234287,
+      "grad_norm": 2.44075083732605,
+      "learning_rate": 9.931398190627278e-05,
+      "loss": 7.2359,
+      "step": 14000
+    },
+    {
+      "epoch": 0.6112211777599798,
+      "grad_norm": 2.195629358291626,
+      "learning_rate": 9.922822964455688e-05,
+      "loss": 7.2285,
+      "step": 14500
+    },
+    {
+      "epoch": 0.6322977700965308,
+      "grad_norm": 2.1929731369018555,
+      "learning_rate": 9.914247738284099e-05,
+      "loss": 7.2275,
+      "step": 15000
+    },
+    {
+      "epoch": 0.6533743624330818,
+      "grad_norm": 3.0161330699920654,
+      "learning_rate": 9.905672512112507e-05,
+      "loss": 7.2187,
+      "step": 15500
+    },
+    {
+      "epoch": 0.6744509547696328,
+      "grad_norm": 1.000184178352356,
+      "learning_rate": 9.897097285940917e-05,
+      "loss": 7.2185,
+      "step": 16000
+    },
+    {
+      "epoch": 0.6955275471061839,
+      "grad_norm": 2.779953718185425,
+      "learning_rate": 9.888522059769327e-05,
+      "loss": 7.2117,
+      "step": 16500
+    },
+    {
+      "epoch": 0.7166041394427349,
+      "grad_norm": 1.8206664323806763,
+      "learning_rate": 9.879946833597737e-05,
+      "loss": 7.2094,
+      "step": 17000
+    },
+    {
+      "epoch": 0.737680731779286,
+      "grad_norm": 1.5893921852111816,
+      "learning_rate": 9.871371607426147e-05,
+      "loss": 7.2025,
+      "step": 17500
+    },
+    {
+      "epoch": 0.758757324115837,
+      "grad_norm": 1.892212986946106,
+      "learning_rate": 9.862796381254556e-05,
+      "loss": 7.2093,
+      "step": 18000
+    },
+    {
+      "epoch": 0.779833916452388,
+      "grad_norm": 0.8703418374061584,
+      "learning_rate": 9.854221155082965e-05,
+      "loss": 7.2012,
+      "step": 18500
+    },
+    {
+      "epoch": 0.800910508788939,
+      "grad_norm": 1.2937694787979126,
+      "learning_rate": 9.845645928911376e-05,
+      "loss": 7.2031,
+      "step": 19000
+    },
+    {
+      "epoch": 0.82198710112549,
+      "grad_norm": 2.5257461071014404,
+      "learning_rate": 9.837070702739786e-05,
+      "loss": 7.199,
+      "step": 19500
+    },
+    {
+      "epoch": 0.843063693462041,
+      "grad_norm": 2.28627610206604,
+      "learning_rate": 9.828495476568194e-05,
+      "loss": 7.1705,
+      "step": 20000
+    },
+    {
+      "epoch": 0.843063693462041,
+      "eval_accuracy": 0.17264274978346936,
+      "eval_loss": 6.979538440704346,
+      "eval_runtime": 170.6429,
+      "eval_samples_per_second": 707.36,
+      "eval_steps_per_second": 7.372,
+      "step": 20000
+    },
+    {
+      "epoch": 0.864140285798592,
+      "grad_norm": 2.5543596744537354,
+      "learning_rate": 9.819920250396604e-05,
+      "loss": 7.1134,
+      "step": 20500
+    },
+    {
+      "epoch": 0.8852168781351432,
+      "grad_norm": 1.5039359331130981,
+      "learning_rate": 9.811345024225015e-05,
+      "loss": 7.0448,
+      "step": 21000
+    },
+    {
+      "epoch": 0.9062934704716942,
+      "grad_norm": 2.339515209197998,
+      "learning_rate": 9.802769798053424e-05,
+      "loss": 6.9528,
+      "step": 21500
+    },
+    {
+      "epoch": 0.9273700628082452,
+      "grad_norm": 3.437640905380249,
+      "learning_rate": 9.794194571881834e-05,
+      "loss": 6.5948,
+      "step": 22000
+    },
+    {
+      "epoch": 0.9484466551447962,
+      "grad_norm": 2.6708824634552,
+      "learning_rate": 9.785619345710244e-05,
+      "loss": 6.1553,
+      "step": 22500
+    },
+    {
+      "epoch": 0.9695232474813472,
+      "grad_norm": 3.4177470207214355,
+      "learning_rate": 9.777044119538653e-05,
+      "loss": 5.8901,
+      "step": 23000
+    },
+    {
+      "epoch": 0.9905998398178982,
+      "grad_norm": 2.3284451961517334,
+      "learning_rate": 9.768468893367063e-05,
+      "loss": 5.656,
+      "step": 23500
+    },
+    {
+      "epoch": 1.0116764321544494,
+      "grad_norm": 2.7531368732452393,
+      "learning_rate": 9.759893667195473e-05,
+      "loss": 5.4351,
+      "step": 24000
+    },
+    {
+      "epoch": 1.0327530244910004,
+      "grad_norm": 4.624035358428955,
+      "learning_rate": 9.751318441023883e-05,
+      "loss": 5.276,
+      "step": 24500
+    },
+    {
+      "epoch": 1.0538296168275514,
+      "grad_norm": 2.2471227645874023,
+      "learning_rate": 9.742743214852293e-05,
+      "loss": 5.1645,
+      "step": 25000
+    },
+    {
+      "epoch": 1.0749062091641024,
+      "grad_norm": 2.358665943145752,
+      "learning_rate": 9.734167988680703e-05,
+      "loss": 5.0724,
+      "step": 25500
+    },
+    {
+      "epoch": 1.0959828015006534,
+      "grad_norm": 1.5727806091308594,
+      "learning_rate": 9.725592762509111e-05,
+      "loss": 5.0013,
+      "step": 26000
+    },
+    {
+      "epoch": 1.1170593938372044,
+      "grad_norm": 1.8627222776412964,
+      "learning_rate": 9.717017536337522e-05,
+      "loss": 4.9376,
+      "step": 26500
+    },
+    {
+      "epoch": 1.1381359861737554,
+      "grad_norm": 3.190422773361206,
+      "learning_rate": 9.708442310165932e-05,
+      "loss": 4.8952,
+      "step": 27000
+    },
+    {
+      "epoch": 1.1592125785103065,
+      "grad_norm": 2.6108388900756836,
+      "learning_rate": 9.69986708399434e-05,
+      "loss": 4.8444,
+      "step": 27500
+    },
+    {
+      "epoch": 1.1802891708468575,
+      "grad_norm": 1.650323748588562,
+      "learning_rate": 9.69129185782275e-05,
+      "loss": 4.8094,
+      "step": 28000
+    },
+    {
+      "epoch": 1.2013657631834085,
+      "grad_norm": 1.5939751863479614,
+      "learning_rate": 9.68271663165116e-05,
+      "loss": 4.7615,
+      "step": 28500
+    },
+    {
+      "epoch": 1.2224423555199595,
+      "grad_norm": 2.2978341579437256,
+      "learning_rate": 9.67414140547957e-05,
+      "loss": 4.7266,
+      "step": 29000
+    },
+    {
+      "epoch": 1.2435189478565105,
+      "grad_norm": 2.4565980434417725,
+      "learning_rate": 9.66556617930798e-05,
+      "loss": 4.6866,
+      "step": 29500
+    },
+    {
+      "epoch": 1.2645955401930615,
+      "grad_norm": 2.503505229949951,
+      "learning_rate": 9.656990953136388e-05,
+      "loss": 4.6634,
+      "step": 30000
+    },
+    {
+      "epoch": 1.2645955401930615,
+      "eval_accuracy": 0.508355550610458,
+      "eval_loss": 4.281208038330078,
+      "eval_runtime": 171.2388,
+      "eval_samples_per_second": 704.899,
+      "eval_steps_per_second": 7.346,
+      "step": 30000
+    },
+    {
+      "epoch": 1.2856721325296125,
+      "grad_norm": 2.1511712074279785,
+      "learning_rate": 9.6484157269648e-05,
+      "loss": 4.6298,
+      "step": 30500
+    },
+    {
+      "epoch": 1.3067487248661636,
+      "grad_norm": 2.377535820007324,
+      "learning_rate": 9.639840500793209e-05,
+      "loss": 4.6097,
+      "step": 31000
+    },
+    {
+      "epoch": 1.3278253172027146,
+      "grad_norm": 2.6877024173736572,
+      "learning_rate": 9.631265274621618e-05,
+      "loss": 4.5824,
+      "step": 31500
+    },
+    {
+      "epoch": 1.3489019095392658,
+      "grad_norm": 2.923597574234009,
+      "learning_rate": 9.622690048450028e-05,
+      "loss": 4.5649,
+      "step": 32000
+    },
+    {
+      "epoch": 1.3699785018758166,
+      "grad_norm": 1.8175469636917114,
+      "learning_rate": 9.614114822278439e-05,
+      "loss": 4.5366,
+      "step": 32500
+    },
+    {
+      "epoch": 1.3910550942123678,
+      "grad_norm": 2.341097116470337,
+      "learning_rate": 9.605539596106847e-05,
+      "loss": 4.5192,
+      "step": 33000
+    },
+    {
+      "epoch": 1.4121316865489189,
+      "grad_norm": 1.964946985244751,
+      "learning_rate": 9.596964369935257e-05,
+      "loss": 4.4956,
+      "step": 33500
+    },
+    {
+      "epoch": 1.4332082788854699,
+      "grad_norm": 1.5739874839782715,
+      "learning_rate": 9.588389143763667e-05,
+      "loss": 4.4802,
+      "step": 34000
+    },
+    {
+      "epoch": 1.4542848712220209,
+      "grad_norm": 2.994678020477295,
+      "learning_rate": 9.579813917592077e-05,
+      "loss": 4.4574,
+      "step": 34500
+    },
+    {
+      "epoch": 1.475361463558572,
+      "grad_norm": 2.064154624938965,
+      "learning_rate": 9.571238691420487e-05,
+      "loss": 4.4394,
+      "step": 35000
+    },
+    {
+      "epoch": 1.496438055895123,
+      "grad_norm": 3.0208098888397217,
+      "learning_rate": 9.562663465248896e-05,
+      "loss": 4.4182,
+      "step": 35500
+    },
+    {
+      "epoch": 1.517514648231674,
+      "grad_norm": 2.008312702178955,
+      "learning_rate": 9.554088239077306e-05,
+      "loss": 4.3975,
+      "step": 36000
+    },
+    {
+      "epoch": 1.538591240568225,
+      "grad_norm": 1.5609052181243896,
+      "learning_rate": 9.545513012905716e-05,
+      "loss": 4.3821,
+      "step": 36500
+    },
+    {
+      "epoch": 1.559667832904776,
+      "grad_norm": 1.9999297857284546,
+      "learning_rate": 9.536937786734126e-05,
+      "loss": 4.3732,
+      "step": 37000
+    },
+    {
+      "epoch": 1.580744425241327,
+      "grad_norm": 2.307922124862671,
+      "learning_rate": 9.528362560562534e-05,
+      "loss": 4.3544,
+      "step": 37500
+    },
+    {
+      "epoch": 1.601821017577878,
+      "grad_norm": 3.480792284011841,
+      "learning_rate": 9.519787334390946e-05,
+      "loss": 4.339,
+      "step": 38000
+    },
+    {
+      "epoch": 1.622897609914429,
+      "grad_norm": 2.9850962162017822,
+      "learning_rate": 9.511212108219355e-05,
+      "loss": 4.3215,
+      "step": 38500
+    },
+    {
+      "epoch": 1.64397420225098,
+      "grad_norm": 1.5874985456466675,
+      "learning_rate": 9.502636882047764e-05,
+      "loss": 4.3074,
+      "step": 39000
+    },
+    {
+      "epoch": 1.6650507945875312,
+      "grad_norm": 3.2197210788726807,
+      "learning_rate": 9.494061655876174e-05,
+      "loss": 4.2944,
+      "step": 39500
+    },
+    {
+      "epoch": 1.686127386924082,
+      "grad_norm": 2.4538028240203857,
+      "learning_rate": 9.485486429704585e-05,
+      "loss": 4.2844,
+      "step": 40000
+    },
+    {
+      "epoch": 1.686127386924082,
+      "eval_accuracy": 0.5556485703559956,
+      "eval_loss": 3.9252429008483887,
+      "eval_runtime": 171.2404,
+      "eval_samples_per_second": 704.892,
+      "eval_steps_per_second": 7.346,
+      "step": 40000
+    },
+    {
+      "epoch": 1.7072039792606333,
+      "grad_norm": 2.2828330993652344,
+      "learning_rate": 9.476911203532993e-05,
+      "loss": 4.2673,
+      "step": 40500
+    },
+    {
+      "epoch": 1.728280571597184,
+      "grad_norm": 1.7881807088851929,
+      "learning_rate": 9.468335977361403e-05,
+      "loss": 4.2599,
+      "step": 41000
+    },
+    {
+      "epoch": 1.7493571639337353,
+      "grad_norm": 1.8580396175384521,
+      "learning_rate": 9.459760751189813e-05,
+      "loss": 4.246,
+      "step": 41500
+    },
+    {
+      "epoch": 1.770433756270286,
+      "grad_norm": 1.7360947132110596,
+      "learning_rate": 9.451185525018223e-05,
+      "loss": 4.2301,
+      "step": 42000
+    },
+    {
+      "epoch": 1.7915103486068373,
+      "grad_norm": 1.4446935653686523,
+      "learning_rate": 9.442610298846633e-05,
+      "loss": 4.2186,
+      "step": 42500
+    },
+    {
+      "epoch": 1.8125869409433881,
+      "grad_norm": 2.3633315563201904,
+      "learning_rate": 9.434035072675043e-05,
+      "loss": 4.2047,
+      "step": 43000
+    },
+    {
+      "epoch": 1.8336635332799394,
+      "grad_norm": 2.330095052719116,
+      "learning_rate": 9.425459846503451e-05,
+      "loss": 4.1908,
+      "step": 43500
+    },
+    {
+      "epoch": 1.8547401256164904,
+      "grad_norm": 3.68125057220459,
+      "learning_rate": 9.416884620331862e-05,
+      "loss": 4.1813,
+      "step": 44000
+    },
+    {
+      "epoch": 1.8758167179530414,
+      "grad_norm": 3.25962495803833,
+      "learning_rate": 9.408309394160272e-05,
+      "loss": 4.1746,
+      "step": 44500
+    },
+    {
+      "epoch": 1.8968933102895924,
+      "grad_norm": 3.0640392303466797,
+      "learning_rate": 9.39973416798868e-05,
+      "loss": 4.1537,
+      "step": 45000
+    },
+    {
+      "epoch": 1.9179699026261434,
+      "grad_norm": 3.2719156742095947,
+      "learning_rate": 9.39115894181709e-05,
+      "loss": 4.1476,
+      "step": 45500
+    },
+    {
+      "epoch": 1.9390464949626944,
+      "grad_norm": 2.7434871196746826,
+      "learning_rate": 9.382583715645502e-05,
+      "loss": 4.1311,
+      "step": 46000
+    },
+    {
+      "epoch": 1.9601230872992454,
+      "grad_norm": 1.7805469036102295,
+      "learning_rate": 9.37400848947391e-05,
+      "loss": 4.1221,
+      "step": 46500
+    },
+    {
+      "epoch": 1.9811996796357965,
+      "grad_norm": 2.7733612060546875,
+      "learning_rate": 9.36543326330232e-05,
+      "loss": 4.1176,
+      "step": 47000
+    },
+    {
+      "epoch": 2.0022762719723475,
+      "grad_norm": 1.8867709636688232,
+      "learning_rate": 9.35685803713073e-05,
+      "loss": 4.1128,
+      "step": 47500
+    },
+    {
+      "epoch": 2.0233528643088987,
+      "grad_norm": 2.148491382598877,
+      "learning_rate": 9.34828281095914e-05,
+      "loss": 4.0929,
+      "step": 48000
+    },
+    {
+      "epoch": 2.0444294566454495,
+      "grad_norm": 2.8966031074523926,
+      "learning_rate": 9.339707584787549e-05,
+      "loss": 4.0843,
+      "step": 48500
+    },
+    {
+      "epoch": 2.0655060489820007,
+      "grad_norm": 2.0389890670776367,
+      "learning_rate": 9.331132358615959e-05,
+      "loss": 4.0713,
+      "step": 49000
+    },
+    {
+      "epoch": 2.0865826413185515,
+      "grad_norm": 2.4401726722717285,
+      "learning_rate": 9.322557132444369e-05,
+      "loss": 4.0618,
+      "step": 49500
+    },
+    {
+      "epoch": 2.1076592336551028,
+      "grad_norm": 2.2224786281585693,
+      "learning_rate": 9.313981906272779e-05,
+      "loss": 4.0555,
+      "step": 50000
+    },
+    {
+      "epoch": 2.1076592336551028,
+      "eval_accuracy": 0.5844534983499117,
+      "eval_loss": 3.7209959030151367,
+      "eval_runtime": 170.9559,
+      "eval_samples_per_second": 706.065,
+      "eval_steps_per_second": 7.359,
+      "step": 50000
+    },
+    {
+      "epoch": 2.1287358259916536,
+      "grad_norm": 2.3282253742218018,
+      "learning_rate": 9.305406680101189e-05,
+      "loss": 4.051,
+      "step": 50500
+    },
+    {
+      "epoch": 2.149812418328205,
+      "grad_norm": 1.9450957775115967,
+      "learning_rate": 9.296831453929597e-05,
+      "loss": 4.0374,
+      "step": 51000
+    },
+    {
+      "epoch": 2.1708890106647556,
+      "grad_norm": 1.3277090787887573,
+      "learning_rate": 9.288256227758008e-05,
+      "loss": 4.0259,
+      "step": 51500
+    },
+    {
+      "epoch": 2.191965603001307,
+      "grad_norm": 2.2632498741149902,
+      "learning_rate": 9.279681001586418e-05,
+      "loss": 4.0189,
+      "step": 52000
+    },
+    {
+      "epoch": 2.2130421953378576,
+      "grad_norm": 3.8596460819244385,
+      "learning_rate": 9.271105775414827e-05,
+      "loss": 4.0082,
+      "step": 52500
+    },
+    {
+      "epoch": 2.234118787674409,
+      "grad_norm": 2.436032295227051,
+      "learning_rate": 9.262530549243236e-05,
+      "loss": 4.0021,
+      "step": 53000
+    },
+    {
+      "epoch": 2.2551953800109596,
+      "grad_norm": 1.6290899515151978,
+      "learning_rate": 9.253955323071646e-05,
+      "loss": 3.9946,
+      "step": 53500
+    },
+    {
+      "epoch": 2.276271972347511,
+      "grad_norm": 1.7445964813232422,
+      "learning_rate": 9.245380096900056e-05,
+      "loss": 3.9917,
+      "step": 54000
+    },
+    {
+      "epoch": 2.2973485646840617,
+      "grad_norm": 2.2918787002563477,
+      "learning_rate": 9.236804870728466e-05,
+      "loss": 3.9768,
+      "step": 54500
+    },
+    {
+      "epoch": 2.318425157020613,
+      "grad_norm": 2.0511977672576904,
+      "learning_rate": 9.228229644556876e-05,
+      "loss": 3.9725,
+      "step": 55000
+    },
+    {
+      "epoch": 2.339501749357164,
+      "grad_norm": 1.4360889196395874,
+      "learning_rate": 9.219654418385286e-05,
+      "loss": 3.9606,
+      "step": 55500
+    },
+    {
+      "epoch": 2.360578341693715,
+      "grad_norm": 2.165125846862793,
+      "learning_rate": 9.211079192213695e-05,
+      "loss": 3.9631,
+      "step": 56000
+    },
+    {
+      "epoch": 2.381654934030266,
+      "grad_norm": 1.3233758211135864,
+      "learning_rate": 9.202503966042105e-05,
+      "loss": 3.9501,
+      "step": 56500
+    },
+    {
+      "epoch": 2.402731526366817,
+      "grad_norm": 1.6704837083816528,
+      "learning_rate": 9.193928739870514e-05,
+      "loss": 3.9416,
+      "step": 57000
+    },
+    {
+      "epoch": 2.423808118703368,
+      "grad_norm": 1.8866082429885864,
+      "learning_rate": 9.185353513698925e-05,
+      "loss": 3.9348,
+      "step": 57500
+    },
+    {
+      "epoch": 2.444884711039919,
+      "grad_norm": 2.8515818119049072,
+      "learning_rate": 9.176778287527335e-05,
+      "loss": 3.9291,
+      "step": 58000
+    },
+    {
+      "epoch": 2.4659613033764702,
+      "grad_norm": 1.25554621219635,
+      "learning_rate": 9.168203061355743e-05,
+      "loss": 3.9252,
+      "step": 58500
+    },
+    {
+      "epoch": 2.487037895713021,
+      "grad_norm": 1.9014805555343628,
+      "learning_rate": 9.159627835184153e-05,
+      "loss": 3.9194,
+      "step": 59000
+    },
+    {
+      "epoch": 2.5081144880495723,
+      "grad_norm": 4.307132720947266,
+      "learning_rate": 9.151052609012564e-05,
+      "loss": 3.9164,
+      "step": 59500
+    },
+    {
+      "epoch": 2.529191080386123,
+      "grad_norm": 1.2848464250564575,
+      "learning_rate": 9.142477382840973e-05,
+      "loss": 3.9068,
+      "step": 60000
+    },
+    {
+      "epoch": 2.529191080386123,
+      "eval_accuracy": 0.6047395967416954,
+      "eval_loss": 3.5822408199310303,
+      "eval_runtime": 172.3346,
+      "eval_samples_per_second": 700.417,
+      "eval_steps_per_second": 7.3,
+      "step": 60000
+    },
+    {
+      "epoch": 2.5502676727226743,
+      "grad_norm": 2.778567314147949,
+      "learning_rate": 9.133902156669383e-05,
+      "loss": 3.8979,
+      "step": 60500
+    },
+    {
+      "epoch": 2.571344265059225,
+      "grad_norm": 2.41713547706604,
+      "learning_rate": 9.125326930497792e-05,
+      "loss": 3.8946,
+      "step": 61000
+    },
+    {
+      "epoch": 2.5924208573957763,
+      "grad_norm": 1.7044235467910767,
+      "learning_rate": 9.116751704326202e-05,
+      "loss": 3.8859,
+      "step": 61500
+    },
+    {
+      "epoch": 2.613497449732327,
+      "grad_norm": 1.8987773656845093,
+      "learning_rate": 9.108176478154612e-05,
+      "loss": 3.8776,
+      "step": 62000
+    },
+    {
+      "epoch": 2.6345740420688784,
+      "grad_norm": 2.9212422370910645,
+      "learning_rate": 9.09960125198302e-05,
+      "loss": 3.8808,
+      "step": 62500
+    },
+    {
+      "epoch": 2.655650634405429,
+      "grad_norm": 1.852678894996643,
+      "learning_rate": 9.091026025811432e-05,
+      "loss": 3.87,
+      "step": 63000
+    },
+    {
+      "epoch": 2.6767272267419804,
+      "grad_norm": 3.1264853477478027,
+      "learning_rate": 9.082450799639842e-05,
+      "loss": 3.8672,
+      "step": 63500
+    },
+    {
+      "epoch": 2.6978038190785316,
+      "grad_norm": 1.9994491338729858,
+      "learning_rate": 9.07387557346825e-05,
+      "loss": 3.8645,
+      "step": 64000
+    },
+    {
+      "epoch": 2.7188804114150824,
+      "grad_norm": 2.4018962383270264,
+      "learning_rate": 9.06530034729666e-05,
+      "loss": 3.8567,
+      "step": 64500
+    },
+    {
+      "epoch": 2.739957003751633,
+      "grad_norm": 1.900780200958252,
+      "learning_rate": 9.05672512112507e-05,
+      "loss": 3.8514,
+      "step": 65000
+    },
+    {
+      "epoch": 2.7610335960881844,
+      "grad_norm": 1.7594151496887207,
+      "learning_rate": 9.04814989495348e-05,
+      "loss": 3.8467,
+      "step": 65500
+    },
+    {
+      "epoch": 2.7821101884247357,
+      "grad_norm": 1.7174185514450073,
+      "learning_rate": 9.03957466878189e-05,
+      "loss": 3.8449,
+      "step": 66000
+    },
+    {
+      "epoch": 2.8031867807612865,
+      "grad_norm": 2.6180167198181152,
+      "learning_rate": 9.030999442610299e-05,
+      "loss": 3.8373,
+      "step": 66500
+    },
+    {
+      "epoch": 2.8242633730978377,
+      "grad_norm": 3.5789287090301514,
+      "learning_rate": 9.022424216438709e-05,
+      "loss": 3.8285,
+      "step": 67000
+    },
+    {
+      "epoch": 2.8453399654343885,
+      "grad_norm": 2.2866082191467285,
+      "learning_rate": 9.013848990267119e-05,
+      "loss": 3.825,
+      "step": 67500
+    },
+    {
+      "epoch": 2.8664165577709397,
+      "grad_norm": 2.7827377319335938,
+      "learning_rate": 9.005273764095529e-05,
+      "loss": 3.8197,
+      "step": 68000
+    },
+    {
+      "epoch": 2.8874931501074905,
+      "grad_norm": 1.645926594734192,
+      "learning_rate": 8.996698537923937e-05,
+      "loss": 3.823,
+      "step": 68500
+    },
+    {
+      "epoch": 2.9085697424440418,
+      "grad_norm": 1.6839686632156372,
+      "learning_rate": 8.988123311752348e-05,
+      "loss": 3.8123,
+      "step": 69000
+    },
+    {
+      "epoch": 2.9296463347805926,
+      "grad_norm": 2.4849064350128174,
+      "learning_rate": 8.979548085580758e-05,
+      "loss": 3.8122,
+      "step": 69500
+    },
+    {
+      "epoch": 2.950722927117144,
+      "grad_norm": 2.5568675994873047,
+      "learning_rate": 8.970972859409167e-05,
+      "loss": 3.8047,
+      "step": 70000
+    },
+    {
+      "epoch": 2.950722927117144,
+      "eval_accuracy": 0.6182514680913767,
+      "eval_loss": 3.4939889907836914,
+      "eval_runtime": 169.7947,
+      "eval_samples_per_second": 710.894,
+      "eval_steps_per_second": 7.409,
+      "step": 70000
+    },
+    {
+      "epoch": 2.9717995194536946,
+      "grad_norm": 1.167945146560669,
+      "learning_rate": 8.962397633237576e-05,
+      "loss": 3.803,
+      "step": 70500
+    },
+    {
+      "epoch": 2.992876111790246,
+      "grad_norm": 2.347172260284424,
+      "learning_rate": 8.953822407065988e-05,
+      "loss": 3.7947,
+      "step": 71000
+    },
+    {
+      "epoch": 3.0139527041267966,
+      "grad_norm": 1.7915819883346558,
+      "learning_rate": 8.945247180894396e-05,
+      "loss": 3.7919,
+      "step": 71500
+    },
+    {
+      "epoch": 3.035029296463348,
+      "grad_norm": 3.2532317638397217,
+      "learning_rate": 8.936671954722806e-05,
+      "loss": 3.7876,
+      "step": 72000
+    },
+    {
+      "epoch": 3.0561058887998986,
+      "grad_norm": 2.8248579502105713,
+      "learning_rate": 8.928096728551216e-05,
+      "loss": 3.7884,
+      "step": 72500
+    },
+    {
+      "epoch": 3.07718248113645,
+      "grad_norm": 2.8527846336364746,
+      "learning_rate": 8.919521502379626e-05,
+      "loss": 3.7813,
+      "step": 73000
+    },
+    {
+      "epoch": 3.0982590734730007,
+      "grad_norm": 1.9902697801589966,
+      "learning_rate": 8.910946276208035e-05,
+      "loss": 3.7808,
+      "step": 73500
+    },
+    {
+      "epoch": 3.119335665809552,
+      "grad_norm": 1.1623148918151855,
+      "learning_rate": 8.902371050036445e-05,
+      "loss": 3.7697,
+      "step": 74000
+    },
+    {
+      "epoch": 3.140412258146103,
+      "grad_norm": 2.2464077472686768,
+      "learning_rate": 8.893795823864855e-05,
+      "loss": 3.7738,
+      "step": 74500
+    },
+    {
+      "epoch": 3.161488850482654,
+      "grad_norm": 1.8677114248275757,
+      "learning_rate": 8.885220597693265e-05,
+      "loss": 3.768,
+      "step": 75000
+    },
+    {
+      "epoch": 3.182565442819205,
+      "grad_norm": 1.455930233001709,
+      "learning_rate": 8.876645371521675e-05,
+      "loss": 3.7618,
+      "step": 75500
+    },
+    {
+      "epoch": 3.203642035155756,
+      "grad_norm": 2.1140687465667725,
+      "learning_rate": 8.868070145350083e-05,
+      "loss": 3.7596,
+      "step": 76000
+    },
+    {
+      "epoch": 3.224718627492307,
+      "grad_norm": 2.3617889881134033,
+      "learning_rate": 8.859494919178493e-05,
+      "loss": 3.7582,
+      "step": 76500
+    },
+    {
+      "epoch": 3.245795219828858,
+      "grad_norm": 2.381812334060669,
+      "learning_rate": 8.850919693006904e-05,
+      "loss": 3.7506,
+      "step": 77000
+    },
+    {
+      "epoch": 3.2668718121654092,
+      "grad_norm": 1.3791255950927734,
+      "learning_rate": 8.842344466835313e-05,
+      "loss": 3.7484,
+      "step": 77500
+    },
+    {
+      "epoch": 3.28794840450196,
+      "grad_norm": 1.5806275606155396,
+      "learning_rate": 8.833769240663723e-05,
+      "loss": 3.7436,
+      "step": 78000
+    },
+    {
+      "epoch": 3.3090249968385113,
+      "grad_norm": 2.932819366455078,
+      "learning_rate": 8.825194014492132e-05,
+      "loss": 3.7418,
+      "step": 78500
+    },
+    {
+      "epoch": 3.330101589175062,
+      "grad_norm": 2.4233949184417725,
+      "learning_rate": 8.816618788320542e-05,
+      "loss": 3.74,
+      "step": 79000
+    },
+    {
+      "epoch": 3.3511781815116133,
+      "grad_norm": 2.1907780170440674,
+      "learning_rate": 8.808043562148952e-05,
+      "loss": 3.7373,
+      "step": 79500
+    },
+    {
+      "epoch": 3.372254773848164,
+      "grad_norm": 1.5484669208526611,
+      "learning_rate": 8.799468335977362e-05,
+      "loss": 3.734,
+      "step": 80000
+    },
+    {
+      "epoch": 3.372254773848164,
+      "eval_accuracy": 0.6280736465053394,
+      "eval_loss": 3.430922269821167,
+      "eval_runtime": 170.4001,
+      "eval_samples_per_second": 708.368,
+      "eval_steps_per_second": 7.383,
+      "step": 80000
+    },
+    {
+      "epoch": 3.3933313661847153,
+      "grad_norm": 1.6316457986831665,
+      "learning_rate": 8.790893109805772e-05,
+      "loss": 3.732,
+      "step": 80500
+    },
+    {
+      "epoch": 3.414407958521266,
+      "grad_norm": 2.568800449371338,
+      "learning_rate": 8.782317883634182e-05,
+      "loss": 3.7271,
+      "step": 81000
+    },
+    {
+      "epoch": 3.4354845508578173,
+      "grad_norm": 2.6804869174957275,
+      "learning_rate": 8.773742657462591e-05,
+      "loss": 3.7238,
+      "step": 81500
+    },
+    {
+      "epoch": 3.456561143194368,
+      "grad_norm": 2.3391616344451904,
+      "learning_rate": 8.765167431291e-05,
+      "loss": 3.7216,
+      "step": 82000
+    },
+    {
+      "epoch": 3.4776377355309194,
+      "grad_norm": 1.7324694395065308,
+      "learning_rate": 8.756592205119411e-05,
+      "loss": 3.7173,
+      "step": 82500
+    },
+    {
+      "epoch": 3.4987143278674706,
+      "grad_norm": 2.6797163486480713,
+      "learning_rate": 8.748016978947821e-05,
+      "loss": 3.7166,
+      "step": 83000
+    },
+    {
+      "epoch": 3.5197909202040214,
+      "grad_norm": 1.618472933769226,
+      "learning_rate": 8.73944175277623e-05,
+      "loss": 3.7135,
+      "step": 83500
+    },
+    {
+      "epoch": 3.540867512540572,
+      "grad_norm": 2.0404040813446045,
+      "learning_rate": 8.730866526604639e-05,
+      "loss": 3.7136,
+      "step": 84000
+    },
+    {
+      "epoch": 3.5619441048771234,
+      "grad_norm": 2.189887523651123,
+      "learning_rate": 8.72229130043305e-05,
+      "loss": 3.7112,
+      "step": 84500
+    },
+    {
+      "epoch": 3.5830206972136747,
+      "grad_norm": 1.735327124595642,
+      "learning_rate": 8.713716074261459e-05,
+      "loss": 3.7081,
+      "step": 85000
+    },
+    {
+      "epoch": 3.6040972895502255,
+      "grad_norm": 2.6286449432373047,
+      "learning_rate": 8.705140848089869e-05,
+      "loss": 3.7041,
+      "step": 85500
+    },
+    {
+      "epoch": 3.6251738818867767,
+      "grad_norm": 2.385605812072754,
+      "learning_rate": 8.696565621918278e-05,
+      "loss": 3.6968,
+      "step": 86000
+    },
+    {
+      "epoch": 3.6462504742233275,
+      "grad_norm": 2.7943646907806396,
+      "learning_rate": 8.687990395746688e-05,
+      "loss": 3.6997,
+      "step": 86500
+    },
+    {
+      "epoch": 3.6673270665598787,
+      "grad_norm": 2.086812973022461,
+      "learning_rate": 8.679415169575098e-05,
+      "loss": 3.6949,
+      "step": 87000
+    },
+    {
+      "epoch": 3.6884036588964295,
+      "grad_norm": 2.6318092346191406,
+      "learning_rate": 8.670839943403508e-05,
+      "loss": 3.6927,
+      "step": 87500
+    },
+    {
+      "epoch": 3.7094802512329808,
+      "grad_norm": 1.3619747161865234,
+      "learning_rate": 8.662264717231916e-05,
+      "loss": 3.6921,
+      "step": 88000
+    },
+    {
+      "epoch": 3.7305568435695315,
+      "grad_norm": 1.7893599271774292,
+      "learning_rate": 8.653689491060328e-05,
+      "loss": 3.6882,
+      "step": 88500
+    },
+    {
+      "epoch": 3.751633435906083,
+      "grad_norm": 1.9847978353500366,
+      "learning_rate": 8.645114264888737e-05,
+      "loss": 3.6873,
+      "step": 89000
+    },
+    {
+      "epoch": 3.7727100282426336,
+      "grad_norm": 3.16633677482605,
+      "learning_rate": 8.636539038717146e-05,
+      "loss": 3.6793,
+      "step": 89500
+    },
+    {
+      "epoch": 3.793786620579185,
+      "grad_norm": 1.7003432512283325,
+      "learning_rate": 8.627963812545556e-05,
+      "loss": 3.678,
+      "step": 90000
+    },
+    {
+      "epoch": 3.793786620579185,
+      "eval_accuracy": 0.6365212027658343,
+      "eval_loss": 3.3726961612701416,
+      "eval_runtime": 169.4646,
+      "eval_samples_per_second": 712.279,
+      "eval_steps_per_second": 7.423,
+      "step": 90000
+    },
+    {
+      "epoch": 3.8148632129157356,
+      "grad_norm": 2.740455150604248,
+      "learning_rate": 8.619388586373967e-05,
+      "loss": 3.6803,
+      "step": 90500
+    },
+    {
+      "epoch": 3.835939805252287,
+      "grad_norm": 1.8273558616638184,
+      "learning_rate": 8.610813360202375e-05,
+      "loss": 3.6722,
+      "step": 91000
+    },
+    {
+      "epoch": 3.857016397588838,
+      "grad_norm": 2.465141773223877,
+      "learning_rate": 8.602238134030785e-05,
+      "loss": 3.6741,
+      "step": 91500
+    },
+    {
+      "epoch": 3.878092989925389,
+      "grad_norm": 2.77006196975708,
+      "learning_rate": 8.593662907859195e-05,
+      "loss": 3.666,
+      "step": 92000
+    },
+    {
+      "epoch": 3.8991695822619397,
+      "grad_norm": 1.5233337879180908,
+      "learning_rate": 8.585087681687605e-05,
+      "loss": 3.668,
+      "step": 92500
+    },
+    {
+      "epoch": 3.920246174598491,
+      "grad_norm": 2.099942445755005,
+      "learning_rate": 8.576512455516015e-05,
+      "loss": 3.6682,
+      "step": 93000
+    },
+    {
+      "epoch": 3.941322766935042,
+      "grad_norm": 1.8762391805648804,
+      "learning_rate": 8.567937229344425e-05,
+      "loss": 3.6639,
+      "step": 93500
+    },
+    {
+      "epoch": 3.962399359271593,
+      "grad_norm": 3.186664581298828,
+      "learning_rate": 8.559362003172834e-05,
+      "loss": 3.6612,
+      "step": 94000
+    },
+    {
+      "epoch": 3.983475951608144,
+      "grad_norm": 2.8411457538604736,
+      "learning_rate": 8.550786777001244e-05,
+      "loss": 3.6609,
+      "step": 94500
+    },
+    {
+      "epoch": 4.004552543944695,
+      "grad_norm": 2.1656606197357178,
+      "learning_rate": 8.542211550829654e-05,
+      "loss": 3.6558,
+      "step": 95000
+    },
+    {
+      "epoch": 4.025629136281246,
+      "grad_norm": 2.647808074951172,
+      "learning_rate": 8.533636324658063e-05,
+      "loss": 3.6475,
+      "step": 95500
+    },
+    {
+      "epoch": 4.046705728617797,
+      "grad_norm": 1.7667932510375977,
+      "learning_rate": 8.525061098486474e-05,
+      "loss": 3.6509,
+      "step": 96000
+    },
+    {
+      "epoch": 4.067782320954348,
+      "grad_norm": 1.8241006135940552,
+      "learning_rate": 8.516485872314884e-05,
+      "loss": 3.6447,
+      "step": 96500
+    },
+    {
+      "epoch": 4.088858913290899,
+      "grad_norm": 2.6353302001953125,
+      "learning_rate": 8.507910646143292e-05,
+      "loss": 3.6443,
+      "step": 97000
+    },
+    {
+      "epoch": 4.10993550562745,
+      "grad_norm": 1.9048806428909302,
+      "learning_rate": 8.499335419971702e-05,
+      "loss": 3.6415,
+      "step": 97500
+    },
+    {
+      "epoch": 4.1310120979640015,
+      "grad_norm": 2.452421188354492,
+      "learning_rate": 8.490760193800112e-05,
+      "loss": 3.6404,
+      "step": 98000
+    },
+    {
+      "epoch": 4.152088690300552,
+      "grad_norm": 2.238510847091675,
+      "learning_rate": 8.482184967628522e-05,
+      "loss": 3.6362,
+      "step": 98500
+    },
+    {
+      "epoch": 4.173165282637103,
+      "grad_norm": 1.6972004175186157,
+      "learning_rate": 8.473609741456931e-05,
+      "loss": 3.6422,
+      "step": 99000
+    },
+    {
+      "epoch": 4.194241874973654,
+      "grad_norm": 2.081347942352295,
+      "learning_rate": 8.46503451528534e-05,
+      "loss": 3.6397,
+      "step": 99500
+    },
+    {
+      "epoch": 4.2153184673102055,
+      "grad_norm": 2.1786227226257324,
+      "learning_rate": 8.456459289113751e-05,
+      "loss": 3.6347,
+      "step": 100000
+    },
+    {
+      "epoch": 4.2153184673102055,
+      "eval_accuracy": 0.6425543223286596,
+      "eval_loss": 3.33632230758667,
+      "eval_runtime": 171.1302,
+      "eval_samples_per_second": 705.346,
+      "eval_steps_per_second": 7.351,
+      "step": 100000
+    },
+    {
+      "epoch": 4.236395059646756,
+      "grad_norm": 1.7079287767410278,
+      "learning_rate": 8.447884062942161e-05,
+      "loss": 3.6343,
+      "step": 100500
+    },
+    {
+      "epoch": 4.257471651983307,
+      "grad_norm": 2.7177727222442627,
+      "learning_rate": 8.43930883677057e-05,
+      "loss": 3.6329,
+      "step": 101000
+    },
+    {
+      "epoch": 4.278548244319858,
+      "grad_norm": 3.191612958908081,
+      "learning_rate": 8.430733610598979e-05,
+      "loss": 3.6288,
+      "step": 101500
+    },
+    {
+      "epoch": 4.29962483665641,
+      "grad_norm": 2.5266671180725098,
+      "learning_rate": 8.42215838442739e-05,
+      "loss": 3.6263,
+      "step": 102000
+    },
+    {
+      "epoch": 4.32070142899296,
+      "grad_norm": 2.1244566440582275,
+      "learning_rate": 8.413583158255799e-05,
+      "loss": 3.6265,
+      "step": 102500
+    },
+    {
+      "epoch": 4.341778021329511,
+      "grad_norm": 1.8111193180084229,
+      "learning_rate": 8.405007932084209e-05,
+      "loss": 3.6219,
+      "step": 103000
+    },
+    {
+      "epoch": 4.362854613666062,
+      "grad_norm": 1.7496929168701172,
+      "learning_rate": 8.396432705912619e-05,
+      "loss": 3.6213,
+      "step": 103500
+    },
+    {
+      "epoch": 4.383931206002614,
+      "grad_norm": 1.4520421028137207,
+      "learning_rate": 8.387857479741028e-05,
+      "loss": 3.6232,
+      "step": 104000
+    },
+    {
+      "epoch": 4.405007798339165,
+      "grad_norm": 1.2445356845855713,
+      "learning_rate": 8.379282253569438e-05,
+      "loss": 3.6196,
+      "step": 104500
+    },
+    {
+      "epoch": 4.426084390675715,
+      "grad_norm": 1.2059781551361084,
+      "learning_rate": 8.370707027397848e-05,
+      "loss": 3.617,
+      "step": 105000
+    },
+    {
+      "epoch": 4.4471609830122665,
+      "grad_norm": 1.9422649145126343,
+      "learning_rate": 8.362131801226258e-05,
+      "loss": 3.6145,
+      "step": 105500
+    },
+    {
+      "epoch": 4.468237575348818,
+      "grad_norm": 1.6592026948928833,
+      "learning_rate": 8.353556575054668e-05,
+      "loss": 3.6121,
+      "step": 106000
+    },
+    {
+      "epoch": 4.489314167685369,
+      "grad_norm": 1.3929331302642822,
+      "learning_rate": 8.344981348883077e-05,
+      "loss": 3.6139,
+      "step": 106500
+    },
+    {
+      "epoch": 4.510390760021919,
+      "grad_norm": 1.597322702407837,
+      "learning_rate": 8.336406122711486e-05,
+      "loss": 3.6101,
+      "step": 107000
+    },
+    {
+      "epoch": 4.5314673523584705,
+      "grad_norm": 2.3094332218170166,
+      "learning_rate": 8.327830896539897e-05,
+      "loss": 3.6143,
+      "step": 107500
+    },
+    {
+      "epoch": 4.552543944695022,
+      "grad_norm": 2.338700294494629,
+      "learning_rate": 8.319255670368307e-05,
+      "loss": 3.6072,
+      "step": 108000
+    },
+    {
+      "epoch": 4.573620537031573,
+      "grad_norm": 2.713604211807251,
+      "learning_rate": 8.310680444196715e-05,
+      "loss": 3.6079,
+      "step": 108500
+    },
+    {
+      "epoch": 4.594697129368123,
+      "grad_norm": 1.7006219625473022,
+      "learning_rate": 8.302105218025125e-05,
+      "loss": 3.5995,
+      "step": 109000
+    },
+    {
+      "epoch": 4.615773721704675,
+      "grad_norm": 2.7594213485717773,
+      "learning_rate": 8.293529991853536e-05,
+      "loss": 3.6061,
+      "step": 109500
+    },
+    {
+      "epoch": 4.636850314041226,
+      "grad_norm": 2.5298995971679688,
+      "learning_rate": 8.284954765681945e-05,
+      "loss": 3.6013,
+      "step": 110000
+    },
+    {
+      "epoch": 4.636850314041226,
+      "eval_accuracy": 0.647126303253824,
+      "eval_loss": 3.30576229095459,
+      "eval_runtime": 171.64,
+      "eval_samples_per_second": 703.251,
+      "eval_steps_per_second": 7.329,
+      "step": 110000
+    },
+    {
+      "epoch": 4.657926906377777,
+      "grad_norm": 2.4554312229156494,
+      "learning_rate": 8.276379539510355e-05,
+      "loss": 3.6003,
+      "step": 110500
+    },
+    {
+      "epoch": 4.679003498714328,
+      "grad_norm": 2.5871071815490723,
+      "learning_rate": 8.267804313338765e-05,
+      "loss": 3.6014,
+      "step": 111000
+    },
+    {
+      "epoch": 4.700080091050879,
+      "grad_norm": 1.8674741983413696,
+      "learning_rate": 8.259229087167174e-05,
+      "loss": 3.5968,
+      "step": 111500
+    },
+    {
+      "epoch": 4.72115668338743,
+      "grad_norm": 1.8471206426620483,
+      "learning_rate": 8.250653860995584e-05,
+      "loss": 3.5942,
+      "step": 112000
+    },
+    {
+      "epoch": 4.742233275723981,
+      "grad_norm": 2.081817388534546,
+      "learning_rate": 8.242078634823994e-05,
+      "loss": 3.5931,
+      "step": 112500
+    },
+    {
+      "epoch": 4.763309868060532,
+      "grad_norm": 2.1382105350494385,
+      "learning_rate": 8.233503408652403e-05,
+      "loss": 3.5937,
+      "step": 113000
+    },
+    {
+      "epoch": 4.784386460397083,
+      "grad_norm": 3.1155643463134766,
+      "learning_rate": 8.224928182480814e-05,
+      "loss": 3.5955,
+      "step": 113500
+    },
+    {
+      "epoch": 4.805463052733634,
+      "grad_norm": 3.0467007160186768,
+      "learning_rate": 8.216352956309224e-05,
+      "loss": 3.5854,
+      "step": 114000
+    },
+    {
+      "epoch": 4.826539645070185,
+      "grad_norm": 2.7216248512268066,
+      "learning_rate": 8.207777730137632e-05,
+      "loss": 3.587,
+      "step": 114500
+    },
+    {
+      "epoch": 4.847616237406736,
+      "grad_norm": 2.245497226715088,
+      "learning_rate": 8.199202503966042e-05,
+      "loss": 3.5867,
+      "step": 115000
+    },
+    {
+      "epoch": 4.868692829743287,
+      "grad_norm": 2.45318341255188,
+      "learning_rate": 8.190627277794453e-05,
+      "loss": 3.5863,
+      "step": 115500
+    },
+    {
+      "epoch": 4.889769422079838,
+      "grad_norm": 2.4969282150268555,
+      "learning_rate": 8.182052051622862e-05,
+      "loss": 3.582,
+      "step": 116000
+    },
+    {
+      "epoch": 4.910846014416389,
+      "grad_norm": 1.959963321685791,
+      "learning_rate": 8.173476825451271e-05,
+      "loss": 3.5827,
+      "step": 116500
+    },
+    {
+      "epoch": 4.9319226067529405,
+      "grad_norm": 1.5817019939422607,
+      "learning_rate": 8.164901599279681e-05,
+      "loss": 3.5821,
+      "step": 117000
+    },
+    {
+      "epoch": 4.952999199089491,
+      "grad_norm": 2.812964916229248,
+      "learning_rate": 8.156326373108091e-05,
+      "loss": 3.5827,
+      "step": 117500
+    },
+    {
+      "epoch": 4.974075791426042,
+      "grad_norm": 2.282419443130493,
+      "learning_rate": 8.147751146936501e-05,
+      "loss": 3.5793,
+      "step": 118000
+    },
+    {
+      "epoch": 4.995152383762593,
+      "grad_norm": 1.9533885717391968,
+      "learning_rate": 8.139175920764911e-05,
+      "loss": 3.5741,
+      "step": 118500
+    },
+    {
+      "epoch": 5.0162289760991445,
+      "grad_norm": 1.9340102672576904,
+      "learning_rate": 8.13060069459332e-05,
+      "loss": 3.5763,
+      "step": 119000
+    },
+    {
+      "epoch": 5.037305568435695,
+      "grad_norm": 2.75980806350708,
+      "learning_rate": 8.12202546842173e-05,
+      "loss": 3.5705,
+      "step": 119500
+    },
+    {
+      "epoch": 5.058382160772246,
+      "grad_norm": 1.5110394954681396,
+      "learning_rate": 8.11345024225014e-05,
+      "loss": 3.5724,
+      "step": 120000
+    },
+    {
+      "epoch": 5.058382160772246,
+      "eval_accuracy": 0.6515163088771775,
+      "eval_loss": 3.2758891582489014,
+      "eval_runtime": 169.9427,
+      "eval_samples_per_second": 710.275,
+      "eval_steps_per_second": 7.402,
+      "step": 120000
+    },
+    {
+      "epoch": 5.079458753108797,
+      "grad_norm": 2.7520923614501953,
+      "learning_rate": 8.104875016078549e-05,
+      "loss": 3.5658,
+      "step": 120500
+    },
+    {
+      "epoch": 5.100535345445349,
+      "grad_norm": 2.4743049144744873,
+      "learning_rate": 8.09629978990696e-05,
+      "loss": 3.5652,
+      "step": 121000
+    },
+    {
+      "epoch": 5.1216119377819,
+      "grad_norm": 1.5048964023590088,
+      "learning_rate": 8.08772456373537e-05,
+      "loss": 3.5681,
+      "step": 121500
+    },
+    {
+      "epoch": 5.14268853011845,
+      "grad_norm": 2.0164084434509277,
+      "learning_rate": 8.079149337563778e-05,
+      "loss": 3.5668,
+      "step": 122000
+    },
+    {
+      "epoch": 5.163765122455001,
+      "grad_norm": 2.504014015197754,
+      "learning_rate": 8.070574111392188e-05,
+      "loss": 3.5617,
+      "step": 122500
+    },
+    {
+      "epoch": 5.184841714791553,
+      "grad_norm": 2.798123836517334,
+      "learning_rate": 8.061998885220599e-05,
+      "loss": 3.5624,
+      "step": 123000
+    },
+    {
+      "epoch": 5.205918307128104,
+      "grad_norm": 1.1612788438796997,
+      "learning_rate": 8.053423659049008e-05,
+      "loss": 3.5629,
+      "step": 123500
+    },
+    {
+      "epoch": 5.226994899464654,
+      "grad_norm": 2.4905834197998047,
+      "learning_rate": 8.044848432877418e-05,
+      "loss": 3.5624,
+      "step": 124000
+    },
+    {
+      "epoch": 5.2480714918012055,
+      "grad_norm": 3.208575963973999,
+      "learning_rate": 8.036273206705827e-05,
+      "loss": 3.5642,
+      "step": 124500
+    },
+    {
+      "epoch": 5.269148084137757,
+      "grad_norm": 2.5786023139953613,
+      "learning_rate": 8.027697980534237e-05,
+      "loss": 3.5566,
+      "step": 125000
+    },
+    {
+      "epoch": 5.290224676474308,
+      "grad_norm": 3.0945849418640137,
+      "learning_rate": 8.019122754362647e-05,
+      "loss": 3.5591,
+      "step": 125500
+    },
+    {
+      "epoch": 5.311301268810858,
+      "grad_norm": 1.6908981800079346,
+      "learning_rate": 8.010547528191057e-05,
+      "loss": 3.5571,
+      "step": 126000
+    },
+    {
+      "epoch": 5.3323778611474095,
+      "grad_norm": 2.745734691619873,
+      "learning_rate": 8.001972302019465e-05,
+      "loss": 3.5528,
+      "step": 126500
+    },
+    {
+      "epoch": 5.353454453483961,
+      "grad_norm": 2.8354687690734863,
+      "learning_rate": 7.993397075847876e-05,
+      "loss": 3.5524,
+      "step": 127000
+    },
+    {
+      "epoch": 5.374531045820512,
+      "grad_norm": 2.145784378051758,
+      "learning_rate": 7.984821849676286e-05,
+      "loss": 3.5561,
+      "step": 127500
+    },
+    {
+      "epoch": 5.395607638157063,
+      "grad_norm": 2.4846346378326416,
+      "learning_rate": 7.976246623504695e-05,
+      "loss": 3.5538,
+      "step": 128000
+    },
+    {
+      "epoch": 5.416684230493614,
+      "grad_norm": 1.974616289138794,
+      "learning_rate": 7.967671397333105e-05,
+      "loss": 3.5477,
+      "step": 128500
+    },
+    {
+      "epoch": 5.437760822830165,
+      "grad_norm": 2.1043343544006348,
+      "learning_rate": 7.959096171161516e-05,
+      "loss": 3.5478,
+      "step": 129000
+    },
+    {
+      "epoch": 5.458837415166716,
+      "grad_norm": 2.256662368774414,
+      "learning_rate": 7.950520944989924e-05,
+      "loss": 3.5484,
+      "step": 129500
+    },
+    {
+      "epoch": 5.479914007503266,
+      "grad_norm": 3.679149866104126,
+      "learning_rate": 7.941945718818334e-05,
+      "loss": 3.5458,
+      "step": 130000
+    },
+    {
+      "epoch": 5.479914007503266,
+      "eval_accuracy": 0.655075434533612,
+      "eval_loss": 3.256190299987793,
+      "eval_runtime": 171.3053,
+      "eval_samples_per_second": 704.625,
+      "eval_steps_per_second": 7.344,
+      "step": 130000
+    },
+    {
+      "epoch": 5.500990599839818,
+      "grad_norm": 2.2602665424346924,
+      "learning_rate": 7.933370492646744e-05,
+      "loss": 3.5518,
+      "step": 130500
+    },
+    {
+      "epoch": 5.522067192176369,
+      "grad_norm": 2.8571791648864746,
+      "learning_rate": 7.924795266475154e-05,
+      "loss": 3.5441,
+      "step": 131000
+    },
+    {
+      "epoch": 5.54314378451292,
+      "grad_norm": 3.013056993484497,
+      "learning_rate": 7.916220040303564e-05,
+      "loss": 3.5397,
+      "step": 131500
+    },
+    {
+      "epoch": 5.564220376849471,
+      "grad_norm": 1.3959802389144897,
+      "learning_rate": 7.907644814131972e-05,
+      "loss": 3.5438,
+      "step": 132000
+    },
+    {
+      "epoch": 5.585296969186022,
+      "grad_norm": 1.8692615032196045,
+      "learning_rate": 7.899069587960383e-05,
+      "loss": 3.5394,
+      "step": 132500
+    },
+    {
+      "epoch": 5.606373561522573,
+      "grad_norm": 2.1380040645599365,
+      "learning_rate": 7.890494361788793e-05,
+      "loss": 3.5398,
+      "step": 133000
+    },
+    {
+      "epoch": 5.627450153859124,
+      "grad_norm": 1.6554392576217651,
+      "learning_rate": 7.881919135617202e-05,
+      "loss": 3.5366,
+      "step": 133500
+    },
+    {
+      "epoch": 5.648526746195675,
+      "grad_norm": 1.6986337900161743,
+      "learning_rate": 7.873343909445611e-05,
+      "loss": 3.5321,
+      "step": 134000
+    },
+    {
+      "epoch": 5.669603338532226,
+      "grad_norm": 2.5319340229034424,
+      "learning_rate": 7.864768683274023e-05,
+      "loss": 3.541,
+      "step": 134500
+    },
+    {
+      "epoch": 5.690679930868777,
+      "grad_norm": 2.888364553451538,
+      "learning_rate": 7.856193457102431e-05,
+      "loss": 3.5411,
+      "step": 135000
+    },
+    {
+      "epoch": 5.711756523205328,
+      "grad_norm": 1.922222375869751,
+      "learning_rate": 7.847618230930841e-05,
+      "loss": 3.5368,
+      "step": 135500
+    },
+    {
+      "epoch": 5.7328331155418795,
+      "grad_norm": 1.859988808631897,
+      "learning_rate": 7.839043004759251e-05,
+      "loss": 3.5361,
+      "step": 136000
+    },
+    {
+      "epoch": 5.75390970787843,
+      "grad_norm": 1.6843948364257812,
+      "learning_rate": 7.83046777858766e-05,
+      "loss": 3.5322,
+      "step": 136500
+    },
+    {
+      "epoch": 5.774986300214981,
+      "grad_norm": 1.9180785417556763,
+      "learning_rate": 7.82189255241607e-05,
+      "loss": 3.5331,
+      "step": 137000
+    },
+    {
+      "epoch": 5.796062892551532,
+      "grad_norm": 2.320843458175659,
+      "learning_rate": 7.81331732624448e-05,
+      "loss": 3.5316,
+      "step": 137500
+    },
+    {
+      "epoch": 5.8171394848880835,
+      "grad_norm": 1.709078311920166,
+      "learning_rate": 7.804742100072889e-05,
+      "loss": 3.5279,
+      "step": 138000
+    },
+    {
+      "epoch": 5.838216077224635,
+      "grad_norm": 1.6784801483154297,
+      "learning_rate": 7.7961668739013e-05,
+      "loss": 3.5334,
+      "step": 138500
+    },
+    {
+      "epoch": 5.859292669561185,
+      "grad_norm": 2.466494083404541,
+      "learning_rate": 7.78759164772971e-05,
+      "loss": 3.5287,
+      "step": 139000
+    },
+    {
+      "epoch": 5.880369261897736,
+      "grad_norm": 1.6847925186157227,
+      "learning_rate": 7.779016421558118e-05,
+      "loss": 3.5264,
+      "step": 139500
+    },
+    {
+      "epoch": 5.901445854234288,
+      "grad_norm": 2.5696985721588135,
+      "learning_rate": 7.770441195386528e-05,
+      "loss": 3.53,
+      "step": 140000
+    },
+    {
+      "epoch": 5.901445854234288,
+      "eval_accuracy": 0.6587684995170529,
+      "eval_loss": 3.2333531379699707,
+      "eval_runtime": 170.5846,
+      "eval_samples_per_second": 707.602,
+      "eval_steps_per_second": 7.375,
+      "step": 140000
+    },
+    {
+      "epoch": 5.922522446570838,
+      "grad_norm": 1.8815934658050537,
+      "learning_rate": 7.761865969214939e-05,
+      "loss": 3.5216,
+      "step": 140500
+    },
+    {
+      "epoch": 5.943599038907389,
+      "grad_norm": 2.7514848709106445,
+      "learning_rate": 7.753290743043348e-05,
+      "loss": 3.5265,
+      "step": 141000
+    },
+    {
+      "epoch": 5.96467563124394,
+      "grad_norm": 2.053565502166748,
+      "learning_rate": 7.744715516871758e-05,
+      "loss": 3.5266,
+      "step": 141500
+    },
+    {
+      "epoch": 5.985752223580492,
+      "grad_norm": 2.6324079036712646,
+      "learning_rate": 7.736140290700167e-05,
+      "loss": 3.5205,
+      "step": 142000
+    },
+    {
+      "epoch": 6.006828815917043,
+      "grad_norm": 4.056976318359375,
+      "learning_rate": 7.727565064528577e-05,
+      "loss": 3.52,
+      "step": 142500
+    },
+    {
+      "epoch": 6.027905408253593,
+      "grad_norm": 2.0082273483276367,
+      "learning_rate": 7.718989838356987e-05,
+      "loss": 3.5188,
+      "step": 143000
+    },
+    {
+      "epoch": 6.0489820005901445,
+      "grad_norm": 3.0564260482788086,
+      "learning_rate": 7.710414612185397e-05,
+      "loss": 3.5208,
+      "step": 143500
+    },
+    {
+      "epoch": 6.070058592926696,
+      "grad_norm": 1.8841111660003662,
+      "learning_rate": 7.701839386013807e-05,
+      "loss": 3.5178,
+      "step": 144000
+    },
+    {
+      "epoch": 6.091135185263247,
+      "grad_norm": 1.9848800897598267,
+      "learning_rate": 7.693264159842217e-05,
+      "loss": 3.5147,
+      "step": 144500
+    },
+    {
+      "epoch": 6.112211777599797,
+      "grad_norm": 1.7279789447784424,
+      "learning_rate": 7.684688933670626e-05,
+      "loss": 3.5143,
+      "step": 145000
+    },
+    {
+      "epoch": 6.1332883699363485,
+      "grad_norm": 2.6928131580352783,
+      "learning_rate": 7.676113707499035e-05,
+      "loss": 3.5149,
+      "step": 145500
+    },
+    {
+      "epoch": 6.1543649622729,
+      "grad_norm": 2.1646580696105957,
+      "learning_rate": 7.667538481327446e-05,
+      "loss": 3.5138,
+      "step": 146000
+    },
+    {
+      "epoch": 6.175441554609451,
+      "grad_norm": 2.470848798751831,
+      "learning_rate": 7.658963255155856e-05,
+      "loss": 3.5153,
+      "step": 146500
+    },
+    {
+      "epoch": 6.196518146946001,
+      "grad_norm": 1.618044376373291,
+      "learning_rate": 7.650388028984264e-05,
+      "loss": 3.5125,
+      "step": 147000
+    },
+    {
+      "epoch": 6.217594739282553,
+      "grad_norm": 3.5621047019958496,
+      "learning_rate": 7.641812802812674e-05,
+      "loss": 3.5113,
+      "step": 147500
+    },
+    {
+      "epoch": 6.238671331619104,
+      "grad_norm": 1.950307011604309,
+      "learning_rate": 7.633237576641084e-05,
+      "loss": 3.509,
+      "step": 148000
+    },
+    {
+      "epoch": 6.259747923955655,
+      "grad_norm": 2.0797901153564453,
+      "learning_rate": 7.624662350469494e-05,
+      "loss": 3.5114,
+      "step": 148500
+    },
+    {
+      "epoch": 6.280824516292206,
+      "grad_norm": 2.415170192718506,
+      "learning_rate": 7.616087124297904e-05,
+      "loss": 3.5109,
+      "step": 149000
+    },
+    {
+      "epoch": 6.301901108628757,
+      "grad_norm": 2.1355020999908447,
+      "learning_rate": 7.607511898126313e-05,
+      "loss": 3.509,
+      "step": 149500
+    },
+    {
+      "epoch": 6.322977700965308,
+      "grad_norm": 1.6038293838500977,
+      "learning_rate": 7.598936671954723e-05,
+      "loss": 3.5054,
+      "step": 150000
+    },
+    {
+      "epoch": 6.322977700965308,
+      "eval_accuracy": 0.6609640034277329,
+      "eval_loss": 3.2178454399108887,
+      "eval_runtime": 171.1727,
+      "eval_samples_per_second": 705.171,
+      "eval_steps_per_second": 7.349,
+      "step": 150000
+    },
+    {
+      "epoch": 6.344054293301859,
+      "grad_norm": 3.664113998413086,
+      "learning_rate": 7.590361445783133e-05,
+      "loss": 3.5073,
+      "step": 150500
+    },
+    {
+      "epoch": 6.36513088563841,
+      "grad_norm": 2.015871524810791,
+      "learning_rate": 7.581786219611543e-05,
+      "loss": 3.5056,
+      "step": 151000
+    },
+    {
+      "epoch": 6.386207477974961,
+      "grad_norm": 1.8370819091796875,
+      "learning_rate": 7.573210993439951e-05,
+      "loss": 3.5069,
+      "step": 151500
+    },
+    {
+      "epoch": 6.407284070311512,
+      "grad_norm": 1.9508640766143799,
+      "learning_rate": 7.564635767268363e-05,
+      "loss": 3.5055,
+      "step": 152000
+    },
+    {
+      "epoch": 6.428360662648063,
+      "grad_norm": 1.6486629247665405,
+      "learning_rate": 7.556060541096772e-05,
+      "loss": 3.5017,
+      "step": 152500
+    },
+    {
+      "epoch": 6.449437254984614,
+      "grad_norm": 1.4163538217544556,
+      "learning_rate": 7.547485314925181e-05,
+      "loss": 3.5035,
+      "step": 153000
+    },
+    {
+      "epoch": 6.470513847321165,
+      "grad_norm": 1.6235407590866089,
+      "learning_rate": 7.538910088753591e-05,
+      "loss": 3.5013,
+      "step": 153500
+    },
+    {
+      "epoch": 6.491590439657716,
+      "grad_norm": 1.7410558462142944,
+      "learning_rate": 7.530334862582002e-05,
+      "loss": 3.5007,
+      "step": 154000
+    },
+    {
+      "epoch": 6.512667031994267,
+      "grad_norm": 2.534162998199463,
+      "learning_rate": 7.52175963641041e-05,
+      "loss": 3.5035,
+      "step": 154500
+    },
+    {
+      "epoch": 6.5337436243308185,
+      "grad_norm": 3.3592731952667236,
+      "learning_rate": 7.51318441023882e-05,
+      "loss": 3.5044,
+      "step": 155000
+    },
+    {
+      "epoch": 6.55482021666737,
+      "grad_norm": 1.6545741558074951,
+      "learning_rate": 7.50460918406723e-05,
+      "loss": 3.5006,
+      "step": 155500
+    },
+    {
+      "epoch": 6.57589680900392,
+      "grad_norm": 1.7850518226623535,
+      "learning_rate": 7.49603395789564e-05,
+      "loss": 3.4986,
+      "step": 156000
+    },
+    {
+      "epoch": 6.596973401340471,
+      "grad_norm": 1.5255476236343384,
+      "learning_rate": 7.48745873172405e-05,
+      "loss": 3.494,
+      "step": 156500
+    },
+    {
+      "epoch": 6.6180499936770225,
+      "grad_norm": 2.448270559310913,
+      "learning_rate": 7.47888350555246e-05,
+      "loss": 3.4905,
+      "step": 157000
+    },
+    {
+      "epoch": 6.639126586013573,
+      "grad_norm": 3.2912964820861816,
+      "learning_rate": 7.47030827938087e-05,
+      "loss": 3.4948,
+      "step": 157500
+    },
+    {
+      "epoch": 6.660203178350124,
+      "grad_norm": 1.9476572275161743,
+      "learning_rate": 7.461733053209279e-05,
+      "loss": 3.497,
+      "step": 158000
+    },
+    {
+      "epoch": 6.681279770686675,
+      "grad_norm": 2.466834783554077,
+      "learning_rate": 7.453157827037689e-05,
+      "loss": 3.4996,
+      "step": 158500
+    },
+    {
+      "epoch": 6.702356363023227,
+      "grad_norm": 1.8931043148040771,
+      "learning_rate": 7.444582600866098e-05,
+      "loss": 3.4904,
+      "step": 159000
+    },
+    {
+      "epoch": 6.723432955359778,
+      "grad_norm": 3.9832231998443604,
+      "learning_rate": 7.436007374694507e-05,
+      "loss": 3.4961,
+      "step": 159500
+    },
+    {
+      "epoch": 6.744509547696328,
+      "grad_norm": 1.7633931636810303,
+      "learning_rate": 7.427432148522919e-05,
+      "loss": 3.4971,
+      "step": 160000
+    },
+    {
+      "epoch": 6.744509547696328,
+      "eval_accuracy": 0.6631916394821478,
+      "eval_loss": 3.2042951583862305,
+      "eval_runtime": 170.3745,
+      "eval_samples_per_second": 708.475,
+      "eval_steps_per_second": 7.384,
+      "step": 160000
+    },
+    {
+      "epoch": 6.765586140032879,
+      "grad_norm": 2.2226905822753906,
+      "learning_rate": 7.418856922351327e-05,
+      "loss": 3.493,
+      "step": 160500
+    },
+    {
+      "epoch": 6.786662732369431,
+      "grad_norm": 2.6337685585021973,
+      "learning_rate": 7.410281696179737e-05,
+      "loss": 3.4898,
+      "step": 161000
+    },
+    {
+      "epoch": 6.807739324705982,
+      "grad_norm": 1.5373756885528564,
+      "learning_rate": 7.401706470008147e-05,
+      "loss": 3.4867,
+      "step": 161500
+    },
+    {
+      "epoch": 6.828815917042532,
+      "grad_norm": 1.7764091491699219,
+      "learning_rate": 7.393131243836557e-05,
+      "loss": 3.488,
+      "step": 162000
+    },
+    {
+      "epoch": 6.8498925093790834,
+      "grad_norm": 3.6729800701141357,
+      "learning_rate": 7.384556017664966e-05,
+      "loss": 3.4864,
+      "step": 162500
+    },
+    {
+      "epoch": 6.870969101715635,
+      "grad_norm": 1.7245464324951172,
+      "learning_rate": 7.375980791493376e-05,
+      "loss": 3.4836,
+      "step": 163000
+    },
+    {
+      "epoch": 6.892045694052186,
+      "grad_norm": 2.27529239654541,
+      "learning_rate": 7.367405565321786e-05,
+      "loss": 3.4848,
+      "step": 163500
+    },
+    {
+      "epoch": 6.913122286388736,
+      "grad_norm": 2.717825412750244,
+      "learning_rate": 7.358830339150196e-05,
+      "loss": 3.4851,
+      "step": 164000
+    },
+    {
+      "epoch": 6.9341988787252875,
+      "grad_norm": 1.4855817556381226,
+      "learning_rate": 7.350255112978606e-05,
+      "loss": 3.4899,
+      "step": 164500
+    },
+    {
+      "epoch": 6.955275471061839,
+      "grad_norm": 2.408026933670044,
+      "learning_rate": 7.341679886807014e-05,
+      "loss": 3.485,
+      "step": 165000
+    },
+    {
+      "epoch": 6.97635206339839,
+      "grad_norm": 2.3751261234283447,
+      "learning_rate": 7.333104660635425e-05,
+      "loss": 3.4833,
+      "step": 165500
+    },
+    {
+      "epoch": 6.997428655734941,
+      "grad_norm": 1.9266947507858276,
+      "learning_rate": 7.324529434463835e-05,
+      "loss": 3.482,
+      "step": 166000
+    },
+    {
+      "epoch": 7.018505248071492,
+      "grad_norm": 2.067725658416748,
+      "learning_rate": 7.315954208292244e-05,
+      "loss": 3.4825,
+      "step": 166500
+    },
+    {
+      "epoch": 7.039581840408043,
+      "grad_norm": 2.5464956760406494,
+      "learning_rate": 7.307378982120653e-05,
+      "loss": 3.4761,
+      "step": 167000
+    },
+    {
+      "epoch": 7.060658432744594,
+      "grad_norm": 2.7005245685577393,
+      "learning_rate": 7.298803755949063e-05,
+      "loss": 3.4773,
+      "step": 167500
+    },
+    {
+      "epoch": 7.081735025081145,
+      "grad_norm": 2.1714351177215576,
+      "learning_rate": 7.290228529777473e-05,
+      "loss": 3.4754,
+      "step": 168000
+    },
+    {
+      "epoch": 7.102811617417696,
+      "grad_norm": 1.805710792541504,
+      "learning_rate": 7.281653303605883e-05,
+      "loss": 3.4787,
+      "step": 168500
+    },
+    {
+      "epoch": 7.123888209754247,
+      "grad_norm": 1.9820823669433594,
+      "learning_rate": 7.273078077434293e-05,
+      "loss": 3.4823,
+      "step": 169000
+    },
+    {
+      "epoch": 7.144964802090798,
+      "grad_norm": 3.9914393424987793,
+      "learning_rate": 7.264502851262703e-05,
+      "loss": 3.4741,
+      "step": 169500
+    },
+    {
+      "epoch": 7.166041394427349,
+      "grad_norm": 2.913053035736084,
+      "learning_rate": 7.255927625091112e-05,
+      "loss": 3.4724,
+      "step": 170000
+    },
+    {
+      "epoch": 7.166041394427349,
+      "eval_accuracy": 0.6659268646324903,
+      "eval_loss": 3.1912574768066406,
+      "eval_runtime": 170.4916,
+      "eval_samples_per_second": 707.988,
+      "eval_steps_per_second": 7.379,
+      "step": 170000
+    },
+    {
+      "epoch": 7.1871179867639,
+      "grad_norm": 1.9531694650650024,
+      "learning_rate": 7.247352398919521e-05,
+      "loss": 3.4776,
+      "step": 170500
+    },
+    {
+      "epoch": 7.208194579100451,
+      "grad_norm": 2.780045509338379,
+      "learning_rate": 7.238777172747931e-05,
+      "loss": 3.475,
+      "step": 171000
+    },
+    {
+      "epoch": 7.229271171437002,
+      "grad_norm": 2.4574947357177734,
+      "learning_rate": 7.230201946576342e-05,
+      "loss": 3.4747,
+      "step": 171500
+    },
+    {
+      "epoch": 7.250347763773553,
+      "grad_norm": 2.1881215572357178,
+      "learning_rate": 7.22162672040475e-05,
+      "loss": 3.4741,
+      "step": 172000
+    },
+    {
+      "epoch": 7.271424356110104,
+      "grad_norm": 2.0233688354492188,
+      "learning_rate": 7.21305149423316e-05,
+      "loss": 3.4715,
+      "step": 172500
+    },
+    {
+      "epoch": 7.292500948446655,
+      "grad_norm": 2.9025232791900635,
+      "learning_rate": 7.20447626806157e-05,
+      "loss": 3.4676,
+      "step": 173000
+    },
+    {
+      "epoch": 7.313577540783206,
+      "grad_norm": 1.5233278274536133,
+      "learning_rate": 7.19590104188998e-05,
+      "loss": 3.4683,
+      "step": 173500
+    },
+    {
+      "epoch": 7.3346541331197574,
+      "grad_norm": 1.9305742979049683,
+      "learning_rate": 7.18732581571839e-05,
+      "loss": 3.4702,
+      "step": 174000
+    },
+    {
+      "epoch": 7.355730725456308,
+      "grad_norm": 2.9489240646362305,
+      "learning_rate": 7.1787505895468e-05,
+      "loss": 3.4682,
+      "step": 174500
+    },
+    {
+      "epoch": 7.376807317792859,
+      "grad_norm": 3.5476725101470947,
+      "learning_rate": 7.17017536337521e-05,
+      "loss": 3.4691,
+      "step": 175000
+    },
+    {
+      "epoch": 7.39788391012941,
+      "grad_norm": 2.627847671508789,
+      "learning_rate": 7.161600137203619e-05,
+      "loss": 3.4713,
+      "step": 175500
+    },
+    {
+      "epoch": 7.4189605024659615,
+      "grad_norm": 1.6323418617248535,
+      "learning_rate": 7.153024911032029e-05,
+      "loss": 3.465,
+      "step": 176000
+    },
+    {
+      "epoch": 7.440037094802513,
+      "grad_norm": 2.443479537963867,
+      "learning_rate": 7.144449684860438e-05,
+      "loss": 3.4683,
+      "step": 176500
+    },
+    {
+      "epoch": 7.461113687139063,
+      "grad_norm": 2.5164949893951416,
+      "learning_rate": 7.135874458688849e-05,
+      "loss": 3.4709,
+      "step": 177000
+    },
+    {
+      "epoch": 7.482190279475614,
+      "grad_norm": 3.3398420810699463,
+      "learning_rate": 7.127299232517259e-05,
+      "loss": 3.4712,
+      "step": 177500
+    },
+    {
+      "epoch": 7.503266871812166,
+      "grad_norm": 2.715989828109741,
+      "learning_rate": 7.118724006345667e-05,
+      "loss": 3.4658,
+      "step": 178000
+    },
+    {
+      "epoch": 7.524343464148717,
+      "grad_norm": 2.8976948261260986,
+      "learning_rate": 7.110148780174077e-05,
+      "loss": 3.4627,
+      "step": 178500
+    },
+    {
+      "epoch": 7.545420056485267,
+      "grad_norm": 2.5288238525390625,
+      "learning_rate": 7.101573554002488e-05,
+      "loss": 3.4629,
+      "step": 179000
+    },
+    {
+      "epoch": 7.566496648821818,
+      "grad_norm": 2.166144847869873,
+      "learning_rate": 7.092998327830897e-05,
+      "loss": 3.4662,
+      "step": 179500
+    },
+    {
+      "epoch": 7.58757324115837,
+      "grad_norm": 1.8500299453735352,
+      "learning_rate": 7.084423101659306e-05,
+      "loss": 3.4615,
+      "step": 180000
+    },
+    {
+      "epoch": 7.58757324115837,
+      "eval_accuracy": 0.667119587968321,
+      "eval_loss": 3.1805150508880615,
+      "eval_runtime": 170.0558,
+      "eval_samples_per_second": 709.802,
+      "eval_steps_per_second": 7.398,
+      "step": 180000
+    },
+    {
+      "epoch": 7.608649833494921,
+      "grad_norm": 3.2349753379821777,
+      "learning_rate": 7.075847875487716e-05,
+      "loss": 3.464,
+      "step": 180500
+    },
+    {
+      "epoch": 7.629726425831471,
+      "grad_norm": 2.023772716522217,
+      "learning_rate": 7.067272649316126e-05,
+      "loss": 3.4639,
+      "step": 181000
+    },
+    {
+      "epoch": 7.650803018168022,
+      "grad_norm": 1.7736855745315552,
+      "learning_rate": 7.058697423144536e-05,
+      "loss": 3.4575,
+      "step": 181500
+    },
+    {
+      "epoch": 7.671879610504574,
+      "grad_norm": 1.7336115837097168,
+      "learning_rate": 7.050122196972946e-05,
+      "loss": 3.458,
+      "step": 182000
+    },
+    {
+      "epoch": 7.692956202841125,
+      "grad_norm": 1.6601215600967407,
+      "learning_rate": 7.041546970801354e-05,
+      "loss": 3.4604,
+      "step": 182500
+    },
+    {
+      "epoch": 7.714032795177676,
+      "grad_norm": 2.5029027462005615,
+      "learning_rate": 7.032971744629765e-05,
+      "loss": 3.4612,
+      "step": 183000
+    },
+    {
+      "epoch": 7.7351093875142265,
+      "grad_norm": 2.9890849590301514,
+      "learning_rate": 7.024396518458175e-05,
+      "loss": 3.4646,
+      "step": 183500
+    },
+    {
+      "epoch": 7.756185979850778,
+      "grad_norm": 2.307880401611328,
+      "learning_rate": 7.015821292286584e-05,
+      "loss": 3.4599,
+      "step": 184000
+    },
+    {
+      "epoch": 7.777262572187329,
+      "grad_norm": 1.771583080291748,
+      "learning_rate": 7.007246066114993e-05,
+      "loss": 3.458,
+      "step": 184500
+    },
+    {
+      "epoch": 7.798339164523879,
+      "grad_norm": 3.9901998043060303,
+      "learning_rate": 6.998670839943405e-05,
+      "loss": 3.4605,
+      "step": 185000
+    },
+    {
+      "epoch": 7.8194157568604306,
+      "grad_norm": 2.877674102783203,
+      "learning_rate": 6.990095613771813e-05,
+      "loss": 3.4566,
+      "step": 185500
+    },
+    {
+      "epoch": 7.840492349196982,
+      "grad_norm": 2.722691774368286,
+      "learning_rate": 6.981520387600223e-05,
+      "loss": 3.4558,
+      "step": 186000
+    },
+    {
+      "epoch": 7.861568941533533,
+      "grad_norm": 2.9399402141571045,
+      "learning_rate": 6.972945161428633e-05,
+      "loss": 3.4575,
+      "step": 186500
+    },
+    {
+      "epoch": 7.882645533870084,
+      "grad_norm": 1.9601466655731201,
+      "learning_rate": 6.964369935257043e-05,
+      "loss": 3.4563,
+      "step": 187000
+    },
+    {
+      "epoch": 7.903722126206635,
+      "grad_norm": 1.7952498197555542,
+      "learning_rate": 6.955794709085452e-05,
+      "loss": 3.4567,
+      "step": 187500
+    },
+    {
+      "epoch": 7.924798718543186,
+      "grad_norm": 1.6690802574157715,
+      "learning_rate": 6.947219482913862e-05,
+      "loss": 3.4452,
+      "step": 188000
+    },
+    {
+      "epoch": 7.945875310879737,
+      "grad_norm": 2.447540283203125,
+      "learning_rate": 6.938644256742272e-05,
+      "loss": 3.4566,
+      "step": 188500
+    },
+    {
+      "epoch": 7.966951903216288,
+      "grad_norm": 2.148979425430298,
+      "learning_rate": 6.930069030570682e-05,
+      "loss": 3.4532,
+      "step": 189000
+    },
+    {
+      "epoch": 7.988028495552839,
+      "grad_norm": 3.1907222270965576,
+      "learning_rate": 6.921493804399092e-05,
+      "loss": 3.449,
+      "step": 189500
+    },
+    {
+      "epoch": 8.00910508788939,
+      "grad_norm": 1.3401950597763062,
+      "learning_rate": 6.9129185782275e-05,
+      "loss": 3.4516,
+      "step": 190000
+    },
+    {
+      "epoch": 8.00910508788939,
+      "eval_accuracy": 0.6690257900726433,
+      "eval_loss": 3.1676809787750244,
+      "eval_runtime": 170.342,
+      "eval_samples_per_second": 708.61,
+      "eval_steps_per_second": 7.385,
+      "step": 190000
+    },
+    {
+      "epoch": 8.03018168022594,
+      "grad_norm": 1.70050847530365,
+      "learning_rate": 6.904343352055911e-05,
+      "loss": 3.447,
+      "step": 190500
+    },
+    {
+      "epoch": 8.051258272562492,
+      "grad_norm": 1.8312853574752808,
+      "learning_rate": 6.895768125884321e-05,
+      "loss": 3.4504,
+      "step": 191000
+    },
+    {
+      "epoch": 8.072334864899043,
+      "grad_norm": 2.555469512939453,
+      "learning_rate": 6.88719289971273e-05,
+      "loss": 3.4485,
+      "step": 191500
+    },
+    {
+      "epoch": 8.093411457235595,
+      "grad_norm": 1.7558132410049438,
+      "learning_rate": 6.87861767354114e-05,
+      "loss": 3.4425,
+      "step": 192000
+    },
+    {
+      "epoch": 8.114488049572145,
+      "grad_norm": 3.5355947017669678,
+      "learning_rate": 6.870042447369551e-05,
+      "loss": 3.4498,
+      "step": 192500
+    },
+    {
+      "epoch": 8.135564641908696,
+      "grad_norm": 2.2370765209198,
+      "learning_rate": 6.861467221197959e-05,
+      "loss": 3.4469,
+      "step": 193000
+    },
+    {
+      "epoch": 8.156641234245248,
+      "grad_norm": 1.830134630203247,
+      "learning_rate": 6.852891995026369e-05,
+      "loss": 3.4487,
+      "step": 193500
+    },
+    {
+      "epoch": 8.177717826581798,
+      "grad_norm": 2.2721779346466064,
+      "learning_rate": 6.844316768854779e-05,
+      "loss": 3.4486,
+      "step": 194000
+    },
+    {
+      "epoch": 8.19879441891835,
+      "grad_norm": 1.363957166671753,
+      "learning_rate": 6.835741542683189e-05,
+      "loss": 3.4457,
+      "step": 194500
+    },
+    {
+      "epoch": 8.2198710112549,
+      "grad_norm": 1.8466328382492065,
+      "learning_rate": 6.827166316511599e-05,
+      "loss": 3.4438,
+      "step": 195000
+    },
+    {
+      "epoch": 8.24094760359145,
+      "grad_norm": 2.2240543365478516,
+      "learning_rate": 6.818591090340008e-05,
+      "loss": 3.4449,
+      "step": 195500
+    },
+    {
+      "epoch": 8.262024195928003,
+      "grad_norm": 3.2572226524353027,
+      "learning_rate": 6.810015864168417e-05,
+      "loss": 3.4418,
+      "step": 196000
+    },
+    {
+      "epoch": 8.283100788264553,
+      "grad_norm": 3.133857011795044,
+      "learning_rate": 6.801440637996828e-05,
+      "loss": 3.4412,
+      "step": 196500
+    },
+    {
+      "epoch": 8.304177380601104,
+      "grad_norm": 2.5218095779418945,
+      "learning_rate": 6.792865411825238e-05,
+      "loss": 3.4441,
+      "step": 197000
+    },
+    {
+      "epoch": 8.325253972937656,
+      "grad_norm": 2.769869804382324,
+      "learning_rate": 6.784290185653646e-05,
+      "loss": 3.4416,
+      "step": 197500
+    },
+    {
+      "epoch": 8.346330565274206,
+      "grad_norm": 2.3545761108398438,
+      "learning_rate": 6.775714959482056e-05,
+      "loss": 3.4401,
+      "step": 198000
+    },
+    {
+      "epoch": 8.367407157610758,
+      "grad_norm": 1.479090929031372,
+      "learning_rate": 6.767139733310467e-05,
+      "loss": 3.4416,
+      "step": 198500
+    },
+    {
+      "epoch": 8.388483749947309,
+      "grad_norm": 2.7855374813079834,
+      "learning_rate": 6.758564507138876e-05,
+      "loss": 3.4445,
+      "step": 199000
+    },
+    {
+      "epoch": 8.409560342283859,
+      "grad_norm": 2.530709981918335,
+      "learning_rate": 6.749989280967286e-05,
+      "loss": 3.4447,
+      "step": 199500
+    },
+    {
+      "epoch": 8.430636934620411,
+      "grad_norm": 2.4114856719970703,
+      "learning_rate": 6.741414054795696e-05,
+      "loss": 3.4405,
+      "step": 200000
+    },
+    {
+      "epoch": 8.430636934620411,
+      "eval_accuracy": 0.6709656507496864,
+      "eval_loss": 3.1585371494293213,
+      "eval_runtime": 170.4665,
+      "eval_samples_per_second": 708.092,
+      "eval_steps_per_second": 7.38,
+      "step": 200000
+    },
+    {
+      "epoch": 8.451713526956961,
+      "grad_norm": 1.9623064994812012,
+      "learning_rate": 6.732838828624105e-05,
+      "loss": 3.439,
+      "step": 200500
+    },
+    {
+      "epoch": 8.472790119293512,
+      "grad_norm": 1.6759023666381836,
+      "learning_rate": 6.724263602452515e-05,
+      "loss": 3.4421,
+      "step": 201000
+    },
+    {
+      "epoch": 8.493866711630064,
+      "grad_norm": 3.8088364601135254,
+      "learning_rate": 6.715688376280924e-05,
+      "loss": 3.4393,
+      "step": 201500
+    },
+    {
+      "epoch": 8.514943303966614,
+      "grad_norm": 1.4799833297729492,
+      "learning_rate": 6.707113150109335e-05,
+      "loss": 3.4381,
+      "step": 202000
+    },
+    {
+      "epoch": 8.536019896303166,
+      "grad_norm": 1.9126983880996704,
+      "learning_rate": 6.698537923937745e-05,
+      "loss": 3.4431,
+      "step": 202500
+    },
+    {
+      "epoch": 8.557096488639717,
+      "grad_norm": 2.3618762493133545,
+      "learning_rate": 6.689962697766153e-05,
+      "loss": 3.4373,
+      "step": 203000
+    },
+    {
+      "epoch": 8.578173080976267,
+      "grad_norm": 1.5178073644638062,
+      "learning_rate": 6.681387471594563e-05,
+      "loss": 3.438,
+      "step": 203500
+    },
+    {
+      "epoch": 8.59924967331282,
+      "grad_norm": 1.8160009384155273,
+      "learning_rate": 6.672812245422974e-05,
+      "loss": 3.4369,
+      "step": 204000
+    },
+    {
+      "epoch": 8.62032626564937,
+      "grad_norm": 2.766925573348999,
+      "learning_rate": 6.664237019251383e-05,
+      "loss": 3.4381,
+      "step": 204500
+    },
+    {
+      "epoch": 8.64140285798592,
+      "grad_norm": 1.6445059776306152,
+      "learning_rate": 6.655661793079792e-05,
+      "loss": 3.4364,
+      "step": 205000
+    },
+    {
+      "epoch": 8.662479450322472,
+      "grad_norm": 1.5063164234161377,
+      "learning_rate": 6.647086566908202e-05,
+      "loss": 3.4316,
+      "step": 205500
+    },
+    {
+      "epoch": 8.683556042659022,
+      "grad_norm": 3.5412802696228027,
+      "learning_rate": 6.638511340736612e-05,
+      "loss": 3.4334,
+      "step": 206000
+    },
+    {
+      "epoch": 8.704632634995574,
+      "grad_norm": 3.2456698417663574,
+      "learning_rate": 6.629936114565022e-05,
+      "loss": 3.4294,
+      "step": 206500
+    },
+    {
+      "epoch": 8.725709227332125,
+      "grad_norm": 2.5221192836761475,
+      "learning_rate": 6.621360888393432e-05,
+      "loss": 3.4321,
+      "step": 207000
+    },
+    {
+      "epoch": 8.746785819668675,
+      "grad_norm": 2.969827890396118,
+      "learning_rate": 6.61278566222184e-05,
+      "loss": 3.4265,
+      "step": 207500
+    },
+    {
+      "epoch": 8.767862412005227,
+      "grad_norm": 3.342434883117676,
+      "learning_rate": 6.604210436050251e-05,
+      "loss": 3.4326,
+      "step": 208000
+    },
+    {
+      "epoch": 8.788939004341778,
+      "grad_norm": 1.7624473571777344,
+      "learning_rate": 6.595635209878661e-05,
+      "loss": 3.4319,
+      "step": 208500
+    },
+    {
+      "epoch": 8.81001559667833,
+      "grad_norm": 1.3666162490844727,
+      "learning_rate": 6.58705998370707e-05,
+      "loss": 3.434,
+      "step": 209000
+    },
+    {
+      "epoch": 8.83109218901488,
+      "grad_norm": 3.529292583465576,
+      "learning_rate": 6.57848475753548e-05,
+      "loss": 3.4311,
+      "step": 209500
+    },
+    {
+      "epoch": 8.85216878135143,
+      "grad_norm": 2.5807464122772217,
+      "learning_rate": 6.569909531363891e-05,
+      "loss": 3.4284,
+      "step": 210000
+    },
+    {
+      "epoch": 8.85216878135143,
+      "eval_accuracy": 0.6722068360336451,
+      "eval_loss": 3.150799036026001,
+      "eval_runtime": 170.4269,
+      "eval_samples_per_second": 708.257,
+      "eval_steps_per_second": 7.381,
+      "step": 210000
+    },
+    {
+      "epoch": 8.873245373687983,
+      "grad_norm": 2.0278961658477783,
+      "learning_rate": 6.561334305192299e-05,
+      "loss": 3.4301,
+      "step": 210500
+    },
+    {
+      "epoch": 8.894321966024533,
+      "grad_norm": 2.5886778831481934,
+      "learning_rate": 6.552759079020709e-05,
+      "loss": 3.4311,
+      "step": 211000
+    },
+    {
+      "epoch": 8.915398558361083,
+      "grad_norm": 2.6455419063568115,
+      "learning_rate": 6.544183852849119e-05,
+      "loss": 3.4278,
+      "step": 211500
+    },
+    {
+      "epoch": 8.936475150697635,
+      "grad_norm": 1.5726028680801392,
+      "learning_rate": 6.535608626677529e-05,
+      "loss": 3.4334,
+      "step": 212000
+    },
+    {
+      "epoch": 8.957551743034186,
+      "grad_norm": 2.4174139499664307,
+      "learning_rate": 6.527033400505939e-05,
+      "loss": 3.433,
+      "step": 212500
+    },
+    {
+      "epoch": 8.978628335370738,
+      "grad_norm": 2.2509894371032715,
+      "learning_rate": 6.518458174334348e-05,
+      "loss": 3.4323,
+      "step": 213000
+    },
+    {
+      "epoch": 8.999704927707288,
+      "grad_norm": 2.8906824588775635,
+      "learning_rate": 6.509882948162758e-05,
+      "loss": 3.4307,
+      "step": 213500
+    },
+    {
+      "epoch": 9.020781520043839,
+      "grad_norm": 2.3862533569335938,
+      "learning_rate": 6.501307721991168e-05,
+      "loss": 3.423,
+      "step": 214000
+    },
+    {
+      "epoch": 9.04185811238039,
+      "grad_norm": 1.9133273363113403,
+      "learning_rate": 6.492732495819578e-05,
+      "loss": 3.4212,
+      "step": 214500
+    },
+    {
+      "epoch": 9.062934704716941,
+      "grad_norm": 2.4166204929351807,
+      "learning_rate": 6.484157269647986e-05,
+      "loss": 3.4235,
+      "step": 215000
+    },
+    {
+      "epoch": 9.084011297053493,
+      "grad_norm": 1.972204566001892,
+      "learning_rate": 6.475582043476398e-05,
+      "loss": 3.4252,
+      "step": 215500
+    },
+    {
+      "epoch": 9.105087889390044,
+      "grad_norm": 2.017542600631714,
+      "learning_rate": 6.467006817304807e-05,
+      "loss": 3.4206,
+      "step": 216000
+    },
+    {
+      "epoch": 9.126164481726594,
+      "grad_norm": 2.25396728515625,
+      "learning_rate": 6.458431591133216e-05,
+      "loss": 3.4166,
+      "step": 216500
+    },
+    {
+      "epoch": 9.147241074063146,
+      "grad_norm": 2.4529523849487305,
+      "learning_rate": 6.449856364961626e-05,
+      "loss": 3.4213,
+      "step": 217000
+    },
+    {
+      "epoch": 9.168317666399696,
+      "grad_norm": 3.830040454864502,
+      "learning_rate": 6.441281138790037e-05,
+      "loss": 3.4198,
+      "step": 217500
+    },
+    {
+      "epoch": 9.189394258736247,
+      "grad_norm": 2.7730979919433594,
+      "learning_rate": 6.432705912618445e-05,
+      "loss": 3.4215,
+      "step": 218000
+    },
+    {
+      "epoch": 9.210470851072799,
+      "grad_norm": 2.499155282974243,
+      "learning_rate": 6.424130686446855e-05,
+      "loss": 3.4236,
+      "step": 218500
+    },
+    {
+      "epoch": 9.23154744340935,
+      "grad_norm": 2.6788418292999268,
+      "learning_rate": 6.415555460275265e-05,
+      "loss": 3.4207,
+      "step": 219000
+    },
+    {
+      "epoch": 9.252624035745901,
+      "grad_norm": 2.8359076976776123,
+      "learning_rate": 6.406980234103675e-05,
+      "loss": 3.4214,
+      "step": 219500
+    },
+    {
+      "epoch": 9.273700628082452,
+      "grad_norm": 2.5217347145080566,
+      "learning_rate": 6.398405007932085e-05,
+      "loss": 3.4212,
+      "step": 220000
+    },
+    {
+      "epoch": 9.273700628082452,
+      "eval_accuracy": 0.6731496578274068,
+      "eval_loss": 3.142566442489624,
+      "eval_runtime": 170.6803,
+      "eval_samples_per_second": 707.205,
+      "eval_steps_per_second": 7.371,
+      "step": 220000
+    },
+    {
+      "epoch": 9.294777220419002,
+      "grad_norm": 2.3356823921203613,
+      "learning_rate": 6.389829781760495e-05,
+      "loss": 3.4193,
+      "step": 220500
+    },
+    {
+      "epoch": 9.315853812755554,
+      "grad_norm": 1.7055457830429077,
+      "learning_rate": 6.381254555588903e-05,
+      "loss": 3.4223,
+      "step": 221000
+    },
+    {
+      "epoch": 9.336930405092104,
+      "grad_norm": 2.5891177654266357,
+      "learning_rate": 6.372679329417314e-05,
+      "loss": 3.4173,
+      "step": 221500
+    },
+    {
+      "epoch": 9.358006997428657,
+      "grad_norm": 1.4155911207199097,
+      "learning_rate": 6.364104103245724e-05,
+      "loss": 3.4155,
+      "step": 222000
+    },
+    {
+      "epoch": 9.379083589765207,
+      "grad_norm": 1.3315377235412598,
+      "learning_rate": 6.355528877074132e-05,
+      "loss": 3.4198,
+      "step": 222500
+    },
+    {
+      "epoch": 9.400160182101757,
+      "grad_norm": 1.7182308435440063,
+      "learning_rate": 6.346953650902542e-05,
+      "loss": 3.4182,
+      "step": 223000
+    },
+    {
+      "epoch": 9.42123677443831,
+      "grad_norm": 2.1110994815826416,
+      "learning_rate": 6.338378424730954e-05,
+      "loss": 3.4176,
+      "step": 223500
+    },
+    {
+      "epoch": 9.44231336677486,
+      "grad_norm": 3.825590133666992,
+      "learning_rate": 6.329803198559362e-05,
+      "loss": 3.4169,
+      "step": 224000
+    },
+    {
+      "epoch": 9.46338995911141,
+      "grad_norm": 2.2922894954681396,
+      "learning_rate": 6.321227972387772e-05,
+      "loss": 3.4171,
+      "step": 224500
+    },
+    {
+      "epoch": 9.484466551447962,
+      "grad_norm": 2.301192283630371,
+      "learning_rate": 6.312652746216182e-05,
+      "loss": 3.4129,
+      "step": 225000
+    },
+    {
+      "epoch": 9.505543143784513,
+      "grad_norm": 2.3466062545776367,
+      "learning_rate": 6.304077520044591e-05,
+      "loss": 3.4216,
+      "step": 225500
+    },
+    {
+      "epoch": 9.526619736121065,
+      "grad_norm": 2.055072784423828,
+      "learning_rate": 6.295502293873001e-05,
+      "loss": 3.4144,
+      "step": 226000
+    },
+    {
+      "epoch": 9.547696328457615,
+      "grad_norm": 2.3598484992980957,
+      "learning_rate": 6.286927067701411e-05,
+      "loss": 3.4128,
+      "step": 226500
+    },
+    {
+      "epoch": 9.568772920794165,
+      "grad_norm": 1.962658405303955,
+      "learning_rate": 6.278351841529821e-05,
+      "loss": 3.4163,
+      "step": 227000
+    },
+    {
+      "epoch": 9.589849513130718,
+      "grad_norm": 1.9100004434585571,
+      "learning_rate": 6.269776615358231e-05,
+      "loss": 3.4143,
+      "step": 227500
+    },
+    {
+      "epoch": 9.610926105467268,
+      "grad_norm": 1.861249566078186,
+      "learning_rate": 6.26120138918664e-05,
+      "loss": 3.413,
+      "step": 228000
+    },
+    {
+      "epoch": 9.63200269780382,
+      "grad_norm": 2.6080896854400635,
+      "learning_rate": 6.252626163015049e-05,
+      "loss": 3.4141,
+      "step": 228500
+    },
+    {
+      "epoch": 9.65307929014037,
+      "grad_norm": 2.047487497329712,
+      "learning_rate": 6.24405093684346e-05,
+      "loss": 3.412,
+      "step": 229000
+    },
+    {
+      "epoch": 9.67415588247692,
+      "grad_norm": 2.567725419998169,
+      "learning_rate": 6.23547571067187e-05,
+      "loss": 3.4109,
+      "step": 229500
+    },
+    {
+      "epoch": 9.695232474813473,
+      "grad_norm": 1.6157793998718262,
+      "learning_rate": 6.226900484500279e-05,
+      "loss": 3.4078,
+      "step": 230000
+    },
+    {
+      "epoch": 9.695232474813473,
+      "eval_accuracy": 0.6745959335667012,
+      "eval_loss": 3.1355597972869873,
+      "eval_runtime": 170.4742,
+      "eval_samples_per_second": 708.06,
+      "eval_steps_per_second": 7.379,
+      "step": 230000
+    },
+    {
+      "epoch": 9.716309067150023,
+      "grad_norm": 2.2575342655181885,
+      "learning_rate": 6.218325258328688e-05,
+      "loss": 3.4155,
+      "step": 230500
+    },
+    {
+      "epoch": 9.737385659486574,
+      "grad_norm": 3.746753215789795,
+      "learning_rate": 6.209750032157098e-05,
+      "loss": 3.4125,
+      "step": 231000
+    },
+    {
+      "epoch": 9.758462251823126,
+      "grad_norm": 2.044355630874634,
+      "learning_rate": 6.201174805985508e-05,
+      "loss": 3.4097,
+      "step": 231500
+    },
+    {
+      "epoch": 9.779538844159676,
+      "grad_norm": 4.164864540100098,
+      "learning_rate": 6.192599579813918e-05,
+      "loss": 3.4112,
+      "step": 232000
+    },
+    {
+      "epoch": 9.800615436496226,
+      "grad_norm": 2.8845841884613037,
+      "learning_rate": 6.184024353642328e-05,
+      "loss": 3.4091,
+      "step": 232500
+    },
+    {
+      "epoch": 9.821692028832778,
+      "grad_norm": 2.2413220405578613,
+      "learning_rate": 6.175449127470738e-05,
+      "loss": 3.4066,
+      "step": 233000
+    },
+    {
+      "epoch": 9.842768621169329,
+      "grad_norm": 2.336934804916382,
+      "learning_rate": 6.166873901299147e-05,
+      "loss": 3.4149,
+      "step": 233500
+    },
+    {
+      "epoch": 9.863845213505881,
+      "grad_norm": 2.381974458694458,
+      "learning_rate": 6.158298675127557e-05,
+      "loss": 3.4075,
+      "step": 234000
+    },
+    {
+      "epoch": 9.884921805842431,
+      "grad_norm": 3.263017177581787,
+      "learning_rate": 6.149723448955966e-05,
+      "loss": 3.4113,
+      "step": 234500
+    },
+    {
+      "epoch": 9.905998398178982,
+      "grad_norm": 2.1938061714172363,
+      "learning_rate": 6.141148222784377e-05,
+      "loss": 3.4053,
+      "step": 235000
+    },
+    {
+      "epoch": 9.927074990515534,
+      "grad_norm": 3.128086805343628,
+      "learning_rate": 6.132572996612787e-05,
+      "loss": 3.4064,
+      "step": 235500
+    },
+    {
+      "epoch": 9.948151582852084,
+      "grad_norm": 2.7213141918182373,
+      "learning_rate": 6.123997770441195e-05,
+      "loss": 3.4088,
+      "step": 236000
+    },
+    {
+      "epoch": 9.969228175188636,
+      "grad_norm": 1.4141939878463745,
+      "learning_rate": 6.115422544269605e-05,
+      "loss": 3.4126,
+      "step": 236500
+    },
+    {
+      "epoch": 9.990304767525187,
+      "grad_norm": 3.4904379844665527,
+      "learning_rate": 6.106847318098015e-05,
+      "loss": 3.4093,
+      "step": 237000
+    },
+    {
+      "epoch": 10.011381359861737,
+      "grad_norm": 2.37056303024292,
+      "learning_rate": 6.098272091926425e-05,
+      "loss": 3.4048,
+      "step": 237500
+    },
+    {
+      "epoch": 10.032457952198289,
+      "grad_norm": 2.0672197341918945,
+      "learning_rate": 6.0896968657548345e-05,
+      "loss": 3.4036,
+      "step": 238000
+    },
+    {
+      "epoch": 10.05353454453484,
+      "grad_norm": 1.508747935295105,
+      "learning_rate": 6.081121639583244e-05,
+      "loss": 3.4012,
+      "step": 238500
+    },
+    {
+      "epoch": 10.07461113687139,
+      "grad_norm": 3.113813638687134,
+      "learning_rate": 6.072546413411654e-05,
+      "loss": 3.4012,
+      "step": 239000
+    },
+    {
+      "epoch": 10.095687729207942,
+      "grad_norm": 1.8261330127716064,
+      "learning_rate": 6.063971187240064e-05,
+      "loss": 3.4014,
+      "step": 239500
+    },
+    {
+      "epoch": 10.116764321544492,
+      "grad_norm": 2.9391748905181885,
+      "learning_rate": 6.055395961068473e-05,
+      "loss": 3.3963,
+      "step": 240000
+    },
+    {
+      "epoch": 10.116764321544492,
+      "eval_accuracy": 0.6762358462992788,
+      "eval_loss": 3.1309874057769775,
+      "eval_runtime": 170.8017,
+      "eval_samples_per_second": 706.702,
+      "eval_steps_per_second": 7.365,
+      "step": 240000
+    },
+    {
+      "epoch": 10.137840913881044,
+      "grad_norm": 2.407372236251831,
+      "learning_rate": 6.046820734896883e-05,
+      "loss": 3.3983,
+      "step": 240500
+    },
+    {
+      "epoch": 10.158917506217595,
+      "grad_norm": 1.699548363685608,
+      "learning_rate": 6.038245508725293e-05,
+      "loss": 3.4002,
+      "step": 241000
+    },
+    {
+      "epoch": 10.179994098554145,
+      "grad_norm": 2.2014219760894775,
+      "learning_rate": 6.029670282553702e-05,
+      "loss": 3.3979,
+      "step": 241500
+    },
+    {
+      "epoch": 10.201070690890697,
+      "grad_norm": 2.6174776554107666,
+      "learning_rate": 6.0210950563821125e-05,
+      "loss": 3.4031,
+      "step": 242000
+    },
+    {
+      "epoch": 10.222147283227248,
+      "grad_norm": 2.962143659591675,
+      "learning_rate": 6.0125198302105223e-05,
+      "loss": 3.4009,
+      "step": 242500
+    },
+    {
+      "epoch": 10.2432238755638,
+      "grad_norm": 1.6970453262329102,
+      "learning_rate": 6.0039446040389315e-05,
+      "loss": 3.3986,
+      "step": 243000
+    },
+    {
+      "epoch": 10.26430046790035,
+      "grad_norm": 1.9613587856292725,
+      "learning_rate": 5.995369377867341e-05,
+      "loss": 3.404,
+      "step": 243500
+    },
+    {
+      "epoch": 10.2853770602369,
+      "grad_norm": 2.640956401824951,
+      "learning_rate": 5.986794151695752e-05,
+      "loss": 3.3938,
+      "step": 244000
+    },
+    {
+      "epoch": 10.306453652573452,
+      "grad_norm": 2.766664981842041,
+      "learning_rate": 5.97821892552416e-05,
+      "loss": 3.4028,
+      "step": 244500
+    },
+    {
+      "epoch": 10.327530244910003,
+      "grad_norm": 2.4877254962921143,
+      "learning_rate": 5.969643699352571e-05,
+      "loss": 3.4014,
+      "step": 245000
+    },
+    {
+      "epoch": 10.348606837246553,
+      "grad_norm": 2.1203157901763916,
+      "learning_rate": 5.9610684731809806e-05,
+      "loss": 3.4006,
+      "step": 245500
+    },
+    {
+      "epoch": 10.369683429583105,
+      "grad_norm": 2.270655870437622,
+      "learning_rate": 5.95249324700939e-05,
+      "loss": 3.3979,
+      "step": 246000
+    },
+    {
+      "epoch": 10.390760021919656,
+      "grad_norm": 1.516566514968872,
+      "learning_rate": 5.9439180208377996e-05,
+      "loss": 3.3984,
+      "step": 246500
+    },
+    {
+      "epoch": 10.411836614256208,
+      "grad_norm": 2.475419521331787,
+      "learning_rate": 5.93534279466621e-05,
+      "loss": 3.3988,
+      "step": 247000
+    },
+    {
+      "epoch": 10.432913206592758,
+      "grad_norm": 1.500043272972107,
+      "learning_rate": 5.9267675684946186e-05,
+      "loss": 3.3996,
+      "step": 247500
+    },
+    {
+      "epoch": 10.453989798929308,
+      "grad_norm": 2.8367066383361816,
+      "learning_rate": 5.918192342323029e-05,
+      "loss": 3.3947,
+      "step": 248000
+    },
+    {
+      "epoch": 10.47506639126586,
+      "grad_norm": 1.6601351499557495,
+      "learning_rate": 5.909617116151439e-05,
+      "loss": 3.3956,
+      "step": 248500
+    },
+    {
+      "epoch": 10.496142983602411,
+      "grad_norm": 2.9257187843322754,
+      "learning_rate": 5.901041889979848e-05,
+      "loss": 3.3957,
+      "step": 249000
+    },
+    {
+      "epoch": 10.517219575938963,
+      "grad_norm": 2.4048988819122314,
+      "learning_rate": 5.892466663808258e-05,
+      "loss": 3.3982,
+      "step": 249500
+    },
+    {
+      "epoch": 10.538296168275513,
+      "grad_norm": 2.5580906867980957,
+      "learning_rate": 5.8838914376366684e-05,
+      "loss": 3.3982,
+      "step": 250000
+    },
+    {
+      "epoch": 10.538296168275513,
+      "eval_accuracy": 0.6770840198438894,
+      "eval_loss": 3.1213865280151367,
+      "eval_runtime": 170.3124,
+      "eval_samples_per_second": 708.733,
+      "eval_steps_per_second": 7.386,
+      "step": 250000
+    },
+    {
+      "epoch": 10.559372760612064,
+      "grad_norm": 2.329874277114868,
+      "learning_rate": 5.8753162114650776e-05,
+      "loss": 3.4019,
+      "step": 250500
+    },
+    {
+      "epoch": 10.580449352948616,
+      "grad_norm": 2.34917950630188,
+      "learning_rate": 5.8667409852934874e-05,
+      "loss": 3.395,
+      "step": 251000
+    },
+    {
+      "epoch": 10.601525945285166,
+      "grad_norm": 3.460050344467163,
+      "learning_rate": 5.858165759121897e-05,
+      "loss": 3.3917,
+      "step": 251500
+    },
+    {
+      "epoch": 10.622602537621717,
+      "grad_norm": 2.5006661415100098,
+      "learning_rate": 5.8495905329503064e-05,
+      "loss": 3.3959,
+      "step": 252000
+    },
+    {
+      "epoch": 10.643679129958269,
+      "grad_norm": 1.8837177753448486,
+      "learning_rate": 5.841015306778716e-05,
+      "loss": 3.3963,
+      "step": 252500
+    },
+    {
+      "epoch": 10.664755722294819,
+      "grad_norm": 2.2483580112457275,
+      "learning_rate": 5.832440080607127e-05,
+      "loss": 3.3973,
+      "step": 253000
+    },
+    {
+      "epoch": 10.685832314631371,
+      "grad_norm": 3.864516258239746,
+      "learning_rate": 5.823864854435536e-05,
+      "loss": 3.3907,
+      "step": 253500
+    },
+    {
+      "epoch": 10.706908906967922,
+      "grad_norm": 1.649067759513855,
+      "learning_rate": 5.815289628263946e-05,
+      "loss": 3.3941,
+      "step": 254000
+    },
+    {
+      "epoch": 10.727985499304472,
+      "grad_norm": 2.371920585632324,
+      "learning_rate": 5.8067144020923556e-05,
+      "loss": 3.3987,
+      "step": 254500
+    },
+    {
+      "epoch": 10.749062091641024,
+      "grad_norm": 1.98582124710083,
+      "learning_rate": 5.798139175920765e-05,
+      "loss": 3.3929,
+      "step": 255000
+    },
+    {
+      "epoch": 10.770138683977574,
+      "grad_norm": 2.4473342895507812,
+      "learning_rate": 5.789563949749175e-05,
+      "loss": 3.3949,
+      "step": 255500
+    },
+    {
+      "epoch": 10.791215276314126,
+      "grad_norm": 2.7707066535949707,
+      "learning_rate": 5.780988723577585e-05,
+      "loss": 3.3942,
+      "step": 256000
+    },
+    {
+      "epoch": 10.812291868650677,
+      "grad_norm": 1.7143223285675049,
+      "learning_rate": 5.772413497405994e-05,
+      "loss": 3.3961,
+      "step": 256500
+    },
+    {
+      "epoch": 10.833368460987227,
+      "grad_norm": 1.9074887037277222,
+      "learning_rate": 5.763838271234404e-05,
+      "loss": 3.3924,
+      "step": 257000
+    },
+    {
+      "epoch": 10.85444505332378,
+      "grad_norm": 3.1995513439178467,
+      "learning_rate": 5.7552630450628146e-05,
+      "loss": 3.3932,
+      "step": 257500
+    },
+    {
+      "epoch": 10.87552164566033,
+      "grad_norm": 2.8397433757781982,
+      "learning_rate": 5.746687818891223e-05,
+      "loss": 3.3909,
+      "step": 258000
+    },
+    {
+      "epoch": 10.89659823799688,
+      "grad_norm": 1.9134889841079712,
+      "learning_rate": 5.7381125927196335e-05,
+      "loss": 3.3926,
+      "step": 258500
+    },
+    {
+      "epoch": 10.917674830333432,
+      "grad_norm": 2.0516281127929688,
+      "learning_rate": 5.7295373665480434e-05,
+      "loss": 3.392,
+      "step": 259000
+    },
+    {
+      "epoch": 10.938751422669982,
+      "grad_norm": 2.3784446716308594,
+      "learning_rate": 5.7209621403764525e-05,
+      "loss": 3.3897,
+      "step": 259500
+    },
+    {
+      "epoch": 10.959828015006533,
+      "grad_norm": 2.5315377712249756,
+      "learning_rate": 5.7123869142048624e-05,
+      "loss": 3.3879,
+      "step": 260000
+    },
+    {
+      "epoch": 10.959828015006533,
+      "eval_accuracy": 0.6781287393542702,
+      "eval_loss": 3.11572003364563,
+      "eval_runtime": 169.4409,
+      "eval_samples_per_second": 712.378,
+      "eval_steps_per_second": 7.424,
+      "step": 260000
+    },
+    {
+      "epoch": 10.980904607343085,
+      "grad_norm": 1.8317375183105469,
+      "learning_rate": 5.703811688033273e-05,
+      "loss": 3.3869,
+      "step": 260500
+    },
+    {
+      "epoch": 11.001981199679635,
+      "grad_norm": 3.8753769397735596,
+      "learning_rate": 5.695236461861681e-05,
+      "loss": 3.3888,
+      "step": 261000
+    },
+    {
+      "epoch": 11.023057792016187,
+      "grad_norm": 2.162994146347046,
+      "learning_rate": 5.686661235690092e-05,
+      "loss": 3.3793,
+      "step": 261500
+    },
+    {
+      "epoch": 11.044134384352738,
+      "grad_norm": 2.1795506477355957,
+      "learning_rate": 5.678086009518502e-05,
+      "loss": 3.3863,
+      "step": 262000
+    },
+    {
+      "epoch": 11.065210976689288,
+      "grad_norm": 2.295637607574463,
+      "learning_rate": 5.669510783346911e-05,
+      "loss": 3.3859,
+      "step": 262500
+    },
+    {
+      "epoch": 11.08628756902584,
+      "grad_norm": 3.8470888137817383,
+      "learning_rate": 5.6609355571753207e-05,
+      "loss": 3.3802,
+      "step": 263000
+    },
+    {
+      "epoch": 11.10736416136239,
+      "grad_norm": 3.033310890197754,
+      "learning_rate": 5.652360331003731e-05,
+      "loss": 3.3835,
+      "step": 263500
+    },
+    {
+      "epoch": 11.128440753698943,
+      "grad_norm": 2.4268572330474854,
+      "learning_rate": 5.6437851048321396e-05,
+      "loss": 3.3795,
+      "step": 264000
+    },
+    {
+      "epoch": 11.149517346035493,
+      "grad_norm": 2.3423731327056885,
+      "learning_rate": 5.63520987866055e-05,
+      "loss": 3.386,
+      "step": 264500
+    },
+    {
+      "epoch": 11.170593938372043,
+      "grad_norm": 3.301711082458496,
+      "learning_rate": 5.62663465248896e-05,
+      "loss": 3.3833,
+      "step": 265000
+    },
+    {
+      "epoch": 11.191670530708596,
+      "grad_norm": 1.815821647644043,
+      "learning_rate": 5.618059426317369e-05,
+      "loss": 3.379,
+      "step": 265500
+    },
+    {
+      "epoch": 11.212747123045146,
+      "grad_norm": 2.718214511871338,
+      "learning_rate": 5.609484200145779e-05,
+      "loss": 3.3855,
+      "step": 266000
+    },
+    {
+      "epoch": 11.233823715381696,
+      "grad_norm": 2.1291420459747314,
+      "learning_rate": 5.6009089739741895e-05,
+      "loss": 3.385,
+      "step": 266500
+    },
+    {
+      "epoch": 11.254900307718248,
+      "grad_norm": 3.0063846111297607,
+      "learning_rate": 5.5923337478025986e-05,
+      "loss": 3.381,
+      "step": 267000
+    },
+    {
+      "epoch": 11.275976900054799,
+      "grad_norm": 3.3212552070617676,
+      "learning_rate": 5.5837585216310085e-05,
+      "loss": 3.3849,
+      "step": 267500
+    },
+    {
+      "epoch": 11.29705349239135,
+      "grad_norm": 2.539224863052368,
+      "learning_rate": 5.575183295459418e-05,
+      "loss": 3.3798,
+      "step": 268000
+    },
+    {
+      "epoch": 11.318130084727901,
+      "grad_norm": 2.3344130516052246,
+      "learning_rate": 5.5666080692878274e-05,
+      "loss": 3.3824,
+      "step": 268500
+    },
+    {
+      "epoch": 11.339206677064452,
+      "grad_norm": 3.473512649536133,
+      "learning_rate": 5.558032843116238e-05,
+      "loss": 3.3831,
+      "step": 269000
+    },
+    {
+      "epoch": 11.360283269401004,
+      "grad_norm": 3.2008068561553955,
+      "learning_rate": 5.549457616944648e-05,
+      "loss": 3.3819,
+      "step": 269500
+    },
+    {
+      "epoch": 11.381359861737554,
+      "grad_norm": 1.7833986282348633,
+      "learning_rate": 5.540882390773057e-05,
+      "loss": 3.3778,
+      "step": 270000
+    },
+    {
+      "epoch": 11.381359861737554,
+      "eval_accuracy": 0.6780095185437774,
+      "eval_loss": 3.117027997970581,
+      "eval_runtime": 170.8211,
+      "eval_samples_per_second": 706.622,
+      "eval_steps_per_second": 7.364,
+      "step": 270000
+    },
+    {
+      "epoch": 11.402436454074106,
+      "grad_norm": 2.1491875648498535,
+      "learning_rate": 5.532307164601467e-05,
+      "loss": 3.3841,
+      "step": 270500
+    },
+    {
+      "epoch": 11.423513046410656,
+      "grad_norm": 1.849218487739563,
+      "learning_rate": 5.5237319384298766e-05,
+      "loss": 3.3792,
+      "step": 271000
+    },
+    {
+      "epoch": 11.444589638747207,
+      "grad_norm": 2.3126378059387207,
+      "learning_rate": 5.515156712258286e-05,
+      "loss": 3.3778,
+      "step": 271500
+    },
+    {
+      "epoch": 11.465666231083759,
+      "grad_norm": 2.970428228378296,
+      "learning_rate": 5.506581486086696e-05,
+      "loss": 3.3801,
+      "step": 272000
+    },
+    {
+      "epoch": 11.48674282342031,
+      "grad_norm": 2.5138731002807617,
+      "learning_rate": 5.498006259915105e-05,
+      "loss": 3.3815,
+      "step": 272500
+    },
+    {
+      "epoch": 11.50781941575686,
+      "grad_norm": 2.6348447799682617,
+      "learning_rate": 5.489431033743515e-05,
+      "loss": 3.3818,
+      "step": 273000
+    },
+    {
+      "epoch": 11.528896008093412,
+      "grad_norm": 1.474714994430542,
+      "learning_rate": 5.480855807571925e-05,
+      "loss": 3.3831,
+      "step": 273500
+    },
+    {
+      "epoch": 11.549972600429962,
+      "grad_norm": 3.386077404022217,
+      "learning_rate": 5.472280581400334e-05,
+      "loss": 3.3791,
+      "step": 274000
+    },
+    {
+      "epoch": 11.571049192766514,
+      "grad_norm": 1.4705528020858765,
+      "learning_rate": 5.463705355228744e-05,
+      "loss": 3.3795,
+      "step": 274500
+    },
+    {
+      "epoch": 11.592125785103065,
+      "grad_norm": 3.055100679397583,
+      "learning_rate": 5.4551301290571546e-05,
+      "loss": 3.3829,
+      "step": 275000
+    },
+    {
+      "epoch": 11.613202377439615,
+      "grad_norm": 1.8801623582839966,
+      "learning_rate": 5.446554902885563e-05,
+      "loss": 3.3811,
+      "step": 275500
+    },
+    {
+      "epoch": 11.634278969776167,
+      "grad_norm": 3.8939247131347656,
+      "learning_rate": 5.4379796767139736e-05,
+      "loss": 3.3788,
+      "step": 276000
+    },
+    {
+      "epoch": 11.655355562112717,
+      "grad_norm": 1.9519978761672974,
+      "learning_rate": 5.4294044505423834e-05,
+      "loss": 3.3778,
+      "step": 276500
+    },
+    {
+      "epoch": 11.67643215444927,
+      "grad_norm": 2.2114455699920654,
+      "learning_rate": 5.4208292243707925e-05,
+      "loss": 3.3789,
+      "step": 277000
+    },
+    {
+      "epoch": 11.69750874678582,
+      "grad_norm": 1.6959408521652222,
+      "learning_rate": 5.4122539981992024e-05,
+      "loss": 3.3805,
+      "step": 277500
+    },
+    {
+      "epoch": 11.71858533912237,
+      "grad_norm": 2.8530588150024414,
+      "learning_rate": 5.403678772027613e-05,
+      "loss": 3.3754,
+      "step": 278000
+    },
+    {
+      "epoch": 11.739661931458922,
+      "grad_norm": 2.2549824714660645,
+      "learning_rate": 5.395103545856022e-05,
+      "loss": 3.3798,
+      "step": 278500
+    },
+    {
+      "epoch": 11.760738523795473,
+      "grad_norm": 2.953165292739868,
+      "learning_rate": 5.386528319684432e-05,
+      "loss": 3.3759,
+      "step": 279000
+    },
+    {
+      "epoch": 11.781815116132023,
+      "grad_norm": 2.352945327758789,
+      "learning_rate": 5.377953093512842e-05,
+      "loss": 3.3754,
+      "step": 279500
+    },
+    {
+      "epoch": 11.802891708468575,
+      "grad_norm": 2.080897569656372,
+      "learning_rate": 5.369377867341251e-05,
+      "loss": 3.3824,
+      "step": 280000
+    },
+    {
+      "epoch": 11.802891708468575,
+      "eval_accuracy": 0.6796368512961544,
+      "eval_loss": 3.10966420173645,
+      "eval_runtime": 169.5579,
+      "eval_samples_per_second": 711.887,
+      "eval_steps_per_second": 7.419,
+      "step": 280000
+    },
+    {
+      "epoch": 11.823968300805126,
+      "grad_norm": 2.5307400226593018,
+      "learning_rate": 5.3608026411696614e-05,
+      "loss": 3.3822,
+      "step": 280500
+    },
+    {
+      "epoch": 11.845044893141678,
+      "grad_norm": 2.172370433807373,
+      "learning_rate": 5.352227414998071e-05,
+      "loss": 3.3724,
+      "step": 281000
+    },
+    {
+      "epoch": 11.866121485478228,
+      "grad_norm": 1.8301374912261963,
+      "learning_rate": 5.34365218882648e-05,
+      "loss": 3.3759,
+      "step": 281500
+    },
+    {
+      "epoch": 11.887198077814778,
+      "grad_norm": 1.5306801795959473,
+      "learning_rate": 5.33507696265489e-05,
+      "loss": 3.3703,
+      "step": 282000
+    },
+    {
+      "epoch": 11.90827467015133,
+      "grad_norm": 2.749340295791626,
+      "learning_rate": 5.3265017364833e-05,
+      "loss": 3.379,
+      "step": 282500
+    },
+    {
+      "epoch": 11.92935126248788,
+      "grad_norm": 2.2288589477539062,
+      "learning_rate": 5.317926510311709e-05,
+      "loss": 3.3752,
+      "step": 283000
+    },
+    {
+      "epoch": 11.950427854824433,
+      "grad_norm": 2.842214822769165,
+      "learning_rate": 5.3093512841401197e-05,
+      "loss": 3.3736,
+      "step": 283500
+    },
+    {
+      "epoch": 11.971504447160983,
+      "grad_norm": 2.6877601146698,
+      "learning_rate": 5.3007760579685295e-05,
+      "loss": 3.3782,
+      "step": 284000
+    },
+    {
+      "epoch": 11.992581039497534,
+      "grad_norm": 1.6082152128219604,
+      "learning_rate": 5.2922008317969386e-05,
+      "loss": 3.3822,
+      "step": 284500
+    },
+    {
+      "epoch": 12.013657631834086,
+      "grad_norm": 2.300049304962158,
+      "learning_rate": 5.2836256056253485e-05,
+      "loss": 3.3689,
+      "step": 285000
+    },
+    {
+      "epoch": 12.034734224170636,
+      "grad_norm": 2.58770751953125,
+      "learning_rate": 5.275050379453759e-05,
+      "loss": 3.3668,
+      "step": 285500
+    },
+    {
+      "epoch": 12.055810816507186,
+      "grad_norm": 1.9373024702072144,
+      "learning_rate": 5.2664751532821675e-05,
+      "loss": 3.3644,
+      "step": 286000
+    },
+    {
+      "epoch": 12.076887408843739,
+      "grad_norm": 2.5758776664733887,
+      "learning_rate": 5.257899927110578e-05,
+      "loss": 3.3658,
+      "step": 286500
+    },
+    {
+      "epoch": 12.097964001180289,
+      "grad_norm": 2.1495437622070312,
+      "learning_rate": 5.249324700938988e-05,
+      "loss": 3.3632,
+      "step": 287000
+    },
+    {
+      "epoch": 12.119040593516841,
+      "grad_norm": 1.8603123426437378,
+      "learning_rate": 5.240749474767397e-05,
+      "loss": 3.3627,
+      "step": 287500
+    },
+    {
+      "epoch": 12.140117185853391,
+      "grad_norm": 2.555169105529785,
+      "learning_rate": 5.232174248595807e-05,
+      "loss": 3.3682,
+      "step": 288000
+    },
+    {
+      "epoch": 12.161193778189942,
+      "grad_norm": 1.6816571950912476,
+      "learning_rate": 5.223599022424217e-05,
+      "loss": 3.3682,
+      "step": 288500
+    },
+    {
+      "epoch": 12.182270370526494,
+      "grad_norm": 1.8761543035507202,
+      "learning_rate": 5.215023796252626e-05,
+      "loss": 3.3717,
+      "step": 289000
+    },
+    {
+      "epoch": 12.203346962863044,
+      "grad_norm": 1.9936304092407227,
+      "learning_rate": 5.206448570081036e-05,
+      "loss": 3.363,
+      "step": 289500
+    },
+    {
+      "epoch": 12.224423555199595,
+      "grad_norm": 2.1429715156555176,
+      "learning_rate": 5.197873343909446e-05,
+      "loss": 3.3619,
+      "step": 290000
+    },
+    {
+      "epoch": 12.224423555199595,
+      "eval_accuracy": 0.6801408694617069,
+      "eval_loss": 3.111128568649292,
+      "eval_runtime": 170.3478,
+      "eval_samples_per_second": 708.586,
+      "eval_steps_per_second": 7.385,
+      "step": 290000
+    },
+    {
+      "epoch": 12.245500147536147,
+      "grad_norm": 1.8628485202789307,
+      "learning_rate": 5.189298117737855e-05,
+      "loss": 3.3658,
+      "step": 290500
+    },
+    {
+      "epoch": 12.266576739872697,
+      "grad_norm": 2.5187036991119385,
+      "learning_rate": 5.180722891566265e-05,
+      "loss": 3.3663,
+      "step": 291000
+    },
+    {
+      "epoch": 12.28765333220925,
+      "grad_norm": 2.8511927127838135,
+      "learning_rate": 5.1721476653946756e-05,
+      "loss": 3.3609,
+      "step": 291500
+    },
+    {
+      "epoch": 12.3087299245458,
+      "grad_norm": 2.7446353435516357,
+      "learning_rate": 5.163572439223085e-05,
+      "loss": 3.3656,
+      "step": 292000
+    },
+    {
+      "epoch": 12.32980651688235,
+      "grad_norm": 1.5571297407150269,
+      "learning_rate": 5.1549972130514946e-05,
+      "loss": 3.3669,
+      "step": 292500
+    },
+    {
+      "epoch": 12.350883109218902,
+      "grad_norm": 2.1858420372009277,
+      "learning_rate": 5.1464219868799044e-05,
+      "loss": 3.3662,
+      "step": 293000
+    },
+    {
+      "epoch": 12.371959701555452,
+      "grad_norm": 3.6336138248443604,
+      "learning_rate": 5.1378467607083136e-05,
+      "loss": 3.3629,
+      "step": 293500
+    },
+    {
+      "epoch": 12.393036293892003,
+      "grad_norm": 2.2869255542755127,
+      "learning_rate": 5.1292715345367234e-05,
+      "loss": 3.3653,
+      "step": 294000
+    },
+    {
+      "epoch": 12.414112886228555,
+      "grad_norm": 3.0545709133148193,
+      "learning_rate": 5.120696308365134e-05,
+      "loss": 3.3637,
+      "step": 294500
+    },
+    {
+      "epoch": 12.435189478565105,
+      "grad_norm": 2.401414632797241,
+      "learning_rate": 5.112121082193543e-05,
+      "loss": 3.3681,
+      "step": 295000
+    },
+    {
+      "epoch": 12.456266070901657,
+      "grad_norm": 3.205618381500244,
+      "learning_rate": 5.103545856021953e-05,
+      "loss": 3.365,
+      "step": 295500
+    },
+    {
+      "epoch": 12.477342663238208,
+      "grad_norm": 2.1464741230010986,
+      "learning_rate": 5.094970629850363e-05,
+      "loss": 3.3648,
+      "step": 296000
+    },
+    {
+      "epoch": 12.498419255574758,
+      "grad_norm": 1.9436228275299072,
+      "learning_rate": 5.086395403678772e-05,
+      "loss": 3.3668,
+      "step": 296500
+    },
+    {
+      "epoch": 12.51949584791131,
+      "grad_norm": 2.8417422771453857,
+      "learning_rate": 5.0778201775071824e-05,
+      "loss": 3.3732,
+      "step": 297000
+    },
+    {
+      "epoch": 12.54057244024786,
+      "grad_norm": 2.2397570610046387,
+      "learning_rate": 5.069244951335592e-05,
+      "loss": 3.3697,
+      "step": 297500
+    },
+    {
+      "epoch": 12.561649032584413,
+      "grad_norm": 3.196122169494629,
+      "learning_rate": 5.0606697251640014e-05,
+      "loss": 3.363,
+      "step": 298000
+    },
+    {
+      "epoch": 12.582725624920963,
+      "grad_norm": 2.1866040229797363,
+      "learning_rate": 5.052094498992411e-05,
+      "loss": 3.3677,
+      "step": 298500
+    },
+    {
+      "epoch": 12.603802217257513,
+      "grad_norm": 2.501305103302002,
+      "learning_rate": 5.043519272820822e-05,
+      "loss": 3.3681,
+      "step": 299000
+    },
+    {
+      "epoch": 12.624878809594065,
+      "grad_norm": 3.0755064487457275,
+      "learning_rate": 5.03494404664923e-05,
+      "loss": 3.3614,
+      "step": 299500
+    },
+    {
+      "epoch": 12.645955401930616,
+      "grad_norm": 3.6725895404815674,
+      "learning_rate": 5.026368820477641e-05,
+      "loss": 3.3641,
+      "step": 300000
+    },
+    {
+      "epoch": 12.645955401930616,
+      "eval_accuracy": 0.6809297276579789,
+      "eval_loss": 3.1049678325653076,
+      "eval_runtime": 170.7562,
+      "eval_samples_per_second": 706.891,
+      "eval_steps_per_second": 7.367,
+      "step": 300000
+    },
+    {
+      "epoch": 12.667031994267166,
+      "grad_norm": 3.133557081222534,
+      "learning_rate": 5.0177935943060505e-05,
+      "loss": 3.3608,
+      "step": 300500
+    },
+    {
+      "epoch": 12.688108586603718,
+      "grad_norm": 2.6127982139587402,
+      "learning_rate": 5.00921836813446e-05,
+      "loss": 3.3672,
+      "step": 301000
+    },
+    {
+      "epoch": 12.709185178940269,
+      "grad_norm": 2.641176462173462,
+      "learning_rate": 5.0006431419628695e-05,
+      "loss": 3.3656,
+      "step": 301500
+    },
+    {
+      "epoch": 12.73026177127682,
+      "grad_norm": 2.3598828315734863,
+      "learning_rate": 4.992067915791279e-05,
+      "loss": 3.3653,
+      "step": 302000
+    },
+    {
+      "epoch": 12.751338363613371,
+      "grad_norm": 2.420578956604004,
+      "learning_rate": 4.9834926896196885e-05,
+      "loss": 3.3604,
+      "step": 302500
+    },
+    {
+      "epoch": 12.772414955949921,
+      "grad_norm": 3.1444644927978516,
+      "learning_rate": 4.974917463448099e-05,
+      "loss": 3.3659,
+      "step": 303000
+    },
+    {
+      "epoch": 12.793491548286474,
+      "grad_norm": 1.9409797191619873,
+      "learning_rate": 4.966342237276508e-05,
+      "loss": 3.3668,
+      "step": 303500
+    },
+    {
+      "epoch": 12.814568140623024,
+      "grad_norm": 2.087050437927246,
+      "learning_rate": 4.957767011104918e-05,
+      "loss": 3.3646,
+      "step": 304000
+    },
+    {
+      "epoch": 12.835644732959576,
+      "grad_norm": 3.4791581630706787,
+      "learning_rate": 4.949191784933328e-05,
+      "loss": 3.3646,
+      "step": 304500
+    },
+    {
+      "epoch": 12.856721325296126,
+      "grad_norm": 1.7854629755020142,
+      "learning_rate": 4.9406165587617376e-05,
+      "loss": 3.3621,
+      "step": 305000
+    },
+    {
+      "epoch": 12.877797917632677,
+      "grad_norm": 2.2141644954681396,
+      "learning_rate": 4.932041332590147e-05,
+      "loss": 3.3676,
+      "step": 305500
+    },
+    {
+      "epoch": 12.898874509969229,
+      "grad_norm": 3.517284393310547,
+      "learning_rate": 4.923466106418557e-05,
+      "loss": 3.3615,
+      "step": 306000
+    },
+    {
+      "epoch": 12.91995110230578,
+      "grad_norm": 1.9898682832717896,
+      "learning_rate": 4.9148908802469665e-05,
+      "loss": 3.3595,
+      "step": 306500
+    },
+    {
+      "epoch": 12.94102769464233,
+      "grad_norm": 2.848742723464966,
+      "learning_rate": 4.906315654075376e-05,
+      "loss": 3.3628,
+      "step": 307000
+    },
+    {
+      "epoch": 12.962104286978882,
+      "grad_norm": 3.502206325531006,
+      "learning_rate": 4.897740427903786e-05,
+      "loss": 3.3677,
+      "step": 307500
+    },
+    {
+      "epoch": 12.983180879315432,
+      "grad_norm": 1.9018526077270508,
+      "learning_rate": 4.889165201732196e-05,
+      "loss": 3.3658,
+      "step": 308000
+    },
+    {
+      "epoch": 13.004257471651984,
+      "grad_norm": 3.7067360877990723,
+      "learning_rate": 4.880589975560606e-05,
+      "loss": 3.3604,
+      "step": 308500
+    },
+    {
+      "epoch": 13.025334063988534,
+      "grad_norm": 2.3630428314208984,
+      "learning_rate": 4.8720147493890156e-05,
+      "loss": 3.3442,
+      "step": 309000
+    },
+    {
+      "epoch": 13.046410656325085,
+      "grad_norm": 1.8998318910598755,
+      "learning_rate": 4.8634395232174254e-05,
+      "loss": 3.3495,
+      "step": 309500
+    },
+    {
+      "epoch": 13.067487248661637,
+      "grad_norm": 3.551257848739624,
+      "learning_rate": 4.8548642970458346e-05,
+      "loss": 3.3471,
+      "step": 310000
+    },
+    {
+      "epoch": 13.067487248661637,
+      "eval_accuracy": 0.6812326028129959,
+      "eval_loss": 3.110882520675659,
+      "eval_runtime": 170.1295,
+      "eval_samples_per_second": 709.495,
+      "eval_steps_per_second": 7.394,
+      "step": 310000
+    },
+    {
+      "epoch": 13.088563840998187,
+      "grad_norm": 2.1664552688598633,
+      "learning_rate": 4.846289070874245e-05,
+      "loss": 3.3487,
+      "step": 310500
+    },
+    {
+      "epoch": 13.109640433334738,
+      "grad_norm": 2.583301544189453,
+      "learning_rate": 4.837713844702654e-05,
+      "loss": 3.351,
+      "step": 311000
+    },
+    {
+      "epoch": 13.13071702567129,
+      "grad_norm": 1.9067200422286987,
+      "learning_rate": 4.829138618531064e-05,
+      "loss": 3.3472,
+      "step": 311500
+    },
+    {
+      "epoch": 13.15179361800784,
+      "grad_norm": 2.3606061935424805,
+      "learning_rate": 4.820563392359474e-05,
+      "loss": 3.3514,
+      "step": 312000
+    },
+    {
+      "epoch": 13.172870210344392,
+      "grad_norm": 2.772138833999634,
+      "learning_rate": 4.811988166187884e-05,
+      "loss": 3.349,
+      "step": 312500
+    },
+    {
+      "epoch": 13.193946802680943,
+      "grad_norm": 2.6407580375671387,
+      "learning_rate": 4.803412940016293e-05,
+      "loss": 3.3473,
+      "step": 313000
+    },
+    {
+      "epoch": 13.215023395017493,
+      "grad_norm": 3.1339502334594727,
+      "learning_rate": 4.7948377138447034e-05,
+      "loss": 3.3533,
+      "step": 313500
+    },
+    {
+      "epoch": 13.236099987354045,
+      "grad_norm": 2.5136163234710693,
+      "learning_rate": 4.7862624876731126e-05,
+      "loss": 3.348,
+      "step": 314000
+    },
+    {
+      "epoch": 13.257176579690595,
+      "grad_norm": 3.064723253250122,
+      "learning_rate": 4.7776872615015224e-05,
+      "loss": 3.351,
+      "step": 314500
+    },
+    {
+      "epoch": 13.278253172027147,
+      "grad_norm": 2.7210912704467773,
+      "learning_rate": 4.7691120353299315e-05,
+      "loss": 3.3482,
+      "step": 315000
+    },
+    {
+      "epoch": 13.299329764363698,
+      "grad_norm": 2.4014081954956055,
+      "learning_rate": 4.760536809158342e-05,
+      "loss": 3.354,
+      "step": 315500
+    },
+    {
+      "epoch": 13.320406356700248,
+      "grad_norm": 3.098799705505371,
+      "learning_rate": 4.751961582986751e-05,
+      "loss": 3.3553,
+      "step": 316000
+    },
+    {
+      "epoch": 13.3414829490368,
+      "grad_norm": 3.319927215576172,
+      "learning_rate": 4.743386356815161e-05,
+      "loss": 3.3527,
+      "step": 316500
+    },
+    {
+      "epoch": 13.36255954137335,
+      "grad_norm": 1.9429969787597656,
+      "learning_rate": 4.734811130643571e-05,
+      "loss": 3.3533,
+      "step": 317000
+    },
+    {
+      "epoch": 13.383636133709901,
+      "grad_norm": 2.553861618041992,
+      "learning_rate": 4.726235904471981e-05,
+      "loss": 3.3479,
+      "step": 317500
+    },
+    {
+      "epoch": 13.404712726046453,
+      "grad_norm": 2.4669454097747803,
+      "learning_rate": 4.71766067830039e-05,
+      "loss": 3.3519,
+      "step": 318000
+    },
+    {
+      "epoch": 13.425789318383003,
+      "grad_norm": 2.1440346240997314,
+      "learning_rate": 4.7090854521288004e-05,
+      "loss": 3.354,
+      "step": 318500
+    },
+    {
+      "epoch": 13.446865910719556,
+      "grad_norm": 2.532210350036621,
+      "learning_rate": 4.7005102259572095e-05,
+      "loss": 3.352,
+      "step": 319000
+    },
+    {
+      "epoch": 13.467942503056106,
+      "grad_norm": 2.347913980484009,
+      "learning_rate": 4.6919349997856193e-05,
+      "loss": 3.3529,
+      "step": 319500
+    },
+    {
+      "epoch": 13.489019095392656,
+      "grad_norm": 2.301966428756714,
+      "learning_rate": 4.683359773614029e-05,
+      "loss": 3.3512,
+      "step": 320000
+    },
+    {
+      "epoch": 13.489019095392656,
+      "eval_accuracy": 0.6817314188073657,
+      "eval_loss": 3.1055829524993896,
+      "eval_runtime": 170.2751,
+      "eval_samples_per_second": 708.888,
+      "eval_steps_per_second": 7.388,
+      "step": 320000
+    },
+    {
+      "epoch": 13.510095687729208,
+      "grad_norm": 2.18611216545105,
+      "learning_rate": 4.674784547442439e-05,
+      "loss": 3.3578,
+      "step": 320500
+    },
+    {
+      "epoch": 13.531172280065759,
+      "grad_norm": 2.8243513107299805,
+      "learning_rate": 4.666209321270849e-05,
+      "loss": 3.3521,
+      "step": 321000
+    },
+    {
+      "epoch": 13.55224887240231,
+      "grad_norm": 2.2865235805511475,
+      "learning_rate": 4.657634095099259e-05,
+      "loss": 3.3539,
+      "step": 321500
+    },
+    {
+      "epoch": 13.573325464738861,
+      "grad_norm": 2.301863670349121,
+      "learning_rate": 4.6490588689276685e-05,
+      "loss": 3.3527,
+      "step": 322000
+    },
+    {
+      "epoch": 13.594402057075412,
+      "grad_norm": 2.6193535327911377,
+      "learning_rate": 4.6404836427560777e-05,
+      "loss": 3.3546,
+      "step": 322500
+    },
+    {
+      "epoch": 13.615478649411964,
+      "grad_norm": 3.129697561264038,
+      "learning_rate": 4.6319084165844875e-05,
+      "loss": 3.3474,
+      "step": 323000
+    },
+    {
+      "epoch": 13.636555241748514,
+      "grad_norm": 2.0105528831481934,
+      "learning_rate": 4.623333190412897e-05,
+      "loss": 3.3537,
+      "step": 323500
+    },
+    {
+      "epoch": 13.657631834085064,
+      "grad_norm": 3.153481960296631,
+      "learning_rate": 4.614757964241307e-05,
+      "loss": 3.3499,
+      "step": 324000
+    },
+    {
+      "epoch": 13.678708426421617,
+      "grad_norm": 3.809739828109741,
+      "learning_rate": 4.606182738069717e-05,
+      "loss": 3.3525,
+      "step": 324500
+    },
+    {
+      "epoch": 13.699785018758167,
+      "grad_norm": 2.1247236728668213,
+      "learning_rate": 4.597607511898127e-05,
+      "loss": 3.3513,
+      "step": 325000
+    },
+    {
+      "epoch": 13.720861611094719,
+      "grad_norm": 3.640238046646118,
+      "learning_rate": 4.589032285726536e-05,
+      "loss": 3.3528,
+      "step": 325500
+    },
+    {
+      "epoch": 13.74193820343127,
+      "grad_norm": 2.309969425201416,
+      "learning_rate": 4.5804570595549465e-05,
+      "loss": 3.353,
+      "step": 326000
+    },
+    {
+      "epoch": 13.76301479576782,
+      "grad_norm": 2.3356688022613525,
+      "learning_rate": 4.5718818333833556e-05,
+      "loss": 3.3547,
+      "step": 326500
+    },
+    {
+      "epoch": 13.784091388104372,
+      "grad_norm": 2.6970977783203125,
+      "learning_rate": 4.5633066072117655e-05,
+      "loss": 3.3536,
+      "step": 327000
+    },
+    {
+      "epoch": 13.805167980440922,
+      "grad_norm": 2.955115556716919,
+      "learning_rate": 4.554731381040175e-05,
+      "loss": 3.3489,
+      "step": 327500
+    },
+    {
+      "epoch": 13.826244572777473,
+      "grad_norm": 2.4655394554138184,
+      "learning_rate": 4.546156154868585e-05,
+      "loss": 3.3507,
+      "step": 328000
+    },
+    {
+      "epoch": 13.847321165114025,
+      "grad_norm": 2.341932535171509,
+      "learning_rate": 4.537580928696994e-05,
+      "loss": 3.3528,
+      "step": 328500
+    },
+    {
+      "epoch": 13.868397757450575,
+      "grad_norm": 2.6296274662017822,
+      "learning_rate": 4.529005702525405e-05,
+      "loss": 3.3517,
+      "step": 329000
+    },
+    {
+      "epoch": 13.889474349787127,
+      "grad_norm": 1.9726841449737549,
+      "learning_rate": 4.520430476353814e-05,
+      "loss": 3.3524,
+      "step": 329500
+    },
+    {
+      "epoch": 13.910550942123677,
+      "grad_norm": 1.9200328588485718,
+      "learning_rate": 4.511855250182224e-05,
+      "loss": 3.3523,
+      "step": 330000
+    },
+    {
+      "epoch": 13.910550942123677,
+      "eval_accuracy": 0.6819591110263561,
+      "eval_loss": 3.1033778190612793,
+      "eval_runtime": 170.4674,
+      "eval_samples_per_second": 708.088,
+      "eval_steps_per_second": 7.38,
+      "step": 330000
+    },
+    {
+      "epoch": 13.931627534460228,
+      "grad_norm": 1.8566782474517822,
+      "learning_rate": 4.5032800240106336e-05,
+      "loss": 3.353,
+      "step": 330500
+    },
+    {
+      "epoch": 13.95270412679678,
+      "grad_norm": 2.9055211544036865,
+      "learning_rate": 4.4947047978390434e-05,
+      "loss": 3.3561,
+      "step": 331000
+    },
+    {
+      "epoch": 13.97378071913333,
+      "grad_norm": 2.3704004287719727,
+      "learning_rate": 4.4861295716674526e-05,
+      "loss": 3.3525,
+      "step": 331500
+    },
+    {
+      "epoch": 13.994857311469882,
+      "grad_norm": 2.5733187198638916,
+      "learning_rate": 4.4775543454958624e-05,
+      "loss": 3.3543,
+      "step": 332000
+    },
+    {
+      "epoch": 14.015933903806433,
+      "grad_norm": 2.3355889320373535,
+      "learning_rate": 4.468979119324272e-05,
+      "loss": 3.3332,
+      "step": 332500
+    },
+    {
+      "epoch": 14.037010496142983,
+      "grad_norm": 2.750149726867676,
+      "learning_rate": 4.460403893152682e-05,
+      "loss": 3.3329,
+      "step": 333000
+    },
+    {
+      "epoch": 14.058087088479535,
+      "grad_norm": 2.7765588760375977,
+      "learning_rate": 4.451828666981092e-05,
+      "loss": 3.3316,
+      "step": 333500
+    },
+    {
+      "epoch": 14.079163680816086,
+      "grad_norm": 2.9302079677581787,
+      "learning_rate": 4.443253440809502e-05,
+      "loss": 3.3333,
+      "step": 334000
+    },
+    {
+      "epoch": 14.100240273152636,
+      "grad_norm": 2.797461986541748,
+      "learning_rate": 4.4346782146379116e-05,
+      "loss": 3.3333,
+      "step": 334500
+    },
+    {
+      "epoch": 14.121316865489188,
+      "grad_norm": 2.7614829540252686,
+      "learning_rate": 4.426102988466321e-05,
+      "loss": 3.3325,
+      "step": 335000
+    },
+    {
+      "epoch": 14.142393457825738,
+      "grad_norm": 3.5114266872406006,
+      "learning_rate": 4.4175277622947305e-05,
+      "loss": 3.3301,
+      "step": 335500
+    },
+    {
+      "epoch": 14.16347005016229,
+      "grad_norm": 2.927475929260254,
+      "learning_rate": 4.4089525361231404e-05,
+      "loss": 3.3335,
+      "step": 336000
+    },
+    {
+      "epoch": 14.184546642498841,
+      "grad_norm": 2.506596803665161,
+      "learning_rate": 4.40037730995155e-05,
+      "loss": 3.3324,
+      "step": 336500
+    },
+    {
+      "epoch": 14.205623234835391,
+      "grad_norm": 2.9631221294403076,
+      "learning_rate": 4.39180208377996e-05,
+      "loss": 3.3343,
+      "step": 337000
+    },
+    {
+      "epoch": 14.226699827171943,
+      "grad_norm": 2.5132553577423096,
+      "learning_rate": 4.38322685760837e-05,
+      "loss": 3.3363,
+      "step": 337500
+    },
+    {
+      "epoch": 14.247776419508494,
+      "grad_norm": 2.054077625274658,
+      "learning_rate": 4.374651631436779e-05,
+      "loss": 3.3359,
+      "step": 338000
+    },
+    {
+      "epoch": 14.268853011845044,
+      "grad_norm": 2.988914966583252,
+      "learning_rate": 4.3660764052651895e-05,
+      "loss": 3.3392,
+      "step": 338500
+    },
+    {
+      "epoch": 14.289929604181596,
+      "grad_norm": 2.9316554069519043,
+      "learning_rate": 4.357501179093599e-05,
+      "loss": 3.3345,
+      "step": 339000
+    },
+    {
+      "epoch": 14.311006196518147,
+      "grad_norm": 3.0771501064300537,
+      "learning_rate": 4.3489259529220085e-05,
+      "loss": 3.3356,
+      "step": 339500
+    },
+    {
+      "epoch": 14.332082788854699,
+      "grad_norm": 3.116081714630127,
+      "learning_rate": 4.3403507267504183e-05,
+      "loss": 3.3385,
+      "step": 340000
+    },
+    {
+      "epoch": 14.332082788854699,
+      "eval_accuracy": 0.6823742210902608,
+      "eval_loss": 3.1032865047454834,
+      "eval_runtime": 169.6925,
+      "eval_samples_per_second": 711.322,
+      "eval_steps_per_second": 7.413,
+      "step": 340000
+    },
+    {
+      "epoch": 14.353159381191249,
+      "grad_norm": 2.2543790340423584,
+      "learning_rate": 4.331775500578828e-05,
+      "loss": 3.3356,
+      "step": 340500
+    },
+    {
+      "epoch": 14.3742359735278,
+      "grad_norm": 3.443559408187866,
+      "learning_rate": 4.323200274407237e-05,
+      "loss": 3.3358,
+      "step": 341000
+    },
+    {
+      "epoch": 14.395312565864351,
+      "grad_norm": 2.635343313217163,
+      "learning_rate": 4.314625048235648e-05,
+      "loss": 3.3353,
+      "step": 341500
+    },
+    {
+      "epoch": 14.416389158200902,
+      "grad_norm": 3.3518848419189453,
+      "learning_rate": 4.306049822064057e-05,
+      "loss": 3.3414,
+      "step": 342000
+    },
+    {
+      "epoch": 14.437465750537454,
+      "grad_norm": 3.0962071418762207,
+      "learning_rate": 4.297474595892467e-05,
+      "loss": 3.3381,
+      "step": 342500
+    },
+    {
+      "epoch": 14.458542342874004,
+      "grad_norm": 2.133889675140381,
+      "learning_rate": 4.2888993697208767e-05,
+      "loss": 3.3357,
+      "step": 343000
+    },
+    {
+      "epoch": 14.479618935210555,
+      "grad_norm": 2.625239610671997,
+      "learning_rate": 4.2803241435492865e-05,
+      "loss": 3.3393,
+      "step": 343500
+    },
+    {
+      "epoch": 14.500695527547107,
+      "grad_norm": 2.7125182151794434,
+      "learning_rate": 4.2717489173776956e-05,
+      "loss": 3.337,
+      "step": 344000
+    },
+    {
+      "epoch": 14.521772119883657,
+      "grad_norm": 3.7416670322418213,
+      "learning_rate": 4.263173691206106e-05,
+      "loss": 3.34,
+      "step": 344500
+    },
+    {
+      "epoch": 14.542848712220207,
+      "grad_norm": 2.4357383251190186,
+      "learning_rate": 4.254598465034515e-05,
+      "loss": 3.3397,
+      "step": 345000
+    },
+    {
+      "epoch": 14.56392530455676,
+      "grad_norm": 4.409759521484375,
+      "learning_rate": 4.246023238862925e-05,
+      "loss": 3.3374,
+      "step": 345500
+    },
+    {
+      "epoch": 14.58500189689331,
+      "grad_norm": 2.467895746231079,
+      "learning_rate": 4.237448012691335e-05,
+      "loss": 3.3385,
+      "step": 346000
+    },
+    {
+      "epoch": 14.606078489229862,
+      "grad_norm": 2.780771255493164,
+      "learning_rate": 4.228872786519745e-05,
+      "loss": 3.3394,
+      "step": 346500
+    },
+    {
+      "epoch": 14.627155081566412,
+      "grad_norm": 2.6083991527557373,
+      "learning_rate": 4.220297560348154e-05,
+      "loss": 3.3412,
+      "step": 347000
+    },
+    {
+      "epoch": 14.648231673902963,
+      "grad_norm": 2.1286678314208984,
+      "learning_rate": 4.2117223341765645e-05,
+      "loss": 3.3396,
+      "step": 347500
+    },
+    {
+      "epoch": 14.669308266239515,
+      "grad_norm": 2.7418437004089355,
+      "learning_rate": 4.2031471080049736e-05,
+      "loss": 3.3365,
+      "step": 348000
+    },
+    {
+      "epoch": 14.690384858576065,
+      "grad_norm": 2.028881788253784,
+      "learning_rate": 4.1945718818333834e-05,
+      "loss": 3.3391,
+      "step": 348500
+    },
+    {
+      "epoch": 14.711461450912616,
+      "grad_norm": 2.2924909591674805,
+      "learning_rate": 4.185996655661793e-05,
+      "loss": 3.3406,
+      "step": 349000
+    },
+    {
+      "epoch": 14.732538043249168,
+      "grad_norm": 2.0082998275756836,
+      "learning_rate": 4.177421429490203e-05,
+      "loss": 3.3445,
+      "step": 349500
+    },
+    {
+      "epoch": 14.753614635585718,
+      "grad_norm": 4.167133331298828,
+      "learning_rate": 4.168846203318613e-05,
+      "loss": 3.3393,
+      "step": 350000
+    },
+    {
+      "epoch": 14.753614635585718,
+      "eval_accuracy": 0.6828001179401756,
+      "eval_loss": 3.1060609817504883,
+      "eval_runtime": 170.6389,
+      "eval_samples_per_second": 707.377,
+      "eval_steps_per_second": 7.372,
+      "step": 350000
+    },
+    {
+      "epoch": 14.77469122792227,
+      "grad_norm": 2.3272616863250732,
+      "learning_rate": 4.160270977147022e-05,
+      "loss": 3.3385,
+      "step": 350500
+    },
+    {
+      "epoch": 14.79576782025882,
+      "grad_norm": 3.4731385707855225,
+      "learning_rate": 4.1516957509754326e-05,
+      "loss": 3.343,
+      "step": 351000
+    },
+    {
+      "epoch": 14.81684441259537,
+      "grad_norm": 2.8600258827209473,
+      "learning_rate": 4.143120524803842e-05,
+      "loss": 3.3364,
+      "step": 351500
+    },
+    {
+      "epoch": 14.837921004931923,
+      "grad_norm": 2.513681173324585,
+      "learning_rate": 4.1345452986322516e-05,
+      "loss": 3.3364,
+      "step": 352000
+    },
+    {
+      "epoch": 14.858997597268473,
+      "grad_norm": 3.447681188583374,
+      "learning_rate": 4.1259700724606614e-05,
+      "loss": 3.3377,
+      "step": 352500
+    },
+    {
+      "epoch": 14.880074189605025,
+      "grad_norm": 3.6882128715515137,
+      "learning_rate": 4.117394846289071e-05,
+      "loss": 3.338,
+      "step": 353000
+    },
+    {
+      "epoch": 14.901150781941576,
+      "grad_norm": 2.7427830696105957,
+      "learning_rate": 4.1088196201174804e-05,
+      "loss": 3.3417,
+      "step": 353500
+    },
+    {
+      "epoch": 14.922227374278126,
+      "grad_norm": 2.587165355682373,
+      "learning_rate": 4.100244393945891e-05,
+      "loss": 3.3385,
+      "step": 354000
+    },
+    {
+      "epoch": 14.943303966614678,
+      "grad_norm": 2.773124933242798,
+      "learning_rate": 4.0916691677743e-05,
+      "loss": 3.3418,
+      "step": 354500
+    },
+    {
+      "epoch": 14.964380558951229,
+      "grad_norm": 2.291010618209839,
+      "learning_rate": 4.08309394160271e-05,
+      "loss": 3.3406,
+      "step": 355000
+    },
+    {
+      "epoch": 14.985457151287779,
+      "grad_norm": 2.690659523010254,
+      "learning_rate": 4.07451871543112e-05,
+      "loss": 3.342,
+      "step": 355500
+    },
+    {
+      "epoch": 15.006533743624331,
+      "grad_norm": 3.178884983062744,
+      "learning_rate": 4.0659434892595295e-05,
+      "loss": 3.3279,
+      "step": 356000
+    },
+    {
+      "epoch": 15.027610335960881,
+      "grad_norm": 3.6222596168518066,
+      "learning_rate": 4.057368263087939e-05,
+      "loss": 3.3094,
+      "step": 356500
+    },
+    {
+      "epoch": 15.048686928297434,
+      "grad_norm": 2.374471426010132,
+      "learning_rate": 4.048793036916349e-05,
+      "loss": 3.3144,
+      "step": 357000
+    },
+    {
+      "epoch": 15.069763520633984,
+      "grad_norm": 2.788585901260376,
+      "learning_rate": 4.0402178107447584e-05,
+      "loss": 3.3193,
+      "step": 357500
+    },
+    {
+      "epoch": 15.090840112970534,
+      "grad_norm": 3.1044700145721436,
+      "learning_rate": 4.031642584573168e-05,
+      "loss": 3.3171,
+      "step": 358000
+    },
+    {
+      "epoch": 15.111916705307086,
+      "grad_norm": 3.6265785694122314,
+      "learning_rate": 4.023067358401578e-05,
+      "loss": 3.314,
+      "step": 358500
+    },
+    {
+      "epoch": 15.132993297643637,
+      "grad_norm": 2.771571159362793,
+      "learning_rate": 4.014492132229988e-05,
+      "loss": 3.3175,
+      "step": 359000
+    },
+    {
+      "epoch": 15.154069889980189,
+      "grad_norm": 4.2338032722473145,
+      "learning_rate": 4.005916906058397e-05,
+      "loss": 3.3167,
+      "step": 359500
+    },
+    {
+      "epoch": 15.17514648231674,
+      "grad_norm": 2.791841983795166,
+      "learning_rate": 3.9973416798868075e-05,
+      "loss": 3.3183,
+      "step": 360000
+    },
+    {
+      "epoch": 15.17514648231674,
+      "eval_accuracy": 0.682873967889179,
+      "eval_loss": 3.1163697242736816,
+      "eval_runtime": 170.3722,
+      "eval_samples_per_second": 708.484,
+      "eval_steps_per_second": 7.384,
+      "step": 360000
+    },
+    {
+      "epoch": 15.19622307465329,
+      "grad_norm": 2.770967721939087,
+      "learning_rate": 3.988766453715217e-05,
+      "loss": 3.3242,
+      "step": 360500
+    },
+    {
+      "epoch": 15.217299666989842,
+      "grad_norm": 3.354207992553711,
+      "learning_rate": 3.9801912275436265e-05,
+      "loss": 3.3189,
+      "step": 361000
+    },
+    {
+      "epoch": 15.238376259326392,
+      "grad_norm": 4.321069717407227,
+      "learning_rate": 3.971616001372036e-05,
+      "loss": 3.3218,
+      "step": 361500
+    },
+    {
+      "epoch": 15.259452851662942,
+      "grad_norm": 3.3897533416748047,
+      "learning_rate": 3.963040775200446e-05,
+      "loss": 3.3193,
+      "step": 362000
+    },
+    {
+      "epoch": 15.280529443999495,
+      "grad_norm": 3.112244129180908,
+      "learning_rate": 3.954465549028856e-05,
+      "loss": 3.3212,
+      "step": 362500
+    },
+    {
+      "epoch": 15.301606036336045,
+      "grad_norm": 2.899683713912964,
+      "learning_rate": 3.945890322857266e-05,
+      "loss": 3.3235,
+      "step": 363000
+    },
+    {
+      "epoch": 15.322682628672597,
+      "grad_norm": 2.8062589168548584,
+      "learning_rate": 3.9373150966856756e-05,
+      "loss": 3.3183,
+      "step": 363500
+    },
+    {
+      "epoch": 15.343759221009147,
+      "grad_norm": 2.419450521469116,
+      "learning_rate": 3.928739870514085e-05,
+      "loss": 3.324,
+      "step": 364000
+    },
+    {
+      "epoch": 15.364835813345698,
+      "grad_norm": 2.887693405151367,
+      "learning_rate": 3.920164644342495e-05,
+      "loss": 3.3197,
+      "step": 364500
+    },
+    {
+      "epoch": 15.38591240568225,
+      "grad_norm": 2.9862475395202637,
+      "learning_rate": 3.9115894181709045e-05,
+      "loss": 3.324,
+      "step": 365000
+    },
+    {
+      "epoch": 15.4069889980188,
+      "grad_norm": 2.7634437084198,
+      "learning_rate": 3.903014191999314e-05,
+      "loss": 3.3161,
+      "step": 365500
+    },
+    {
+      "epoch": 15.42806559035535,
+      "grad_norm": 2.270430326461792,
+      "learning_rate": 3.894438965827724e-05,
+      "loss": 3.3224,
+      "step": 366000
+    },
+    {
+      "epoch": 15.449142182691903,
+      "grad_norm": 3.033433198928833,
+      "learning_rate": 3.885863739656134e-05,
+      "loss": 3.3237,
+      "step": 366500
+    },
+    {
+      "epoch": 15.470218775028453,
+      "grad_norm": 2.837465763092041,
+      "learning_rate": 3.877288513484543e-05,
+      "loss": 3.3224,
+      "step": 367000
+    },
+    {
+      "epoch": 15.491295367365005,
+      "grad_norm": 2.5261902809143066,
+      "learning_rate": 3.868713287312953e-05,
+      "loss": 3.3249,
+      "step": 367500
+    },
+    {
+      "epoch": 15.512371959701555,
+      "grad_norm": 2.939676523208618,
+      "learning_rate": 3.860138061141363e-05,
+      "loss": 3.3232,
+      "step": 368000
+    },
+    {
+      "epoch": 15.533448552038106,
+      "grad_norm": 2.6163270473480225,
+      "learning_rate": 3.8515628349697726e-05,
+      "loss": 3.3269,
+      "step": 368500
+    },
+    {
+      "epoch": 15.554525144374658,
+      "grad_norm": 2.850574493408203,
+      "learning_rate": 3.842987608798182e-05,
+      "loss": 3.3246,
+      "step": 369000
+    },
+    {
+      "epoch": 15.575601736711208,
+      "grad_norm": 2.5736196041107178,
+      "learning_rate": 3.834412382626592e-05,
+      "loss": 3.3222,
+      "step": 369500
+    },
+    {
+      "epoch": 15.596678329047759,
+      "grad_norm": 2.356257200241089,
+      "learning_rate": 3.8258371564550014e-05,
+      "loss": 3.3261,
+      "step": 370000
+    },
+    {
+      "epoch": 15.596678329047759,
+      "eval_accuracy": 0.6833068547501796,
+      "eval_loss": 3.1081740856170654,
+      "eval_runtime": 169.7462,
+      "eval_samples_per_second": 711.097,
+      "eval_steps_per_second": 7.411,
+      "step": 370000
+    },
+    {
+      "epoch": 15.61775492138431,
+      "grad_norm": 2.3240933418273926,
+      "learning_rate": 3.817261930283411e-05,
+      "loss": 3.3219,
+      "step": 370500
+    },
+    {
+      "epoch": 15.638831513720861,
+      "grad_norm": 3.337608575820923,
+      "learning_rate": 3.808686704111821e-05,
+      "loss": 3.3239,
+      "step": 371000
+    },
+    {
+      "epoch": 15.659908106057413,
+      "grad_norm": 2.8807895183563232,
+      "learning_rate": 3.800111477940231e-05,
+      "loss": 3.3252,
+      "step": 371500
+    },
+    {
+      "epoch": 15.680984698393964,
+      "grad_norm": 2.3612146377563477,
+      "learning_rate": 3.79153625176864e-05,
+      "loss": 3.3265,
+      "step": 372000
+    },
+    {
+      "epoch": 15.702061290730514,
+      "grad_norm": 3.3425514698028564,
+      "learning_rate": 3.7829610255970506e-05,
+      "loss": 3.3304,
+      "step": 372500
+    },
+    {
+      "epoch": 15.723137883067066,
+      "grad_norm": 3.067561626434326,
+      "learning_rate": 3.77438579942546e-05,
+      "loss": 3.3236,
+      "step": 373000
+    },
+    {
+      "epoch": 15.744214475403616,
+      "grad_norm": 3.2922592163085938,
+      "learning_rate": 3.7658105732538696e-05,
+      "loss": 3.3197,
+      "step": 373500
+    },
+    {
+      "epoch": 15.765291067740169,
+      "grad_norm": 2.8093509674072266,
+      "learning_rate": 3.7572353470822794e-05,
+      "loss": 3.3268,
+      "step": 374000
+    },
+    {
+      "epoch": 15.786367660076719,
+      "grad_norm": 2.381267786026001,
+      "learning_rate": 3.748660120910689e-05,
+      "loss": 3.3242,
+      "step": 374500
+    },
+    {
+      "epoch": 15.80744425241327,
+      "grad_norm": 3.9539859294891357,
+      "learning_rate": 3.740084894739099e-05,
+      "loss": 3.3202,
+      "step": 375000
+    },
+    {
+      "epoch": 15.828520844749821,
+      "grad_norm": 3.488492012023926,
+      "learning_rate": 3.731509668567509e-05,
+      "loss": 3.3232,
+      "step": 375500
+    },
+    {
+      "epoch": 15.849597437086372,
+      "grad_norm": 2.8233420848846436,
+      "learning_rate": 3.722934442395919e-05,
+      "loss": 3.3288,
+      "step": 376000
+    },
+    {
+      "epoch": 15.870674029422922,
+      "grad_norm": 3.63529109954834,
+      "learning_rate": 3.714359216224328e-05,
+      "loss": 3.3266,
+      "step": 376500
+    },
+    {
+      "epoch": 15.891750621759474,
+      "grad_norm": 2.4166970252990723,
+      "learning_rate": 3.705783990052738e-05,
+      "loss": 3.3254,
+      "step": 377000
+    },
+    {
+      "epoch": 15.912827214096025,
+      "grad_norm": 3.4087672233581543,
+      "learning_rate": 3.6972087638811475e-05,
+      "loss": 3.3231,
+      "step": 377500
+    },
+    {
+      "epoch": 15.933903806432577,
+      "grad_norm": 4.156463623046875,
+      "learning_rate": 3.6886335377095574e-05,
+      "loss": 3.3269,
+      "step": 378000
+    },
+    {
+      "epoch": 15.954980398769127,
+      "grad_norm": 2.6670656204223633,
+      "learning_rate": 3.680058311537967e-05,
+      "loss": 3.3247,
+      "step": 378500
+    },
+    {
+      "epoch": 15.976056991105677,
+      "grad_norm": 2.7266945838928223,
+      "learning_rate": 3.671483085366377e-05,
+      "loss": 3.3272,
+      "step": 379000
+    },
+    {
+      "epoch": 15.99713358344223,
+      "grad_norm": 2.2741847038269043,
+      "learning_rate": 3.662907859194786e-05,
+      "loss": 3.3217,
+      "step": 379500
+    },
+    {
+      "epoch": 16.01821017577878,
+      "grad_norm": 3.330495595932007,
+      "learning_rate": 3.654332633023197e-05,
+      "loss": 3.2993,
+      "step": 380000
+    },
+    {
+      "epoch": 16.01821017577878,
+      "eval_accuracy": 0.683746544544867,
+      "eval_loss": 3.1229801177978516,
+      "eval_runtime": 170.2962,
+      "eval_samples_per_second": 708.8,
+      "eval_steps_per_second": 7.387,
+      "step": 380000
+    },
+    {
+      "epoch": 16.039286768115332,
+      "grad_norm": 2.614999532699585,
+      "learning_rate": 3.645757406851606e-05,
+      "loss": 3.2957,
+      "step": 380500
+    },
+    {
+      "epoch": 16.06036336045188,
+      "grad_norm": 3.3745853900909424,
+      "learning_rate": 3.6371821806800157e-05,
+      "loss": 3.295,
+      "step": 381000
+    },
+    {
+      "epoch": 16.081439952788433,
+      "grad_norm": 2.984684705734253,
+      "learning_rate": 3.6286069545084255e-05,
+      "loss": 3.3003,
+      "step": 381500
+    },
+    {
+      "epoch": 16.102516545124985,
+      "grad_norm": 3.076197862625122,
+      "learning_rate": 3.620031728336835e-05,
+      "loss": 3.2996,
+      "step": 382000
+    },
+    {
+      "epoch": 16.123593137461537,
+      "grad_norm": 3.6839046478271484,
+      "learning_rate": 3.6114565021652445e-05,
+      "loss": 3.2917,
+      "step": 382500
+    },
+    {
+      "epoch": 16.144669729798085,
+      "grad_norm": 2.494629144668579,
+      "learning_rate": 3.602881275993655e-05,
+      "loss": 3.2988,
+      "step": 383000
+    },
+    {
+      "epoch": 16.165746322134638,
+      "grad_norm": 3.2421014308929443,
+      "learning_rate": 3.594306049822064e-05,
+      "loss": 3.3015,
+      "step": 383500
+    },
+    {
+      "epoch": 16.18682291447119,
+      "grad_norm": 3.225808620452881,
+      "learning_rate": 3.585730823650474e-05,
+      "loss": 3.2969,
+      "step": 384000
+    },
+    {
+      "epoch": 16.20789950680774,
+      "grad_norm": 3.7454004287719727,
+      "learning_rate": 3.577155597478883e-05,
+      "loss": 3.3004,
+      "step": 384500
+    },
+    {
+      "epoch": 16.22897609914429,
+      "grad_norm": 2.819350004196167,
+      "learning_rate": 3.5685803713072936e-05,
+      "loss": 3.2973,
+      "step": 385000
+    },
+    {
+      "epoch": 16.250052691480843,
+      "grad_norm": 3.053962230682373,
+      "learning_rate": 3.560005145135703e-05,
+      "loss": 3.3022,
+      "step": 385500
+    },
+    {
+      "epoch": 16.27112928381739,
+      "grad_norm": 3.5590388774871826,
+      "learning_rate": 3.5514299189641126e-05,
+      "loss": 3.2969,
+      "step": 386000
+    },
+    {
+      "epoch": 16.292205876153943,
+      "grad_norm": 2.316598415374756,
+      "learning_rate": 3.5428546927925224e-05,
+      "loss": 3.3058,
+      "step": 386500
+    },
+    {
+      "epoch": 16.313282468490495,
+      "grad_norm": 3.3866915702819824,
+      "learning_rate": 3.534279466620932e-05,
+      "loss": 3.2998,
+      "step": 387000
+    },
+    {
+      "epoch": 16.334359060827044,
+      "grad_norm": 3.135079860687256,
+      "learning_rate": 3.525704240449342e-05,
+      "loss": 3.3032,
+      "step": 387500
+    },
+    {
+      "epoch": 16.355435653163596,
+      "grad_norm": 3.351874828338623,
+      "learning_rate": 3.517129014277752e-05,
+      "loss": 3.3007,
+      "step": 388000
+    },
+    {
+      "epoch": 16.376512245500148,
+      "grad_norm": 3.098057985305786,
+      "learning_rate": 3.508553788106161e-05,
+      "loss": 3.3038,
+      "step": 388500
+    },
+    {
+      "epoch": 16.3975888378367,
+      "grad_norm": 3.1159565448760986,
+      "learning_rate": 3.499978561934571e-05,
+      "loss": 3.3015,
+      "step": 389000
+    },
+    {
+      "epoch": 16.41866543017325,
+      "grad_norm": 3.3988399505615234,
+      "learning_rate": 3.491403335762981e-05,
+      "loss": 3.3029,
+      "step": 389500
+    },
+    {
+      "epoch": 16.4397420225098,
+      "grad_norm": 2.4276998043060303,
+      "learning_rate": 3.4828281095913906e-05,
+      "loss": 3.3079,
+      "step": 390000
+    },
+    {
+      "epoch": 16.4397420225098,
+      "eval_accuracy": 0.6836240310077519,
+      "eval_loss": 3.1179006099700928,
+      "eval_runtime": 170.9751,
+      "eval_samples_per_second": 705.986,
+      "eval_steps_per_second": 7.358,
+      "step": 390000
+    },
+    {
+      "epoch": 16.460818614846353,
+      "grad_norm": 3.078444719314575,
+      "learning_rate": 3.4742528834198004e-05,
+      "loss": 3.3029,
+      "step": 390500
+    },
+    {
+      "epoch": 16.4818952071829,
+      "grad_norm": 2.98537540435791,
+      "learning_rate": 3.46567765724821e-05,
+      "loss": 3.309,
+      "step": 391000
+    },
+    {
+      "epoch": 16.502971799519454,
+      "grad_norm": 3.862744092941284,
+      "learning_rate": 3.45710243107662e-05,
+      "loss": 3.3039,
+      "step": 391500
+    },
+    {
+      "epoch": 16.524048391856006,
+      "grad_norm": 2.548874855041504,
+      "learning_rate": 3.448527204905029e-05,
+      "loss": 3.3097,
+      "step": 392000
+    },
+    {
+      "epoch": 16.545124984192555,
+      "grad_norm": 2.5996193885803223,
+      "learning_rate": 3.43995197873344e-05,
+      "loss": 3.3032,
+      "step": 392500
+    },
+    {
+      "epoch": 16.566201576529107,
+      "grad_norm": 3.4460301399230957,
+      "learning_rate": 3.431376752561849e-05,
+      "loss": 3.3068,
+      "step": 393000
+    },
+    {
+      "epoch": 16.58727816886566,
+      "grad_norm": 3.088254928588867,
+      "learning_rate": 3.422801526390259e-05,
+      "loss": 3.3105,
+      "step": 393500
+    },
+    {
+      "epoch": 16.608354761202207,
+      "grad_norm": 3.3189446926116943,
+      "learning_rate": 3.4142263002186686e-05,
+      "loss": 3.3092,
+      "step": 394000
+    },
+    {
+      "epoch": 16.62943135353876,
+      "grad_norm": 2.438575029373169,
+      "learning_rate": 3.4056510740470784e-05,
+      "loss": 3.3134,
+      "step": 394500
+    },
+    {
+      "epoch": 16.65050794587531,
+      "grad_norm": 2.1685843467712402,
+      "learning_rate": 3.3970758478754875e-05,
+      "loss": 3.3107,
+      "step": 395000
+    },
+    {
+      "epoch": 16.67158453821186,
+      "grad_norm": 2.4223270416259766,
+      "learning_rate": 3.388500621703898e-05,
+      "loss": 3.3104,
+      "step": 395500
+    },
+    {
+      "epoch": 16.692661130548412,
+      "grad_norm": 3.374260902404785,
+      "learning_rate": 3.379925395532307e-05,
+      "loss": 3.3111,
+      "step": 396000
+    },
+    {
+      "epoch": 16.713737722884964,
+      "grad_norm": 3.0407509803771973,
+      "learning_rate": 3.371350169360717e-05,
+      "loss": 3.3047,
+      "step": 396500
+    },
+    {
+      "epoch": 16.734814315221517,
+      "grad_norm": 2.592278480529785,
+      "learning_rate": 3.362774943189127e-05,
+      "loss": 3.3126,
+      "step": 397000
+    },
+    {
+      "epoch": 16.755890907558065,
+      "grad_norm": 2.8476998805999756,
+      "learning_rate": 3.354199717017537e-05,
+      "loss": 3.31,
+      "step": 397500
+    },
+    {
+      "epoch": 16.776967499894617,
+      "grad_norm": 2.965733766555786,
+      "learning_rate": 3.345624490845946e-05,
+      "loss": 3.3084,
+      "step": 398000
+    },
+    {
+      "epoch": 16.79804409223117,
+      "grad_norm": 3.0942959785461426,
+      "learning_rate": 3.3370492646743564e-05,
+      "loss": 3.3087,
+      "step": 398500
+    },
+    {
+      "epoch": 16.819120684567718,
+      "grad_norm": 2.264943838119507,
+      "learning_rate": 3.3284740385027655e-05,
+      "loss": 3.3113,
+      "step": 399000
+    },
+    {
+      "epoch": 16.84019727690427,
+      "grad_norm": 2.9896159172058105,
+      "learning_rate": 3.319898812331175e-05,
+      "loss": 3.3093,
+      "step": 399500
+    },
+    {
+      "epoch": 16.861273869240822,
+      "grad_norm": 2.483379602432251,
+      "learning_rate": 3.311323586159585e-05,
+      "loss": 3.3071,
+      "step": 400000
+    },
+    {
+      "epoch": 16.861273869240822,
+      "eval_accuracy": 0.6837188977160888,
+      "eval_loss": 3.1099298000335693,
+      "eval_runtime": 170.1849,
+      "eval_samples_per_second": 709.264,
+      "eval_steps_per_second": 7.392,
+      "step": 400000
+    },
+    {
+      "epoch": 16.88235046157737,
+      "grad_norm": 2.8032302856445312,
+      "learning_rate": 3.302748359987995e-05,
+      "loss": 3.3034,
+      "step": 400500
+    },
+    {
+      "epoch": 16.903427053913923,
+      "grad_norm": 2.7127344608306885,
+      "learning_rate": 3.294173133816404e-05,
+      "loss": 3.3104,
+      "step": 401000
+    },
+    {
+      "epoch": 16.924503646250475,
+      "grad_norm": 2.525390625,
+      "learning_rate": 3.2855979076448147e-05,
+      "loss": 3.3115,
+      "step": 401500
+    },
+    {
+      "epoch": 16.945580238587024,
+      "grad_norm": 3.892774820327759,
+      "learning_rate": 3.277022681473224e-05,
+      "loss": 3.3107,
+      "step": 402000
+    },
+    {
+      "epoch": 16.966656830923576,
+      "grad_norm": 3.205115795135498,
+      "learning_rate": 3.2684474553016336e-05,
+      "loss": 3.3112,
+      "step": 402500
+    },
+    {
+      "epoch": 16.987733423260128,
+      "grad_norm": 3.153106212615967,
+      "learning_rate": 3.2598722291300435e-05,
+      "loss": 3.3127,
+      "step": 403000
+    },
+    {
+      "epoch": 17.00881001559668,
+      "grad_norm": 3.274756669998169,
+      "learning_rate": 3.251297002958453e-05,
+      "loss": 3.2948,
+      "step": 403500
+    },
+    {
+      "epoch": 17.02988660793323,
+      "grad_norm": 2.702843189239502,
+      "learning_rate": 3.242721776786863e-05,
+      "loss": 3.2757,
+      "step": 404000
+    },
+    {
+      "epoch": 17.05096320026978,
+      "grad_norm": 3.87237548828125,
+      "learning_rate": 3.234146550615272e-05,
+      "loss": 3.2753,
+      "step": 404500
+    },
+    {
+      "epoch": 17.072039792606333,
+      "grad_norm": 4.515840530395508,
+      "learning_rate": 3.225571324443683e-05,
+      "loss": 3.2781,
+      "step": 405000
+    },
+    {
+      "epoch": 17.09311638494288,
+      "grad_norm": 3.5361855030059814,
+      "learning_rate": 3.216996098272092e-05,
+      "loss": 3.2783,
+      "step": 405500
+    },
+    {
+      "epoch": 17.114192977279433,
+      "grad_norm": 3.55424427986145,
+      "learning_rate": 3.208420872100502e-05,
+      "loss": 3.2781,
+      "step": 406000
+    },
+    {
+      "epoch": 17.135269569615986,
+      "grad_norm": 3.8088538646698,
+      "learning_rate": 3.1998456459289116e-05,
+      "loss": 3.2794,
+      "step": 406500
+    },
+    {
+      "epoch": 17.156346161952534,
+      "grad_norm": 3.78381609916687,
+      "learning_rate": 3.1912704197573214e-05,
+      "loss": 3.276,
+      "step": 407000
+    },
+    {
+      "epoch": 17.177422754289086,
+      "grad_norm": 3.333416223526001,
+      "learning_rate": 3.1826951935857306e-05,
+      "loss": 3.2766,
+      "step": 407500
+    },
+    {
+      "epoch": 17.19849934662564,
+      "grad_norm": 2.9532582759857178,
+      "learning_rate": 3.174119967414141e-05,
+      "loss": 3.2855,
+      "step": 408000
+    },
+    {
+      "epoch": 17.219575938962187,
+      "grad_norm": 3.4870810508728027,
+      "learning_rate": 3.16554474124255e-05,
+      "loss": 3.2816,
+      "step": 408500
+    },
+    {
+      "epoch": 17.24065253129874,
+      "grad_norm": 3.011261463165283,
+      "learning_rate": 3.15696951507096e-05,
+      "loss": 3.2862,
+      "step": 409000
+    },
+    {
+      "epoch": 17.26172912363529,
+      "grad_norm": 3.4503183364868164,
+      "learning_rate": 3.14839428889937e-05,
+      "loss": 3.28,
+      "step": 409500
+    },
+    {
+      "epoch": 17.282805715971843,
+      "grad_norm": 3.40216064453125,
+      "learning_rate": 3.13981906272778e-05,
+      "loss": 3.2867,
+      "step": 410000
+    },
+    {
+      "epoch": 17.282805715971843,
+      "eval_accuracy": 0.6842659418496722,
+      "eval_loss": 3.129236936569214,
+      "eval_runtime": 169.5557,
+      "eval_samples_per_second": 711.896,
+      "eval_steps_per_second": 7.419,
+      "step": 410000
+    },
+    {
+      "epoch": 17.303882308308392,
+      "grad_norm": 3.196157693862915,
+      "learning_rate": 3.131243836556189e-05,
+      "loss": 3.2915,
+      "step": 410500
+    },
+    {
+      "epoch": 17.324958900644944,
+      "grad_norm": 3.163710355758667,
+      "learning_rate": 3.1226686103845994e-05,
+      "loss": 3.2807,
+      "step": 411000
+    },
+    {
+      "epoch": 17.346035492981496,
+      "grad_norm": 3.039942502975464,
+      "learning_rate": 3.1140933842130086e-05,
+      "loss": 3.2835,
+      "step": 411500
+    },
+    {
+      "epoch": 17.367112085318045,
+      "grad_norm": 3.512453556060791,
+      "learning_rate": 3.1055181580414184e-05,
+      "loss": 3.2883,
+      "step": 412000
+    },
+    {
+      "epoch": 17.388188677654597,
+      "grad_norm": 3.227476119995117,
+      "learning_rate": 3.096942931869828e-05,
+      "loss": 3.2907,
+      "step": 412500
+    },
+    {
+      "epoch": 17.40926526999115,
+      "grad_norm": 3.441025972366333,
+      "learning_rate": 3.088367705698238e-05,
+      "loss": 3.2842,
+      "step": 413000
+    },
+    {
+      "epoch": 17.430341862327698,
+      "grad_norm": 3.8537466526031494,
+      "learning_rate": 3.079792479526647e-05,
+      "loss": 3.2882,
+      "step": 413500
+    },
+    {
+      "epoch": 17.45141845466425,
+      "grad_norm": 3.4829189777374268,
+      "learning_rate": 3.071217253355058e-05,
+      "loss": 3.2862,
+      "step": 414000
+    },
+    {
+      "epoch": 17.472495047000802,
+      "grad_norm": 3.919830560684204,
+      "learning_rate": 3.062642027183467e-05,
+      "loss": 3.2858,
+      "step": 414500
+    },
+    {
+      "epoch": 17.49357163933735,
+      "grad_norm": 3.927372932434082,
+      "learning_rate": 3.054066801011877e-05,
+      "loss": 3.2874,
+      "step": 415000
+    },
+    {
+      "epoch": 17.514648231673903,
+      "grad_norm": 3.6875627040863037,
+      "learning_rate": 3.045491574840287e-05,
+      "loss": 3.2916,
+      "step": 415500
+    },
+    {
+      "epoch": 17.535724824010455,
+      "grad_norm": 3.297685384750366,
+      "learning_rate": 3.0369163486686964e-05,
+      "loss": 3.2878,
+      "step": 416000
+    },
+    {
+      "epoch": 17.556801416347007,
+      "grad_norm": 3.185832977294922,
+      "learning_rate": 3.028341122497106e-05,
+      "loss": 3.2885,
+      "step": 416500
+    },
+    {
+      "epoch": 17.577878008683555,
+      "grad_norm": 3.592794895172119,
+      "learning_rate": 3.019765896325516e-05,
+      "loss": 3.2925,
+      "step": 417000
+    },
+    {
+      "epoch": 17.598954601020107,
+      "grad_norm": 3.360342264175415,
+      "learning_rate": 3.0111906701539255e-05,
+      "loss": 3.291,
+      "step": 417500
+    },
+    {
+      "epoch": 17.62003119335666,
+      "grad_norm": 3.4382853507995605,
+      "learning_rate": 3.002615443982335e-05,
+      "loss": 3.2877,
+      "step": 418000
+    },
+    {
+      "epoch": 17.641107785693208,
+      "grad_norm": 3.272459030151367,
+      "learning_rate": 2.9940402178107452e-05,
+      "loss": 3.2895,
+      "step": 418500
+    },
+    {
+      "epoch": 17.66218437802976,
+      "grad_norm": 2.695430040359497,
+      "learning_rate": 2.9854649916391547e-05,
+      "loss": 3.288,
+      "step": 419000
+    },
+    {
+      "epoch": 17.683260970366312,
+      "grad_norm": 3.237405776977539,
+      "learning_rate": 2.976889765467564e-05,
+      "loss": 3.2949,
+      "step": 419500
+    },
+    {
+      "epoch": 17.70433756270286,
+      "grad_norm": 3.3073835372924805,
+      "learning_rate": 2.9683145392959737e-05,
+      "loss": 3.2879,
+      "step": 420000
+    },
+    {
+      "epoch": 17.70433756270286,
+      "eval_accuracy": 0.6840768687876021,
+      "eval_loss": 3.1273975372314453,
+      "eval_runtime": 170.5419,
+      "eval_samples_per_second": 707.779,
+      "eval_steps_per_second": 7.376,
+      "step": 420000
+    },
+    {
+      "epoch": 17.725414155039413,
+      "grad_norm": 4.059576511383057,
+      "learning_rate": 2.9597393131243838e-05,
+      "loss": 3.2911,
+      "step": 420500
+    },
+    {
+      "epoch": 17.746490747375965,
+      "grad_norm": 3.774427890777588,
+      "learning_rate": 2.9511640869527933e-05,
+      "loss": 3.2908,
+      "step": 421000
+    },
+    {
+      "epoch": 17.767567339712514,
+      "grad_norm": 3.2842395305633545,
+      "learning_rate": 2.942588860781203e-05,
+      "loss": 3.2894,
+      "step": 421500
+    },
+    {
+      "epoch": 17.788643932049066,
+      "grad_norm": 3.387479782104492,
+      "learning_rate": 2.934013634609613e-05,
+      "loss": 3.2933,
+      "step": 422000
+    },
+    {
+      "epoch": 17.809720524385618,
+      "grad_norm": 4.478725433349609,
+      "learning_rate": 2.9254384084380225e-05,
+      "loss": 3.2955,
+      "step": 422500
+    },
+    {
+      "epoch": 17.830797116722167,
+      "grad_norm": 2.984009027481079,
+      "learning_rate": 2.9168631822664323e-05,
+      "loss": 3.2909,
+      "step": 423000
+    },
+    {
+      "epoch": 17.85187370905872,
+      "grad_norm": 4.35100793838501,
+      "learning_rate": 2.908287956094842e-05,
+      "loss": 3.2941,
+      "step": 423500
+    },
+    {
+      "epoch": 17.87295030139527,
+      "grad_norm": 3.322648286819458,
+      "learning_rate": 2.899712729923252e-05,
+      "loss": 3.2945,
+      "step": 424000
+    },
+    {
+      "epoch": 17.894026893731823,
+      "grad_norm": 3.9791224002838135,
+      "learning_rate": 2.8911375037516615e-05,
+      "loss": 3.2939,
+      "step": 424500
+    },
+    {
+      "epoch": 17.91510348606837,
+      "grad_norm": 2.774501085281372,
+      "learning_rate": 2.8825622775800716e-05,
+      "loss": 3.2957,
+      "step": 425000
+    },
+    {
+      "epoch": 17.936180078404924,
+      "grad_norm": 3.1015520095825195,
+      "learning_rate": 2.873987051408481e-05,
+      "loss": 3.2941,
+      "step": 425500
+    },
+    {
+      "epoch": 17.957256670741476,
+      "grad_norm": 3.339456558227539,
+      "learning_rate": 2.8654118252368906e-05,
+      "loss": 3.292,
+      "step": 426000
+    },
+    {
+      "epoch": 17.978333263078024,
+      "grad_norm": 3.968292474746704,
+      "learning_rate": 2.8568365990653008e-05,
+      "loss": 3.2899,
+      "step": 426500
+    },
+    {
+      "epoch": 17.999409855414576,
+      "grad_norm": 3.918501853942871,
+      "learning_rate": 2.8482613728937103e-05,
+      "loss": 3.2958,
+      "step": 427000
+    },
+    {
+      "epoch": 18.02048644775113,
+      "grad_norm": 4.164634704589844,
+      "learning_rate": 2.8396861467221198e-05,
+      "loss": 3.2574,
+      "step": 427500
+    },
+    {
+      "epoch": 18.041563040087677,
+      "grad_norm": 3.878692865371704,
+      "learning_rate": 2.83111092055053e-05,
+      "loss": 3.2529,
+      "step": 428000
+    },
+    {
+      "epoch": 18.06263963242423,
+      "grad_norm": 3.773500919342041,
+      "learning_rate": 2.8225356943789394e-05,
+      "loss": 3.2533,
+      "step": 428500
+    },
+    {
+      "epoch": 18.08371622476078,
+      "grad_norm": 4.820009708404541,
+      "learning_rate": 2.813960468207349e-05,
+      "loss": 3.2545,
+      "step": 429000
+    },
+    {
+      "epoch": 18.10479281709733,
+      "grad_norm": 3.718006134033203,
+      "learning_rate": 2.805385242035759e-05,
+      "loss": 3.2573,
+      "step": 429500
+    },
+    {
+      "epoch": 18.125869409433882,
+      "grad_norm": 3.680863857269287,
+      "learning_rate": 2.7968100158641686e-05,
+      "loss": 3.2591,
+      "step": 430000
+    },
+    {
+      "epoch": 18.125869409433882,
+      "eval_accuracy": 0.684055379649511,
+      "eval_loss": 3.1492443084716797,
+      "eval_runtime": 170.796,
+      "eval_samples_per_second": 706.726,
+      "eval_steps_per_second": 7.366,
+      "step": 430000
+    },
+    {
+      "epoch": 18.146946001770434,
+      "grad_norm": 3.637221097946167,
+      "learning_rate": 2.788234789692578e-05,
+      "loss": 3.2572,
+      "step": 430500
+    },
+    {
+      "epoch": 18.168022594106986,
+      "grad_norm": 4.023508548736572,
+      "learning_rate": 2.7796595635209882e-05,
+      "loss": 3.262,
+      "step": 431000
+    },
+    {
+      "epoch": 18.189099186443535,
+      "grad_norm": 3.752685546875,
+      "learning_rate": 2.7710843373493977e-05,
+      "loss": 3.2613,
+      "step": 431500
+    },
+    {
+      "epoch": 18.210175778780087,
+      "grad_norm": 4.996819972991943,
+      "learning_rate": 2.7625091111778072e-05,
+      "loss": 3.264,
+      "step": 432000
+    },
+    {
+      "epoch": 18.23125237111664,
+      "grad_norm": 4.04128360748291,
+      "learning_rate": 2.7539338850062174e-05,
+      "loss": 3.2573,
+      "step": 432500
+    },
+    {
+      "epoch": 18.252328963453188,
+      "grad_norm": 4.287837028503418,
+      "learning_rate": 2.745358658834627e-05,
+      "loss": 3.2685,
+      "step": 433000
+    },
+    {
+      "epoch": 18.27340555578974,
+      "grad_norm": 3.3309459686279297,
+      "learning_rate": 2.7367834326630364e-05,
+      "loss": 3.261,
+      "step": 433500
+    },
+    {
+      "epoch": 18.294482148126292,
+      "grad_norm": 4.1754469871521,
+      "learning_rate": 2.7282082064914465e-05,
+      "loss": 3.2611,
+      "step": 434000
+    },
+    {
+      "epoch": 18.31555874046284,
+      "grad_norm": 3.925184726715088,
+      "learning_rate": 2.719632980319856e-05,
+      "loss": 3.2614,
+      "step": 434500
+    },
+    {
+      "epoch": 18.336635332799393,
+      "grad_norm": 3.9996166229248047,
+      "learning_rate": 2.7110577541482655e-05,
+      "loss": 3.262,
+      "step": 435000
+    },
+    {
+      "epoch": 18.357711925135945,
+      "grad_norm": 4.0513200759887695,
+      "learning_rate": 2.7024825279766757e-05,
+      "loss": 3.2708,
+      "step": 435500
+    },
+    {
+      "epoch": 18.378788517472493,
+      "grad_norm": 4.114017009735107,
+      "learning_rate": 2.6939073018050852e-05,
+      "loss": 3.2682,
+      "step": 436000
+    },
+    {
+      "epoch": 18.399865109809046,
+      "grad_norm": 3.170576572418213,
+      "learning_rate": 2.685332075633495e-05,
+      "loss": 3.2659,
+      "step": 436500
+    },
+    {
+      "epoch": 18.420941702145598,
+      "grad_norm": 5.085720539093018,
+      "learning_rate": 2.6767568494619045e-05,
+      "loss": 3.2698,
+      "step": 437000
+    },
+    {
+      "epoch": 18.44201829448215,
+      "grad_norm": 4.064079761505127,
+      "learning_rate": 2.6681816232903143e-05,
+      "loss": 3.2665,
+      "step": 437500
+    },
+    {
+      "epoch": 18.4630948868187,
+      "grad_norm": 4.315755367279053,
+      "learning_rate": 2.6596063971187242e-05,
+      "loss": 3.2665,
+      "step": 438000
+    },
+    {
+      "epoch": 18.48417147915525,
+      "grad_norm": 3.711355686187744,
+      "learning_rate": 2.6510311709471337e-05,
+      "loss": 3.2643,
+      "step": 438500
+    },
+    {
+      "epoch": 18.505248071491803,
+      "grad_norm": 3.931490659713745,
+      "learning_rate": 2.642455944775544e-05,
+      "loss": 3.2711,
+      "step": 439000
+    },
+    {
+      "epoch": 18.52632466382835,
+      "grad_norm": 3.3598387241363525,
+      "learning_rate": 2.6338807186039533e-05,
+      "loss": 3.2682,
+      "step": 439500
+    },
+    {
+      "epoch": 18.547401256164903,
+      "grad_norm": 3.9856948852539062,
+      "learning_rate": 2.6253054924323628e-05,
+      "loss": 3.265,
+      "step": 440000
+    },
+    {
+      "epoch": 18.547401256164903,
+      "eval_accuracy": 0.683926602224607,
+      "eval_loss": 3.1468591690063477,
+      "eval_runtime": 170.6889,
+      "eval_samples_per_second": 707.17,
+      "eval_steps_per_second": 7.37,
+      "step": 440000
+    },
+    {
+      "epoch": 18.568477848501455,
+      "grad_norm": 4.438536167144775,
+      "learning_rate": 2.616730266260773e-05,
+      "loss": 3.2665,
+      "step": 440500
+    },
+    {
+      "epoch": 18.589554440838004,
+      "grad_norm": 3.2423741817474365,
+      "learning_rate": 2.6081550400891825e-05,
+      "loss": 3.2686,
+      "step": 441000
+    },
+    {
+      "epoch": 18.610631033174556,
+      "grad_norm": 4.126953125,
+      "learning_rate": 2.599579813917592e-05,
+      "loss": 3.2671,
+      "step": 441500
+    },
+    {
+      "epoch": 18.63170762551111,
+      "grad_norm": 3.4374070167541504,
+      "learning_rate": 2.591004587746002e-05,
+      "loss": 3.2698,
+      "step": 442000
+    },
+    {
+      "epoch": 18.652784217847657,
+      "grad_norm": 4.37463903427124,
+      "learning_rate": 2.5824293615744116e-05,
+      "loss": 3.2702,
+      "step": 442500
+    },
+    {
+      "epoch": 18.67386081018421,
+      "grad_norm": 4.58735990524292,
+      "learning_rate": 2.573854135402821e-05,
+      "loss": 3.2715,
+      "step": 443000
+    },
+    {
+      "epoch": 18.69493740252076,
+      "grad_norm": 3.698725461959839,
+      "learning_rate": 2.5652789092312313e-05,
+      "loss": 3.2698,
+      "step": 443500
+    },
+    {
+      "epoch": 18.716013994857313,
+      "grad_norm": 5.114134311676025,
+      "learning_rate": 2.5567036830596408e-05,
+      "loss": 3.2678,
+      "step": 444000
+    },
+    {
+      "epoch": 18.737090587193862,
+      "grad_norm": 4.050961971282959,
+      "learning_rate": 2.5481284568880503e-05,
+      "loss": 3.2698,
+      "step": 444500
+    },
+    {
+      "epoch": 18.758167179530414,
+      "grad_norm": 3.7773776054382324,
+      "learning_rate": 2.5395532307164605e-05,
+      "loss": 3.2654,
+      "step": 445000
+    },
+    {
+      "epoch": 18.779243771866966,
+      "grad_norm": 4.1582441329956055,
+      "learning_rate": 2.53097800454487e-05,
+      "loss": 3.2758,
+      "step": 445500
+    },
+    {
+      "epoch": 18.800320364203515,
+      "grad_norm": 3.3719358444213867,
+      "learning_rate": 2.5224027783732794e-05,
+      "loss": 3.2727,
+      "step": 446000
+    },
+    {
+      "epoch": 18.821396956540067,
+      "grad_norm": 4.29610538482666,
+      "learning_rate": 2.5138275522016896e-05,
+      "loss": 3.2634,
+      "step": 446500
+    },
+    {
+      "epoch": 18.84247354887662,
+      "grad_norm": 3.3512744903564453,
+      "learning_rate": 2.505252326030099e-05,
+      "loss": 3.2683,
+      "step": 447000
+    },
+    {
+      "epoch": 18.863550141213167,
+      "grad_norm": 4.351622104644775,
+      "learning_rate": 2.496677099858509e-05,
+      "loss": 3.2704,
+      "step": 447500
+    },
+    {
+      "epoch": 18.88462673354972,
+      "grad_norm": 4.247168064117432,
+      "learning_rate": 2.4881018736869184e-05,
+      "loss": 3.2742,
+      "step": 448000
+    },
+    {
+      "epoch": 18.90570332588627,
+      "grad_norm": 3.575975179672241,
+      "learning_rate": 2.4795266475153283e-05,
+      "loss": 3.2722,
+      "step": 448500
+    },
+    {
+      "epoch": 18.92677991822282,
+      "grad_norm": 4.169772624969482,
+      "learning_rate": 2.470951421343738e-05,
+      "loss": 3.2748,
+      "step": 449000
+    },
+    {
+      "epoch": 18.947856510559372,
+      "grad_norm": 3.4350407123565674,
+      "learning_rate": 2.4623761951721476e-05,
+      "loss": 3.2693,
+      "step": 449500
+    },
+    {
+      "epoch": 18.968933102895924,
+      "grad_norm": 3.747549295425415,
+      "learning_rate": 2.4538009690005574e-05,
+      "loss": 3.2726,
+      "step": 450000
+    },
+    {
+      "epoch": 18.968933102895924,
+      "eval_accuracy": 0.6845765015363443,
+      "eval_loss": 3.143242835998535,
+      "eval_runtime": 170.3174,
+      "eval_samples_per_second": 708.712,
+      "eval_steps_per_second": 7.386,
+      "step": 450000
+    },
+    {
+      "epoch": 18.990009695232473,
+      "grad_norm": 4.648780345916748,
+      "learning_rate": 2.4452257428289672e-05,
+      "loss": 3.2721,
+      "step": 450500
+    },
+    {
+      "epoch": 19.011086287569025,
+      "grad_norm": 4.312752723693848,
+      "learning_rate": 2.436650516657377e-05,
+      "loss": 3.25,
+      "step": 451000
+    },
+    {
+      "epoch": 19.032162879905577,
+      "grad_norm": 4.037884712219238,
+      "learning_rate": 2.428075290485787e-05,
+      "loss": 3.232,
+      "step": 451500
+    },
+    {
+      "epoch": 19.05323947224213,
+      "grad_norm": 4.347498416900635,
+      "learning_rate": 2.4195000643141964e-05,
+      "loss": 3.2325,
+      "step": 452000
+    },
+    {
+      "epoch": 19.074316064578678,
+      "grad_norm": 4.133699893951416,
+      "learning_rate": 2.4109248381426062e-05,
+      "loss": 3.2362,
+      "step": 452500
+    },
+    {
+      "epoch": 19.09539265691523,
+      "grad_norm": 4.085947513580322,
+      "learning_rate": 2.402349611971016e-05,
+      "loss": 3.2347,
+      "step": 453000
+    },
+    {
+      "epoch": 19.116469249251782,
+      "grad_norm": 3.81748366355896,
+      "learning_rate": 2.3937743857994255e-05,
+      "loss": 3.2342,
+      "step": 453500
+    },
+    {
+      "epoch": 19.13754584158833,
+      "grad_norm": 3.960181951522827,
+      "learning_rate": 2.3851991596278354e-05,
+      "loss": 3.2381,
+      "step": 454000
+    },
+    {
+      "epoch": 19.158622433924883,
+      "grad_norm": 5.008077621459961,
+      "learning_rate": 2.3766239334562452e-05,
+      "loss": 3.2326,
+      "step": 454500
+    },
+    {
+      "epoch": 19.179699026261435,
+      "grad_norm": 4.1802849769592285,
+      "learning_rate": 2.3680487072846547e-05,
+      "loss": 3.2357,
+      "step": 455000
+    },
+    {
+      "epoch": 19.200775618597984,
+      "grad_norm": 4.099165439605713,
+      "learning_rate": 2.3594734811130645e-05,
+      "loss": 3.2338,
+      "step": 455500
+    },
+    {
+      "epoch": 19.221852210934536,
+      "grad_norm": 3.9187498092651367,
+      "learning_rate": 2.3508982549414744e-05,
+      "loss": 3.2406,
+      "step": 456000
+    },
+    {
+      "epoch": 19.242928803271088,
+      "grad_norm": 5.7032999992370605,
+      "learning_rate": 2.342323028769884e-05,
+      "loss": 3.2383,
+      "step": 456500
+    },
+    {
+      "epoch": 19.264005395607636,
+      "grad_norm": 3.5878989696502686,
+      "learning_rate": 2.3337478025982937e-05,
+      "loss": 3.2355,
+      "step": 457000
+    },
+    {
+      "epoch": 19.28508198794419,
+      "grad_norm": 3.6120500564575195,
+      "learning_rate": 2.3251725764267035e-05,
+      "loss": 3.2392,
+      "step": 457500
+    },
+    {
+      "epoch": 19.30615858028074,
+      "grad_norm": 4.108083248138428,
+      "learning_rate": 2.316597350255113e-05,
+      "loss": 3.2471,
+      "step": 458000
+    },
+    {
+      "epoch": 19.327235172617293,
+      "grad_norm": 4.111003875732422,
+      "learning_rate": 2.308022124083523e-05,
+      "loss": 3.2412,
+      "step": 458500
+    },
+    {
+      "epoch": 19.34831176495384,
+      "grad_norm": 4.6436028480529785,
+      "learning_rate": 2.2994468979119323e-05,
+      "loss": 3.2445,
+      "step": 459000
+    },
+    {
+      "epoch": 19.369388357290394,
+      "grad_norm": 4.312155246734619,
+      "learning_rate": 2.290871671740342e-05,
+      "loss": 3.2387,
+      "step": 459500
+    },
+    {
+      "epoch": 19.390464949626946,
+      "grad_norm": 3.894751787185669,
+      "learning_rate": 2.282296445568752e-05,
+      "loss": 3.2429,
+      "step": 460000
+    },
+    {
+      "epoch": 19.390464949626946,
+      "eval_accuracy": 0.6847457390766554,
+      "eval_loss": 3.1708741188049316,
+      "eval_runtime": 170.5301,
+      "eval_samples_per_second": 707.828,
+      "eval_steps_per_second": 7.377,
+      "step": 460000
+    },
+    {
+      "epoch": 19.411541541963494,
+      "grad_norm": 4.2461090087890625,
+      "learning_rate": 2.2737212193971615e-05,
+      "loss": 3.2461,
+      "step": 460500
+    },
+    {
+      "epoch": 19.432618134300046,
+      "grad_norm": 3.5640110969543457,
+      "learning_rate": 2.2651459932255713e-05,
+      "loss": 3.2457,
+      "step": 461000
+    },
+    {
+      "epoch": 19.4536947266366,
+      "grad_norm": 4.043247222900391,
+      "learning_rate": 2.256570767053981e-05,
+      "loss": 3.2398,
+      "step": 461500
+    },
+    {
+      "epoch": 19.474771318973147,
+      "grad_norm": 4.585749626159668,
+      "learning_rate": 2.2479955408823906e-05,
+      "loss": 3.2398,
+      "step": 462000
+    },
+    {
+      "epoch": 19.4958479113097,
+      "grad_norm": 4.122952461242676,
+      "learning_rate": 2.2394203147108005e-05,
+      "loss": 3.2413,
+      "step": 462500
+    },
+    {
+      "epoch": 19.51692450364625,
+      "grad_norm": 4.112588882446289,
+      "learning_rate": 2.2308450885392103e-05,
+      "loss": 3.248,
+      "step": 463000
+    },
+    {
+      "epoch": 19.5380010959828,
+      "grad_norm": 3.56839919090271,
+      "learning_rate": 2.22226986236762e-05,
+      "loss": 3.242,
+      "step": 463500
+    },
+    {
+      "epoch": 19.559077688319352,
+      "grad_norm": 4.39258337020874,
+      "learning_rate": 2.21369463619603e-05,
+      "loss": 3.2424,
+      "step": 464000
+    },
+    {
+      "epoch": 19.580154280655904,
+      "grad_norm": 4.252934455871582,
+      "learning_rate": 2.2051194100244394e-05,
+      "loss": 3.2478,
+      "step": 464500
+    },
+    {
+      "epoch": 19.601230872992456,
+      "grad_norm": 4.219910621643066,
+      "learning_rate": 2.1965441838528493e-05,
+      "loss": 3.249,
+      "step": 465000
+    },
+    {
+      "epoch": 19.622307465329005,
+      "grad_norm": 3.3714215755462646,
+      "learning_rate": 2.187968957681259e-05,
+      "loss": 3.2467,
+      "step": 465500
+    },
+    {
+      "epoch": 19.643384057665557,
+      "grad_norm": 4.69956111907959,
+      "learning_rate": 2.179393731509669e-05,
+      "loss": 3.2477,
+      "step": 466000
+    },
+    {
+      "epoch": 19.66446065000211,
+      "grad_norm": 3.8777003288269043,
+      "learning_rate": 2.1708185053380784e-05,
+      "loss": 3.2481,
+      "step": 466500
+    },
+    {
+      "epoch": 19.685537242338658,
+      "grad_norm": 3.8825223445892334,
+      "learning_rate": 2.1622432791664883e-05,
+      "loss": 3.2513,
+      "step": 467000
+    },
+    {
+      "epoch": 19.70661383467521,
+      "grad_norm": 4.216036319732666,
+      "learning_rate": 2.1536680529948978e-05,
+      "loss": 3.244,
+      "step": 467500
+    },
+    {
+      "epoch": 19.727690427011762,
+      "grad_norm": 3.7775492668151855,
+      "learning_rate": 2.1450928268233076e-05,
+      "loss": 3.2462,
+      "step": 468000
+    },
+    {
+      "epoch": 19.74876701934831,
+      "grad_norm": 4.436530590057373,
+      "learning_rate": 2.1365176006517174e-05,
+      "loss": 3.2486,
+      "step": 468500
+    },
+    {
+      "epoch": 19.769843611684863,
+      "grad_norm": 4.05308198928833,
+      "learning_rate": 2.127942374480127e-05,
+      "loss": 3.2466,
+      "step": 469000
+    },
+    {
+      "epoch": 19.790920204021415,
+      "grad_norm": 4.378019332885742,
+      "learning_rate": 2.1193671483085367e-05,
+      "loss": 3.2529,
+      "step": 469500
+    },
+    {
+      "epoch": 19.811996796357963,
+      "grad_norm": 4.108922004699707,
+      "learning_rate": 2.1107919221369466e-05,
+      "loss": 3.2518,
+      "step": 470000
+    },
+    {
+      "epoch": 19.811996796357963,
+      "eval_accuracy": 0.6845928069728188,
+      "eval_loss": 3.159818649291992,
+      "eval_runtime": 170.8194,
+      "eval_samples_per_second": 706.629,
+      "eval_steps_per_second": 7.365,
+      "step": 470000
+    },
+    {
+      "epoch": 19.833073388694515,
+      "grad_norm": 3.38533616065979,
+      "learning_rate": 2.102216695965356e-05,
+      "loss": 3.2526,
+      "step": 470500
+    },
+    {
+      "epoch": 19.854149981031068,
+      "grad_norm": 4.119932174682617,
+      "learning_rate": 2.093641469793766e-05,
+      "loss": 3.2488,
+      "step": 471000
+    },
+    {
+      "epoch": 19.87522657336762,
+      "grad_norm": 4.643124103546143,
+      "learning_rate": 2.0850662436221757e-05,
+      "loss": 3.2479,
+      "step": 471500
+    },
+    {
+      "epoch": 19.89630316570417,
+      "grad_norm": 4.0324931144714355,
+      "learning_rate": 2.0764910174505852e-05,
+      "loss": 3.2459,
+      "step": 472000
+    },
+    {
+      "epoch": 19.91737975804072,
+      "grad_norm": 4.815215587615967,
+      "learning_rate": 2.067915791278995e-05,
+      "loss": 3.2513,
+      "step": 472500
+    },
+    {
+      "epoch": 19.938456350377272,
+      "grad_norm": 4.339061260223389,
+      "learning_rate": 2.059340565107405e-05,
+      "loss": 3.2479,
+      "step": 473000
+    },
+    {
+      "epoch": 19.95953294271382,
+      "grad_norm": 3.8695290088653564,
+      "learning_rate": 2.0507653389358144e-05,
+      "loss": 3.2508,
+      "step": 473500
+    },
+    {
+      "epoch": 19.980609535050373,
+      "grad_norm": 4.041666507720947,
+      "learning_rate": 2.0421901127642242e-05,
+      "loss": 3.2483,
+      "step": 474000
+    },
+    {
+      "epoch": 20.001686127386925,
+      "grad_norm": 4.477975368499756,
+      "learning_rate": 2.033614886592634e-05,
+      "loss": 3.2543,
+      "step": 474500
+    },
+    {
+      "epoch": 20.022762719723474,
+      "grad_norm": 4.476532936096191,
+      "learning_rate": 2.0250396604210435e-05,
+      "loss": 3.2121,
+      "step": 475000
+    },
+    {
+      "epoch": 20.043839312060026,
+      "grad_norm": 4.109957218170166,
+      "learning_rate": 2.0164644342494534e-05,
+      "loss": 3.211,
+      "step": 475500
+    },
+    {
+      "epoch": 20.064915904396578,
+      "grad_norm": 5.721868991851807,
+      "learning_rate": 2.007889208077863e-05,
+      "loss": 3.2016,
+      "step": 476000
+    },
+    {
+      "epoch": 20.085992496733127,
+      "grad_norm": 4.6245198249816895,
+      "learning_rate": 1.9993139819062727e-05,
+      "loss": 3.2099,
+      "step": 476500
+    },
+    {
+      "epoch": 20.10706908906968,
+      "grad_norm": 5.032809734344482,
+      "learning_rate": 1.9907387557346825e-05,
+      "loss": 3.2136,
+      "step": 477000
+    },
+    {
+      "epoch": 20.12814568140623,
+      "grad_norm": 4.981314182281494,
+      "learning_rate": 1.9821635295630923e-05,
+      "loss": 3.212,
+      "step": 477500
+    },
+    {
+      "epoch": 20.14922227374278,
+      "grad_norm": 4.635950088500977,
+      "learning_rate": 1.9735883033915022e-05,
+      "loss": 3.2099,
+      "step": 478000
+    },
+    {
+      "epoch": 20.17029886607933,
+      "grad_norm": 4.22322940826416,
+      "learning_rate": 1.965013077219912e-05,
+      "loss": 3.2168,
+      "step": 478500
+    },
+    {
+      "epoch": 20.191375458415884,
+      "grad_norm": 4.1681108474731445,
+      "learning_rate": 1.9564378510483215e-05,
+      "loss": 3.2144,
+      "step": 479000
+    },
+    {
+      "epoch": 20.212452050752436,
+      "grad_norm": 5.723038196563721,
+      "learning_rate": 1.9478626248767313e-05,
+      "loss": 3.2187,
+      "step": 479500
+    },
+    {
+      "epoch": 20.233528643088984,
+      "grad_norm": 5.504665374755859,
+      "learning_rate": 1.939287398705141e-05,
+      "loss": 3.2136,
+      "step": 480000
+    },
+    {
+      "epoch": 20.233528643088984,
+      "eval_accuracy": 0.6845690430146579,
+      "eval_loss": 3.1931612491607666,
+      "eval_runtime": 171.4264,
+      "eval_samples_per_second": 704.127,
+      "eval_steps_per_second": 7.338,
+      "step": 480000
+    },
+    {
+      "epoch": 20.254605235425537,
+      "grad_norm": 5.115634918212891,
+      "learning_rate": 1.9307121725335506e-05,
+      "loss": 3.2141,
+      "step": 480500
+    },
+    {
+      "epoch": 20.27568182776209,
+      "grad_norm": 4.559135437011719,
+      "learning_rate": 1.9221369463619605e-05,
+      "loss": 3.2138,
+      "step": 481000
+    },
+    {
+      "epoch": 20.296758420098637,
+      "grad_norm": 4.420338153839111,
+      "learning_rate": 1.9135617201903703e-05,
+      "loss": 3.2141,
+      "step": 481500
+    },
+    {
+      "epoch": 20.31783501243519,
+      "grad_norm": 4.991296291351318,
+      "learning_rate": 1.9049864940187798e-05,
+      "loss": 3.2189,
+      "step": 482000
+    },
+    {
+      "epoch": 20.33891160477174,
+      "grad_norm": 5.635188579559326,
+      "learning_rate": 1.8964112678471896e-05,
+      "loss": 3.2174,
+      "step": 482500
+    },
+    {
+      "epoch": 20.35998819710829,
+      "grad_norm": 5.333300590515137,
+      "learning_rate": 1.8878360416755995e-05,
+      "loss": 3.2248,
+      "step": 483000
+    },
+    {
+      "epoch": 20.381064789444842,
+      "grad_norm": 4.8347907066345215,
+      "learning_rate": 1.879260815504009e-05,
+      "loss": 3.2227,
+      "step": 483500
+    },
+    {
+      "epoch": 20.402141381781394,
+      "grad_norm": 4.219213008880615,
+      "learning_rate": 1.8706855893324188e-05,
+      "loss": 3.2166,
+      "step": 484000
+    },
+    {
+      "epoch": 20.423217974117943,
+      "grad_norm": 3.7311036586761475,
+      "learning_rate": 1.8621103631608283e-05,
+      "loss": 3.2195,
+      "step": 484500
+    },
+    {
+      "epoch": 20.444294566454495,
+      "grad_norm": 3.829192638397217,
+      "learning_rate": 1.853535136989238e-05,
+      "loss": 3.2207,
+      "step": 485000
+    },
+    {
+      "epoch": 20.465371158791047,
+      "grad_norm": 4.77553653717041,
+      "learning_rate": 1.844959910817648e-05,
+      "loss": 3.2196,
+      "step": 485500
+    },
+    {
+      "epoch": 20.4864477511276,
+      "grad_norm": 4.871058464050293,
+      "learning_rate": 1.8363846846460574e-05,
+      "loss": 3.222,
+      "step": 486000
+    },
+    {
+      "epoch": 20.507524343464148,
+      "grad_norm": 5.606616497039795,
+      "learning_rate": 1.8278094584744673e-05,
+      "loss": 3.2184,
+      "step": 486500
+    },
+    {
+      "epoch": 20.5286009358007,
+      "grad_norm": 4.276630878448486,
+      "learning_rate": 1.819234232302877e-05,
+      "loss": 3.2252,
+      "step": 487000
+    },
+    {
+      "epoch": 20.549677528137252,
+      "grad_norm": 5.017322540283203,
+      "learning_rate": 1.8106590061312866e-05,
+      "loss": 3.2198,
+      "step": 487500
+    },
+    {
+      "epoch": 20.5707541204738,
+      "grad_norm": 5.141661167144775,
+      "learning_rate": 1.8020837799596964e-05,
+      "loss": 3.2243,
+      "step": 488000
+    },
+    {
+      "epoch": 20.591830712810353,
+      "grad_norm": 4.082401752471924,
+      "learning_rate": 1.7935085537881062e-05,
+      "loss": 3.2277,
+      "step": 488500
+    },
+    {
+      "epoch": 20.612907305146905,
+      "grad_norm": 3.846029043197632,
+      "learning_rate": 1.7849333276165157e-05,
+      "loss": 3.2217,
+      "step": 489000
+    },
+    {
+      "epoch": 20.633983897483454,
+      "grad_norm": 4.768825054168701,
+      "learning_rate": 1.7763581014449256e-05,
+      "loss": 3.2223,
+      "step": 489500
+    },
+    {
+      "epoch": 20.655060489820006,
+      "grad_norm": 4.781210899353027,
+      "learning_rate": 1.7677828752733354e-05,
+      "loss": 3.2214,
+      "step": 490000
+    },
+    {
+      "epoch": 20.655060489820006,
+      "eval_accuracy": 0.6848229375940145,
+      "eval_loss": 3.1829471588134766,
+      "eval_runtime": 170.6938,
+      "eval_samples_per_second": 707.149,
+      "eval_steps_per_second": 7.37,
+      "step": 490000
+    },
+    {
+      "epoch": 20.676137082156558,
+      "grad_norm": 4.5583367347717285,
+      "learning_rate": 1.7592076491017452e-05,
+      "loss": 3.2212,
+      "step": 490500
+    },
+    {
+      "epoch": 20.697213674493106,
+      "grad_norm": 4.421870231628418,
+      "learning_rate": 1.7506324229301547e-05,
+      "loss": 3.2263,
+      "step": 491000
+    },
+    {
+      "epoch": 20.71829026682966,
+      "grad_norm": 4.701058864593506,
+      "learning_rate": 1.7420571967585646e-05,
+      "loss": 3.2272,
+      "step": 491500
+    },
+    {
+      "epoch": 20.73936685916621,
+      "grad_norm": 4.232299327850342,
+      "learning_rate": 1.7334819705869744e-05,
+      "loss": 3.2234,
+      "step": 492000
+    },
+    {
+      "epoch": 20.760443451502763,
+      "grad_norm": 4.774412155151367,
+      "learning_rate": 1.7249067444153842e-05,
+      "loss": 3.2286,
+      "step": 492500
+    },
+    {
+      "epoch": 20.78152004383931,
+      "grad_norm": 4.754651069641113,
+      "learning_rate": 1.7163315182437937e-05,
+      "loss": 3.2184,
+      "step": 493000
+    },
+    {
+      "epoch": 20.802596636175863,
+      "grad_norm": 3.7205655574798584,
+      "learning_rate": 1.7077562920722035e-05,
+      "loss": 3.2225,
+      "step": 493500
+    },
+    {
+      "epoch": 20.823673228512416,
+      "grad_norm": 4.571173667907715,
+      "learning_rate": 1.6991810659006134e-05,
+      "loss": 3.2257,
+      "step": 494000
+    },
+    {
+      "epoch": 20.844749820848964,
+      "grad_norm": 4.7466254234313965,
+      "learning_rate": 1.690605839729023e-05,
+      "loss": 3.2268,
+      "step": 494500
+    },
+    {
+      "epoch": 20.865826413185516,
+      "grad_norm": 4.926101207733154,
+      "learning_rate": 1.6820306135574327e-05,
+      "loss": 3.2308,
+      "step": 495000
+    },
+    {
+      "epoch": 20.88690300552207,
+      "grad_norm": 4.948836803436279,
+      "learning_rate": 1.6734553873858425e-05,
+      "loss": 3.2273,
+      "step": 495500
+    },
+    {
+      "epoch": 20.907979597858617,
+      "grad_norm": 4.47009801864624,
+      "learning_rate": 1.664880161214252e-05,
+      "loss": 3.2269,
+      "step": 496000
+    },
+    {
+      "epoch": 20.92905619019517,
+      "grad_norm": 5.178150177001953,
+      "learning_rate": 1.656304935042662e-05,
+      "loss": 3.2291,
+      "step": 496500
+    },
+    {
+      "epoch": 20.95013278253172,
+      "grad_norm": 5.286802768707275,
+      "learning_rate": 1.6477297088710717e-05,
+      "loss": 3.2278,
+      "step": 497000
+    },
+    {
+      "epoch": 20.97120937486827,
+      "grad_norm": 4.866032600402832,
+      "learning_rate": 1.639154482699481e-05,
+      "loss": 3.2222,
+      "step": 497500
+    },
+    {
+      "epoch": 20.992285967204822,
+      "grad_norm": 4.26856803894043,
+      "learning_rate": 1.630579256527891e-05,
+      "loss": 3.2254,
+      "step": 498000
+    },
+    {
+      "epoch": 21.013362559541374,
+      "grad_norm": 4.24958610534668,
+      "learning_rate": 1.6220040303563008e-05,
+      "loss": 3.2023,
+      "step": 498500
+    },
+    {
+      "epoch": 21.034439151877926,
+      "grad_norm": 5.562870502471924,
+      "learning_rate": 1.6134288041847103e-05,
+      "loss": 3.1871,
+      "step": 499000
+    },
+    {
+      "epoch": 21.055515744214475,
+      "grad_norm": 5.064261436462402,
+      "learning_rate": 1.60485357801312e-05,
+      "loss": 3.1829,
+      "step": 499500
+    },
+    {
+      "epoch": 21.076592336551027,
+      "grad_norm": 5.453797817230225,
+      "learning_rate": 1.59627835184153e-05,
+      "loss": 3.1855,
+      "step": 500000
+    },
+    {
+      "epoch": 21.076592336551027,
+      "eval_accuracy": 0.6848303179138335,
+      "eval_loss": 3.1981201171875,
+      "eval_runtime": 169.8119,
+      "eval_samples_per_second": 710.822,
+      "eval_steps_per_second": 7.408,
+      "step": 500000
+    },
+    {
+      "epoch": 21.09766892888758,
+      "grad_norm": 4.877757549285889,
+      "learning_rate": 1.5877031256699395e-05,
+      "loss": 3.1916,
+      "step": 500500
+    },
+    {
+      "epoch": 21.118745521224128,
+      "grad_norm": 5.0154900550842285,
+      "learning_rate": 1.5791278994983493e-05,
+      "loss": 3.1919,
+      "step": 501000
+    },
+    {
+      "epoch": 21.13982211356068,
+      "grad_norm": 4.875800609588623,
+      "learning_rate": 1.570552673326759e-05,
+      "loss": 3.184,
+      "step": 501500
+    },
+    {
+      "epoch": 21.16089870589723,
+      "grad_norm": 4.926912784576416,
+      "learning_rate": 1.5619774471551686e-05,
+      "loss": 3.1896,
+      "step": 502000
+    },
+    {
+      "epoch": 21.18197529823378,
+      "grad_norm": 4.5805277824401855,
+      "learning_rate": 1.5534022209835785e-05,
+      "loss": 3.1872,
+      "step": 502500
+    },
+    {
+      "epoch": 21.203051890570332,
+      "grad_norm": 4.029383182525635,
+      "learning_rate": 1.544826994811988e-05,
+      "loss": 3.1914,
+      "step": 503000
+    },
+    {
+      "epoch": 21.224128482906885,
+      "grad_norm": 4.803980827331543,
+      "learning_rate": 1.5362517686403978e-05,
+      "loss": 3.1966,
+      "step": 503500
+    },
+    {
+      "epoch": 21.245205075243433,
+      "grad_norm": 6.157120227813721,
+      "learning_rate": 1.5276765424688076e-05,
+      "loss": 3.1948,
+      "step": 504000
+    },
+    {
+      "epoch": 21.266281667579985,
+      "grad_norm": 5.6742048263549805,
+      "learning_rate": 1.5191013162972173e-05,
+      "loss": 3.1914,
+      "step": 504500
+    },
+    {
+      "epoch": 21.287358259916537,
+      "grad_norm": 5.427289009094238,
+      "learning_rate": 1.5105260901256271e-05,
+      "loss": 3.1982,
+      "step": 505000
+    },
+    {
+      "epoch": 21.308434852253086,
+      "grad_norm": 5.275989532470703,
+      "learning_rate": 1.501950863954037e-05,
+      "loss": 3.196,
+      "step": 505500
+    },
+    {
+      "epoch": 21.329511444589638,
+      "grad_norm": 5.165039539337158,
+      "learning_rate": 1.4933756377824464e-05,
+      "loss": 3.192,
+      "step": 506000
+    },
+    {
+      "epoch": 21.35058803692619,
+      "grad_norm": 4.751610279083252,
+      "learning_rate": 1.4848004116108563e-05,
+      "loss": 3.1891,
+      "step": 506500
+    },
+    {
+      "epoch": 21.371664629262742,
+      "grad_norm": 4.658975124359131,
+      "learning_rate": 1.4762251854392661e-05,
+      "loss": 3.2004,
+      "step": 507000
+    },
+    {
+      "epoch": 21.39274122159929,
+      "grad_norm": 4.797718048095703,
+      "learning_rate": 1.4676499592676758e-05,
+      "loss": 3.2018,
+      "step": 507500
+    },
+    {
+      "epoch": 21.413817813935843,
+      "grad_norm": 4.8728718757629395,
+      "learning_rate": 1.4590747330960856e-05,
+      "loss": 3.1952,
+      "step": 508000
+    },
+    {
+      "epoch": 21.434894406272395,
+      "grad_norm": 5.120675086975098,
+      "learning_rate": 1.4504995069244952e-05,
+      "loss": 3.1949,
+      "step": 508500
+    },
+    {
+      "epoch": 21.455970998608944,
+      "grad_norm": 5.3744330406188965,
+      "learning_rate": 1.4419242807529049e-05,
+      "loss": 3.1962,
+      "step": 509000
+    },
+    {
+      "epoch": 21.477047590945496,
+      "grad_norm": 4.529819011688232,
+      "learning_rate": 1.4333490545813147e-05,
+      "loss": 3.2027,
+      "step": 509500
+    },
+    {
+      "epoch": 21.498124183282048,
+      "grad_norm": 6.3125996589660645,
+      "learning_rate": 1.4247738284097246e-05,
+      "loss": 3.1918,
+      "step": 510000
+    },
+    {
+      "epoch": 21.498124183282048,
+      "eval_accuracy": 0.6850588633365172,
+      "eval_loss": 3.211841344833374,
+      "eval_runtime": 165.4568,
+      "eval_samples_per_second": 729.532,
+      "eval_steps_per_second": 7.603,
+      "step": 510000
+    },
+    {
+      "epoch": 21.519200775618597,
+      "grad_norm": 4.498312473297119,
+      "learning_rate": 1.416198602238134e-05,
+      "loss": 3.2026,
+      "step": 510500
+    },
+    {
+      "epoch": 21.54027736795515,
+      "grad_norm": 5.6001176834106445,
+      "learning_rate": 1.4076233760665439e-05,
+      "loss": 3.198,
+      "step": 511000
+    },
+    {
+      "epoch": 21.5613539602917,
+      "grad_norm": 4.619207859039307,
+      "learning_rate": 1.3990481498949534e-05,
+      "loss": 3.1966,
+      "step": 511500
+    },
+    {
+      "epoch": 21.58243055262825,
+      "grad_norm": 5.158178806304932,
+      "learning_rate": 1.3904729237233632e-05,
+      "loss": 3.2015,
+      "step": 512000
+    },
+    {
+      "epoch": 21.6035071449648,
+      "grad_norm": 4.5401811599731445,
+      "learning_rate": 1.381897697551773e-05,
+      "loss": 3.1998,
+      "step": 512500
+    },
+    {
+      "epoch": 21.624583737301354,
+      "grad_norm": 4.69637393951416,
+      "learning_rate": 1.3733224713801825e-05,
+      "loss": 3.202,
+      "step": 513000
+    },
+    {
+      "epoch": 21.645660329637906,
+      "grad_norm": 5.348447322845459,
+      "learning_rate": 1.3647472452085924e-05,
+      "loss": 3.2063,
+      "step": 513500
+    },
+    {
+      "epoch": 21.666736921974454,
+      "grad_norm": 5.2119059562683105,
+      "learning_rate": 1.3561720190370022e-05,
+      "loss": 3.1985,
+      "step": 514000
+    },
+    {
+      "epoch": 21.687813514311006,
+      "grad_norm": 4.997492790222168,
+      "learning_rate": 1.3475967928654119e-05,
+      "loss": 3.197,
+      "step": 514500
+    },
+    {
+      "epoch": 21.70889010664756,
+      "grad_norm": 4.931380748748779,
+      "learning_rate": 1.3390215666938217e-05,
+      "loss": 3.2032,
+      "step": 515000
+    },
+    {
+      "epoch": 21.729966698984107,
+      "grad_norm": 4.813687324523926,
+      "learning_rate": 1.3304463405222315e-05,
+      "loss": 3.2039,
+      "step": 515500
+    },
+    {
+      "epoch": 21.75104329132066,
+      "grad_norm": 4.975936412811279,
+      "learning_rate": 1.321871114350641e-05,
+      "loss": 3.2045,
+      "step": 516000
+    },
+    {
+      "epoch": 21.77211988365721,
+      "grad_norm": 4.288062572479248,
+      "learning_rate": 1.3132958881790508e-05,
+      "loss": 3.2016,
+      "step": 516500
+    },
+    {
+      "epoch": 21.79319647599376,
+      "grad_norm": 5.69228458404541,
+      "learning_rate": 1.3047206620074607e-05,
+      "loss": 3.2004,
+      "step": 517000
+    },
+    {
+      "epoch": 21.814273068330312,
+      "grad_norm": 4.891899108886719,
+      "learning_rate": 1.2961454358358702e-05,
+      "loss": 3.2027,
+      "step": 517500
+    },
+    {
+      "epoch": 21.835349660666864,
+      "grad_norm": 5.452838897705078,
+      "learning_rate": 1.28757020966428e-05,
+      "loss": 3.2006,
+      "step": 518000
+    },
+    {
+      "epoch": 21.856426253003413,
+      "grad_norm": 5.107034206390381,
+      "learning_rate": 1.2789949834926898e-05,
+      "loss": 3.1998,
+      "step": 518500
+    },
+    {
+      "epoch": 21.877502845339965,
+      "grad_norm": 4.643390655517578,
+      "learning_rate": 1.2704197573210993e-05,
+      "loss": 3.2024,
+      "step": 519000
+    },
+    {
+      "epoch": 21.898579437676517,
+      "grad_norm": 5.448083400726318,
+      "learning_rate": 1.2618445311495092e-05,
+      "loss": 3.2027,
+      "step": 519500
+    },
+    {
+      "epoch": 21.91965603001307,
+      "grad_norm": 5.660414218902588,
+      "learning_rate": 1.2532693049779186e-05,
+      "loss": 3.2026,
+      "step": 520000
+    },
+    {
+      "epoch": 21.91965603001307,
+      "eval_accuracy": 0.6850510005868433,
+      "eval_loss": 3.1941893100738525,
+      "eval_runtime": 165.6303,
+      "eval_samples_per_second": 728.768,
+      "eval_steps_per_second": 7.595,
+      "step": 520000
+    },
+    {
+      "epoch": 21.940732622349618,
+      "grad_norm": 5.486734867095947,
+      "learning_rate": 1.2446940788063285e-05,
+      "loss": 3.2009,
+      "step": 520500
+    },
+    {
+      "epoch": 21.96180921468617,
+      "grad_norm": 4.463052272796631,
+      "learning_rate": 1.2361188526347383e-05,
+      "loss": 3.2034,
+      "step": 521000
+    },
+    {
+      "epoch": 21.982885807022722,
+      "grad_norm": 5.204930782318115,
+      "learning_rate": 1.227543626463148e-05,
+      "loss": 3.2027,
+      "step": 521500
+    },
+    {
+      "epoch": 22.00396239935927,
+      "grad_norm": 4.586575508117676,
+      "learning_rate": 1.2189684002915578e-05,
+      "loss": 3.1889,
+      "step": 522000
+    },
+    {
+      "epoch": 22.025038991695823,
+      "grad_norm": 5.440699577331543,
+      "learning_rate": 1.2103931741199675e-05,
+      "loss": 3.1579,
+      "step": 522500
+    },
+    {
+      "epoch": 22.046115584032375,
+      "grad_norm": 4.934328079223633,
+      "learning_rate": 1.2018179479483773e-05,
+      "loss": 3.1673,
+      "step": 523000
+    },
+    {
+      "epoch": 22.067192176368923,
+      "grad_norm": 5.601432800292969,
+      "learning_rate": 1.193242721776787e-05,
+      "loss": 3.1716,
+      "step": 523500
+    },
+    {
+      "epoch": 22.088268768705476,
+      "grad_norm": 4.88502311706543,
+      "learning_rate": 1.1846674956051966e-05,
+      "loss": 3.1718,
+      "step": 524000
+    },
+    {
+      "epoch": 22.109345361042028,
+      "grad_norm": 4.73794412612915,
+      "learning_rate": 1.1760922694336064e-05,
+      "loss": 3.1684,
+      "step": 524500
+    },
+    {
+      "epoch": 22.130421953378576,
+      "grad_norm": 6.459784507751465,
+      "learning_rate": 1.1675170432620161e-05,
+      "loss": 3.1683,
+      "step": 525000
+    },
+    {
+      "epoch": 22.15149854571513,
+      "grad_norm": 6.056376934051514,
+      "learning_rate": 1.1589418170904258e-05,
+      "loss": 3.1656,
+      "step": 525500
+    },
+    {
+      "epoch": 22.17257513805168,
+      "grad_norm": 6.270998001098633,
+      "learning_rate": 1.1503665909188354e-05,
+      "loss": 3.1685,
+      "step": 526000
+    },
+    {
+      "epoch": 22.193651730388233,
+      "grad_norm": 5.268072128295898,
+      "learning_rate": 1.1417913647472453e-05,
+      "loss": 3.1729,
+      "step": 526500
+    },
+    {
+      "epoch": 22.21472832272478,
+      "grad_norm": 5.000064849853516,
+      "learning_rate": 1.133216138575655e-05,
+      "loss": 3.1728,
+      "step": 527000
+    },
+    {
+      "epoch": 22.235804915061333,
+      "grad_norm": 4.135090351104736,
+      "learning_rate": 1.1246409124040646e-05,
+      "loss": 3.1699,
+      "step": 527500
+    },
+    {
+      "epoch": 22.256881507397885,
+      "grad_norm": 4.9894609451293945,
+      "learning_rate": 1.1160656862324744e-05,
+      "loss": 3.1745,
+      "step": 528000
+    },
+    {
+      "epoch": 22.277958099734434,
+      "grad_norm": 5.3042707443237305,
+      "learning_rate": 1.1074904600608842e-05,
+      "loss": 3.1763,
+      "step": 528500
+    },
+    {
+      "epoch": 22.299034692070986,
+      "grad_norm": 6.436113357543945,
+      "learning_rate": 1.0989152338892939e-05,
+      "loss": 3.1731,
+      "step": 529000
+    },
+    {
+      "epoch": 22.320111284407538,
+      "grad_norm": 4.843597888946533,
+      "learning_rate": 1.0903400077177037e-05,
+      "loss": 3.1724,
+      "step": 529500
+    },
+    {
+      "epoch": 22.341187876744087,
+      "grad_norm": 5.15333890914917,
+      "learning_rate": 1.0817647815461134e-05,
+      "loss": 3.1785,
+      "step": 530000
+    },
+    {
+      "epoch": 22.341187876744087,
+      "eval_accuracy": 0.6851311543315363,
+      "eval_loss": 3.223714828491211,
+      "eval_runtime": 168.2124,
+      "eval_samples_per_second": 717.581,
+      "eval_steps_per_second": 7.479,
+      "step": 530000
+    },
+    {
+      "epoch": 22.36226446908064,
+      "grad_norm": 5.487493991851807,
+      "learning_rate": 1.073189555374523e-05,
+      "loss": 3.1724,
+      "step": 530500
+    },
+    {
+      "epoch": 22.38334106141719,
+      "grad_norm": 5.864981174468994,
+      "learning_rate": 1.0646143292029327e-05,
+      "loss": 3.1792,
+      "step": 531000
+    },
+    {
+      "epoch": 22.40441765375374,
+      "grad_norm": 4.940119743347168,
+      "learning_rate": 1.0560391030313425e-05,
+      "loss": 3.1708,
+      "step": 531500
+    },
+    {
+      "epoch": 22.42549424609029,
+      "grad_norm": 5.398608684539795,
+      "learning_rate": 1.0474638768597522e-05,
+      "loss": 3.17,
+      "step": 532000
+    },
+    {
+      "epoch": 22.446570838426844,
+      "grad_norm": 5.617795467376709,
+      "learning_rate": 1.0388886506881619e-05,
+      "loss": 3.1688,
+      "step": 532500
+    },
+    {
+      "epoch": 22.467647430763392,
+      "grad_norm": 6.463131904602051,
+      "learning_rate": 1.0303134245165717e-05,
+      "loss": 3.175,
+      "step": 533000
+    },
+    {
+      "epoch": 22.488724023099945,
+      "grad_norm": 5.9649577140808105,
+      "learning_rate": 1.0217381983449814e-05,
+      "loss": 3.1742,
+      "step": 533500
+    },
+    {
+      "epoch": 22.509800615436497,
+      "grad_norm": 4.905406951904297,
+      "learning_rate": 1.013162972173391e-05,
+      "loss": 3.1726,
+      "step": 534000
+    },
+    {
+      "epoch": 22.53087720777305,
+      "grad_norm": 4.908738613128662,
+      "learning_rate": 1.0045877460018009e-05,
+      "loss": 3.1803,
+      "step": 534500
+    },
+    {
+      "epoch": 22.551953800109597,
+      "grad_norm": 6.330564498901367,
+      "learning_rate": 9.960125198302105e-06,
+      "loss": 3.18,
+      "step": 535000
+    },
+    {
+      "epoch": 22.57303039244615,
+      "grad_norm": 5.547333240509033,
+      "learning_rate": 9.874372936586203e-06,
+      "loss": 3.1769,
+      "step": 535500
+    },
+    {
+      "epoch": 22.5941069847827,
+      "grad_norm": 4.69191312789917,
+      "learning_rate": 9.7886206748703e-06,
+      "loss": 3.1741,
+      "step": 536000
+    },
+    {
+      "epoch": 22.61518357711925,
+      "grad_norm": 4.491515159606934,
+      "learning_rate": 9.702868413154398e-06,
+      "loss": 3.1719,
+      "step": 536500
+    },
+    {
+      "epoch": 22.636260169455802,
+      "grad_norm": 4.397480487823486,
+      "learning_rate": 9.617116151438495e-06,
+      "loss": 3.1747,
+      "step": 537000
+    },
+    {
+      "epoch": 22.657336761792354,
+      "grad_norm": 4.929766654968262,
+      "learning_rate": 9.531363889722592e-06,
+      "loss": 3.1764,
+      "step": 537500
+    },
+    {
+      "epoch": 22.678413354128903,
+      "grad_norm": 4.644393444061279,
+      "learning_rate": 9.44561162800669e-06,
+      "loss": 3.1776,
+      "step": 538000
+    },
+    {
+      "epoch": 22.699489946465455,
+      "grad_norm": 5.393256664276123,
+      "learning_rate": 9.359859366290787e-06,
+      "loss": 3.1749,
+      "step": 538500
+    },
+    {
+      "epoch": 22.720566538802007,
+      "grad_norm": 5.509423732757568,
+      "learning_rate": 9.274107104574883e-06,
+      "loss": 3.179,
+      "step": 539000
+    },
+    {
+      "epoch": 22.741643131138556,
+      "grad_norm": 5.442287445068359,
+      "learning_rate": 9.18835484285898e-06,
+      "loss": 3.173,
+      "step": 539500
+    },
+    {
+      "epoch": 22.762719723475108,
+      "grad_norm": 6.117405414581299,
+      "learning_rate": 9.102602581143078e-06,
+      "loss": 3.1744,
+      "step": 540000
+    },
+    {
+      "epoch": 22.762719723475108,
+      "eval_accuracy": 0.6852774781929489,
+      "eval_loss": 3.2269108295440674,
+      "eval_runtime": 170.8349,
+      "eval_samples_per_second": 706.565,
+      "eval_steps_per_second": 7.364,
+      "step": 540000
+    },
+    {
+      "epoch": 22.78379631581166,
+      "grad_norm": 5.455935001373291,
+      "learning_rate": 9.016850319427175e-06,
+      "loss": 3.1776,
+      "step": 540500
+    },
+    {
+      "epoch": 22.804872908148212,
+      "grad_norm": 5.297155857086182,
+      "learning_rate": 8.931098057711271e-06,
+      "loss": 3.1791,
+      "step": 541000
+    },
+    {
+      "epoch": 22.82594950048476,
+      "grad_norm": 5.2517008781433105,
+      "learning_rate": 8.84534579599537e-06,
+      "loss": 3.1786,
+      "step": 541500
+    },
+    {
+      "epoch": 22.847026092821313,
+      "grad_norm": 5.268688678741455,
+      "learning_rate": 8.759593534279468e-06,
+      "loss": 3.1789,
+      "step": 542000
+    },
+    {
+      "epoch": 22.868102685157865,
+      "grad_norm": 5.350543022155762,
+      "learning_rate": 8.673841272563565e-06,
+      "loss": 3.1735,
+      "step": 542500
+    },
+    {
+      "epoch": 22.889179277494414,
+      "grad_norm": 4.294102191925049,
+      "learning_rate": 8.588089010847663e-06,
+      "loss": 3.1778,
+      "step": 543000
+    },
+    {
+      "epoch": 22.910255869830966,
+      "grad_norm": 5.507205009460449,
+      "learning_rate": 8.50233674913176e-06,
+      "loss": 3.1777,
+      "step": 543500
+    },
+    {
+      "epoch": 22.931332462167518,
+      "grad_norm": 5.61820650100708,
+      "learning_rate": 8.416584487415856e-06,
+      "loss": 3.1724,
+      "step": 544000
+    },
+    {
+      "epoch": 22.952409054504066,
+      "grad_norm": 5.269683837890625,
+      "learning_rate": 8.330832225699953e-06,
+      "loss": 3.1758,
+      "step": 544500
+    },
+    {
+      "epoch": 22.97348564684062,
+      "grad_norm": 6.310712814331055,
+      "learning_rate": 8.245079963984051e-06,
+      "loss": 3.1732,
+      "step": 545000
+    },
+    {
+      "epoch": 22.99456223917717,
+      "grad_norm": 4.972056865692139,
+      "learning_rate": 8.159327702268148e-06,
+      "loss": 3.1818,
+      "step": 545500
+    },
+    {
+      "epoch": 23.01563883151372,
+      "grad_norm": 5.364410877227783,
+      "learning_rate": 8.073575440552244e-06,
+      "loss": 3.157,
+      "step": 546000
+    },
+    {
+      "epoch": 23.03671542385027,
+      "grad_norm": 5.340466499328613,
+      "learning_rate": 7.987823178836343e-06,
+      "loss": 3.1463,
+      "step": 546500
+    },
+    {
+      "epoch": 23.057792016186824,
+      "grad_norm": 5.486800193786621,
+      "learning_rate": 7.90207091712044e-06,
+      "loss": 3.1492,
+      "step": 547000
+    },
+    {
+      "epoch": 23.078868608523376,
+      "grad_norm": 6.088263988494873,
+      "learning_rate": 7.816318655404536e-06,
+      "loss": 3.1507,
+      "step": 547500
+    },
+    {
+      "epoch": 23.099945200859924,
+      "grad_norm": 5.486906051635742,
+      "learning_rate": 7.730566393688634e-06,
+      "loss": 3.1503,
+      "step": 548000
+    },
+    {
+      "epoch": 23.121021793196476,
+      "grad_norm": 5.962146759033203,
+      "learning_rate": 7.64481413197273e-06,
+      "loss": 3.1523,
+      "step": 548500
+    },
+    {
+      "epoch": 23.14209838553303,
+      "grad_norm": 4.951182842254639,
+      "learning_rate": 7.559061870256828e-06,
+      "loss": 3.1505,
+      "step": 549000
+    },
+    {
+      "epoch": 23.163174977869577,
+      "grad_norm": 5.514127731323242,
+      "learning_rate": 7.473309608540925e-06,
+      "loss": 3.1501,
+      "step": 549500
+    },
+    {
+      "epoch": 23.18425157020613,
+      "grad_norm": 6.135988712310791,
+      "learning_rate": 7.387557346825023e-06,
+      "loss": 3.1528,
+      "step": 550000
+    },
+    {
+      "epoch": 23.18425157020613,
+      "eval_accuracy": 0.6853225200617573,
+      "eval_loss": 3.241945266723633,
+      "eval_runtime": 166.1049,
+      "eval_samples_per_second": 726.685,
+      "eval_steps_per_second": 7.574,
+      "step": 550000
+    },
+    {
+      "epoch": 23.20532816254268,
+      "grad_norm": 4.78519868850708,
+      "learning_rate": 7.3018050851091205e-06,
+      "loss": 3.1468,
+      "step": 550500
+    },
+    {
+      "epoch": 23.22640475487923,
+      "grad_norm": 6.132195949554443,
+      "learning_rate": 7.216052823393217e-06,
+      "loss": 3.1611,
+      "step": 551000
+    },
+    {
+      "epoch": 23.247481347215782,
+      "grad_norm": 5.638672351837158,
+      "learning_rate": 7.1303005616773155e-06,
+      "loss": 3.1538,
+      "step": 551500
+    },
+    {
+      "epoch": 23.268557939552334,
+      "grad_norm": 5.02998161315918,
+      "learning_rate": 7.044548299961412e-06,
+      "loss": 3.1578,
+      "step": 552000
+    },
+    {
+      "epoch": 23.289634531888883,
+      "grad_norm": 6.7584733963012695,
+      "learning_rate": 6.958796038245509e-06,
+      "loss": 3.1469,
+      "step": 552500
+    },
+    {
+      "epoch": 23.310711124225435,
+      "grad_norm": 5.262241840362549,
+      "learning_rate": 6.873043776529606e-06,
+      "loss": 3.1564,
+      "step": 553000
+    },
+    {
+      "epoch": 23.331787716561987,
+      "grad_norm": 4.942476272583008,
+      "learning_rate": 6.787291514813704e-06,
+      "loss": 3.1475,
+      "step": 553500
+    },
+    {
+      "epoch": 23.35286430889854,
+      "grad_norm": 5.01283073425293,
+      "learning_rate": 6.701539253097801e-06,
+      "loss": 3.1492,
+      "step": 554000
+    },
+    {
+      "epoch": 23.373940901235088,
+      "grad_norm": 7.218055248260498,
+      "learning_rate": 6.615786991381898e-06,
+      "loss": 3.1502,
+      "step": 554500
+    },
+    {
+      "epoch": 23.39501749357164,
+      "grad_norm": 5.898569583892822,
+      "learning_rate": 6.530034729665996e-06,
+      "loss": 3.1517,
+      "step": 555000
+    },
+    {
+      "epoch": 23.416094085908192,
+      "grad_norm": 5.227570056915283,
+      "learning_rate": 6.444282467950093e-06,
+      "loss": 3.1498,
+      "step": 555500
+    },
+    {
+      "epoch": 23.43717067824474,
+      "grad_norm": 4.997844219207764,
+      "learning_rate": 6.358530206234189e-06,
+      "loss": 3.1508,
+      "step": 556000
+    },
+    {
+      "epoch": 23.458247270581293,
+      "grad_norm": 5.416109561920166,
+      "learning_rate": 6.272777944518287e-06,
+      "loss": 3.1567,
+      "step": 556500
+    },
+    {
+      "epoch": 23.479323862917845,
+      "grad_norm": 6.433823585510254,
+      "learning_rate": 6.187025682802384e-06,
+      "loss": 3.1536,
+      "step": 557000
+    },
+    {
+      "epoch": 23.500400455254393,
+      "grad_norm": 5.854318618774414,
+      "learning_rate": 6.101273421086482e-06,
+      "loss": 3.1488,
+      "step": 557500
+    },
+    {
+      "epoch": 23.521477047590945,
+      "grad_norm": 4.039520740509033,
+      "learning_rate": 6.015521159370579e-06,
+      "loss": 3.1528,
+      "step": 558000
+    },
+    {
+      "epoch": 23.542553639927497,
+      "grad_norm": 5.101160049438477,
+      "learning_rate": 5.929768897654676e-06,
+      "loss": 3.1504,
+      "step": 558500
+    },
+    {
+      "epoch": 23.563630232264046,
+      "grad_norm": 5.288351535797119,
+      "learning_rate": 5.844016635938773e-06,
+      "loss": 3.1499,
+      "step": 559000
+    },
+    {
+      "epoch": 23.584706824600598,
+      "grad_norm": 4.054171085357666,
+      "learning_rate": 5.75826437422287e-06,
+      "loss": 3.1573,
+      "step": 559500
+    },
+    {
+      "epoch": 23.60578341693715,
+      "grad_norm": 6.486856937408447,
+      "learning_rate": 5.672512112506967e-06,
+      "loss": 3.155,
+      "step": 560000
+    },
+    {
+      "epoch": 23.60578341693715,
+      "eval_accuracy": 0.6858476815843514,
+      "eval_loss": 3.2405009269714355,
+      "eval_runtime": 163.8917,
+      "eval_samples_per_second": 736.498,
+      "eval_steps_per_second": 7.676,
+      "step": 560000
+    },
+    {
+      "epoch": 23.6268600092737,
+      "grad_norm": 5.373769760131836,
+      "learning_rate": 5.5867598507910655e-06,
+      "loss": 3.1583,
+      "step": 560500
+    },
+    {
+      "epoch": 23.64793660161025,
+      "grad_norm": 6.451066017150879,
+      "learning_rate": 5.501007589075162e-06,
+      "loss": 3.1575,
+      "step": 561000
+    },
+    {
+      "epoch": 23.669013193946803,
+      "grad_norm": 4.7933125495910645,
+      "learning_rate": 5.41525532735926e-06,
+      "loss": 3.156,
+      "step": 561500
+    },
+    {
+      "epoch": 23.690089786283355,
+      "grad_norm": 4.360853672027588,
+      "learning_rate": 5.329503065643356e-06,
+      "loss": 3.1543,
+      "step": 562000
+    },
+    {
+      "epoch": 23.711166378619904,
+      "grad_norm": 6.058946132659912,
+      "learning_rate": 5.243750803927454e-06,
+      "loss": 3.1527,
+      "step": 562500
+    },
+    {
+      "epoch": 23.732242970956456,
+      "grad_norm": 4.753995895385742,
+      "learning_rate": 5.157998542211551e-06,
+      "loss": 3.1523,
+      "step": 563000
+    },
+    {
+      "epoch": 23.753319563293008,
+      "grad_norm": 7.127832889556885,
+      "learning_rate": 5.072246280495649e-06,
+      "loss": 3.1488,
+      "step": 563500
+    },
+    {
+      "epoch": 23.774396155629557,
+      "grad_norm": 4.377109050750732,
+      "learning_rate": 4.986494018779746e-06,
+      "loss": 3.1566,
+      "step": 564000
+    },
+    {
+      "epoch": 23.79547274796611,
+      "grad_norm": 6.577995777130127,
+      "learning_rate": 4.900741757063843e-06,
+      "loss": 3.1497,
+      "step": 564500
+    },
+    {
+      "epoch": 23.81654934030266,
+      "grad_norm": 5.56707239151001,
+      "learning_rate": 4.81498949534794e-06,
+      "loss": 3.1489,
+      "step": 565000
+    },
+    {
+      "epoch": 23.83762593263921,
+      "grad_norm": 6.672051906585693,
+      "learning_rate": 4.729237233632038e-06,
+      "loss": 3.1545,
+      "step": 565500
+    },
+    {
+      "epoch": 23.85870252497576,
+      "grad_norm": 4.8806633949279785,
+      "learning_rate": 4.643484971916134e-06,
+      "loss": 3.1555,
+      "step": 566000
+    },
+    {
+      "epoch": 23.879779117312314,
+      "grad_norm": 5.836051940917969,
+      "learning_rate": 4.557732710200232e-06,
+      "loss": 3.151,
+      "step": 566500
+    },
+    {
+      "epoch": 23.900855709648862,
+      "grad_norm": 5.394222259521484,
+      "learning_rate": 4.471980448484329e-06,
+      "loss": 3.1546,
+      "step": 567000
+    },
+    {
+      "epoch": 23.921932301985414,
+      "grad_norm": 5.1816606521606445,
+      "learning_rate": 4.386228186768427e-06,
+      "loss": 3.1503,
+      "step": 567500
+    },
+    {
+      "epoch": 23.943008894321967,
+      "grad_norm": 5.873056411743164,
+      "learning_rate": 4.300475925052524e-06,
+      "loss": 3.155,
+      "step": 568000
+    },
+    {
+      "epoch": 23.96408548665852,
+      "grad_norm": 5.052947521209717,
+      "learning_rate": 4.214723663336621e-06,
+      "loss": 3.1532,
+      "step": 568500
+    },
+    {
+      "epoch": 23.985162078995067,
+      "grad_norm": 5.476485252380371,
+      "learning_rate": 4.128971401620718e-06,
+      "loss": 3.1509,
+      "step": 569000
+    },
+    {
+      "epoch": 24.00623867133162,
+      "grad_norm": 5.572786331176758,
+      "learning_rate": 4.043219139904815e-06,
+      "loss": 3.1466,
+      "step": 569500
+    },
+    {
+      "epoch": 24.02731526366817,
+      "grad_norm": 5.841972351074219,
+      "learning_rate": 3.957466878188912e-06,
+      "loss": 3.1321,
+      "step": 570000
+    },
+    {
+      "epoch": 24.02731526366817,
+      "eval_accuracy": 0.6858839066913045,
+      "eval_loss": 3.257392168045044,
+      "eval_runtime": 166.4622,
+      "eval_samples_per_second": 725.126,
+      "eval_steps_per_second": 7.557,
+      "step": 570000
+    },
+    {
+      "epoch": 24.04839185600472,
+      "grad_norm": 4.969844341278076,
+      "learning_rate": 3.87171461647301e-06,
+      "loss": 3.1372,
+      "step": 570500
+    },
+    {
+      "epoch": 24.069468448341272,
+      "grad_norm": 4.788638114929199,
+      "learning_rate": 3.7859623547571067e-06,
+      "loss": 3.1345,
+      "step": 571000
+    },
+    {
+      "epoch": 24.090545040677824,
+      "grad_norm": 5.74620246887207,
+      "learning_rate": 3.7002100930412046e-06,
+      "loss": 3.1294,
+      "step": 571500
+    },
+    {
+      "epoch": 24.111621633014373,
+      "grad_norm": 6.707075119018555,
+      "learning_rate": 3.614457831325301e-06,
+      "loss": 3.141,
+      "step": 572000
+    },
+    {
+      "epoch": 24.132698225350925,
+      "grad_norm": 5.399980068206787,
+      "learning_rate": 3.5287055696093987e-06,
+      "loss": 3.1338,
+      "step": 572500
+    },
+    {
+      "epoch": 24.153774817687477,
+      "grad_norm": 6.343264102935791,
+      "learning_rate": 3.4429533078934957e-06,
+      "loss": 3.1319,
+      "step": 573000
+    },
+    {
+      "epoch": 24.174851410024026,
+      "grad_norm": 6.0124101638793945,
+      "learning_rate": 3.357201046177593e-06,
+      "loss": 3.1416,
+      "step": 573500
+    },
+    {
+      "epoch": 24.195928002360578,
+      "grad_norm": 6.94915771484375,
+      "learning_rate": 3.2714487844616906e-06,
+      "loss": 3.1407,
+      "step": 574000
+    },
+    {
+      "epoch": 24.21700459469713,
+      "grad_norm": 6.396584987640381,
+      "learning_rate": 3.1856965227457872e-06,
+      "loss": 3.1369,
+      "step": 574500
+    },
+    {
+      "epoch": 24.238081187033682,
+      "grad_norm": 5.3992509841918945,
+      "learning_rate": 3.099944261029885e-06,
+      "loss": 3.1413,
+      "step": 575000
+    },
+    {
+      "epoch": 24.25915777937023,
+      "grad_norm": 5.128297805786133,
+      "learning_rate": 3.014191999313982e-06,
+      "loss": 3.1353,
+      "step": 575500
+    },
+    {
+      "epoch": 24.280234371706783,
+      "grad_norm": 5.4213337898254395,
+      "learning_rate": 2.928439737598079e-06,
+      "loss": 3.1333,
+      "step": 576000
+    },
+    {
+      "epoch": 24.301310964043335,
+      "grad_norm": 5.59174919128418,
+      "learning_rate": 2.8426874758821767e-06,
+      "loss": 3.1375,
+      "step": 576500
+    },
+    {
+      "epoch": 24.322387556379883,
+      "grad_norm": 6.2897844314575195,
+      "learning_rate": 2.7569352141662737e-06,
+      "loss": 3.1282,
+      "step": 577000
+    },
+    {
+      "epoch": 24.343464148716436,
+      "grad_norm": 5.552764415740967,
+      "learning_rate": 2.6711829524503707e-06,
+      "loss": 3.1289,
+      "step": 577500
+    },
+    {
+      "epoch": 24.364540741052988,
+      "grad_norm": 5.757107257843018,
+      "learning_rate": 2.585430690734468e-06,
+      "loss": 3.1333,
+      "step": 578000
+    },
+    {
+      "epoch": 24.385617333389536,
+      "grad_norm": 5.809369087219238,
+      "learning_rate": 2.4996784290185657e-06,
+      "loss": 3.1354,
+      "step": 578500
+    },
+    {
+      "epoch": 24.40669392572609,
+      "grad_norm": 7.0864644050598145,
+      "learning_rate": 2.4139261673026627e-06,
+      "loss": 3.1323,
+      "step": 579000
+    },
+    {
+      "epoch": 24.42777051806264,
+      "grad_norm": 5.688948631286621,
+      "learning_rate": 2.3281739055867597e-06,
+      "loss": 3.1352,
+      "step": 579500
+    },
+    {
+      "epoch": 24.44884711039919,
+      "grad_norm": 5.715628147125244,
+      "learning_rate": 2.242421643870857e-06,
+      "loss": 3.1351,
+      "step": 580000
+    },
+    {
+      "epoch": 24.44884711039919,
+      "eval_accuracy": 0.6862414160254735,
+      "eval_loss": 3.249751567840576,
+      "eval_runtime": 166.0768,
+      "eval_samples_per_second": 726.808,
+      "eval_steps_per_second": 7.575,
+      "step": 580000
+    },
+    {
+      "epoch": 24.46992370273574,
+      "grad_norm": 5.787480354309082,
+      "learning_rate": 2.1566693821549547e-06,
+      "loss": 3.1304,
+      "step": 580500
+    },
+    {
+      "epoch": 24.491000295072293,
+      "grad_norm": 6.171711444854736,
+      "learning_rate": 2.0709171204390517e-06,
+      "loss": 3.1378,
+      "step": 581000
+    },
+    {
+      "epoch": 24.512076887408845,
+      "grad_norm": 4.055683612823486,
+      "learning_rate": 1.985164858723149e-06,
+      "loss": 3.127,
+      "step": 581500
+    },
+    {
+      "epoch": 24.533153479745394,
+      "grad_norm": 5.515174388885498,
+      "learning_rate": 1.8994125970072462e-06,
+      "loss": 3.1365,
+      "step": 582000
+    },
+    {
+      "epoch": 24.554230072081946,
+      "grad_norm": 5.443107604980469,
+      "learning_rate": 1.8136603352913432e-06,
+      "loss": 3.1293,
+      "step": 582500
+    },
+    {
+      "epoch": 24.5753066644185,
+      "grad_norm": 5.300081253051758,
+      "learning_rate": 1.7279080735754405e-06,
+      "loss": 3.1353,
+      "step": 583000
+    },
+    {
+      "epoch": 24.596383256755047,
+      "grad_norm": 5.778362274169922,
+      "learning_rate": 1.642155811859538e-06,
+      "loss": 3.1385,
+      "step": 583500
+    },
+    {
+      "epoch": 24.6174598490916,
+      "grad_norm": 5.719929218292236,
+      "learning_rate": 1.5564035501436352e-06,
+      "loss": 3.1341,
+      "step": 584000
+    },
+    {
+      "epoch": 24.63853644142815,
+      "grad_norm": 6.920734405517578,
+      "learning_rate": 1.4706512884277324e-06,
+      "loss": 3.1364,
+      "step": 584500
+    },
+    {
+      "epoch": 24.6596130337647,
+      "grad_norm": 4.936020374298096,
+      "learning_rate": 1.3848990267118295e-06,
+      "loss": 3.1357,
+      "step": 585000
+    },
+    {
+      "epoch": 24.680689626101252,
+      "grad_norm": 5.8474040031433105,
+      "learning_rate": 1.299146764995927e-06,
+      "loss": 3.1343,
+      "step": 585500
+    },
+    {
+      "epoch": 24.701766218437804,
+      "grad_norm": 5.402904033660889,
+      "learning_rate": 1.213394503280024e-06,
+      "loss": 3.1335,
+      "step": 586000
+    },
+    {
+      "epoch": 24.722842810774353,
+      "grad_norm": 4.910691738128662,
+      "learning_rate": 1.1276422415641212e-06,
+      "loss": 3.1285,
+      "step": 586500
+    },
+    {
+      "epoch": 24.743919403110905,
+      "grad_norm": 5.613698959350586,
+      "learning_rate": 1.0418899798482187e-06,
+      "loss": 3.1407,
+      "step": 587000
+    },
+    {
+      "epoch": 24.764995995447457,
+      "grad_norm": 5.1039814949035645,
+      "learning_rate": 9.561377181323157e-07,
+      "loss": 3.1355,
+      "step": 587500
+    },
+    {
+      "epoch": 24.786072587784005,
+      "grad_norm": 5.673089027404785,
+      "learning_rate": 8.70385456416413e-07,
+      "loss": 3.1327,
+      "step": 588000
+    },
+    {
+      "epoch": 24.807149180120557,
+      "grad_norm": 5.770620346069336,
+      "learning_rate": 7.846331947005103e-07,
+      "loss": 3.1382,
+      "step": 588500
+    },
+    {
+      "epoch": 24.82822577245711,
+      "grad_norm": 6.562462329864502,
+      "learning_rate": 6.988809329846075e-07,
+      "loss": 3.1327,
+      "step": 589000
+    },
+    {
+      "epoch": 24.84930236479366,
+      "grad_norm": 5.800909042358398,
+      "learning_rate": 6.131286712687047e-07,
+      "loss": 3.1329,
+      "step": 589500
+    },
+    {
+      "epoch": 24.87037895713021,
+      "grad_norm": 5.737913608551025,
+      "learning_rate": 5.27376409552802e-07,
+      "loss": 3.133,
+      "step": 590000
+    },
+    {
+      "epoch": 24.87037895713021,
+      "eval_accuracy": 0.6858570754118856,
+      "eval_loss": 3.251901388168335,
+      "eval_runtime": 164.9081,
+      "eval_samples_per_second": 731.959,
+      "eval_steps_per_second": 7.628,
+      "step": 590000
+    },
+    {
+      "epoch": 24.891455549466762,
+      "grad_norm": 5.944506645202637,
+      "learning_rate": 4.4162414783689917e-07,
+      "loss": 3.1342,
+      "step": 590500
+    },
+    {
+      "epoch": 24.912532141803315,
+      "grad_norm": 6.638115406036377,
+      "learning_rate": 3.558718861209964e-07,
+      "loss": 3.1317,
+      "step": 591000
+    },
+    {
+      "epoch": 24.933608734139863,
+      "grad_norm": 6.674060344696045,
+      "learning_rate": 2.701196244050937e-07,
+      "loss": 3.138,
+      "step": 591500
+    },
+    {
+      "epoch": 24.954685326476415,
+      "grad_norm": 5.454610824584961,
+      "learning_rate": 1.8436736268919094e-07,
+      "loss": 3.1355,
+      "step": 592000
+    },
+    {
+      "epoch": 24.975761918812967,
+      "grad_norm": 5.823305606842041,
+      "learning_rate": 9.861510097328816e-08,
+      "loss": 3.1362,
+      "step": 592500
+    },
+    {
+      "epoch": 24.996838511149516,
+      "grad_norm": 5.8226094245910645,
+      "learning_rate": 1.2862839257385414e-08,
+      "loss": 3.1357,
+      "step": 593000
+    },
+    {
+      "epoch": 25.0,
+      "step": 593075,
+      "total_flos": 3.037727852389325e+18,
+      "train_loss": 3.5820748045259263,
+      "train_runtime": 127167.6489,
+      "train_samples_per_second": 447.705,
+      "train_steps_per_second": 4.664
+    }
+  ],
+  "logging_steps": 500,
+  "max_steps": 593075,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 25,
+  "save_steps": 10000,
+  "total_flos": 3.037727852389325e+18,
+  "train_batch_size": 96,
+  "trial_name": null,
+  "trial_params": null
+}