diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,8721 @@
+{
+  "best_metric": 0.8558232931726908,
+  "best_model_checkpoint": "/scratch/camembertv2/runs/results/xnli/camembertav2-base-bf16-p2-17000/max_seq_length-160-gradient_accumulation_steps-4-precision-fp32-learning_rate-1e-05-epochs-10-lr_scheduler-polynomial-warmup_steps-0.1/SEED-1/checkpoint-36816",
+  "epoch": 10.0,
+  "eval_steps": 500,
+  "global_step": 122720,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.008148631029986962,
+      "grad_norm": 1.3529162406921387,
+      "learning_rate": 8.148631029986963e-08,
+      "loss": 1.1024,
+      "step": 100
+    },
+    {
+      "epoch": 0.016297262059973925,
+      "grad_norm": 1.2678381204605103,
+      "learning_rate": 1.6297262059973925e-07,
+      "loss": 1.1006,
+      "step": 200
+    },
+    {
+      "epoch": 0.024445893089960886,
+      "grad_norm": 0.9964377880096436,
+      "learning_rate": 2.4445893089960885e-07,
+      "loss": 1.1,
+      "step": 300
+    },
+    {
+      "epoch": 0.03259452411994785,
+      "grad_norm": 2.1048221588134766,
+      "learning_rate": 3.259452411994785e-07,
+      "loss": 1.1015,
+      "step": 400
+    },
+    {
+      "epoch": 0.04074315514993481,
+      "grad_norm": 1.244388222694397,
+      "learning_rate": 4.0743155149934816e-07,
+      "loss": 1.0995,
+      "step": 500
+    },
+    {
+      "epoch": 0.04889178617992177,
+      "grad_norm": 1.3034368753433228,
+      "learning_rate": 4.889178617992177e-07,
+      "loss": 1.0988,
+      "step": 600
+    },
+    {
+      "epoch": 0.05704041720990873,
+      "grad_norm": 1.0444238185882568,
+      "learning_rate": 5.704041720990874e-07,
+      "loss": 1.0963,
+      "step": 700
+    },
+    {
+      "epoch": 0.0651890482398957,
+      "grad_norm": 3.489915370941162,
+      "learning_rate": 6.51890482398957e-07,
+      "loss": 1.0967,
+      "step": 800
+    },
+    {
+      "epoch": 0.07333767926988266,
+      "grad_norm": 1.0847524404525757,
+      "learning_rate": 7.333767926988267e-07,
+      "loss": 1.0952,
+      "step": 900
+    },
+    {
+      "epoch": 0.08148631029986962,
+      "grad_norm": 1.0315290689468384,
+      "learning_rate": 8.148631029986963e-07,
+      "loss": 1.0935,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08963494132985658,
+      "grad_norm": 1.655386209487915,
+      "learning_rate": 8.963494132985659e-07,
+      "loss": 1.0883,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09778357235984354,
+      "grad_norm": 1.4588720798492432,
+      "learning_rate": 9.778357235984354e-07,
+      "loss": 1.0823,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1059322033898305,
+      "grad_norm": 1.3579902648925781,
+      "learning_rate": 1.059322033898305e-06,
+      "loss": 1.0728,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11408083441981746,
+      "grad_norm": 1.7215408086776733,
+      "learning_rate": 1.1408083441981747e-06,
+      "loss": 1.0528,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12222946544980444,
+      "grad_norm": 2.2282140254974365,
+      "learning_rate": 1.2222946544980446e-06,
+      "loss": 1.0169,
+      "step": 1500
+    },
+    {
+      "epoch": 0.1303780964797914,
+      "grad_norm": 2.632673501968384,
+      "learning_rate": 1.303780964797914e-06,
+      "loss": 0.9727,
+      "step": 1600
+    },
+    {
+      "epoch": 0.13852672750977835,
+      "grad_norm": 5.786465167999268,
+      "learning_rate": 1.3852672750977837e-06,
+      "loss": 0.9276,
+      "step": 1700
+    },
+    {
+      "epoch": 0.14667535853976532,
+      "grad_norm": 5.518361568450928,
+      "learning_rate": 1.4667535853976533e-06,
+      "loss": 0.8697,
+      "step": 1800
+    },
+    {
+      "epoch": 0.15482398956975227,
+      "grad_norm": 6.2041192054748535,
+      "learning_rate": 1.5482398956975228e-06,
+      "loss": 0.8348,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16297262059973924,
+      "grad_norm": 4.39878511428833,
+      "learning_rate": 1.6297262059973926e-06,
+      "loss": 0.784,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17112125162972622,
+      "grad_norm": 8.355758666992188,
+      "learning_rate": 1.7112125162972623e-06,
+      "loss": 0.7816,
+      "step": 2100
+    },
+    {
+      "epoch": 0.17926988265971316,
+      "grad_norm": 7.7773518562316895,
+      "learning_rate": 1.7926988265971317e-06,
+      "loss": 0.7391,
+      "step": 2200
+    },
+    {
+      "epoch": 0.18741851368970014,
+      "grad_norm": 10.548506736755371,
+      "learning_rate": 1.8741851368970016e-06,
+      "loss": 0.7357,
+      "step": 2300
+    },
+    {
+      "epoch": 0.19556714471968709,
+      "grad_norm": 7.029742240905762,
+      "learning_rate": 1.955671447196871e-06,
+      "loss": 0.7138,
+      "step": 2400
+    },
+    {
+      "epoch": 0.20371577574967406,
+      "grad_norm": 7.368396759033203,
+      "learning_rate": 2.037157757496741e-06,
+      "loss": 0.6928,
+      "step": 2500
+    },
+    {
+      "epoch": 0.211864406779661,
+      "grad_norm": 5.480055809020996,
+      "learning_rate": 2.11864406779661e-06,
+      "loss": 0.7032,
+      "step": 2600
+    },
+    {
+      "epoch": 0.22001303780964798,
+      "grad_norm": 9.542736053466797,
+      "learning_rate": 2.20013037809648e-06,
+      "loss": 0.6752,
+      "step": 2700
+    },
+    {
+      "epoch": 0.22816166883963493,
+      "grad_norm": 5.912474632263184,
+      "learning_rate": 2.2816166883963494e-06,
+      "loss": 0.6566,
+      "step": 2800
+    },
+    {
+      "epoch": 0.2363102998696219,
+      "grad_norm": 6.762691020965576,
+      "learning_rate": 2.363102998696219e-06,
+      "loss": 0.6671,
+      "step": 2900
+    },
+    {
+      "epoch": 0.24445893089960888,
+      "grad_norm": 9.684687614440918,
+      "learning_rate": 2.444589308996089e-06,
+      "loss": 0.6612,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2526075619295958,
+      "grad_norm": 8.38856029510498,
+      "learning_rate": 2.5260756192959584e-06,
+      "loss": 0.6441,
+      "step": 3100
+    },
+    {
+      "epoch": 0.2607561929595828,
+      "grad_norm": 8.438015937805176,
+      "learning_rate": 2.607561929595828e-06,
+      "loss": 0.6314,
+      "step": 3200
+    },
+    {
+      "epoch": 0.2689048239895698,
+      "grad_norm": 8.523892402648926,
+      "learning_rate": 2.689048239895698e-06,
+      "loss": 0.6382,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2770534550195567,
+      "grad_norm": 8.889211654663086,
+      "learning_rate": 2.7705345501955674e-06,
+      "loss": 0.6364,
+      "step": 3400
+    },
+    {
+      "epoch": 0.28520208604954367,
+      "grad_norm": 441.1722717285156,
+      "learning_rate": 2.852020860495437e-06,
+      "loss": 0.6033,
+      "step": 3500
+    },
+    {
+      "epoch": 0.29335071707953064,
+      "grad_norm": 10.166702270507812,
+      "learning_rate": 2.9335071707953067e-06,
+      "loss": 0.6282,
+      "step": 3600
+    },
+    {
+      "epoch": 0.3014993481095176,
+      "grad_norm": 10.398645401000977,
+      "learning_rate": 3.0149934810951763e-06,
+      "loss": 0.6072,
+      "step": 3700
+    },
+    {
+      "epoch": 0.30964797913950454,
+      "grad_norm": 12.474647521972656,
+      "learning_rate": 3.0964797913950456e-06,
+      "loss": 0.6154,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3177966101694915,
+      "grad_norm": 7.278645038604736,
+      "learning_rate": 3.1779661016949152e-06,
+      "loss": 0.6028,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3259452411994785,
+      "grad_norm": 8.92871379852295,
+      "learning_rate": 3.2594524119947853e-06,
+      "loss": 0.61,
+      "step": 4000
+    },
+    {
+      "epoch": 0.33409387222946546,
+      "grad_norm": 7.182549476623535,
+      "learning_rate": 3.340938722294655e-06,
+      "loss": 0.6053,
+      "step": 4100
+    },
+    {
+      "epoch": 0.34224250325945244,
+      "grad_norm": 8.213513374328613,
+      "learning_rate": 3.4224250325945246e-06,
+      "loss": 0.5985,
+      "step": 4200
+    },
+    {
+      "epoch": 0.35039113428943935,
+      "grad_norm": 8.712896347045898,
+      "learning_rate": 3.503911342894394e-06,
+      "loss": 0.6186,
+      "step": 4300
+    },
+    {
+      "epoch": 0.35853976531942633,
+      "grad_norm": 11.862103462219238,
+      "learning_rate": 3.5853976531942635e-06,
+      "loss": 0.5893,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3666883963494133,
+      "grad_norm": 10.739442825317383,
+      "learning_rate": 3.666883963494133e-06,
+      "loss": 0.5883,
+      "step": 4500
+    },
+    {
+      "epoch": 0.3748370273794003,
+      "grad_norm": 7.735357761383057,
+      "learning_rate": 3.748370273794003e-06,
+      "loss": 0.5942,
+      "step": 4600
+    },
+    {
+      "epoch": 0.3829856584093872,
+      "grad_norm": 7.83574914932251,
+      "learning_rate": 3.829856584093872e-06,
+      "loss": 0.5715,
+      "step": 4700
+    },
+    {
+      "epoch": 0.39113428943937417,
+      "grad_norm": 8.37131118774414,
+      "learning_rate": 3.911342894393742e-06,
+      "loss": 0.5481,
+      "step": 4800
+    },
+    {
+      "epoch": 0.39928292046936115,
+      "grad_norm": 8.036620140075684,
+      "learning_rate": 3.992829204693612e-06,
+      "loss": 0.5712,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4074315514993481,
+      "grad_norm": 13.716651916503906,
+      "learning_rate": 4.074315514993482e-06,
+      "loss": 0.5569,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4155801825293351,
+      "grad_norm": 8.926308631896973,
+      "learning_rate": 4.1558018252933515e-06,
+      "loss": 0.5404,
+      "step": 5100
+    },
+    {
+      "epoch": 0.423728813559322,
+      "grad_norm": 10.423807144165039,
+      "learning_rate": 4.23728813559322e-06,
+      "loss": 0.5535,
+      "step": 5200
+    },
+    {
+      "epoch": 0.431877444589309,
+      "grad_norm": 7.072977542877197,
+      "learning_rate": 4.31877444589309e-06,
+      "loss": 0.5529,
+      "step": 5300
+    },
+    {
+      "epoch": 0.44002607561929596,
+      "grad_norm": 9.907496452331543,
+      "learning_rate": 4.40026075619296e-06,
+      "loss": 0.5459,
+      "step": 5400
+    },
+    {
+      "epoch": 0.44817470664928294,
+      "grad_norm": 10.329442977905273,
+      "learning_rate": 4.48174706649283e-06,
+      "loss": 0.5777,
+      "step": 5500
+    },
+    {
+      "epoch": 0.45632333767926986,
+      "grad_norm": 9.943466186523438,
+      "learning_rate": 4.563233376792699e-06,
+      "loss": 0.5686,
+      "step": 5600
+    },
+    {
+      "epoch": 0.46447196870925683,
+      "grad_norm": 6.4508056640625,
+      "learning_rate": 4.6447196870925686e-06,
+      "loss": 0.5461,
+      "step": 5700
+    },
+    {
+      "epoch": 0.4726205997392438,
+      "grad_norm": 7.132810115814209,
+      "learning_rate": 4.726205997392438e-06,
+      "loss": 0.5566,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4807692307692308,
+      "grad_norm": 7.723414421081543,
+      "learning_rate": 4.807692307692308e-06,
+      "loss": 0.5661,
+      "step": 5900
+    },
+    {
+      "epoch": 0.48891786179921776,
+      "grad_norm": 11.582465171813965,
+      "learning_rate": 4.889178617992178e-06,
+      "loss": 0.5517,
+      "step": 6000
+    },
+    {
+      "epoch": 0.4970664928292047,
+      "grad_norm": 7.484246730804443,
+      "learning_rate": 4.970664928292047e-06,
+      "loss": 0.568,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5052151238591917,
+      "grad_norm": 10.276209831237793,
+      "learning_rate": 5.052151238591917e-06,
+      "loss": 0.525,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5133637548891786,
+      "grad_norm": 8.953728675842285,
+      "learning_rate": 5.1336375488917865e-06,
+      "loss": 0.5534,
+      "step": 6300
+    },
+    {
+      "epoch": 0.5215123859191656,
+      "grad_norm": 8.184313774108887,
+      "learning_rate": 5.215123859191656e-06,
+      "loss": 0.5467,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5296610169491526,
+      "grad_norm": 8.448795318603516,
+      "learning_rate": 5.296610169491526e-06,
+      "loss": 0.5411,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5378096479791395,
+      "grad_norm": 9.622394561767578,
+      "learning_rate": 5.378096479791396e-06,
+      "loss": 0.5656,
+      "step": 6600
+    },
+    {
+      "epoch": 0.5459582790091264,
+      "grad_norm": 13.388031959533691,
+      "learning_rate": 5.459582790091264e-06,
+      "loss": 0.5433,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5541069100391134,
+      "grad_norm": 8.313009262084961,
+      "learning_rate": 5.541069100391135e-06,
+      "loss": 0.5373,
+      "step": 6800
+    },
+    {
+      "epoch": 0.5622555410691004,
+      "grad_norm": 6.502901077270508,
+      "learning_rate": 5.622555410691004e-06,
+      "loss": 0.5443,
+      "step": 6900
+    },
+    {
+      "epoch": 0.5704041720990873,
+      "grad_norm": 6.7939863204956055,
+      "learning_rate": 5.704041720990874e-06,
+      "loss": 0.5691,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5785528031290743,
+      "grad_norm": 8.51882266998291,
+      "learning_rate": 5.785528031290744e-06,
+      "loss": 0.5477,
+      "step": 7100
+    },
+    {
+      "epoch": 0.5867014341590613,
+      "grad_norm": 7.1099324226379395,
+      "learning_rate": 5.867014341590613e-06,
+      "loss": 0.5287,
+      "step": 7200
+    },
+    {
+      "epoch": 0.5948500651890483,
+      "grad_norm": 12.773337364196777,
+      "learning_rate": 5.948500651890483e-06,
+      "loss": 0.5188,
+      "step": 7300
+    },
+    {
+      "epoch": 0.6029986962190352,
+      "grad_norm": 7.300539016723633,
+      "learning_rate": 6.029986962190353e-06,
+      "loss": 0.5259,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6111473272490222,
+      "grad_norm": 8.644438743591309,
+      "learning_rate": 6.111473272490222e-06,
+      "loss": 0.5299,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6192959582790091,
+      "grad_norm": 10.571152687072754,
+      "learning_rate": 6.192959582790091e-06,
+      "loss": 0.5238,
+      "step": 7600
+    },
+    {
+      "epoch": 0.627444589308996,
+      "grad_norm": 7.20805549621582,
+      "learning_rate": 6.274445893089961e-06,
+      "loss": 0.5184,
+      "step": 7700
+    },
+    {
+      "epoch": 0.635593220338983,
+      "grad_norm": 6.403540134429932,
+      "learning_rate": 6.3559322033898304e-06,
+      "loss": 0.5069,
+      "step": 7800
+    },
+    {
+      "epoch": 0.64374185136897,
+      "grad_norm": 5.379425525665283,
+      "learning_rate": 6.4374185136897e-06,
+      "loss": 0.5149,
+      "step": 7900
+    },
+    {
+      "epoch": 0.651890482398957,
+      "grad_norm": 10.648305892944336,
+      "learning_rate": 6.518904823989571e-06,
+      "loss": 0.5319,
+      "step": 8000
+    },
+    {
+      "epoch": 0.660039113428944,
+      "grad_norm": 7.570769786834717,
+      "learning_rate": 6.60039113428944e-06,
+      "loss": 0.5556,
+      "step": 8100
+    },
+    {
+      "epoch": 0.6681877444589309,
+      "grad_norm": 7.133244037628174,
+      "learning_rate": 6.68187744458931e-06,
+      "loss": 0.5016,
+      "step": 8200
+    },
+    {
+      "epoch": 0.6763363754889179,
+      "grad_norm": 9.345438003540039,
+      "learning_rate": 6.7633637548891795e-06,
+      "loss": 0.515,
+      "step": 8300
+    },
+    {
+      "epoch": 0.6844850065189049,
+      "grad_norm": 7.831772327423096,
+      "learning_rate": 6.844850065189049e-06,
+      "loss": 0.5035,
+      "step": 8400
+    },
+    {
+      "epoch": 0.6926336375488917,
+      "grad_norm": 4.8923468589782715,
+      "learning_rate": 6.926336375488918e-06,
+      "loss": 0.5379,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7007822685788787,
+      "grad_norm": 5.273585319519043,
+      "learning_rate": 7.007822685788788e-06,
+      "loss": 0.5299,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7089308996088657,
+      "grad_norm": 5.884276390075684,
+      "learning_rate": 7.089308996088657e-06,
+      "loss": 0.5212,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7170795306388527,
+      "grad_norm": 11.059792518615723,
+      "learning_rate": 7.170795306388527e-06,
+      "loss": 0.5156,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7252281616688396,
+      "grad_norm": 9.073988914489746,
+      "learning_rate": 7.252281616688397e-06,
+      "loss": 0.516,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7333767926988266,
+      "grad_norm": 7.554878234863281,
+      "learning_rate": 7.333767926988266e-06,
+      "loss": 0.5189,
+      "step": 9000
+    },
+    {
+      "epoch": 0.7415254237288136,
+      "grad_norm": 7.445247173309326,
+      "learning_rate": 7.415254237288137e-06,
+      "loss": 0.5257,
+      "step": 9100
+    },
+    {
+      "epoch": 0.7496740547588006,
+      "grad_norm": 6.337717056274414,
+      "learning_rate": 7.496740547588006e-06,
+      "loss": 0.5297,
+      "step": 9200
+    },
+    {
+      "epoch": 0.7578226857887875,
+      "grad_norm": 8.99385929107666,
+      "learning_rate": 7.578226857887876e-06,
+      "loss": 0.516,
+      "step": 9300
+    },
+    {
+      "epoch": 0.7659713168187744,
+      "grad_norm": 6.633099555969238,
+      "learning_rate": 7.659713168187744e-06,
+      "loss": 0.5012,
+      "step": 9400
+    },
+    {
+      "epoch": 0.7741199478487614,
+      "grad_norm": 5.951461315155029,
+      "learning_rate": 7.741199478487615e-06,
+      "loss": 0.5004,
+      "step": 9500
+    },
+    {
+      "epoch": 0.7822685788787483,
+      "grad_norm": 8.544897079467773,
+      "learning_rate": 7.822685788787483e-06,
+      "loss": 0.5181,
+      "step": 9600
+    },
+    {
+      "epoch": 0.7904172099087353,
+      "grad_norm": 9.489747047424316,
+      "learning_rate": 7.904172099087354e-06,
+      "loss": 0.4967,
+      "step": 9700
+    },
+    {
+      "epoch": 0.7985658409387223,
+      "grad_norm": 8.092077255249023,
+      "learning_rate": 7.985658409387224e-06,
+      "loss": 0.5054,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8067144719687093,
+      "grad_norm": 6.389376163482666,
+      "learning_rate": 8.067144719687093e-06,
+      "loss": 0.4826,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8148631029986962,
+      "grad_norm": 8.780875205993652,
+      "learning_rate": 8.148631029986964e-06,
+      "loss": 0.5211,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8230117340286832,
+      "grad_norm": 5.853449821472168,
+      "learning_rate": 8.230117340286832e-06,
+      "loss": 0.4993,
+      "step": 10100
+    },
+    {
+      "epoch": 0.8311603650586702,
+      "grad_norm": 6.378355026245117,
+      "learning_rate": 8.311603650586703e-06,
+      "loss": 0.4901,
+      "step": 10200
+    },
+    {
+      "epoch": 0.8393089960886571,
+      "grad_norm": 6.240468502044678,
+      "learning_rate": 8.393089960886572e-06,
+      "loss": 0.5026,
+      "step": 10300
+    },
+    {
+      "epoch": 0.847457627118644,
+      "grad_norm": 8.406333923339844,
+      "learning_rate": 8.47457627118644e-06,
+      "loss": 0.5139,
+      "step": 10400
+    },
+    {
+      "epoch": 0.855606258148631,
+      "grad_norm": 9.096386909484863,
+      "learning_rate": 8.556062581486311e-06,
+      "loss": 0.4982,
+      "step": 10500
+    },
+    {
+      "epoch": 0.863754889178618,
+      "grad_norm": 8.424356460571289,
+      "learning_rate": 8.63754889178618e-06,
+      "loss": 0.5216,
+      "step": 10600
+    },
+    {
+      "epoch": 0.871903520208605,
+      "grad_norm": 5.484040260314941,
+      "learning_rate": 8.71903520208605e-06,
+      "loss": 0.4975,
+      "step": 10700
+    },
+    {
+      "epoch": 0.8800521512385919,
+      "grad_norm": 6.343160152435303,
+      "learning_rate": 8.80052151238592e-06,
+      "loss": 0.5016,
+      "step": 10800
+    },
+    {
+      "epoch": 0.8882007822685789,
+      "grad_norm": 4.8916497230529785,
+      "learning_rate": 8.88200782268579e-06,
+      "loss": 0.5241,
+      "step": 10900
+    },
+    {
+      "epoch": 0.8963494132985659,
+      "grad_norm": 11.608317375183105,
+      "learning_rate": 8.96349413298566e-06,
+      "loss": 0.5009,
+      "step": 11000
+    },
+    {
+      "epoch": 0.9044980443285529,
+      "grad_norm": 6.503841876983643,
+      "learning_rate": 9.044980443285529e-06,
+      "loss": 0.491,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9126466753585397,
+      "grad_norm": 5.118412494659424,
+      "learning_rate": 9.126466753585398e-06,
+      "loss": 0.5199,
+      "step": 11200
+    },
+    {
+      "epoch": 0.9207953063885267,
+      "grad_norm": 4.837841510772705,
+      "learning_rate": 9.207953063885268e-06,
+      "loss": 0.5021,
+      "step": 11300
+    },
+    {
+      "epoch": 0.9289439374185137,
+      "grad_norm": 6.082594394683838,
+      "learning_rate": 9.289439374185137e-06,
+      "loss": 0.4881,
+      "step": 11400
+    },
+    {
+      "epoch": 0.9370925684485006,
+      "grad_norm": 8.028156280517578,
+      "learning_rate": 9.370925684485008e-06,
+      "loss": 0.4741,
+      "step": 11500
+    },
+    {
+      "epoch": 0.9452411994784876,
+      "grad_norm": 7.433921813964844,
+      "learning_rate": 9.452411994784876e-06,
+      "loss": 0.5031,
+      "step": 11600
+    },
+    {
+      "epoch": 0.9533898305084746,
+      "grad_norm": 6.140247821807861,
+      "learning_rate": 9.533898305084747e-06,
+      "loss": 0.4971,
+      "step": 11700
+    },
+    {
+      "epoch": 0.9615384615384616,
+      "grad_norm": 9.514617919921875,
+      "learning_rate": 9.615384615384616e-06,
+      "loss": 0.4782,
+      "step": 11800
+    },
+    {
+      "epoch": 0.9696870925684485,
+      "grad_norm": 6.993307590484619,
+      "learning_rate": 9.696870925684486e-06,
+      "loss": 0.481,
+      "step": 11900
+    },
+    {
+      "epoch": 0.9778357235984355,
+      "grad_norm": 5.9556989669799805,
+      "learning_rate": 9.778357235984357e-06,
+      "loss": 0.4979,
+      "step": 12000
+    },
+    {
+      "epoch": 0.9859843546284224,
+      "grad_norm": 9.89676570892334,
+      "learning_rate": 9.859843546284224e-06,
+      "loss": 0.489,
+      "step": 12100
+    },
+    {
+      "epoch": 0.9941329856584094,
+      "grad_norm": 6.464805603027344,
+      "learning_rate": 9.941329856584094e-06,
+      "loss": 0.5016,
+      "step": 12200
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.8433734939759037,
+      "eval_loss": 0.42833781242370605,
+      "eval_runtime": 6.2968,
+      "eval_samples_per_second": 395.439,
+      "eval_steps_per_second": 49.549,
+      "step": 12272
+    },
+    {
+      "epoch": 1.0022816166883963,
+      "grad_norm": 5.668147563934326,
+      "learning_rate": 9.997490221642765e-06,
+      "loss": 0.4942,
+      "step": 12300
+    },
+    {
+      "epoch": 1.0104302477183833,
+      "grad_norm": 6.818889141082764,
+      "learning_rate": 9.988526727509779e-06,
+      "loss": 0.4668,
+      "step": 12400
+    },
+    {
+      "epoch": 1.0185788787483703,
+      "grad_norm": 8.424429893493652,
+      "learning_rate": 9.979563233376793e-06,
+      "loss": 0.4502,
+      "step": 12500
+    },
+    {
+      "epoch": 1.0267275097783573,
+      "grad_norm": 6.944746971130371,
+      "learning_rate": 9.970599739243808e-06,
+      "loss": 0.4248,
+      "step": 12600
+    },
+    {
+      "epoch": 1.0348761408083442,
+      "grad_norm": 7.109178066253662,
+      "learning_rate": 9.961636245110821e-06,
+      "loss": 0.4612,
+      "step": 12700
+    },
+    {
+      "epoch": 1.0430247718383312,
+      "grad_norm": 7.571024417877197,
+      "learning_rate": 9.952672750977836e-06,
+      "loss": 0.4519,
+      "step": 12800
+    },
+    {
+      "epoch": 1.0511734028683182,
+      "grad_norm": 5.856245517730713,
+      "learning_rate": 9.943709256844851e-06,
+      "loss": 0.4469,
+      "step": 12900
+    },
+    {
+      "epoch": 1.0593220338983051,
+      "grad_norm": 7.169976234436035,
+      "learning_rate": 9.934745762711866e-06,
+      "loss": 0.4542,
+      "step": 13000
+    },
+    {
+      "epoch": 1.0674706649282921,
+      "grad_norm": 7.9827351570129395,
+      "learning_rate": 9.92578226857888e-06,
+      "loss": 0.4486,
+      "step": 13100
+    },
+    {
+      "epoch": 1.075619295958279,
+      "grad_norm": 5.570741176605225,
+      "learning_rate": 9.916818774445894e-06,
+      "loss": 0.4432,
+      "step": 13200
+    },
+    {
+      "epoch": 1.083767926988266,
+      "grad_norm": 5.507940769195557,
+      "learning_rate": 9.907855280312908e-06,
+      "loss": 0.4817,
+      "step": 13300
+    },
+    {
+      "epoch": 1.0919165580182528,
+      "grad_norm": 4.281393527984619,
+      "learning_rate": 9.898891786179923e-06,
+      "loss": 0.444,
+      "step": 13400
+    },
+    {
+      "epoch": 1.1000651890482398,
+      "grad_norm": 8.615004539489746,
+      "learning_rate": 9.889928292046936e-06,
+      "loss": 0.4556,
+      "step": 13500
+    },
+    {
+      "epoch": 1.1082138200782268,
+      "grad_norm": 6.27708625793457,
+      "learning_rate": 9.880964797913951e-06,
+      "loss": 0.4776,
+      "step": 13600
+    },
+    {
+      "epoch": 1.1163624511082137,
+      "grad_norm": 7.115694999694824,
+      "learning_rate": 9.872001303780966e-06,
+      "loss": 0.4491,
+      "step": 13700
+    },
+    {
+      "epoch": 1.1245110821382007,
+      "grad_norm": 9.465932846069336,
+      "learning_rate": 9.86303780964798e-06,
+      "loss": 0.4242,
+      "step": 13800
+    },
+    {
+      "epoch": 1.1326597131681877,
+      "grad_norm": 6.352951526641846,
+      "learning_rate": 9.854074315514994e-06,
+      "loss": 0.4598,
+      "step": 13900
+    },
+    {
+      "epoch": 1.1408083441981747,
+      "grad_norm": 6.058132171630859,
+      "learning_rate": 9.845110821382009e-06,
+      "loss": 0.4477,
+      "step": 14000
+    },
+    {
+      "epoch": 1.1489569752281616,
+      "grad_norm": 11.479142189025879,
+      "learning_rate": 9.836147327249023e-06,
+      "loss": 0.451,
+      "step": 14100
+    },
+    {
+      "epoch": 1.1571056062581486,
+      "grad_norm": 6.964011192321777,
+      "learning_rate": 9.827183833116037e-06,
+      "loss": 0.4676,
+      "step": 14200
+    },
+    {
+      "epoch": 1.1652542372881356,
+      "grad_norm": 5.015002250671387,
+      "learning_rate": 9.818220338983051e-06,
+      "loss": 0.4485,
+      "step": 14300
+    },
+    {
+      "epoch": 1.1734028683181226,
+      "grad_norm": 9.673796653747559,
+      "learning_rate": 9.809256844850066e-06,
+      "loss": 0.457,
+      "step": 14400
+    },
+    {
+      "epoch": 1.1815514993481095,
+      "grad_norm": 7.423110485076904,
+      "learning_rate": 9.80029335071708e-06,
+      "loss": 0.4452,
+      "step": 14500
+    },
+    {
+      "epoch": 1.1897001303780965,
+      "grad_norm": 7.399952411651611,
+      "learning_rate": 9.791329856584096e-06,
+      "loss": 0.445,
+      "step": 14600
+    },
+    {
+      "epoch": 1.1978487614080835,
+      "grad_norm": 7.837041854858398,
+      "learning_rate": 9.782366362451109e-06,
+      "loss": 0.4348,
+      "step": 14700
+    },
+    {
+      "epoch": 1.2059973924380705,
+      "grad_norm": 5.745814800262451,
+      "learning_rate": 9.773402868318124e-06,
+      "loss": 0.4383,
+      "step": 14800
+    },
+    {
+      "epoch": 1.2141460234680574,
+      "grad_norm": 5.527169227600098,
+      "learning_rate": 9.764439374185138e-06,
+      "loss": 0.4304,
+      "step": 14900
+    },
+    {
+      "epoch": 1.2222946544980444,
+      "grad_norm": 5.521317005157471,
+      "learning_rate": 9.755475880052152e-06,
+      "loss": 0.4348,
+      "step": 15000
+    },
+    {
+      "epoch": 1.2304432855280312,
+      "grad_norm": 6.435342788696289,
+      "learning_rate": 9.746512385919166e-06,
+      "loss": 0.4407,
+      "step": 15100
+    },
+    {
+      "epoch": 1.2385919165580184,
+      "grad_norm": 7.2671637535095215,
+      "learning_rate": 9.737548891786181e-06,
+      "loss": 0.4543,
+      "step": 15200
+    },
+    {
+      "epoch": 1.2467405475880051,
+      "grad_norm": 7.949999809265137,
+      "learning_rate": 9.728585397653194e-06,
+      "loss": 0.4422,
+      "step": 15300
+    },
+    {
+      "epoch": 1.254889178617992,
+      "grad_norm": 6.410602569580078,
+      "learning_rate": 9.719621903520209e-06,
+      "loss": 0.4526,
+      "step": 15400
+    },
+    {
+      "epoch": 1.263037809647979,
+      "grad_norm": 5.39459753036499,
+      "learning_rate": 9.710658409387224e-06,
+      "loss": 0.4536,
+      "step": 15500
+    },
+    {
+      "epoch": 1.271186440677966,
+      "grad_norm": 6.411767482757568,
+      "learning_rate": 9.701694915254239e-06,
+      "loss": 0.4434,
+      "step": 15600
+    },
+    {
+      "epoch": 1.279335071707953,
+      "grad_norm": 4.861929416656494,
+      "learning_rate": 9.692731421121252e-06,
+      "loss": 0.4281,
+      "step": 15700
+    },
+    {
+      "epoch": 1.28748370273794,
+      "grad_norm": 5.593100547790527,
+      "learning_rate": 9.683767926988267e-06,
+      "loss": 0.4557,
+      "step": 15800
+    },
+    {
+      "epoch": 1.295632333767927,
+      "grad_norm": 8.539257049560547,
+      "learning_rate": 9.674804432855281e-06,
+      "loss": 0.4487,
+      "step": 15900
+    },
+    {
+      "epoch": 1.303780964797914,
+      "grad_norm": 6.900200366973877,
+      "learning_rate": 9.665840938722295e-06,
+      "loss": 0.4285,
+      "step": 16000
+    },
+    {
+      "epoch": 1.311929595827901,
+      "grad_norm": 5.426370143890381,
+      "learning_rate": 9.65687744458931e-06,
+      "loss": 0.434,
+      "step": 16100
+    },
+    {
+      "epoch": 1.320078226857888,
+      "grad_norm": 8.011603355407715,
+      "learning_rate": 9.647913950456324e-06,
+      "loss": 0.429,
+      "step": 16200
+    },
+    {
+      "epoch": 1.3282268578878749,
+      "grad_norm": 5.003970146179199,
+      "learning_rate": 9.638950456323339e-06,
+      "loss": 0.4046,
+      "step": 16300
+    },
+    {
+      "epoch": 1.3363754889178618,
+      "grad_norm": 6.891529560089111,
+      "learning_rate": 9.629986962190354e-06,
+      "loss": 0.439,
+      "step": 16400
+    },
+    {
+      "epoch": 1.3445241199478488,
+      "grad_norm": 7.005788803100586,
+      "learning_rate": 9.621023468057367e-06,
+      "loss": 0.4422,
+      "step": 16500
+    },
+    {
+      "epoch": 1.3526727509778358,
+      "grad_norm": 9.583046913146973,
+      "learning_rate": 9.612059973924382e-06,
+      "loss": 0.4351,
+      "step": 16600
+    },
+    {
+      "epoch": 1.3608213820078228,
+      "grad_norm": 6.386882305145264,
+      "learning_rate": 9.603096479791396e-06,
+      "loss": 0.4435,
+      "step": 16700
+    },
+    {
+      "epoch": 1.3689700130378095,
+      "grad_norm": 6.294077396392822,
+      "learning_rate": 9.59413298565841e-06,
+      "loss": 0.4404,
+      "step": 16800
+    },
+    {
+      "epoch": 1.3771186440677967,
+      "grad_norm": 9.978828430175781,
+      "learning_rate": 9.585169491525424e-06,
+      "loss": 0.4334,
+      "step": 16900
+    },
+    {
+      "epoch": 1.3852672750977835,
+      "grad_norm": 9.329121589660645,
+      "learning_rate": 9.57620599739244e-06,
+      "loss": 0.4593,
+      "step": 17000
+    },
+    {
+      "epoch": 1.3934159061277707,
+      "grad_norm": 6.96733283996582,
+      "learning_rate": 9.567242503259454e-06,
+      "loss": 0.4398,
+      "step": 17100
+    },
+    {
+      "epoch": 1.4015645371577574,
+      "grad_norm": 6.355087757110596,
+      "learning_rate": 9.558279009126467e-06,
+      "loss": 0.4198,
+      "step": 17200
+    },
+    {
+      "epoch": 1.4097131681877444,
+      "grad_norm": 4.224381446838379,
+      "learning_rate": 9.549315514993482e-06,
+      "loss": 0.4598,
+      "step": 17300
+    },
+    {
+      "epoch": 1.4178617992177314,
+      "grad_norm": 5.903212070465088,
+      "learning_rate": 9.540352020860497e-06,
+      "loss": 0.4474,
+      "step": 17400
+    },
+    {
+      "epoch": 1.4260104302477183,
+      "grad_norm": 7.5130438804626465,
+      "learning_rate": 9.53138852672751e-06,
+      "loss": 0.4576,
+      "step": 17500
+    },
+    {
+      "epoch": 1.4341590612777053,
+      "grad_norm": 7.3006086349487305,
+      "learning_rate": 9.522425032594525e-06,
+      "loss": 0.4133,
+      "step": 17600
+    },
+    {
+      "epoch": 1.4423076923076923,
+      "grad_norm": 4.964842796325684,
+      "learning_rate": 9.51346153846154e-06,
+      "loss": 0.4291,
+      "step": 17700
+    },
+    {
+      "epoch": 1.4504563233376793,
+      "grad_norm": 5.327952861785889,
+      "learning_rate": 9.504498044328554e-06,
+      "loss": 0.4257,
+      "step": 17800
+    },
+    {
+      "epoch": 1.4586049543676662,
+      "grad_norm": 8.153892517089844,
+      "learning_rate": 9.495534550195569e-06,
+      "loss": 0.4369,
+      "step": 17900
+    },
+    {
+      "epoch": 1.4667535853976532,
+      "grad_norm": 6.842040061950684,
+      "learning_rate": 9.486571056062582e-06,
+      "loss": 0.434,
+      "step": 18000
+    },
+    {
+      "epoch": 1.4749022164276402,
+      "grad_norm": 10.316542625427246,
+      "learning_rate": 9.477607561929597e-06,
+      "loss": 0.4263,
+      "step": 18100
+    },
+    {
+      "epoch": 1.4830508474576272,
+      "grad_norm": 7.404006481170654,
+      "learning_rate": 9.468644067796612e-06,
+      "loss": 0.4419,
+      "step": 18200
+    },
+    {
+      "epoch": 1.4911994784876141,
+      "grad_norm": 7.187209129333496,
+      "learning_rate": 9.459680573663625e-06,
+      "loss": 0.4371,
+      "step": 18300
+    },
+    {
+      "epoch": 1.4993481095176011,
+      "grad_norm": 6.297025203704834,
+      "learning_rate": 9.45071707953064e-06,
+      "loss": 0.4355,
+      "step": 18400
+    },
+    {
+      "epoch": 1.5074967405475879,
+      "grad_norm": 5.6740875244140625,
+      "learning_rate": 9.441753585397653e-06,
+      "loss": 0.4562,
+      "step": 18500
+    },
+    {
+      "epoch": 1.515645371577575,
+      "grad_norm": 6.485991954803467,
+      "learning_rate": 9.43279009126467e-06,
+      "loss": 0.4446,
+      "step": 18600
+    },
+    {
+      "epoch": 1.5237940026075618,
+      "grad_norm": 6.18002986907959,
+      "learning_rate": 9.423826597131682e-06,
+      "loss": 0.4388,
+      "step": 18700
+    },
+    {
+      "epoch": 1.531942633637549,
+      "grad_norm": 7.745771884918213,
+      "learning_rate": 9.414863102998697e-06,
+      "loss": 0.4376,
+      "step": 18800
+    },
+    {
+      "epoch": 1.5400912646675358,
+      "grad_norm": 10.22498607635498,
+      "learning_rate": 9.405899608865712e-06,
+      "loss": 0.4334,
+      "step": 18900
+    },
+    {
+      "epoch": 1.548239895697523,
+      "grad_norm": 8.568212509155273,
+      "learning_rate": 9.396936114732725e-06,
+      "loss": 0.4492,
+      "step": 19000
+    },
+    {
+      "epoch": 1.5563885267275097,
+      "grad_norm": 7.228001117706299,
+      "learning_rate": 9.38797262059974e-06,
+      "loss": 0.4232,
+      "step": 19100
+    },
+    {
+      "epoch": 1.5645371577574967,
+      "grad_norm": 8.333541870117188,
+      "learning_rate": 9.379009126466755e-06,
+      "loss": 0.4303,
+      "step": 19200
+    },
+    {
+      "epoch": 1.5726857887874837,
+      "grad_norm": 7.778701305389404,
+      "learning_rate": 9.370045632333768e-06,
+      "loss": 0.4518,
+      "step": 19300
+    },
+    {
+      "epoch": 1.5808344198174706,
+      "grad_norm": 6.472428798675537,
+      "learning_rate": 9.361082138200784e-06,
+      "loss": 0.4354,
+      "step": 19400
+    },
+    {
+      "epoch": 1.5889830508474576,
+      "grad_norm": 8.317501068115234,
+      "learning_rate": 9.352118644067797e-06,
+      "loss": 0.4554,
+      "step": 19500
+    },
+    {
+      "epoch": 1.5971316818774446,
+      "grad_norm": 8.868863105773926,
+      "learning_rate": 9.343155149934812e-06,
+      "loss": 0.435,
+      "step": 19600
+    },
+    {
+      "epoch": 1.6052803129074316,
+      "grad_norm": 6.662153244018555,
+      "learning_rate": 9.334191655801827e-06,
+      "loss": 0.4498,
+      "step": 19700
+    },
+    {
+      "epoch": 1.6134289439374185,
+      "grad_norm": 4.642838001251221,
+      "learning_rate": 9.32522816166884e-06,
+      "loss": 0.4496,
+      "step": 19800
+    },
+    {
+      "epoch": 1.6215775749674055,
+      "grad_norm": 6.16021203994751,
+      "learning_rate": 9.316264667535855e-06,
+      "loss": 0.4477,
+      "step": 19900
+    },
+    {
+      "epoch": 1.6297262059973925,
+      "grad_norm": 6.225278377532959,
+      "learning_rate": 9.307301173402868e-06,
+      "loss": 0.4356,
+      "step": 20000
+    },
+    {
+      "epoch": 1.6378748370273795,
+      "grad_norm": 5.955763339996338,
+      "learning_rate": 9.298337679269883e-06,
+      "loss": 0.439,
+      "step": 20100
+    },
+    {
+      "epoch": 1.6460234680573662,
+      "grad_norm": 6.927220344543457,
+      "learning_rate": 9.289374185136898e-06,
+      "loss": 0.4371,
+      "step": 20200
+    },
+    {
+      "epoch": 1.6541720990873534,
+      "grad_norm": 7.151987075805664,
+      "learning_rate": 9.280410691003912e-06,
+      "loss": 0.4334,
+      "step": 20300
+    },
+    {
+      "epoch": 1.6623207301173402,
+      "grad_norm": 7.74386739730835,
+      "learning_rate": 9.271447196870927e-06,
+      "loss": 0.425,
+      "step": 20400
+    },
+    {
+      "epoch": 1.6704693611473274,
+      "grad_norm": 5.687450408935547,
+      "learning_rate": 9.26248370273794e-06,
+      "loss": 0.412,
+      "step": 20500
+    },
+    {
+      "epoch": 1.6786179921773141,
+      "grad_norm": 6.604902744293213,
+      "learning_rate": 9.253520208604955e-06,
+      "loss": 0.4534,
+      "step": 20600
+    },
+    {
+      "epoch": 1.6867666232073013,
+      "grad_norm": 6.3008623123168945,
+      "learning_rate": 9.24455671447197e-06,
+      "loss": 0.443,
+      "step": 20700
+    },
+    {
+      "epoch": 1.694915254237288,
+      "grad_norm": 4.483551502227783,
+      "learning_rate": 9.235593220338983e-06,
+      "loss": 0.4397,
+      "step": 20800
+    },
+    {
+      "epoch": 1.7030638852672753,
+      "grad_norm": 6.908230304718018,
+      "learning_rate": 9.226629726205998e-06,
+      "loss": 0.4533,
+      "step": 20900
+    },
+    {
+      "epoch": 1.711212516297262,
+      "grad_norm": 6.523016929626465,
+      "learning_rate": 9.217666232073013e-06,
+      "loss": 0.4352,
+      "step": 21000
+    },
+    {
+      "epoch": 1.719361147327249,
+      "grad_norm": 5.211183547973633,
+      "learning_rate": 9.208702737940028e-06,
+      "loss": 0.4485,
+      "step": 21100
+    },
+    {
+      "epoch": 1.727509778357236,
+      "grad_norm": 4.074219703674316,
+      "learning_rate": 9.199739243807042e-06,
+      "loss": 0.4536,
+      "step": 21200
+    },
+    {
+      "epoch": 1.735658409387223,
+      "grad_norm": 5.244410037994385,
+      "learning_rate": 9.190775749674055e-06,
+      "loss": 0.4311,
+      "step": 21300
+    },
+    {
+      "epoch": 1.74380704041721,
+      "grad_norm": 5.48605489730835,
+      "learning_rate": 9.18181225554107e-06,
+      "loss": 0.4464,
+      "step": 21400
+    },
+    {
+      "epoch": 1.7519556714471969,
+      "grad_norm": 6.30487060546875,
+      "learning_rate": 9.172848761408083e-06,
+      "loss": 0.4119,
+      "step": 21500
+    },
+    {
+      "epoch": 1.7601043024771839,
+      "grad_norm": 7.300415992736816,
+      "learning_rate": 9.163885267275098e-06,
+      "loss": 0.4118,
+      "step": 21600
+    },
+    {
+      "epoch": 1.7682529335071708,
+      "grad_norm": 7.198819160461426,
+      "learning_rate": 9.154921773142113e-06,
+      "loss": 0.425,
+      "step": 21700
+    },
+    {
+      "epoch": 1.7764015645371578,
+      "grad_norm": 7.625805854797363,
+      "learning_rate": 9.145958279009126e-06,
+      "loss": 0.454,
+      "step": 21800
+    },
+    {
+      "epoch": 1.7845501955671446,
+      "grad_norm": 5.602317810058594,
+      "learning_rate": 9.136994784876143e-06,
+      "loss": 0.4336,
+      "step": 21900
+    },
+    {
+      "epoch": 1.7926988265971318,
+      "grad_norm": 4.4945478439331055,
+      "learning_rate": 9.128031290743156e-06,
+      "loss": 0.4295,
+      "step": 22000
+    },
+    {
+      "epoch": 1.8008474576271185,
+      "grad_norm": 8.505398750305176,
+      "learning_rate": 9.11906779661017e-06,
+      "loss": 0.4481,
+      "step": 22100
+    },
+    {
+      "epoch": 1.8089960886571057,
+      "grad_norm": 6.470762729644775,
+      "learning_rate": 9.110104302477185e-06,
+      "loss": 0.4142,
+      "step": 22200
+    },
+    {
+      "epoch": 1.8171447196870925,
+      "grad_norm": 9.113410949707031,
+      "learning_rate": 9.101140808344198e-06,
+      "loss": 0.4169,
+      "step": 22300
+    },
+    {
+      "epoch": 1.8252933507170797,
+      "grad_norm": 5.110347270965576,
+      "learning_rate": 9.092177314211213e-06,
+      "loss": 0.4283,
+      "step": 22400
+    },
+    {
+      "epoch": 1.8334419817470664,
+      "grad_norm": 7.161343574523926,
+      "learning_rate": 9.083213820078228e-06,
+      "loss": 0.4499,
+      "step": 22500
+    },
+    {
+      "epoch": 1.8415906127770536,
+      "grad_norm": 7.881711959838867,
+      "learning_rate": 9.074250325945243e-06,
+      "loss": 0.4069,
+      "step": 22600
+    },
+    {
+      "epoch": 1.8497392438070404,
+      "grad_norm": 6.091899871826172,
+      "learning_rate": 9.065286831812258e-06,
+      "loss": 0.4315,
+      "step": 22700
+    },
+    {
+      "epoch": 1.8578878748370273,
+      "grad_norm": 6.5380167961120605,
+      "learning_rate": 9.05632333767927e-06,
+      "loss": 0.431,
+      "step": 22800
+    },
+    {
+      "epoch": 1.8660365058670143,
+      "grad_norm": 6.479442119598389,
+      "learning_rate": 9.047359843546286e-06,
+      "loss": 0.4299,
+      "step": 22900
+    },
+    {
+      "epoch": 1.8741851368970013,
+      "grad_norm": 8.060819625854492,
+      "learning_rate": 9.038396349413299e-06,
+      "loss": 0.4204,
+      "step": 23000
+    },
+    {
+      "epoch": 1.8823337679269883,
+      "grad_norm": 4.994590759277344,
+      "learning_rate": 9.029432855280313e-06,
+      "loss": 0.44,
+      "step": 23100
+    },
+    {
+      "epoch": 1.8904823989569752,
+      "grad_norm": 6.113391876220703,
+      "learning_rate": 9.020469361147328e-06,
+      "loss": 0.448,
+      "step": 23200
+    },
+    {
+      "epoch": 1.8986310299869622,
+      "grad_norm": 5.091489315032959,
+      "learning_rate": 9.011505867014341e-06,
+      "loss": 0.4615,
+      "step": 23300
+    },
+    {
+      "epoch": 1.9067796610169492,
+      "grad_norm": 5.466245651245117,
+      "learning_rate": 9.002542372881356e-06,
+      "loss": 0.4456,
+      "step": 23400
+    },
+    {
+      "epoch": 1.9149282920469362,
+      "grad_norm": 7.680745601654053,
+      "learning_rate": 8.993578878748371e-06,
+      "loss": 0.4207,
+      "step": 23500
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 6.748586177825928,
+      "learning_rate": 8.984615384615386e-06,
+      "loss": 0.4234,
+      "step": 23600
+    },
+    {
+      "epoch": 1.93122555410691,
+      "grad_norm": 7.408194541931152,
+      "learning_rate": 8.9756518904824e-06,
+      "loss": 0.4345,
+      "step": 23700
+    },
+    {
+      "epoch": 1.9393741851368969,
+      "grad_norm": 9.398754119873047,
+      "learning_rate": 8.966688396349414e-06,
+      "loss": 0.4267,
+      "step": 23800
+    },
+    {
+      "epoch": 1.947522816166884,
+      "grad_norm": 8.72804069519043,
+      "learning_rate": 8.957724902216428e-06,
+      "loss": 0.4222,
+      "step": 23900
+    },
+    {
+      "epoch": 1.9556714471968708,
+      "grad_norm": 3.8285768032073975,
+      "learning_rate": 8.948761408083443e-06,
+      "loss": 0.4295,
+      "step": 24000
+    },
+    {
+      "epoch": 1.963820078226858,
+      "grad_norm": 7.150258541107178,
+      "learning_rate": 8.939797913950456e-06,
+      "loss": 0.4308,
+      "step": 24100
+    },
+    {
+      "epoch": 1.9719687092568448,
+      "grad_norm": 5.6762495040893555,
+      "learning_rate": 8.930834419817473e-06,
+      "loss": 0.4315,
+      "step": 24200
+    },
+    {
+      "epoch": 1.980117340286832,
+      "grad_norm": 6.01600980758667,
+      "learning_rate": 8.921870925684486e-06,
+      "loss": 0.4479,
+      "step": 24300
+    },
+    {
+      "epoch": 1.9882659713168187,
+      "grad_norm": 5.277040958404541,
+      "learning_rate": 8.9129074315515e-06,
+      "loss": 0.4357,
+      "step": 24400
+    },
+    {
+      "epoch": 1.996414602346806,
+      "grad_norm": 7.153831958770752,
+      "learning_rate": 8.903943937418514e-06,
+      "loss": 0.427,
+      "step": 24500
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.8325301204819278,
+      "eval_loss": 0.4437418580055237,
+      "eval_runtime": 5.5267,
+      "eval_samples_per_second": 450.541,
+      "eval_steps_per_second": 56.453,
+      "step": 24544
+    },
+    {
+      "epoch": 2.0045632333767927,
+      "grad_norm": 8.098788261413574,
+      "learning_rate": 8.894980443285529e-06,
+      "loss": 0.3801,
+      "step": 24600
+    },
+    {
+      "epoch": 2.01271186440678,
+      "grad_norm": 6.156032085418701,
+      "learning_rate": 8.886016949152544e-06,
+      "loss": 0.3587,
+      "step": 24700
+    },
+    {
+      "epoch": 2.0208604954367666,
+      "grad_norm": 3.556076765060425,
+      "learning_rate": 8.877053455019557e-06,
+      "loss": 0.3615,
+      "step": 24800
+    },
+    {
+      "epoch": 2.029009126466754,
+      "grad_norm": 7.757440090179443,
+      "learning_rate": 8.868089960886571e-06,
+      "loss": 0.3457,
+      "step": 24900
+    },
+    {
+      "epoch": 2.0371577574967406,
+      "grad_norm": 3.515918493270874,
+      "learning_rate": 8.859126466753586e-06,
+      "loss": 0.363,
+      "step": 25000
+    },
+    {
+      "epoch": 2.0453063885267273,
+      "grad_norm": 4.51107931137085,
+      "learning_rate": 8.850162972620601e-06,
+      "loss": 0.3352,
+      "step": 25100
+    },
+    {
+      "epoch": 2.0534550195567145,
+      "grad_norm": 5.263424396514893,
+      "learning_rate": 8.841199478487616e-06,
+      "loss": 0.3589,
+      "step": 25200
+    },
+    {
+      "epoch": 2.0616036505867013,
+      "grad_norm": 8.238348960876465,
+      "learning_rate": 8.832235984354629e-06,
+      "loss": 0.3312,
+      "step": 25300
+    },
+    {
+      "epoch": 2.0697522816166884,
+      "grad_norm": 6.862210750579834,
+      "learning_rate": 8.823272490221644e-06,
+      "loss": 0.3555,
+      "step": 25400
+    },
+    {
+      "epoch": 2.077900912646675,
+      "grad_norm": 11.949597358703613,
+      "learning_rate": 8.814308996088659e-06,
+      "loss": 0.3522,
+      "step": 25500
+    },
+    {
+      "epoch": 2.0860495436766624,
+      "grad_norm": 6.46090030670166,
+      "learning_rate": 8.805345501955672e-06,
+      "loss": 0.3373,
+      "step": 25600
+    },
+    {
+      "epoch": 2.094198174706649,
+      "grad_norm": 6.2601637840271,
+      "learning_rate": 8.796382007822686e-06,
+      "loss": 0.3587,
+      "step": 25700
+    },
+    {
+      "epoch": 2.1023468057366363,
+      "grad_norm": 6.387700080871582,
+      "learning_rate": 8.787418513689701e-06,
+      "loss": 0.3746,
+      "step": 25800
+    },
+    {
+      "epoch": 2.110495436766623,
+      "grad_norm": 8.692789077758789,
+      "learning_rate": 8.778455019556714e-06,
+      "loss": 0.3499,
+      "step": 25900
+    },
+    {
+      "epoch": 2.1186440677966103,
+      "grad_norm": 4.821929454803467,
+      "learning_rate": 8.76949152542373e-06,
+      "loss": 0.3605,
+      "step": 26000
+    },
+    {
+      "epoch": 2.126792698826597,
+      "grad_norm": 8.666610717773438,
+      "learning_rate": 8.760528031290744e-06,
+      "loss": 0.3462,
+      "step": 26100
+    },
+    {
+      "epoch": 2.1349413298565842,
+      "grad_norm": 7.047764301300049,
+      "learning_rate": 8.751564537157759e-06,
+      "loss": 0.3573,
+      "step": 26200
+    },
+    {
+      "epoch": 2.143089960886571,
+      "grad_norm": 6.5750837326049805,
+      "learning_rate": 8.742601043024772e-06,
+      "loss": 0.3625,
+      "step": 26300
+    },
+    {
+      "epoch": 2.151238591916558,
+      "grad_norm": 8.66965389251709,
+      "learning_rate": 8.733637548891787e-06,
+      "loss": 0.3656,
+      "step": 26400
+    },
+    {
+      "epoch": 2.159387222946545,
+      "grad_norm": 8.956038475036621,
+      "learning_rate": 8.724674054758802e-06,
+      "loss": 0.3553,
+      "step": 26500
+    },
+    {
+      "epoch": 2.167535853976532,
+      "grad_norm": 9.39869499206543,
+      "learning_rate": 8.715710560625815e-06,
+      "loss": 0.3388,
+      "step": 26600
+    },
+    {
+      "epoch": 2.175684485006519,
+      "grad_norm": 7.84521484375,
+      "learning_rate": 8.706747066492831e-06,
+      "loss": 0.3521,
+      "step": 26700
+    },
+    {
+      "epoch": 2.1838331160365057,
+      "grad_norm": 8.594714164733887,
+      "learning_rate": 8.697783572359844e-06,
+      "loss": 0.3451,
+      "step": 26800
+    },
+    {
+      "epoch": 2.191981747066493,
+      "grad_norm": 5.132546424865723,
+      "learning_rate": 8.688820078226859e-06,
+      "loss": 0.3568,
+      "step": 26900
+    },
+    {
+      "epoch": 2.2001303780964796,
+      "grad_norm": 6.885127067565918,
+      "learning_rate": 8.679856584093874e-06,
+      "loss": 0.3603,
+      "step": 27000
+    },
+    {
+      "epoch": 2.208279009126467,
+      "grad_norm": 3.2750186920166016,
+      "learning_rate": 8.670893089960887e-06,
+      "loss": 0.3548,
+      "step": 27100
+    },
+    {
+      "epoch": 2.2164276401564535,
+      "grad_norm": 6.619633674621582,
+      "learning_rate": 8.661929595827902e-06,
+      "loss": 0.3442,
+      "step": 27200
+    },
+    {
+      "epoch": 2.2245762711864407,
+      "grad_norm": 5.827129364013672,
+      "learning_rate": 8.652966101694917e-06,
+      "loss": 0.3641,
+      "step": 27300
+    },
+    {
+      "epoch": 2.2327249022164275,
+      "grad_norm": 5.905134677886963,
+      "learning_rate": 8.644002607561931e-06,
+      "loss": 0.3738,
+      "step": 27400
+    },
+    {
+      "epoch": 2.2408735332464147,
+      "grad_norm": 7.653305530548096,
+      "learning_rate": 8.635039113428944e-06,
+      "loss": 0.3315,
+      "step": 27500
+    },
+    {
+      "epoch": 2.2490221642764014,
+      "grad_norm": 4.101062297821045,
+      "learning_rate": 8.62607561929596e-06,
+      "loss": 0.3647,
+      "step": 27600
+    },
+    {
+      "epoch": 2.2571707953063886,
+      "grad_norm": 5.449923992156982,
+      "learning_rate": 8.617112125162974e-06,
+      "loss": 0.3537,
+      "step": 27700
+    },
+    {
+      "epoch": 2.2653194263363754,
+      "grad_norm": 7.924961566925049,
+      "learning_rate": 8.608148631029987e-06,
+      "loss": 0.3494,
+      "step": 27800
+    },
+    {
+      "epoch": 2.2734680573663626,
+      "grad_norm": 7.311993598937988,
+      "learning_rate": 8.599185136897002e-06,
+      "loss": 0.3791,
+      "step": 27900
+    },
+    {
+      "epoch": 2.2816166883963493,
+      "grad_norm": 9.550232887268066,
+      "learning_rate": 8.590221642764017e-06,
+      "loss": 0.3758,
+      "step": 28000
+    },
+    {
+      "epoch": 2.2897653194263365,
+      "grad_norm": 8.781049728393555,
+      "learning_rate": 8.58125814863103e-06,
+      "loss": 0.3508,
+      "step": 28100
+    },
+    {
+      "epoch": 2.2979139504563233,
+      "grad_norm": 12.556912422180176,
+      "learning_rate": 8.572294654498045e-06,
+      "loss": 0.3781,
+      "step": 28200
+    },
+    {
+      "epoch": 2.3060625814863105,
+      "grad_norm": 7.472509860992432,
+      "learning_rate": 8.56333116036506e-06,
+      "loss": 0.3532,
+      "step": 28300
+    },
+    {
+      "epoch": 2.3142112125162972,
+      "grad_norm": 7.974531173706055,
+      "learning_rate": 8.554367666232074e-06,
+      "loss": 0.3736,
+      "step": 28400
+    },
+    {
+      "epoch": 2.322359843546284,
+      "grad_norm": 3.605121374130249,
+      "learning_rate": 8.545404172099089e-06,
+      "loss": 0.3735,
+      "step": 28500
+    },
+    {
+      "epoch": 2.330508474576271,
+      "grad_norm": 6.731493949890137,
+      "learning_rate": 8.536440677966102e-06,
+      "loss": 0.3744,
+      "step": 28600
+    },
+    {
+      "epoch": 2.3386571056062584,
+      "grad_norm": 4.8290839195251465,
+      "learning_rate": 8.527477183833117e-06,
+      "loss": 0.3313,
+      "step": 28700
+    },
+    {
+      "epoch": 2.346805736636245,
+      "grad_norm": 7.126736640930176,
+      "learning_rate": 8.518513689700132e-06,
+      "loss": 0.3748,
+      "step": 28800
+    },
+    {
+      "epoch": 2.354954367666232,
+      "grad_norm": 8.92186450958252,
+      "learning_rate": 8.509550195567145e-06,
+      "loss": 0.3766,
+      "step": 28900
+    },
+    {
+      "epoch": 2.363102998696219,
+      "grad_norm": 4.0069146156311035,
+      "learning_rate": 8.50058670143416e-06,
+      "loss": 0.3667,
+      "step": 29000
+    },
+    {
+      "epoch": 2.371251629726206,
+      "grad_norm": 4.4215168952941895,
+      "learning_rate": 8.491623207301173e-06,
+      "loss": 0.366,
+      "step": 29100
+    },
+    {
+      "epoch": 2.379400260756193,
+      "grad_norm": 6.785125255584717,
+      "learning_rate": 8.48265971316819e-06,
+      "loss": 0.3718,
+      "step": 29200
+    },
+    {
+      "epoch": 2.38754889178618,
+      "grad_norm": 5.78675651550293,
+      "learning_rate": 8.473696219035202e-06,
+      "loss": 0.3529,
+      "step": 29300
+    },
+    {
+      "epoch": 2.395697522816167,
+      "grad_norm": 8.054706573486328,
+      "learning_rate": 8.464732724902217e-06,
+      "loss": 0.3755,
+      "step": 29400
+    },
+    {
+      "epoch": 2.4038461538461537,
+      "grad_norm": 6.803676605224609,
+      "learning_rate": 8.455769230769232e-06,
+      "loss": 0.3157,
+      "step": 29500
+    },
+    {
+      "epoch": 2.411994784876141,
+      "grad_norm": 8.307502746582031,
+      "learning_rate": 8.446805736636245e-06,
+      "loss": 0.3489,
+      "step": 29600
+    },
+    {
+      "epoch": 2.4201434159061277,
+      "grad_norm": 8.928215026855469,
+      "learning_rate": 8.43784224250326e-06,
+      "loss": 0.3793,
+      "step": 29700
+    },
+    {
+      "epoch": 2.428292046936115,
+      "grad_norm": 6.0352396965026855,
+      "learning_rate": 8.428878748370275e-06,
+      "loss": 0.3612,
+      "step": 29800
+    },
+    {
+      "epoch": 2.4364406779661016,
+      "grad_norm": 8.676813125610352,
+      "learning_rate": 8.41991525423729e-06,
+      "loss": 0.3548,
+      "step": 29900
+    },
+    {
+      "epoch": 2.444589308996089,
+      "grad_norm": 8.409051895141602,
+      "learning_rate": 8.410951760104304e-06,
+      "loss": 0.3733,
+      "step": 30000
+    },
+    {
+      "epoch": 2.4527379400260756,
+      "grad_norm": 12.27320671081543,
+      "learning_rate": 8.401988265971318e-06,
+      "loss": 0.3656,
+      "step": 30100
+    },
+    {
+      "epoch": 2.4608865710560623,
+      "grad_norm": 5.953177452087402,
+      "learning_rate": 8.393024771838332e-06,
+      "loss": 0.3513,
+      "step": 30200
+    },
+    {
+      "epoch": 2.4690352020860495,
+      "grad_norm": 3.4694623947143555,
+      "learning_rate": 8.384061277705347e-06,
+      "loss": 0.3439,
+      "step": 30300
+    },
+    {
+      "epoch": 2.4771838331160367,
+      "grad_norm": 7.406796455383301,
+      "learning_rate": 8.37509778357236e-06,
+      "loss": 0.3431,
+      "step": 30400
+    },
+    {
+      "epoch": 2.4853324641460235,
+      "grad_norm": 11.077946662902832,
+      "learning_rate": 8.366134289439375e-06,
+      "loss": 0.3387,
+      "step": 30500
+    },
+    {
+      "epoch": 2.4934810951760102,
+      "grad_norm": 5.576053142547607,
+      "learning_rate": 8.357170795306388e-06,
+      "loss": 0.3635,
+      "step": 30600
+    },
+    {
+      "epoch": 2.5016297262059974,
+      "grad_norm": 5.933220863342285,
+      "learning_rate": 8.348207301173403e-06,
+      "loss": 0.3555,
+      "step": 30700
+    },
+    {
+      "epoch": 2.509778357235984,
+      "grad_norm": 5.958785533905029,
+      "learning_rate": 8.339243807040418e-06,
+      "loss": 0.3586,
+      "step": 30800
+    },
+    {
+      "epoch": 2.5179269882659714,
+      "grad_norm": 6.1259446144104,
+      "learning_rate": 8.330280312907433e-06,
+      "loss": 0.3526,
+      "step": 30900
+    },
+    {
+      "epoch": 2.526075619295958,
+      "grad_norm": 4.747119426727295,
+      "learning_rate": 8.321316818774447e-06,
+      "loss": 0.3525,
+      "step": 31000
+    },
+    {
+      "epoch": 2.5342242503259453,
+      "grad_norm": 5.299834251403809,
+      "learning_rate": 8.31235332464146e-06,
+      "loss": 0.3445,
+      "step": 31100
+    },
+    {
+      "epoch": 2.542372881355932,
+      "grad_norm": 8.24792194366455,
+      "learning_rate": 8.303389830508475e-06,
+      "loss": 0.3562,
+      "step": 31200
+    },
+    {
+      "epoch": 2.5505215123859193,
+      "grad_norm": 4.663332462310791,
+      "learning_rate": 8.29442633637549e-06,
+      "loss": 0.3497,
+      "step": 31300
+    },
+    {
+      "epoch": 2.558670143415906,
+      "grad_norm": 6.58770227432251,
+      "learning_rate": 8.285462842242503e-06,
+      "loss": 0.3528,
+      "step": 31400
+    },
+    {
+      "epoch": 2.5668187744458932,
+      "grad_norm": 6.949512481689453,
+      "learning_rate": 8.27649934810952e-06,
+      "loss": 0.3743,
+      "step": 31500
+    },
+    {
+      "epoch": 2.57496740547588,
+      "grad_norm": 7.117619514465332,
+      "learning_rate": 8.267535853976533e-06,
+      "loss": 0.369,
+      "step": 31600
+    },
+    {
+      "epoch": 2.583116036505867,
+      "grad_norm": 5.4686689376831055,
+      "learning_rate": 8.258572359843548e-06,
+      "loss": 0.3577,
+      "step": 31700
+    },
+    {
+      "epoch": 2.591264667535854,
+      "grad_norm": 7.302579879760742,
+      "learning_rate": 8.249608865710562e-06,
+      "loss": 0.3454,
+      "step": 31800
+    },
+    {
+      "epoch": 2.5994132985658407,
+      "grad_norm": 3.9816572666168213,
+      "learning_rate": 8.240645371577576e-06,
+      "loss": 0.3506,
+      "step": 31900
+    },
+    {
+      "epoch": 2.607561929595828,
+      "grad_norm": 11.21993350982666,
+      "learning_rate": 8.23168187744459e-06,
+      "loss": 0.3461,
+      "step": 32000
+    },
+    {
+      "epoch": 2.615710560625815,
+      "grad_norm": 6.287281513214111,
+      "learning_rate": 8.222718383311603e-06,
+      "loss": 0.336,
+      "step": 32100
+    },
+    {
+      "epoch": 2.623859191655802,
+      "grad_norm": 6.645712852478027,
+      "learning_rate": 8.213754889178618e-06,
+      "loss": 0.3675,
+      "step": 32200
+    },
+    {
+      "epoch": 2.6320078226857886,
+      "grad_norm": 4.677251815795898,
+      "learning_rate": 8.204791395045633e-06,
+      "loss": 0.3517,
+      "step": 32300
+    },
+    {
+      "epoch": 2.640156453715776,
+      "grad_norm": 5.261903285980225,
+      "learning_rate": 8.195827900912648e-06,
+      "loss": 0.3595,
+      "step": 32400
+    },
+    {
+      "epoch": 2.648305084745763,
+      "grad_norm": 6.755844593048096,
+      "learning_rate": 8.186864406779663e-06,
+      "loss": 0.3592,
+      "step": 32500
+    },
+    {
+      "epoch": 2.6564537157757497,
+      "grad_norm": 9.253024101257324,
+      "learning_rate": 8.177900912646676e-06,
+      "loss": 0.3655,
+      "step": 32600
+    },
+    {
+      "epoch": 2.6646023468057365,
+      "grad_norm": 7.2145490646362305,
+      "learning_rate": 8.16893741851369e-06,
+      "loss": 0.3573,
+      "step": 32700
+    },
+    {
+      "epoch": 2.6727509778357237,
+      "grad_norm": 5.882996082305908,
+      "learning_rate": 8.159973924380705e-06,
+      "loss": 0.3555,
+      "step": 32800
+    },
+    {
+      "epoch": 2.6808996088657104,
+      "grad_norm": 13.761658668518066,
+      "learning_rate": 8.151010430247718e-06,
+      "loss": 0.3631,
+      "step": 32900
+    },
+    {
+      "epoch": 2.6890482398956976,
+      "grad_norm": 6.742971420288086,
+      "learning_rate": 8.142046936114733e-06,
+      "loss": 0.3714,
+      "step": 33000
+    },
+    {
+      "epoch": 2.6971968709256844,
+      "grad_norm": 5.441837787628174,
+      "learning_rate": 8.133083441981748e-06,
+      "loss": 0.35,
+      "step": 33100
+    },
+    {
+      "epoch": 2.7053455019556716,
+      "grad_norm": 8.888238906860352,
+      "learning_rate": 8.124119947848761e-06,
+      "loss": 0.349,
+      "step": 33200
+    },
+    {
+      "epoch": 2.7134941329856583,
+      "grad_norm": 6.0085368156433105,
+      "learning_rate": 8.115156453715778e-06,
+      "loss": 0.3555,
+      "step": 33300
+    },
+    {
+      "epoch": 2.7216427640156455,
+      "grad_norm": 8.16164493560791,
+      "learning_rate": 8.10619295958279e-06,
+      "loss": 0.361,
+      "step": 33400
+    },
+    {
+      "epoch": 2.7297913950456323,
+      "grad_norm": 7.598118782043457,
+      "learning_rate": 8.097229465449806e-06,
+      "loss": 0.3655,
+      "step": 33500
+    },
+    {
+      "epoch": 2.737940026075619,
+      "grad_norm": 7.139162540435791,
+      "learning_rate": 8.088265971316819e-06,
+      "loss": 0.3604,
+      "step": 33600
+    },
+    {
+      "epoch": 2.7460886571056062,
+      "grad_norm": 6.311704158782959,
+      "learning_rate": 8.079302477183834e-06,
+      "loss": 0.354,
+      "step": 33700
+    },
+    {
+      "epoch": 2.7542372881355934,
+      "grad_norm": 5.32002592086792,
+      "learning_rate": 8.070338983050848e-06,
+      "loss": 0.3655,
+      "step": 33800
+    },
+    {
+      "epoch": 2.76238591916558,
+      "grad_norm": 4.203720569610596,
+      "learning_rate": 8.061375488917861e-06,
+      "loss": 0.3611,
+      "step": 33900
+    },
+    {
+      "epoch": 2.770534550195567,
+      "grad_norm": 9.162140846252441,
+      "learning_rate": 8.052411994784878e-06,
+      "loss": 0.3633,
+      "step": 34000
+    },
+    {
+      "epoch": 2.778683181225554,
+      "grad_norm": 5.322396755218506,
+      "learning_rate": 8.043448500651891e-06,
+      "loss": 0.3614,
+      "step": 34100
+    },
+    {
+      "epoch": 2.7868318122555413,
+      "grad_norm": 8.66873550415039,
+      "learning_rate": 8.034485006518906e-06,
+      "loss": 0.3522,
+      "step": 34200
+    },
+    {
+      "epoch": 2.794980443285528,
+      "grad_norm": 7.859837055206299,
+      "learning_rate": 8.02552151238592e-06,
+      "loss": 0.3562,
+      "step": 34300
+    },
+    {
+      "epoch": 2.803129074315515,
+      "grad_norm": 6.7927398681640625,
+      "learning_rate": 8.016558018252934e-06,
+      "loss": 0.3662,
+      "step": 34400
+    },
+    {
+      "epoch": 2.811277705345502,
+      "grad_norm": 5.269212245941162,
+      "learning_rate": 8.007594524119949e-06,
+      "loss": 0.3509,
+      "step": 34500
+    },
+    {
+      "epoch": 2.819426336375489,
+      "grad_norm": 5.3910017013549805,
+      "learning_rate": 7.998631029986963e-06,
+      "loss": 0.3618,
+      "step": 34600
+    },
+    {
+      "epoch": 2.827574967405476,
+      "grad_norm": 6.497340202331543,
+      "learning_rate": 7.989667535853978e-06,
+      "loss": 0.338,
+      "step": 34700
+    },
+    {
+      "epoch": 2.8357235984354627,
+      "grad_norm": 9.944086074829102,
+      "learning_rate": 7.980704041720991e-06,
+      "loss": 0.3523,
+      "step": 34800
+    },
+    {
+      "epoch": 2.84387222946545,
+      "grad_norm": 4.036046504974365,
+      "learning_rate": 7.971740547588006e-06,
+      "loss": 0.3458,
+      "step": 34900
+    },
+    {
+      "epoch": 2.8520208604954367,
+      "grad_norm": 7.9690327644348145,
+      "learning_rate": 7.962777053455021e-06,
+      "loss": 0.3469,
+      "step": 35000
+    },
+    {
+      "epoch": 2.860169491525424,
+      "grad_norm": 5.498697757720947,
+      "learning_rate": 7.953813559322034e-06,
+      "loss": 0.3703,
+      "step": 35100
+    },
+    {
+      "epoch": 2.8683181225554106,
+      "grad_norm": 8.990461349487305,
+      "learning_rate": 7.944850065189049e-06,
+      "loss": 0.3463,
+      "step": 35200
+    },
+    {
+      "epoch": 2.8764667535853974,
+      "grad_norm": 5.683210372924805,
+      "learning_rate": 7.935886571056064e-06,
+      "loss": 0.3285,
+      "step": 35300
+    },
+    {
+      "epoch": 2.8846153846153846,
+      "grad_norm": 11.34899616241455,
+      "learning_rate": 7.926923076923077e-06,
+      "loss": 0.3638,
+      "step": 35400
+    },
+    {
+      "epoch": 2.8927640156453718,
+      "grad_norm": 8.744894027709961,
+      "learning_rate": 7.917959582790092e-06,
+      "loss": 0.3575,
+      "step": 35500
+    },
+    {
+      "epoch": 2.9009126466753585,
+      "grad_norm": 8.961909294128418,
+      "learning_rate": 7.908996088657106e-06,
+      "loss": 0.3367,
+      "step": 35600
+    },
+    {
+      "epoch": 2.9090612777053453,
+      "grad_norm": 9.418771743774414,
+      "learning_rate": 7.900032594524121e-06,
+      "loss": 0.3575,
+      "step": 35700
+    },
+    {
+      "epoch": 2.9172099087353325,
+      "grad_norm": 10.538002967834473,
+      "learning_rate": 7.891069100391136e-06,
+      "loss": 0.376,
+      "step": 35800
+    },
+    {
+      "epoch": 2.9253585397653197,
+      "grad_norm": 4.822318077087402,
+      "learning_rate": 7.882105606258149e-06,
+      "loss": 0.3468,
+      "step": 35900
+    },
+    {
+      "epoch": 2.9335071707953064,
+      "grad_norm": 9.379559516906738,
+      "learning_rate": 7.873142112125164e-06,
+      "loss": 0.3506,
+      "step": 36000
+    },
+    {
+      "epoch": 2.941655801825293,
+      "grad_norm": 11.871332168579102,
+      "learning_rate": 7.864178617992179e-06,
+      "loss": 0.3494,
+      "step": 36100
+    },
+    {
+      "epoch": 2.9498044328552804,
+      "grad_norm": 5.834349632263184,
+      "learning_rate": 7.855215123859192e-06,
+      "loss": 0.3951,
+      "step": 36200
+    },
+    {
+      "epoch": 2.957953063885267,
+      "grad_norm": 9.025989532470703,
+      "learning_rate": 7.846251629726208e-06,
+      "loss": 0.3642,
+      "step": 36300
+    },
+    {
+      "epoch": 2.9661016949152543,
+      "grad_norm": 7.520548343658447,
+      "learning_rate": 7.837288135593221e-06,
+      "loss": 0.3668,
+      "step": 36400
+    },
+    {
+      "epoch": 2.974250325945241,
+      "grad_norm": 7.14879035949707,
+      "learning_rate": 7.828324641460236e-06,
+      "loss": 0.3696,
+      "step": 36500
+    },
+    {
+      "epoch": 2.9823989569752283,
+      "grad_norm": 8.017966270446777,
+      "learning_rate": 7.81936114732725e-06,
+      "loss": 0.3637,
+      "step": 36600
+    },
+    {
+      "epoch": 2.990547588005215,
+      "grad_norm": 7.470760345458984,
+      "learning_rate": 7.810397653194264e-06,
+      "loss": 0.3586,
+      "step": 36700
+    },
+    {
+      "epoch": 2.9986962190352022,
+      "grad_norm": 4.955636501312256,
+      "learning_rate": 7.801434159061279e-06,
+      "loss": 0.3493,
+      "step": 36800
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.8558232931726908,
+      "eval_loss": 0.4229515790939331,
+      "eval_runtime": 5.5172,
+      "eval_samples_per_second": 451.314,
+      "eval_steps_per_second": 56.55,
+      "step": 36816
+    },
+    {
+      "epoch": 3.006844850065189,
+      "grad_norm": 11.195860862731934,
+      "learning_rate": 7.792470664928292e-06,
+      "loss": 0.303,
+      "step": 36900
+    },
+    {
+      "epoch": 3.014993481095176,
+      "grad_norm": 8.619274139404297,
+      "learning_rate": 7.783507170795307e-06,
+      "loss": 0.2999,
+      "step": 37000
+    },
+    {
+      "epoch": 3.023142112125163,
+      "grad_norm": 7.2149200439453125,
+      "learning_rate": 7.774543676662322e-06,
+      "loss": 0.2997,
+      "step": 37100
+    },
+    {
+      "epoch": 3.03129074315515,
+      "grad_norm": 10.171060562133789,
+      "learning_rate": 7.765580182529336e-06,
+      "loss": 0.302,
+      "step": 37200
+    },
+    {
+      "epoch": 3.039439374185137,
+      "grad_norm": 10.720246315002441,
+      "learning_rate": 7.756616688396351e-06,
+      "loss": 0.2884,
+      "step": 37300
+    },
+    {
+      "epoch": 3.047588005215124,
+      "grad_norm": 7.545303821563721,
+      "learning_rate": 7.747653194263364e-06,
+      "loss": 0.3076,
+      "step": 37400
+    },
+    {
+      "epoch": 3.055736636245111,
+      "grad_norm": 7.973757266998291,
+      "learning_rate": 7.73868970013038e-06,
+      "loss": 0.2762,
+      "step": 37500
+    },
+    {
+      "epoch": 3.0638852672750976,
+      "grad_norm": 7.613004207611084,
+      "learning_rate": 7.729726205997394e-06,
+      "loss": 0.285,
+      "step": 37600
+    },
+    {
+      "epoch": 3.0720338983050848,
+      "grad_norm": 5.3108367919921875,
+      "learning_rate": 7.720762711864407e-06,
+      "loss": 0.2777,
+      "step": 37700
+    },
+    {
+      "epoch": 3.0801825293350715,
+      "grad_norm": 4.986673355102539,
+      "learning_rate": 7.711799217731422e-06,
+      "loss": 0.2718,
+      "step": 37800
+    },
+    {
+      "epoch": 3.0883311603650587,
+      "grad_norm": 8.627853393554688,
+      "learning_rate": 7.702835723598437e-06,
+      "loss": 0.3053,
+      "step": 37900
+    },
+    {
+      "epoch": 3.0964797913950455,
+      "grad_norm": 5.650912284851074,
+      "learning_rate": 7.69387222946545e-06,
+      "loss": 0.2761,
+      "step": 38000
+    },
+    {
+      "epoch": 3.1046284224250327,
+      "grad_norm": 8.650264739990234,
+      "learning_rate": 7.684908735332465e-06,
+      "loss": 0.3097,
+      "step": 38100
+    },
+    {
+      "epoch": 3.1127770534550194,
+      "grad_norm": 3.7856078147888184,
+      "learning_rate": 7.67594524119948e-06,
+      "loss": 0.2875,
+      "step": 38200
+    },
+    {
+      "epoch": 3.1209256844850066,
+      "grad_norm": 10.533018112182617,
+      "learning_rate": 7.666981747066493e-06,
+      "loss": 0.2876,
+      "step": 38300
+    },
+    {
+      "epoch": 3.1290743155149934,
+      "grad_norm": 5.538301467895508,
+      "learning_rate": 7.658018252933507e-06,
+      "loss": 0.2934,
+      "step": 38400
+    },
+    {
+      "epoch": 3.1372229465449806,
+      "grad_norm": 12.393535614013672,
+      "learning_rate": 7.649054758800522e-06,
+      "loss": 0.2992,
+      "step": 38500
+    },
+    {
+      "epoch": 3.1453715775749673,
+      "grad_norm": 6.047006130218506,
+      "learning_rate": 7.640091264667535e-06,
+      "loss": 0.2723,
+      "step": 38600
+    },
+    {
+      "epoch": 3.1535202086049545,
+      "grad_norm": 7.563205718994141,
+      "learning_rate": 7.63112777053455e-06,
+      "loss": 0.2927,
+      "step": 38700
+    },
+    {
+      "epoch": 3.1616688396349413,
+      "grad_norm": 10.292147636413574,
+      "learning_rate": 7.622164276401565e-06,
+      "loss": 0.267,
+      "step": 38800
+    },
+    {
+      "epoch": 3.1698174706649285,
+      "grad_norm": 8.638325691223145,
+      "learning_rate": 7.613200782268578e-06,
+      "loss": 0.292,
+      "step": 38900
+    },
+    {
+      "epoch": 3.1779661016949152,
+      "grad_norm": 8.556358337402344,
+      "learning_rate": 7.6042372881355944e-06,
+      "loss": 0.2935,
+      "step": 39000
+    },
+    {
+      "epoch": 3.1861147327249024,
+      "grad_norm": 9.609768867492676,
+      "learning_rate": 7.5952737940026075e-06,
+      "loss": 0.2804,
+      "step": 39100
+    },
+    {
+      "epoch": 3.194263363754889,
+      "grad_norm": 7.0451507568359375,
+      "learning_rate": 7.5863102998696215e-06,
+      "loss": 0.315,
+      "step": 39200
+    },
+    {
+      "epoch": 3.2024119947848764,
+      "grad_norm": 13.499476432800293,
+      "learning_rate": 7.577346805736636e-06,
+      "loss": 0.2866,
+      "step": 39300
+    },
+    {
+      "epoch": 3.210560625814863,
+      "grad_norm": 12.766288757324219,
+      "learning_rate": 7.568383311603649e-06,
+      "loss": 0.2743,
+      "step": 39400
+    },
+    {
+      "epoch": 3.21870925684485,
+      "grad_norm": 5.220301151275635,
+      "learning_rate": 7.559419817470664e-06,
+      "loss": 0.2879,
+      "step": 39500
+    },
+    {
+      "epoch": 3.226857887874837,
+      "grad_norm": 5.3805060386657715,
+      "learning_rate": 7.550456323337679e-06,
+      "loss": 0.2884,
+      "step": 39600
+    },
+    {
+      "epoch": 3.235006518904824,
+      "grad_norm": 5.386475086212158,
+      "learning_rate": 7.541492829204694e-06,
+      "loss": 0.2806,
+      "step": 39700
+    },
+    {
+      "epoch": 3.243155149934811,
+      "grad_norm": 9.031757354736328,
+      "learning_rate": 7.532529335071708e-06,
+      "loss": 0.286,
+      "step": 39800
+    },
+    {
+      "epoch": 3.2513037809647978,
+      "grad_norm": 6.262955665588379,
+      "learning_rate": 7.523565840938723e-06,
+      "loss": 0.2869,
+      "step": 39900
+    },
+    {
+      "epoch": 3.259452411994785,
+      "grad_norm": 8.278800964355469,
+      "learning_rate": 7.514602346805736e-06,
+      "loss": 0.3113,
+      "step": 40000
+    },
+    {
+      "epoch": 3.2676010430247717,
+      "grad_norm": 6.366271495819092,
+      "learning_rate": 7.5056388526727505e-06,
+      "loss": 0.2828,
+      "step": 40100
+    },
+    {
+      "epoch": 3.275749674054759,
+      "grad_norm": 9.194992065429688,
+      "learning_rate": 7.496675358539765e-06,
+      "loss": 0.2878,
+      "step": 40200
+    },
+    {
+      "epoch": 3.2838983050847457,
+      "grad_norm": 10.264578819274902,
+      "learning_rate": 7.487711864406779e-06,
+      "loss": 0.3111,
+      "step": 40300
+    },
+    {
+      "epoch": 3.292046936114733,
+      "grad_norm": 10.604935646057129,
+      "learning_rate": 7.478748370273795e-06,
+      "loss": 0.2721,
+      "step": 40400
+    },
+    {
+      "epoch": 3.3001955671447196,
+      "grad_norm": 7.705831050872803,
+      "learning_rate": 7.469784876140808e-06,
+      "loss": 0.3051,
+      "step": 40500
+    },
+    {
+      "epoch": 3.308344198174707,
+      "grad_norm": 10.131879806518555,
+      "learning_rate": 7.460821382007822e-06,
+      "loss": 0.2912,
+      "step": 40600
+    },
+    {
+      "epoch": 3.3164928292046936,
+      "grad_norm": 7.966042995452881,
+      "learning_rate": 7.451857887874837e-06,
+      "loss": 0.3067,
+      "step": 40700
+    },
+    {
+      "epoch": 3.3246414602346808,
+      "grad_norm": 7.855635643005371,
+      "learning_rate": 7.442894393741852e-06,
+      "loss": 0.2991,
+      "step": 40800
+    },
+    {
+      "epoch": 3.3327900912646675,
+      "grad_norm": 6.461732864379883,
+      "learning_rate": 7.4339308996088655e-06,
+      "loss": 0.2724,
+      "step": 40900
+    },
+    {
+      "epoch": 3.3409387222946547,
+      "grad_norm": 7.149626731872559,
+      "learning_rate": 7.4249674054758795e-06,
+      "loss": 0.3029,
+      "step": 41000
+    },
+    {
+      "epoch": 3.3490873533246415,
+      "grad_norm": 8.23928451538086,
+      "learning_rate": 7.416003911342895e-06,
+      "loss": 0.2957,
+      "step": 41100
+    },
+    {
+      "epoch": 3.3572359843546282,
+      "grad_norm": 7.156587600708008,
+      "learning_rate": 7.407040417209908e-06,
+      "loss": 0.3053,
+      "step": 41200
+    },
+    {
+      "epoch": 3.3653846153846154,
+      "grad_norm": 9.947956085205078,
+      "learning_rate": 7.398076923076923e-06,
+      "loss": 0.308,
+      "step": 41300
+    },
+    {
+      "epoch": 3.373533246414602,
+      "grad_norm": 11.746365547180176,
+      "learning_rate": 7.389113428943936e-06,
+      "loss": 0.2818,
+      "step": 41400
+    },
+    {
+      "epoch": 3.3816818774445894,
+      "grad_norm": 11.161590576171875,
+      "learning_rate": 7.380149934810952e-06,
+      "loss": 0.2969,
+      "step": 41500
+    },
+    {
+      "epoch": 3.389830508474576,
+      "grad_norm": 7.389200687408447,
+      "learning_rate": 7.371186440677966e-06,
+      "loss": 0.2892,
+      "step": 41600
+    },
+    {
+      "epoch": 3.3979791395045633,
+      "grad_norm": 8.784459114074707,
+      "learning_rate": 7.36222294654498e-06,
+      "loss": 0.2976,
+      "step": 41700
+    },
+    {
+      "epoch": 3.40612777053455,
+      "grad_norm": 7.9105072021484375,
+      "learning_rate": 7.3532594524119945e-06,
+      "loss": 0.2904,
+      "step": 41800
+    },
+    {
+      "epoch": 3.4142764015645373,
+      "grad_norm": 4.8031229972839355,
+      "learning_rate": 7.3442959582790085e-06,
+      "loss": 0.2928,
+      "step": 41900
+    },
+    {
+      "epoch": 3.422425032594524,
+      "grad_norm": 7.598479747772217,
+      "learning_rate": 7.335332464146024e-06,
+      "loss": 0.315,
+      "step": 42000
+    },
+    {
+      "epoch": 3.430573663624511,
+      "grad_norm": 4.397596836090088,
+      "learning_rate": 7.326368970013037e-06,
+      "loss": 0.278,
+      "step": 42100
+    },
+    {
+      "epoch": 3.438722294654498,
+      "grad_norm": 5.5082316398620605,
+      "learning_rate": 7.317405475880052e-06,
+      "loss": 0.3155,
+      "step": 42200
+    },
+    {
+      "epoch": 3.446870925684485,
+      "grad_norm": 1.8329218626022339,
+      "learning_rate": 7.308441981747066e-06,
+      "loss": 0.2834,
+      "step": 42300
+    },
+    {
+      "epoch": 3.455019556714472,
+      "grad_norm": 5.782153129577637,
+      "learning_rate": 7.299478487614081e-06,
+      "loss": 0.2999,
+      "step": 42400
+    },
+    {
+      "epoch": 3.463168187744459,
+      "grad_norm": 6.500859260559082,
+      "learning_rate": 7.290514993481095e-06,
+      "loss": 0.3001,
+      "step": 42500
+    },
+    {
+      "epoch": 3.471316818774446,
+      "grad_norm": 5.2227396965026855,
+      "learning_rate": 7.28155149934811e-06,
+      "loss": 0.3155,
+      "step": 42600
+    },
+    {
+      "epoch": 3.479465449804433,
+      "grad_norm": 7.104318618774414,
+      "learning_rate": 7.272588005215124e-06,
+      "loss": 0.3103,
+      "step": 42700
+    },
+    {
+      "epoch": 3.48761408083442,
+      "grad_norm": 8.62105941772461,
+      "learning_rate": 7.263624511082138e-06,
+      "loss": 0.2892,
+      "step": 42800
+    },
+    {
+      "epoch": 3.4957627118644066,
+      "grad_norm": 11.45067310333252,
+      "learning_rate": 7.254661016949152e-06,
+      "loss": 0.2889,
+      "step": 42900
+    },
+    {
+      "epoch": 3.5039113428943938,
+      "grad_norm": 11.803699493408203,
+      "learning_rate": 7.245697522816166e-06,
+      "loss": 0.2755,
+      "step": 43000
+    },
+    {
+      "epoch": 3.512059973924381,
+      "grad_norm": 6.8339762687683105,
+      "learning_rate": 7.236734028683181e-06,
+      "loss": 0.3083,
+      "step": 43100
+    },
+    {
+      "epoch": 3.5202086049543677,
+      "grad_norm": 5.603298187255859,
+      "learning_rate": 7.227770534550196e-06,
+      "loss": 0.3034,
+      "step": 43200
+    },
+    {
+      "epoch": 3.5283572359843545,
+      "grad_norm": 8.332951545715332,
+      "learning_rate": 7.21880704041721e-06,
+      "loss": 0.3187,
+      "step": 43300
+    },
+    {
+      "epoch": 3.5365058670143417,
+      "grad_norm": 10.166672706604004,
+      "learning_rate": 7.209843546284225e-06,
+      "loss": 0.2966,
+      "step": 43400
+    },
+    {
+      "epoch": 3.5446544980443284,
+      "grad_norm": 8.069250106811523,
+      "learning_rate": 7.200880052151239e-06,
+      "loss": 0.3033,
+      "step": 43500
+    },
+    {
+      "epoch": 3.5528031290743156,
+      "grad_norm": 6.117469787597656,
+      "learning_rate": 7.191916558018253e-06,
+      "loss": 0.3169,
+      "step": 43600
+    },
+    {
+      "epoch": 3.5609517601043024,
+      "grad_norm": 9.44518756866455,
+      "learning_rate": 7.1829530638852665e-06,
+      "loss": 0.3097,
+      "step": 43700
+    },
+    {
+      "epoch": 3.5691003911342896,
+      "grad_norm": 12.208475112915039,
+      "learning_rate": 7.173989569752282e-06,
+      "loss": 0.2995,
+      "step": 43800
+    },
+    {
+      "epoch": 3.5772490221642763,
+      "grad_norm": 3.2542121410369873,
+      "learning_rate": 7.165026075619296e-06,
+      "loss": 0.2853,
+      "step": 43900
+    },
+    {
+      "epoch": 3.5853976531942635,
+      "grad_norm": 5.231081962585449,
+      "learning_rate": 7.156062581486311e-06,
+      "loss": 0.2811,
+      "step": 44000
+    },
+    {
+      "epoch": 3.5935462842242503,
+      "grad_norm": 8.1747407913208,
+      "learning_rate": 7.147099087353325e-06,
+      "loss": 0.2902,
+      "step": 44100
+    },
+    {
+      "epoch": 3.601694915254237,
+      "grad_norm": 5.907063961029053,
+      "learning_rate": 7.138135593220339e-06,
+      "loss": 0.2955,
+      "step": 44200
+    },
+    {
+      "epoch": 3.609843546284224,
+      "grad_norm": 8.774656295776367,
+      "learning_rate": 7.129172099087354e-06,
+      "loss": 0.2976,
+      "step": 44300
+    },
+    {
+      "epoch": 3.6179921773142114,
+      "grad_norm": 6.138618469238281,
+      "learning_rate": 7.120208604954368e-06,
+      "loss": 0.3101,
+      "step": 44400
+    },
+    {
+      "epoch": 3.626140808344198,
+      "grad_norm": 3.3446667194366455,
+      "learning_rate": 7.111245110821382e-06,
+      "loss": 0.2896,
+      "step": 44500
+    },
+    {
+      "epoch": 3.634289439374185,
+      "grad_norm": 4.759154319763184,
+      "learning_rate": 7.1022816166883955e-06,
+      "loss": 0.2952,
+      "step": 44600
+    },
+    {
+      "epoch": 3.642438070404172,
+      "grad_norm": 5.970653533935547,
+      "learning_rate": 7.093318122555411e-06,
+      "loss": 0.2943,
+      "step": 44700
+    },
+    {
+      "epoch": 3.6505867014341593,
+      "grad_norm": 7.213112831115723,
+      "learning_rate": 7.084354628422425e-06,
+      "loss": 0.304,
+      "step": 44800
+    },
+    {
+      "epoch": 3.658735332464146,
+      "grad_norm": 6.11391019821167,
+      "learning_rate": 7.075391134289439e-06,
+      "loss": 0.2935,
+      "step": 44900
+    },
+    {
+      "epoch": 3.666883963494133,
+      "grad_norm": 6.258280277252197,
+      "learning_rate": 7.066427640156454e-06,
+      "loss": 0.2933,
+      "step": 45000
+    },
+    {
+      "epoch": 3.67503259452412,
+      "grad_norm": 11.640277862548828,
+      "learning_rate": 7.057464146023469e-06,
+      "loss": 0.318,
+      "step": 45100
+    },
+    {
+      "epoch": 3.6831812255541068,
+      "grad_norm": 6.384451866149902,
+      "learning_rate": 7.048500651890483e-06,
+      "loss": 0.3106,
+      "step": 45200
+    },
+    {
+      "epoch": 3.691329856584094,
+      "grad_norm": 6.5114054679870605,
+      "learning_rate": 7.039537157757496e-06,
+      "loss": 0.3065,
+      "step": 45300
+    },
+    {
+      "epoch": 3.6994784876140807,
+      "grad_norm": 7.666008949279785,
+      "learning_rate": 7.030573663624511e-06,
+      "loss": 0.2934,
+      "step": 45400
+    },
+    {
+      "epoch": 3.707627118644068,
+      "grad_norm": 8.854146003723145,
+      "learning_rate": 7.021610169491525e-06,
+      "loss": 0.2986,
+      "step": 45500
+    },
+    {
+      "epoch": 3.7157757496740547,
+      "grad_norm": 8.994945526123047,
+      "learning_rate": 7.01264667535854e-06,
+      "loss": 0.2904,
+      "step": 45600
+    },
+    {
+      "epoch": 3.723924380704042,
+      "grad_norm": 6.301919460296631,
+      "learning_rate": 7.003683181225555e-06,
+      "loss": 0.3042,
+      "step": 45700
+    },
+    {
+      "epoch": 3.7320730117340286,
+      "grad_norm": 6.444594860076904,
+      "learning_rate": 6.994719687092568e-06,
+      "loss": 0.2934,
+      "step": 45800
+    },
+    {
+      "epoch": 3.740221642764016,
+      "grad_norm": 7.342471122741699,
+      "learning_rate": 6.985756192959584e-06,
+      "loss": 0.2573,
+      "step": 45900
+    },
+    {
+      "epoch": 3.7483702737940026,
+      "grad_norm": 10.449690818786621,
+      "learning_rate": 6.976792698826597e-06,
+      "loss": 0.2992,
+      "step": 46000
+    },
+    {
+      "epoch": 3.7565189048239898,
+      "grad_norm": 6.386083602905273,
+      "learning_rate": 6.967829204693612e-06,
+      "loss": 0.3197,
+      "step": 46100
+    },
+    {
+      "epoch": 3.7646675358539765,
+      "grad_norm": 5.020893573760986,
+      "learning_rate": 6.958865710560625e-06,
+      "loss": 0.2898,
+      "step": 46200
+    },
+    {
+      "epoch": 3.7728161668839633,
+      "grad_norm": 13.005038261413574,
+      "learning_rate": 6.949902216427641e-06,
+      "loss": 0.2775,
+      "step": 46300
+    },
+    {
+      "epoch": 3.7809647979139505,
+      "grad_norm": 9.009956359863281,
+      "learning_rate": 6.940938722294654e-06,
+      "loss": 0.3111,
+      "step": 46400
+    },
+    {
+      "epoch": 3.7891134289439377,
+      "grad_norm": 11.799562454223633,
+      "learning_rate": 6.931975228161669e-06,
+      "loss": 0.3086,
+      "step": 46500
+    },
+    {
+      "epoch": 3.7972620599739244,
+      "grad_norm": 5.661113262176514,
+      "learning_rate": 6.923011734028683e-06,
+      "loss": 0.2922,
+      "step": 46600
+    },
+    {
+      "epoch": 3.805410691003911,
+      "grad_norm": 7.668259143829346,
+      "learning_rate": 6.914048239895698e-06,
+      "loss": 0.3089,
+      "step": 46700
+    },
+    {
+      "epoch": 3.8135593220338984,
+      "grad_norm": 5.768048286437988,
+      "learning_rate": 6.905084745762713e-06,
+      "loss": 0.316,
+      "step": 46800
+    },
+    {
+      "epoch": 3.821707953063885,
+      "grad_norm": 10.127528190612793,
+      "learning_rate": 6.896121251629726e-06,
+      "loss": 0.3055,
+      "step": 46900
+    },
+    {
+      "epoch": 3.8298565840938723,
+      "grad_norm": 8.772488594055176,
+      "learning_rate": 6.887157757496741e-06,
+      "loss": 0.3021,
+      "step": 47000
+    },
+    {
+      "epoch": 3.838005215123859,
+      "grad_norm": 5.210601806640625,
+      "learning_rate": 6.8781942633637554e-06,
+      "loss": 0.2871,
+      "step": 47100
+    },
+    {
+      "epoch": 3.8461538461538463,
+      "grad_norm": 6.755288124084473,
+      "learning_rate": 6.869230769230769e-06,
+      "loss": 0.3067,
+      "step": 47200
+    },
+    {
+      "epoch": 3.854302477183833,
+      "grad_norm": 5.8384833335876465,
+      "learning_rate": 6.860267275097784e-06,
+      "loss": 0.3026,
+      "step": 47300
+    },
+    {
+      "epoch": 3.86245110821382,
+      "grad_norm": 10.123713493347168,
+      "learning_rate": 6.851303780964797e-06,
+      "loss": 0.2925,
+      "step": 47400
+    },
+    {
+      "epoch": 3.870599739243807,
+      "grad_norm": 6.272671699523926,
+      "learning_rate": 6.842340286831814e-06,
+      "loss": 0.3202,
+      "step": 47500
+    },
+    {
+      "epoch": 3.878748370273794,
+      "grad_norm": 6.816521167755127,
+      "learning_rate": 6.833376792698826e-06,
+      "loss": 0.3031,
+      "step": 47600
+    },
+    {
+      "epoch": 3.886897001303781,
+      "grad_norm": 6.4613752365112305,
+      "learning_rate": 6.824413298565842e-06,
+      "loss": 0.3067,
+      "step": 47700
+    },
+    {
+      "epoch": 3.895045632333768,
+      "grad_norm": 10.189191818237305,
+      "learning_rate": 6.815449804432854e-06,
+      "loss": 0.2946,
+      "step": 47800
+    },
+    {
+      "epoch": 3.903194263363755,
+      "grad_norm": 12.38447093963623,
+      "learning_rate": 6.8064863102998705e-06,
+      "loss": 0.3023,
+      "step": 47900
+    },
+    {
+      "epoch": 3.9113428943937416,
+      "grad_norm": 9.731588363647461,
+      "learning_rate": 6.797522816166885e-06,
+      "loss": 0.2932,
+      "step": 48000
+    },
+    {
+      "epoch": 3.919491525423729,
+      "grad_norm": 9.75956916809082,
+      "learning_rate": 6.788559322033898e-06,
+      "loss": 0.3004,
+      "step": 48100
+    },
+    {
+      "epoch": 3.927640156453716,
+      "grad_norm": 9.017115592956543,
+      "learning_rate": 6.779595827900912e-06,
+      "loss": 0.3083,
+      "step": 48200
+    },
+    {
+      "epoch": 3.9357887874837028,
+      "grad_norm": 4.012237071990967,
+      "learning_rate": 6.7706323337679255e-06,
+      "loss": 0.2975,
+      "step": 48300
+    },
+    {
+      "epoch": 3.9439374185136895,
+      "grad_norm": 9.703700065612793,
+      "learning_rate": 6.761668839634942e-06,
+      "loss": 0.3007,
+      "step": 48400
+    },
+    {
+      "epoch": 3.9520860495436767,
+      "grad_norm": 5.018248558044434,
+      "learning_rate": 6.752705345501955e-06,
+      "loss": 0.2835,
+      "step": 48500
+    },
+    {
+      "epoch": 3.960234680573664,
+      "grad_norm": 8.908330917358398,
+      "learning_rate": 6.7437418513689716e-06,
+      "loss": 0.2885,
+      "step": 48600
+    },
+    {
+      "epoch": 3.9683833116036507,
+      "grad_norm": 7.982217788696289,
+      "learning_rate": 6.734778357235984e-06,
+      "loss": 0.2939,
+      "step": 48700
+    },
+    {
+      "epoch": 3.9765319426336374,
+      "grad_norm": 6.468784809112549,
+      "learning_rate": 6.725814863102999e-06,
+      "loss": 0.2973,
+      "step": 48800
+    },
+    {
+      "epoch": 3.9846805736636246,
+      "grad_norm": 11.741927146911621,
+      "learning_rate": 6.716851368970012e-06,
+      "loss": 0.3115,
+      "step": 48900
+    },
+    {
+      "epoch": 3.9928292046936114,
+      "grad_norm": 6.260389804840088,
+      "learning_rate": 6.7078878748370266e-06,
+      "loss": 0.3033,
+      "step": 49000
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.8365461847389558,
+      "eval_loss": 0.49297481775283813,
+      "eval_runtime": 5.5068,
+      "eval_samples_per_second": 452.166,
+      "eval_steps_per_second": 56.657,
+      "step": 49088
+    },
+    {
+      "epoch": 4.0009778357235986,
+      "grad_norm": 6.392615795135498,
+      "learning_rate": 6.698924380704041e-06,
+      "loss": 0.29,
+      "step": 49100
+    },
+    {
+      "epoch": 4.009126466753585,
+      "grad_norm": 8.285822868347168,
+      "learning_rate": 6.689960886571055e-06,
+      "loss": 0.2284,
+      "step": 49200
+    },
+    {
+      "epoch": 4.017275097783572,
+      "grad_norm": 8.578091621398926,
+      "learning_rate": 6.680997392438071e-06,
+      "loss": 0.2389,
+      "step": 49300
+    },
+    {
+      "epoch": 4.02542372881356,
+      "grad_norm": 7.808691024780273,
+      "learning_rate": 6.672033898305083e-06,
+      "loss": 0.237,
+      "step": 49400
+    },
+    {
+      "epoch": 4.0335723598435465,
+      "grad_norm": 11.781246185302734,
+      "learning_rate": 6.6630704041721e-06,
+      "loss": 0.2207,
+      "step": 49500
+    },
+    {
+      "epoch": 4.041720990873533,
+      "grad_norm": 11.581729888916016,
+      "learning_rate": 6.654106910039113e-06,
+      "loss": 0.2456,
+      "step": 49600
+    },
+    {
+      "epoch": 4.04986962190352,
+      "grad_norm": 11.857730865478516,
+      "learning_rate": 6.645143415906128e-06,
+      "loss": 0.2381,
+      "step": 49700
+    },
+    {
+      "epoch": 4.058018252933508,
+      "grad_norm": 5.7172722816467285,
+      "learning_rate": 6.636179921773142e-06,
+      "loss": 0.2306,
+      "step": 49800
+    },
+    {
+      "epoch": 4.066166883963494,
+      "grad_norm": 9.22181224822998,
+      "learning_rate": 6.627216427640156e-06,
+      "loss": 0.2418,
+      "step": 49900
+    },
+    {
+      "epoch": 4.074315514993481,
+      "grad_norm": 9.777718544006348,
+      "learning_rate": 6.618252933507171e-06,
+      "loss": 0.2391,
+      "step": 50000
+    },
+    {
+      "epoch": 4.082464146023468,
+      "grad_norm": 11.828337669372559,
+      "learning_rate": 6.609289439374184e-06,
+      "loss": 0.234,
+      "step": 50100
+    },
+    {
+      "epoch": 4.090612777053455,
+      "grad_norm": 8.534415245056152,
+      "learning_rate": 6.600325945241201e-06,
+      "loss": 0.2199,
+      "step": 50200
+    },
+    {
+      "epoch": 4.098761408083442,
+      "grad_norm": 8.329660415649414,
+      "learning_rate": 6.591362451108213e-06,
+      "loss": 0.261,
+      "step": 50300
+    },
+    {
+      "epoch": 4.106910039113429,
+      "grad_norm": 6.748033046722412,
+      "learning_rate": 6.582398956975228e-06,
+      "loss": 0.2412,
+      "step": 50400
+    },
+    {
+      "epoch": 4.115058670143416,
+      "grad_norm": 3.2746846675872803,
+      "learning_rate": 6.573435462842243e-06,
+      "loss": 0.2509,
+      "step": 50500
+    },
+    {
+      "epoch": 4.1232073011734025,
+      "grad_norm": 5.985006809234619,
+      "learning_rate": 6.564471968709256e-06,
+      "loss": 0.2412,
+      "step": 50600
+    },
+    {
+      "epoch": 4.13135593220339,
+      "grad_norm": 10.339166641235352,
+      "learning_rate": 6.555508474576271e-06,
+      "loss": 0.2439,
+      "step": 50700
+    },
+    {
+      "epoch": 4.139504563233377,
+      "grad_norm": 3.1601479053497314,
+      "learning_rate": 6.546544980443285e-06,
+      "loss": 0.2332,
+      "step": 50800
+    },
+    {
+      "epoch": 4.147653194263364,
+      "grad_norm": 7.7636308670043945,
+      "learning_rate": 6.5375814863103e-06,
+      "loss": 0.2428,
+      "step": 50900
+    },
+    {
+      "epoch": 4.15580182529335,
+      "grad_norm": 5.945723056793213,
+      "learning_rate": 6.528617992177313e-06,
+      "loss": 0.235,
+      "step": 51000
+    },
+    {
+      "epoch": 4.163950456323338,
+      "grad_norm": 6.053178787231445,
+      "learning_rate": 6.519654498044329e-06,
+      "loss": 0.258,
+      "step": 51100
+    },
+    {
+      "epoch": 4.172099087353325,
+      "grad_norm": 7.7220845222473145,
+      "learning_rate": 6.510691003911342e-06,
+      "loss": 0.2202,
+      "step": 51200
+    },
+    {
+      "epoch": 4.1802477183833116,
+      "grad_norm": 10.34273910522461,
+      "learning_rate": 6.501727509778357e-06,
+      "loss": 0.2308,
+      "step": 51300
+    },
+    {
+      "epoch": 4.188396349413298,
+      "grad_norm": 11.372400283813477,
+      "learning_rate": 6.492764015645372e-06,
+      "loss": 0.2403,
+      "step": 51400
+    },
+    {
+      "epoch": 4.196544980443286,
+      "grad_norm": 5.102409839630127,
+      "learning_rate": 6.483800521512386e-06,
+      "loss": 0.2423,
+      "step": 51500
+    },
+    {
+      "epoch": 4.204693611473273,
+      "grad_norm": 5.650859832763672,
+      "learning_rate": 6.474837027379401e-06,
+      "loss": 0.2239,
+      "step": 51600
+    },
+    {
+      "epoch": 4.2128422425032594,
+      "grad_norm": 10.150152206420898,
+      "learning_rate": 6.4658735332464136e-06,
+      "loss": 0.2458,
+      "step": 51700
+    },
+    {
+      "epoch": 4.220990873533246,
+      "grad_norm": 13.111183166503906,
+      "learning_rate": 6.456910039113428e-06,
+      "loss": 0.2542,
+      "step": 51800
+    },
+    {
+      "epoch": 4.229139504563233,
+      "grad_norm": 10.616249084472656,
+      "learning_rate": 6.447946544980442e-06,
+      "loss": 0.2643,
+      "step": 51900
+    },
+    {
+      "epoch": 4.237288135593221,
+      "grad_norm": 8.26252269744873,
+      "learning_rate": 6.438983050847458e-06,
+      "loss": 0.2476,
+      "step": 52000
+    },
+    {
+      "epoch": 4.245436766623207,
+      "grad_norm": 3.8655807971954346,
+      "learning_rate": 6.430019556714472e-06,
+      "loss": 0.2421,
+      "step": 52100
+    },
+    {
+      "epoch": 4.253585397653194,
+      "grad_norm": 12.139904975891113,
+      "learning_rate": 6.421056062581486e-06,
+      "loss": 0.2498,
+      "step": 52200
+    },
+    {
+      "epoch": 4.261734028683181,
+      "grad_norm": 6.828706741333008,
+      "learning_rate": 6.4120925684485e-06,
+      "loss": 0.2607,
+      "step": 52300
+    },
+    {
+      "epoch": 4.2698826597131685,
+      "grad_norm": 6.083545684814453,
+      "learning_rate": 6.403129074315515e-06,
+      "loss": 0.2371,
+      "step": 52400
+    },
+    {
+      "epoch": 4.278031290743155,
+      "grad_norm": 11.325626373291016,
+      "learning_rate": 6.3941655801825294e-06,
+      "loss": 0.2487,
+      "step": 52500
+    },
+    {
+      "epoch": 4.286179921773142,
+      "grad_norm": 5.367312431335449,
+      "learning_rate": 6.385202086049543e-06,
+      "loss": 0.2254,
+      "step": 52600
+    },
+    {
+      "epoch": 4.294328552803129,
+      "grad_norm": 14.246549606323242,
+      "learning_rate": 6.376238591916558e-06,
+      "loss": 0.2237,
+      "step": 52700
+    },
+    {
+      "epoch": 4.302477183833116,
+      "grad_norm": 11.854809761047363,
+      "learning_rate": 6.367275097783572e-06,
+      "loss": 0.2446,
+      "step": 52800
+    },
+    {
+      "epoch": 4.310625814863103,
+      "grad_norm": 10.6870698928833,
+      "learning_rate": 6.358311603650586e-06,
+      "loss": 0.2219,
+      "step": 52900
+    },
+    {
+      "epoch": 4.31877444589309,
+      "grad_norm": 11.290742874145508,
+      "learning_rate": 6.349348109517601e-06,
+      "loss": 0.2757,
+      "step": 53000
+    },
+    {
+      "epoch": 4.326923076923077,
+      "grad_norm": 5.060425758361816,
+      "learning_rate": 6.340384615384615e-06,
+      "loss": 0.2279,
+      "step": 53100
+    },
+    {
+      "epoch": 4.335071707953064,
+      "grad_norm": 11.129130363464355,
+      "learning_rate": 6.3314211212516305e-06,
+      "loss": 0.2352,
+      "step": 53200
+    },
+    {
+      "epoch": 4.343220338983051,
+      "grad_norm": 9.912896156311035,
+      "learning_rate": 6.322457627118644e-06,
+      "loss": 0.2378,
+      "step": 53300
+    },
+    {
+      "epoch": 4.351368970013038,
+      "grad_norm": 6.451144218444824,
+      "learning_rate": 6.3134941329856584e-06,
+      "loss": 0.229,
+      "step": 53400
+    },
+    {
+      "epoch": 4.3595176010430245,
+      "grad_norm": 8.81298828125,
+      "learning_rate": 6.3045306388526716e-06,
+      "loss": 0.248,
+      "step": 53500
+    },
+    {
+      "epoch": 4.367666232073011,
+      "grad_norm": 10.046360969543457,
+      "learning_rate": 6.295567144719687e-06,
+      "loss": 0.2236,
+      "step": 53600
+    },
+    {
+      "epoch": 4.375814863102999,
+      "grad_norm": 7.006089687347412,
+      "learning_rate": 6.286603650586701e-06,
+      "loss": 0.2522,
+      "step": 53700
+    },
+    {
+      "epoch": 4.383963494132986,
+      "grad_norm": 8.601024627685547,
+      "learning_rate": 6.277640156453716e-06,
+      "loss": 0.2595,
+      "step": 53800
+    },
+    {
+      "epoch": 4.3921121251629724,
+      "grad_norm": 7.744350433349609,
+      "learning_rate": 6.26867666232073e-06,
+      "loss": 0.2433,
+      "step": 53900
+    },
+    {
+      "epoch": 4.400260756192959,
+      "grad_norm": 9.870165824890137,
+      "learning_rate": 6.259713168187745e-06,
+      "loss": 0.256,
+      "step": 54000
+    },
+    {
+      "epoch": 4.408409387222947,
+      "grad_norm": 16.055753707885742,
+      "learning_rate": 6.250749674054759e-06,
+      "loss": 0.2626,
+      "step": 54100
+    },
+    {
+      "epoch": 4.416558018252934,
+      "grad_norm": 7.4111104011535645,
+      "learning_rate": 6.241786179921773e-06,
+      "loss": 0.2403,
+      "step": 54200
+    },
+    {
+      "epoch": 4.42470664928292,
+      "grad_norm": 9.383624076843262,
+      "learning_rate": 6.2328226857887874e-06,
+      "loss": 0.2528,
+      "step": 54300
+    },
+    {
+      "epoch": 4.432855280312907,
+      "grad_norm": 12.037814140319824,
+      "learning_rate": 6.223859191655802e-06,
+      "loss": 0.2618,
+      "step": 54400
+    },
+    {
+      "epoch": 4.441003911342895,
+      "grad_norm": 11.180290222167969,
+      "learning_rate": 6.214895697522816e-06,
+      "loss": 0.2619,
+      "step": 54500
+    },
+    {
+      "epoch": 4.4491525423728815,
+      "grad_norm": 16.229679107666016,
+      "learning_rate": 6.205932203389831e-06,
+      "loss": 0.2113,
+      "step": 54600
+    },
+    {
+      "epoch": 4.457301173402868,
+      "grad_norm": 13.413905143737793,
+      "learning_rate": 6.196968709256845e-06,
+      "loss": 0.2294,
+      "step": 54700
+    },
+    {
+      "epoch": 4.465449804432855,
+      "grad_norm": 4.0000739097595215,
+      "learning_rate": 6.18800521512386e-06,
+      "loss": 0.2492,
+      "step": 54800
+    },
+    {
+      "epoch": 4.473598435462843,
+      "grad_norm": 6.232034206390381,
+      "learning_rate": 6.179041720990873e-06,
+      "loss": 0.2257,
+      "step": 54900
+    },
+    {
+      "epoch": 4.481747066492829,
+      "grad_norm": 10.516459465026855,
+      "learning_rate": 6.1700782268578885e-06,
+      "loss": 0.2411,
+      "step": 55000
+    },
+    {
+      "epoch": 4.489895697522816,
+      "grad_norm": 1.9097963571548462,
+      "learning_rate": 6.1611147327249025e-06,
+      "loss": 0.2397,
+      "step": 55100
+    },
+    {
+      "epoch": 4.498044328552803,
+      "grad_norm": 8.531264305114746,
+      "learning_rate": 6.152151238591917e-06,
+      "loss": 0.2496,
+      "step": 55200
+    },
+    {
+      "epoch": 4.5061929595827905,
+      "grad_norm": 11.344954490661621,
+      "learning_rate": 6.143187744458931e-06,
+      "loss": 0.2376,
+      "step": 55300
+    },
+    {
+      "epoch": 4.514341590612777,
+      "grad_norm": 11.570788383483887,
+      "learning_rate": 6.134224250325945e-06,
+      "loss": 0.238,
+      "step": 55400
+    },
+    {
+      "epoch": 4.522490221642764,
+      "grad_norm": 7.503643035888672,
+      "learning_rate": 6.125260756192959e-06,
+      "loss": 0.2492,
+      "step": 55500
+    },
+    {
+      "epoch": 4.530638852672751,
+      "grad_norm": 10.721003532409668,
+      "learning_rate": 6.116297262059974e-06,
+      "loss": 0.2508,
+      "step": 55600
+    },
+    {
+      "epoch": 4.5387874837027375,
+      "grad_norm": 7.251551151275635,
+      "learning_rate": 6.107333767926989e-06,
+      "loss": 0.2661,
+      "step": 55700
+    },
+    {
+      "epoch": 4.546936114732725,
+      "grad_norm": 5.292377948760986,
+      "learning_rate": 6.098370273794002e-06,
+      "loss": 0.2253,
+      "step": 55800
+    },
+    {
+      "epoch": 4.555084745762712,
+      "grad_norm": 8.165968894958496,
+      "learning_rate": 6.0894067796610175e-06,
+      "loss": 0.2401,
+      "step": 55900
+    },
+    {
+      "epoch": 4.563233376792699,
+      "grad_norm": 5.5696001052856445,
+      "learning_rate": 6.0804432855280315e-06,
+      "loss": 0.2565,
+      "step": 56000
+    },
+    {
+      "epoch": 4.5713820078226854,
+      "grad_norm": 4.321511745452881,
+      "learning_rate": 6.0714797913950455e-06,
+      "loss": 0.2468,
+      "step": 56100
+    },
+    {
+      "epoch": 4.579530638852673,
+      "grad_norm": 7.747223854064941,
+      "learning_rate": 6.06251629726206e-06,
+      "loss": 0.2677,
+      "step": 56200
+    },
+    {
+      "epoch": 4.58767926988266,
+      "grad_norm": 6.845622539520264,
+      "learning_rate": 6.053552803129075e-06,
+      "loss": 0.2399,
+      "step": 56300
+    },
+    {
+      "epoch": 4.595827900912647,
+      "grad_norm": 7.972433567047119,
+      "learning_rate": 6.044589308996088e-06,
+      "loss": 0.2538,
+      "step": 56400
+    },
+    {
+      "epoch": 4.603976531942633,
+      "grad_norm": 7.560089111328125,
+      "learning_rate": 6.035625814863102e-06,
+      "loss": 0.2417,
+      "step": 56500
+    },
+    {
+      "epoch": 4.612125162972621,
+      "grad_norm": 12.906600952148438,
+      "learning_rate": 6.026662320730118e-06,
+      "loss": 0.2534,
+      "step": 56600
+    },
+    {
+      "epoch": 4.620273794002608,
+      "grad_norm": 6.646966934204102,
+      "learning_rate": 6.01769882659713e-06,
+      "loss": 0.2229,
+      "step": 56700
+    },
+    {
+      "epoch": 4.6284224250325945,
+      "grad_norm": 4.299482822418213,
+      "learning_rate": 6.0087353324641465e-06,
+      "loss": 0.2451,
+      "step": 56800
+    },
+    {
+      "epoch": 4.636571056062581,
+      "grad_norm": 2.2282984256744385,
+      "learning_rate": 5.999771838331161e-06,
+      "loss": 0.2441,
+      "step": 56900
+    },
+    {
+      "epoch": 4.644719687092568,
+      "grad_norm": 6.4274210929870605,
+      "learning_rate": 5.9908083441981745e-06,
+      "loss": 0.248,
+      "step": 57000
+    },
+    {
+      "epoch": 4.652868318122556,
+      "grad_norm": 11.95504379272461,
+      "learning_rate": 5.981844850065189e-06,
+      "loss": 0.2479,
+      "step": 57100
+    },
+    {
+      "epoch": 4.661016949152542,
+      "grad_norm": 5.159088134765625,
+      "learning_rate": 5.972881355932203e-06,
+      "loss": 0.2408,
+      "step": 57200
+    },
+    {
+      "epoch": 4.669165580182529,
+      "grad_norm": 13.74659252166748,
+      "learning_rate": 5.963917861799218e-06,
+      "loss": 0.2596,
+      "step": 57300
+    },
+    {
+      "epoch": 4.677314211212517,
+      "grad_norm": 7.052792072296143,
+      "learning_rate": 5.954954367666231e-06,
+      "loss": 0.251,
+      "step": 57400
+    },
+    {
+      "epoch": 4.6854628422425035,
+      "grad_norm": 7.998805999755859,
+      "learning_rate": 5.945990873533248e-06,
+      "loss": 0.2517,
+      "step": 57500
+    },
+    {
+      "epoch": 4.69361147327249,
+      "grad_norm": 8.009773254394531,
+      "learning_rate": 5.93702737940026e-06,
+      "loss": 0.2549,
+      "step": 57600
+    },
+    {
+      "epoch": 4.701760104302477,
+      "grad_norm": 16.998640060424805,
+      "learning_rate": 5.9280638852672755e-06,
+      "loss": 0.2387,
+      "step": 57700
+    },
+    {
+      "epoch": 4.709908735332464,
+      "grad_norm": 5.554251670837402,
+      "learning_rate": 5.9191003911342895e-06,
+      "loss": 0.2342,
+      "step": 57800
+    },
+    {
+      "epoch": 4.718057366362451,
+      "grad_norm": 11.730195999145508,
+      "learning_rate": 5.910136897001304e-06,
+      "loss": 0.2536,
+      "step": 57900
+    },
+    {
+      "epoch": 4.726205997392438,
+      "grad_norm": 8.52701473236084,
+      "learning_rate": 5.901173402868317e-06,
+      "loss": 0.2552,
+      "step": 58000
+    },
+    {
+      "epoch": 4.734354628422425,
+      "grad_norm": 8.105571746826172,
+      "learning_rate": 5.892209908735332e-06,
+      "loss": 0.2512,
+      "step": 58100
+    },
+    {
+      "epoch": 4.742503259452412,
+      "grad_norm": 8.212213516235352,
+      "learning_rate": 5.883246414602347e-06,
+      "loss": 0.2517,
+      "step": 58200
+    },
+    {
+      "epoch": 4.750651890482399,
+      "grad_norm": 7.248411178588867,
+      "learning_rate": 5.874282920469359e-06,
+      "loss": 0.243,
+      "step": 58300
+    },
+    {
+      "epoch": 4.758800521512386,
+      "grad_norm": 7.534228801727295,
+      "learning_rate": 5.865319426336376e-06,
+      "loss": 0.2345,
+      "step": 58400
+    },
+    {
+      "epoch": 4.766949152542373,
+      "grad_norm": 9.132428169250488,
+      "learning_rate": 5.856355932203391e-06,
+      "loss": 0.2378,
+      "step": 58500
+    },
+    {
+      "epoch": 4.77509778357236,
+      "grad_norm": 10.025369644165039,
+      "learning_rate": 5.847392438070405e-06,
+      "loss": 0.2674,
+      "step": 58600
+    },
+    {
+      "epoch": 4.783246414602347,
+      "grad_norm": 10.384153366088867,
+      "learning_rate": 5.8384289439374185e-06,
+      "loss": 0.2391,
+      "step": 58700
+    },
+    {
+      "epoch": 4.791395045632334,
+      "grad_norm": 15.859262466430664,
+      "learning_rate": 5.8294654498044325e-06,
+      "loss": 0.2417,
+      "step": 58800
+    },
+    {
+      "epoch": 4.799543676662321,
+      "grad_norm": 7.405439376831055,
+      "learning_rate": 5.820501955671448e-06,
+      "loss": 0.2475,
+      "step": 58900
+    },
+    {
+      "epoch": 4.8076923076923075,
+      "grad_norm": 9.894548416137695,
+      "learning_rate": 5.81153846153846e-06,
+      "loss": 0.242,
+      "step": 59000
+    },
+    {
+      "epoch": 4.815840938722294,
+      "grad_norm": 4.984899997711182,
+      "learning_rate": 5.802574967405477e-06,
+      "loss": 0.2367,
+      "step": 59100
+    },
+    {
+      "epoch": 4.823989569752282,
+      "grad_norm": 8.233868598937988,
+      "learning_rate": 5.793611473272489e-06,
+      "loss": 0.2422,
+      "step": 59200
+    },
+    {
+      "epoch": 4.832138200782269,
+      "grad_norm": 3.405486583709717,
+      "learning_rate": 5.784647979139505e-06,
+      "loss": 0.2363,
+      "step": 59300
+    },
+    {
+      "epoch": 4.840286831812255,
+      "grad_norm": 10.75444221496582,
+      "learning_rate": 5.775684485006519e-06,
+      "loss": 0.2387,
+      "step": 59400
+    },
+    {
+      "epoch": 4.848435462842242,
+      "grad_norm": 8.05445671081543,
+      "learning_rate": 5.766720990873532e-06,
+      "loss": 0.2501,
+      "step": 59500
+    },
+    {
+      "epoch": 4.85658409387223,
+      "grad_norm": 9.405356407165527,
+      "learning_rate": 5.757757496740547e-06,
+      "loss": 0.2535,
+      "step": 59600
+    },
+    {
+      "epoch": 4.8647327249022165,
+      "grad_norm": 7.889901638031006,
+      "learning_rate": 5.7487940026075615e-06,
+      "loss": 0.2472,
+      "step": 59700
+    },
+    {
+      "epoch": 4.872881355932203,
+      "grad_norm": 12.93803596496582,
+      "learning_rate": 5.739830508474578e-06,
+      "loss": 0.2484,
+      "step": 59800
+    },
+    {
+      "epoch": 4.88102998696219,
+      "grad_norm": 10.640684127807617,
+      "learning_rate": 5.73086701434159e-06,
+      "loss": 0.257,
+      "step": 59900
+    },
+    {
+      "epoch": 4.889178617992178,
+      "grad_norm": 6.97421932220459,
+      "learning_rate": 5.721903520208605e-06,
+      "loss": 0.2271,
+      "step": 60000
+    },
+    {
+      "epoch": 4.897327249022164,
+      "grad_norm": 8.209015846252441,
+      "learning_rate": 5.71294002607562e-06,
+      "loss": 0.261,
+      "step": 60100
+    },
+    {
+      "epoch": 4.905475880052151,
+      "grad_norm": 10.535201072692871,
+      "learning_rate": 5.703976531942633e-06,
+      "loss": 0.2282,
+      "step": 60200
+    },
+    {
+      "epoch": 4.913624511082138,
+      "grad_norm": 7.5695013999938965,
+      "learning_rate": 5.695013037809648e-06,
+      "loss": 0.2598,
+      "step": 60300
+    },
+    {
+      "epoch": 4.921773142112125,
+      "grad_norm": 6.751789569854736,
+      "learning_rate": 5.686049543676662e-06,
+      "loss": 0.2683,
+      "step": 60400
+    },
+    {
+      "epoch": 4.929921773142112,
+      "grad_norm": 8.825233459472656,
+      "learning_rate": 5.677086049543677e-06,
+      "loss": 0.2468,
+      "step": 60500
+    },
+    {
+      "epoch": 4.938070404172099,
+      "grad_norm": 6.470550060272217,
+      "learning_rate": 5.66812255541069e-06,
+      "loss": 0.2296,
+      "step": 60600
+    },
+    {
+      "epoch": 4.946219035202086,
+      "grad_norm": 6.545186996459961,
+      "learning_rate": 5.659159061277706e-06,
+      "loss": 0.2595,
+      "step": 60700
+    },
+    {
+      "epoch": 4.9543676662320735,
+      "grad_norm": 11.159351348876953,
+      "learning_rate": 5.650195567144719e-06,
+      "loss": 0.2515,
+      "step": 60800
+    },
+    {
+      "epoch": 4.96251629726206,
+      "grad_norm": 7.022935390472412,
+      "learning_rate": 5.641232073011734e-06,
+      "loss": 0.2569,
+      "step": 60900
+    },
+    {
+      "epoch": 4.970664928292047,
+      "grad_norm": 6.251514911651611,
+      "learning_rate": 5.632268578878748e-06,
+      "loss": 0.2317,
+      "step": 61000
+    },
+    {
+      "epoch": 4.978813559322034,
+      "grad_norm": 5.2566657066345215,
+      "learning_rate": 5.623305084745763e-06,
+      "loss": 0.2413,
+      "step": 61100
+    },
+    {
+      "epoch": 4.9869621903520205,
+      "grad_norm": 8.16510009765625,
+      "learning_rate": 5.614341590612776e-06,
+      "loss": 0.2392,
+      "step": 61200
+    },
+    {
+      "epoch": 4.995110821382008,
+      "grad_norm": 4.102538585662842,
+      "learning_rate": 5.605378096479791e-06,
+      "loss": 0.257,
+      "step": 61300
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.8502008032128514,
+      "eval_loss": 0.5025619864463806,
+      "eval_runtime": 5.5086,
+      "eval_samples_per_second": 452.021,
+      "eval_steps_per_second": 56.639,
+      "step": 61360
+    },
+    {
+      "epoch": 5.003259452411995,
+      "grad_norm": 3.8134822845458984,
+      "learning_rate": 5.596414602346807e-06,
+      "loss": 0.2316,
+      "step": 61400
+    },
+    {
+      "epoch": 5.011408083441982,
+      "grad_norm": 6.683708667755127,
+      "learning_rate": 5.5874511082138195e-06,
+      "loss": 0.1937,
+      "step": 61500
+    },
+    {
+      "epoch": 5.019556714471968,
+      "grad_norm": 9.566272735595703,
+      "learning_rate": 5.578487614080835e-06,
+      "loss": 0.2288,
+      "step": 61600
+    },
+    {
+      "epoch": 5.027705345501956,
+      "grad_norm": 10.972317695617676,
+      "learning_rate": 5.569524119947849e-06,
+      "loss": 0.2247,
+      "step": 61700
+    },
+    {
+      "epoch": 5.035853976531943,
+      "grad_norm": 7.630614280700684,
+      "learning_rate": 5.560560625814862e-06,
+      "loss": 0.1842,
+      "step": 61800
+    },
+    {
+      "epoch": 5.0440026075619295,
+      "grad_norm": 9.422782897949219,
+      "learning_rate": 5.551597131681877e-06,
+      "loss": 0.1899,
+      "step": 61900
+    },
+    {
+      "epoch": 5.052151238591916,
+      "grad_norm": 8.5192289352417,
+      "learning_rate": 5.542633637548892e-06,
+      "loss": 0.1951,
+      "step": 62000
+    },
+    {
+      "epoch": 5.060299869621904,
+      "grad_norm": 8.805126190185547,
+      "learning_rate": 5.533670143415907e-06,
+      "loss": 0.2086,
+      "step": 62100
+    },
+    {
+      "epoch": 5.068448500651891,
+      "grad_norm": 20.683149337768555,
+      "learning_rate": 5.52470664928292e-06,
+      "loss": 0.183,
+      "step": 62200
+    },
+    {
+      "epoch": 5.076597131681877,
+      "grad_norm": 4.523620128631592,
+      "learning_rate": 5.515743155149935e-06,
+      "loss": 0.2044,
+      "step": 62300
+    },
+    {
+      "epoch": 5.084745762711864,
+      "grad_norm": 9.480749130249023,
+      "learning_rate": 5.5067796610169485e-06,
+      "loss": 0.2056,
+      "step": 62400
+    },
+    {
+      "epoch": 5.092894393741851,
+      "grad_norm": 5.888172149658203,
+      "learning_rate": 5.497816166883963e-06,
+      "loss": 0.2056,
+      "step": 62500
+    },
+    {
+      "epoch": 5.101043024771839,
+      "grad_norm": 5.114003658294678,
+      "learning_rate": 5.488852672750978e-06,
+      "loss": 0.2073,
+      "step": 62600
+    },
+    {
+      "epoch": 5.109191655801825,
+      "grad_norm": 2.9696881771087646,
+      "learning_rate": 5.479889178617992e-06,
+      "loss": 0.2026,
+      "step": 62700
+    },
+    {
+      "epoch": 5.117340286831812,
+      "grad_norm": 3.7792017459869385,
+      "learning_rate": 5.470925684485006e-06,
+      "loss": 0.2032,
+      "step": 62800
+    },
+    {
+      "epoch": 5.125488917861799,
+      "grad_norm": 8.12082290649414,
+      "learning_rate": 5.46196219035202e-06,
+      "loss": 0.2016,
+      "step": 62900
+    },
+    {
+      "epoch": 5.1336375488917865,
+      "grad_norm": 10.067122459411621,
+      "learning_rate": 5.452998696219035e-06,
+      "loss": 0.2199,
+      "step": 63000
+    },
+    {
+      "epoch": 5.141786179921773,
+      "grad_norm": 8.625639915466309,
+      "learning_rate": 5.444035202086049e-06,
+      "loss": 0.197,
+      "step": 63100
+    },
+    {
+      "epoch": 5.14993481095176,
+      "grad_norm": 5.854794502258301,
+      "learning_rate": 5.435071707953064e-06,
+      "loss": 0.2087,
+      "step": 63200
+    },
+    {
+      "epoch": 5.158083441981747,
+      "grad_norm": 25.03852653503418,
+      "learning_rate": 5.426108213820078e-06,
+      "loss": 0.198,
+      "step": 63300
+    },
+    {
+      "epoch": 5.166232073011734,
+      "grad_norm": 8.568482398986816,
+      "learning_rate": 5.417144719687092e-06,
+      "loss": 0.2056,
+      "step": 63400
+    },
+    {
+      "epoch": 5.174380704041721,
+      "grad_norm": 13.770308494567871,
+      "learning_rate": 5.408181225554106e-06,
+      "loss": 0.2177,
+      "step": 63500
+    },
+    {
+      "epoch": 5.182529335071708,
+      "grad_norm": 5.927313804626465,
+      "learning_rate": 5.399217731421121e-06,
+      "loss": 0.2116,
+      "step": 63600
+    },
+    {
+      "epoch": 5.190677966101695,
+      "grad_norm": 8.017006874084473,
+      "learning_rate": 5.390254237288136e-06,
+      "loss": 0.2024,
+      "step": 63700
+    },
+    {
+      "epoch": 5.198826597131682,
+      "grad_norm": 12.88976001739502,
+      "learning_rate": 5.38129074315515e-06,
+      "loss": 0.2093,
+      "step": 63800
+    },
+    {
+      "epoch": 5.206975228161669,
+      "grad_norm": 7.503077507019043,
+      "learning_rate": 5.372327249022165e-06,
+      "loss": 0.2086,
+      "step": 63900
+    },
+    {
+      "epoch": 5.215123859191656,
+      "grad_norm": 5.918655872344971,
+      "learning_rate": 5.3633637548891785e-06,
+      "loss": 0.2132,
+      "step": 64000
+    },
+    {
+      "epoch": 5.2232724902216425,
+      "grad_norm": 14.097025871276855,
+      "learning_rate": 5.3544002607561925e-06,
+      "loss": 0.1944,
+      "step": 64100
+    },
+    {
+      "epoch": 5.23142112125163,
+      "grad_norm": 5.404310703277588,
+      "learning_rate": 5.345436766623207e-06,
+      "loss": 0.2086,
+      "step": 64200
+    },
+    {
+      "epoch": 5.239569752281617,
+      "grad_norm": 12.598578453063965,
+      "learning_rate": 5.336473272490221e-06,
+      "loss": 0.201,
+      "step": 64300
+    },
+    {
+      "epoch": 5.247718383311604,
+      "grad_norm": 7.65584135055542,
+      "learning_rate": 5.327509778357236e-06,
+      "loss": 0.1958,
+      "step": 64400
+    },
+    {
+      "epoch": 5.25586701434159,
+      "grad_norm": 6.702199459075928,
+      "learning_rate": 5.31854628422425e-06,
+      "loss": 0.2,
+      "step": 64500
+    },
+    {
+      "epoch": 5.264015645371577,
+      "grad_norm": 7.551068305969238,
+      "learning_rate": 5.309582790091266e-06,
+      "loss": 0.2015,
+      "step": 64600
+    },
+    {
+      "epoch": 5.272164276401565,
+      "grad_norm": 7.219007968902588,
+      "learning_rate": 5.300619295958279e-06,
+      "loss": 0.2151,
+      "step": 64700
+    },
+    {
+      "epoch": 5.280312907431552,
+      "grad_norm": 11.942194938659668,
+      "learning_rate": 5.291655801825294e-06,
+      "loss": 0.2096,
+      "step": 64800
+    },
+    {
+      "epoch": 5.288461538461538,
+      "grad_norm": 11.055607795715332,
+      "learning_rate": 5.2826923076923075e-06,
+      "loss": 0.1832,
+      "step": 64900
+    },
+    {
+      "epoch": 5.296610169491525,
+      "grad_norm": 9.625134468078613,
+      "learning_rate": 5.273728813559322e-06,
+      "loss": 0.2222,
+      "step": 65000
+    },
+    {
+      "epoch": 5.304758800521513,
+      "grad_norm": 8.117315292358398,
+      "learning_rate": 5.264765319426336e-06,
+      "loss": 0.2024,
+      "step": 65100
+    },
+    {
+      "epoch": 5.3129074315514995,
+      "grad_norm": 16.055788040161133,
+      "learning_rate": 5.255801825293351e-06,
+      "loss": 0.2045,
+      "step": 65200
+    },
+    {
+      "epoch": 5.321056062581486,
+      "grad_norm": 11.91893482208252,
+      "learning_rate": 5.246838331160365e-06,
+      "loss": 0.1999,
+      "step": 65300
+    },
+    {
+      "epoch": 5.329204693611473,
+      "grad_norm": 8.290375709533691,
+      "learning_rate": 5.237874837027379e-06,
+      "loss": 0.2037,
+      "step": 65400
+    },
+    {
+      "epoch": 5.337353324641461,
+      "grad_norm": 9.186257362365723,
+      "learning_rate": 5.228911342894394e-06,
+      "loss": 0.2167,
+      "step": 65500
+    },
+    {
+      "epoch": 5.345501955671447,
+      "grad_norm": 13.397846221923828,
+      "learning_rate": 5.219947848761408e-06,
+      "loss": 0.2201,
+      "step": 65600
+    },
+    {
+      "epoch": 5.353650586701434,
+      "grad_norm": 13.287492752075195,
+      "learning_rate": 5.210984354628423e-06,
+      "loss": 0.2168,
+      "step": 65700
+    },
+    {
+      "epoch": 5.361799217731421,
+      "grad_norm": 14.4291410446167,
+      "learning_rate": 5.202020860495437e-06,
+      "loss": 0.2131,
+      "step": 65800
+    },
+    {
+      "epoch": 5.369947848761408,
+      "grad_norm": 12.810196876525879,
+      "learning_rate": 5.193057366362451e-06,
+      "loss": 0.1926,
+      "step": 65900
+    },
+    {
+      "epoch": 5.378096479791395,
+      "grad_norm": 7.368422985076904,
+      "learning_rate": 5.184093872229465e-06,
+      "loss": 0.2103,
+      "step": 66000
+    },
+    {
+      "epoch": 5.386245110821382,
+      "grad_norm": 7.721367359161377,
+      "learning_rate": 5.175130378096479e-06,
+      "loss": 0.1999,
+      "step": 66100
+    },
+    {
+      "epoch": 5.394393741851369,
+      "grad_norm": 12.670356750488281,
+      "learning_rate": 5.166166883963495e-06,
+      "loss": 0.216,
+      "step": 66200
+    },
+    {
+      "epoch": 5.4025423728813555,
+      "grad_norm": 11.775321960449219,
+      "learning_rate": 5.157203389830509e-06,
+      "loss": 0.1863,
+      "step": 66300
+    },
+    {
+      "epoch": 5.410691003911343,
+      "grad_norm": 3.6226017475128174,
+      "learning_rate": 5.148239895697523e-06,
+      "loss": 0.1994,
+      "step": 66400
+    },
+    {
+      "epoch": 5.41883963494133,
+      "grad_norm": 19.766469955444336,
+      "learning_rate": 5.139276401564536e-06,
+      "loss": 0.2046,
+      "step": 66500
+    },
+    {
+      "epoch": 5.426988265971317,
+      "grad_norm": 14.729117393493652,
+      "learning_rate": 5.130312907431552e-06,
+      "loss": 0.2117,
+      "step": 66600
+    },
+    {
+      "epoch": 5.435136897001303,
+      "grad_norm": 5.150448322296143,
+      "learning_rate": 5.1213494132985655e-06,
+      "loss": 0.2051,
+      "step": 66700
+    },
+    {
+      "epoch": 5.443285528031291,
+      "grad_norm": 17.367290496826172,
+      "learning_rate": 5.11238591916558e-06,
+      "loss": 0.2148,
+      "step": 66800
+    },
+    {
+      "epoch": 5.451434159061278,
+      "grad_norm": 10.725879669189453,
+      "learning_rate": 5.103422425032595e-06,
+      "loss": 0.2168,
+      "step": 66900
+    },
+    {
+      "epoch": 5.459582790091265,
+      "grad_norm": 12.010479927062988,
+      "learning_rate": 5.094458930899608e-06,
+      "loss": 0.2032,
+      "step": 67000
+    },
+    {
+      "epoch": 5.467731421121251,
+      "grad_norm": 10.557023048400879,
+      "learning_rate": 5.085495436766624e-06,
+      "loss": 0.2118,
+      "step": 67100
+    },
+    {
+      "epoch": 5.475880052151239,
+      "grad_norm": 14.533101081848145,
+      "learning_rate": 5.076531942633637e-06,
+      "loss": 0.2186,
+      "step": 67200
+    },
+    {
+      "epoch": 5.484028683181226,
+      "grad_norm": 21.979625701904297,
+      "learning_rate": 5.067568448500652e-06,
+      "loss": 0.1842,
+      "step": 67300
+    },
+    {
+      "epoch": 5.4921773142112125,
+      "grad_norm": 12.857218742370605,
+      "learning_rate": 5.058604954367667e-06,
+      "loss": 0.2051,
+      "step": 67400
+    },
+    {
+      "epoch": 5.500325945241199,
+      "grad_norm": 9.391901969909668,
+      "learning_rate": 5.0496414602346814e-06,
+      "loss": 0.21,
+      "step": 67500
+    },
+    {
+      "epoch": 5.508474576271187,
+      "grad_norm": 11.181014060974121,
+      "learning_rate": 5.0406779661016945e-06,
+      "loss": 0.207,
+      "step": 67600
+    },
+    {
+      "epoch": 5.516623207301174,
+      "grad_norm": 11.92509651184082,
+      "learning_rate": 5.031714471968709e-06,
+      "loss": 0.191,
+      "step": 67700
+    },
+    {
+      "epoch": 5.52477183833116,
+      "grad_norm": 11.574973106384277,
+      "learning_rate": 5.022750977835723e-06,
+      "loss": 0.2046,
+      "step": 67800
+    },
+    {
+      "epoch": 5.532920469361147,
+      "grad_norm": 10.391201972961426,
+      "learning_rate": 5.013787483702736e-06,
+      "loss": 0.1916,
+      "step": 67900
+    },
+    {
+      "epoch": 5.541069100391134,
+      "grad_norm": 11.391480445861816,
+      "learning_rate": 5.004823989569753e-06,
+      "loss": 0.2191,
+      "step": 68000
+    },
+    {
+      "epoch": 5.5492177314211215,
+      "grad_norm": 4.8835015296936035,
+      "learning_rate": 4.995860495436767e-06,
+      "loss": 0.1916,
+      "step": 68100
+    },
+    {
+      "epoch": 5.557366362451108,
+      "grad_norm": 15.419282913208008,
+      "learning_rate": 4.986897001303781e-06,
+      "loss": 0.2146,
+      "step": 68200
+    },
+    {
+      "epoch": 5.565514993481095,
+      "grad_norm": 7.127319812774658,
+      "learning_rate": 4.977933507170796e-06,
+      "loss": 0.2111,
+      "step": 68300
+    },
+    {
+      "epoch": 5.573663624511082,
+      "grad_norm": 12.268532752990723,
+      "learning_rate": 4.96897001303781e-06,
+      "loss": 0.1966,
+      "step": 68400
+    },
+    {
+      "epoch": 5.581812255541069,
+      "grad_norm": 8.935240745544434,
+      "learning_rate": 4.9600065189048235e-06,
+      "loss": 0.2063,
+      "step": 68500
+    },
+    {
+      "epoch": 5.589960886571056,
+      "grad_norm": 7.016481876373291,
+      "learning_rate": 4.9510430247718375e-06,
+      "loss": 0.2156,
+      "step": 68600
+    },
+    {
+      "epoch": 5.598109517601043,
+      "grad_norm": 12.118579864501953,
+      "learning_rate": 4.942079530638852e-06,
+      "loss": 0.2162,
+      "step": 68700
+    },
+    {
+      "epoch": 5.60625814863103,
+      "grad_norm": 15.719651222229004,
+      "learning_rate": 4.933116036505867e-06,
+      "loss": 0.1974,
+      "step": 68800
+    },
+    {
+      "epoch": 5.614406779661017,
+      "grad_norm": 1.1470224857330322,
+      "learning_rate": 4.924152542372882e-06,
+      "loss": 0.1947,
+      "step": 68900
+    },
+    {
+      "epoch": 5.622555410691004,
+      "grad_norm": 9.488736152648926,
+      "learning_rate": 4.915189048239896e-06,
+      "loss": 0.207,
+      "step": 69000
+    },
+    {
+      "epoch": 5.630704041720991,
+      "grad_norm": 11.961145401000977,
+      "learning_rate": 4.90622555410691e-06,
+      "loss": 0.2143,
+      "step": 69100
+    },
+    {
+      "epoch": 5.638852672750978,
+      "grad_norm": 4.816403388977051,
+      "learning_rate": 4.897262059973925e-06,
+      "loss": 0.2119,
+      "step": 69200
+    },
+    {
+      "epoch": 5.647001303780964,
+      "grad_norm": 9.472099304199219,
+      "learning_rate": 4.888298565840939e-06,
+      "loss": 0.1907,
+      "step": 69300
+    },
+    {
+      "epoch": 5.655149934810952,
+      "grad_norm": 12.95845890045166,
+      "learning_rate": 4.8793350717079525e-06,
+      "loss": 0.2015,
+      "step": 69400
+    },
+    {
+      "epoch": 5.663298565840939,
+      "grad_norm": 17.84122657775879,
+      "learning_rate": 4.8703715775749665e-06,
+      "loss": 0.1871,
+      "step": 69500
+    },
+    {
+      "epoch": 5.6714471968709255,
+      "grad_norm": 12.41882610321045,
+      "learning_rate": 4.861408083441982e-06,
+      "loss": 0.2017,
+      "step": 69600
+    },
+    {
+      "epoch": 5.679595827900913,
+      "grad_norm": 2.849573850631714,
+      "learning_rate": 4.852444589308996e-06,
+      "loss": 0.1933,
+      "step": 69700
+    },
+    {
+      "epoch": 5.6877444589309,
+      "grad_norm": 8.450750350952148,
+      "learning_rate": 4.843481095176011e-06,
+      "loss": 0.2051,
+      "step": 69800
+    },
+    {
+      "epoch": 5.695893089960887,
+      "grad_norm": 2.862558126449585,
+      "learning_rate": 4.834517601043025e-06,
+      "loss": 0.217,
+      "step": 69900
+    },
+    {
+      "epoch": 5.704041720990873,
+      "grad_norm": 7.673337936401367,
+      "learning_rate": 4.825554106910039e-06,
+      "loss": 0.1976,
+      "step": 70000
+    },
+    {
+      "epoch": 5.71219035202086,
+      "grad_norm": 12.854650497436523,
+      "learning_rate": 4.816590612777053e-06,
+      "loss": 0.2092,
+      "step": 70100
+    },
+    {
+      "epoch": 5.720338983050848,
+      "grad_norm": 14.666707038879395,
+      "learning_rate": 4.807627118644068e-06,
+      "loss": 0.2147,
+      "step": 70200
+    },
+    {
+      "epoch": 5.7284876140808345,
+      "grad_norm": 12.781336784362793,
+      "learning_rate": 4.7986636245110815e-06,
+      "loss": 0.2145,
+      "step": 70300
+    },
+    {
+      "epoch": 5.736636245110821,
+      "grad_norm": 15.079524993896484,
+      "learning_rate": 4.789700130378097e-06,
+      "loss": 0.1991,
+      "step": 70400
+    },
+    {
+      "epoch": 5.744784876140808,
+      "grad_norm": 10.632801055908203,
+      "learning_rate": 4.780736636245111e-06,
+      "loss": 0.2046,
+      "step": 70500
+    },
+    {
+      "epoch": 5.752933507170796,
+      "grad_norm": 4.691617488861084,
+      "learning_rate": 4.771773142112125e-06,
+      "loss": 0.2143,
+      "step": 70600
+    },
+    {
+      "epoch": 5.761082138200782,
+      "grad_norm": 5.9435648918151855,
+      "learning_rate": 4.76280964797914e-06,
+      "loss": 0.211,
+      "step": 70700
+    },
+    {
+      "epoch": 5.769230769230769,
+      "grad_norm": 4.910902500152588,
+      "learning_rate": 4.753846153846154e-06,
+      "loss": 0.2491,
+      "step": 70800
+    },
+    {
+      "epoch": 5.777379400260756,
+      "grad_norm": 6.0848708152771,
+      "learning_rate": 4.744882659713168e-06,
+      "loss": 0.1952,
+      "step": 70900
+    },
+    {
+      "epoch": 5.7855280312907436,
+      "grad_norm": 7.89244270324707,
+      "learning_rate": 4.735919165580182e-06,
+      "loss": 0.2118,
+      "step": 71000
+    },
+    {
+      "epoch": 5.79367666232073,
+      "grad_norm": 3.871981382369995,
+      "learning_rate": 4.726955671447197e-06,
+      "loss": 0.2104,
+      "step": 71100
+    },
+    {
+      "epoch": 5.801825293350717,
+      "grad_norm": 13.02302074432373,
+      "learning_rate": 4.717992177314211e-06,
+      "loss": 0.203,
+      "step": 71200
+    },
+    {
+      "epoch": 5.809973924380704,
+      "grad_norm": 6.594838619232178,
+      "learning_rate": 4.709028683181226e-06,
+      "loss": 0.2154,
+      "step": 71300
+    },
+    {
+      "epoch": 5.818122555410691,
+      "grad_norm": 12.507122993469238,
+      "learning_rate": 4.70006518904824e-06,
+      "loss": 0.223,
+      "step": 71400
+    },
+    {
+      "epoch": 5.826271186440678,
+      "grad_norm": 3.6633384227752686,
+      "learning_rate": 4.691101694915254e-06,
+      "loss": 0.2032,
+      "step": 71500
+    },
+    {
+      "epoch": 5.834419817470665,
+      "grad_norm": 9.926380157470703,
+      "learning_rate": 4.682138200782268e-06,
+      "loss": 0.2046,
+      "step": 71600
+    },
+    {
+      "epoch": 5.842568448500652,
+      "grad_norm": 14.155366897583008,
+      "learning_rate": 4.673174706649283e-06,
+      "loss": 0.1853,
+      "step": 71700
+    },
+    {
+      "epoch": 5.8507170795306385,
+      "grad_norm": 9.108920097351074,
+      "learning_rate": 4.664211212516297e-06,
+      "loss": 0.2252,
+      "step": 71800
+    },
+    {
+      "epoch": 5.858865710560626,
+      "grad_norm": 7.253769397735596,
+      "learning_rate": 4.655247718383311e-06,
+      "loss": 0.196,
+      "step": 71900
+    },
+    {
+      "epoch": 5.867014341590613,
+      "grad_norm": 9.013677597045898,
+      "learning_rate": 4.6462842242503264e-06,
+      "loss": 0.2264,
+      "step": 72000
+    },
+    {
+      "epoch": 5.8751629726206,
+      "grad_norm": 8.925735473632812,
+      "learning_rate": 4.63732073011734e-06,
+      "loss": 0.2142,
+      "step": 72100
+    },
+    {
+      "epoch": 5.883311603650586,
+      "grad_norm": 9.083677291870117,
+      "learning_rate": 4.628357235984355e-06,
+      "loss": 0.225,
+      "step": 72200
+    },
+    {
+      "epoch": 5.891460234680574,
+      "grad_norm": 12.813594818115234,
+      "learning_rate": 4.619393741851369e-06,
+      "loss": 0.2093,
+      "step": 72300
+    },
+    {
+      "epoch": 5.899608865710561,
+      "grad_norm": 13.768078804016113,
+      "learning_rate": 4.610430247718383e-06,
+      "loss": 0.2031,
+      "step": 72400
+    },
+    {
+      "epoch": 5.9077574967405475,
+      "grad_norm": 8.680143356323242,
+      "learning_rate": 4.601466753585397e-06,
+      "loss": 0.2147,
+      "step": 72500
+    },
+    {
+      "epoch": 5.915906127770534,
+      "grad_norm": 17.846006393432617,
+      "learning_rate": 4.592503259452412e-06,
+      "loss": 0.2096,
+      "step": 72600
+    },
+    {
+      "epoch": 5.924054758800521,
+      "grad_norm": 8.560224533081055,
+      "learning_rate": 4.583539765319426e-06,
+      "loss": 0.1881,
+      "step": 72700
+    },
+    {
+      "epoch": 5.932203389830509,
+      "grad_norm": 5.689293384552002,
+      "learning_rate": 4.5745762711864415e-06,
+      "loss": 0.1971,
+      "step": 72800
+    },
+    {
+      "epoch": 5.940352020860495,
+      "grad_norm": 3.36490535736084,
+      "learning_rate": 4.5656127770534554e-06,
+      "loss": 0.2107,
+      "step": 72900
+    },
+    {
+      "epoch": 5.948500651890482,
+      "grad_norm": 6.769710063934326,
+      "learning_rate": 4.556649282920469e-06,
+      "loss": 0.2062,
+      "step": 73000
+    },
+    {
+      "epoch": 5.95664928292047,
+      "grad_norm": 4.0076775550842285,
+      "learning_rate": 4.547685788787483e-06,
+      "loss": 0.202,
+      "step": 73100
+    },
+    {
+      "epoch": 5.9647979139504566,
+      "grad_norm": 15.15523910522461,
+      "learning_rate": 4.538722294654498e-06,
+      "loss": 0.2186,
+      "step": 73200
+    },
+    {
+      "epoch": 5.972946544980443,
+      "grad_norm": 8.619361877441406,
+      "learning_rate": 4.529758800521512e-06,
+      "loss": 0.2087,
+      "step": 73300
+    },
+    {
+      "epoch": 5.98109517601043,
+      "grad_norm": 13.271175384521484,
+      "learning_rate": 4.520795306388526e-06,
+      "loss": 0.2069,
+      "step": 73400
+    },
+    {
+      "epoch": 5.989243807040417,
+      "grad_norm": 10.295662879943848,
+      "learning_rate": 4.511831812255541e-06,
+      "loss": 0.2219,
+      "step": 73500
+    },
+    {
+      "epoch": 5.9973924380704045,
+      "grad_norm": 11.802294731140137,
+      "learning_rate": 4.502868318122556e-06,
+      "loss": 0.1998,
+      "step": 73600
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.840562248995984,
+      "eval_loss": 0.6011542677879333,
+      "eval_runtime": 5.4949,
+      "eval_samples_per_second": 453.151,
+      "eval_steps_per_second": 56.78,
+      "step": 73632
+    },
+    {
+      "epoch": 6.005541069100391,
+      "grad_norm": 6.388237953186035,
+      "learning_rate": 4.4939048239895705e-06,
+      "loss": 0.1839,
+      "step": 73700
+    },
+    {
+      "epoch": 6.013689700130378,
+      "grad_norm": 10.506056785583496,
+      "learning_rate": 4.4849413298565844e-06,
+      "loss": 0.1734,
+      "step": 73800
+    },
+    {
+      "epoch": 6.021838331160365,
+      "grad_norm": 8.927146911621094,
+      "learning_rate": 4.475977835723598e-06,
+      "loss": 0.19,
+      "step": 73900
+    },
+    {
+      "epoch": 6.029986962190352,
+      "grad_norm": 7.502388000488281,
+      "learning_rate": 4.467014341590612e-06,
+      "loss": 0.1928,
+      "step": 74000
+    },
+    {
+      "epoch": 6.038135593220339,
+      "grad_norm": 3.7142844200134277,
+      "learning_rate": 4.458050847457627e-06,
+      "loss": 0.1615,
+      "step": 74100
+    },
+    {
+      "epoch": 6.046284224250326,
+      "grad_norm": 8.377524375915527,
+      "learning_rate": 4.449087353324641e-06,
+      "loss": 0.1725,
+      "step": 74200
+    },
+    {
+      "epoch": 6.054432855280313,
+      "grad_norm": 12.744202613830566,
+      "learning_rate": 4.440123859191655e-06,
+      "loss": 0.1858,
+      "step": 74300
+    },
+    {
+      "epoch": 6.0625814863103,
+      "grad_norm": 6.460068225860596,
+      "learning_rate": 4.43116036505867e-06,
+      "loss": 0.1941,
+      "step": 74400
+    },
+    {
+      "epoch": 6.070730117340287,
+      "grad_norm": 5.896920204162598,
+      "learning_rate": 4.422196870925685e-06,
+      "loss": 0.1656,
+      "step": 74500
+    },
+    {
+      "epoch": 6.078878748370274,
+      "grad_norm": 4.367869853973389,
+      "learning_rate": 4.413233376792699e-06,
+      "loss": 0.1582,
+      "step": 74600
+    },
+    {
+      "epoch": 6.0870273794002605,
+      "grad_norm": 18.771028518676758,
+      "learning_rate": 4.4042698826597134e-06,
+      "loss": 0.1585,
+      "step": 74700
+    },
+    {
+      "epoch": 6.095176010430248,
+      "grad_norm": 9.596067428588867,
+      "learning_rate": 4.395306388526727e-06,
+      "loss": 0.1734,
+      "step": 74800
+    },
+    {
+      "epoch": 6.103324641460235,
+      "grad_norm": 5.74586820602417,
+      "learning_rate": 4.386342894393741e-06,
+      "loss": 0.1721,
+      "step": 74900
+    },
+    {
+      "epoch": 6.111473272490222,
+      "grad_norm": 14.977295875549316,
+      "learning_rate": 4.377379400260756e-06,
+      "loss": 0.1664,
+      "step": 75000
+    },
+    {
+      "epoch": 6.119621903520208,
+      "grad_norm": 15.801685333251953,
+      "learning_rate": 4.36841590612777e-06,
+      "loss": 0.1658,
+      "step": 75100
+    },
+    {
+      "epoch": 6.127770534550195,
+      "grad_norm": 9.82263469696045,
+      "learning_rate": 4.359452411994784e-06,
+      "loss": 0.1708,
+      "step": 75200
+    },
+    {
+      "epoch": 6.135919165580183,
+      "grad_norm": 5.305476188659668,
+      "learning_rate": 4.3504889178618e-06,
+      "loss": 0.1879,
+      "step": 75300
+    },
+    {
+      "epoch": 6.1440677966101696,
+      "grad_norm": 11.85977554321289,
+      "learning_rate": 4.341525423728814e-06,
+      "loss": 0.1725,
+      "step": 75400
+    },
+    {
+      "epoch": 6.152216427640156,
+      "grad_norm": 12.436184883117676,
+      "learning_rate": 4.332561929595828e-06,
+      "loss": 0.1645,
+      "step": 75500
+    },
+    {
+      "epoch": 6.160365058670143,
+      "grad_norm": 4.3290324211120605,
+      "learning_rate": 4.3235984354628424e-06,
+      "loss": 0.1893,
+      "step": 75600
+    },
+    {
+      "epoch": 6.168513689700131,
+      "grad_norm": 3.737994432449341,
+      "learning_rate": 4.314634941329856e-06,
+      "loss": 0.189,
+      "step": 75700
+    },
+    {
+      "epoch": 6.1766623207301175,
+      "grad_norm": 7.4377241134643555,
+      "learning_rate": 4.30567144719687e-06,
+      "loss": 0.1878,
+      "step": 75800
+    },
+    {
+      "epoch": 6.184810951760104,
+      "grad_norm": 16.438753128051758,
+      "learning_rate": 4.296707953063885e-06,
+      "loss": 0.1941,
+      "step": 75900
+    },
+    {
+      "epoch": 6.192959582790091,
+      "grad_norm": 14.083968162536621,
+      "learning_rate": 4.287744458930899e-06,
+      "loss": 0.1814,
+      "step": 76000
+    },
+    {
+      "epoch": 6.201108213820079,
+      "grad_norm": 12.704230308532715,
+      "learning_rate": 4.278780964797914e-06,
+      "loss": 0.1782,
+      "step": 76100
+    },
+    {
+      "epoch": 6.209256844850065,
+      "grad_norm": 7.3511481285095215,
+      "learning_rate": 4.269817470664929e-06,
+      "loss": 0.1882,
+      "step": 76200
+    },
+    {
+      "epoch": 6.217405475880052,
+      "grad_norm": 9.752764701843262,
+      "learning_rate": 4.260853976531943e-06,
+      "loss": 0.1596,
+      "step": 76300
+    },
+    {
+      "epoch": 6.225554106910039,
+      "grad_norm": 9.257549285888672,
+      "learning_rate": 4.251890482398957e-06,
+      "loss": 0.1569,
+      "step": 76400
+    },
+    {
+      "epoch": 6.2337027379400265,
+      "grad_norm": 11.098052978515625,
+      "learning_rate": 4.2429269882659714e-06,
+      "loss": 0.1709,
+      "step": 76500
+    },
+    {
+      "epoch": 6.241851368970013,
+      "grad_norm": 14.279288291931152,
+      "learning_rate": 4.233963494132985e-06,
+      "loss": 0.1708,
+      "step": 76600
+    },
+    {
+      "epoch": 6.25,
+      "grad_norm": 10.652061462402344,
+      "learning_rate": 4.224999999999999e-06,
+      "loss": 0.1838,
+      "step": 76700
+    },
+    {
+      "epoch": 6.258148631029987,
+      "grad_norm": 19.59062385559082,
+      "learning_rate": 4.216036505867014e-06,
+      "loss": 0.1601,
+      "step": 76800
+    },
+    {
+      "epoch": 6.2662972620599735,
+      "grad_norm": 11.830570220947266,
+      "learning_rate": 4.207073011734029e-06,
+      "loss": 0.1888,
+      "step": 76900
+    },
+    {
+      "epoch": 6.274445893089961,
+      "grad_norm": 12.277135848999023,
+      "learning_rate": 4.198109517601043e-06,
+      "loss": 0.1573,
+      "step": 77000
+    },
+    {
+      "epoch": 6.282594524119948,
+      "grad_norm": 12.045758247375488,
+      "learning_rate": 4.189146023468058e-06,
+      "loss": 0.2031,
+      "step": 77100
+    },
+    {
+      "epoch": 6.290743155149935,
+      "grad_norm": 9.055584907531738,
+      "learning_rate": 4.180182529335072e-06,
+      "loss": 0.1797,
+      "step": 77200
+    },
+    {
+      "epoch": 6.298891786179921,
+      "grad_norm": 16.008596420288086,
+      "learning_rate": 4.171219035202086e-06,
+      "loss": 0.1665,
+      "step": 77300
+    },
+    {
+      "epoch": 6.307040417209909,
+      "grad_norm": 2.769876480102539,
+      "learning_rate": 4.1622555410691004e-06,
+      "loss": 0.1788,
+      "step": 77400
+    },
+    {
+      "epoch": 6.315189048239896,
+      "grad_norm": 13.493385314941406,
+      "learning_rate": 4.153292046936114e-06,
+      "loss": 0.1727,
+      "step": 77500
+    },
+    {
+      "epoch": 6.3233376792698825,
+      "grad_norm": 17.533010482788086,
+      "learning_rate": 4.144328552803128e-06,
+      "loss": 0.1961,
+      "step": 77600
+    },
+    {
+      "epoch": 6.331486310299869,
+      "grad_norm": 9.357499122619629,
+      "learning_rate": 4.135365058670144e-06,
+      "loss": 0.1772,
+      "step": 77700
+    },
+    {
+      "epoch": 6.339634941329857,
+      "grad_norm": 13.053030967712402,
+      "learning_rate": 4.126401564537158e-06,
+      "loss": 0.1685,
+      "step": 77800
+    },
+    {
+      "epoch": 6.347783572359844,
+      "grad_norm": 7.006876468658447,
+      "learning_rate": 4.117438070404172e-06,
+      "loss": 0.17,
+      "step": 77900
+    },
+    {
+      "epoch": 6.3559322033898304,
+      "grad_norm": 14.491143226623535,
+      "learning_rate": 4.108474576271187e-06,
+      "loss": 0.1748,
+      "step": 78000
+    },
+    {
+      "epoch": 6.364080834419817,
+      "grad_norm": 6.185390472412109,
+      "learning_rate": 4.099511082138201e-06,
+      "loss": 0.1654,
+      "step": 78100
+    },
+    {
+      "epoch": 6.372229465449805,
+      "grad_norm": 23.27703857421875,
+      "learning_rate": 4.090547588005215e-06,
+      "loss": 0.1704,
+      "step": 78200
+    },
+    {
+      "epoch": 6.380378096479792,
+      "grad_norm": 6.468029022216797,
+      "learning_rate": 4.0815840938722294e-06,
+      "loss": 0.1781,
+      "step": 78300
+    },
+    {
+      "epoch": 6.388526727509778,
+      "grad_norm": 12.720208168029785,
+      "learning_rate": 4.072620599739243e-06,
+      "loss": 0.1678,
+      "step": 78400
+    },
+    {
+      "epoch": 6.396675358539765,
+      "grad_norm": 2.81718373298645,
+      "learning_rate": 4.063657105606258e-06,
+      "loss": 0.1785,
+      "step": 78500
+    },
+    {
+      "epoch": 6.404823989569753,
+      "grad_norm": 18.40774154663086,
+      "learning_rate": 4.054693611473273e-06,
+      "loss": 0.1648,
+      "step": 78600
+    },
+    {
+      "epoch": 6.4129726205997395,
+      "grad_norm": 4.896127223968506,
+      "learning_rate": 4.045730117340287e-06,
+      "loss": 0.1961,
+      "step": 78700
+    },
+    {
+      "epoch": 6.421121251629726,
+      "grad_norm": 7.5092387199401855,
+      "learning_rate": 4.036766623207301e-06,
+      "loss": 0.1703,
+      "step": 78800
+    },
+    {
+      "epoch": 6.429269882659713,
+      "grad_norm": 6.800247669219971,
+      "learning_rate": 4.027803129074316e-06,
+      "loss": 0.1715,
+      "step": 78900
+    },
+    {
+      "epoch": 6.4374185136897,
+      "grad_norm": 12.569574356079102,
+      "learning_rate": 4.01883963494133e-06,
+      "loss": 0.1771,
+      "step": 79000
+    },
+    {
+      "epoch": 6.445567144719687,
+      "grad_norm": 7.636009216308594,
+      "learning_rate": 4.009876140808344e-06,
+      "loss": 0.1736,
+      "step": 79100
+    },
+    {
+      "epoch": 6.453715775749674,
+      "grad_norm": 7.997006893157959,
+      "learning_rate": 4.000912646675358e-06,
+      "loss": 0.169,
+      "step": 79200
+    },
+    {
+      "epoch": 6.461864406779661,
+      "grad_norm": 7.453688144683838,
+      "learning_rate": 3.991949152542373e-06,
+      "loss": 0.1636,
+      "step": 79300
+    },
+    {
+      "epoch": 6.470013037809648,
+      "grad_norm": 9.414899826049805,
+      "learning_rate": 3.982985658409387e-06,
+      "loss": 0.1678,
+      "step": 79400
+    },
+    {
+      "epoch": 6.478161668839635,
+      "grad_norm": 13.018781661987305,
+      "learning_rate": 3.974022164276402e-06,
+      "loss": 0.1668,
+      "step": 79500
+    },
+    {
+      "epoch": 6.486310299869622,
+      "grad_norm": 10.45935344696045,
+      "learning_rate": 3.965058670143416e-06,
+      "loss": 0.1866,
+      "step": 79600
+    },
+    {
+      "epoch": 6.494458930899609,
+      "grad_norm": 19.200838088989258,
+      "learning_rate": 3.95609517601043e-06,
+      "loss": 0.2034,
+      "step": 79700
+    },
+    {
+      "epoch": 6.5026075619295955,
+      "grad_norm": 15.825751304626465,
+      "learning_rate": 3.947131681877445e-06,
+      "loss": 0.1656,
+      "step": 79800
+    },
+    {
+      "epoch": 6.510756192959583,
+      "grad_norm": 2.095191240310669,
+      "learning_rate": 3.938168187744459e-06,
+      "loss": 0.1774,
+      "step": 79900
+    },
+    {
+      "epoch": 6.51890482398957,
+      "grad_norm": 3.431582450866699,
+      "learning_rate": 3.929204693611473e-06,
+      "loss": 0.1812,
+      "step": 80000
+    },
+    {
+      "epoch": 6.527053455019557,
+      "grad_norm": 10.681953430175781,
+      "learning_rate": 3.920241199478488e-06,
+      "loss": 0.1935,
+      "step": 80100
+    },
+    {
+      "epoch": 6.5352020860495434,
+      "grad_norm": 15.143997192382812,
+      "learning_rate": 3.911277705345502e-06,
+      "loss": 0.1792,
+      "step": 80200
+    },
+    {
+      "epoch": 6.54335071707953,
+      "grad_norm": 7.197531223297119,
+      "learning_rate": 3.902314211212517e-06,
+      "loss": 0.1718,
+      "step": 80300
+    },
+    {
+      "epoch": 6.551499348109518,
+      "grad_norm": 6.244622707366943,
+      "learning_rate": 3.893350717079531e-06,
+      "loss": 0.1591,
+      "step": 80400
+    },
+    {
+      "epoch": 6.559647979139505,
+      "grad_norm": 8.115144729614258,
+      "learning_rate": 3.884387222946545e-06,
+      "loss": 0.1619,
+      "step": 80500
+    },
+    {
+      "epoch": 6.567796610169491,
+      "grad_norm": 4.486932754516602,
+      "learning_rate": 3.875423728813559e-06,
+      "loss": 0.192,
+      "step": 80600
+    },
+    {
+      "epoch": 6.575945241199479,
+      "grad_norm": 12.3147611618042,
+      "learning_rate": 3.866460234680574e-06,
+      "loss": 0.1869,
+      "step": 80700
+    },
+    {
+      "epoch": 6.584093872229466,
+      "grad_norm": 8.49996280670166,
+      "learning_rate": 3.857496740547588e-06,
+      "loss": 0.1689,
+      "step": 80800
+    },
+    {
+      "epoch": 6.5922425032594525,
+      "grad_norm": 19.81438446044922,
+      "learning_rate": 3.848533246414603e-06,
+      "loss": 0.1918,
+      "step": 80900
+    },
+    {
+      "epoch": 6.600391134289439,
+      "grad_norm": 8.60556697845459,
+      "learning_rate": 3.839569752281617e-06,
+      "loss": 0.1951,
+      "step": 81000
+    },
+    {
+      "epoch": 6.608539765319426,
+      "grad_norm": 11.23202133178711,
+      "learning_rate": 3.830606258148631e-06,
+      "loss": 0.1531,
+      "step": 81100
+    },
+    {
+      "epoch": 6.616688396349414,
+      "grad_norm": 6.247717380523682,
+      "learning_rate": 3.821642764015645e-06,
+      "loss": 0.17,
+      "step": 81200
+    },
+    {
+      "epoch": 6.6248370273794,
+      "grad_norm": 5.568502902984619,
+      "learning_rate": 3.8126792698826596e-06,
+      "loss": 0.1765,
+      "step": 81300
+    },
+    {
+      "epoch": 6.632985658409387,
+      "grad_norm": 9.164602279663086,
+      "learning_rate": 3.803715775749674e-06,
+      "loss": 0.1467,
+      "step": 81400
+    },
+    {
+      "epoch": 6.641134289439374,
+      "grad_norm": 26.45659065246582,
+      "learning_rate": 3.7947522816166884e-06,
+      "loss": 0.1778,
+      "step": 81500
+    },
+    {
+      "epoch": 6.6492829204693615,
+      "grad_norm": 15.123627662658691,
+      "learning_rate": 3.7857887874837023e-06,
+      "loss": 0.171,
+      "step": 81600
+    },
+    {
+      "epoch": 6.657431551499348,
+      "grad_norm": 14.836517333984375,
+      "learning_rate": 3.7768252933507175e-06,
+      "loss": 0.175,
+      "step": 81700
+    },
+    {
+      "epoch": 6.665580182529335,
+      "grad_norm": 7.995541572570801,
+      "learning_rate": 3.767861799217732e-06,
+      "loss": 0.2045,
+      "step": 81800
+    },
+    {
+      "epoch": 6.673728813559322,
+      "grad_norm": 15.063085556030273,
+      "learning_rate": 3.758898305084746e-06,
+      "loss": 0.1797,
+      "step": 81900
+    },
+    {
+      "epoch": 6.681877444589309,
+      "grad_norm": 8.916457176208496,
+      "learning_rate": 3.7499348109517603e-06,
+      "loss": 0.1525,
+      "step": 82000
+    },
+    {
+      "epoch": 6.690026075619296,
+      "grad_norm": 8.451703071594238,
+      "learning_rate": 3.7409713168187746e-06,
+      "loss": 0.1755,
+      "step": 82100
+    },
+    {
+      "epoch": 6.698174706649283,
+      "grad_norm": 7.143068790435791,
+      "learning_rate": 3.7320078226857886e-06,
+      "loss": 0.1613,
+      "step": 82200
+    },
+    {
+      "epoch": 6.70632333767927,
+      "grad_norm": 12.28380012512207,
+      "learning_rate": 3.723044328552803e-06,
+      "loss": 0.1797,
+      "step": 82300
+    },
+    {
+      "epoch": 6.7144719687092564,
+      "grad_norm": 7.404759407043457,
+      "learning_rate": 3.714080834419817e-06,
+      "loss": 0.193,
+      "step": 82400
+    },
+    {
+      "epoch": 6.722620599739244,
+      "grad_norm": 14.824822425842285,
+      "learning_rate": 3.705117340286832e-06,
+      "loss": 0.1669,
+      "step": 82500
+    },
+    {
+      "epoch": 6.730769230769231,
+      "grad_norm": 8.087990760803223,
+      "learning_rate": 3.6961538461538465e-06,
+      "loss": 0.1909,
+      "step": 82600
+    },
+    {
+      "epoch": 6.738917861799218,
+      "grad_norm": 8.700764656066895,
+      "learning_rate": 3.687190352020861e-06,
+      "loss": 0.1634,
+      "step": 82700
+    },
+    {
+      "epoch": 6.747066492829204,
+      "grad_norm": 15.996380805969238,
+      "learning_rate": 3.678226857887875e-06,
+      "loss": 0.1715,
+      "step": 82800
+    },
+    {
+      "epoch": 6.755215123859192,
+      "grad_norm": 4.605205059051514,
+      "learning_rate": 3.6692633637548893e-06,
+      "loss": 0.1875,
+      "step": 82900
+    },
+    {
+      "epoch": 6.763363754889179,
+      "grad_norm": 12.218436241149902,
+      "learning_rate": 3.660299869621903e-06,
+      "loss": 0.1471,
+      "step": 83000
+    },
+    {
+      "epoch": 6.7715123859191655,
+      "grad_norm": 7.971900939941406,
+      "learning_rate": 3.6513363754889176e-06,
+      "loss": 0.1613,
+      "step": 83100
+    },
+    {
+      "epoch": 6.779661016949152,
+      "grad_norm": 15.262982368469238,
+      "learning_rate": 3.6423728813559315e-06,
+      "loss": 0.1711,
+      "step": 83200
+    },
+    {
+      "epoch": 6.78780964797914,
+      "grad_norm": 14.941751480102539,
+      "learning_rate": 3.633409387222946e-06,
+      "loss": 0.17,
+      "step": 83300
+    },
+    {
+      "epoch": 6.795958279009127,
+      "grad_norm": 18.507015228271484,
+      "learning_rate": 3.624445893089961e-06,
+      "loss": 0.1555,
+      "step": 83400
+    },
+    {
+      "epoch": 6.804106910039113,
+      "grad_norm": 9.77978801727295,
+      "learning_rate": 3.6154823989569755e-06,
+      "loss": 0.189,
+      "step": 83500
+    },
+    {
+      "epoch": 6.8122555410691,
+      "grad_norm": 6.868803977966309,
+      "learning_rate": 3.6065189048239895e-06,
+      "loss": 0.1675,
+      "step": 83600
+    },
+    {
+      "epoch": 6.820404172099087,
+      "grad_norm": 7.371277809143066,
+      "learning_rate": 3.597555410691004e-06,
+      "loss": 0.196,
+      "step": 83700
+    },
+    {
+      "epoch": 6.8285528031290745,
+      "grad_norm": 2.901946783065796,
+      "learning_rate": 3.5885919165580187e-06,
+      "loss": 0.1852,
+      "step": 83800
+    },
+    {
+      "epoch": 6.836701434159061,
+      "grad_norm": 15.448030471801758,
+      "learning_rate": 3.5796284224250322e-06,
+      "loss": 0.1678,
+      "step": 83900
+    },
+    {
+      "epoch": 6.844850065189048,
+      "grad_norm": 9.891990661621094,
+      "learning_rate": 3.570664928292046e-06,
+      "loss": 0.1885,
+      "step": 84000
+    },
+    {
+      "epoch": 6.852998696219036,
+      "grad_norm": 13.63137149810791,
+      "learning_rate": 3.5617014341590605e-06,
+      "loss": 0.1902,
+      "step": 84100
+    },
+    {
+      "epoch": 6.861147327249022,
+      "grad_norm": 4.282588481903076,
+      "learning_rate": 3.5527379400260758e-06,
+      "loss": 0.1888,
+      "step": 84200
+    },
+    {
+      "epoch": 6.869295958279009,
+      "grad_norm": 7.632056713104248,
+      "learning_rate": 3.54377444589309e-06,
+      "loss": 0.1703,
+      "step": 84300
+    },
+    {
+      "epoch": 6.877444589308996,
+      "grad_norm": 2.5631256103515625,
+      "learning_rate": 3.534810951760105e-06,
+      "loss": 0.1863,
+      "step": 84400
+    },
+    {
+      "epoch": 6.885593220338983,
+      "grad_norm": 11.351937294006348,
+      "learning_rate": 3.5258474576271185e-06,
+      "loss": 0.1753,
+      "step": 84500
+    },
+    {
+      "epoch": 6.89374185136897,
+      "grad_norm": 9.421881675720215,
+      "learning_rate": 3.5168839634941325e-06,
+      "loss": 0.1752,
+      "step": 84600
+    },
+    {
+      "epoch": 6.901890482398957,
+      "grad_norm": 7.395664691925049,
+      "learning_rate": 3.507920469361147e-06,
+      "loss": 0.1636,
+      "step": 84700
+    },
+    {
+      "epoch": 6.910039113428944,
+      "grad_norm": 10.84174633026123,
+      "learning_rate": 3.4989569752281612e-06,
+      "loss": 0.1781,
+      "step": 84800
+    },
+    {
+      "epoch": 6.918187744458931,
+      "grad_norm": 6.427243709564209,
+      "learning_rate": 3.4899934810951756e-06,
+      "loss": 0.1751,
+      "step": 84900
+    },
+    {
+      "epoch": 6.926336375488918,
+      "grad_norm": 5.036814212799072,
+      "learning_rate": 3.4810299869621912e-06,
+      "loss": 0.1758,
+      "step": 85000
+    },
+    {
+      "epoch": 6.934485006518905,
+      "grad_norm": 12.158060073852539,
+      "learning_rate": 3.4720664928292048e-06,
+      "loss": 0.1846,
+      "step": 85100
+    },
+    {
+      "epoch": 6.942633637548892,
+      "grad_norm": 9.157272338867188,
+      "learning_rate": 3.4631029986962187e-06,
+      "loss": 0.1807,
+      "step": 85200
+    },
+    {
+      "epoch": 6.9507822685788785,
+      "grad_norm": 19.283143997192383,
+      "learning_rate": 3.4541395045632335e-06,
+      "loss": 0.1778,
+      "step": 85300
+    },
+    {
+      "epoch": 6.958930899608866,
+      "grad_norm": 25.128320693969727,
+      "learning_rate": 3.4451760104302475e-06,
+      "loss": 0.1785,
+      "step": 85400
+    },
+    {
+      "epoch": 6.967079530638853,
+      "grad_norm": 16.265281677246094,
+      "learning_rate": 3.436212516297262e-06,
+      "loss": 0.1577,
+      "step": 85500
+    },
+    {
+      "epoch": 6.97522816166884,
+      "grad_norm": 26.714344024658203,
+      "learning_rate": 3.427249022164276e-06,
+      "loss": 0.1859,
+      "step": 85600
+    },
+    {
+      "epoch": 6.983376792698826,
+      "grad_norm": 6.852247714996338,
+      "learning_rate": 3.4182855280312906e-06,
+      "loss": 0.1724,
+      "step": 85700
+    },
+    {
+      "epoch": 6.991525423728813,
+      "grad_norm": 12.064892768859863,
+      "learning_rate": 3.409322033898305e-06,
+      "loss": 0.1944,
+      "step": 85800
+    },
+    {
+      "epoch": 6.999674054758801,
+      "grad_norm": 8.689165115356445,
+      "learning_rate": 3.40035853976532e-06,
+      "loss": 0.1737,
+      "step": 85900
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.8393574297188755,
+      "eval_loss": 0.636776864528656,
+      "eval_runtime": 5.4695,
+      "eval_samples_per_second": 455.252,
+      "eval_steps_per_second": 57.044,
+      "step": 85904
+    },
+    {
+      "epoch": 7.0078226857887875,
+      "grad_norm": 7.449398517608643,
+      "learning_rate": 3.3913950456323338e-06,
+      "loss": 0.1534,
+      "step": 86000
+    },
+    {
+      "epoch": 7.015971316818774,
+      "grad_norm": 4.026369571685791,
+      "learning_rate": 3.382431551499348e-06,
+      "loss": 0.1476,
+      "step": 86100
+    },
+    {
+      "epoch": 7.024119947848761,
+      "grad_norm": 12.344401359558105,
+      "learning_rate": 3.373468057366362e-06,
+      "loss": 0.1352,
+      "step": 86200
+    },
+    {
+      "epoch": 7.032268578878749,
+      "grad_norm": 2.723181962966919,
+      "learning_rate": 3.364504563233377e-06,
+      "loss": 0.1449,
+      "step": 86300
+    },
+    {
+      "epoch": 7.040417209908735,
+      "grad_norm": 13.284101486206055,
+      "learning_rate": 3.355541069100391e-06,
+      "loss": 0.1424,
+      "step": 86400
+    },
+    {
+      "epoch": 7.048565840938722,
+      "grad_norm": 7.289649963378906,
+      "learning_rate": 3.3465775749674053e-06,
+      "loss": 0.1556,
+      "step": 86500
+    },
+    {
+      "epoch": 7.056714471968709,
+      "grad_norm": 11.415511131286621,
+      "learning_rate": 3.3376140808344196e-06,
+      "loss": 0.1537,
+      "step": 86600
+    },
+    {
+      "epoch": 7.064863102998697,
+      "grad_norm": 10.688644409179688,
+      "learning_rate": 3.3286505867014344e-06,
+      "loss": 0.1539,
+      "step": 86700
+    },
+    {
+      "epoch": 7.073011734028683,
+      "grad_norm": 0.7387442588806152,
+      "learning_rate": 3.319687092568449e-06,
+      "loss": 0.152,
+      "step": 86800
+    },
+    {
+      "epoch": 7.08116036505867,
+      "grad_norm": 11.108467102050781,
+      "learning_rate": 3.3107235984354628e-06,
+      "loss": 0.1561,
+      "step": 86900
+    },
+    {
+      "epoch": 7.089308996088657,
+      "grad_norm": 3.977663040161133,
+      "learning_rate": 3.301760104302477e-06,
+      "loss": 0.1588,
+      "step": 87000
+    },
+    {
+      "epoch": 7.0974576271186445,
+      "grad_norm": 15.751721382141113,
+      "learning_rate": 3.2927966101694915e-06,
+      "loss": 0.1555,
+      "step": 87100
+    },
+    {
+      "epoch": 7.105606258148631,
+      "grad_norm": 8.952390670776367,
+      "learning_rate": 3.2838331160365055e-06,
+      "loss": 0.1501,
+      "step": 87200
+    },
+    {
+      "epoch": 7.113754889178618,
+      "grad_norm": 8.698328971862793,
+      "learning_rate": 3.274869621903519e-06,
+      "loss": 0.1579,
+      "step": 87300
+    },
+    {
+      "epoch": 7.121903520208605,
+      "grad_norm": 6.493157386779785,
+      "learning_rate": 3.265906127770535e-06,
+      "loss": 0.1585,
+      "step": 87400
+    },
+    {
+      "epoch": 7.130052151238592,
+      "grad_norm": 12.06603717803955,
+      "learning_rate": 3.256942633637549e-06,
+      "loss": 0.1498,
+      "step": 87500
+    },
+    {
+      "epoch": 7.138200782268579,
+      "grad_norm": 8.994619369506836,
+      "learning_rate": 3.2479791395045634e-06,
+      "loss": 0.1561,
+      "step": 87600
+    },
+    {
+      "epoch": 7.146349413298566,
+      "grad_norm": 8.27117919921875,
+      "learning_rate": 3.239015645371578e-06,
+      "loss": 0.1532,
+      "step": 87700
+    },
+    {
+      "epoch": 7.154498044328553,
+      "grad_norm": 14.190399169921875,
+      "learning_rate": 3.2300521512385918e-06,
+      "loss": 0.1447,
+      "step": 87800
+    },
+    {
+      "epoch": 7.162646675358539,
+      "grad_norm": 5.721526145935059,
+      "learning_rate": 3.221088657105606e-06,
+      "loss": 0.1525,
+      "step": 87900
+    },
+    {
+      "epoch": 7.170795306388527,
+      "grad_norm": 9.777070999145508,
+      "learning_rate": 3.21212516297262e-06,
+      "loss": 0.154,
+      "step": 88000
+    },
+    {
+      "epoch": 7.178943937418514,
+      "grad_norm": 9.652997016906738,
+      "learning_rate": 3.2031616688396345e-06,
+      "loss": 0.1634,
+      "step": 88100
+    },
+    {
+      "epoch": 7.1870925684485005,
+      "grad_norm": 18.047399520874023,
+      "learning_rate": 3.1941981747066497e-06,
+      "loss": 0.1412,
+      "step": 88200
+    },
+    {
+      "epoch": 7.195241199478487,
+      "grad_norm": 28.265060424804688,
+      "learning_rate": 3.185234680573664e-06,
+      "loss": 0.1419,
+      "step": 88300
+    },
+    {
+      "epoch": 7.203389830508475,
+      "grad_norm": 13.160205841064453,
+      "learning_rate": 3.176271186440678e-06,
+      "loss": 0.1427,
+      "step": 88400
+    },
+    {
+      "epoch": 7.211538461538462,
+      "grad_norm": 10.281450271606445,
+      "learning_rate": 3.1673076923076924e-06,
+      "loss": 0.1447,
+      "step": 88500
+    },
+    {
+      "epoch": 7.219687092568448,
+      "grad_norm": 13.10381031036377,
+      "learning_rate": 3.1583441981747064e-06,
+      "loss": 0.1358,
+      "step": 88600
+    },
+    {
+      "epoch": 7.227835723598435,
+      "grad_norm": 6.493006706237793,
+      "learning_rate": 3.1493807040417208e-06,
+      "loss": 0.1317,
+      "step": 88700
+    },
+    {
+      "epoch": 7.235984354628423,
+      "grad_norm": 3.7273976802825928,
+      "learning_rate": 3.1404172099087347e-06,
+      "loss": 0.1464,
+      "step": 88800
+    },
+    {
+      "epoch": 7.24413298565841,
+      "grad_norm": 17.353458404541016,
+      "learning_rate": 3.131453715775749e-06,
+      "loss": 0.1413,
+      "step": 88900
+    },
+    {
+      "epoch": 7.252281616688396,
+      "grad_norm": 17.094566345214844,
+      "learning_rate": 3.1224902216427643e-06,
+      "loss": 0.163,
+      "step": 89000
+    },
+    {
+      "epoch": 7.260430247718383,
+      "grad_norm": 17.34813117980957,
+      "learning_rate": 3.1135267275097787e-06,
+      "loss": 0.1454,
+      "step": 89100
+    },
+    {
+      "epoch": 7.26857887874837,
+      "grad_norm": 2.7473742961883545,
+      "learning_rate": 3.1045632333767927e-06,
+      "loss": 0.1515,
+      "step": 89200
+    },
+    {
+      "epoch": 7.2767275097783575,
+      "grad_norm": 10.833642959594727,
+      "learning_rate": 3.095599739243807e-06,
+      "loss": 0.1445,
+      "step": 89300
+    },
+    {
+      "epoch": 7.284876140808344,
+      "grad_norm": 10.2700777053833,
+      "learning_rate": 3.086636245110822e-06,
+      "loss": 0.1595,
+      "step": 89400
+    },
+    {
+      "epoch": 7.293024771838331,
+      "grad_norm": 7.091324329376221,
+      "learning_rate": 3.0776727509778354e-06,
+      "loss": 0.1601,
+      "step": 89500
+    },
+    {
+      "epoch": 7.301173402868318,
+      "grad_norm": 14.033100128173828,
+      "learning_rate": 3.0687092568448494e-06,
+      "loss": 0.1565,
+      "step": 89600
+    },
+    {
+      "epoch": 7.309322033898305,
+      "grad_norm": 1.7879611253738403,
+      "learning_rate": 3.0597457627118637e-06,
+      "loss": 0.1334,
+      "step": 89700
+    },
+    {
+      "epoch": 7.317470664928292,
+      "grad_norm": 15.966545104980469,
+      "learning_rate": 3.05078226857888e-06,
+      "loss": 0.1591,
+      "step": 89800
+    },
+    {
+      "epoch": 7.325619295958279,
+      "grad_norm": 17.925477981567383,
+      "learning_rate": 3.0418187744458933e-06,
+      "loss": 0.1611,
+      "step": 89900
+    },
+    {
+      "epoch": 7.333767926988266,
+      "grad_norm": 28.116106033325195,
+      "learning_rate": 3.032855280312908e-06,
+      "loss": 0.1397,
+      "step": 90000
+    },
+    {
+      "epoch": 7.341916558018253,
+      "grad_norm": 22.7088565826416,
+      "learning_rate": 3.0238917861799217e-06,
+      "loss": 0.1345,
+      "step": 90100
+    },
+    {
+      "epoch": 7.35006518904824,
+      "grad_norm": 7.538867473602295,
+      "learning_rate": 3.0149282920469356e-06,
+      "loss": 0.1332,
+      "step": 90200
+    },
+    {
+      "epoch": 7.358213820078227,
+      "grad_norm": 15.904698371887207,
+      "learning_rate": 3.00596479791395e-06,
+      "loss": 0.1734,
+      "step": 90300
+    },
+    {
+      "epoch": 7.3663624511082135,
+      "grad_norm": 17.251426696777344,
+      "learning_rate": 2.9970013037809644e-06,
+      "loss": 0.1524,
+      "step": 90400
+    },
+    {
+      "epoch": 7.374511082138201,
+      "grad_norm": 18.56187629699707,
+      "learning_rate": 2.9880378096479788e-06,
+      "loss": 0.1484,
+      "step": 90500
+    },
+    {
+      "epoch": 7.382659713168188,
+      "grad_norm": 9.472875595092773,
+      "learning_rate": 2.9790743155149944e-06,
+      "loss": 0.1491,
+      "step": 90600
+    },
+    {
+      "epoch": 7.390808344198175,
+      "grad_norm": 11.260825157165527,
+      "learning_rate": 2.970110821382008e-06,
+      "loss": 0.1522,
+      "step": 90700
+    },
+    {
+      "epoch": 7.398956975228161,
+      "grad_norm": 23.518583297729492,
+      "learning_rate": 2.961147327249022e-06,
+      "loss": 0.1507,
+      "step": 90800
+    },
+    {
+      "epoch": 7.407105606258149,
+      "grad_norm": 6.331607818603516,
+      "learning_rate": 2.9521838331160367e-06,
+      "loss": 0.1562,
+      "step": 90900
+    },
+    {
+      "epoch": 7.415254237288136,
+      "grad_norm": 23.582700729370117,
+      "learning_rate": 2.9432203389830507e-06,
+      "loss": 0.1432,
+      "step": 91000
+    },
+    {
+      "epoch": 7.423402868318123,
+      "grad_norm": 14.635252952575684,
+      "learning_rate": 2.934256844850065e-06,
+      "loss": 0.1496,
+      "step": 91100
+    },
+    {
+      "epoch": 7.431551499348109,
+      "grad_norm": 8.255620002746582,
+      "learning_rate": 2.925293350717079e-06,
+      "loss": 0.1602,
+      "step": 91200
+    },
+    {
+      "epoch": 7.439700130378096,
+      "grad_norm": 3.5353052616119385,
+      "learning_rate": 2.916329856584094e-06,
+      "loss": 0.1508,
+      "step": 91300
+    },
+    {
+      "epoch": 7.447848761408084,
+      "grad_norm": 10.22754192352295,
+      "learning_rate": 2.9073663624511078e-06,
+      "loss": 0.1417,
+      "step": 91400
+    },
+    {
+      "epoch": 7.4559973924380705,
+      "grad_norm": 8.802057266235352,
+      "learning_rate": 2.898402868318123e-06,
+      "loss": 0.1555,
+      "step": 91500
+    },
+    {
+      "epoch": 7.464146023468057,
+      "grad_norm": 7.474407196044922,
+      "learning_rate": 2.889439374185137e-06,
+      "loss": 0.1658,
+      "step": 91600
+    },
+    {
+      "epoch": 7.472294654498044,
+      "grad_norm": 18.78350830078125,
+      "learning_rate": 2.8804758800521513e-06,
+      "loss": 0.1482,
+      "step": 91700
+    },
+    {
+      "epoch": 7.480443285528032,
+      "grad_norm": 8.468660354614258,
+      "learning_rate": 2.8715123859191653e-06,
+      "loss": 0.1704,
+      "step": 91800
+    },
+    {
+      "epoch": 7.488591916558018,
+      "grad_norm": 0.4872637391090393,
+      "learning_rate": 2.8625488917861797e-06,
+      "loss": 0.1515,
+      "step": 91900
+    },
+    {
+      "epoch": 7.496740547588005,
+      "grad_norm": 8.285701751708984,
+      "learning_rate": 2.853585397653194e-06,
+      "loss": 0.1685,
+      "step": 92000
+    },
+    {
+      "epoch": 7.504889178617992,
+      "grad_norm": 19.152118682861328,
+      "learning_rate": 2.8446219035202084e-06,
+      "loss": 0.1515,
+      "step": 92100
+    },
+    {
+      "epoch": 7.5130378096479795,
+      "grad_norm": 22.134498596191406,
+      "learning_rate": 2.8356584093872224e-06,
+      "loss": 0.1412,
+      "step": 92200
+    },
+    {
+      "epoch": 7.521186440677966,
+      "grad_norm": 22.581012725830078,
+      "learning_rate": 2.8266949152542376e-06,
+      "loss": 0.155,
+      "step": 92300
+    },
+    {
+      "epoch": 7.529335071707953,
+      "grad_norm": 9.176002502441406,
+      "learning_rate": 2.817731421121252e-06,
+      "loss": 0.154,
+      "step": 92400
+    },
+    {
+      "epoch": 7.53748370273794,
+      "grad_norm": 12.96983814239502,
+      "learning_rate": 2.808767926988266e-06,
+      "loss": 0.1656,
+      "step": 92500
+    },
+    {
+      "epoch": 7.5456323337679265,
+      "grad_norm": 6.628359317779541,
+      "learning_rate": 2.7998044328552803e-06,
+      "loss": 0.1413,
+      "step": 92600
+    },
+    {
+      "epoch": 7.553780964797914,
+      "grad_norm": 4.591996192932129,
+      "learning_rate": 2.7908409387222947e-06,
+      "loss": 0.1544,
+      "step": 92700
+    },
+    {
+      "epoch": 7.561929595827901,
+      "grad_norm": 13.306089401245117,
+      "learning_rate": 2.7818774445893087e-06,
+      "loss": 0.1523,
+      "step": 92800
+    },
+    {
+      "epoch": 7.570078226857888,
+      "grad_norm": 13.552392959594727,
+      "learning_rate": 2.772913950456323e-06,
+      "loss": 0.1417,
+      "step": 92900
+    },
+    {
+      "epoch": 7.578226857887875,
+      "grad_norm": 10.100014686584473,
+      "learning_rate": 2.763950456323337e-06,
+      "loss": 0.1491,
+      "step": 93000
+    },
+    {
+      "epoch": 7.586375488917862,
+      "grad_norm": 13.423697471618652,
+      "learning_rate": 2.7549869621903523e-06,
+      "loss": 0.1638,
+      "step": 93100
+    },
+    {
+      "epoch": 7.594524119947849,
+      "grad_norm": 16.027891159057617,
+      "learning_rate": 2.7460234680573666e-06,
+      "loss": 0.1548,
+      "step": 93200
+    },
+    {
+      "epoch": 7.602672750977836,
+      "grad_norm": 15.648063659667969,
+      "learning_rate": 2.737059973924381e-06,
+      "loss": 0.1601,
+      "step": 93300
+    },
+    {
+      "epoch": 7.610821382007822,
+      "grad_norm": 17.28276824951172,
+      "learning_rate": 2.728096479791395e-06,
+      "loss": 0.1455,
+      "step": 93400
+    },
+    {
+      "epoch": 7.61897001303781,
+      "grad_norm": 1.7760809659957886,
+      "learning_rate": 2.7191329856584093e-06,
+      "loss": 0.1523,
+      "step": 93500
+    },
+    {
+      "epoch": 7.627118644067797,
+      "grad_norm": 12.409880638122559,
+      "learning_rate": 2.7101694915254233e-06,
+      "loss": 0.17,
+      "step": 93600
+    },
+    {
+      "epoch": 7.6352672750977835,
+      "grad_norm": 8.9781494140625,
+      "learning_rate": 2.7012059973924377e-06,
+      "loss": 0.1628,
+      "step": 93700
+    },
+    {
+      "epoch": 7.64341590612777,
+      "grad_norm": 11.374709129333496,
+      "learning_rate": 2.6922425032594516e-06,
+      "loss": 0.1584,
+      "step": 93800
+    },
+    {
+      "epoch": 7.651564537157758,
+      "grad_norm": 8.572677612304688,
+      "learning_rate": 2.6832790091264673e-06,
+      "loss": 0.1581,
+      "step": 93900
+    },
+    {
+      "epoch": 7.659713168187745,
+      "grad_norm": 15.630659103393555,
+      "learning_rate": 2.6743155149934813e-06,
+      "loss": 0.1547,
+      "step": 94000
+    },
+    {
+      "epoch": 7.667861799217731,
+      "grad_norm": 2.3736350536346436,
+      "learning_rate": 2.6653520208604956e-06,
+      "loss": 0.1588,
+      "step": 94100
+    },
+    {
+      "epoch": 7.676010430247718,
+      "grad_norm": 6.438554286956787,
+      "learning_rate": 2.6563885267275096e-06,
+      "loss": 0.1483,
+      "step": 94200
+    },
+    {
+      "epoch": 7.684159061277706,
+      "grad_norm": 3.4226784706115723,
+      "learning_rate": 2.647425032594524e-06,
+      "loss": 0.1543,
+      "step": 94300
+    },
+    {
+      "epoch": 7.6923076923076925,
+      "grad_norm": 0.7672721147537231,
+      "learning_rate": 2.638461538461538e-06,
+      "loss": 0.1519,
+      "step": 94400
+    },
+    {
+      "epoch": 7.700456323337679,
+      "grad_norm": 14.213494300842285,
+      "learning_rate": 2.6294980443285523e-06,
+      "loss": 0.1406,
+      "step": 94500
+    },
+    {
+      "epoch": 7.708604954367666,
+      "grad_norm": 0.7124353647232056,
+      "learning_rate": 2.6205345501955663e-06,
+      "loss": 0.144,
+      "step": 94600
+    },
+    {
+      "epoch": 7.716753585397653,
+      "grad_norm": 23.662931442260742,
+      "learning_rate": 2.611571056062582e-06,
+      "loss": 0.1617,
+      "step": 94700
+    },
+    {
+      "epoch": 7.72490221642764,
+      "grad_norm": 11.896281242370605,
+      "learning_rate": 2.602607561929596e-06,
+      "loss": 0.1404,
+      "step": 94800
+    },
+    {
+      "epoch": 7.733050847457627,
+      "grad_norm": 21.615901947021484,
+      "learning_rate": 2.5936440677966103e-06,
+      "loss": 0.156,
+      "step": 94900
+    },
+    {
+      "epoch": 7.741199478487614,
+      "grad_norm": 1.9456912279129028,
+      "learning_rate": 2.584680573663625e-06,
+      "loss": 0.1591,
+      "step": 95000
+    },
+    {
+      "epoch": 7.749348109517601,
+      "grad_norm": 12.462078094482422,
+      "learning_rate": 2.5757170795306386e-06,
+      "loss": 0.1598,
+      "step": 95100
+    },
+    {
+      "epoch": 7.757496740547588,
+      "grad_norm": 6.921799182891846,
+      "learning_rate": 2.5667535853976525e-06,
+      "loss": 0.1479,
+      "step": 95200
+    },
+    {
+      "epoch": 7.765645371577575,
+      "grad_norm": 5.263434410095215,
+      "learning_rate": 2.557790091264667e-06,
+      "loss": 0.1522,
+      "step": 95300
+    },
+    {
+      "epoch": 7.773794002607562,
+      "grad_norm": 6.944530963897705,
+      "learning_rate": 2.5488265971316813e-06,
+      "loss": 0.1598,
+      "step": 95400
+    },
+    {
+      "epoch": 7.781942633637549,
+      "grad_norm": 13.571086883544922,
+      "learning_rate": 2.5398631029986965e-06,
+      "loss": 0.1611,
+      "step": 95500
+    },
+    {
+      "epoch": 7.790091264667536,
+      "grad_norm": 17.36863899230957,
+      "learning_rate": 2.5308996088657113e-06,
+      "loss": 0.153,
+      "step": 95600
+    },
+    {
+      "epoch": 7.798239895697523,
+      "grad_norm": 8.735814094543457,
+      "learning_rate": 2.521936114732725e-06,
+      "loss": 0.138,
+      "step": 95700
+    },
+    {
+      "epoch": 7.80638852672751,
+      "grad_norm": 15.301424980163574,
+      "learning_rate": 2.512972620599739e-06,
+      "loss": 0.1572,
+      "step": 95800
+    },
+    {
+      "epoch": 7.8145371577574965,
+      "grad_norm": 15.560503005981445,
+      "learning_rate": 2.5040091264667536e-06,
+      "loss": 0.1446,
+      "step": 95900
+    },
+    {
+      "epoch": 7.822685788787483,
+      "grad_norm": 16.19012451171875,
+      "learning_rate": 2.4950456323337676e-06,
+      "loss": 0.1742,
+      "step": 96000
+    },
+    {
+      "epoch": 7.830834419817471,
+      "grad_norm": 4.544064998626709,
+      "learning_rate": 2.486082138200782e-06,
+      "loss": 0.1516,
+      "step": 96100
+    },
+    {
+      "epoch": 7.838983050847458,
+      "grad_norm": 5.693359375,
+      "learning_rate": 2.477118644067796e-06,
+      "loss": 0.1588,
+      "step": 96200
+    },
+    {
+      "epoch": 7.847131681877444,
+      "grad_norm": 14.035545349121094,
+      "learning_rate": 2.468155149934811e-06,
+      "loss": 0.1373,
+      "step": 96300
+    },
+    {
+      "epoch": 7.855280312907432,
+      "grad_norm": 15.699686050415039,
+      "learning_rate": 2.4591916558018255e-06,
+      "loss": 0.143,
+      "step": 96400
+    },
+    {
+      "epoch": 7.863428943937419,
+      "grad_norm": 17.58596420288086,
+      "learning_rate": 2.45022816166884e-06,
+      "loss": 0.1504,
+      "step": 96500
+    },
+    {
+      "epoch": 7.8715775749674055,
+      "grad_norm": 12.520322799682617,
+      "learning_rate": 2.441264667535854e-06,
+      "loss": 0.1642,
+      "step": 96600
+    },
+    {
+      "epoch": 7.879726205997392,
+      "grad_norm": 6.03079080581665,
+      "learning_rate": 2.4323011734028683e-06,
+      "loss": 0.1493,
+      "step": 96700
+    },
+    {
+      "epoch": 7.887874837027379,
+      "grad_norm": 11.687211990356445,
+      "learning_rate": 2.423337679269882e-06,
+      "loss": 0.1567,
+      "step": 96800
+    },
+    {
+      "epoch": 7.896023468057367,
+      "grad_norm": 16.79334259033203,
+      "learning_rate": 2.4143741851368966e-06,
+      "loss": 0.1635,
+      "step": 96900
+    },
+    {
+      "epoch": 7.904172099087353,
+      "grad_norm": 8.044095039367676,
+      "learning_rate": 2.405410691003911e-06,
+      "loss": 0.1485,
+      "step": 97000
+    },
+    {
+      "epoch": 7.91232073011734,
+      "grad_norm": 10.20647144317627,
+      "learning_rate": 2.396447196870926e-06,
+      "loss": 0.163,
+      "step": 97100
+    },
+    {
+      "epoch": 7.920469361147327,
+      "grad_norm": 7.316319465637207,
+      "learning_rate": 2.38748370273794e-06,
+      "loss": 0.165,
+      "step": 97200
+    },
+    {
+      "epoch": 7.9286179921773146,
+      "grad_norm": 6.296102523803711,
+      "learning_rate": 2.3785202086049545e-06,
+      "loss": 0.1644,
+      "step": 97300
+    },
+    {
+      "epoch": 7.936766623207301,
+      "grad_norm": 7.60091495513916,
+      "learning_rate": 2.369556714471969e-06,
+      "loss": 0.1559,
+      "step": 97400
+    },
+    {
+      "epoch": 7.944915254237288,
+      "grad_norm": 0.3591231405735016,
+      "learning_rate": 2.360593220338983e-06,
+      "loss": 0.1637,
+      "step": 97500
+    },
+    {
+      "epoch": 7.953063885267275,
+      "grad_norm": 13.358650207519531,
+      "learning_rate": 2.3516297262059973e-06,
+      "loss": 0.1576,
+      "step": 97600
+    },
+    {
+      "epoch": 7.9612125162972625,
+      "grad_norm": 3.420527696609497,
+      "learning_rate": 2.3426662320730112e-06,
+      "loss": 0.1362,
+      "step": 97700
+    },
+    {
+      "epoch": 7.969361147327249,
+      "grad_norm": 14.245848655700684,
+      "learning_rate": 2.3337027379400256e-06,
+      "loss": 0.1487,
+      "step": 97800
+    },
+    {
+      "epoch": 7.977509778357236,
+      "grad_norm": 12.376141548156738,
+      "learning_rate": 2.324739243807041e-06,
+      "loss": 0.14,
+      "step": 97900
+    },
+    {
+      "epoch": 7.985658409387223,
+      "grad_norm": 0.6416822671890259,
+      "learning_rate": 2.315775749674055e-06,
+      "loss": 0.1541,
+      "step": 98000
+    },
+    {
+      "epoch": 7.9938070404172095,
+      "grad_norm": 10.687934875488281,
+      "learning_rate": 2.306812255541069e-06,
+      "loss": 0.1485,
+      "step": 98100
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.8417670682730923,
+      "eval_loss": 0.6814994215965271,
+      "eval_runtime": 5.5348,
+      "eval_samples_per_second": 449.877,
+      "eval_steps_per_second": 56.37,
+      "step": 98176
+    },
+    {
+      "epoch": 8.001955671447197,
+      "grad_norm": 1.1167025566101074,
+      "learning_rate": 2.2978487614080835e-06,
+      "loss": 0.1672,
+      "step": 98200
+    },
+    {
+      "epoch": 8.010104302477183,
+      "grad_norm": 18.145008087158203,
+      "learning_rate": 2.2888852672750975e-06,
+      "loss": 0.116,
+      "step": 98300
+    },
+    {
+      "epoch": 8.01825293350717,
+      "grad_norm": 9.317021369934082,
+      "learning_rate": 2.279921773142112e-06,
+      "loss": 0.138,
+      "step": 98400
+    },
+    {
+      "epoch": 8.026401564537158,
+      "grad_norm": 9.0396089553833,
+      "learning_rate": 2.2709582790091263e-06,
+      "loss": 0.1428,
+      "step": 98500
+    },
+    {
+      "epoch": 8.034550195567144,
+      "grad_norm": 13.613592147827148,
+      "learning_rate": 2.2619947848761402e-06,
+      "loss": 0.1222,
+      "step": 98600
+    },
+    {
+      "epoch": 8.042698826597132,
+      "grad_norm": 8.281913757324219,
+      "learning_rate": 2.2530312907431554e-06,
+      "loss": 0.1309,
+      "step": 98700
+    },
+    {
+      "epoch": 8.05084745762712,
+      "grad_norm": 2.797579050064087,
+      "learning_rate": 2.24406779661017e-06,
+      "loss": 0.1313,
+      "step": 98800
+    },
+    {
+      "epoch": 8.058996088657105,
+      "grad_norm": 13.423441886901855,
+      "learning_rate": 2.235104302477184e-06,
+      "loss": 0.1223,
+      "step": 98900
+    },
+    {
+      "epoch": 8.067144719687093,
+      "grad_norm": 2.5191502571105957,
+      "learning_rate": 2.226140808344198e-06,
+      "loss": 0.1175,
+      "step": 99000
+    },
+    {
+      "epoch": 8.075293350717079,
+      "grad_norm": 9.729893684387207,
+      "learning_rate": 2.2171773142112125e-06,
+      "loss": 0.1281,
+      "step": 99100
+    },
+    {
+      "epoch": 8.083441981747066,
+      "grad_norm": 10.935778617858887,
+      "learning_rate": 2.2082138200782265e-06,
+      "loss": 0.1274,
+      "step": 99200
+    },
+    {
+      "epoch": 8.091590612777054,
+      "grad_norm": 14.475391387939453,
+      "learning_rate": 2.199250325945241e-06,
+      "loss": 0.1434,
+      "step": 99300
+    },
+    {
+      "epoch": 8.09973924380704,
+      "grad_norm": 5.676713466644287,
+      "learning_rate": 2.190286831812255e-06,
+      "loss": 0.1282,
+      "step": 99400
+    },
+    {
+      "epoch": 8.107887874837028,
+      "grad_norm": 3.881793260574341,
+      "learning_rate": 2.1813233376792705e-06,
+      "loss": 0.113,
+      "step": 99500
+    },
+    {
+      "epoch": 8.116036505867015,
+      "grad_norm": 18.958267211914062,
+      "learning_rate": 2.1723598435462844e-06,
+      "loss": 0.1462,
+      "step": 99600
+    },
+    {
+      "epoch": 8.124185136897001,
+      "grad_norm": 3.9751124382019043,
+      "learning_rate": 2.163396349413299e-06,
+      "loss": 0.1304,
+      "step": 99700
+    },
+    {
+      "epoch": 8.132333767926989,
+      "grad_norm": 20.424734115600586,
+      "learning_rate": 2.1544328552803128e-06,
+      "loss": 0.1509,
+      "step": 99800
+    },
+    {
+      "epoch": 8.140482398956975,
+      "grad_norm": 17.55602264404297,
+      "learning_rate": 2.145469361147327e-06,
+      "loss": 0.1458,
+      "step": 99900
+    },
+    {
+      "epoch": 8.148631029986962,
+      "grad_norm": 9.35766315460205,
+      "learning_rate": 2.1365058670143415e-06,
+      "loss": 0.1335,
+      "step": 100000
+    },
+    {
+      "epoch": 8.15677966101695,
+      "grad_norm": 0.5167970657348633,
+      "learning_rate": 2.1275423728813555e-06,
+      "loss": 0.1336,
+      "step": 100100
+    },
+    {
+      "epoch": 8.164928292046936,
+      "grad_norm": 9.043828964233398,
+      "learning_rate": 2.11857887874837e-06,
+      "loss": 0.1242,
+      "step": 100200
+    },
+    {
+      "epoch": 8.173076923076923,
+      "grad_norm": 8.647052764892578,
+      "learning_rate": 2.109615384615384e-06,
+      "loss": 0.131,
+      "step": 100300
+    },
+    {
+      "epoch": 8.18122555410691,
+      "grad_norm": 16.887643814086914,
+      "learning_rate": 2.1006518904823995e-06,
+      "loss": 0.1475,
+      "step": 100400
+    },
+    {
+      "epoch": 8.189374185136897,
+      "grad_norm": 24.23810386657715,
+      "learning_rate": 2.0916883963494134e-06,
+      "loss": 0.1256,
+      "step": 100500
+    },
+    {
+      "epoch": 8.197522816166884,
+      "grad_norm": 13.14070987701416,
+      "learning_rate": 2.082724902216428e-06,
+      "loss": 0.137,
+      "step": 100600
+    },
+    {
+      "epoch": 8.20567144719687,
+      "grad_norm": 6.199899673461914,
+      "learning_rate": 2.0737614080834418e-06,
+      "loss": 0.1363,
+      "step": 100700
+    },
+    {
+      "epoch": 8.213820078226858,
+      "grad_norm": 3.194542646408081,
+      "learning_rate": 2.064797913950456e-06,
+      "loss": 0.1584,
+      "step": 100800
+    },
+    {
+      "epoch": 8.221968709256846,
+      "grad_norm": 16.422868728637695,
+      "learning_rate": 2.05583441981747e-06,
+      "loss": 0.1464,
+      "step": 100900
+    },
+    {
+      "epoch": 8.230117340286832,
+      "grad_norm": 18.549264907836914,
+      "learning_rate": 2.0468709256844845e-06,
+      "loss": 0.1392,
+      "step": 101000
+    },
+    {
+      "epoch": 8.23826597131682,
+      "grad_norm": 3.2117178440093994,
+      "learning_rate": 2.037907431551499e-06,
+      "loss": 0.1431,
+      "step": 101100
+    },
+    {
+      "epoch": 8.246414602346805,
+      "grad_norm": 23.774818420410156,
+      "learning_rate": 2.028943937418514e-06,
+      "loss": 0.1308,
+      "step": 101200
+    },
+    {
+      "epoch": 8.254563233376793,
+      "grad_norm": 1.8825786113739014,
+      "learning_rate": 2.019980443285528e-06,
+      "loss": 0.1409,
+      "step": 101300
+    },
+    {
+      "epoch": 8.26271186440678,
+      "grad_norm": 4.661303520202637,
+      "learning_rate": 2.0110169491525424e-06,
+      "loss": 0.1357,
+      "step": 101400
+    },
+    {
+      "epoch": 8.270860495436766,
+      "grad_norm": 18.980335235595703,
+      "learning_rate": 2.002053455019557e-06,
+      "loss": 0.1446,
+      "step": 101500
+    },
+    {
+      "epoch": 8.279009126466754,
+      "grad_norm": 9.78714656829834,
+      "learning_rate": 1.9930899608865708e-06,
+      "loss": 0.1197,
+      "step": 101600
+    },
+    {
+      "epoch": 8.28715775749674,
+      "grad_norm": 9.437408447265625,
+      "learning_rate": 1.984126466753585e-06,
+      "loss": 0.1296,
+      "step": 101700
+    },
+    {
+      "epoch": 8.295306388526727,
+      "grad_norm": 11.14543628692627,
+      "learning_rate": 1.9751629726205995e-06,
+      "loss": 0.1402,
+      "step": 101800
+    },
+    {
+      "epoch": 8.303455019556715,
+      "grad_norm": 10.319045066833496,
+      "learning_rate": 1.9661994784876135e-06,
+      "loss": 0.1278,
+      "step": 101900
+    },
+    {
+      "epoch": 8.3116036505867,
+      "grad_norm": 20.33431053161621,
+      "learning_rate": 1.9572359843546287e-06,
+      "loss": 0.1667,
+      "step": 102000
+    },
+    {
+      "epoch": 8.319752281616688,
+      "grad_norm": 6.3067145347595215,
+      "learning_rate": 1.948272490221643e-06,
+      "loss": 0.1346,
+      "step": 102100
+    },
+    {
+      "epoch": 8.327900912646676,
+      "grad_norm": 15.225107192993164,
+      "learning_rate": 1.9393089960886575e-06,
+      "loss": 0.1464,
+      "step": 102200
+    },
+    {
+      "epoch": 8.336049543676662,
+      "grad_norm": 17.693836212158203,
+      "learning_rate": 1.9303455019556714e-06,
+      "loss": 0.1479,
+      "step": 102300
+    },
+    {
+      "epoch": 8.34419817470665,
+      "grad_norm": 28.224361419677734,
+      "learning_rate": 1.921382007822686e-06,
+      "loss": 0.1315,
+      "step": 102400
+    },
+    {
+      "epoch": 8.352346805736635,
+      "grad_norm": 1.448736548423767,
+      "learning_rate": 1.9124185136896998e-06,
+      "loss": 0.1433,
+      "step": 102500
+    },
+    {
+      "epoch": 8.360495436766623,
+      "grad_norm": 24.642520904541016,
+      "learning_rate": 1.9034550195567142e-06,
+      "loss": 0.1155,
+      "step": 102600
+    },
+    {
+      "epoch": 8.36864406779661,
+      "grad_norm": 18.223909378051758,
+      "learning_rate": 1.8944915254237281e-06,
+      "loss": 0.147,
+      "step": 102700
+    },
+    {
+      "epoch": 8.376792698826597,
+      "grad_norm": 14.59261417388916,
+      "learning_rate": 1.8855280312907438e-06,
+      "loss": 0.1158,
+      "step": 102800
+    },
+    {
+      "epoch": 8.384941329856584,
+      "grad_norm": 7.161222457885742,
+      "learning_rate": 1.8765645371577575e-06,
+      "loss": 0.1316,
+      "step": 102900
+    },
+    {
+      "epoch": 8.393089960886572,
+      "grad_norm": 6.764099597930908,
+      "learning_rate": 1.8676010430247721e-06,
+      "loss": 0.1377,
+      "step": 103000
+    },
+    {
+      "epoch": 8.401238591916558,
+      "grad_norm": 3.5225605964660645,
+      "learning_rate": 1.858637548891786e-06,
+      "loss": 0.1354,
+      "step": 103100
+    },
+    {
+      "epoch": 8.409387222946545,
+      "grad_norm": 28.121694564819336,
+      "learning_rate": 1.8496740547588004e-06,
+      "loss": 0.1351,
+      "step": 103200
+    },
+    {
+      "epoch": 8.417535853976531,
+      "grad_norm": 8.749202728271484,
+      "learning_rate": 1.840710560625815e-06,
+      "loss": 0.1318,
+      "step": 103300
+    },
+    {
+      "epoch": 8.425684485006519,
+      "grad_norm": 18.74785804748535,
+      "learning_rate": 1.8317470664928288e-06,
+      "loss": 0.1453,
+      "step": 103400
+    },
+    {
+      "epoch": 8.433833116036507,
+      "grad_norm": 29.5967960357666,
+      "learning_rate": 1.822783572359843e-06,
+      "loss": 0.1347,
+      "step": 103500
+    },
+    {
+      "epoch": 8.441981747066492,
+      "grad_norm": 13.819513320922852,
+      "learning_rate": 1.8138200782268584e-06,
+      "loss": 0.1453,
+      "step": 103600
+    },
+    {
+      "epoch": 8.45013037809648,
+      "grad_norm": 14.125680923461914,
+      "learning_rate": 1.8048565840938728e-06,
+      "loss": 0.1325,
+      "step": 103700
+    },
+    {
+      "epoch": 8.458279009126466,
+      "grad_norm": 20.32880973815918,
+      "learning_rate": 1.7958930899608867e-06,
+      "loss": 0.1489,
+      "step": 103800
+    },
+    {
+      "epoch": 8.466427640156454,
+      "grad_norm": 1.8261317014694214,
+      "learning_rate": 1.7869295958279011e-06,
+      "loss": 0.1209,
+      "step": 103900
+    },
+    {
+      "epoch": 8.474576271186441,
+      "grad_norm": 18.03254508972168,
+      "learning_rate": 1.777966101694915e-06,
+      "loss": 0.1549,
+      "step": 104000
+    },
+    {
+      "epoch": 8.482724902216427,
+      "grad_norm": 14.922330856323242,
+      "learning_rate": 1.7690026075619292e-06,
+      "loss": 0.1532,
+      "step": 104100
+    },
+    {
+      "epoch": 8.490873533246415,
+      "grad_norm": 8.974449157714844,
+      "learning_rate": 1.7600391134289436e-06,
+      "loss": 0.1596,
+      "step": 104200
+    },
+    {
+      "epoch": 8.499022164276402,
+      "grad_norm": 13.620256423950195,
+      "learning_rate": 1.7510756192959578e-06,
+      "loss": 0.1236,
+      "step": 104300
+    },
+    {
+      "epoch": 8.507170795306388,
+      "grad_norm": 9.534696578979492,
+      "learning_rate": 1.742112125162973e-06,
+      "loss": 0.1421,
+      "step": 104400
+    },
+    {
+      "epoch": 8.515319426336376,
+      "grad_norm": 12.891791343688965,
+      "learning_rate": 1.7331486310299874e-06,
+      "loss": 0.1395,
+      "step": 104500
+    },
+    {
+      "epoch": 8.523468057366362,
+      "grad_norm": 0.4351995587348938,
+      "learning_rate": 1.7241851368970013e-06,
+      "loss": 0.1237,
+      "step": 104600
+    },
+    {
+      "epoch": 8.53161668839635,
+      "grad_norm": 7.187576770782471,
+      "learning_rate": 1.7152216427640155e-06,
+      "loss": 0.1346,
+      "step": 104700
+    },
+    {
+      "epoch": 8.539765319426337,
+      "grad_norm": 8.591465950012207,
+      "learning_rate": 1.7062581486310301e-06,
+      "loss": 0.1565,
+      "step": 104800
+    },
+    {
+      "epoch": 8.547913950456323,
+      "grad_norm": 17.407093048095703,
+      "learning_rate": 1.697294654498044e-06,
+      "loss": 0.1426,
+      "step": 104900
+    },
+    {
+      "epoch": 8.55606258148631,
+      "grad_norm": 11.454471588134766,
+      "learning_rate": 1.6883311603650584e-06,
+      "loss": 0.1436,
+      "step": 105000
+    },
+    {
+      "epoch": 8.564211212516298,
+      "grad_norm": 21.674543380737305,
+      "learning_rate": 1.6793676662320724e-06,
+      "loss": 0.1309,
+      "step": 105100
+    },
+    {
+      "epoch": 8.572359843546284,
+      "grad_norm": 16.86087417602539,
+      "learning_rate": 1.670404172099088e-06,
+      "loss": 0.1348,
+      "step": 105200
+    },
+    {
+      "epoch": 8.580508474576272,
+      "grad_norm": 9.330849647521973,
+      "learning_rate": 1.6614406779661022e-06,
+      "loss": 0.1409,
+      "step": 105300
+    },
+    {
+      "epoch": 8.588657105606258,
+      "grad_norm": 1.9600093364715576,
+      "learning_rate": 1.6524771838331164e-06,
+      "loss": 0.1344,
+      "step": 105400
+    },
+    {
+      "epoch": 8.596805736636245,
+      "grad_norm": 3.9729743003845215,
+      "learning_rate": 1.6435136897001303e-06,
+      "loss": 0.1456,
+      "step": 105500
+    },
+    {
+      "epoch": 8.604954367666233,
+      "grad_norm": 13.668156623840332,
+      "learning_rate": 1.6345501955671447e-06,
+      "loss": 0.1293,
+      "step": 105600
+    },
+    {
+      "epoch": 8.613102998696219,
+      "grad_norm": 13.970529556274414,
+      "learning_rate": 1.6255867014341587e-06,
+      "loss": 0.1527,
+      "step": 105700
+    },
+    {
+      "epoch": 8.621251629726206,
+      "grad_norm": 8.949882507324219,
+      "learning_rate": 1.616623207301173e-06,
+      "loss": 0.1363,
+      "step": 105800
+    },
+    {
+      "epoch": 8.629400260756192,
+      "grad_norm": 15.386001586914062,
+      "learning_rate": 1.6076597131681872e-06,
+      "loss": 0.1318,
+      "step": 105900
+    },
+    {
+      "epoch": 8.63754889178618,
+      "grad_norm": 11.052765846252441,
+      "learning_rate": 1.5986962190352027e-06,
+      "loss": 0.1407,
+      "step": 106000
+    },
+    {
+      "epoch": 8.645697522816167,
+      "grad_norm": 21.316442489624023,
+      "learning_rate": 1.5897327249022166e-06,
+      "loss": 0.1389,
+      "step": 106100
+    },
+    {
+      "epoch": 8.653846153846153,
+      "grad_norm": 14.646417617797852,
+      "learning_rate": 1.580769230769231e-06,
+      "loss": 0.1378,
+      "step": 106200
+    },
+    {
+      "epoch": 8.661994784876141,
+      "grad_norm": 20.964698791503906,
+      "learning_rate": 1.5718057366362454e-06,
+      "loss": 0.1391,
+      "step": 106300
+    },
+    {
+      "epoch": 8.670143415906129,
+      "grad_norm": 8.166078567504883,
+      "learning_rate": 1.5628422425032593e-06,
+      "loss": 0.1551,
+      "step": 106400
+    },
+    {
+      "epoch": 8.678292046936114,
+      "grad_norm": 20.075531005859375,
+      "learning_rate": 1.553878748370274e-06,
+      "loss": 0.1457,
+      "step": 106500
+    },
+    {
+      "epoch": 8.686440677966102,
+      "grad_norm": 19.019393920898438,
+      "learning_rate": 1.5449152542372877e-06,
+      "loss": 0.1311,
+      "step": 106600
+    },
+    {
+      "epoch": 8.694589308996088,
+      "grad_norm": 7.855512619018555,
+      "learning_rate": 1.535951760104302e-06,
+      "loss": 0.134,
+      "step": 106700
+    },
+    {
+      "epoch": 8.702737940026076,
+      "grad_norm": 15.56972885131836,
+      "learning_rate": 1.5269882659713173e-06,
+      "loss": 0.1322,
+      "step": 106800
+    },
+    {
+      "epoch": 8.710886571056063,
+      "grad_norm": 7.912954807281494,
+      "learning_rate": 1.5180247718383317e-06,
+      "loss": 0.1264,
+      "step": 106900
+    },
+    {
+      "epoch": 8.719035202086049,
+      "grad_norm": 12.312883377075195,
+      "learning_rate": 1.5090612777053456e-06,
+      "loss": 0.1357,
+      "step": 107000
+    },
+    {
+      "epoch": 8.727183833116037,
+      "grad_norm": 7.040205001831055,
+      "learning_rate": 1.5000977835723602e-06,
+      "loss": 0.1411,
+      "step": 107100
+    },
+    {
+      "epoch": 8.735332464146023,
+      "grad_norm": 16.882474899291992,
+      "learning_rate": 1.491134289439374e-06,
+      "loss": 0.1258,
+      "step": 107200
+    },
+    {
+      "epoch": 8.74348109517601,
+      "grad_norm": 13.230895042419434,
+      "learning_rate": 1.4821707953063883e-06,
+      "loss": 0.1371,
+      "step": 107300
+    },
+    {
+      "epoch": 8.751629726205998,
+      "grad_norm": 1.2307300567626953,
+      "learning_rate": 1.4732073011734027e-06,
+      "loss": 0.1303,
+      "step": 107400
+    },
+    {
+      "epoch": 8.759778357235984,
+      "grad_norm": 2.0919406414031982,
+      "learning_rate": 1.4642438070404167e-06,
+      "loss": 0.1358,
+      "step": 107500
+    },
+    {
+      "epoch": 8.767926988265971,
+      "grad_norm": 9.384042739868164,
+      "learning_rate": 1.455280312907432e-06,
+      "loss": 0.1434,
+      "step": 107600
+    },
+    {
+      "epoch": 8.776075619295959,
+      "grad_norm": 3.808473587036133,
+      "learning_rate": 1.4463168187744465e-06,
+      "loss": 0.1337,
+      "step": 107700
+    },
+    {
+      "epoch": 8.784224250325945,
+      "grad_norm": 9.139209747314453,
+      "learning_rate": 1.4373533246414607e-06,
+      "loss": 0.1237,
+      "step": 107800
+    },
+    {
+      "epoch": 8.792372881355933,
+      "grad_norm": 23.689847946166992,
+      "learning_rate": 1.4283898305084744e-06,
+      "loss": 0.1506,
+      "step": 107900
+    },
+    {
+      "epoch": 8.800521512385918,
+      "grad_norm": 17.127538681030273,
+      "learning_rate": 1.419426336375489e-06,
+      "loss": 0.1308,
+      "step": 108000
+    },
+    {
+      "epoch": 8.808670143415906,
+      "grad_norm": 10.87045669555664,
+      "learning_rate": 1.410462842242503e-06,
+      "loss": 0.1366,
+      "step": 108100
+    },
+    {
+      "epoch": 8.816818774445894,
+      "grad_norm": 8.058269500732422,
+      "learning_rate": 1.4014993481095174e-06,
+      "loss": 0.122,
+      "step": 108200
+    },
+    {
+      "epoch": 8.82496740547588,
+      "grad_norm": 30.747976303100586,
+      "learning_rate": 1.392535853976532e-06,
+      "loss": 0.1476,
+      "step": 108300
+    },
+    {
+      "epoch": 8.833116036505867,
+      "grad_norm": 2.1181864738464355,
+      "learning_rate": 1.383572359843547e-06,
+      "loss": 0.1389,
+      "step": 108400
+    },
+    {
+      "epoch": 8.841264667535853,
+      "grad_norm": 16.122636795043945,
+      "learning_rate": 1.3746088657105607e-06,
+      "loss": 0.1494,
+      "step": 108500
+    },
+    {
+      "epoch": 8.84941329856584,
+      "grad_norm": 1.7896504402160645,
+      "learning_rate": 1.3656453715775753e-06,
+      "loss": 0.1429,
+      "step": 108600
+    },
+    {
+      "epoch": 8.857561929595828,
+      "grad_norm": 14.50307559967041,
+      "learning_rate": 1.3566818774445893e-06,
+      "loss": 0.1391,
+      "step": 108700
+    },
+    {
+      "epoch": 8.865710560625814,
+      "grad_norm": 12.862863540649414,
+      "learning_rate": 1.3477183833116036e-06,
+      "loss": 0.1449,
+      "step": 108800
+    },
+    {
+      "epoch": 8.873859191655802,
+      "grad_norm": 4.946181774139404,
+      "learning_rate": 1.3387548891786182e-06,
+      "loss": 0.1547,
+      "step": 108900
+    },
+    {
+      "epoch": 8.88200782268579,
+      "grad_norm": 10.275555610656738,
+      "learning_rate": 1.329791395045632e-06,
+      "loss": 0.1231,
+      "step": 109000
+    },
+    {
+      "epoch": 8.890156453715775,
+      "grad_norm": 12.205680847167969,
+      "learning_rate": 1.3208279009126461e-06,
+      "loss": 0.1273,
+      "step": 109100
+    },
+    {
+      "epoch": 8.898305084745763,
+      "grad_norm": 9.211130142211914,
+      "learning_rate": 1.3118644067796605e-06,
+      "loss": 0.1376,
+      "step": 109200
+    },
+    {
+      "epoch": 8.906453715775749,
+      "grad_norm": 3.3791849613189697,
+      "learning_rate": 1.302900912646676e-06,
+      "loss": 0.1403,
+      "step": 109300
+    },
+    {
+      "epoch": 8.914602346805736,
+      "grad_norm": 1.7787587642669678,
+      "learning_rate": 1.29393741851369e-06,
+      "loss": 0.1497,
+      "step": 109400
+    },
+    {
+      "epoch": 8.922750977835724,
+      "grad_norm": 19.89299964904785,
+      "learning_rate": 1.2849739243807043e-06,
+      "loss": 0.1231,
+      "step": 109500
+    },
+    {
+      "epoch": 8.93089960886571,
+      "grad_norm": 1.886210322380066,
+      "learning_rate": 1.2760104302477183e-06,
+      "loss": 0.1336,
+      "step": 109600
+    },
+    {
+      "epoch": 8.939048239895698,
+      "grad_norm": 8.365442276000977,
+      "learning_rate": 1.2670469361147324e-06,
+      "loss": 0.1502,
+      "step": 109700
+    },
+    {
+      "epoch": 8.947196870925685,
+      "grad_norm": 7.866243839263916,
+      "learning_rate": 1.258083441981747e-06,
+      "loss": 0.1369,
+      "step": 109800
+    },
+    {
+      "epoch": 8.955345501955671,
+      "grad_norm": 13.674020767211914,
+      "learning_rate": 1.249119947848761e-06,
+      "loss": 0.1232,
+      "step": 109900
+    },
+    {
+      "epoch": 8.963494132985659,
+      "grad_norm": 10.142361640930176,
+      "learning_rate": 1.2401564537157754e-06,
+      "loss": 0.123,
+      "step": 110000
+    },
+    {
+      "epoch": 8.971642764015645,
+      "grad_norm": 9.111577987670898,
+      "learning_rate": 1.2311929595827906e-06,
+      "loss": 0.1184,
+      "step": 110100
+    },
+    {
+      "epoch": 8.979791395045632,
+      "grad_norm": 5.888847827911377,
+      "learning_rate": 1.2222294654498045e-06,
+      "loss": 0.1449,
+      "step": 110200
+    },
+    {
+      "epoch": 8.98794002607562,
+      "grad_norm": 11.367358207702637,
+      "learning_rate": 1.213265971316819e-06,
+      "loss": 0.1326,
+      "step": 110300
+    },
+    {
+      "epoch": 8.996088657105606,
+      "grad_norm": 7.464927673339844,
+      "learning_rate": 1.2043024771838333e-06,
+      "loss": 0.1315,
+      "step": 110400
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.8365461847389558,
+      "eval_loss": 0.7411776185035706,
+      "eval_runtime": 5.509,
+      "eval_samples_per_second": 451.985,
+      "eval_steps_per_second": 56.634,
+      "step": 110448
+    },
+    {
+      "epoch": 9.004237288135593,
+      "grad_norm": 15.54743766784668,
+      "learning_rate": 1.1953389830508473e-06,
+      "loss": 0.133,
+      "step": 110500
+    },
+    {
+      "epoch": 9.01238591916558,
+      "grad_norm": 2.1653244495391846,
+      "learning_rate": 1.1863754889178616e-06,
+      "loss": 0.1328,
+      "step": 110600
+    },
+    {
+      "epoch": 9.020534550195567,
+      "grad_norm": 0.2608337104320526,
+      "learning_rate": 1.1774119947848756e-06,
+      "loss": 0.1013,
+      "step": 110700
+    },
+    {
+      "epoch": 9.028683181225555,
+      "grad_norm": 4.152002334594727,
+      "learning_rate": 1.16844850065189e-06,
+      "loss": 0.1265,
+      "step": 110800
+    },
+    {
+      "epoch": 9.03683181225554,
+      "grad_norm": 6.718245029449463,
+      "learning_rate": 1.1594850065189052e-06,
+      "loss": 0.1255,
+      "step": 110900
+    },
+    {
+      "epoch": 9.044980443285528,
+      "grad_norm": 19.890954971313477,
+      "learning_rate": 1.1505215123859196e-06,
+      "loss": 0.1221,
+      "step": 111000
+    },
+    {
+      "epoch": 9.053129074315516,
+      "grad_norm": 11.731401443481445,
+      "learning_rate": 1.1415580182529335e-06,
+      "loss": 0.127,
+      "step": 111100
+    },
+    {
+      "epoch": 9.061277705345502,
+      "grad_norm": 11.024049758911133,
+      "learning_rate": 1.132594524119948e-06,
+      "loss": 0.1247,
+      "step": 111200
+    },
+    {
+      "epoch": 9.06942633637549,
+      "grad_norm": 20.952821731567383,
+      "learning_rate": 1.1236310299869623e-06,
+      "loss": 0.1189,
+      "step": 111300
+    },
+    {
+      "epoch": 9.077574967405475,
+      "grad_norm": 20.7772216796875,
+      "learning_rate": 1.1146675358539763e-06,
+      "loss": 0.1381,
+      "step": 111400
+    },
+    {
+      "epoch": 9.085723598435463,
+      "grad_norm": 6.1934027671813965,
+      "learning_rate": 1.1057040417209906e-06,
+      "loss": 0.134,
+      "step": 111500
+    },
+    {
+      "epoch": 9.09387222946545,
+      "grad_norm": 10.067231178283691,
+      "learning_rate": 1.0967405475880046e-06,
+      "loss": 0.1136,
+      "step": 111600
+    },
+    {
+      "epoch": 9.102020860495436,
+      "grad_norm": 7.726376533508301,
+      "learning_rate": 1.0877770534550198e-06,
+      "loss": 0.1192,
+      "step": 111700
+    },
+    {
+      "epoch": 9.110169491525424,
+      "grad_norm": 11.030654907226562,
+      "learning_rate": 1.0788135593220342e-06,
+      "loss": 0.1129,
+      "step": 111800
+    },
+    {
+      "epoch": 9.118318122555412,
+      "grad_norm": 0.4144195020198822,
+      "learning_rate": 1.0698500651890486e-06,
+      "loss": 0.1309,
+      "step": 111900
+    },
+    {
+      "epoch": 9.126466753585397,
+      "grad_norm": 2.212282180786133,
+      "learning_rate": 1.0608865710560625e-06,
+      "loss": 0.1237,
+      "step": 112000
+    },
+    {
+      "epoch": 9.134615384615385,
+      "grad_norm": 8.832420349121094,
+      "learning_rate": 1.051923076923077e-06,
+      "loss": 0.1244,
+      "step": 112100
+    },
+    {
+      "epoch": 9.142764015645371,
+      "grad_norm": 4.031465530395508,
+      "learning_rate": 1.042959582790091e-06,
+      "loss": 0.1374,
+      "step": 112200
+    },
+    {
+      "epoch": 9.150912646675359,
+      "grad_norm": 4.740699768066406,
+      "learning_rate": 1.0339960886571053e-06,
+      "loss": 0.1416,
+      "step": 112300
+    },
+    {
+      "epoch": 9.159061277705346,
+      "grad_norm": 8.297629356384277,
+      "learning_rate": 1.0250325945241194e-06,
+      "loss": 0.1216,
+      "step": 112400
+    },
+    {
+      "epoch": 9.167209908735332,
+      "grad_norm": 20.47978401184082,
+      "learning_rate": 1.0160691003911349e-06,
+      "loss": 0.1124,
+      "step": 112500
+    },
+    {
+      "epoch": 9.17535853976532,
+      "grad_norm": 6.031740665435791,
+      "learning_rate": 1.0071056062581488e-06,
+      "loss": 0.1074,
+      "step": 112600
+    },
+    {
+      "epoch": 9.183507170795306,
+      "grad_norm": 14.264923095703125,
+      "learning_rate": 9.981421121251632e-07,
+      "loss": 0.1197,
+      "step": 112700
+    },
+    {
+      "epoch": 9.191655801825293,
+      "grad_norm": 11.424922943115234,
+      "learning_rate": 9.891786179921774e-07,
+      "loss": 0.132,
+      "step": 112800
+    },
+    {
+      "epoch": 9.19980443285528,
+      "grad_norm": 8.677164077758789,
+      "learning_rate": 9.802151238591915e-07,
+      "loss": 0.1338,
+      "step": 112900
+    },
+    {
+      "epoch": 9.207953063885267,
+      "grad_norm": 12.700013160705566,
+      "learning_rate": 9.71251629726206e-07,
+      "loss": 0.1312,
+      "step": 113000
+    },
+    {
+      "epoch": 9.216101694915254,
+      "grad_norm": 4.308102607727051,
+      "learning_rate": 9.622881355932199e-07,
+      "loss": 0.1241,
+      "step": 113100
+    },
+    {
+      "epoch": 9.224250325945242,
+      "grad_norm": 2.213650703430176,
+      "learning_rate": 9.533246414602343e-07,
+      "loss": 0.1362,
+      "step": 113200
+    },
+    {
+      "epoch": 9.232398956975228,
+      "grad_norm": 30.006113052368164,
+      "learning_rate": 9.443611473272495e-07,
+      "loss": 0.129,
+      "step": 113300
+    },
+    {
+      "epoch": 9.240547588005215,
+      "grad_norm": 9.764779090881348,
+      "learning_rate": 9.353976531942638e-07,
+      "loss": 0.1289,
+      "step": 113400
+    },
+    {
+      "epoch": 9.248696219035201,
+      "grad_norm": 3.6789746284484863,
+      "learning_rate": 9.26434159061278e-07,
+      "loss": 0.1192,
+      "step": 113500
+    },
+    {
+      "epoch": 9.256844850065189,
+      "grad_norm": 11.824788093566895,
+      "learning_rate": 9.17470664928292e-07,
+      "loss": 0.1263,
+      "step": 113600
+    },
+    {
+      "epoch": 9.264993481095177,
+      "grad_norm": 8.798393249511719,
+      "learning_rate": 9.085071707953064e-07,
+      "loss": 0.1135,
+      "step": 113700
+    },
+    {
+      "epoch": 9.273142112125162,
+      "grad_norm": 2.8183655738830566,
+      "learning_rate": 8.995436766623205e-07,
+      "loss": 0.1376,
+      "step": 113800
+    },
+    {
+      "epoch": 9.28129074315515,
+      "grad_norm": 3.4520905017852783,
+      "learning_rate": 8.905801825293347e-07,
+      "loss": 0.1314,
+      "step": 113900
+    },
+    {
+      "epoch": 9.289439374185136,
+      "grad_norm": 10.228925704956055,
+      "learning_rate": 8.81616688396349e-07,
+      "loss": 0.1292,
+      "step": 114000
+    },
+    {
+      "epoch": 9.297588005215124,
+      "grad_norm": 0.910640299320221,
+      "learning_rate": 8.726531942633642e-07,
+      "loss": 0.1238,
+      "step": 114100
+    },
+    {
+      "epoch": 9.305736636245111,
+      "grad_norm": 1.382246732711792,
+      "learning_rate": 8.636897001303785e-07,
+      "loss": 0.1037,
+      "step": 114200
+    },
+    {
+      "epoch": 9.313885267275097,
+      "grad_norm": 23.052778244018555,
+      "learning_rate": 8.547262059973927e-07,
+      "loss": 0.1124,
+      "step": 114300
+    },
+    {
+      "epoch": 9.322033898305085,
+      "grad_norm": 33.121097564697266,
+      "learning_rate": 8.457627118644068e-07,
+      "loss": 0.1403,
+      "step": 114400
+    },
+    {
+      "epoch": 9.330182529335072,
+      "grad_norm": 13.370987892150879,
+      "learning_rate": 8.36799217731421e-07,
+      "loss": 0.1258,
+      "step": 114500
+    },
+    {
+      "epoch": 9.338331160365058,
+      "grad_norm": 21.812013626098633,
+      "learning_rate": 8.278357235984353e-07,
+      "loss": 0.1288,
+      "step": 114600
+    },
+    {
+      "epoch": 9.346479791395046,
+      "grad_norm": 2.375351667404175,
+      "learning_rate": 8.188722294654495e-07,
+      "loss": 0.1349,
+      "step": 114700
+    },
+    {
+      "epoch": 9.354628422425032,
+      "grad_norm": 14.761313438415527,
+      "learning_rate": 8.099087353324637e-07,
+      "loss": 0.1334,
+      "step": 114800
+    },
+    {
+      "epoch": 9.36277705345502,
+      "grad_norm": 14.228364944458008,
+      "learning_rate": 8.00945241199479e-07,
+      "loss": 0.1239,
+      "step": 114900
+    },
+    {
+      "epoch": 9.370925684485007,
+      "grad_norm": 7.746068954467773,
+      "learning_rate": 7.919817470664932e-07,
+      "loss": 0.1049,
+      "step": 115000
+    },
+    {
+      "epoch": 9.379074315514993,
+      "grad_norm": 0.5430285334587097,
+      "learning_rate": 7.830182529335075e-07,
+      "loss": 0.1293,
+      "step": 115100
+    },
+    {
+      "epoch": 9.38722294654498,
+      "grad_norm": 9.228418350219727,
+      "learning_rate": 7.740547588005217e-07,
+      "loss": 0.1234,
+      "step": 115200
+    },
+    {
+      "epoch": 9.395371577574968,
+      "grad_norm": 22.232402801513672,
+      "learning_rate": 7.650912646675358e-07,
+      "loss": 0.141,
+      "step": 115300
+    },
+    {
+      "epoch": 9.403520208604954,
+      "grad_norm": 2.0123260021209717,
+      "learning_rate": 7.5612777053455e-07,
+      "loss": 0.1063,
+      "step": 115400
+    },
+    {
+      "epoch": 9.411668839634942,
+      "grad_norm": 0.9207943081855774,
+      "learning_rate": 7.471642764015643e-07,
+      "loss": 0.1377,
+      "step": 115500
+    },
+    {
+      "epoch": 9.419817470664928,
+      "grad_norm": 7.680292129516602,
+      "learning_rate": 7.382007822685784e-07,
+      "loss": 0.1131,
+      "step": 115600
+    },
+    {
+      "epoch": 9.427966101694915,
+      "grad_norm": 10.872476577758789,
+      "learning_rate": 7.292372881355938e-07,
+      "loss": 0.1266,
+      "step": 115700
+    },
+    {
+      "epoch": 9.436114732724903,
+      "grad_norm": 18.4665584564209,
+      "learning_rate": 7.202737940026079e-07,
+      "loss": 0.131,
+      "step": 115800
+    },
+    {
+      "epoch": 9.444263363754889,
+      "grad_norm": 23.102298736572266,
+      "learning_rate": 7.113102998696221e-07,
+      "loss": 0.1099,
+      "step": 115900
+    },
+    {
+      "epoch": 9.452411994784876,
+      "grad_norm": 1.6242234706878662,
+      "learning_rate": 7.023468057366365e-07,
+      "loss": 0.1231,
+      "step": 116000
+    },
+    {
+      "epoch": 9.460560625814864,
+      "grad_norm": 8.302433013916016,
+      "learning_rate": 6.933833116036504e-07,
+      "loss": 0.1121,
+      "step": 116100
+    },
+    {
+      "epoch": 9.46870925684485,
+      "grad_norm": 25.818700790405273,
+      "learning_rate": 6.844198174706648e-07,
+      "loss": 0.1175,
+      "step": 116200
+    },
+    {
+      "epoch": 9.476857887874838,
+      "grad_norm": 18.826948165893555,
+      "learning_rate": 6.75456323337679e-07,
+      "loss": 0.1205,
+      "step": 116300
+    },
+    {
+      "epoch": 9.485006518904823,
+      "grad_norm": 0.6583182215690613,
+      "learning_rate": 6.664928292046933e-07,
+      "loss": 0.1358,
+      "step": 116400
+    },
+    {
+      "epoch": 9.493155149934811,
+      "grad_norm": 12.578688621520996,
+      "learning_rate": 6.575293350717085e-07,
+      "loss": 0.1096,
+      "step": 116500
+    },
+    {
+      "epoch": 9.501303780964799,
+      "grad_norm": 13.54981803894043,
+      "learning_rate": 6.485658409387227e-07,
+      "loss": 0.1368,
+      "step": 116600
+    },
+    {
+      "epoch": 9.509452411994785,
+      "grad_norm": 6.809188365936279,
+      "learning_rate": 6.396023468057369e-07,
+      "loss": 0.1188,
+      "step": 116700
+    },
+    {
+      "epoch": 9.517601043024772,
+      "grad_norm": 11.08338737487793,
+      "learning_rate": 6.306388526727511e-07,
+      "loss": 0.1203,
+      "step": 116800
+    },
+    {
+      "epoch": 9.525749674054758,
+      "grad_norm": 4.479588031768799,
+      "learning_rate": 6.216753585397653e-07,
+      "loss": 0.1297,
+      "step": 116900
+    },
+    {
+      "epoch": 9.533898305084746,
+      "grad_norm": 7.722095966339111,
+      "learning_rate": 6.127118644067795e-07,
+      "loss": 0.1192,
+      "step": 117000
+    },
+    {
+      "epoch": 9.542046936114733,
+      "grad_norm": 10.053470611572266,
+      "learning_rate": 6.037483702737937e-07,
+      "loss": 0.1222,
+      "step": 117100
+    },
+    {
+      "epoch": 9.55019556714472,
+      "grad_norm": 12.395706176757812,
+      "learning_rate": 5.947848761408079e-07,
+      "loss": 0.1207,
+      "step": 117200
+    },
+    {
+      "epoch": 9.558344198174707,
+      "grad_norm": 7.065868854522705,
+      "learning_rate": 5.858213820078232e-07,
+      "loss": 0.1262,
+      "step": 117300
+    },
+    {
+      "epoch": 9.566492829204694,
+      "grad_norm": 5.896761417388916,
+      "learning_rate": 5.768578878748374e-07,
+      "loss": 0.1258,
+      "step": 117400
+    },
+    {
+      "epoch": 9.57464146023468,
+      "grad_norm": 11.613407135009766,
+      "learning_rate": 5.678943937418517e-07,
+      "loss": 0.1234,
+      "step": 117500
+    },
+    {
+      "epoch": 9.582790091264668,
+      "grad_norm": 20.267986297607422,
+      "learning_rate": 5.589308996088658e-07,
+      "loss": 0.1136,
+      "step": 117600
+    },
+    {
+      "epoch": 9.590938722294654,
+      "grad_norm": 7.217312335968018,
+      "learning_rate": 5.499674054758801e-07,
+      "loss": 0.1125,
+      "step": 117700
+    },
+    {
+      "epoch": 9.599087353324641,
+      "grad_norm": 13.285292625427246,
+      "learning_rate": 5.410039113428943e-07,
+      "loss": 0.1157,
+      "step": 117800
+    },
+    {
+      "epoch": 9.607235984354629,
+      "grad_norm": 22.7592716217041,
+      "learning_rate": 5.320404172099084e-07,
+      "loss": 0.1219,
+      "step": 117900
+    },
+    {
+      "epoch": 9.615384615384615,
+      "grad_norm": 17.77228546142578,
+      "learning_rate": 5.230769230769226e-07,
+      "loss": 0.1423,
+      "step": 118000
+    },
+    {
+      "epoch": 9.623533246414603,
+      "grad_norm": 35.28404235839844,
+      "learning_rate": 5.141134289439369e-07,
+      "loss": 0.1254,
+      "step": 118100
+    },
+    {
+      "epoch": 9.631681877444588,
+      "grad_norm": 5.466562747955322,
+      "learning_rate": 5.051499348109522e-07,
+      "loss": 0.1481,
+      "step": 118200
+    },
+    {
+      "epoch": 9.639830508474576,
+      "grad_norm": 14.727246284484863,
+      "learning_rate": 4.961864406779664e-07,
+      "loss": 0.1369,
+      "step": 118300
+    },
+    {
+      "epoch": 9.647979139504564,
+      "grad_norm": 15.746825218200684,
+      "learning_rate": 4.872229465449806e-07,
+      "loss": 0.1176,
+      "step": 118400
+    },
+    {
+      "epoch": 9.65612777053455,
+      "grad_norm": 19.238407135009766,
+      "learning_rate": 4.782594524119947e-07,
+      "loss": 0.1178,
+      "step": 118500
+    },
+    {
+      "epoch": 9.664276401564537,
+      "grad_norm": 13.003064155578613,
+      "learning_rate": 4.69295958279009e-07,
+      "loss": 0.1273,
+      "step": 118600
+    },
+    {
+      "epoch": 9.672425032594525,
+      "grad_norm": 5.110761642456055,
+      "learning_rate": 4.603324641460232e-07,
+      "loss": 0.1242,
+      "step": 118700
+    },
+    {
+      "epoch": 9.68057366362451,
+      "grad_norm": 12.598255157470703,
+      "learning_rate": 4.513689700130374e-07,
+      "loss": 0.1186,
+      "step": 118800
+    },
+    {
+      "epoch": 9.688722294654498,
+      "grad_norm": 7.440487384796143,
+      "learning_rate": 4.424054758800516e-07,
+      "loss": 0.1145,
+      "step": 118900
+    },
+    {
+      "epoch": 9.696870925684484,
+      "grad_norm": 10.29443073272705,
+      "learning_rate": 4.33441981747067e-07,
+      "loss": 0.1179,
+      "step": 119000
+    },
+    {
+      "epoch": 9.705019556714472,
+      "grad_norm": 0.5482966899871826,
+      "learning_rate": 4.244784876140811e-07,
+      "loss": 0.1214,
+      "step": 119100
+    },
+    {
+      "epoch": 9.71316818774446,
+      "grad_norm": 8.483821868896484,
+      "learning_rate": 4.1551499348109534e-07,
+      "loss": 0.1443,
+      "step": 119200
+    },
+    {
+      "epoch": 9.721316818774445,
+      "grad_norm": 12.351343154907227,
+      "learning_rate": 4.0655149934810956e-07,
+      "loss": 0.13,
+      "step": 119300
+    },
+    {
+      "epoch": 9.729465449804433,
+      "grad_norm": 11.550556182861328,
+      "learning_rate": 3.9758800521512373e-07,
+      "loss": 0.1387,
+      "step": 119400
+    },
+    {
+      "epoch": 9.737614080834419,
+      "grad_norm": 7.8005690574646,
+      "learning_rate": 3.8862451108213795e-07,
+      "loss": 0.1144,
+      "step": 119500
+    },
+    {
+      "epoch": 9.745762711864407,
+      "grad_norm": 1.0686675310134888,
+      "learning_rate": 3.796610169491521e-07,
+      "loss": 0.1166,
+      "step": 119600
+    },
+    {
+      "epoch": 9.753911342894394,
+      "grad_norm": 27.589384078979492,
+      "learning_rate": 3.706975228161664e-07,
+      "loss": 0.1322,
+      "step": 119700
+    },
+    {
+      "epoch": 9.76205997392438,
+      "grad_norm": 4.429134845733643,
+      "learning_rate": 3.6173402868318173e-07,
+      "loss": 0.1331,
+      "step": 119800
+    },
+    {
+      "epoch": 9.770208604954368,
+      "grad_norm": 24.076099395751953,
+      "learning_rate": 3.527705345501959e-07,
+      "loss": 0.1159,
+      "step": 119900
+    },
+    {
+      "epoch": 9.778357235984355,
+      "grad_norm": 11.259610176086426,
+      "learning_rate": 3.438070404172101e-07,
+      "loss": 0.1492,
+      "step": 120000
+    },
+    {
+      "epoch": 9.786505867014341,
+      "grad_norm": 4.419936656951904,
+      "learning_rate": 3.348435462842243e-07,
+      "loss": 0.1174,
+      "step": 120100
+    },
+    {
+      "epoch": 9.794654498044329,
+      "grad_norm": 16.751590728759766,
+      "learning_rate": 3.258800521512385e-07,
+      "loss": 0.1293,
+      "step": 120200
+    },
+    {
+      "epoch": 9.802803129074315,
+      "grad_norm": 21.935205459594727,
+      "learning_rate": 3.1691655801825273e-07,
+      "loss": 0.1432,
+      "step": 120300
+    },
+    {
+      "epoch": 9.810951760104302,
+      "grad_norm": 8.330902099609375,
+      "learning_rate": 3.079530638852669e-07,
+      "loss": 0.1162,
+      "step": 120400
+    },
+    {
+      "epoch": 9.81910039113429,
+      "grad_norm": 2.7674975395202637,
+      "learning_rate": 2.989895697522811e-07,
+      "loss": 0.1237,
+      "step": 120500
+    },
+    {
+      "epoch": 9.827249022164276,
+      "grad_norm": 13.894972801208496,
+      "learning_rate": 2.900260756192964e-07,
+      "loss": 0.1323,
+      "step": 120600
+    },
+    {
+      "epoch": 9.835397653194264,
+      "grad_norm": 15.896739959716797,
+      "learning_rate": 2.810625814863106e-07,
+      "loss": 0.1213,
+      "step": 120700
+    },
+    {
+      "epoch": 9.843546284224251,
+      "grad_norm": 14.188883781433105,
+      "learning_rate": 2.7209908735332484e-07,
+      "loss": 0.1383,
+      "step": 120800
+    },
+    {
+      "epoch": 9.851694915254237,
+      "grad_norm": 17.167747497558594,
+      "learning_rate": 2.63135593220339e-07,
+      "loss": 0.1198,
+      "step": 120900
+    },
+    {
+      "epoch": 9.859843546284225,
+      "grad_norm": 18.93288803100586,
+      "learning_rate": 2.5417209908735323e-07,
+      "loss": 0.1223,
+      "step": 121000
+    },
+    {
+      "epoch": 9.86799217731421,
+      "grad_norm": 8.871918678283691,
+      "learning_rate": 2.4520860495436746e-07,
+      "loss": 0.1253,
+      "step": 121100
+    },
+    {
+      "epoch": 9.876140808344198,
+      "grad_norm": 23.355209350585938,
+      "learning_rate": 2.3624511082138162e-07,
+      "loss": 0.1208,
+      "step": 121200
+    },
+    {
+      "epoch": 9.884289439374186,
+      "grad_norm": 8.497678756713867,
+      "learning_rate": 2.2728161668839582e-07,
+      "loss": 0.1173,
+      "step": 121300
+    },
+    {
+      "epoch": 9.892438070404172,
+      "grad_norm": 7.700555801391602,
+      "learning_rate": 2.1831812255541118e-07,
+      "loss": 0.1193,
+      "step": 121400
+    },
+    {
+      "epoch": 9.90058670143416,
+      "grad_norm": 7.455496788024902,
+      "learning_rate": 2.0935462842242537e-07,
+      "loss": 0.1155,
+      "step": 121500
+    },
+    {
+      "epoch": 9.908735332464147,
+      "grad_norm": 24.733013153076172,
+      "learning_rate": 2.0039113428943957e-07,
+      "loss": 0.1084,
+      "step": 121600
+    },
+    {
+      "epoch": 9.916883963494133,
+      "grad_norm": 8.922455787658691,
+      "learning_rate": 1.914276401564538e-07,
+      "loss": 0.1094,
+      "step": 121700
+    },
+    {
+      "epoch": 9.92503259452412,
+      "grad_norm": 9.614680290222168,
+      "learning_rate": 1.8246414602346799e-07,
+      "loss": 0.1221,
+      "step": 121800
+    },
+    {
+      "epoch": 9.933181225554106,
+      "grad_norm": 16.748014450073242,
+      "learning_rate": 1.7350065189048218e-07,
+      "loss": 0.1274,
+      "step": 121900
+    },
+    {
+      "epoch": 9.941329856584094,
+      "grad_norm": 5.290912628173828,
+      "learning_rate": 1.645371577574964e-07,
+      "loss": 0.1304,
+      "step": 122000
+    },
+    {
+      "epoch": 9.949478487614082,
+      "grad_norm": 16.15484619140625,
+      "learning_rate": 1.555736636245106e-07,
+      "loss": 0.1335,
+      "step": 122100
+    },
+    {
+      "epoch": 9.957627118644067,
+      "grad_norm": 12.658509254455566,
+      "learning_rate": 1.466101694915259e-07,
+      "loss": 0.1219,
+      "step": 122200
+    },
+    {
+      "epoch": 9.965775749674055,
+      "grad_norm": 12.304189682006836,
+      "learning_rate": 1.3764667535854013e-07,
+      "loss": 0.1119,
+      "step": 122300
+    },
+    {
+      "epoch": 9.973924380704041,
+      "grad_norm": 21.67201805114746,
+      "learning_rate": 1.2868318122555432e-07,
+      "loss": 0.135,
+      "step": 122400
+    },
+    {
+      "epoch": 9.982073011734029,
+      "grad_norm": 18.46645164489746,
+      "learning_rate": 1.1971968709256852e-07,
+      "loss": 0.1222,
+      "step": 122500
+    },
+    {
+      "epoch": 9.990221642764016,
+      "grad_norm": 8.396200180053711,
+      "learning_rate": 1.1075619295958273e-07,
+      "loss": 0.1213,
+      "step": 122600
+    },
+    {
+      "epoch": 9.998370273794002,
+      "grad_norm": 15.41945743560791,
+      "learning_rate": 1.0179269882659693e-07,
+      "loss": 0.1102,
+      "step": 122700
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.8369477911646587,
+      "eval_loss": 0.7886391282081604,
+      "eval_runtime": 5.5306,
+      "eval_samples_per_second": 450.222,
+      "eval_steps_per_second": 56.413,
+      "step": 122720
+    },
+    {
+      "epoch": 10.0,
+      "step": 122720,
+      "total_flos": 1.6169329248859574e+17,
+      "train_loss": 0.2765254594356482,
+      "train_runtime": 31314.1287,
+      "train_samples_per_second": 125.407,
+      "train_steps_per_second": 3.919
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 122720,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 10,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.6169329248859574e+17,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}