End of training

README.md

@@ -23,7 +23,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.692
+      value: 0.7025
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -33,8 +33,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [microsoft/swin-tiny-patch4-window7-224](https://huggingface.co/microsoft/swin-tiny-patch4-window7-224) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8708
-- Accuracy: 0.692
+- Loss: 0.8305
+- Accuracy: 0.7025
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 9.984,
-    "eval_accuracy": 0.80625,
-    "eval_loss": 0.662419319152832,
-    "eval_runtime": 157.5998,
-    "eval_samples_per_second": 25.381,
-    "eval_steps_per_second": 0.793,
-    "total_flos": 9.926487761391452e+18,
-    "train_loss": 0.24441122439427254,
-    "train_runtime": 137433.0815,
-    "train_samples_per_second": 2.911,
+    "epoch": 2.9952,
+    "eval_accuracy": 0.7025,
+    "eval_loss": 0.830498456954956,
+    "eval_runtime": 156.0429,
+    "eval_samples_per_second": 25.634,
+    "eval_steps_per_second": 0.801,
+    "total_flos": 2.9779463284174356e+18,
+    "train_loss": 0.36125070786374247,
+    "train_runtime": 41238.7802,
+    "train_samples_per_second": 2.91,
     "train_steps_per_second": 0.023
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.984,
-    "eval_accuracy": 0.80625,
-    "eval_loss": 0.662419319152832,
-    "eval_runtime": 157.5998,
-    "eval_samples_per_second": 25.381,
-    "eval_steps_per_second": 0.793
+    "epoch": 2.9952,
+    "eval_accuracy": 0.7025,
+    "eval_loss": 0.830498456954956,
+    "eval_runtime": 156.0429,
+    "eval_samples_per_second": 25.634,
+    "eval_steps_per_second": 0.801
 }

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 9.984,
-    "total_flos": 9.926487761391452e+18,
-    "train_loss": 0.24441122439427254,
-    "train_runtime": 137433.0815,
-    "train_samples_per_second": 2.911,
+    "epoch": 2.9952,
+    "total_flos": 2.9779463284174356e+18,
+    "train_loss": 0.36125070786374247,
+    "train_runtime": 41238.7802,
+    "train_samples_per_second": 2.91,
     "train_steps_per_second": 0.023
 }

trainer_state.json

@@ -1,2301 +1,705 @@
 {
-  "best_metric": 0.80625,
-  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-leukemia-08-2024.v1.2\\checkpoint-2500",
-  "epoch": 9.984,
+  "best_metric": 0.7025,
+  "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-leukemia-08-2024.v1.2\\checkpoint-625",
+  "epoch": 2.9952,
   "eval_steps": 500,
-  "global_step": 3120,
+  "global_step": 936,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.032,
-      "grad_norm": 5.175816059112549,
-      "learning_rate": 1.6025641025641025e-06,
-      "loss": 0.701,
+      "grad_norm": 5.205770969390869,
+      "learning_rate": 5.319148936170213e-06,
+      "loss": 0.6922,
       "step": 10
     },
     {
       "epoch": 0.064,
-      "grad_norm": 4.15721321105957,
-      "learning_rate": 3.205128205128205e-06,
-      "loss": 0.6915,
+      "grad_norm": 5.621664524078369,
+      "learning_rate": 1.0638297872340426e-05,
+      "loss": 0.6735,
       "step": 20
     },
     {
       "epoch": 0.096,
-      "grad_norm": 3.4217724800109863,
-      "learning_rate": 4.807692307692308e-06,
-      "loss": 0.6905,
+      "grad_norm": 4.5251078605651855,
+      "learning_rate": 1.595744680851064e-05,
+      "loss": 0.6799,
       "step": 30
     },
     {
       "epoch": 0.128,
-      "grad_norm": 5.180055141448975,
-      "learning_rate": 6.41025641025641e-06,
-      "loss": 0.6768,
+      "grad_norm": 6.694261074066162,
+      "learning_rate": 2.1276595744680852e-05,
+      "loss": 0.6703,
       "step": 40
     },
     {
       "epoch": 0.16,
-      "grad_norm": 5.26149320602417,
-      "learning_rate": 8.012820512820515e-06,
-      "loss": 0.6593,
+      "grad_norm": 5.9614176750183105,
+      "learning_rate": 2.6595744680851064e-05,
+      "loss": 0.6335,
       "step": 50
     },
     {
       "epoch": 0.192,
-      "grad_norm": 9.300067901611328,
-      "learning_rate": 9.615384615384616e-06,
-      "loss": 0.6601,
+      "grad_norm": 9.891090393066406,
+      "learning_rate": 3.191489361702128e-05,
+      "loss": 0.6442,
       "step": 60
     },
     {
       "epoch": 0.224,
-      "grad_norm": 7.7086100578308105,
-      "learning_rate": 1.1217948717948719e-05,
-      "loss": 0.6399,
+      "grad_norm": 5.152803897857666,
+      "learning_rate": 3.723404255319149e-05,
+      "loss": 0.6192,
       "step": 70
     },
     {
       "epoch": 0.256,
-      "grad_norm": 9.903676986694336,
-      "learning_rate": 1.282051282051282e-05,
-      "loss": 0.6239,
+      "grad_norm": 19.81842613220215,
+      "learning_rate": 4.2553191489361704e-05,
+      "loss": 0.6061,
       "step": 80
     },
     {
       "epoch": 0.288,
-      "grad_norm": 8.374137878417969,
-      "learning_rate": 1.4423076923076923e-05,
-      "loss": 0.6283,
+      "grad_norm": 12.223305702209473,
+      "learning_rate": 4.787234042553192e-05,
+      "loss": 0.5728,
       "step": 90
     },
     {
       "epoch": 0.32,
-      "grad_norm": 13.180610656738281,
-      "learning_rate": 1.602564102564103e-05,
-      "loss": 0.6182,
+      "grad_norm": 9.165103912353516,
+      "learning_rate": 4.96437054631829e-05,
+      "loss": 0.532,
       "step": 100
     },
     {
       "epoch": 0.352,
-      "grad_norm": 16.700641632080078,
-      "learning_rate": 1.762820512820513e-05,
-      "loss": 0.6064,
+      "grad_norm": 11.49010181427002,
+      "learning_rate": 4.90498812351544e-05,
+      "loss": 0.5192,
       "step": 110
     },
     {
       "epoch": 0.384,
-      "grad_norm": 7.996766567230225,
-      "learning_rate": 1.923076923076923e-05,
-      "loss": 0.6113,
+      "grad_norm": 16.154022216796875,
+      "learning_rate": 4.851543942992874e-05,
+      "loss": 0.5428,
       "step": 120
     },
     {
       "epoch": 0.416,
-      "grad_norm": 11.852957725524902,
-      "learning_rate": 2.0833333333333336e-05,
-      "loss": 0.5819,
+      "grad_norm": 15.223276138305664,
+      "learning_rate": 4.792161520190024e-05,
+      "loss": 0.4874,
       "step": 130
     },
     {
       "epoch": 0.448,
-      "grad_norm": 14.05556869506836,
-      "learning_rate": 2.2435897435897437e-05,
-      "loss": 0.5803,
+      "grad_norm": 15.795281410217285,
+      "learning_rate": 4.732779097387174e-05,
+      "loss": 0.4917,
       "step": 140
     },
     {
       "epoch": 0.48,
-      "grad_norm": 9.655508041381836,
-      "learning_rate": 2.4038461538461542e-05,
-      "loss": 0.5554,
+      "grad_norm": 15.048002243041992,
+      "learning_rate": 4.673396674584323e-05,
+      "loss": 0.4914,
       "step": 150
     },
     {
       "epoch": 0.512,
-      "grad_norm": 8.562996864318848,
-      "learning_rate": 2.564102564102564e-05,
-      "loss": 0.5555,
+      "grad_norm": 9.113037109375,
+      "learning_rate": 4.6140142517814724e-05,
+      "loss": 0.4794,
       "step": 160
     },
     {
       "epoch": 0.544,
-      "grad_norm": 25.228897094726562,
-      "learning_rate": 2.724358974358974e-05,
-      "loss": 0.5462,
+      "grad_norm": 8.940505027770996,
+      "learning_rate": 4.5546318289786225e-05,
+      "loss": 0.467,
       "step": 170
     },
     {
       "epoch": 0.576,
-      "grad_norm": 28.388383865356445,
-      "learning_rate": 2.8685897435897437e-05,
-      "loss": 0.591,
+      "grad_norm": 10.99445915222168,
+      "learning_rate": 4.501187648456057e-05,
+      "loss": 0.4666,
       "step": 180
     },
     {
       "epoch": 0.608,
-      "grad_norm": 25.528270721435547,
-      "learning_rate": 3.0288461538461538e-05,
-      "loss": 0.5561,
+      "grad_norm": 11.823775291442871,
+      "learning_rate": 4.441805225653207e-05,
+      "loss": 0.4256,
       "step": 190
     },
     {
       "epoch": 0.64,
-      "grad_norm": 9.540146827697754,
-      "learning_rate": 3.189102564102564e-05,
-      "loss": 0.5586,
+      "grad_norm": 32.22935485839844,
+      "learning_rate": 4.382422802850357e-05,
+      "loss": 0.412,
       "step": 200
     },
     {
       "epoch": 0.672,
-      "grad_norm": 12.264636993408203,
-      "learning_rate": 3.3493589743589744e-05,
-      "loss": 0.5514,
+      "grad_norm": 31.271358489990234,
+      "learning_rate": 4.323040380047506e-05,
+      "loss": 0.493,
       "step": 210
     },
     {
       "epoch": 0.704,
-      "grad_norm": 6.773402214050293,
-      "learning_rate": 3.5096153846153845e-05,
-      "loss": 0.5146,
+      "grad_norm": 7.7791900634765625,
+      "learning_rate": 4.263657957244656e-05,
+      "loss": 0.4073,
       "step": 220
     },
     {
       "epoch": 0.736,
-      "grad_norm": 11.84838581085205,
-      "learning_rate": 3.6698717948717946e-05,
-      "loss": 0.474,
+      "grad_norm": 10.995684623718262,
+      "learning_rate": 4.204275534441806e-05,
+      "loss": 0.3982,
       "step": 230
     },
     {
       "epoch": 0.768,
-      "grad_norm": 7.139883041381836,
-      "learning_rate": 3.8301282051282054e-05,
-      "loss": 0.461,
+      "grad_norm": 17.266887664794922,
+      "learning_rate": 4.144893111638955e-05,
+      "loss": 0.3794,
       "step": 240
     },
     {
       "epoch": 0.8,
-      "grad_norm": 7.588615417480469,
-      "learning_rate": 3.9903846153846155e-05,
-      "loss": 0.4871,
+      "grad_norm": 7.906798362731934,
+      "learning_rate": 4.0855106888361044e-05,
+      "loss": 0.3911,
       "step": 250
     },
     {
       "epoch": 0.832,
-      "grad_norm": 10.481608390808105,
-      "learning_rate": 4.150641025641026e-05,
-      "loss": 0.4536,
+      "grad_norm": 13.777398109436035,
+      "learning_rate": 4.0261282660332545e-05,
+      "loss": 0.3829,
       "step": 260
     },
     {
       "epoch": 0.864,
-      "grad_norm": 19.465951919555664,
-      "learning_rate": 4.3108974358974364e-05,
-      "loss": 0.4818,
+      "grad_norm": 15.481444358825684,
+      "learning_rate": 3.966745843230404e-05,
+      "loss": 0.3937,
       "step": 270
     },
     {
       "epoch": 0.896,
-      "grad_norm": 13.416147232055664,
-      "learning_rate": 4.4711538461538466e-05,
-      "loss": 0.433,
+      "grad_norm": 11.324766159057617,
+      "learning_rate": 3.907363420427554e-05,
+      "loss": 0.3503,
       "step": 280
     },
     {
       "epoch": 0.928,
-      "grad_norm": 8.518133163452148,
-      "learning_rate": 4.615384615384616e-05,
-      "loss": 0.4145,
+      "grad_norm": 19.757225036621094,
+      "learning_rate": 3.847980997624703e-05,
+      "loss": 0.3484,
       "step": 290
     },
     {
       "epoch": 0.96,
-      "grad_norm": 53.73661804199219,
-      "learning_rate": 4.775641025641026e-05,
-      "loss": 0.5544,
+      "grad_norm": 9.243185997009277,
+      "learning_rate": 3.7885985748218526e-05,
+      "loss": 0.3803,
       "step": 300
     },
     {
       "epoch": 0.992,
-      "grad_norm": 12.88679313659668,
-      "learning_rate": 4.935897435897436e-05,
-      "loss": 0.471,
+      "grad_norm": 9.08983325958252,
+      "learning_rate": 3.7292161520190026e-05,
+      "loss": 0.3704,
       "step": 310
     },
     {
       "epoch": 0.9984,
-      "eval_accuracy": 0.6715,
-      "eval_loss": 0.590698778629303,
-      "eval_runtime": 168.7555,
-      "eval_samples_per_second": 23.703,
-      "eval_steps_per_second": 0.741,
+      "eval_accuracy": 0.63875,
+      "eval_loss": 0.7880752682685852,
+      "eval_runtime": 187.0148,
+      "eval_samples_per_second": 21.389,
+      "eval_steps_per_second": 0.668,
       "step": 312
     },
     {
       "epoch": 1.024,
-      "grad_norm": 17.80354881286621,
-      "learning_rate": 4.98931623931624e-05,
-      "loss": 0.5001,
+      "grad_norm": 45.11425018310547,
+      "learning_rate": 3.669833729216152e-05,
+      "loss": 0.3855,
       "step": 320
     },
     {
       "epoch": 1.056,
-      "grad_norm": 9.920671463012695,
-      "learning_rate": 4.971509971509972e-05,
-      "loss": 0.4407,
+      "grad_norm": 26.53282928466797,
+      "learning_rate": 3.6104513064133013e-05,
+      "loss": 0.3922,
       "step": 330
     },
     {
       "epoch": 1.088,
-      "grad_norm": 15.50145149230957,
-      "learning_rate": 4.9537037037037035e-05,
-      "loss": 0.416,
+      "grad_norm": 9.646315574645996,
+      "learning_rate": 3.5510688836104514e-05,
+      "loss": 0.4238,
       "step": 340
     },
     {
       "epoch": 1.12,
-      "grad_norm": 11.008417129516602,
-      "learning_rate": 4.935897435897436e-05,
-      "loss": 0.4409,
+      "grad_norm": 21.654006958007812,
+      "learning_rate": 3.4916864608076014e-05,
+      "loss": 0.3569,
       "step": 350
     },
     {
       "epoch": 1.152,
-      "grad_norm": 17.966873168945312,
-      "learning_rate": 4.9180911680911686e-05,
-      "loss": 0.4154,
+      "grad_norm": 10.348204612731934,
+      "learning_rate": 3.432304038004751e-05,
+      "loss": 0.3122,
       "step": 360
     },
     {
       "epoch": 1.184,
-      "grad_norm": 35.332679748535156,
-      "learning_rate": 4.9002849002849004e-05,
-      "loss": 0.4375,
+      "grad_norm": 9.6500883102417,
+      "learning_rate": 3.372921615201901e-05,
+      "loss": 0.327,
       "step": 370
     },
     {
       "epoch": 1.216,
-      "grad_norm": 13.41420841217041,
-      "learning_rate": 4.882478632478633e-05,
-      "loss": 0.4239,
+      "grad_norm": 16.52385902404785,
+      "learning_rate": 3.31353919239905e-05,
+      "loss": 0.4007,
       "step": 380
     },
     {
       "epoch": 1.248,
-      "grad_norm": 42.17304611206055,
-      "learning_rate": 4.864672364672365e-05,
-      "loss": 0.4071,
+      "grad_norm": 13.382709503173828,
+      "learning_rate": 3.2541567695961995e-05,
+      "loss": 0.3414,
       "step": 390
     },
     {
       "epoch": 1.28,
-      "grad_norm": 30.35808753967285,
-      "learning_rate": 4.846866096866097e-05,
-      "loss": 0.3735,
+      "grad_norm": 58.999324798583984,
+      "learning_rate": 3.1947743467933496e-05,
+      "loss": 0.3501,
       "step": 400
     },
     {
       "epoch": 1.312,
-      "grad_norm": 24.77705955505371,
-      "learning_rate": 4.829059829059829e-05,
-      "loss": 0.4135,
+      "grad_norm": 7.981963157653809,
+      "learning_rate": 3.135391923990499e-05,
+      "loss": 0.3822,
       "step": 410
     },
     {
       "epoch": 1.3439999999999999,
-      "grad_norm": 11.592647552490234,
-      "learning_rate": 4.8112535612535616e-05,
-      "loss": 0.3879,
+      "grad_norm": 9.790070533752441,
+      "learning_rate": 3.076009501187649e-05,
+      "loss": 0.3383,
       "step": 420
     },
     {
       "epoch": 1.376,
-      "grad_norm": 13.36685562133789,
-      "learning_rate": 4.7934472934472934e-05,
-      "loss": 0.383,
+      "grad_norm": 8.631843566894531,
+      "learning_rate": 3.0166270783847983e-05,
+      "loss": 0.335,
       "step": 430
     },
     {
       "epoch": 1.408,
-      "grad_norm": 9.121360778808594,
-      "learning_rate": 4.775641025641026e-05,
-      "loss": 0.3852,
+      "grad_norm": 9.060861587524414,
+      "learning_rate": 2.9572446555819477e-05,
+      "loss": 0.3282,
       "step": 440
     },
     {
       "epoch": 1.44,
-      "grad_norm": 20.881765365600586,
-      "learning_rate": 4.7578347578347584e-05,
-      "loss": 0.3584,
+      "grad_norm": 15.991608619689941,
+      "learning_rate": 2.8978622327790977e-05,
+      "loss": 0.3037,
       "step": 450
     },
     {
       "epoch": 1.472,
-      "grad_norm": 41.44031524658203,
-      "learning_rate": 4.74002849002849e-05,
-      "loss": 0.4047,
+      "grad_norm": 35.242740631103516,
+      "learning_rate": 2.838479809976247e-05,
+      "loss": 0.3321,
       "step": 460
     },
     {
       "epoch": 1.504,
-      "grad_norm": 6.26457405090332,
-      "learning_rate": 4.722222222222222e-05,
-      "loss": 0.3623,
+      "grad_norm": 10.073174476623535,
+      "learning_rate": 2.7790973871733968e-05,
+      "loss": 0.3023,
       "step": 470
     },
     {
       "epoch": 1.536,
-      "grad_norm": 18.60125732421875,
-      "learning_rate": 4.7044159544159546e-05,
-      "loss": 0.3609,
+      "grad_norm": 19.67376708984375,
+      "learning_rate": 2.7197149643705465e-05,
+      "loss": 0.3168,
       "step": 480
     },
     {
       "epoch": 1.568,
-      "grad_norm": 17.937673568725586,
-      "learning_rate": 4.686609686609687e-05,
-      "loss": 0.3832,
+      "grad_norm": 6.408135890960693,
+      "learning_rate": 2.6603325415676962e-05,
+      "loss": 0.306,
       "step": 490
     },
     {
       "epoch": 1.6,
-      "grad_norm": 11.792247772216797,
-      "learning_rate": 4.668803418803419e-05,
-      "loss": 0.3499,
+      "grad_norm": 8.863042831420898,
+      "learning_rate": 2.6009501187648455e-05,
+      "loss": 0.3103,
       "step": 500
     },
     {
       "epoch": 1.6320000000000001,
-      "grad_norm": 35.20759201049805,
-      "learning_rate": 4.6509971509971515e-05,
-      "loss": 0.3703,
+      "grad_norm": 17.778520584106445,
+      "learning_rate": 2.5415676959619956e-05,
+      "loss": 0.3123,
       "step": 510
     },
     {
       "epoch": 1.6640000000000001,
-      "grad_norm": 9.141999244689941,
-      "learning_rate": 4.633190883190883e-05,
-      "loss": 0.3433,
+      "grad_norm": 11.152641296386719,
+      "learning_rate": 2.482185273159145e-05,
+      "loss": 0.304,
       "step": 520
     },
     {
       "epoch": 1.696,
-      "grad_norm": 9.520153999328613,
-      "learning_rate": 4.615384615384616e-05,
-      "loss": 0.3638,
+      "grad_norm": 18.749303817749023,
+      "learning_rate": 2.4228028503562946e-05,
+      "loss": 0.2905,
       "step": 530
     },
     {
       "epoch": 1.728,
-      "grad_norm": 20.68021011352539,
-      "learning_rate": 4.5975783475783476e-05,
-      "loss": 0.4226,
+      "grad_norm": 26.219440460205078,
+      "learning_rate": 2.3634204275534443e-05,
+      "loss": 0.3163,
       "step": 540
     },
     {
       "epoch": 1.76,
-      "grad_norm": Infinity,
-      "learning_rate": 4.581552706552707e-05,
-      "loss": 0.3537,
+      "grad_norm": 16.212156295776367,
+      "learning_rate": 2.3040380047505937e-05,
+      "loss": 0.3073,
       "step": 550
     },
     {
       "epoch": 1.792,
-      "grad_norm": 29.779680252075195,
-      "learning_rate": 4.563746438746439e-05,
-      "loss": 0.4041,
+      "grad_norm": 11.084125518798828,
+      "learning_rate": 2.2446555819477437e-05,
+      "loss": 0.3166,
       "step": 560
     },
     {
       "epoch": 1.8239999999999998,
-      "grad_norm": 36.394161224365234,
-      "learning_rate": 4.545940170940171e-05,
-      "loss": 0.3585,
+      "grad_norm": 12.52534008026123,
+      "learning_rate": 2.1852731591448934e-05,
+      "loss": 0.293,
       "step": 570
     },
     {
       "epoch": 1.8559999999999999,
-      "grad_norm": 27.704442977905273,
-      "learning_rate": 4.528133903133903e-05,
-      "loss": 0.3484,
+      "grad_norm": 20.346576690673828,
+      "learning_rate": 2.1258907363420428e-05,
+      "loss": 0.2948,
       "step": 580
     },
     {
       "epoch": 1.888,
-      "grad_norm": 9.653308868408203,
-      "learning_rate": 4.510327635327636e-05,
-      "loss": 0.3234,
+      "grad_norm": 9.408613204956055,
+      "learning_rate": 2.0665083135391925e-05,
+      "loss": 0.3081,
       "step": 590
     },
     {
       "epoch": 1.92,
-      "grad_norm": 15.85473346710205,
-      "learning_rate": 4.492521367521368e-05,
-      "loss": 0.3486,
+      "grad_norm": 27.265464782714844,
+      "learning_rate": 2.0071258907363422e-05,
+      "loss": 0.3084,
       "step": 600
     },
     {
       "epoch": 1.952,
-      "grad_norm": 18.47919464111328,
-      "learning_rate": 4.4747150997151e-05,
-      "loss": 0.3643,
+      "grad_norm": 8.094114303588867,
+      "learning_rate": 1.947743467933492e-05,
+      "loss": 0.3047,
       "step": 610
     },
     {
       "epoch": 1.984,
-      "grad_norm": 25.203697204589844,
-      "learning_rate": 4.456908831908832e-05,
-      "loss": 0.3376,
+      "grad_norm": 13.893214225769043,
+      "learning_rate": 1.8883610451306412e-05,
+      "loss": 0.2726,
       "step": 620
     },
     {
       "epoch": 2.0,
-      "eval_accuracy": 0.702,
-      "eval_loss": 0.8903971910476685,
-      "eval_runtime": 155.8146,
-      "eval_samples_per_second": 25.672,
-      "eval_steps_per_second": 0.802,
+      "eval_accuracy": 0.7025,
+      "eval_loss": 0.830498456954956,
+      "eval_runtime": 154.4758,
+      "eval_samples_per_second": 25.894,
+      "eval_steps_per_second": 0.809,
       "step": 625
     },
     {
       "epoch": 2.016,
-      "grad_norm": 15.359530448913574,
-      "learning_rate": 4.439102564102564e-05,
-      "loss": 0.3328,
+      "grad_norm": 5.93212890625,
+      "learning_rate": 1.828978622327791e-05,
+      "loss": 0.2716,
       "step": 630
     },
     {
       "epoch": 2.048,
-      "grad_norm": 32.16645812988281,
-      "learning_rate": 4.4212962962962966e-05,
-      "loss": 0.3114,
+      "grad_norm": 33.18698501586914,
+      "learning_rate": 1.7695961995249406e-05,
+      "loss": 0.2845,
       "step": 640
     },
     {
       "epoch": 2.08,
-      "grad_norm": 8.913246154785156,
-      "learning_rate": 4.4034900284900285e-05,
-      "loss": 0.3503,
+      "grad_norm": 23.357540130615234,
+      "learning_rate": 1.7102137767220903e-05,
+      "loss": 0.3093,
       "step": 650
     },
     {
       "epoch": 2.112,
-      "grad_norm": 33.13675308227539,
-      "learning_rate": 4.385683760683761e-05,
-      "loss": 0.3231,
+      "grad_norm": 14.696449279785156,
+      "learning_rate": 1.65083135391924e-05,
+      "loss": 0.2886,
       "step": 660
     },
     {
       "epoch": 2.144,
-      "grad_norm": 24.989564895629883,
-      "learning_rate": 4.367877492877493e-05,
-      "loss": 0.299,
+      "grad_norm": 32.7164192199707,
+      "learning_rate": 1.5914489311163897e-05,
+      "loss": 0.2865,
       "step": 670
     },
     {
       "epoch": 2.176,
-      "grad_norm": 19.014232635498047,
-      "learning_rate": 4.350071225071225e-05,
-      "loss": 0.3708,
+      "grad_norm": 11.33267879486084,
+      "learning_rate": 1.5320665083135394e-05,
+      "loss": 0.2616,
       "step": 680
     },
     {
       "epoch": 2.208,
-      "grad_norm": 12.924874305725098,
-      "learning_rate": 4.332264957264958e-05,
-      "loss": 0.2897,
+      "grad_norm": 10.119343757629395,
+      "learning_rate": 1.4726840855106888e-05,
+      "loss": 0.2768,
       "step": 690
     },
     {
       "epoch": 2.24,
-      "grad_norm": 8.708645820617676,
-      "learning_rate": 4.31445868945869e-05,
-      "loss": 0.2907,
+      "grad_norm": 12.684237480163574,
+      "learning_rate": 1.4133016627078385e-05,
+      "loss": 0.2332,
       "step": 700
     },
     {
       "epoch": 2.2720000000000002,
-      "grad_norm": 23.476911544799805,
-      "learning_rate": 4.2966524216524215e-05,
-      "loss": 0.3338,
+      "grad_norm": 17.10093879699707,
+      "learning_rate": 1.3539192399049882e-05,
+      "loss": 0.2648,
       "step": 710
     },
     {
       "epoch": 2.304,
-      "grad_norm": 12.567790985107422,
-      "learning_rate": 4.278846153846154e-05,
-      "loss": 0.3332,
+      "grad_norm": 22.309677124023438,
+      "learning_rate": 1.2945368171021377e-05,
+      "loss": 0.2794,
       "step": 720
     },
     {
       "epoch": 2.336,
-      "grad_norm": 9.736689567565918,
-      "learning_rate": 4.2610398860398865e-05,
-      "loss": 0.3076,
+      "grad_norm": 15.592386245727539,
+      "learning_rate": 1.2351543942992874e-05,
+      "loss": 0.2826,
       "step": 730
     },
     {
       "epoch": 2.368,
-      "grad_norm": 8.446788787841797,
-      "learning_rate": 4.2432336182336184e-05,
-      "loss": 0.3052,
+      "grad_norm": 10.027396202087402,
+      "learning_rate": 1.1757719714964371e-05,
+      "loss": 0.2662,
       "step": 740
     },
     {
       "epoch": 2.4,
-      "grad_norm": 24.329153060913086,
-      "learning_rate": 4.225427350427351e-05,
-      "loss": 0.3356,
+      "grad_norm": 7.71236515045166,
+      "learning_rate": 1.1163895486935868e-05,
+      "loss": 0.2802,
       "step": 750
     },
     {
       "epoch": 2.432,
-      "grad_norm": 6.862959861755371,
-      "learning_rate": 4.207621082621083e-05,
-      "loss": 0.381,
+      "grad_norm": 8.326642036437988,
+      "learning_rate": 1.0570071258907365e-05,
+      "loss": 0.2644,
       "step": 760
     },
     {
       "epoch": 2.464,
-      "grad_norm": 12.4033842086792,
-      "learning_rate": 4.1898148148148145e-05,
-      "loss": 0.2917,
+      "grad_norm": 17.820472717285156,
+      "learning_rate": 9.97624703087886e-06,
+      "loss": 0.2649,
       "step": 770
     },
     {
       "epoch": 2.496,
-      "grad_norm": 11.243486404418945,
-      "learning_rate": 4.172008547008547e-05,
-      "loss": 0.2766,
+      "grad_norm": 10.056536674499512,
+      "learning_rate": 9.382422802850356e-06,
+      "loss": 0.248,
       "step": 780
     },
     {
       "epoch": 2.528,
-      "grad_norm": 28.551401138305664,
-      "learning_rate": 4.1542022792022796e-05,
-      "loss": 0.2965,
+      "grad_norm": 30.569965362548828,
+      "learning_rate": 8.788598574821852e-06,
+      "loss": 0.2303,
       "step": 790
     },
     {
       "epoch": 2.56,
-      "grad_norm": 7.9521484375,
-      "learning_rate": 4.1363960113960114e-05,
-      "loss": 0.2912,
+      "grad_norm": 7.8709797859191895,
+      "learning_rate": 8.19477434679335e-06,
+      "loss": 0.2537,
       "step": 800
     },
     {
       "epoch": 2.592,
-      "grad_norm": 14.33103084564209,
-      "learning_rate": 4.118589743589744e-05,
-      "loss": 0.3085,
+      "grad_norm": 15.642313957214355,
+      "learning_rate": 7.6009501187648464e-06,
+      "loss": 0.2726,
       "step": 810
     },
     {
       "epoch": 2.624,
-      "grad_norm": 14.015169143676758,
-      "learning_rate": 4.1007834757834764e-05,
-      "loss": 0.2754,
+      "grad_norm": 8.542279243469238,
+      "learning_rate": 7.007125890736342e-06,
+      "loss": 0.27,
       "step": 820
     },
     {
       "epoch": 2.656,
-      "grad_norm": 27.55406379699707,
-      "learning_rate": 4.082977207977208e-05,
-      "loss": 0.2825,
+      "grad_norm": 7.526516914367676,
+      "learning_rate": 6.4133016627078396e-06,
+      "loss": 0.2742,
       "step": 830
     },
     {
       "epoch": 2.6879999999999997,
-      "grad_norm": 24.08679962158203,
-      "learning_rate": 4.06517094017094e-05,
-      "loss": 0.3177,
+      "grad_norm": 9.183687210083008,
+      "learning_rate": 5.819477434679335e-06,
+      "loss": 0.2477,
       "step": 840
     },
     {
       "epoch": 2.7199999999999998,
-      "grad_norm": 15.628003120422363,
-      "learning_rate": 4.0473646723646726e-05,
-      "loss": 0.289,
+      "grad_norm": 13.24605655670166,
+      "learning_rate": 5.225653206650832e-06,
+      "loss": 0.2462,
       "step": 850
     },
     {
       "epoch": 2.752,
-      "grad_norm": 33.23855972290039,
-      "learning_rate": 4.0295584045584044e-05,
-      "loss": 0.2952,
+      "grad_norm": 14.802075386047363,
+      "learning_rate": 4.631828978622328e-06,
+      "loss": 0.2451,
       "step": 860
     },
     {
       "epoch": 2.784,
-      "grad_norm": 28.434730529785156,
-      "learning_rate": 4.011752136752137e-05,
-      "loss": 0.2608,
+      "grad_norm": 8.528770446777344,
+      "learning_rate": 4.038004750593825e-06,
+      "loss": 0.2221,
       "step": 870
     },
     {
       "epoch": 2.816,
-      "grad_norm": 12.023883819580078,
-      "learning_rate": 3.9939458689458694e-05,
-      "loss": 0.2821,
+      "grad_norm": 12.269448280334473,
+      "learning_rate": 3.4441805225653207e-06,
+      "loss": 0.2522,
       "step": 880
     },
     {
       "epoch": 2.848,
-      "grad_norm": 18.148351669311523,
-      "learning_rate": 3.976139601139601e-05,
-      "loss": 0.2998,
+      "grad_norm": 30.11684226989746,
+      "learning_rate": 2.850356294536817e-06,
+      "loss": 0.2509,
       "step": 890
     },
     {
       "epoch": 2.88,
-      "grad_norm": 52.181968688964844,
-      "learning_rate": 3.958333333333333e-05,
-      "loss": 0.3114,
+      "grad_norm": 25.639724731445312,
+      "learning_rate": 2.2565320665083133e-06,
+      "loss": 0.2423,
       "step": 900
     },
     {
       "epoch": 2.912,
-      "grad_norm": 39.47223663330078,
-      "learning_rate": 3.940527065527066e-05,
-      "loss": 0.3137,
+      "grad_norm": 10.527947425842285,
+      "learning_rate": 1.6627078384798101e-06,
+      "loss": 0.2528,
       "step": 910
     },
     {
       "epoch": 2.944,
-      "grad_norm": 13.708846092224121,
-      "learning_rate": 3.922720797720798e-05,
-      "loss": 0.2849,
+      "grad_norm": 18.0466365814209,
+      "learning_rate": 1.0688836104513065e-06,
+      "loss": 0.257,
       "step": 920
     },
     {
       "epoch": 2.976,
-      "grad_norm": 7.4183735847473145,
-      "learning_rate": 3.90491452991453e-05,
-      "loss": 0.2266,
+      "grad_norm": 12.547955513000488,
+      "learning_rate": 4.750593824228029e-07,
+      "loss": 0.2103,
       "step": 930
     },
     {
-      "epoch": 2.9984,
-      "eval_accuracy": 0.556,
-      "eval_loss": 1.8065074682235718,
-      "eval_runtime": 155.7933,
-      "eval_samples_per_second": 25.675,
-      "eval_steps_per_second": 0.802,
-      "step": 937
-    },
-    {
-      "epoch": 3.008,
-      "grad_norm": 51.52549743652344,
-      "learning_rate": 3.8871082621082625e-05,
-      "loss": 0.3457,
-      "step": 940
-    },
-    {
-      "epoch": 3.04,
-      "grad_norm": 22.345596313476562,
-      "learning_rate": 3.869301994301994e-05,
-      "loss": 0.3204,
-      "step": 950
-    },
-    {
-      "epoch": 3.072,
-      "grad_norm": 16.317461013793945,
-      "learning_rate": 3.851495726495727e-05,
-      "loss": 0.3101,
-      "step": 960
-    },
-    {
-      "epoch": 3.104,
-      "grad_norm": 27.974485397338867,
-      "learning_rate": 3.8336894586894586e-05,
-      "loss": 0.2918,
-      "step": 970
-    },
-    {
-      "epoch": 3.136,
-      "grad_norm": 23.675912857055664,
-      "learning_rate": 3.815883190883191e-05,
-      "loss": 0.2817,
-      "step": 980
-    },
-    {
-      "epoch": 3.168,
-      "grad_norm": 30.348739624023438,
-      "learning_rate": 3.798076923076923e-05,
-      "loss": 0.2431,
-      "step": 990
-    },
-    {
-      "epoch": 3.2,
-      "grad_norm": 11.50307846069336,
-      "learning_rate": 3.7802706552706555e-05,
-      "loss": 0.2584,
-      "step": 1000
-    },
-    {
-      "epoch": 3.232,
-      "grad_norm": 19.67528533935547,
-      "learning_rate": 3.762464387464388e-05,
-      "loss": 0.2714,
-      "step": 1010
-    },
-    {
-      "epoch": 3.2640000000000002,
-      "grad_norm": 22.57136344909668,
-      "learning_rate": 3.74465811965812e-05,
-      "loss": 0.2544,
-      "step": 1020
-    },
-    {
-      "epoch": 3.296,
-      "grad_norm": 16.600223541259766,
-      "learning_rate": 3.726851851851852e-05,
-      "loss": 0.2469,
-      "step": 1030
-    },
-    {
-      "epoch": 3.328,
-      "grad_norm": 13.466314315795898,
-      "learning_rate": 3.709045584045584e-05,
-      "loss": 0.3036,
-      "step": 1040
-    },
-    {
-      "epoch": 3.36,
-      "grad_norm": 21.5009765625,
-      "learning_rate": 3.691239316239317e-05,
-      "loss": 0.2864,
-      "step": 1050
-    },
-    {
-      "epoch": 3.392,
-      "grad_norm": 7.8777008056640625,
-      "learning_rate": 3.6734330484330485e-05,
-      "loss": 0.2907,
-      "step": 1060
-    },
-    {
-      "epoch": 3.424,
-      "grad_norm": 14.131542205810547,
-      "learning_rate": 3.655626780626781e-05,
-      "loss": 0.2849,
-      "step": 1070
-    },
-    {
-      "epoch": 3.456,
-      "grad_norm": 11.86939525604248,
-      "learning_rate": 3.637820512820513e-05,
-      "loss": 0.2659,
-      "step": 1080
-    },
-    {
-      "epoch": 3.488,
-      "grad_norm": 8.989656448364258,
-      "learning_rate": 3.6200142450142454e-05,
-      "loss": 0.2714,
-      "step": 1090
-    },
-    {
-      "epoch": 3.52,
-      "grad_norm": 22.53978729248047,
-      "learning_rate": 3.602207977207977e-05,
-      "loss": 0.2592,
-      "step": 1100
-    },
-    {
-      "epoch": 3.552,
-      "grad_norm": 11.309256553649902,
-      "learning_rate": 3.58440170940171e-05,
-      "loss": 0.248,
-      "step": 1110
-    },
-    {
-      "epoch": 3.584,
-      "grad_norm": 8.17591381072998,
-      "learning_rate": 3.5665954415954415e-05,
-      "loss": 0.2481,
-      "step": 1120
-    },
-    {
-      "epoch": 3.616,
-      "grad_norm": 44.116905212402344,
-      "learning_rate": 3.548789173789174e-05,
-      "loss": 0.2424,
-      "step": 1130
-    },
-    {
-      "epoch": 3.648,
-      "grad_norm": 15.891951560974121,
-      "learning_rate": 3.5309829059829066e-05,
-      "loss": 0.2467,
-      "step": 1140
-    },
-    {
-      "epoch": 3.68,
-      "grad_norm": 9.102974891662598,
-      "learning_rate": 3.5131766381766384e-05,
-      "loss": 0.2456,
-      "step": 1150
-    },
-    {
-      "epoch": 3.7119999999999997,
-      "grad_norm": 14.138261795043945,
-      "learning_rate": 3.49537037037037e-05,
-      "loss": 0.2479,
-      "step": 1160
-    },
-    {
-      "epoch": 3.7439999999999998,
-      "grad_norm": 37.55530548095703,
-      "learning_rate": 3.477564102564103e-05,
-      "loss": 0.2604,
-      "step": 1170
-    },
-    {
-      "epoch": 3.776,
-      "grad_norm": 20.72405242919922,
-      "learning_rate": 3.4597578347578346e-05,
-      "loss": 0.2392,
-      "step": 1180
-    },
-    {
-      "epoch": 3.808,
-      "grad_norm": 9.120800018310547,
-      "learning_rate": 3.441951566951567e-05,
-      "loss": 0.2243,
-      "step": 1190
-    },
-    {
-      "epoch": 3.84,
-      "grad_norm": 5.410057067871094,
-      "learning_rate": 3.4241452991452996e-05,
-      "loss": 0.2796,
-      "step": 1200
-    },
-    {
-      "epoch": 3.872,
-      "grad_norm": 13.668625831604004,
-      "learning_rate": 3.4063390313390314e-05,
-      "loss": 0.2386,
-      "step": 1210
-    },
-    {
-      "epoch": 3.904,
-      "grad_norm": 15.398672103881836,
-      "learning_rate": 3.388532763532763e-05,
-      "loss": 0.2461,
-      "step": 1220
-    },
-    {
-      "epoch": 3.936,
-      "grad_norm": 21.709699630737305,
-      "learning_rate": 3.3707264957264964e-05,
-      "loss": 0.2225,
-      "step": 1230
-    },
-    {
-      "epoch": 3.968,
-      "grad_norm": 6.977370262145996,
-      "learning_rate": 3.352920227920228e-05,
-      "loss": 0.223,
-      "step": 1240
-    },
-    {
-      "epoch": 4.0,
-      "grad_norm": 8.210868835449219,
-      "learning_rate": 3.33511396011396e-05,
-      "loss": 0.2529,
-      "step": 1250
-    },
-    {
-      "epoch": 4.0,
-      "eval_accuracy": 0.713,
-      "eval_loss": 0.8170278668403625,
-      "eval_runtime": 160.9705,
-      "eval_samples_per_second": 24.849,
-      "eval_steps_per_second": 0.777,
-      "step": 1250
-    },
-    {
-      "epoch": 4.032,
-      "grad_norm": 10.20702075958252,
-      "learning_rate": 3.3173076923076926e-05,
-      "loss": 0.2091,
-      "step": 1260
-    },
-    {
-      "epoch": 4.064,
-      "grad_norm": 10.824769020080566,
-      "learning_rate": 3.2995014245014244e-05,
-      "loss": 0.2298,
-      "step": 1270
-    },
-    {
-      "epoch": 4.096,
-      "grad_norm": 10.660799026489258,
-      "learning_rate": 3.281695156695157e-05,
-      "loss": 0.2172,
-      "step": 1280
-    },
-    {
-      "epoch": 4.128,
-      "grad_norm": 12.774826049804688,
-      "learning_rate": 3.263888888888889e-05,
-      "loss": 0.2204,
-      "step": 1290
-    },
-    {
-      "epoch": 4.16,
-      "grad_norm": 9.919928550720215,
-      "learning_rate": 3.246082621082621e-05,
-      "loss": 0.2355,
-      "step": 1300
-    },
-    {
-      "epoch": 4.192,
-      "grad_norm": 10.62849235534668,
-      "learning_rate": 3.228276353276353e-05,
-      "loss": 0.2046,
-      "step": 1310
-    },
-    {
-      "epoch": 4.224,
-      "grad_norm": 11.318085670471191,
-      "learning_rate": 3.2104700854700856e-05,
-      "loss": 0.2058,
-      "step": 1320
-    },
-    {
-      "epoch": 4.256,
-      "grad_norm": 23.96332359313965,
-      "learning_rate": 3.192663817663818e-05,
-      "loss": 0.2158,
-      "step": 1330
-    },
-    {
-      "epoch": 4.288,
-      "grad_norm": 12.120290756225586,
-      "learning_rate": 3.17485754985755e-05,
-      "loss": 0.1955,
-      "step": 1340
-    },
-    {
-      "epoch": 4.32,
-      "grad_norm": 22.462135314941406,
-      "learning_rate": 3.157051282051282e-05,
-      "loss": 0.2213,
-      "step": 1350
-    },
-    {
-      "epoch": 4.352,
-      "grad_norm": 10.982671737670898,
-      "learning_rate": 3.139245014245014e-05,
-      "loss": 0.204,
-      "step": 1360
-    },
-    {
-      "epoch": 4.384,
-      "grad_norm": 9.690351486206055,
-      "learning_rate": 3.121438746438747e-05,
-      "loss": 0.2072,
-      "step": 1370
-    },
-    {
-      "epoch": 4.416,
-      "grad_norm": 21.02212142944336,
-      "learning_rate": 3.103632478632479e-05,
-      "loss": 0.2084,
-      "step": 1380
-    },
-    {
-      "epoch": 4.448,
-      "grad_norm": 11.301972389221191,
-      "learning_rate": 3.085826210826211e-05,
-      "loss": 0.2046,
-      "step": 1390
-    },
-    {
-      "epoch": 4.48,
-      "grad_norm": 13.271730422973633,
-      "learning_rate": 3.068019943019943e-05,
-      "loss": 0.2341,
-      "step": 1400
-    },
-    {
-      "epoch": 4.5120000000000005,
-      "grad_norm": 12.261125564575195,
-      "learning_rate": 3.0502136752136755e-05,
-      "loss": 0.2151,
-      "step": 1410
-    },
-    {
-      "epoch": 4.5440000000000005,
-      "grad_norm": 5.64617919921875,
-      "learning_rate": 3.0324074074074077e-05,
-      "loss": 0.2046,
-      "step": 1420
-    },
-    {
-      "epoch": 4.576,
-      "grad_norm": 17.69122314453125,
-      "learning_rate": 3.01460113960114e-05,
-      "loss": 0.2058,
-      "step": 1430
-    },
-    {
-      "epoch": 4.608,
-      "grad_norm": 10.206043243408203,
-      "learning_rate": 2.9967948717948717e-05,
-      "loss": 0.2235,
-      "step": 1440
-    },
-    {
-      "epoch": 4.64,
-      "grad_norm": 13.48324203491211,
-      "learning_rate": 2.978988603988604e-05,
-      "loss": 0.1907,
-      "step": 1450
-    },
-    {
-      "epoch": 4.672,
-      "grad_norm": 7.78882360458374,
-      "learning_rate": 2.9611823361823364e-05,
-      "loss": 0.2012,
-      "step": 1460
-    },
-    {
-      "epoch": 4.704,
-      "grad_norm": 10.230093002319336,
-      "learning_rate": 2.9433760683760685e-05,
-      "loss": 0.1803,
-      "step": 1470
-    },
-    {
-      "epoch": 4.736,
-      "grad_norm": 37.21369934082031,
-      "learning_rate": 2.9255698005698007e-05,
-      "loss": 0.1995,
-      "step": 1480
-    },
-    {
-      "epoch": 4.768,
-      "grad_norm": 10.13237190246582,
-      "learning_rate": 2.907763532763533e-05,
-      "loss": 0.2096,
-      "step": 1490
-    },
-    {
-      "epoch": 4.8,
-      "grad_norm": 6.978107929229736,
-      "learning_rate": 2.8899572649572647e-05,
-      "loss": 0.1905,
-      "step": 1500
-    },
-    {
-      "epoch": 4.832,
-      "grad_norm": 17.814905166625977,
-      "learning_rate": 2.8721509971509976e-05,
-      "loss": 0.1818,
-      "step": 1510
-    },
-    {
-      "epoch": 4.864,
-      "grad_norm": 10.511373519897461,
-      "learning_rate": 2.8543447293447294e-05,
-      "loss": 0.2339,
-      "step": 1520
-    },
-    {
-      "epoch": 4.896,
-      "grad_norm": 16.811450958251953,
-      "learning_rate": 2.8365384615384616e-05,
-      "loss": 0.1969,
-      "step": 1530
-    },
-    {
-      "epoch": 4.928,
-      "grad_norm": 8.571528434753418,
-      "learning_rate": 2.8187321937321937e-05,
-      "loss": 0.1951,
-      "step": 1540
-    },
-    {
-      "epoch": 4.96,
-      "grad_norm": 15.971531867980957,
-      "learning_rate": 2.8009259259259263e-05,
-      "loss": 0.2012,
-      "step": 1550
-    },
-    {
-      "epoch": 4.992,
-      "grad_norm": 22.862905502319336,
-      "learning_rate": 2.7831196581196584e-05,
-      "loss": 0.1925,
-      "step": 1560
-    },
-    {
-      "epoch": 4.9984,
-      "eval_accuracy": 0.69075,
-      "eval_loss": 1.0642794370651245,
-      "eval_runtime": 164.1188,
-      "eval_samples_per_second": 24.373,
-      "eval_steps_per_second": 0.762,
-      "step": 1562
-    },
-    {
-      "epoch": 5.024,
-      "grad_norm": 14.107147216796875,
-      "learning_rate": 2.7653133903133903e-05,
-      "loss": 0.2248,
-      "step": 1570
-    },
-    {
-      "epoch": 5.056,
-      "grad_norm": 16.5964412689209,
-      "learning_rate": 2.7475071225071224e-05,
-      "loss": 0.1826,
-      "step": 1580
-    },
-    {
-      "epoch": 5.088,
-      "grad_norm": 9.056426048278809,
-      "learning_rate": 2.7297008547008546e-05,
-      "loss": 0.1975,
-      "step": 1590
-    },
-    {
-      "epoch": 5.12,
-      "grad_norm": 11.496504783630371,
-      "learning_rate": 2.711894586894587e-05,
-      "loss": 0.1818,
-      "step": 1600
-    },
-    {
-      "epoch": 5.152,
-      "grad_norm": 14.043261528015137,
-      "learning_rate": 2.6940883190883193e-05,
-      "loss": 0.1744,
-      "step": 1610
-    },
-    {
-      "epoch": 5.184,
-      "grad_norm": 8.685189247131348,
-      "learning_rate": 2.6762820512820515e-05,
-      "loss": 0.2042,
-      "step": 1620
-    },
-    {
-      "epoch": 5.216,
-      "grad_norm": 7.994570732116699,
-      "learning_rate": 2.6584757834757833e-05,
-      "loss": 0.1703,
-      "step": 1630
-    },
-    {
-      "epoch": 5.248,
-      "grad_norm": 14.493515968322754,
-      "learning_rate": 2.640669515669516e-05,
-      "loss": 0.1736,
-      "step": 1640
-    },
-    {
-      "epoch": 5.28,
-      "grad_norm": 7.479248523712158,
-      "learning_rate": 2.622863247863248e-05,
-      "loss": 0.1868,
-      "step": 1650
-    },
-    {
-      "epoch": 5.312,
-      "grad_norm": 12.55150318145752,
-      "learning_rate": 2.60505698005698e-05,
-      "loss": 0.2088,
-      "step": 1660
-    },
-    {
-      "epoch": 5.344,
-      "grad_norm": 8.034307479858398,
-      "learning_rate": 2.5872507122507123e-05,
-      "loss": 0.18,
-      "step": 1670
-    },
-    {
-      "epoch": 5.376,
-      "grad_norm": 7.336935997009277,
-      "learning_rate": 2.5694444444444445e-05,
-      "loss": 0.173,
-      "step": 1680
-    },
-    {
-      "epoch": 5.408,
-      "grad_norm": 7.364587306976318,
-      "learning_rate": 2.551638176638177e-05,
-      "loss": 0.1739,
-      "step": 1690
-    },
-    {
-      "epoch": 5.44,
-      "grad_norm": 12.251140594482422,
-      "learning_rate": 2.533831908831909e-05,
-      "loss": 0.1766,
-      "step": 1700
-    },
-    {
-      "epoch": 5.4719999999999995,
-      "grad_norm": 8.115775108337402,
-      "learning_rate": 2.516025641025641e-05,
-      "loss": 0.1862,
-      "step": 1710
-    },
-    {
-      "epoch": 5.504,
-      "grad_norm": 12.475907325744629,
-      "learning_rate": 2.4982193732193735e-05,
-      "loss": 0.1988,
-      "step": 1720
-    },
-    {
-      "epoch": 5.536,
-      "grad_norm": 7.327911376953125,
-      "learning_rate": 2.4804131054131057e-05,
-      "loss": 0.1828,
-      "step": 1730
-    },
-    {
-      "epoch": 5.568,
-      "grad_norm": 8.060572624206543,
-      "learning_rate": 2.462606837606838e-05,
-      "loss": 0.1692,
-      "step": 1740
-    },
-    {
-      "epoch": 5.6,
-      "grad_norm": 28.06664276123047,
-      "learning_rate": 2.44480056980057e-05,
-      "loss": 0.1345,
-      "step": 1750
-    },
-    {
-      "epoch": 5.632,
-      "grad_norm": 12.124792098999023,
-      "learning_rate": 2.426994301994302e-05,
-      "loss": 0.2026,
-      "step": 1760
-    },
-    {
-      "epoch": 5.664,
-      "grad_norm": 12.650993347167969,
-      "learning_rate": 2.4091880341880344e-05,
-      "loss": 0.1861,
-      "step": 1770
-    },
-    {
-      "epoch": 5.696,
-      "grad_norm": 15.481111526489258,
-      "learning_rate": 2.3913817663817665e-05,
-      "loss": 0.1963,
-      "step": 1780
-    },
-    {
-      "epoch": 5.728,
-      "grad_norm": 12.246734619140625,
-      "learning_rate": 2.3735754985754987e-05,
-      "loss": 0.162,
-      "step": 1790
-    },
-    {
-      "epoch": 5.76,
-      "grad_norm": 11.539756774902344,
-      "learning_rate": 2.355769230769231e-05,
-      "loss": 0.2141,
-      "step": 1800
-    },
-    {
-      "epoch": 5.792,
-      "grad_norm": 7.406561374664307,
-      "learning_rate": 2.337962962962963e-05,
-      "loss": 0.1877,
-      "step": 1810
-    },
-    {
-      "epoch": 5.824,
-      "grad_norm": 14.728952407836914,
-      "learning_rate": 2.3201566951566952e-05,
-      "loss": 0.1796,
-      "step": 1820
-    },
-    {
-      "epoch": 5.856,
-      "grad_norm": 7.532665729522705,
-      "learning_rate": 2.3023504273504274e-05,
-      "loss": 0.157,
-      "step": 1830
-    },
-    {
-      "epoch": 5.888,
-      "grad_norm": 16.279264450073242,
-      "learning_rate": 2.2845441595441596e-05,
-      "loss": 0.1836,
-      "step": 1840
-    },
-    {
-      "epoch": 5.92,
-      "grad_norm": 7.029083728790283,
-      "learning_rate": 2.2667378917378917e-05,
-      "loss": 0.1788,
-      "step": 1850
-    },
-    {
-      "epoch": 5.952,
-      "grad_norm": 7.0097174644470215,
-      "learning_rate": 2.2489316239316242e-05,
-      "loss": 0.1658,
-      "step": 1860
-    },
-    {
-      "epoch": 5.984,
-      "grad_norm": 8.254197120666504,
-      "learning_rate": 2.231125356125356e-05,
-      "loss": 0.177,
-      "step": 1870
-    },
-    {
-      "epoch": 6.0,
-      "eval_accuracy": 0.68425,
-      "eval_loss": 1.2558131217956543,
-      "eval_runtime": 148.6431,
-      "eval_samples_per_second": 26.91,
-      "eval_steps_per_second": 0.841,
-      "step": 1875
-    },
-    {
-      "epoch": 6.016,
-      "grad_norm": 10.65648078918457,
-      "learning_rate": 2.2133190883190886e-05,
-      "loss": 0.197,
-      "step": 1880
-    },
-    {
-      "epoch": 6.048,
-      "grad_norm": 12.431746482849121,
-      "learning_rate": 2.1955128205128208e-05,
-      "loss": 0.1876,
-      "step": 1890
-    },
-    {
-      "epoch": 6.08,
-      "grad_norm": 20.30584716796875,
-      "learning_rate": 2.177706552706553e-05,
-      "loss": 0.1651,
-      "step": 1900
-    },
-    {
-      "epoch": 6.112,
-      "grad_norm": 13.690505981445312,
-      "learning_rate": 2.159900284900285e-05,
-      "loss": 0.1646,
-      "step": 1910
-    },
-    {
-      "epoch": 6.144,
-      "grad_norm": 23.424381256103516,
-      "learning_rate": 2.142094017094017e-05,
-      "loss": 0.1732,
-      "step": 1920
-    },
-    {
-      "epoch": 6.176,
-      "grad_norm": 8.711691856384277,
-      "learning_rate": 2.1242877492877494e-05,
-      "loss": 0.1537,
-      "step": 1930
-    },
-    {
-      "epoch": 6.208,
-      "grad_norm": 10.551396369934082,
-      "learning_rate": 2.1064814814814816e-05,
-      "loss": 0.1512,
-      "step": 1940
-    },
-    {
-      "epoch": 6.24,
-      "grad_norm": 31.970632553100586,
-      "learning_rate": 2.0886752136752138e-05,
-      "loss": 0.1607,
-      "step": 1950
-    },
-    {
-      "epoch": 6.272,
-      "grad_norm": 14.962709426879883,
-      "learning_rate": 2.070868945868946e-05,
-      "loss": 0.1476,
-      "step": 1960
-    },
-    {
-      "epoch": 6.304,
-      "grad_norm": 11.659952163696289,
-      "learning_rate": 2.053062678062678e-05,
-      "loss": 0.1627,
-      "step": 1970
-    },
-    {
-      "epoch": 6.336,
-      "grad_norm": 8.625643730163574,
-      "learning_rate": 2.0352564102564103e-05,
-      "loss": 0.1644,
-      "step": 1980
-    },
-    {
-      "epoch": 6.368,
-      "grad_norm": 20.484344482421875,
-      "learning_rate": 2.0174501424501425e-05,
-      "loss": 0.1605,
-      "step": 1990
-    },
-    {
-      "epoch": 6.4,
-      "grad_norm": 8.137100219726562,
-      "learning_rate": 1.9996438746438746e-05,
-      "loss": 0.1482,
-      "step": 2000
-    },
-    {
-      "epoch": 6.432,
-      "grad_norm": 9.601692199707031,
-      "learning_rate": 1.9818376068376068e-05,
-      "loss": 0.1654,
-      "step": 2010
-    },
-    {
-      "epoch": 6.464,
-      "grad_norm": 10.012831687927246,
-      "learning_rate": 1.9640313390313393e-05,
-      "loss": 0.1925,
-      "step": 2020
-    },
-    {
-      "epoch": 6.496,
-      "grad_norm": 7.507844924926758,
-      "learning_rate": 1.946225071225071e-05,
-      "loss": 0.1735,
-      "step": 2030
-    },
-    {
-      "epoch": 6.5280000000000005,
-      "grad_norm": 19.662033081054688,
-      "learning_rate": 1.9284188034188037e-05,
-      "loss": 0.1547,
-      "step": 2040
-    },
-    {
-      "epoch": 6.5600000000000005,
-      "grad_norm": 7.522573947906494,
-      "learning_rate": 1.910612535612536e-05,
-      "loss": 0.1602,
-      "step": 2050
-    },
-    {
-      "epoch": 6.592,
-      "grad_norm": 9.965826034545898,
-      "learning_rate": 1.892806267806268e-05,
-      "loss": 0.1488,
-      "step": 2060
-    },
-    {
-      "epoch": 6.624,
-      "grad_norm": 6.560938358306885,
-      "learning_rate": 1.8750000000000002e-05,
-      "loss": 0.1439,
-      "step": 2070
-    },
-    {
-      "epoch": 6.656,
-      "grad_norm": 16.683101654052734,
-      "learning_rate": 1.857193732193732e-05,
-      "loss": 0.1384,
-      "step": 2080
-    },
-    {
-      "epoch": 6.688,
-      "grad_norm": 7.641848564147949,
-      "learning_rate": 1.8393874643874645e-05,
-      "loss": 0.1429,
-      "step": 2090
-    },
-    {
-      "epoch": 6.72,
-      "grad_norm": 11.998226165771484,
-      "learning_rate": 1.8215811965811967e-05,
-      "loss": 0.1485,
-      "step": 2100
-    },
-    {
-      "epoch": 6.752,
-      "grad_norm": 7.238443374633789,
-      "learning_rate": 1.803774928774929e-05,
-      "loss": 0.1544,
-      "step": 2110
-    },
-    {
-      "epoch": 6.784,
-      "grad_norm": 7.6015777587890625,
-      "learning_rate": 1.785968660968661e-05,
-      "loss": 0.1864,
-      "step": 2120
-    },
-    {
-      "epoch": 6.816,
-      "grad_norm": 12.038315773010254,
-      "learning_rate": 1.7681623931623932e-05,
-      "loss": 0.1479,
-      "step": 2130
-    },
-    {
-      "epoch": 6.848,
-      "grad_norm": 25.752042770385742,
-      "learning_rate": 1.7503561253561254e-05,
-      "loss": 0.1479,
-      "step": 2140
-    },
-    {
-      "epoch": 6.88,
-      "grad_norm": 12.058791160583496,
-      "learning_rate": 1.732549857549858e-05,
-      "loss": 0.1342,
-      "step": 2150
-    },
-    {
-      "epoch": 6.912,
-      "grad_norm": 18.634387969970703,
-      "learning_rate": 1.7147435897435897e-05,
-      "loss": 0.165,
-      "step": 2160
-    },
-    {
-      "epoch": 6.944,
-      "grad_norm": 8.238666534423828,
-      "learning_rate": 1.696937321937322e-05,
-      "loss": 0.1452,
-      "step": 2170
-    },
-    {
-      "epoch": 6.976,
-      "grad_norm": 8.537055015563965,
-      "learning_rate": 1.6791310541310544e-05,
-      "loss": 0.1563,
-      "step": 2180
-    },
-    {
-      "epoch": 6.9984,
-      "eval_accuracy": 0.7445,
-      "eval_loss": 0.9205208420753479,
-      "eval_runtime": 156.8331,
-      "eval_samples_per_second": 25.505,
-      "eval_steps_per_second": 0.797,
-      "step": 2187
-    },
-    {
-      "epoch": 7.008,
-      "grad_norm": 21.315475463867188,
-      "learning_rate": 1.6613247863247862e-05,
-      "loss": 0.1621,
-      "step": 2190
-    },
-    {
-      "epoch": 7.04,
-      "grad_norm": 4.930037975311279,
-      "learning_rate": 1.6435185185185187e-05,
-      "loss": 0.143,
-      "step": 2200
-    },
-    {
-      "epoch": 7.072,
-      "grad_norm": 20.74496078491211,
-      "learning_rate": 1.625712250712251e-05,
-      "loss": 0.1334,
-      "step": 2210
-    },
-    {
-      "epoch": 7.104,
-      "grad_norm": 32.969154357910156,
-      "learning_rate": 1.607905982905983e-05,
-      "loss": 0.1502,
-      "step": 2220
-    },
-    {
-      "epoch": 7.136,
-      "grad_norm": 10.03094482421875,
-      "learning_rate": 1.5900997150997153e-05,
-      "loss": 0.1611,
-      "step": 2230
-    },
-    {
-      "epoch": 7.168,
-      "grad_norm": 10.585143089294434,
-      "learning_rate": 1.572293447293447e-05,
-      "loss": 0.1583,
-      "step": 2240
-    },
-    {
-      "epoch": 7.2,
-      "grad_norm": 11.96883487701416,
-      "learning_rate": 1.5544871794871796e-05,
-      "loss": 0.1433,
-      "step": 2250
-    },
-    {
-      "epoch": 7.232,
-      "grad_norm": 17.068143844604492,
-      "learning_rate": 1.5366809116809118e-05,
-      "loss": 0.1641,
-      "step": 2260
-    },
-    {
-      "epoch": 7.264,
-      "grad_norm": 9.417765617370605,
-      "learning_rate": 1.518874643874644e-05,
-      "loss": 0.1266,
-      "step": 2270
-    },
-    {
-      "epoch": 7.296,
-      "grad_norm": 12.315200805664062,
-      "learning_rate": 1.5010683760683761e-05,
-      "loss": 0.1337,
-      "step": 2280
-    },
-    {
-      "epoch": 7.328,
-      "grad_norm": 20.546875,
-      "learning_rate": 1.4832621082621084e-05,
-      "loss": 0.1312,
-      "step": 2290
-    },
-    {
-      "epoch": 7.36,
-      "grad_norm": 7.2887468338012695,
-      "learning_rate": 1.4654558404558405e-05,
-      "loss": 0.1534,
-      "step": 2300
-    },
-    {
-      "epoch": 7.392,
-      "grad_norm": 8.704584121704102,
-      "learning_rate": 1.4476495726495728e-05,
-      "loss": 0.1317,
-      "step": 2310
-    },
-    {
-      "epoch": 7.424,
-      "grad_norm": 9.408025741577148,
-      "learning_rate": 1.429843304843305e-05,
-      "loss": 0.1336,
-      "step": 2320
-    },
-    {
-      "epoch": 7.456,
-      "grad_norm": 16.08376121520996,
-      "learning_rate": 1.412037037037037e-05,
-      "loss": 0.1361,
-      "step": 2330
-    },
-    {
-      "epoch": 7.4879999999999995,
-      "grad_norm": 13.386305809020996,
-      "learning_rate": 1.3942307692307693e-05,
-      "loss": 0.1254,
-      "step": 2340
-    },
-    {
-      "epoch": 7.52,
-      "grad_norm": 8.729687690734863,
-      "learning_rate": 1.3764245014245015e-05,
-      "loss": 0.1477,
-      "step": 2350
-    },
-    {
-      "epoch": 7.552,
-      "grad_norm": 7.066646099090576,
-      "learning_rate": 1.3586182336182338e-05,
-      "loss": 0.1266,
-      "step": 2360
-    },
-    {
-      "epoch": 7.584,
-      "grad_norm": 11.6298828125,
-      "learning_rate": 1.3408119658119658e-05,
-      "loss": 0.1303,
-      "step": 2370
-    },
-    {
-      "epoch": 7.616,
-      "grad_norm": 5.477276802062988,
-      "learning_rate": 1.3230056980056982e-05,
-      "loss": 0.126,
-      "step": 2380
-    },
-    {
-      "epoch": 7.648,
-      "grad_norm": 9.086950302124023,
-      "learning_rate": 1.3051994301994303e-05,
-      "loss": 0.1383,
-      "step": 2390
-    },
-    {
-      "epoch": 7.68,
-      "grad_norm": 7.193227291107178,
-      "learning_rate": 1.2873931623931623e-05,
-      "loss": 0.1371,
-      "step": 2400
-    },
-    {
-      "epoch": 7.712,
-      "grad_norm": 11.82983684539795,
-      "learning_rate": 1.2695868945868947e-05,
-      "loss": 0.1406,
-      "step": 2410
-    },
-    {
-      "epoch": 7.744,
-      "grad_norm": 9.807685852050781,
-      "learning_rate": 1.2517806267806267e-05,
-      "loss": 0.1349,
-      "step": 2420
-    },
-    {
-      "epoch": 7.776,
-      "grad_norm": 13.509228706359863,
-      "learning_rate": 1.233974358974359e-05,
-      "loss": 0.1324,
-      "step": 2430
-    },
-    {
-      "epoch": 7.808,
-      "grad_norm": 20.25900650024414,
-      "learning_rate": 1.2161680911680912e-05,
-      "loss": 0.1337,
-      "step": 2440
-    },
-    {
-      "epoch": 7.84,
-      "grad_norm": 8.688305854797363,
-      "learning_rate": 1.1983618233618234e-05,
-      "loss": 0.1206,
-      "step": 2450
-    },
-    {
-      "epoch": 7.872,
-      "grad_norm": 16.35320472717285,
-      "learning_rate": 1.1805555555555555e-05,
-      "loss": 0.133,
-      "step": 2460
-    },
-    {
-      "epoch": 7.904,
-      "grad_norm": 35.276092529296875,
-      "learning_rate": 1.1627492877492879e-05,
-      "loss": 0.1425,
-      "step": 2470
-    },
-    {
-      "epoch": 7.936,
-      "grad_norm": 16.226844787597656,
-      "learning_rate": 1.14494301994302e-05,
-      "loss": 0.1421,
-      "step": 2480
-    },
-    {
-      "epoch": 7.968,
-      "grad_norm": 4.65103006362915,
-      "learning_rate": 1.1271367521367522e-05,
-      "loss": 0.1225,
-      "step": 2490
-    },
-    {
-      "epoch": 8.0,
-      "grad_norm": 10.949529647827148,
-      "learning_rate": 1.1093304843304844e-05,
-      "loss": 0.1417,
-      "step": 2500
-    },
-    {
-      "epoch": 8.0,
-      "eval_accuracy": 0.80625,
-      "eval_loss": 0.662419319152832,
-      "eval_runtime": 156.8602,
-      "eval_samples_per_second": 25.5,
-      "eval_steps_per_second": 0.797,
-      "step": 2500
-    },
-    {
-      "epoch": 8.032,
-      "grad_norm": 14.452116966247559,
-      "learning_rate": 1.0915242165242166e-05,
-      "loss": 0.1422,
-      "step": 2510
-    },
-    {
-      "epoch": 8.064,
-      "grad_norm": 26.71641731262207,
-      "learning_rate": 1.0737179487179487e-05,
-      "loss": 0.1337,
-      "step": 2520
-    },
-    {
-      "epoch": 8.096,
-      "grad_norm": 8.862665176391602,
-      "learning_rate": 1.0559116809116809e-05,
-      "loss": 0.1349,
-      "step": 2530
-    },
-    {
-      "epoch": 8.128,
-      "grad_norm": 12.3915433883667,
-      "learning_rate": 1.038105413105413e-05,
-      "loss": 0.1479,
-      "step": 2540
-    },
-    {
-      "epoch": 8.16,
-      "grad_norm": 11.4617280960083,
-      "learning_rate": 1.0202991452991454e-05,
-      "loss": 0.1295,
-      "step": 2550
-    },
-    {
-      "epoch": 8.192,
-      "grad_norm": 19.975990295410156,
-      "learning_rate": 1.0024928774928776e-05,
-      "loss": 0.1382,
-      "step": 2560
-    },
-    {
-      "epoch": 8.224,
-      "grad_norm": 9.245658874511719,
-      "learning_rate": 9.846866096866097e-06,
-      "loss": 0.1308,
-      "step": 2570
-    },
-    {
-      "epoch": 8.256,
-      "grad_norm": 5.326711177825928,
-      "learning_rate": 9.66880341880342e-06,
-      "loss": 0.1023,
-      "step": 2580
-    },
-    {
-      "epoch": 8.288,
-      "grad_norm": 11.740446090698242,
-      "learning_rate": 9.490740740740741e-06,
-      "loss": 0.1262,
-      "step": 2590
-    },
-    {
-      "epoch": 8.32,
-      "grad_norm": 10.618571281433105,
-      "learning_rate": 9.312678062678064e-06,
-      "loss": 0.1306,
-      "step": 2600
-    },
-    {
-      "epoch": 8.352,
-      "grad_norm": 21.31345558166504,
-      "learning_rate": 9.134615384615384e-06,
-      "loss": 0.136,
-      "step": 2610
-    },
-    {
-      "epoch": 8.384,
-      "grad_norm": 6.4540863037109375,
-      "learning_rate": 8.956552706552706e-06,
-      "loss": 0.1063,
-      "step": 2620
-    },
-    {
-      "epoch": 8.416,
-      "grad_norm": 12.291879653930664,
-      "learning_rate": 8.77849002849003e-06,
-      "loss": 0.1222,
-      "step": 2630
-    },
-    {
-      "epoch": 8.448,
-      "grad_norm": 7.860499858856201,
-      "learning_rate": 8.600427350427351e-06,
-      "loss": 0.1152,
-      "step": 2640
-    },
-    {
-      "epoch": 8.48,
-      "grad_norm": 7.345857620239258,
-      "learning_rate": 8.422364672364673e-06,
-      "loss": 0.1194,
-      "step": 2650
-    },
-    {
-      "epoch": 8.512,
-      "grad_norm": 25.202924728393555,
-      "learning_rate": 8.244301994301995e-06,
-      "loss": 0.1214,
-      "step": 2660
-    },
-    {
-      "epoch": 8.544,
-      "grad_norm": 4.337811470031738,
-      "learning_rate": 8.066239316239316e-06,
-      "loss": 0.1316,
-      "step": 2670
-    },
-    {
-      "epoch": 8.576,
-      "grad_norm": 9.332151412963867,
-      "learning_rate": 7.88817663817664e-06,
-      "loss": 0.1208,
-      "step": 2680
-    },
-    {
-      "epoch": 8.608,
-      "grad_norm": 15.260229110717773,
-      "learning_rate": 7.71011396011396e-06,
-      "loss": 0.1223,
-      "step": 2690
-    },
-    {
-      "epoch": 8.64,
-      "grad_norm": 4.976259708404541,
-      "learning_rate": 7.532051282051282e-06,
-      "loss": 0.111,
-      "step": 2700
-    },
-    {
-      "epoch": 8.672,
-      "grad_norm": 18.121919631958008,
-      "learning_rate": 7.353988603988604e-06,
-      "loss": 0.1184,
-      "step": 2710
-    },
-    {
-      "epoch": 8.704,
-      "grad_norm": 15.61289119720459,
-      "learning_rate": 7.1759259259259266e-06,
-      "loss": 0.125,
-      "step": 2720
-    },
-    {
-      "epoch": 8.736,
-      "grad_norm": 14.504264831542969,
-      "learning_rate": 6.997863247863248e-06,
-      "loss": 0.1165,
-      "step": 2730
-    },
-    {
-      "epoch": 8.768,
-      "grad_norm": 10.373973846435547,
-      "learning_rate": 6.819800569800571e-06,
-      "loss": 0.1117,
-      "step": 2740
-    },
-    {
-      "epoch": 8.8,
-      "grad_norm": 32.67918014526367,
-      "learning_rate": 6.6417378917378925e-06,
-      "loss": 0.1345,
-      "step": 2750
-    },
-    {
-      "epoch": 8.832,
-      "grad_norm": 11.788604736328125,
-      "learning_rate": 6.463675213675214e-06,
-      "loss": 0.1194,
-      "step": 2760
-    },
-    {
-      "epoch": 8.864,
-      "grad_norm": 10.273115158081055,
-      "learning_rate": 6.285612535612535e-06,
-      "loss": 0.1089,
-      "step": 2770
-    },
-    {
-      "epoch": 8.896,
-      "grad_norm": 18.808115005493164,
-      "learning_rate": 6.1075498575498585e-06,
-      "loss": 0.1106,
-      "step": 2780
-    },
-    {
-      "epoch": 8.928,
-      "grad_norm": 8.078729629516602,
-      "learning_rate": 5.929487179487179e-06,
-      "loss": 0.1407,
-      "step": 2790
-    },
-    {
-      "epoch": 8.96,
-      "grad_norm": 6.521687984466553,
-      "learning_rate": 5.751424501424502e-06,
-      "loss": 0.1195,
-      "step": 2800
-    },
-    {
-      "epoch": 8.992,
-      "grad_norm": 7.088119029998779,
-      "learning_rate": 5.573361823361824e-06,
-      "loss": 0.1284,
-      "step": 2810
-    },
-    {
-      "epoch": 8.9984,
-      "eval_accuracy": 0.739,
-      "eval_loss": 1.1648138761520386,
-      "eval_runtime": 157.4642,
-      "eval_samples_per_second": 25.403,
-      "eval_steps_per_second": 0.794,
-      "step": 2812
-    },
-    {
-      "epoch": 9.024,
-      "grad_norm": 19.490074157714844,
-      "learning_rate": 5.395299145299146e-06,
-      "loss": 0.1084,
-      "step": 2820
-    },
-    {
-      "epoch": 9.056,
-      "grad_norm": 6.817428112030029,
-      "learning_rate": 5.217236467236467e-06,
-      "loss": 0.1194,
-      "step": 2830
-    },
-    {
-      "epoch": 9.088,
-      "grad_norm": 8.714266777038574,
-      "learning_rate": 5.03917378917379e-06,
-      "loss": 0.1126,
-      "step": 2840
-    },
-    {
-      "epoch": 9.12,
-      "grad_norm": 15.373977661132812,
-      "learning_rate": 4.861111111111111e-06,
-      "loss": 0.1147,
-      "step": 2850
-    },
-    {
-      "epoch": 9.152,
-      "grad_norm": 8.224656105041504,
-      "learning_rate": 4.683048433048434e-06,
-      "loss": 0.1023,
-      "step": 2860
-    },
-    {
-      "epoch": 9.184,
-      "grad_norm": 14.560908317565918,
-      "learning_rate": 4.504985754985755e-06,
-      "loss": 0.1156,
-      "step": 2870
-    },
-    {
-      "epoch": 9.216,
-      "grad_norm": 9.510125160217285,
-      "learning_rate": 4.326923076923077e-06,
-      "loss": 0.1123,
-      "step": 2880
-    },
-    {
-      "epoch": 9.248,
-      "grad_norm": 11.263538360595703,
-      "learning_rate": 4.148860398860399e-06,
-      "loss": 0.1126,
-      "step": 2890
-    },
-    {
-      "epoch": 9.28,
-      "grad_norm": 7.910732269287109,
-      "learning_rate": 3.970797720797722e-06,
-      "loss": 0.1075,
-      "step": 2900
-    },
-    {
-      "epoch": 9.312,
-      "grad_norm": 6.68363618850708,
-      "learning_rate": 3.7927350427350425e-06,
-      "loss": 0.1001,
-      "step": 2910
-    },
-    {
-      "epoch": 9.344,
-      "grad_norm": 7.76242733001709,
-      "learning_rate": 3.6146723646723646e-06,
-      "loss": 0.0995,
-      "step": 2920
-    },
-    {
-      "epoch": 9.376,
-      "grad_norm": 12.8687744140625,
-      "learning_rate": 3.4366096866096867e-06,
-      "loss": 0.1449,
-      "step": 2930
-    },
-    {
-      "epoch": 9.408,
-      "grad_norm": 12.708325386047363,
-      "learning_rate": 3.258547008547009e-06,
-      "loss": 0.1299,
-      "step": 2940
-    },
-    {
-      "epoch": 9.44,
-      "grad_norm": 14.307499885559082,
-      "learning_rate": 3.0804843304843306e-06,
-      "loss": 0.1259,
-      "step": 2950
-    },
-    {
-      "epoch": 9.472,
-      "grad_norm": 16.907052993774414,
-      "learning_rate": 2.9024216524216523e-06,
-      "loss": 0.1014,
-      "step": 2960
-    },
-    {
-      "epoch": 9.504,
-      "grad_norm": 20.69292640686035,
-      "learning_rate": 2.7243589743589744e-06,
-      "loss": 0.1059,
-      "step": 2970
-    },
-    {
-      "epoch": 9.536,
-      "grad_norm": 7.013891696929932,
-      "learning_rate": 2.546296296296296e-06,
-      "loss": 0.1025,
-      "step": 2980
-    },
-    {
-      "epoch": 9.568,
-      "grad_norm": 10.161940574645996,
-      "learning_rate": 2.3682336182336183e-06,
-      "loss": 0.1132,
-      "step": 2990
-    },
-    {
-      "epoch": 9.6,
-      "grad_norm": 9.757913589477539,
-      "learning_rate": 2.19017094017094e-06,
-      "loss": 0.1053,
-      "step": 3000
-    },
-    {
-      "epoch": 9.632,
-      "grad_norm": 13.03531265258789,
-      "learning_rate": 2.012108262108262e-06,
-      "loss": 0.0961,
-      "step": 3010
-    },
-    {
-      "epoch": 9.664,
-      "grad_norm": 8.091846466064453,
-      "learning_rate": 1.834045584045584e-06,
-      "loss": 0.0904,
-      "step": 3020
-    },
-    {
-      "epoch": 9.696,
-      "grad_norm": 6.219699382781982,
-      "learning_rate": 1.6559829059829062e-06,
-      "loss": 0.0947,
-      "step": 3030
-    },
-    {
-      "epoch": 9.728,
-      "grad_norm": 8.676128387451172,
-      "learning_rate": 1.477920227920228e-06,
-      "loss": 0.095,
-      "step": 3040
-    },
-    {
-      "epoch": 9.76,
-      "grad_norm": 12.65188980102539,
-      "learning_rate": 1.29985754985755e-06,
-      "loss": 0.1245,
-      "step": 3050
-    },
-    {
-      "epoch": 9.792,
-      "grad_norm": 9.408976554870605,
-      "learning_rate": 1.121794871794872e-06,
-      "loss": 0.112,
-      "step": 3060
-    },
-    {
-      "epoch": 9.824,
-      "grad_norm": 5.554480075836182,
-      "learning_rate": 9.437321937321938e-07,
-      "loss": 0.0919,
-      "step": 3070
-    },
-    {
-      "epoch": 9.856,
-      "grad_norm": 20.830698013305664,
-      "learning_rate": 7.656695156695157e-07,
-      "loss": 0.1212,
-      "step": 3080
-    },
-    {
-      "epoch": 9.888,
-      "grad_norm": 8.195901870727539,
-      "learning_rate": 5.876068376068376e-07,
-      "loss": 0.1122,
-      "step": 3090
-    },
-    {
-      "epoch": 9.92,
-      "grad_norm": 14.159358024597168,
-      "learning_rate": 4.0954415954415953e-07,
-      "loss": 0.1236,
-      "step": 3100
-    },
-    {
-      "epoch": 9.952,
-      "grad_norm": 8.97334098815918,
-      "learning_rate": 2.3148148148148148e-07,
-      "loss": 0.1078,
-      "step": 3110
-    },
-    {
-      "epoch": 9.984,
-      "grad_norm": 8.013399124145508,
-      "learning_rate": 5.341880341880343e-08,
-      "loss": 0.0805,
-      "step": 3120
-    },
-    {
-      "epoch": 9.984,
-      "eval_accuracy": 0.74475,
-      "eval_loss": 1.1793317794799805,
-      "eval_runtime": 151.3727,
-      "eval_samples_per_second": 26.425,
-      "eval_steps_per_second": 0.826,
-      "step": 3120
+      "epoch": 2.9952,
+      "eval_accuracy": 0.692,
+      "eval_loss": 0.870841920375824,
+      "eval_runtime": 155.4401,
+      "eval_samples_per_second": 25.733,
+      "eval_steps_per_second": 0.804,
+      "step": 936
     },
     {
-      "epoch": 9.984,
-      "step": 3120,
-      "total_flos": 9.926487761391452e+18,
-      "train_loss": 0.24441122439427254,
-      "train_runtime": 137433.0815,
-      "train_samples_per_second": 2.911,
+      "epoch": 2.9952,
+      "step": 936,
+      "total_flos": 2.9779463284174356e+18,
+      "train_loss": 0.36125070786374247,
+      "train_runtime": 41238.7802,
+      "train_samples_per_second": 2.91,
       "train_steps_per_second": 0.023
     }
   ],
   "logging_steps": 10,
-  "max_steps": 3120,
+  "max_steps": 936,
   "num_input_tokens_seen": 0,
-  "num_train_epochs": 10,
+  "num_train_epochs": 3,
   "save_steps": 500,
   "stateful_callbacks": {
     "TrainerControl": {
@@ -2309,7 +713,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 9.926487761391452e+18,
+  "total_flos": 2.9779463284174356e+18,
   "train_batch_size": 32,
   "trial_name": null,
   "trial_params": null