End of training

README.md

@@ -23,7 +23,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.8823529411764706
+      value: 0.8970588235294118
 ---
 
 <!-- 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: 0.3140
-- Accuracy: 0.8824
+- Loss: 0.3057
+- Accuracy: 0.8971
 
 ## Model description
 

all_results.json

@@ -1,13 +1,13 @@
 {
     "epoch": 14.901960784313726,
-    "eval_accuracy": 0.9044117647058824,
-    "eval_loss": 0.20269162952899933,
-    "eval_runtime": 30.8962,
-    "eval_samples_per_second": 4.402,
-    "eval_steps_per_second": 0.55,
+    "eval_accuracy": 0.8970588235294118,
+    "eval_loss": 0.30571919679641724,
+    "eval_runtime": 30.5904,
+    "eval_samples_per_second": 4.446,
+    "eval_steps_per_second": 0.556,
     "total_flos": 4.5342391246258176e+17,
-    "train_loss": 0.4384694022044801,
-    "train_runtime": 4524.6194,
-    "train_samples_per_second": 4.058,
-    "train_steps_per_second": 0.126
+    "train_loss": 0.4511167878644508,
+    "train_runtime": 4476.1321,
+    "train_samples_per_second": 4.102,
+    "train_steps_per_second": 0.127
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 14.901960784313726,
-    "eval_accuracy": 0.9044117647058824,
-    "eval_loss": 0.20269162952899933,
-    "eval_runtime": 30.8962,
-    "eval_samples_per_second": 4.402,
-    "eval_steps_per_second": 0.55
+    "eval_accuracy": 0.8970588235294118,
+    "eval_loss": 0.30571919679641724,
+    "eval_runtime": 30.5904,
+    "eval_samples_per_second": 4.446,
+    "eval_steps_per_second": 0.556
 }

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 14.901960784313726,
     "total_flos": 4.5342391246258176e+17,
-    "train_loss": 0.4384694022044801,
-    "train_runtime": 4524.6194,
-    "train_samples_per_second": 4.058,
-    "train_steps_per_second": 0.126
+    "train_loss": 0.4511167878644508,
+    "train_runtime": 4476.1321,
+    "train_samples_per_second": 4.102,
+    "train_steps_per_second": 0.127
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.9044117647058824,
+  "best_metric": 0.8970588235294118,
   "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-papsmear/checkpoint-497",
   "epoch": 14.901960784313726,
   "eval_steps": 500,
@@ -10,546 +10,546 @@
   "log_history": [
     {
       "epoch": 0.26143790849673204,
-      "grad_norm": 12.742263793945312,
+      "grad_norm": 15.00976276397705,
       "learning_rate": 8.771929824561403e-06,
-      "loss": 1.84,
+      "loss": 1.7737,
       "step": 10
     },
     {
       "epoch": 0.5228758169934641,
-      "grad_norm": 12.05686092376709,
+      "grad_norm": 25.949995040893555,
       "learning_rate": 1.7543859649122806e-05,
-      "loss": 1.6658,
+      "loss": 1.633,
       "step": 20
     },
     {
       "epoch": 0.7843137254901961,
-      "grad_norm": 14.378918647766113,
+      "grad_norm": 18.413894653320312,
       "learning_rate": 2.6315789473684212e-05,
-      "loss": 1.4844,
+      "loss": 1.4898,
       "step": 30
     },
     {
       "epoch": 0.9934640522875817,
-      "eval_accuracy": 0.4117647058823529,
-      "eval_loss": 1.429846167564392,
-      "eval_runtime": 31.3549,
-      "eval_samples_per_second": 4.337,
-      "eval_steps_per_second": 0.542,
+      "eval_accuracy": 0.4632352941176471,
+      "eval_loss": 1.3708577156066895,
+      "eval_runtime": 31.5324,
+      "eval_samples_per_second": 4.313,
+      "eval_steps_per_second": 0.539,
       "step": 38
     },
     {
       "epoch": 1.0457516339869282,
-      "grad_norm": 26.927827835083008,
+      "grad_norm": 57.17748260498047,
       "learning_rate": 3.508771929824561e-05,
-      "loss": 1.2683,
+      "loss": 1.3556,
       "step": 40
     },
     {
       "epoch": 1.3071895424836601,
-      "grad_norm": 17.2440185546875,
+      "grad_norm": 32.09730911254883,
       "learning_rate": 4.3859649122807014e-05,
-      "loss": 1.0979,
+      "loss": 1.2202,
       "step": 50
     },
     {
       "epoch": 1.5686274509803921,
-      "grad_norm": 30.65620994567871,
+      "grad_norm": 22.1444034576416,
       "learning_rate": 4.970760233918128e-05,
-      "loss": 1.0747,
+      "loss": 1.08,
       "step": 60
     },
     {
       "epoch": 1.8300653594771243,
-      "grad_norm": 30.137895584106445,
+      "grad_norm": 79.14645385742188,
       "learning_rate": 4.8732943469785574e-05,
-      "loss": 0.8905,
+      "loss": 0.8902,
       "step": 70
     },
     {
       "epoch": 1.9869281045751634,
-      "eval_accuracy": 0.6544117647058824,
-      "eval_loss": 0.8109874129295349,
-      "eval_runtime": 30.5109,
-      "eval_samples_per_second": 4.457,
-      "eval_steps_per_second": 0.557,
+      "eval_accuracy": 0.6323529411764706,
+      "eval_loss": 0.9260798096656799,
+      "eval_runtime": 31.6169,
+      "eval_samples_per_second": 4.301,
+      "eval_steps_per_second": 0.538,
       "step": 76
     },
     {
       "epoch": 2.0915032679738563,
-      "grad_norm": 35.162410736083984,
+      "grad_norm": 36.75314712524414,
       "learning_rate": 4.7758284600389865e-05,
-      "loss": 0.7577,
+      "loss": 0.9146,
       "step": 80
     },
     {
       "epoch": 2.3529411764705883,
-      "grad_norm": 20.342853546142578,
+      "grad_norm": 57.47659683227539,
       "learning_rate": 4.678362573099415e-05,
-      "loss": 0.737,
+      "loss": 0.8231,
       "step": 90
     },
     {
       "epoch": 2.6143790849673203,
-      "grad_norm": 20.470102310180664,
+      "grad_norm": 25.39302635192871,
       "learning_rate": 4.580896686159844e-05,
-      "loss": 0.7883,
+      "loss": 0.8616,
       "step": 100
     },
     {
       "epoch": 2.8758169934640523,
-      "grad_norm": 46.66905212402344,
+      "grad_norm": 17.885847091674805,
       "learning_rate": 4.483430799220273e-05,
-      "loss": 0.9097,
+      "loss": 0.9107,
       "step": 110
     },
     {
       "epoch": 2.980392156862745,
-      "eval_accuracy": 0.7132352941176471,
-      "eval_loss": 0.7109317779541016,
-      "eval_runtime": 31.3915,
-      "eval_samples_per_second": 4.332,
-      "eval_steps_per_second": 0.542,
+      "eval_accuracy": 0.6397058823529411,
+      "eval_loss": 0.8400399088859558,
+      "eval_runtime": 32.195,
+      "eval_samples_per_second": 4.224,
+      "eval_steps_per_second": 0.528,
       "step": 114
     },
     {
       "epoch": 3.1372549019607843,
-      "grad_norm": 20.6967716217041,
+      "grad_norm": 26.590669631958008,
       "learning_rate": 4.3859649122807014e-05,
-      "loss": 0.6869,
+      "loss": 0.7438,
       "step": 120
     },
     {
       "epoch": 3.3986928104575163,
-      "grad_norm": 17.672218322753906,
+      "grad_norm": 28.96077537536621,
       "learning_rate": 4.2884990253411305e-05,
-      "loss": 0.6266,
+      "loss": 0.6823,
       "step": 130
     },
     {
       "epoch": 3.6601307189542482,
-      "grad_norm": 13.8719482421875,
+      "grad_norm": 13.954778671264648,
       "learning_rate": 4.1910331384015596e-05,
-      "loss": 0.521,
+      "loss": 0.5726,
       "step": 140
     },
     {
       "epoch": 3.9215686274509802,
-      "grad_norm": 17.780046463012695,
+      "grad_norm": 25.086530685424805,
       "learning_rate": 4.093567251461988e-05,
-      "loss": 0.6238,
+      "loss": 0.564,
       "step": 150
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.6544117647058824,
-      "eval_loss": 0.9196979403495789,
-      "eval_runtime": 30.0783,
-      "eval_samples_per_second": 4.522,
-      "eval_steps_per_second": 0.565,
+      "eval_accuracy": 0.7279411764705882,
+      "eval_loss": 0.6937077641487122,
+      "eval_runtime": 30.7804,
+      "eval_samples_per_second": 4.418,
+      "eval_steps_per_second": 0.552,
       "step": 153
     },
     {
       "epoch": 4.183006535947713,
-      "grad_norm": 28.253620147705078,
+      "grad_norm": 35.74240493774414,
       "learning_rate": 3.996101364522417e-05,
-      "loss": 0.6683,
+      "loss": 0.5648,
       "step": 160
     },
     {
       "epoch": 4.444444444444445,
-      "grad_norm": 14.647994995117188,
+      "grad_norm": 25.511213302612305,
       "learning_rate": 3.898635477582846e-05,
-      "loss": 0.5111,
+      "loss": 0.5403,
       "step": 170
     },
     {
       "epoch": 4.705882352941177,
-      "grad_norm": 21.987884521484375,
+      "grad_norm": 17.154970169067383,
       "learning_rate": 3.8011695906432746e-05,
-      "loss": 0.4989,
+      "loss": 0.4648,
       "step": 180
     },
     {
       "epoch": 4.967320261437909,
-      "grad_norm": 14.68101692199707,
+      "grad_norm": 13.71211051940918,
       "learning_rate": 3.7037037037037037e-05,
-      "loss": 0.4456,
+      "loss": 0.5563,
       "step": 190
     },
     {
       "epoch": 4.993464052287582,
-      "eval_accuracy": 0.8014705882352942,
-      "eval_loss": 0.46520400047302246,
-      "eval_runtime": 29.9903,
-      "eval_samples_per_second": 4.535,
-      "eval_steps_per_second": 0.567,
+      "eval_accuracy": 0.7647058823529411,
+      "eval_loss": 0.5622314810752869,
+      "eval_runtime": 31.1711,
+      "eval_samples_per_second": 4.363,
+      "eval_steps_per_second": 0.545,
       "step": 191
     },
     {
       "epoch": 5.228758169934641,
-      "grad_norm": 10.403129577636719,
+      "grad_norm": 17.74187469482422,
       "learning_rate": 3.606237816764133e-05,
-      "loss": 0.3867,
+      "loss": 0.4352,
       "step": 200
     },
     {
       "epoch": 5.490196078431373,
-      "grad_norm": 22.698213577270508,
+      "grad_norm": 19.112417221069336,
       "learning_rate": 3.508771929824561e-05,
-      "loss": 0.3797,
+      "loss": 0.4013,
       "step": 210
     },
     {
       "epoch": 5.751633986928105,
-      "grad_norm": 15.229155540466309,
+      "grad_norm": 33.436283111572266,
       "learning_rate": 3.41130604288499e-05,
-      "loss": 0.4394,
+      "loss": 0.3851,
       "step": 220
     },
     {
       "epoch": 5.9869281045751634,
       "eval_accuracy": 0.8014705882352942,
-      "eval_loss": 0.5187686681747437,
-      "eval_runtime": 31.0968,
-      "eval_samples_per_second": 4.373,
-      "eval_steps_per_second": 0.547,
+      "eval_loss": 0.5238231420516968,
+      "eval_runtime": 30.6722,
+      "eval_samples_per_second": 4.434,
+      "eval_steps_per_second": 0.554,
       "step": 229
     },
     {
       "epoch": 6.0130718954248366,
-      "grad_norm": 15.33947467803955,
+      "grad_norm": 16.213756561279297,
       "learning_rate": 3.313840155945419e-05,
-      "loss": 0.3482,
+      "loss": 0.4214,
       "step": 230
     },
     {
       "epoch": 6.2745098039215685,
-      "grad_norm": 27.592214584350586,
+      "grad_norm": 20.761796951293945,
       "learning_rate": 3.216374269005848e-05,
-      "loss": 0.3717,
+      "loss": 0.3362,
       "step": 240
     },
     {
       "epoch": 6.5359477124183005,
-      "grad_norm": 16.805540084838867,
+      "grad_norm": 15.858070373535156,
       "learning_rate": 3.118908382066277e-05,
-      "loss": 0.3361,
+      "loss": 0.318,
       "step": 250
     },
     {
       "epoch": 6.7973856209150325,
-      "grad_norm": 19.967639923095703,
+      "grad_norm": 16.319021224975586,
       "learning_rate": 3.0214424951267055e-05,
-      "loss": 0.3156,
+      "loss": 0.3327,
       "step": 260
     },
     {
       "epoch": 6.980392156862745,
-      "eval_accuracy": 0.8529411764705882,
-      "eval_loss": 0.34466856718063354,
-      "eval_runtime": 30.7808,
-      "eval_samples_per_second": 4.418,
-      "eval_steps_per_second": 0.552,
+      "eval_accuracy": 0.7941176470588235,
+      "eval_loss": 0.638187050819397,
+      "eval_runtime": 31.2725,
+      "eval_samples_per_second": 4.349,
+      "eval_steps_per_second": 0.544,
       "step": 267
     },
     {
       "epoch": 7.0588235294117645,
-      "grad_norm": 15.27043628692627,
+      "grad_norm": 19.036224365234375,
       "learning_rate": 2.9239766081871346e-05,
-      "loss": 0.284,
+      "loss": 0.3181,
       "step": 270
     },
     {
       "epoch": 7.3202614379084965,
-      "grad_norm": 20.704370498657227,
+      "grad_norm": 17.282264709472656,
       "learning_rate": 2.8265107212475634e-05,
-      "loss": 0.2777,
+      "loss": 0.2694,
       "step": 280
     },
     {
       "epoch": 7.5816993464052285,
-      "grad_norm": 19.92112922668457,
+      "grad_norm": 16.846771240234375,
       "learning_rate": 2.729044834307992e-05,
-      "loss": 0.2911,
+      "loss": 0.2652,
       "step": 290
     },
     {
       "epoch": 7.8431372549019605,
-      "grad_norm": 23.951862335205078,
+      "grad_norm": 44.78040313720703,
       "learning_rate": 2.6315789473684212e-05,
-      "loss": 0.2212,
+      "loss": 0.2469,
       "step": 300
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.8382352941176471,
-      "eval_loss": 0.3509288728237152,
-      "eval_runtime": 30.877,
-      "eval_samples_per_second": 4.405,
-      "eval_steps_per_second": 0.551,
+      "eval_accuracy": 0.8455882352941176,
+      "eval_loss": 0.4329652488231659,
+      "eval_runtime": 31.8839,
+      "eval_samples_per_second": 4.265,
+      "eval_steps_per_second": 0.533,
       "step": 306
     },
     {
       "epoch": 8.104575163398692,
-      "grad_norm": 28.307418823242188,
+      "grad_norm": 14.120306968688965,
       "learning_rate": 2.53411306042885e-05,
-      "loss": 0.221,
+      "loss": 0.2101,
       "step": 310
     },
     {
       "epoch": 8.366013071895425,
-      "grad_norm": 12.33401107788086,
+      "grad_norm": 27.170852661132812,
       "learning_rate": 2.4366471734892787e-05,
-      "loss": 0.1945,
+      "loss": 0.2611,
       "step": 320
     },
     {
       "epoch": 8.627450980392156,
-      "grad_norm": 26.391366958618164,
+      "grad_norm": 15.851548194885254,
       "learning_rate": 2.3391812865497074e-05,
-      "loss": 0.2758,
+      "loss": 0.192,
       "step": 330
     },
     {
       "epoch": 8.88888888888889,
-      "grad_norm": 10.806180000305176,
+      "grad_norm": 21.090044021606445,
       "learning_rate": 2.2417153996101365e-05,
-      "loss": 0.2402,
+      "loss": 0.2903,
       "step": 340
     },
     {
       "epoch": 8.993464052287582,
-      "eval_accuracy": 0.8235294117647058,
-      "eval_loss": 0.393919974565506,
-      "eval_runtime": 31.1996,
-      "eval_samples_per_second": 4.359,
-      "eval_steps_per_second": 0.545,
+      "eval_accuracy": 0.8308823529411765,
+      "eval_loss": 0.42120417952537537,
+      "eval_runtime": 30.9494,
+      "eval_samples_per_second": 4.394,
+      "eval_steps_per_second": 0.549,
       "step": 344
     },
     {
       "epoch": 9.15032679738562,
-      "grad_norm": 16.904836654663086,
+      "grad_norm": 11.140714645385742,
       "learning_rate": 2.1442495126705653e-05,
-      "loss": 0.1835,
+      "loss": 0.2339,
       "step": 350
     },
     {
       "epoch": 9.411764705882353,
-      "grad_norm": 30.91939926147461,
+      "grad_norm": 17.719518661499023,
       "learning_rate": 2.046783625730994e-05,
-      "loss": 0.1562,
+      "loss": 0.1923,
       "step": 360
     },
     {
       "epoch": 9.673202614379084,
-      "grad_norm": 14.6829833984375,
+      "grad_norm": 17.916824340820312,
       "learning_rate": 1.949317738791423e-05,
-      "loss": 0.195,
+      "loss": 0.2337,
       "step": 370
     },
     {
       "epoch": 9.934640522875817,
-      "grad_norm": 26.035619735717773,
+      "grad_norm": 14.586275100708008,
       "learning_rate": 1.8518518518518518e-05,
-      "loss": 0.1733,
+      "loss": 0.1861,
       "step": 380
     },
     {
       "epoch": 9.986928104575163,
-      "eval_accuracy": 0.8897058823529411,
-      "eval_loss": 0.24441416561603546,
-      "eval_runtime": 30.8179,
-      "eval_samples_per_second": 4.413,
-      "eval_steps_per_second": 0.552,
+      "eval_accuracy": 0.8529411764705882,
+      "eval_loss": 0.413960725069046,
+      "eval_runtime": 30.9585,
+      "eval_samples_per_second": 4.393,
+      "eval_steps_per_second": 0.549,
       "step": 382
     },
     {
       "epoch": 10.196078431372548,
-      "grad_norm": 14.318477630615234,
+      "grad_norm": 17.971166610717773,
       "learning_rate": 1.7543859649122806e-05,
-      "loss": 0.1458,
+      "loss": 0.2086,
       "step": 390
     },
     {
       "epoch": 10.457516339869281,
-      "grad_norm": 9.527152061462402,
+      "grad_norm": 24.829071044921875,
       "learning_rate": 1.6569200779727097e-05,
-      "loss": 0.1752,
+      "loss": 0.1588,
       "step": 400
     },
     {
       "epoch": 10.718954248366012,
-      "grad_norm": 22.481767654418945,
+      "grad_norm": 52.23923873901367,
       "learning_rate": 1.5594541910331384e-05,
-      "loss": 0.1969,
+      "loss": 0.1667,
       "step": 410
     },
     {
       "epoch": 10.980392156862745,
-      "grad_norm": 42.72972869873047,
+      "grad_norm": 11.44646167755127,
       "learning_rate": 1.4619883040935673e-05,
-      "loss": 0.1953,
+      "loss": 0.1533,
       "step": 420
     },
     {
       "epoch": 10.980392156862745,
-      "eval_accuracy": 0.8676470588235294,
-      "eval_loss": 0.26393771171569824,
-      "eval_runtime": 31.9368,
-      "eval_samples_per_second": 4.258,
-      "eval_steps_per_second": 0.532,
+      "eval_accuracy": 0.8602941176470589,
+      "eval_loss": 0.38100317120552063,
+      "eval_runtime": 31.4123,
+      "eval_samples_per_second": 4.33,
+      "eval_steps_per_second": 0.541,
       "step": 420
     },
     {
       "epoch": 11.241830065359476,
-      "grad_norm": 10.306249618530273,
+      "grad_norm": 5.547791004180908,
       "learning_rate": 1.364522417153996e-05,
-      "loss": 0.1626,
+      "loss": 0.121,
       "step": 430
     },
     {
       "epoch": 11.50326797385621,
-      "grad_norm": 9.890297889709473,
+      "grad_norm": 10.899214744567871,
       "learning_rate": 1.267056530214425e-05,
-      "loss": 0.1885,
+      "loss": 0.1578,
       "step": 440
     },
     {
       "epoch": 11.764705882352942,
-      "grad_norm": 13.813959121704102,
+      "grad_norm": 11.240522384643555,
       "learning_rate": 1.1695906432748537e-05,
-      "loss": 0.1363,
+      "loss": 0.1017,
       "step": 450
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.8823529411764706,
-      "eval_loss": 0.264517605304718,
-      "eval_runtime": 30.4905,
-      "eval_samples_per_second": 4.46,
-      "eval_steps_per_second": 0.558,
+      "eval_accuracy": 0.8602941176470589,
+      "eval_loss": 0.35648760199546814,
+      "eval_runtime": 30.8241,
+      "eval_samples_per_second": 4.412,
+      "eval_steps_per_second": 0.552,
       "step": 459
     },
     {
       "epoch": 12.026143790849673,
-      "grad_norm": 8.809865951538086,
+      "grad_norm": 9.531538963317871,
       "learning_rate": 1.0721247563352826e-05,
-      "loss": 0.0897,
+      "loss": 0.1435,
       "step": 460
     },
     {
       "epoch": 12.287581699346406,
-      "grad_norm": 9.379281997680664,
+      "grad_norm": 17.950599670410156,
       "learning_rate": 9.746588693957115e-06,
-      "loss": 0.0972,
+      "loss": 0.1336,
       "step": 470
     },
     {
       "epoch": 12.549019607843137,
-      "grad_norm": 42.292972564697266,
+      "grad_norm": 16.953811645507812,
       "learning_rate": 8.771929824561403e-06,
-      "loss": 0.1417,
+      "loss": 0.0987,
       "step": 480
     },
     {
       "epoch": 12.81045751633987,
-      "grad_norm": 16.6181697845459,
+      "grad_norm": 11.203367233276367,
       "learning_rate": 7.797270955165692e-06,
-      "loss": 0.1234,
+      "loss": 0.1285,
       "step": 490
     },
     {
       "epoch": 12.993464052287582,
-      "eval_accuracy": 0.9044117647058824,
-      "eval_loss": 0.20269162952899933,
-      "eval_runtime": 30.5106,
-      "eval_samples_per_second": 4.457,
-      "eval_steps_per_second": 0.557,
+      "eval_accuracy": 0.8970588235294118,
+      "eval_loss": 0.30571919679641724,
+      "eval_runtime": 31.345,
+      "eval_samples_per_second": 4.339,
+      "eval_steps_per_second": 0.542,
       "step": 497
     },
     {
       "epoch": 13.071895424836601,
-      "grad_norm": 28.557069778442383,
+      "grad_norm": 12.97030258178711,
       "learning_rate": 6.82261208576998e-06,
-      "loss": 0.1402,
+      "loss": 0.1266,
       "step": 500
     },
     {
       "epoch": 13.333333333333334,
-      "grad_norm": 22.95408058166504,
+      "grad_norm": 22.279695510864258,
       "learning_rate": 5.8479532163742686e-06,
-      "loss": 0.1396,
+      "loss": 0.0987,
       "step": 510
     },
     {
       "epoch": 13.594771241830065,
-      "grad_norm": 17.663022994995117,
+      "grad_norm": 20.224885940551758,
       "learning_rate": 4.873294346978558e-06,
-      "loss": 0.0941,
+      "loss": 0.135,
       "step": 520
     },
     {
       "epoch": 13.856209150326798,
-      "grad_norm": 37.5713996887207,
+      "grad_norm": 19.785856246948242,
       "learning_rate": 3.898635477582846e-06,
-      "loss": 0.1282,
+      "loss": 0.1377,
       "step": 530
     },
     {
       "epoch": 13.986928104575163,
-      "eval_accuracy": 0.9044117647058824,
-      "eval_loss": 0.20270010828971863,
-      "eval_runtime": 30.5276,
-      "eval_samples_per_second": 4.455,
-      "eval_steps_per_second": 0.557,
+      "eval_accuracy": 0.8823529411764706,
+      "eval_loss": 0.3058066666126251,
+      "eval_runtime": 30.6745,
+      "eval_samples_per_second": 4.434,
+      "eval_steps_per_second": 0.554,
       "step": 535
     },
     {
       "epoch": 14.117647058823529,
-      "grad_norm": 15.502400398254395,
+      "grad_norm": 24.332889556884766,
       "learning_rate": 2.9239766081871343e-06,
-      "loss": 0.1222,
+      "loss": 0.128,
       "step": 540
     },
     {
       "epoch": 14.379084967320262,
-      "grad_norm": 30.627925872802734,
+      "grad_norm": 18.584924697875977,
       "learning_rate": 1.949317738791423e-06,
-      "loss": 0.1129,
+      "loss": 0.1116,
       "step": 550
     },
     {
       "epoch": 14.640522875816993,
-      "grad_norm": 19.545286178588867,
+      "grad_norm": 45.670799255371094,
       "learning_rate": 9.746588693957115e-07,
-      "loss": 0.1122,
+      "loss": 0.1302,
       "step": 560
     },
     {
       "epoch": 14.901960784313726,
-      "grad_norm": 8.540277481079102,
+      "grad_norm": 13.613533973693848,
       "learning_rate": 0.0,
-      "loss": 0.1036,
+      "loss": 0.1033,
       "step": 570
     },
     {
       "epoch": 14.901960784313726,
-      "eval_accuracy": 0.8897058823529411,
-      "eval_loss": 0.20258459448814392,
-      "eval_runtime": 30.3873,
-      "eval_samples_per_second": 4.476,
-      "eval_steps_per_second": 0.559,
+      "eval_accuracy": 0.8823529411764706,
+      "eval_loss": 0.3140135705471039,
+      "eval_runtime": 30.9583,
+      "eval_samples_per_second": 4.393,
+      "eval_steps_per_second": 0.549,
       "step": 570
     },
     {
       "epoch": 14.901960784313726,
       "step": 570,
       "total_flos": 4.5342391246258176e+17,
-      "train_loss": 0.4384694022044801,
-      "train_runtime": 4524.6194,
-      "train_samples_per_second": 4.058,
-      "train_steps_per_second": 0.126
+      "train_loss": 0.4511167878644508,
+      "train_runtime": 4476.1321,
+      "train_samples_per_second": 4.102,
+      "train_steps_per_second": 0.127
     }
   ],
   "logging_steps": 10,