🍻 cheers

README.md

@@ -2,6 +2,7 @@
 license: apache-2.0
 base_model: google/vit-base-patch16-224-in21k
 tags:
+- image-classification
 - generated_from_trainer
 metrics:
 - accuracy
@@ -15,10 +16,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # Action_all_10_class
 
-This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset.
+This model is a fine-tuned version of [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the Action_small_dataset dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3598
-- Accuracy: 0.9017
+- Loss: 0.3587
+- Accuracy: 0.8946
 
 ## Model description
 

all_results.json

@@ -1,8 +1,13 @@
 {
     "epoch": 5.0,
+    "eval_accuracy": 0.8945868945868946,
+    "eval_loss": 0.3586599826812744,
+    "eval_runtime": 12.4844,
+    "eval_samples_per_second": 56.23,
+    "eval_steps_per_second": 7.049,
     "total_flos": 1.539101261655982e+18,
-    "train_loss": 0.6643937945844658,
-    "train_runtime": 748.7697,
-    "train_samples_per_second": 26.524,
-    "train_steps_per_second": 1.663
+    "train_loss": 0.6706935805968013,
+    "train_runtime": 749.324,
+    "train_samples_per_second": 26.504,
+    "train_steps_per_second": 1.661
 }

eval_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 5.0,
+    "eval_accuracy": 0.8945868945868946,
+    "eval_loss": 0.3586599826812744,
+    "eval_runtime": 12.4844,
+    "eval_samples_per_second": 56.23,
+    "eval_steps_per_second": 7.049
+}

runs/Nov06_15-48-14_dfb0c3665f64/events.out.tfevents.1699286522.dfb0c3665f64.18450.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:cb34ed9597a7a8ee3a53f1be04b579f3923496da01a7c2b12e565483683c97cf
+size 411

train_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 5.0,
     "total_flos": 1.539101261655982e+18,
-    "train_loss": 0.6643937945844658,
-    "train_runtime": 748.7697,
-    "train_samples_per_second": 26.524,
-    "train_steps_per_second": 1.663
+    "train_loss": 0.6706935805968013,
+    "train_runtime": 749.324,
+    "train_samples_per_second": 26.504,
+    "train_steps_per_second": 1.661
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.39916983246803284,
-  "best_model_checkpoint": "Action_all_10_class/checkpoint-1200",
+  "best_metric": 0.3586599826812744,
+  "best_model_checkpoint": "Action_all_10_class/checkpoint-1000",
   "epoch": 5.0,
   "eval_steps": 100,
   "global_step": 1245,
@@ -11,863 +11,863 @@
     {
       "epoch": 0.04,
       "learning_rate": 0.00019839357429718877,
-      "loss": 2.2294,
+      "loss": 2.1986,
       "step": 10
     },
     {
       "epoch": 0.08,
       "learning_rate": 0.00019678714859437752,
-      "loss": 2.1036,
+      "loss": 2.0053,
       "step": 20
     },
     {
       "epoch": 0.12,
       "learning_rate": 0.00019518072289156628,
-      "loss": 1.9019,
+      "loss": 1.8025,
       "step": 30
     },
     {
       "epoch": 0.16,
       "learning_rate": 0.00019357429718875504,
-      "loss": 1.6396,
+      "loss": 1.577,
       "step": 40
     },
     {
       "epoch": 0.2,
       "learning_rate": 0.00019196787148594377,
-      "loss": 1.5942,
+      "loss": 1.5746,
       "step": 50
     },
     {
       "epoch": 0.24,
       "learning_rate": 0.00019036144578313252,
-      "loss": 1.3722,
+      "loss": 1.4402,
       "step": 60
     },
     {
       "epoch": 0.28,
       "learning_rate": 0.00018875502008032128,
-      "loss": 1.2927,
+      "loss": 1.3624,
       "step": 70
     },
     {
       "epoch": 0.32,
       "learning_rate": 0.00018714859437751004,
-      "loss": 1.2947,
+      "loss": 1.2849,
       "step": 80
     },
     {
       "epoch": 0.36,
       "learning_rate": 0.0001855421686746988,
-      "loss": 1.2353,
+      "loss": 1.2064,
       "step": 90
     },
     {
       "epoch": 0.4,
       "learning_rate": 0.00018393574297188755,
-      "loss": 1.1348,
+      "loss": 1.1524,
       "step": 100
     },
     {
       "epoch": 0.4,
-      "eval_accuracy": 0.698005698005698,
-      "eval_loss": 1.0964313745498657,
-      "eval_runtime": 13.3411,
-      "eval_samples_per_second": 52.619,
-      "eval_steps_per_second": 6.596,
+      "eval_accuracy": 0.6951566951566952,
+      "eval_loss": 1.0612293481826782,
+      "eval_runtime": 13.0271,
+      "eval_samples_per_second": 53.888,
+      "eval_steps_per_second": 6.755,
       "step": 100
     },
     {
       "epoch": 0.44,
       "learning_rate": 0.0001823293172690763,
-      "loss": 1.0305,
+      "loss": 1.085,
       "step": 110
     },
     {
       "epoch": 0.48,
       "learning_rate": 0.00018072289156626507,
-      "loss": 1.0352,
+      "loss": 0.9779,
       "step": 120
     },
     {
       "epoch": 0.52,
       "learning_rate": 0.00017911646586345382,
-      "loss": 1.0957,
+      "loss": 1.0527,
       "step": 130
     },
     {
       "epoch": 0.56,
       "learning_rate": 0.00017751004016064258,
-      "loss": 1.1747,
+      "loss": 1.2653,
       "step": 140
     },
     {
       "epoch": 0.6,
       "learning_rate": 0.00017590361445783134,
-      "loss": 1.0863,
+      "loss": 1.0224,
       "step": 150
     },
     {
       "epoch": 0.64,
       "learning_rate": 0.0001742971887550201,
-      "loss": 1.0802,
+      "loss": 1.0313,
       "step": 160
     },
     {
       "epoch": 0.68,
       "learning_rate": 0.00017269076305220885,
-      "loss": 1.0912,
+      "loss": 1.1361,
       "step": 170
     },
     {
       "epoch": 0.72,
       "learning_rate": 0.0001710843373493976,
-      "loss": 1.0056,
+      "loss": 1.0008,
       "step": 180
     },
     {
       "epoch": 0.76,
       "learning_rate": 0.00016947791164658636,
-      "loss": 0.9641,
+      "loss": 1.0386,
       "step": 190
     },
     {
       "epoch": 0.8,
       "learning_rate": 0.00016787148594377512,
-      "loss": 0.9992,
+      "loss": 0.9818,
       "step": 200
     },
     {
       "epoch": 0.8,
-      "eval_accuracy": 0.7948717948717948,
-      "eval_loss": 0.7362096309661865,
-      "eval_runtime": 13.2093,
-      "eval_samples_per_second": 53.145,
-      "eval_steps_per_second": 6.662,
+      "eval_accuracy": 0.7877492877492878,
+      "eval_loss": 0.7399429082870483,
+      "eval_runtime": 13.2513,
+      "eval_samples_per_second": 52.976,
+      "eval_steps_per_second": 6.641,
       "step": 200
     },
     {
       "epoch": 0.84,
       "learning_rate": 0.00016626506024096388,
-      "loss": 0.9851,
+      "loss": 1.058,
       "step": 210
     },
     {
       "epoch": 0.88,
       "learning_rate": 0.00016465863453815263,
-      "loss": 0.886,
+      "loss": 0.9421,
       "step": 220
     },
     {
       "epoch": 0.92,
       "learning_rate": 0.0001630522088353414,
-      "loss": 0.9604,
+      "loss": 0.9055,
       "step": 230
     },
     {
       "epoch": 0.96,
       "learning_rate": 0.00016144578313253015,
-      "loss": 0.7707,
+      "loss": 0.7195,
       "step": 240
     },
     {
       "epoch": 1.0,
-      "learning_rate": 0.00016,
-      "loss": 0.9153,
+      "learning_rate": 0.00015983935742971888,
+      "loss": 0.9547,
       "step": 250
     },
     {
       "epoch": 1.04,
-      "learning_rate": 0.00015839357429718874,
-      "loss": 0.7581,
+      "learning_rate": 0.00015823293172690763,
+      "loss": 0.7786,
       "step": 260
     },
     {
       "epoch": 1.08,
-      "learning_rate": 0.0001567871485943775,
-      "loss": 0.6622,
+      "learning_rate": 0.0001566265060240964,
+      "loss": 0.7334,
       "step": 270
     },
     {
       "epoch": 1.12,
-      "learning_rate": 0.00015518072289156626,
-      "loss": 0.7251,
+      "learning_rate": 0.00015502008032128515,
+      "loss": 0.8422,
       "step": 280
     },
     {
       "epoch": 1.16,
-      "learning_rate": 0.00015357429718875501,
-      "loss": 0.7955,
+      "learning_rate": 0.0001534136546184739,
+      "loss": 0.7638,
       "step": 290
     },
     {
       "epoch": 1.2,
-      "learning_rate": 0.00015196787148594377,
-      "loss": 0.8314,
+      "learning_rate": 0.00015180722891566266,
+      "loss": 0.844,
       "step": 300
     },
     {
       "epoch": 1.2,
-      "eval_accuracy": 0.8205128205128205,
-      "eval_loss": 0.6410418748855591,
-      "eval_runtime": 12.6655,
-      "eval_samples_per_second": 55.426,
-      "eval_steps_per_second": 6.948,
+      "eval_accuracy": 0.7692307692307693,
+      "eval_loss": 0.7196186780929565,
+      "eval_runtime": 12.9868,
+      "eval_samples_per_second": 54.055,
+      "eval_steps_per_second": 6.776,
       "step": 300
     },
     {
       "epoch": 1.24,
-      "learning_rate": 0.00015036144578313253,
-      "loss": 0.845,
+      "learning_rate": 0.00015020080321285142,
+      "loss": 0.812,
       "step": 310
     },
     {
       "epoch": 1.29,
-      "learning_rate": 0.00014875502008032128,
-      "loss": 0.7132,
+      "learning_rate": 0.00014859437751004018,
+      "loss": 0.801,
       "step": 320
     },
     {
       "epoch": 1.33,
-      "learning_rate": 0.00014714859437751004,
-      "loss": 0.7161,
+      "learning_rate": 0.00014698795180722893,
+      "loss": 0.72,
       "step": 330
     },
     {
       "epoch": 1.37,
-      "learning_rate": 0.0001455421686746988,
-      "loss": 0.7773,
+      "learning_rate": 0.0001453815261044177,
+      "loss": 0.7235,
       "step": 340
     },
     {
       "epoch": 1.41,
-      "learning_rate": 0.00014393574297188756,
-      "loss": 0.7676,
+      "learning_rate": 0.00014377510040160642,
+      "loss": 0.7429,
       "step": 350
     },
     {
       "epoch": 1.45,
-      "learning_rate": 0.0001423293172690763,
-      "loss": 0.8516,
+      "learning_rate": 0.00014216867469879518,
+      "loss": 0.8045,
       "step": 360
     },
     {
       "epoch": 1.49,
-      "learning_rate": 0.00014072289156626507,
-      "loss": 0.6597,
+      "learning_rate": 0.00014056224899598393,
+      "loss": 0.6534,
       "step": 370
     },
     {
       "epoch": 1.53,
-      "learning_rate": 0.00013911646586345383,
-      "loss": 0.6998,
+      "learning_rate": 0.0001389558232931727,
+      "loss": 0.6554,
       "step": 380
     },
     {
       "epoch": 1.57,
-      "learning_rate": 0.00013751004016064258,
-      "loss": 0.8191,
+      "learning_rate": 0.00013734939759036145,
+      "loss": 0.7605,
       "step": 390
     },
     {
       "epoch": 1.61,
-      "learning_rate": 0.00013590361445783134,
-      "loss": 0.7359,
+      "learning_rate": 0.0001357429718875502,
+      "loss": 0.7992,
       "step": 400
     },
     {
       "epoch": 1.61,
-      "eval_accuracy": 0.8247863247863247,
-      "eval_loss": 0.5803518891334534,
-      "eval_runtime": 12.2034,
-      "eval_samples_per_second": 57.525,
-      "eval_steps_per_second": 7.211,
+      "eval_accuracy": 0.8376068376068376,
+      "eval_loss": 0.5383489727973938,
+      "eval_runtime": 13.2408,
+      "eval_samples_per_second": 53.018,
+      "eval_steps_per_second": 6.646,
       "step": 400
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.0001342971887550201,
-      "loss": 0.7354,
+      "learning_rate": 0.00013413654618473896,
+      "loss": 0.8029,
       "step": 410
     },
     {
       "epoch": 1.69,
-      "learning_rate": 0.00013269076305220885,
-      "loss": 0.7507,
+      "learning_rate": 0.00013253012048192772,
+      "loss": 0.7173,
       "step": 420
     },
     {
       "epoch": 1.73,
-      "learning_rate": 0.0001310843373493976,
-      "loss": 0.8088,
+      "learning_rate": 0.00013092369477911648,
+      "loss": 0.7658,
       "step": 430
     },
     {
       "epoch": 1.77,
-      "learning_rate": 0.00012947791164658637,
-      "loss": 0.7279,
+      "learning_rate": 0.00012931726907630523,
+      "loss": 0.6767,
       "step": 440
     },
     {
       "epoch": 1.81,
-      "learning_rate": 0.00012787148594377512,
-      "loss": 0.7254,
+      "learning_rate": 0.00012771084337349396,
+      "loss": 0.7436,
       "step": 450
     },
     {
       "epoch": 1.85,
-      "learning_rate": 0.00012626506024096385,
-      "loss": 0.6707,
+      "learning_rate": 0.00012610441767068272,
+      "loss": 0.679,
       "step": 460
     },
     {
       "epoch": 1.89,
-      "learning_rate": 0.0001246586345381526,
-      "loss": 0.6482,
+      "learning_rate": 0.00012449799196787148,
+      "loss": 0.6192,
       "step": 470
     },
     {
       "epoch": 1.93,
-      "learning_rate": 0.00012305220883534137,
-      "loss": 0.7238,
+      "learning_rate": 0.00012289156626506023,
+      "loss": 0.7447,
       "step": 480
     },
     {
       "epoch": 1.97,
-      "learning_rate": 0.00012144578313253012,
-      "loss": 0.6626,
+      "learning_rate": 0.000121285140562249,
+      "loss": 0.7223,
       "step": 490
     },
     {
       "epoch": 2.01,
-      "learning_rate": 0.00011983935742971888,
-      "loss": 0.776,
+      "learning_rate": 0.00011967871485943776,
+      "loss": 0.7203,
       "step": 500
     },
     {
       "epoch": 2.01,
-      "eval_accuracy": 0.8376068376068376,
-      "eval_loss": 0.5489934086799622,
-      "eval_runtime": 12.4349,
-      "eval_samples_per_second": 56.454,
-      "eval_steps_per_second": 7.077,
+      "eval_accuracy": 0.8390313390313391,
+      "eval_loss": 0.5222235321998596,
+      "eval_runtime": 12.8642,
+      "eval_samples_per_second": 54.57,
+      "eval_steps_per_second": 6.841,
       "step": 500
     },
     {
       "epoch": 2.05,
-      "learning_rate": 0.00011823293172690764,
-      "loss": 0.8069,
+      "learning_rate": 0.00011807228915662652,
+      "loss": 0.7294,
       "step": 510
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.0001166265060240964,
-      "loss": 0.5448,
+      "learning_rate": 0.00011646586345381527,
+      "loss": 0.5726,
       "step": 520
     },
     {
       "epoch": 2.13,
-      "learning_rate": 0.00011502008032128515,
-      "loss": 0.7318,
+      "learning_rate": 0.00011485943775100403,
+      "loss": 0.6862,
       "step": 530
     },
     {
       "epoch": 2.17,
-      "learning_rate": 0.00011341365461847391,
-      "loss": 0.5083,
+      "learning_rate": 0.00011325301204819279,
+      "loss": 0.5806,
       "step": 540
     },
     {
       "epoch": 2.21,
-      "learning_rate": 0.00011180722891566267,
-      "loss": 0.6493,
+      "learning_rate": 0.00011164658634538152,
+      "loss": 0.7621,
       "step": 550
     },
     {
       "epoch": 2.25,
-      "learning_rate": 0.0001102008032128514,
-      "loss": 0.4928,
+      "learning_rate": 0.00011004016064257027,
+      "loss": 0.4579,
       "step": 560
     },
     {
       "epoch": 2.29,
-      "learning_rate": 0.00010859437751004015,
-      "loss": 0.5395,
+      "learning_rate": 0.00010843373493975903,
+      "loss": 0.5173,
       "step": 570
     },
     {
       "epoch": 2.33,
-      "learning_rate": 0.00010698795180722891,
-      "loss": 0.5588,
+      "learning_rate": 0.00010682730923694779,
+      "loss": 0.5604,
       "step": 580
     },
     {
       "epoch": 2.37,
-      "learning_rate": 0.00010538152610441767,
-      "loss": 0.5892,
+      "learning_rate": 0.00010522088353413654,
+      "loss": 0.5459,
       "step": 590
     },
     {
       "epoch": 2.41,
-      "learning_rate": 0.00010377510040160642,
-      "loss": 0.614,
+      "learning_rate": 0.0001036144578313253,
+      "loss": 0.5527,
       "step": 600
     },
     {
       "epoch": 2.41,
-      "eval_accuracy": 0.8504273504273504,
-      "eval_loss": 0.5006864070892334,
-      "eval_runtime": 12.2103,
-      "eval_samples_per_second": 57.493,
-      "eval_steps_per_second": 7.207,
+      "eval_accuracy": 0.8376068376068376,
+      "eval_loss": 0.5394238233566284,
+      "eval_runtime": 12.6495,
+      "eval_samples_per_second": 55.496,
+      "eval_steps_per_second": 6.957,
       "step": 600
     },
     {
       "epoch": 2.45,
-      "learning_rate": 0.00010216867469879518,
-      "loss": 0.7658,
+      "learning_rate": 0.00010200803212851406,
+      "loss": 0.7029,
       "step": 610
     },
     {
       "epoch": 2.49,
-      "learning_rate": 0.00010056224899598394,
-      "loss": 0.5152,
+      "learning_rate": 0.00010040160642570282,
+      "loss": 0.5672,
       "step": 620
     },
     {
       "epoch": 2.53,
-      "learning_rate": 9.89558232931727e-05,
-      "loss": 0.4941,
+      "learning_rate": 9.879518072289157e-05,
+      "loss": 0.5614,
       "step": 630
     },
     {
       "epoch": 2.57,
-      "learning_rate": 9.734939759036145e-05,
-      "loss": 0.6831,
+      "learning_rate": 9.718875502008033e-05,
+      "loss": 0.6897,
       "step": 640
     },
     {
       "epoch": 2.61,
-      "learning_rate": 9.574297188755021e-05,
-      "loss": 0.5971,
+      "learning_rate": 9.558232931726909e-05,
+      "loss": 0.5952,
       "step": 650
     },
     {
       "epoch": 2.65,
-      "learning_rate": 9.413654618473896e-05,
-      "loss": 0.5842,
+      "learning_rate": 9.397590361445784e-05,
+      "loss": 0.6168,
       "step": 660
     },
     {
       "epoch": 2.69,
-      "learning_rate": 9.253012048192772e-05,
-      "loss": 0.5976,
+      "learning_rate": 9.23694779116466e-05,
+      "loss": 0.5881,
       "step": 670
     },
     {
       "epoch": 2.73,
-      "learning_rate": 9.092369477911648e-05,
-      "loss": 0.5051,
+      "learning_rate": 9.076305220883534e-05,
+      "loss": 0.464,
       "step": 680
     },
     {
       "epoch": 2.77,
-      "learning_rate": 8.931726907630522e-05,
-      "loss": 0.4737,
+      "learning_rate": 8.91566265060241e-05,
+      "loss": 0.541,
       "step": 690
     },
     {
       "epoch": 2.81,
-      "learning_rate": 8.771084337349398e-05,
-      "loss": 0.5484,
+      "learning_rate": 8.755020080321286e-05,
+      "loss": 0.5682,
       "step": 700
     },
     {
       "epoch": 2.81,
-      "eval_accuracy": 0.8333333333333334,
-      "eval_loss": 0.5322346091270447,
-      "eval_runtime": 12.3228,
-      "eval_samples_per_second": 56.968,
-      "eval_steps_per_second": 7.141,
+      "eval_accuracy": 0.8461538461538461,
+      "eval_loss": 0.4943247437477112,
+      "eval_runtime": 12.6964,
+      "eval_samples_per_second": 55.291,
+      "eval_steps_per_second": 6.931,
       "step": 700
     },
     {
       "epoch": 2.85,
-      "learning_rate": 8.610441767068274e-05,
-      "loss": 0.4436,
+      "learning_rate": 8.594377510040161e-05,
+      "loss": 0.4844,
       "step": 710
     },
     {
       "epoch": 2.89,
-      "learning_rate": 8.449799196787149e-05,
-      "loss": 0.6452,
+      "learning_rate": 8.433734939759037e-05,
+      "loss": 0.6745,
       "step": 720
     },
     {
       "epoch": 2.93,
-      "learning_rate": 8.289156626506025e-05,
-      "loss": 0.5724,
+      "learning_rate": 8.273092369477911e-05,
+      "loss": 0.5597,
       "step": 730
     },
     {
       "epoch": 2.97,
-      "learning_rate": 8.128514056224899e-05,
-      "loss": 0.3933,
+      "learning_rate": 8.112449799196787e-05,
+      "loss": 0.4626,
       "step": 740
     },
     {
       "epoch": 3.01,
-      "learning_rate": 7.967871485943775e-05,
-      "loss": 0.5753,
+      "learning_rate": 7.951807228915663e-05,
+      "loss": 0.6571,
       "step": 750
     },
     {
       "epoch": 3.05,
-      "learning_rate": 7.80722891566265e-05,
-      "loss": 0.4426,
+      "learning_rate": 7.791164658634539e-05,
+      "loss": 0.4811,
       "step": 760
     },
     {
       "epoch": 3.09,
-      "learning_rate": 7.646586345381526e-05,
-      "loss": 0.5442,
+      "learning_rate": 7.630522088353414e-05,
+      "loss": 0.5323,
       "step": 770
     },
     {
       "epoch": 3.13,
-      "learning_rate": 7.485943775100402e-05,
-      "loss": 0.4839,
+      "learning_rate": 7.469879518072289e-05,
+      "loss": 0.5492,
       "step": 780
     },
     {
       "epoch": 3.17,
-      "learning_rate": 7.325301204819278e-05,
-      "loss": 0.3711,
+      "learning_rate": 7.309236947791164e-05,
+      "loss": 0.4278,
       "step": 790
     },
     {
       "epoch": 3.21,
-      "learning_rate": 7.164658634538153e-05,
-      "loss": 0.3844,
+      "learning_rate": 7.14859437751004e-05,
+      "loss": 0.4238,
       "step": 800
     },
     {
       "epoch": 3.21,
-      "eval_accuracy": 0.8518518518518519,
-      "eval_loss": 0.5012323260307312,
-      "eval_runtime": 12.9006,
-      "eval_samples_per_second": 54.416,
-      "eval_steps_per_second": 6.821,
+      "eval_accuracy": 0.8675213675213675,
+      "eval_loss": 0.4441128373146057,
+      "eval_runtime": 12.9024,
+      "eval_samples_per_second": 54.409,
+      "eval_steps_per_second": 6.82,
       "step": 800
     },
     {
       "epoch": 3.25,
-      "learning_rate": 7.004016064257029e-05,
-      "loss": 0.513,
+      "learning_rate": 6.987951807228917e-05,
+      "loss": 0.5329,
       "step": 810
     },
     {
       "epoch": 3.29,
-      "learning_rate": 6.843373493975905e-05,
-      "loss": 0.3505,
+      "learning_rate": 6.827309236947793e-05,
+      "loss": 0.3775,
       "step": 820
     },
     {
       "epoch": 3.33,
-      "learning_rate": 6.68273092369478e-05,
-      "loss": 0.4022,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 0.4408,
       "step": 830
     },
     {
       "epoch": 3.37,
-      "learning_rate": 6.522088353413655e-05,
-      "loss": 0.3574,
+      "learning_rate": 6.506024096385543e-05,
+      "loss": 0.3515,
       "step": 840
     },
     {
       "epoch": 3.41,
-      "learning_rate": 6.36144578313253e-05,
-      "loss": 0.5703,
+      "learning_rate": 6.345381526104418e-05,
+      "loss": 0.5296,
       "step": 850
     },
     {
       "epoch": 3.45,
-      "learning_rate": 6.200803212851406e-05,
-      "loss": 0.4585,
+      "learning_rate": 6.184738955823294e-05,
+      "loss": 0.4531,
       "step": 860
     },
     {
       "epoch": 3.49,
-      "learning_rate": 6.040160642570282e-05,
-      "loss": 0.3114,
+      "learning_rate": 6.02409638554217e-05,
+      "loss": 0.3878,
       "step": 870
     },
     {
       "epoch": 3.53,
-      "learning_rate": 5.8795180722891576e-05,
-      "loss": 0.3508,
+      "learning_rate": 5.863453815261044e-05,
+      "loss": 0.4233,
       "step": 880
     },
     {
       "epoch": 3.57,
-      "learning_rate": 5.718875502008032e-05,
-      "loss": 0.3614,
+      "learning_rate": 5.70281124497992e-05,
+      "loss": 0.3732,
       "step": 890
     },
     {
       "epoch": 3.61,
-      "learning_rate": 5.5582329317269076e-05,
-      "loss": 0.5681,
+      "learning_rate": 5.5421686746987955e-05,
+      "loss": 0.4684,
       "step": 900
     },
     {
       "epoch": 3.61,
-      "eval_accuracy": 0.8589743589743589,
-      "eval_loss": 0.49228861927986145,
-      "eval_runtime": 13.1999,
-      "eval_samples_per_second": 53.182,
-      "eval_steps_per_second": 6.667,
+      "eval_accuracy": 0.8846153846153846,
+      "eval_loss": 0.41125282645225525,
+      "eval_runtime": 13.1688,
+      "eval_samples_per_second": 53.308,
+      "eval_steps_per_second": 6.682,
       "step": 900
     },
     {
       "epoch": 3.65,
-      "learning_rate": 5.397590361445783e-05,
-      "loss": 0.5074,
+      "learning_rate": 5.381526104417671e-05,
+      "loss": 0.4347,
       "step": 910
     },
     {
       "epoch": 3.69,
-      "learning_rate": 5.236947791164659e-05,
-      "loss": 0.3371,
+      "learning_rate": 5.220883534136547e-05,
+      "loss": 0.3793,
       "step": 920
     },
     {
       "epoch": 3.73,
-      "learning_rate": 5.076305220883535e-05,
-      "loss": 0.5226,
+      "learning_rate": 5.060240963855422e-05,
+      "loss": 0.4689,
       "step": 930
     },
     {
       "epoch": 3.78,
       "learning_rate": 4.9156626506024104e-05,
-      "loss": 0.4462,
+      "loss": 0.4459,
       "step": 940
     },
     {
       "epoch": 3.82,
       "learning_rate": 4.7550200803212854e-05,
-      "loss": 0.4781,
+      "loss": 0.4529,
       "step": 950
     },
     {
       "epoch": 3.86,
       "learning_rate": 4.594377510040161e-05,
-      "loss": 0.4001,
+      "loss": 0.3795,
       "step": 960
     },
     {
       "epoch": 3.9,
       "learning_rate": 4.433734939759036e-05,
-      "loss": 0.4788,
+      "loss": 0.4545,
       "step": 970
     },
     {
       "epoch": 3.94,
       "learning_rate": 4.273092369477912e-05,
-      "loss": 0.4561,
+      "loss": 0.4298,
       "step": 980
     },
     {
       "epoch": 3.98,
       "learning_rate": 4.1124497991967875e-05,
-      "loss": 0.2563,
+      "loss": 0.2978,
       "step": 990
     },
     {
       "epoch": 4.02,
       "learning_rate": 3.9518072289156625e-05,
-      "loss": 0.4315,
+      "loss": 0.4239,
       "step": 1000
     },
     {
       "epoch": 4.02,
-      "eval_accuracy": 0.8774928774928775,
-      "eval_loss": 0.42575880885124207,
-      "eval_runtime": 13.0968,
-      "eval_samples_per_second": 53.601,
-      "eval_steps_per_second": 6.719,
+      "eval_accuracy": 0.8945868945868946,
+      "eval_loss": 0.3586599826812744,
+      "eval_runtime": 12.9226,
+      "eval_samples_per_second": 54.323,
+      "eval_steps_per_second": 6.81,
       "step": 1000
     },
     {
       "epoch": 4.06,
       "learning_rate": 3.791164658634538e-05,
-      "loss": 0.2898,
+      "loss": 0.2987,
       "step": 1010
     },
     {
       "epoch": 4.1,
       "learning_rate": 3.630522088353414e-05,
-      "loss": 0.3803,
+      "loss": 0.3289,
       "step": 1020
     },
     {
       "epoch": 4.14,
       "learning_rate": 3.4698795180722896e-05,
-      "loss": 0.2827,
+      "loss": 0.3527,
       "step": 1030
     },
     {
       "epoch": 4.18,
       "learning_rate": 3.309236947791165e-05,
-      "loss": 0.3382,
+      "loss": 0.3239,
       "step": 1040
     },
     {
       "epoch": 4.22,
       "learning_rate": 3.14859437751004e-05,
-      "loss": 0.3808,
+      "loss": 0.4095,
       "step": 1050
     },
     {
       "epoch": 4.26,
       "learning_rate": 2.987951807228916e-05,
-      "loss": 0.2551,
+      "loss": 0.2874,
       "step": 1060
     },
     {
       "epoch": 4.3,
       "learning_rate": 2.827309236947791e-05,
-      "loss": 0.3011,
+      "loss": 0.3045,
       "step": 1070
     },
     {
       "epoch": 4.34,
       "learning_rate": 2.6666666666666667e-05,
-      "loss": 0.4091,
+      "loss": 0.3476,
       "step": 1080
     },
     {
       "epoch": 4.38,
       "learning_rate": 2.5060240963855423e-05,
-      "loss": 0.286,
+      "loss": 0.3654,
       "step": 1090
     },
     {
       "epoch": 4.42,
       "learning_rate": 2.345381526104418e-05,
-      "loss": 0.3804,
+      "loss": 0.4083,
       "step": 1100
     },
     {
       "epoch": 4.42,
-      "eval_accuracy": 0.8675213675213675,
-      "eval_loss": 0.4324430227279663,
-      "eval_runtime": 12.1257,
-      "eval_samples_per_second": 57.894,
-      "eval_steps_per_second": 7.257,
+      "eval_accuracy": 0.8831908831908832,
+      "eval_loss": 0.3767581582069397,
+      "eval_runtime": 12.931,
+      "eval_samples_per_second": 54.288,
+      "eval_steps_per_second": 6.805,
       "step": 1100
     },
     {
       "epoch": 4.46,
       "learning_rate": 2.1847389558232934e-05,
-      "loss": 0.2672,
+      "loss": 0.3791,
       "step": 1110
     },
     {
       "epoch": 4.5,
       "learning_rate": 2.0240963855421687e-05,
-      "loss": 0.3661,
+      "loss": 0.4041,
       "step": 1120
     },
     {
       "epoch": 4.54,
       "learning_rate": 1.863453815261044e-05,
-      "loss": 0.3126,
+      "loss": 0.3001,
       "step": 1130
     },
     {
       "epoch": 4.58,
       "learning_rate": 1.7028112449799198e-05,
-      "loss": 0.3455,
+      "loss": 0.4239,
       "step": 1140
     },
     {
       "epoch": 4.62,
       "learning_rate": 1.5421686746987955e-05,
-      "loss": 0.3604,
+      "loss": 0.3491,
       "step": 1150
     },
     {
       "epoch": 4.66,
       "learning_rate": 1.3815261044176708e-05,
-      "loss": 0.4628,
+      "loss": 0.4658,
       "step": 1160
     },
     {
       "epoch": 4.7,
       "learning_rate": 1.2208835341365463e-05,
-      "loss": 0.4074,
+      "loss": 0.4182,
       "step": 1170
     },
     {
       "epoch": 4.74,
       "learning_rate": 1.0602409638554217e-05,
-      "loss": 0.2512,
+      "loss": 0.2625,
       "step": 1180
     },
     {
       "epoch": 4.78,
       "learning_rate": 8.995983935742972e-06,
-      "loss": 0.2974,
+      "loss": 0.3003,
       "step": 1190
     },
     {
       "epoch": 4.82,
       "learning_rate": 7.389558232931727e-06,
-      "loss": 0.2887,
+      "loss": 0.3541,
       "step": 1200
     },
     {
       "epoch": 4.82,
-      "eval_accuracy": 0.8803418803418803,
-      "eval_loss": 0.39916983246803284,
-      "eval_runtime": 12.2383,
-      "eval_samples_per_second": 57.361,
-      "eval_steps_per_second": 7.191,
+      "eval_accuracy": 0.9017094017094017,
+      "eval_loss": 0.35980701446533203,
+      "eval_runtime": 13.1716,
+      "eval_samples_per_second": 53.297,
+      "eval_steps_per_second": 6.681,
       "step": 1200
     },
     {
       "epoch": 4.86,
       "learning_rate": 5.783132530120483e-06,
-      "loss": 0.3073,
+      "loss": 0.3299,
       "step": 1210
     },
     {
       "epoch": 4.9,
       "learning_rate": 4.176706827309238e-06,
-      "loss": 0.2825,
+      "loss": 0.2998,
       "step": 1220
     },
     {
       "epoch": 4.94,
       "learning_rate": 2.570281124497992e-06,
-      "loss": 0.288,
+      "loss": 0.3439,
       "step": 1230
     },
     {
       "epoch": 4.98,
       "learning_rate": 9.638554216867472e-07,
-      "loss": 0.3198,
+      "loss": 0.308,
       "step": 1240
     },
     {
       "epoch": 5.0,
       "step": 1245,
       "total_flos": 1.539101261655982e+18,
-      "train_loss": 0.6643937945844658,
-      "train_runtime": 748.7697,
-      "train_samples_per_second": 26.524,
-      "train_steps_per_second": 1.663
+      "train_loss": 0.6706935805968013,
+      "train_runtime": 749.324,
+      "train_samples_per_second": 26.504,
+      "train_steps_per_second": 1.661
     }
   ],
   "logging_steps": 10,