End of training

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 2.99,
-    "eval_accuracy": 0.872196261682243,
-    "eval_loss": 0.4010835587978363,
-    "eval_runtime": 85.9556,
-    "eval_samples_per_second": 49.793,
-    "eval_steps_per_second": 1.559,
-    "total_flos": 3.9721912052731453e+18,
-    "train_loss": 0.5681018877148927,
-    "train_runtime": 3778.1283,
-    "train_samples_per_second": 13.592,
-    "train_steps_per_second": 0.106
+    "epoch": 9.99,
+    "eval_accuracy": 0.8644859813084113,
+    "eval_loss": 0.4172683358192444,
+    "eval_runtime": 69.3033,
+    "eval_samples_per_second": 61.758,
+    "eval_steps_per_second": 1.934,
+    "total_flos": 1.3257453564799912e+19,
+    "train_loss": 0.3000902961967583,
+    "train_runtime": 6004.4248,
+    "train_samples_per_second": 28.507,
+    "train_steps_per_second": 0.222
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.99,
-    "eval_accuracy": 0.872196261682243,
-    "eval_loss": 0.4010835587978363,
-    "eval_runtime": 85.9556,
-    "eval_samples_per_second": 49.793,
-    "eval_steps_per_second": 1.559
+    "epoch": 9.99,
+    "eval_accuracy": 0.8644859813084113,
+    "eval_loss": 0.4172683358192444,
+    "eval_runtime": 69.3033,
+    "eval_samples_per_second": 61.758,
+    "eval_steps_per_second": 1.934
 }

runs/Oct10_16-19-39_2398f85786b1/events.out.tfevents.1665425512.2398f85786b1.17.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:777aec781425c5538177a93e24b2fe9d2f738a73529131025cece60e471552ea
+size 363

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 2.99,
-    "total_flos": 3.9721912052731453e+18,
-    "train_loss": 0.5681018877148927,
-    "train_runtime": 3778.1283,
-    "train_samples_per_second": 13.592,
-    "train_steps_per_second": 0.106
+    "epoch": 9.99,
+    "total_flos": 1.3257453564799912e+19,
+    "train_loss": 0.3000902961967583,
+    "train_runtime": 6004.4248,
+    "train_samples_per_second": 28.507,
+    "train_steps_per_second": 0.222
 }

trainer_state.json

@@ -1,286 +1,913 @@
 {
-  "best_metric": 0.872196261682243,
+  "best_metric": 0.8644859813084113,
   "best_model_checkpoint": "vit-base-patch16-224-in21k-finetuned-cassava/checkpoint-399",
-  "epoch": 2.994392523364486,
-  "global_step": 399,
+  "epoch": 9.994392523364485,
+  "global_step": 1330,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.07,
-      "learning_rate": 1.25e-05,
-      "loss": 1.6059,
+      "learning_rate": 3.7593984962406014e-06,
+      "loss": 1.6081,
       "step": 10
     },
     {
       "epoch": 0.15,
-      "learning_rate": 2.5e-05,
-      "loss": 1.381,
+      "learning_rate": 7.518796992481203e-06,
+      "loss": 1.5302,
       "step": 20
     },
     {
       "epoch": 0.22,
-      "learning_rate": 3.7500000000000003e-05,
-      "loss": 1.1472,
+      "learning_rate": 1.1278195488721805e-05,
+      "loss": 1.3659,
       "step": 30
     },
     {
       "epoch": 0.3,
-      "learning_rate": 5e-05,
-      "loss": 1.0363,
+      "learning_rate": 1.5037593984962406e-05,
+      "loss": 1.2237,
       "step": 40
     },
     {
       "epoch": 0.37,
-      "learning_rate": 4.860724233983287e-05,
-      "loss": 0.9067,
+      "learning_rate": 1.8796992481203007e-05,
+      "loss": 1.0832,
       "step": 50
     },
     {
       "epoch": 0.45,
-      "learning_rate": 4.721448467966574e-05,
-      "loss": 0.7733,
+      "learning_rate": 2.255639097744361e-05,
+      "loss": 1.02,
       "step": 60
     },
     {
       "epoch": 0.52,
-      "learning_rate": 4.582172701949861e-05,
-      "loss": 0.7452,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.8901,
       "step": 70
     },
     {
       "epoch": 0.6,
-      "learning_rate": 4.442896935933148e-05,
-      "loss": 0.6585,
+      "learning_rate": 3.007518796992481e-05,
+      "loss": 0.8331,
       "step": 80
     },
     {
       "epoch": 0.67,
-      "learning_rate": 4.303621169916435e-05,
-      "loss": 0.6018,
+      "learning_rate": 3.3834586466165414e-05,
+      "loss": 0.751,
       "step": 90
     },
     {
       "epoch": 0.75,
-      "learning_rate": 4.164345403899722e-05,
-      "loss": 0.5398,
+      "learning_rate": 3.759398496240601e-05,
+      "loss": 0.7102,
       "step": 100
     },
     {
       "epoch": 0.82,
-      "learning_rate": 4.0250696378830085e-05,
-      "loss": 0.5506,
+      "learning_rate": 4.135338345864662e-05,
+      "loss": 0.6395,
       "step": 110
     },
     {
       "epoch": 0.9,
-      "learning_rate": 3.885793871866296e-05,
-      "loss": 0.5615,
+      "learning_rate": 4.511278195488722e-05,
+      "loss": 0.6145,
       "step": 120
     },
     {
       "epoch": 0.97,
-      "learning_rate": 3.746518105849583e-05,
-      "loss": 0.5242,
+      "learning_rate": 4.887218045112782e-05,
+      "loss": 0.5832,
       "step": 130
     },
     {
       "epoch": 0.99,
-      "eval_accuracy": 0.8502336448598131,
-      "eval_loss": 0.4905190169811249,
-      "eval_runtime": 89.5244,
-      "eval_samples_per_second": 47.808,
-      "eval_steps_per_second": 1.497,
+      "eval_accuracy": 0.8299065420560747,
+      "eval_loss": 0.5485312342643738,
+      "eval_runtime": 73.2179,
+      "eval_samples_per_second": 58.456,
+      "eval_steps_per_second": 1.83,
       "step": 133
     },
     {
       "epoch": 1.05,
-      "learning_rate": 3.607242339832869e-05,
-      "loss": 0.5078,
+      "learning_rate": 4.970760233918128e-05,
+      "loss": 0.5927,
       "step": 140
     },
     {
       "epoch": 1.13,
-      "learning_rate": 3.4679665738161556e-05,
-      "loss": 0.5059,
+      "learning_rate": 4.928989139515455e-05,
+      "loss": 0.4566,
       "step": 150
     },
     {
       "epoch": 1.2,
-      "learning_rate": 3.328690807799443e-05,
-      "loss": 0.4914,
+      "learning_rate": 4.887218045112782e-05,
+      "loss": 0.5143,
       "step": 160
     },
     {
       "epoch": 1.28,
-      "learning_rate": 3.18941504178273e-05,
-      "loss": 0.4694,
+      "learning_rate": 4.8454469507101085e-05,
+      "loss": 0.4993,
       "step": 170
     },
     {
       "epoch": 1.35,
-      "learning_rate": 3.050139275766017e-05,
-      "loss": 0.4538,
+      "learning_rate": 4.803675856307435e-05,
+      "loss": 0.4912,
       "step": 180
     },
     {
       "epoch": 1.43,
-      "learning_rate": 2.9108635097493035e-05,
-      "loss": 0.4467,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 0.4608,
       "step": 190
     },
     {
       "epoch": 1.5,
-      "learning_rate": 2.7715877437325903e-05,
-      "loss": 0.4958,
+      "learning_rate": 4.720133667502089e-05,
+      "loss": 0.4569,
       "step": 200
     },
     {
       "epoch": 1.58,
-      "learning_rate": 2.6323119777158778e-05,
-      "loss": 0.4647,
+      "learning_rate": 4.678362573099415e-05,
+      "loss": 0.4649,
       "step": 210
     },
     {
       "epoch": 1.65,
-      "learning_rate": 2.4930362116991646e-05,
-      "loss": 0.413,
+      "learning_rate": 4.6365914786967416e-05,
+      "loss": 0.4428,
       "step": 220
     },
     {
       "epoch": 1.73,
-      "learning_rate": 2.3537604456824514e-05,
-      "loss": 0.4005,
+      "learning_rate": 4.5948203842940684e-05,
+      "loss": 0.4654,
       "step": 230
     },
     {
       "epoch": 1.8,
-      "learning_rate": 2.214484679665738e-05,
-      "loss": 0.4374,
+      "learning_rate": 4.553049289891395e-05,
+      "loss": 0.4404,
       "step": 240
     },
     {
       "epoch": 1.87,
-      "learning_rate": 2.0752089136490253e-05,
-      "loss": 0.4366,
+      "learning_rate": 4.511278195488722e-05,
+      "loss": 0.4231,
       "step": 250
     },
     {
       "epoch": 1.95,
-      "learning_rate": 1.935933147632312e-05,
-      "loss": 0.4405,
+      "learning_rate": 4.4695071010860486e-05,
+      "loss": 0.4638,
       "step": 260
     },
     {
       "epoch": 1.99,
-      "eval_accuracy": 0.8696261682242991,
-      "eval_loss": 0.41539567708969116,
-      "eval_runtime": 92.6526,
-      "eval_samples_per_second": 46.194,
-      "eval_steps_per_second": 1.446,
+      "eval_accuracy": 0.8574766355140186,
+      "eval_loss": 0.44362181425094604,
+      "eval_runtime": 75.8626,
+      "eval_samples_per_second": 56.418,
+      "eval_steps_per_second": 1.766,
       "step": 266
     },
     {
       "epoch": 2.03,
-      "learning_rate": 1.7966573816155992e-05,
-      "loss": 0.4296,
+      "learning_rate": 4.4277360066833754e-05,
+      "loss": 0.4435,
       "step": 270
     },
     {
       "epoch": 2.1,
-      "learning_rate": 1.6573816155988857e-05,
-      "loss": 0.4433,
+      "learning_rate": 4.3859649122807014e-05,
+      "loss": 0.3596,
       "step": 280
     },
     {
       "epoch": 2.18,
-      "learning_rate": 1.518105849582173e-05,
-      "loss": 0.4323,
+      "learning_rate": 4.344193817878028e-05,
+      "loss": 0.3644,
       "step": 290
     },
     {
       "epoch": 2.25,
-      "learning_rate": 1.3788300835654596e-05,
-      "loss": 0.4169,
+      "learning_rate": 4.302422723475355e-05,
+      "loss": 0.3731,
       "step": 300
     },
     {
       "epoch": 2.33,
-      "learning_rate": 1.2395543175487466e-05,
-      "loss": 0.395,
+      "learning_rate": 4.260651629072682e-05,
+      "loss": 0.3682,
       "step": 310
     },
     {
       "epoch": 2.4,
-      "learning_rate": 1.1002785515320335e-05,
-      "loss": 0.4045,
+      "learning_rate": 4.2188805346700084e-05,
+      "loss": 0.3619,
       "step": 320
     },
     {
       "epoch": 2.48,
-      "learning_rate": 9.610027855153205e-06,
-      "loss": 0.3718,
+      "learning_rate": 4.177109440267335e-05,
+      "loss": 0.3729,
       "step": 330
     },
     {
       "epoch": 2.55,
-      "learning_rate": 8.217270194986073e-06,
-      "loss": 0.4601,
+      "learning_rate": 4.135338345864662e-05,
+      "loss": 0.4054,
       "step": 340
     },
     {
       "epoch": 2.63,
-      "learning_rate": 6.8245125348189415e-06,
-      "loss": 0.3969,
+      "learning_rate": 4.093567251461988e-05,
+      "loss": 0.339,
       "step": 350
     },
     {
       "epoch": 2.7,
-      "learning_rate": 5.43175487465181e-06,
-      "loss": 0.397,
+      "learning_rate": 4.051796157059315e-05,
+      "loss": 0.3598,
       "step": 360
     },
     {
       "epoch": 2.78,
-      "learning_rate": 4.03899721448468e-06,
-      "loss": 0.3909,
+      "learning_rate": 4.0100250626566415e-05,
+      "loss": 0.3417,
       "step": 370
     },
     {
       "epoch": 2.85,
-      "learning_rate": 2.6462395543175487e-06,
-      "loss": 0.4032,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 0.3379,
       "step": 380
     },
     {
       "epoch": 2.93,
-      "learning_rate": 1.253481894150418e-06,
-      "loss": 0.3938,
+      "learning_rate": 3.926482873851295e-05,
+      "loss": 0.3115,
       "step": 390
     },
     {
       "epoch": 2.99,
-      "eval_accuracy": 0.872196261682243,
-      "eval_loss": 0.4010835587978363,
-      "eval_runtime": 89.5202,
-      "eval_samples_per_second": 47.81,
-      "eval_steps_per_second": 1.497,
+      "eval_accuracy": 0.8644859813084113,
+      "eval_loss": 0.4172683358192444,
+      "eval_runtime": 70.472,
+      "eval_samples_per_second": 60.733,
+      "eval_steps_per_second": 1.901,
       "step": 399
     },
     {
-      "epoch": 2.99,
-      "step": 399,
-      "total_flos": 3.9721912052731453e+18,
-      "train_loss": 0.5681018877148927,
-      "train_runtime": 3778.1283,
-      "train_samples_per_second": 13.592,
-      "train_steps_per_second": 0.106
+      "epoch": 3.01,
+      "learning_rate": 3.884711779448622e-05,
+      "loss": 0.4131,
+      "step": 400
+    },
+    {
+      "epoch": 3.08,
+      "learning_rate": 3.8429406850459485e-05,
+      "loss": 0.2941,
+      "step": 410
+    },
+    {
+      "epoch": 3.16,
+      "learning_rate": 3.8011695906432746e-05,
+      "loss": 0.3021,
+      "step": 420
+    },
+    {
+      "epoch": 3.23,
+      "learning_rate": 3.759398496240601e-05,
+      "loss": 0.2967,
+      "step": 430
+    },
+    {
+      "epoch": 3.31,
+      "learning_rate": 3.717627401837928e-05,
+      "loss": 0.2906,
+      "step": 440
+    },
+    {
+      "epoch": 3.38,
+      "learning_rate": 3.675856307435255e-05,
+      "loss": 0.2741,
+      "step": 450
+    },
+    {
+      "epoch": 3.46,
+      "learning_rate": 3.6340852130325816e-05,
+      "loss": 0.3069,
+      "step": 460
+    },
+    {
+      "epoch": 3.53,
+      "learning_rate": 3.592314118629908e-05,
+      "loss": 0.2697,
+      "step": 470
+    },
+    {
+      "epoch": 3.61,
+      "learning_rate": 3.5505430242272344e-05,
+      "loss": 0.2777,
+      "step": 480
+    },
+    {
+      "epoch": 3.68,
+      "learning_rate": 3.508771929824561e-05,
+      "loss": 0.2863,
+      "step": 490
+    },
+    {
+      "epoch": 3.76,
+      "learning_rate": 3.467000835421888e-05,
+      "loss": 0.3172,
+      "step": 500
+    },
+    {
+      "epoch": 3.83,
+      "learning_rate": 3.4252297410192146e-05,
+      "loss": 0.2982,
+      "step": 510
+    },
+    {
+      "epoch": 3.9,
+      "learning_rate": 3.3834586466165414e-05,
+      "loss": 0.2902,
+      "step": 520
+    },
+    {
+      "epoch": 3.98,
+      "learning_rate": 3.341687552213868e-05,
+      "loss": 0.2926,
+      "step": 530
+    },
+    {
+      "epoch": 3.99,
+      "eval_accuracy": 0.8476635514018691,
+      "eval_loss": 0.44750651717185974,
+      "eval_runtime": 69.0564,
+      "eval_samples_per_second": 61.978,
+      "eval_steps_per_second": 1.94,
+      "step": 532
+    },
+    {
+      "epoch": 4.06,
+      "learning_rate": 3.299916457811195e-05,
+      "loss": 0.2541,
+      "step": 540
+    },
+    {
+      "epoch": 4.13,
+      "learning_rate": 3.258145363408521e-05,
+      "loss": 0.2287,
+      "step": 550
+    },
+    {
+      "epoch": 4.21,
+      "learning_rate": 3.216374269005848e-05,
+      "loss": 0.248,
+      "step": 560
+    },
+    {
+      "epoch": 4.28,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 0.2502,
+      "step": 570
+    },
+    {
+      "epoch": 4.36,
+      "learning_rate": 3.132832080200501e-05,
+      "loss": 0.2053,
+      "step": 580
+    },
+    {
+      "epoch": 4.43,
+      "learning_rate": 3.091060985797828e-05,
+      "loss": 0.2359,
+      "step": 590
+    },
+    {
+      "epoch": 4.51,
+      "learning_rate": 3.0492898913951544e-05,
+      "loss": 0.2538,
+      "step": 600
+    },
+    {
+      "epoch": 4.58,
+      "learning_rate": 3.007518796992481e-05,
+      "loss": 0.2371,
+      "step": 610
+    },
+    {
+      "epoch": 4.66,
+      "learning_rate": 2.965747702589808e-05,
+      "loss": 0.2402,
+      "step": 620
+    },
+    {
+      "epoch": 4.73,
+      "learning_rate": 2.9239766081871346e-05,
+      "loss": 0.2418,
+      "step": 630
+    },
+    {
+      "epoch": 4.81,
+      "learning_rate": 2.882205513784461e-05,
+      "loss": 0.2465,
+      "step": 640
+    },
+    {
+      "epoch": 4.88,
+      "learning_rate": 2.8404344193817878e-05,
+      "loss": 0.2241,
+      "step": 650
+    },
+    {
+      "epoch": 4.96,
+      "learning_rate": 2.7986633249791145e-05,
+      "loss": 0.2127,
+      "step": 660
+    },
+    {
+      "epoch": 4.99,
+      "eval_accuracy": 0.8574766355140186,
+      "eval_loss": 0.4497027099132538,
+      "eval_runtime": 74.3222,
+      "eval_samples_per_second": 57.587,
+      "eval_steps_per_second": 1.803,
+      "step": 665
+    },
+    {
+      "epoch": 5.04,
+      "learning_rate": 2.756892230576441e-05,
+      "loss": 0.2127,
+      "step": 670
+    },
+    {
+      "epoch": 5.11,
+      "learning_rate": 2.7151211361737677e-05,
+      "loss": 0.1995,
+      "step": 680
+    },
+    {
+      "epoch": 5.19,
+      "learning_rate": 2.6733500417710944e-05,
+      "loss": 0.184,
+      "step": 690
+    },
+    {
+      "epoch": 5.26,
+      "learning_rate": 2.6315789473684212e-05,
+      "loss": 0.1686,
+      "step": 700
+    },
+    {
+      "epoch": 5.34,
+      "learning_rate": 2.5898078529657476e-05,
+      "loss": 0.1944,
+      "step": 710
+    },
+    {
+      "epoch": 5.41,
+      "learning_rate": 2.5480367585630744e-05,
+      "loss": 0.1827,
+      "step": 720
+    },
+    {
+      "epoch": 5.49,
+      "learning_rate": 2.506265664160401e-05,
+      "loss": 0.2056,
+      "step": 730
+    },
+    {
+      "epoch": 5.56,
+      "learning_rate": 2.4644945697577275e-05,
+      "loss": 0.1706,
+      "step": 740
+    },
+    {
+      "epoch": 5.64,
+      "learning_rate": 2.4227234753550543e-05,
+      "loss": 0.1585,
+      "step": 750
+    },
+    {
+      "epoch": 5.71,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.1805,
+      "step": 760
+    },
+    {
+      "epoch": 5.79,
+      "learning_rate": 2.3391812865497074e-05,
+      "loss": 0.2051,
+      "step": 770
+    },
+    {
+      "epoch": 5.86,
+      "learning_rate": 2.2974101921470342e-05,
+      "loss": 0.1997,
+      "step": 780
+    },
+    {
+      "epoch": 5.93,
+      "learning_rate": 2.255639097744361e-05,
+      "loss": 0.1934,
+      "step": 790
+    },
+    {
+      "epoch": 5.99,
+      "eval_accuracy": 0.858177570093458,
+      "eval_loss": 0.4547787010669708,
+      "eval_runtime": 69.4508,
+      "eval_samples_per_second": 61.626,
+      "eval_steps_per_second": 1.929,
+      "step": 798
+    },
+    {
+      "epoch": 6.01,
+      "learning_rate": 2.2138680033416877e-05,
+      "loss": 0.2103,
+      "step": 800
+    },
+    {
+      "epoch": 6.09,
+      "learning_rate": 2.172096908939014e-05,
+      "loss": 0.1442,
+      "step": 810
+    },
+    {
+      "epoch": 6.16,
+      "learning_rate": 2.130325814536341e-05,
+      "loss": 0.144,
+      "step": 820
+    },
+    {
+      "epoch": 6.24,
+      "learning_rate": 2.0885547201336676e-05,
+      "loss": 0.1844,
+      "step": 830
+    },
+    {
+      "epoch": 6.31,
+      "learning_rate": 2.046783625730994e-05,
+      "loss": 0.1574,
+      "step": 840
+    },
+    {
+      "epoch": 6.39,
+      "learning_rate": 2.0050125313283208e-05,
+      "loss": 0.1594,
+      "step": 850
+    },
+    {
+      "epoch": 6.46,
+      "learning_rate": 1.9632414369256475e-05,
+      "loss": 0.1394,
+      "step": 860
+    },
+    {
+      "epoch": 6.54,
+      "learning_rate": 1.9214703425229743e-05,
+      "loss": 0.1556,
+      "step": 870
+    },
+    {
+      "epoch": 6.61,
+      "learning_rate": 1.8796992481203007e-05,
+      "loss": 0.1181,
+      "step": 880
+    },
+    {
+      "epoch": 6.69,
+      "learning_rate": 1.8379281537176274e-05,
+      "loss": 0.1504,
+      "step": 890
+    },
+    {
+      "epoch": 6.76,
+      "learning_rate": 1.796157059314954e-05,
+      "loss": 0.1296,
+      "step": 900
+    },
+    {
+      "epoch": 6.84,
+      "learning_rate": 1.7543859649122806e-05,
+      "loss": 0.1444,
+      "step": 910
+    },
+    {
+      "epoch": 6.91,
+      "learning_rate": 1.7126148705096073e-05,
+      "loss": 0.1789,
+      "step": 920
+    },
+    {
+      "epoch": 6.99,
+      "learning_rate": 1.670843776106934e-05,
+      "loss": 0.1777,
+      "step": 930
+    },
+    {
+      "epoch": 6.99,
+      "eval_accuracy": 0.8560747663551402,
+      "eval_loss": 0.46795374155044556,
+      "eval_runtime": 69.8404,
+      "eval_samples_per_second": 61.283,
+      "eval_steps_per_second": 1.919,
+      "step": 931
+    },
+    {
+      "epoch": 7.07,
+      "learning_rate": 1.6290726817042605e-05,
+      "loss": 0.132,
+      "step": 940
+    },
+    {
+      "epoch": 7.14,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.1483,
+      "step": 950
+    },
+    {
+      "epoch": 7.22,
+      "learning_rate": 1.545530492898914e-05,
+      "loss": 0.1322,
+      "step": 960
+    },
+    {
+      "epoch": 7.29,
+      "learning_rate": 1.5037593984962406e-05,
+      "loss": 0.1382,
+      "step": 970
+    },
+    {
+      "epoch": 7.37,
+      "learning_rate": 1.4619883040935673e-05,
+      "loss": 0.1041,
+      "step": 980
+    },
+    {
+      "epoch": 7.44,
+      "learning_rate": 1.4202172096908939e-05,
+      "loss": 0.096,
+      "step": 990
+    },
+    {
+      "epoch": 7.52,
+      "learning_rate": 1.3784461152882205e-05,
+      "loss": 0.1336,
+      "step": 1000
+    },
+    {
+      "epoch": 7.59,
+      "learning_rate": 1.3366750208855472e-05,
+      "loss": 0.1269,
+      "step": 1010
+    },
+    {
+      "epoch": 7.67,
+      "learning_rate": 1.2949039264828738e-05,
+      "loss": 0.116,
+      "step": 1020
+    },
+    {
+      "epoch": 7.74,
+      "learning_rate": 1.2531328320802006e-05,
+      "loss": 0.1386,
+      "step": 1030
+    },
+    {
+      "epoch": 7.81,
+      "learning_rate": 1.2113617376775271e-05,
+      "loss": 0.1228,
+      "step": 1040
+    },
+    {
+      "epoch": 7.89,
+      "learning_rate": 1.1695906432748537e-05,
+      "loss": 0.103,
+      "step": 1050
+    },
+    {
+      "epoch": 7.96,
+      "learning_rate": 1.1278195488721805e-05,
+      "loss": 0.1187,
+      "step": 1060
+    },
+    {
+      "epoch": 7.99,
+      "eval_accuracy": 0.8591121495327103,
+      "eval_loss": 0.4879584014415741,
+      "eval_runtime": 69.5594,
+      "eval_samples_per_second": 61.53,
+      "eval_steps_per_second": 1.926,
+      "step": 1064
+    },
+    {
+      "epoch": 8.04,
+      "learning_rate": 1.086048454469507e-05,
+      "loss": 0.1422,
+      "step": 1070
+    },
+    {
+      "epoch": 8.12,
+      "learning_rate": 1.0442773600668338e-05,
+      "loss": 0.0984,
+      "step": 1080
+    },
+    {
+      "epoch": 8.19,
+      "learning_rate": 1.0025062656641604e-05,
+      "loss": 0.1016,
+      "step": 1090
+    },
+    {
+      "epoch": 8.27,
+      "learning_rate": 9.607351712614871e-06,
+      "loss": 0.097,
+      "step": 1100
+    },
+    {
+      "epoch": 8.34,
+      "learning_rate": 9.189640768588137e-06,
+      "loss": 0.1138,
+      "step": 1110
+    },
+    {
+      "epoch": 8.42,
+      "learning_rate": 8.771929824561403e-06,
+      "loss": 0.0806,
+      "step": 1120
+    },
+    {
+      "epoch": 8.49,
+      "learning_rate": 8.35421888053467e-06,
+      "loss": 0.0968,
+      "step": 1130
+    },
+    {
+      "epoch": 8.57,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.0855,
+      "step": 1140
+    },
+    {
+      "epoch": 8.64,
+      "learning_rate": 7.518796992481203e-06,
+      "loss": 0.1295,
+      "step": 1150
+    },
+    {
+      "epoch": 8.72,
+      "learning_rate": 7.1010860484544695e-06,
+      "loss": 0.1062,
+      "step": 1160
+    },
+    {
+      "epoch": 8.79,
+      "learning_rate": 6.683375104427736e-06,
+      "loss": 0.0931,
+      "step": 1170
+    },
+    {
+      "epoch": 8.87,
+      "learning_rate": 6.265664160401003e-06,
+      "loss": 0.118,
+      "step": 1180
+    },
+    {
+      "epoch": 8.94,
+      "learning_rate": 5.8479532163742686e-06,
+      "loss": 0.0801,
+      "step": 1190
+    },
+    {
+      "epoch": 8.99,
+      "eval_accuracy": 0.8556074766355141,
+      "eval_loss": 0.5013762712478638,
+      "eval_runtime": 69.6837,
+      "eval_samples_per_second": 61.42,
+      "eval_steps_per_second": 1.923,
+      "step": 1197
+    },
+    {
+      "epoch": 9.02,
+      "learning_rate": 5.430242272347535e-06,
+      "loss": 0.1025,
+      "step": 1200
+    },
+    {
+      "epoch": 9.1,
+      "learning_rate": 5.012531328320802e-06,
+      "loss": 0.0854,
+      "step": 1210
+    },
+    {
+      "epoch": 9.17,
+      "learning_rate": 4.5948203842940685e-06,
+      "loss": 0.1032,
+      "step": 1220
+    },
+    {
+      "epoch": 9.25,
+      "learning_rate": 4.177109440267335e-06,
+      "loss": 0.108,
+      "step": 1230
+    },
+    {
+      "epoch": 9.32,
+      "learning_rate": 3.7593984962406014e-06,
+      "loss": 0.0933,
+      "step": 1240
+    },
+    {
+      "epoch": 9.4,
+      "learning_rate": 3.341687552213868e-06,
+      "loss": 0.0795,
+      "step": 1250
+    },
+    {
+      "epoch": 9.47,
+      "learning_rate": 2.9239766081871343e-06,
+      "loss": 0.0882,
+      "step": 1260
+    },
+    {
+      "epoch": 9.55,
+      "learning_rate": 2.506265664160401e-06,
+      "loss": 0.0907,
+      "step": 1270
+    },
+    {
+      "epoch": 9.62,
+      "learning_rate": 2.0885547201336676e-06,
+      "loss": 0.0775,
+      "step": 1280
+    },
+    {
+      "epoch": 9.7,
+      "learning_rate": 1.670843776106934e-06,
+      "loss": 0.091,
+      "step": 1290
+    },
+    {
+      "epoch": 9.77,
+      "learning_rate": 1.2531328320802005e-06,
+      "loss": 0.0741,
+      "step": 1300
+    },
+    {
+      "epoch": 9.84,
+      "learning_rate": 8.35421888053467e-07,
+      "loss": 0.0728,
+      "step": 1310
+    },
+    {
+      "epoch": 9.92,
+      "learning_rate": 4.177109440267335e-07,
+      "loss": 0.1036,
+      "step": 1320
+    },
+    {
+      "epoch": 9.99,
+      "learning_rate": 0.0,
+      "loss": 0.088,
+      "step": 1330
+    },
+    {
+      "epoch": 9.99,
+      "eval_accuracy": 0.8558411214953271,
+      "eval_loss": 0.5052832365036011,
+      "eval_runtime": 69.3744,
+      "eval_samples_per_second": 61.694,
+      "eval_steps_per_second": 1.932,
+      "step": 1330
+    },
+    {
+      "epoch": 9.99,
+      "step": 1330,
+      "total_flos": 1.3257453564799912e+19,
+      "train_loss": 0.3000902961967583,
+      "train_runtime": 6004.4248,
+      "train_samples_per_second": 28.507,
+      "train_steps_per_second": 0.222
     }
   ],
-  "max_steps": 399,
-  "num_train_epochs": 3,
-  "total_flos": 3.9721912052731453e+18,
+  "max_steps": 1330,
+  "num_train_epochs": 10,
+  "total_flos": 1.3257453564799912e+19,
   "trial_name": null,
   "trial_params": null
 }