🍻 cheers

all_results.json

@@ -1,13 +1,13 @@
 {
-    "epoch": 15.0,
-    "eval_accuracy": 0.8058823529411765,
-    "eval_loss": 0.6480256915092468,
-    "eval_runtime": 1.4895,
-    "eval_samples_per_second": 114.129,
-    "eval_steps_per_second": 14.77,
-    "total_flos": 1.6866433871220326e+18,
-    "train_loss": 0.11297949704276773,
-    "train_runtime": 502.5996,
-    "train_samples_per_second": 43.305,
-    "train_steps_per_second": 2.716
+    "epoch": 6.0,
+    "eval_accuracy": 0.7411764705882353,
+    "eval_loss": 0.6701839566230774,
+    "eval_runtime": 1.6216,
+    "eval_samples_per_second": 104.832,
+    "eval_steps_per_second": 13.566,
+    "total_flos": 6.74657354848813e+17,
+    "train_loss": 0.6752170354892046,
+    "train_runtime": 241.286,
+    "train_samples_per_second": 36.082,
+    "train_steps_per_second": 2.263
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 15.0,
-    "eval_accuracy": 0.8058823529411765,
-    "eval_loss": 0.6480256915092468,
-    "eval_runtime": 1.4895,
-    "eval_samples_per_second": 114.129,
-    "eval_steps_per_second": 14.77
+    "epoch": 6.0,
+    "eval_accuracy": 0.7411764705882353,
+    "eval_loss": 0.6701839566230774,
+    "eval_runtime": 1.6216,
+    "eval_samples_per_second": 104.832,
+    "eval_steps_per_second": 13.566
 }

runs/Aug06_21-20-02_a3e0ec26426d/events.out.tfevents.1659821844.a3e0ec26426d.71.2

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:7e26146e08f2ba7ffa703d52426c436d046af948cf7d25d3ece570e07971a3a9
+size 363

train_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 15.0,
-    "total_flos": 1.6866433871220326e+18,
-    "train_loss": 0.11297949704276773,
-    "train_runtime": 502.5996,
-    "train_samples_per_second": 43.305,
-    "train_steps_per_second": 2.716
+    "epoch": 6.0,
+    "total_flos": 6.74657354848813e+17,
+    "train_loss": 0.6752170354892046,
+    "train_runtime": 241.286,
+    "train_samples_per_second": 36.082,
+    "train_steps_per_second": 2.263
 }

trainer_state.json

@@ -1,958 +1,394 @@
 {
-  "best_metric": 0.6480256915092468,
-  "best_model_checkpoint": "./croupier-creature-classifier/checkpoint-100",
-  "epoch": 15.0,
-  "global_step": 1365,
+  "best_metric": 0.7184367179870605,
+  "best_model_checkpoint": "./croupier-creature-classifier/checkpoint-500",
+  "epoch": 6.0,
+  "global_step": 546,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
       "epoch": 0.11,
-      "learning_rate": 0.00019853479853479855,
-      "loss": 0.2843,
+      "learning_rate": 0.00019633699633699633,
+      "loss": 1.3778,
       "step": 10
     },
     {
       "epoch": 0.22,
-      "learning_rate": 0.0001970695970695971,
-      "loss": 0.3895,
+      "learning_rate": 0.00019267399267399268,
+      "loss": 1.3655,
       "step": 20
     },
     {
       "epoch": 0.33,
-      "learning_rate": 0.00019560439560439562,
-      "loss": 0.3592,
+      "learning_rate": 0.00018901098901098903,
+      "loss": 1.4055,
       "step": 30
     },
     {
       "epoch": 0.44,
-      "learning_rate": 0.00019413919413919413,
-      "loss": 0.3891,
+      "learning_rate": 0.00018534798534798537,
+      "loss": 1.3073,
       "step": 40
     },
     {
       "epoch": 0.55,
-      "learning_rate": 0.00019267399267399268,
-      "loss": 0.2862,
+      "learning_rate": 0.0001816849816849817,
+      "loss": 1.1786,
       "step": 50
     },
     {
       "epoch": 0.66,
-      "learning_rate": 0.00019120879120879122,
-      "loss": 0.4721,
+      "learning_rate": 0.00017802197802197802,
+      "loss": 1.1462,
       "step": 60
     },
     {
       "epoch": 0.77,
-      "learning_rate": 0.00018974358974358974,
-      "loss": 0.4275,
+      "learning_rate": 0.00017435897435897436,
+      "loss": 1.171,
       "step": 70
     },
     {
       "epoch": 0.88,
-      "learning_rate": 0.00018827838827838828,
-      "loss": 0.2297,
+      "learning_rate": 0.00017069597069597068,
+      "loss": 1.1439,
       "step": 80
     },
     {
       "epoch": 0.99,
-      "learning_rate": 0.00018681318681318683,
-      "loss": 0.3175,
+      "learning_rate": 0.00016703296703296706,
+      "loss": 1.1168,
       "step": 90
     },
     {
       "epoch": 1.1,
-      "learning_rate": 0.00018534798534798537,
-      "loss": 0.1967,
+      "learning_rate": 0.00016336996336996338,
+      "loss": 0.8932,
       "step": 100
     },
     {
       "epoch": 1.1,
-      "eval_accuracy": 0.8058823529411765,
-      "eval_loss": 0.6480256915092468,
-      "eval_runtime": 1.3833,
-      "eval_samples_per_second": 122.894,
-      "eval_steps_per_second": 15.904,
+      "eval_accuracy": 0.6058823529411764,
+      "eval_loss": 0.9914249181747437,
+      "eval_runtime": 1.5014,
+      "eval_samples_per_second": 113.229,
+      "eval_steps_per_second": 14.653,
       "step": 100
     },
     {
       "epoch": 1.21,
-      "learning_rate": 0.0001838827838827839,
-      "loss": 0.2783,
+      "learning_rate": 0.00015970695970695973,
+      "loss": 0.9693,
       "step": 110
     },
     {
       "epoch": 1.32,
-      "learning_rate": 0.0001824175824175824,
-      "loss": 0.3196,
+      "learning_rate": 0.00015604395604395605,
+      "loss": 0.8965,
       "step": 120
     },
     {
       "epoch": 1.43,
-      "learning_rate": 0.00018095238095238095,
-      "loss": 0.2813,
+      "learning_rate": 0.00015238095238095237,
+      "loss": 0.8848,
       "step": 130
     },
     {
       "epoch": 1.54,
-      "learning_rate": 0.0001794871794871795,
-      "loss": 0.2686,
+      "learning_rate": 0.00014871794871794872,
+      "loss": 0.8311,
       "step": 140
     },
     {
       "epoch": 1.65,
-      "learning_rate": 0.00017802197802197802,
-      "loss": 0.2325,
+      "learning_rate": 0.00014505494505494506,
+      "loss": 0.8909,
       "step": 150
     },
     {
       "epoch": 1.76,
-      "learning_rate": 0.00017655677655677656,
-      "loss": 0.3969,
+      "learning_rate": 0.0001413919413919414,
+      "loss": 0.8739,
       "step": 160
     },
     {
       "epoch": 1.87,
-      "learning_rate": 0.0001750915750915751,
-      "loss": 0.1645,
+      "learning_rate": 0.00013772893772893773,
+      "loss": 0.9564,
       "step": 170
     },
     {
       "epoch": 1.98,
-      "learning_rate": 0.00017362637362637365,
-      "loss": 0.3043,
+      "learning_rate": 0.00013406593406593405,
+      "loss": 0.9058,
       "step": 180
     },
     {
       "epoch": 2.09,
-      "learning_rate": 0.00017216117216117217,
-      "loss": 0.1483,
+      "learning_rate": 0.0001304029304029304,
+      "loss": 0.6668,
       "step": 190
     },
     {
       "epoch": 2.2,
-      "learning_rate": 0.00017069597069597068,
-      "loss": 0.1047,
+      "learning_rate": 0.00012673992673992675,
+      "loss": 0.6608,
       "step": 200
     },
     {
       "epoch": 2.2,
-      "eval_accuracy": 0.7529411764705882,
-      "eval_loss": 0.8702762126922607,
-      "eval_runtime": 1.4808,
-      "eval_samples_per_second": 114.8,
-      "eval_steps_per_second": 14.856,
+      "eval_accuracy": 0.6588235294117647,
+      "eval_loss": 0.8645371198654175,
+      "eval_runtime": 1.6411,
+      "eval_samples_per_second": 103.586,
+      "eval_steps_per_second": 13.405,
       "step": 200
     },
     {
       "epoch": 2.31,
-      "learning_rate": 0.00016923076923076923,
-      "loss": 0.1261,
+      "learning_rate": 0.0001230769230769231,
+      "loss": 0.6599,
       "step": 210
     },
     {
       "epoch": 2.42,
-      "learning_rate": 0.00016776556776556777,
-      "loss": 0.282,
+      "learning_rate": 0.00011941391941391942,
+      "loss": 0.6528,
       "step": 220
     },
     {
       "epoch": 2.53,
-      "learning_rate": 0.00016630036630036632,
-      "loss": 0.1773,
+      "learning_rate": 0.00011575091575091577,
+      "loss": 0.79,
       "step": 230
     },
     {
       "epoch": 2.64,
-      "learning_rate": 0.00016483516483516484,
-      "loss": 0.1566,
+      "learning_rate": 0.0001120879120879121,
+      "loss": 0.633,
       "step": 240
     },
     {
       "epoch": 2.75,
-      "learning_rate": 0.00016336996336996338,
-      "loss": 0.2896,
+      "learning_rate": 0.00010842490842490842,
+      "loss": 0.7036,
       "step": 250
     },
     {
       "epoch": 2.86,
-      "learning_rate": 0.00016190476190476192,
-      "loss": 0.2573,
+      "learning_rate": 0.00010476190476190477,
+      "loss": 0.693,
       "step": 260
     },
     {
       "epoch": 2.97,
-      "learning_rate": 0.00016043956043956044,
-      "loss": 0.1943,
+      "learning_rate": 0.0001010989010989011,
+      "loss": 0.6463,
       "step": 270
     },
     {
       "epoch": 3.08,
-      "learning_rate": 0.00015897435897435896,
-      "loss": 0.1993,
+      "learning_rate": 9.743589743589744e-05,
+      "loss": 0.6102,
       "step": 280
     },
     {
       "epoch": 3.19,
-      "learning_rate": 0.0001575091575091575,
-      "loss": 0.1106,
+      "learning_rate": 9.377289377289377e-05,
+      "loss": 0.4745,
       "step": 290
     },
     {
       "epoch": 3.3,
-      "learning_rate": 0.00015604395604395605,
-      "loss": 0.2249,
+      "learning_rate": 9.010989010989012e-05,
+      "loss": 0.6084,
       "step": 300
     },
     {
       "epoch": 3.3,
-      "eval_accuracy": 0.7588235294117647,
-      "eval_loss": 0.9538877606391907,
-      "eval_runtime": 1.4679,
-      "eval_samples_per_second": 115.811,
-      "eval_steps_per_second": 14.987,
+      "eval_accuracy": 0.7294117647058823,
+      "eval_loss": 0.7325618863105774,
+      "eval_runtime": 2.0857,
+      "eval_samples_per_second": 81.509,
+      "eval_steps_per_second": 10.548,
       "step": 300
     },
     {
       "epoch": 3.41,
-      "learning_rate": 0.0001545787545787546,
-      "loss": 0.2232,
+      "learning_rate": 8.644688644688645e-05,
+      "loss": 0.5139,
       "step": 310
     },
     {
       "epoch": 3.52,
-      "learning_rate": 0.0001531135531135531,
-      "loss": 0.084,
+      "learning_rate": 8.278388278388279e-05,
+      "loss": 0.4558,
       "step": 320
     },
     {
       "epoch": 3.63,
-      "learning_rate": 0.00015164835164835165,
-      "loss": 0.1352,
+      "learning_rate": 7.912087912087912e-05,
+      "loss": 0.6333,
       "step": 330
     },
     {
       "epoch": 3.74,
-      "learning_rate": 0.0001501831501831502,
-      "loss": 0.1632,
+      "learning_rate": 7.545787545787546e-05,
+      "loss": 0.4555,
       "step": 340
     },
     {
       "epoch": 3.85,
-      "learning_rate": 0.00014871794871794872,
-      "loss": 0.2161,
+      "learning_rate": 7.17948717948718e-05,
+      "loss": 0.5633,
       "step": 350
     },
     {
       "epoch": 3.96,
-      "learning_rate": 0.00014725274725274726,
-      "loss": 0.1823,
+      "learning_rate": 6.813186813186814e-05,
+      "loss": 0.6345,
       "step": 360
     },
     {
       "epoch": 4.07,
-      "learning_rate": 0.00014578754578754578,
-      "loss": 0.0965,
+      "learning_rate": 6.446886446886447e-05,
+      "loss": 0.4467,
       "step": 370
     },
     {
       "epoch": 4.18,
-      "learning_rate": 0.00014432234432234432,
-      "loss": 0.135,
+      "learning_rate": 6.08058608058608e-05,
+      "loss": 0.4054,
       "step": 380
     },
     {
       "epoch": 4.29,
-      "learning_rate": 0.00014285714285714287,
-      "loss": 0.0782,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 0.3802,
       "step": 390
     },
     {
       "epoch": 4.4,
-      "learning_rate": 0.0001413919413919414,
-      "loss": 0.0984,
+      "learning_rate": 5.347985347985348e-05,
+      "loss": 0.5261,
       "step": 400
     },
     {
       "epoch": 4.4,
-      "eval_accuracy": 0.7529411764705882,
-      "eval_loss": 0.9318849444389343,
-      "eval_runtime": 1.4691,
-      "eval_samples_per_second": 115.719,
-      "eval_steps_per_second": 14.975,
+      "eval_accuracy": 0.6941176470588235,
+      "eval_loss": 0.7683902382850647,
+      "eval_runtime": 1.6322,
+      "eval_samples_per_second": 104.156,
+      "eval_steps_per_second": 13.479,
       "step": 400
     },
     {
       "epoch": 4.51,
-      "learning_rate": 0.00013992673992673993,
-      "loss": 0.1553,
+      "learning_rate": 4.981684981684982e-05,
+      "loss": 0.3335,
       "step": 410
     },
     {
       "epoch": 4.62,
-      "learning_rate": 0.00013846153846153847,
-      "loss": 0.0954,
+      "learning_rate": 4.615384615384616e-05,
+      "loss": 0.3908,
       "step": 420
     },
     {
       "epoch": 4.73,
-      "learning_rate": 0.000136996336996337,
-      "loss": 0.1225,
+      "learning_rate": 4.249084249084249e-05,
+      "loss": 0.3603,
       "step": 430
     },
     {
       "epoch": 4.84,
-      "learning_rate": 0.00013553113553113554,
-      "loss": 0.2155,
+      "learning_rate": 3.8827838827838833e-05,
+      "loss": 0.3774,
       "step": 440
     },
     {
       "epoch": 4.95,
-      "learning_rate": 0.00013406593406593405,
-      "loss": 0.1496,
+      "learning_rate": 3.516483516483517e-05,
+      "loss": 0.3759,
       "step": 450
     },
     {
       "epoch": 5.05,
-      "learning_rate": 0.0001326007326007326,
-      "loss": 0.0837,
+      "learning_rate": 3.15018315018315e-05,
+      "loss": 0.3021,
       "step": 460
     },
     {
       "epoch": 5.16,
-      "learning_rate": 0.00013113553113553114,
-      "loss": 0.0693,
+      "learning_rate": 2.783882783882784e-05,
+      "loss": 0.3494,
       "step": 470
     },
     {
       "epoch": 5.27,
-      "learning_rate": 0.0001296703296703297,
-      "loss": 0.08,
+      "learning_rate": 2.4175824175824177e-05,
+      "loss": 0.33,
       "step": 480
     },
     {
       "epoch": 5.38,
-      "learning_rate": 0.00012820512820512823,
-      "loss": 0.0989,
+      "learning_rate": 2.0512820512820512e-05,
+      "loss": 0.3074,
       "step": 490
     },
     {
       "epoch": 5.49,
-      "learning_rate": 0.00012673992673992675,
-      "loss": 0.086,
+      "learning_rate": 1.6849816849816853e-05,
+      "loss": 0.2511,
       "step": 500
     },
     {
       "epoch": 5.49,
-      "eval_accuracy": 0.7705882352941177,
-      "eval_loss": 0.9061105251312256,
-      "eval_runtime": 1.4763,
-      "eval_samples_per_second": 115.156,
-      "eval_steps_per_second": 14.903,
+      "eval_accuracy": 0.7058823529411765,
+      "eval_loss": 0.7184367179870605,
+      "eval_runtime": 1.6349,
+      "eval_samples_per_second": 103.983,
+      "eval_steps_per_second": 13.457,
       "step": 500
     },
     {
       "epoch": 5.6,
-      "learning_rate": 0.00012527472527472527,
-      "loss": 0.1005,
+      "learning_rate": 1.3186813186813187e-05,
+      "loss": 0.2773,
       "step": 510
     },
     {
       "epoch": 5.71,
-      "learning_rate": 0.0001238095238095238,
-      "loss": 0.0685,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 0.3286,
       "step": 520
     },
     {
       "epoch": 5.82,
-      "learning_rate": 0.00012234432234432236,
-      "loss": 0.1174,
+      "learning_rate": 5.860805860805861e-06,
+      "loss": 0.2832,
       "step": 530
     },
     {
       "epoch": 5.93,
-      "learning_rate": 0.00012087912087912087,
-      "loss": 0.1728,
-      "step": 540
-    },
-    {
-      "epoch": 6.04,
-      "learning_rate": 0.00011941391941391942,
-      "loss": 0.086,
-      "step": 550
-    },
-    {
-      "epoch": 6.15,
-      "learning_rate": 0.00011794871794871796,
-      "loss": 0.0738,
-      "step": 560
-    },
-    {
-      "epoch": 6.26,
-      "learning_rate": 0.0001164835164835165,
-      "loss": 0.04,
-      "step": 570
-    },
-    {
-      "epoch": 6.37,
-      "learning_rate": 0.00011501831501831501,
-      "loss": 0.1145,
-      "step": 580
-    },
-    {
-      "epoch": 6.48,
-      "learning_rate": 0.00011355311355311356,
-      "loss": 0.0571,
-      "step": 590
-    },
-    {
-      "epoch": 6.59,
-      "learning_rate": 0.0001120879120879121,
-      "loss": 0.1164,
-      "step": 600
-    },
-    {
-      "epoch": 6.59,
-      "eval_accuracy": 0.8176470588235294,
-      "eval_loss": 0.7493334412574768,
-      "eval_runtime": 1.4541,
-      "eval_samples_per_second": 116.908,
-      "eval_steps_per_second": 15.129,
-      "step": 600
-    },
-    {
-      "epoch": 6.7,
-      "learning_rate": 0.00011062271062271063,
-      "loss": 0.0479,
-      "step": 610
-    },
-    {
-      "epoch": 6.81,
-      "learning_rate": 0.00010915750915750915,
-      "loss": 0.0899,
-      "step": 620
-    },
-    {
-      "epoch": 6.92,
-      "learning_rate": 0.0001076923076923077,
-      "loss": 0.0746,
-      "step": 630
-    },
-    {
-      "epoch": 7.03,
-      "learning_rate": 0.00010622710622710624,
-      "loss": 0.0826,
-      "step": 640
-    },
-    {
-      "epoch": 7.14,
-      "learning_rate": 0.00010476190476190477,
-      "loss": 0.0382,
-      "step": 650
-    },
-    {
-      "epoch": 7.25,
-      "learning_rate": 0.00010329670329670331,
-      "loss": 0.0724,
-      "step": 660
-    },
-    {
-      "epoch": 7.36,
-      "learning_rate": 0.00010183150183150183,
-      "loss": 0.036,
-      "step": 670
-    },
-    {
-      "epoch": 7.47,
-      "learning_rate": 0.00010036630036630038,
-      "loss": 0.0579,
-      "step": 680
-    },
-    {
-      "epoch": 7.58,
-      "learning_rate": 9.89010989010989e-05,
-      "loss": 0.089,
-      "step": 690
-    },
-    {
-      "epoch": 7.69,
-      "learning_rate": 9.743589743589744e-05,
-      "loss": 0.0518,
-      "step": 700
-    },
-    {
-      "epoch": 7.69,
-      "eval_accuracy": 0.7764705882352941,
-      "eval_loss": 0.8781054615974426,
-      "eval_runtime": 1.4674,
-      "eval_samples_per_second": 115.852,
-      "eval_steps_per_second": 14.993,
-      "step": 700
-    },
-    {
-      "epoch": 7.8,
-      "learning_rate": 9.597069597069598e-05,
-      "loss": 0.0603,
-      "step": 710
-    },
-    {
-      "epoch": 7.91,
-      "learning_rate": 9.450549450549451e-05,
-      "loss": 0.0565,
-      "step": 720
-    },
-    {
-      "epoch": 8.02,
-      "learning_rate": 9.304029304029304e-05,
-      "loss": 0.0689,
-      "step": 730
-    },
-    {
-      "epoch": 8.13,
-      "learning_rate": 9.157509157509158e-05,
-      "loss": 0.0295,
-      "step": 740
-    },
-    {
-      "epoch": 8.24,
-      "learning_rate": 9.010989010989012e-05,
-      "loss": 0.0555,
-      "step": 750
-    },
-    {
-      "epoch": 8.35,
-      "learning_rate": 8.864468864468865e-05,
-      "loss": 0.0501,
-      "step": 760
-    },
-    {
-      "epoch": 8.46,
-      "learning_rate": 8.717948717948718e-05,
-      "loss": 0.0657,
-      "step": 770
-    },
-    {
-      "epoch": 8.57,
-      "learning_rate": 8.571428571428571e-05,
-      "loss": 0.0479,
-      "step": 780
-    },
-    {
-      "epoch": 8.68,
-      "learning_rate": 8.424908424908426e-05,
-      "loss": 0.0665,
-      "step": 790
-    },
-    {
-      "epoch": 8.79,
-      "learning_rate": 8.278388278388279e-05,
-      "loss": 0.0458,
-      "step": 800
-    },
-    {
-      "epoch": 8.79,
-      "eval_accuracy": 0.7823529411764706,
-      "eval_loss": 0.8851077556610107,
-      "eval_runtime": 1.4711,
-      "eval_samples_per_second": 115.561,
-      "eval_steps_per_second": 14.955,
-      "step": 800
-    },
-    {
-      "epoch": 8.9,
-      "learning_rate": 8.131868131868132e-05,
-      "loss": 0.0704,
-      "step": 810
-    },
-    {
-      "epoch": 9.01,
-      "learning_rate": 7.985347985347986e-05,
-      "loss": 0.0779,
-      "step": 820
-    },
-    {
-      "epoch": 9.12,
-      "learning_rate": 7.83882783882784e-05,
-      "loss": 0.0575,
-      "step": 830
-    },
-    {
-      "epoch": 9.23,
-      "learning_rate": 7.692307692307693e-05,
-      "loss": 0.0541,
-      "step": 840
-    },
-    {
-      "epoch": 9.34,
-      "learning_rate": 7.545787545787546e-05,
-      "loss": 0.0413,
-      "step": 850
-    },
-    {
-      "epoch": 9.45,
-      "learning_rate": 7.3992673992674e-05,
-      "loss": 0.0367,
-      "step": 860
-    },
-    {
-      "epoch": 9.56,
-      "learning_rate": 7.252747252747253e-05,
-      "loss": 0.0857,
-      "step": 870
-    },
-    {
-      "epoch": 9.67,
-      "learning_rate": 7.106227106227106e-05,
-      "loss": 0.061,
-      "step": 880
-    },
-    {
-      "epoch": 9.78,
-      "learning_rate": 6.95970695970696e-05,
-      "loss": 0.0408,
-      "step": 890
-    },
-    {
-      "epoch": 9.89,
-      "learning_rate": 6.813186813186814e-05,
-      "loss": 0.0521,
-      "step": 900
-    },
-    {
-      "epoch": 9.89,
-      "eval_accuracy": 0.788235294117647,
-      "eval_loss": 0.9448409080505371,
-      "eval_runtime": 1.3677,
-      "eval_samples_per_second": 124.3,
-      "eval_steps_per_second": 16.086,
-      "step": 900
-    },
-    {
-      "epoch": 10.0,
-      "learning_rate": 6.666666666666667e-05,
-      "loss": 0.0452,
-      "step": 910
-    },
-    {
-      "epoch": 10.11,
-      "learning_rate": 6.52014652014652e-05,
-      "loss": 0.0573,
-      "step": 920
-    },
-    {
-      "epoch": 10.22,
-      "learning_rate": 6.373626373626373e-05,
-      "loss": 0.0333,
-      "step": 930
-    },
-    {
-      "epoch": 10.33,
-      "learning_rate": 6.227106227106228e-05,
-      "loss": 0.035,
-      "step": 940
-    },
-    {
-      "epoch": 10.44,
-      "learning_rate": 6.08058608058608e-05,
-      "loss": 0.0513,
-      "step": 950
-    },
-    {
-      "epoch": 10.55,
-      "learning_rate": 5.9340659340659345e-05,
-      "loss": 0.0412,
-      "step": 960
-    },
-    {
-      "epoch": 10.66,
-      "learning_rate": 5.787545787545788e-05,
-      "loss": 0.044,
-      "step": 970
-    },
-    {
-      "epoch": 10.77,
-      "learning_rate": 5.6410256410256414e-05,
-      "loss": 0.0559,
-      "step": 980
-    },
-    {
-      "epoch": 10.88,
-      "learning_rate": 5.494505494505495e-05,
-      "loss": 0.0889,
-      "step": 990
-    },
-    {
-      "epoch": 10.99,
-      "learning_rate": 5.347985347985348e-05,
-      "loss": 0.0576,
-      "step": 1000
-    },
-    {
-      "epoch": 10.99,
-      "eval_accuracy": 0.7823529411764706,
-      "eval_loss": 0.8883835673332214,
-      "eval_runtime": 1.3691,
-      "eval_samples_per_second": 124.171,
-      "eval_steps_per_second": 16.069,
-      "step": 1000
-    },
-    {
-      "epoch": 11.1,
-      "learning_rate": 5.201465201465202e-05,
-      "loss": 0.0591,
-      "step": 1010
-    },
-    {
-      "epoch": 11.21,
-      "learning_rate": 5.054945054945055e-05,
-      "loss": 0.0649,
-      "step": 1020
-    },
-    {
-      "epoch": 11.32,
-      "learning_rate": 4.908424908424908e-05,
-      "loss": 0.0425,
-      "step": 1030
-    },
-    {
-      "epoch": 11.43,
-      "learning_rate": 4.761904761904762e-05,
-      "loss": 0.0471,
-      "step": 1040
-    },
-    {
-      "epoch": 11.54,
-      "learning_rate": 4.615384615384616e-05,
-      "loss": 0.0366,
-      "step": 1050
-    },
-    {
-      "epoch": 11.65,
-      "learning_rate": 4.4688644688644696e-05,
-      "loss": 0.0414,
-      "step": 1060
-    },
-    {
-      "epoch": 11.76,
-      "learning_rate": 4.322344322344323e-05,
-      "loss": 0.0393,
-      "step": 1070
-    },
-    {
-      "epoch": 11.87,
-      "learning_rate": 4.1758241758241765e-05,
-      "loss": 0.0551,
-      "step": 1080
-    },
-    {
-      "epoch": 11.98,
-      "learning_rate": 4.0293040293040296e-05,
-      "loss": 0.0503,
-      "step": 1090
-    },
-    {
-      "epoch": 12.09,
-      "learning_rate": 3.8827838827838833e-05,
-      "loss": 0.0442,
-      "step": 1100
-    },
-    {
-      "epoch": 12.09,
-      "eval_accuracy": 0.788235294117647,
-      "eval_loss": 0.896532416343689,
-      "eval_runtime": 1.5128,
-      "eval_samples_per_second": 112.377,
-      "eval_steps_per_second": 14.543,
-      "step": 1100
-    },
-    {
-      "epoch": 12.2,
-      "learning_rate": 3.7362637362637365e-05,
-      "loss": 0.0595,
-      "step": 1110
-    },
-    {
-      "epoch": 12.31,
-      "learning_rate": 3.58974358974359e-05,
-      "loss": 0.0304,
-      "step": 1120
-    },
-    {
-      "epoch": 12.42,
-      "learning_rate": 3.443223443223443e-05,
-      "loss": 0.0365,
-      "step": 1130
-    },
-    {
-      "epoch": 12.53,
-      "learning_rate": 3.296703296703297e-05,
-      "loss": 0.0917,
-      "step": 1140
-    },
-    {
-      "epoch": 12.64,
-      "learning_rate": 3.15018315018315e-05,
-      "loss": 0.0549,
-      "step": 1150
-    },
-    {
-      "epoch": 12.75,
-      "learning_rate": 3.0036630036630036e-05,
-      "loss": 0.0616,
-      "step": 1160
-    },
-    {
-      "epoch": 12.86,
-      "learning_rate": 2.857142857142857e-05,
-      "loss": 0.0526,
-      "step": 1170
-    },
-    {
-      "epoch": 12.97,
-      "learning_rate": 2.7106227106227105e-05,
-      "loss": 0.0297,
-      "step": 1180
-    },
-    {
-      "epoch": 13.08,
-      "learning_rate": 2.564102564102564e-05,
-      "loss": 0.0409,
-      "step": 1190
-    },
-    {
-      "epoch": 13.19,
-      "learning_rate": 2.4175824175824177e-05,
-      "loss": 0.0254,
-      "step": 1200
-    },
-    {
-      "epoch": 13.19,
-      "eval_accuracy": 0.788235294117647,
-      "eval_loss": 0.9140186905860901,
-      "eval_runtime": 1.3923,
-      "eval_samples_per_second": 122.1,
-      "eval_steps_per_second": 15.801,
-      "step": 1200
-    },
-    {
-      "epoch": 13.3,
-      "learning_rate": 2.2710622710622712e-05,
-      "loss": 0.0373,
-      "step": 1210
-    },
-    {
-      "epoch": 13.41,
-      "learning_rate": 2.1245421245421246e-05,
-      "loss": 0.0325,
-      "step": 1220
-    },
-    {
-      "epoch": 13.52,
-      "learning_rate": 1.978021978021978e-05,
-      "loss": 0.0373,
-      "step": 1230
-    },
-    {
-      "epoch": 13.63,
-      "learning_rate": 1.8315018315018315e-05,
-      "loss": 0.0545,
-      "step": 1240
-    },
-    {
-      "epoch": 13.74,
-      "learning_rate": 1.6849816849816853e-05,
-      "loss": 0.0613,
-      "step": 1250
-    },
-    {
-      "epoch": 13.85,
-      "learning_rate": 1.5384615384615387e-05,
-      "loss": 0.0515,
-      "step": 1260
-    },
-    {
-      "epoch": 13.96,
-      "learning_rate": 1.391941391941392e-05,
-      "loss": 0.0708,
-      "step": 1270
-    },
-    {
-      "epoch": 14.07,
-      "learning_rate": 1.2454212454212454e-05,
-      "loss": 0.0615,
-      "step": 1280
-    },
-    {
-      "epoch": 14.18,
-      "learning_rate": 1.0989010989010989e-05,
-      "loss": 0.0495,
-      "step": 1290
-    },
-    {
-      "epoch": 14.29,
-      "learning_rate": 9.523809523809523e-06,
-      "loss": 0.0426,
-      "step": 1300
-    },
-    {
-      "epoch": 14.29,
-      "eval_accuracy": 0.788235294117647,
-      "eval_loss": 0.927388072013855,
-      "eval_runtime": 1.3816,
-      "eval_samples_per_second": 123.046,
-      "eval_steps_per_second": 15.924,
-      "step": 1300
-    },
-    {
-      "epoch": 14.4,
-      "learning_rate": 8.058608058608059e-06,
-      "loss": 0.0437,
-      "step": 1310
-    },
-    {
-      "epoch": 14.51,
-      "learning_rate": 6.5934065934065935e-06,
-      "loss": 0.028,
-      "step": 1320
-    },
-    {
-      "epoch": 14.62,
-      "learning_rate": 5.128205128205128e-06,
-      "loss": 0.0345,
-      "step": 1330
-    },
-    {
-      "epoch": 14.73,
-      "learning_rate": 3.663003663003663e-06,
-      "loss": 0.0565,
-      "step": 1340
-    },
-    {
-      "epoch": 14.84,
       "learning_rate": 2.197802197802198e-06,
-      "loss": 0.0758,
-      "step": 1350
-    },
-    {
-      "epoch": 14.95,
-      "learning_rate": 7.326007326007326e-07,
-      "loss": 0.047,
-      "step": 1360
+      "loss": 0.2439,
+      "step": 540
     },
     {
-      "epoch": 15.0,
-      "step": 1365,
-      "total_flos": 1.6866433871220326e+18,
-      "train_loss": 0.11297949704276773,
-      "train_runtime": 502.5996,
-      "train_samples_per_second": 43.305,
-      "train_steps_per_second": 2.716
+      "epoch": 6.0,
+      "step": 546,
+      "total_flos": 6.74657354848813e+17,
+      "train_loss": 0.6752170354892046,
+      "train_runtime": 241.286,
+      "train_samples_per_second": 36.082,
+      "train_steps_per_second": 2.263
     }
   ],
-  "max_steps": 1365,
-  "num_train_epochs": 15,
-  "total_flos": 1.6866433871220326e+18,
+  "max_steps": 546,
+  "num_train_epochs": 6,
+  "total_flos": 6.74657354848813e+17,
   "trial_name": null,
   "trial_params": null
 }