End of training

README.md

@@ -22,7 +22,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.8166666666666667
+      value: 0.8333333333333334
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -32,8 +32,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [google/vit-base-patch16-224](https://huggingface.co/google/vit-base-patch16-224) on the imagefolder dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.9448
-- Accuracy: 0.8167
+- Loss: 0.6467
+- Accuracy: 0.8333
 
 ## Model description
 

all_results.json

@@ -1,12 +1,12 @@
 {
     "epoch": 39.48,
-    "eval_accuracy": 0.7833333333333333,
-    "eval_loss": 4.986918644609257e+34,
-    "eval_runtime": 1.1271,
-    "eval_samples_per_second": 53.232,
-    "eval_steps_per_second": 1.774,
-    "train_loss": 5.490440419440805e+34,
-    "train_runtime": 1489.0102,
-    "train_samples_per_second": 65.816,
-    "train_steps_per_second": 0.51
+    "eval_accuracy": 0.8333333333333334,
+    "eval_loss": 0.6466542482376099,
+    "eval_runtime": 1.2993,
+    "eval_samples_per_second": 46.179,
+    "eval_steps_per_second": 1.539,
+    "train_loss": 0.1845952101443943,
+    "train_runtime": 1526.0361,
+    "train_samples_per_second": 64.219,
+    "train_steps_per_second": 0.498
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 39.48,
-    "eval_accuracy": 0.7833333333333333,
-    "eval_loss": 4.986918644609257e+34,
-    "eval_runtime": 1.1271,
-    "eval_samples_per_second": 53.232,
-    "eval_steps_per_second": 1.774
+    "eval_accuracy": 0.8333333333333334,
+    "eval_loss": 0.6466542482376099,
+    "eval_runtime": 1.2993,
+    "eval_samples_per_second": 46.179,
+    "eval_steps_per_second": 1.539
 }

runs/May24_07-19-00_DESKTOP-SKBE9FB/events.out.tfevents.1716558289.DESKTOP-SKBE9FB.7992.1

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4c3e002f8e6ae58941d1ae4e08ab1c4714de7e858846b4eeee9bc699db012686
+size 411

train_results.json

@@ -1,7 +1,7 @@
 {
     "epoch": 39.48,
-    "train_loss": 5.490440419440805e+34,
-    "train_runtime": 1489.0102,
-    "train_samples_per_second": 65.816,
-    "train_steps_per_second": 0.51
+    "train_loss": 0.1845952101443943,
+    "train_runtime": 1526.0361,
+    "train_samples_per_second": 64.219,
+    "train_steps_per_second": 0.498
 }

trainer_state.json

@@ -1,6 +1,6 @@
 {
-  "best_metric": 0.7833333333333333,
-  "best_model_checkpoint": "vit-base-patch16-224-RU4-40\\checkpoint-192",
+  "best_metric": 0.8333333333333334,
+  "best_model_checkpoint": "vit-base-patch16-224-RU4-40\\checkpoint-134",
   "epoch": 39.48051948051948,
   "eval_steps": 500,
   "global_step": 760,
@@ -11,827 +11,827 @@
     {
       "epoch": 0.52,
       "learning_rate": 1.4473684210526315e-05,
-      "loss": 6.233648058173564e+34,
+      "loss": 1.3822,
       "step": 10
     },
     {
       "epoch": 0.99,
-      "eval_accuracy": 0.4666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.2219,
-      "eval_samples_per_second": 49.104,
-      "eval_steps_per_second": 1.637,
+      "eval_accuracy": 0.48333333333333334,
+      "eval_loss": 1.3130170106887817,
+      "eval_runtime": 1.2415,
+      "eval_samples_per_second": 48.328,
+      "eval_steps_per_second": 1.611,
       "step": 19
     },
     {
       "epoch": 1.04,
       "learning_rate": 2.894736842105263e-05,
-      "loss": 4.900340287388139e+34,
+      "loss": 1.3403,
       "step": 20
     },
     {
       "epoch": 1.56,
       "learning_rate": 4.342105263157895e-05,
-      "loss": 5.844043965150482e+34,
+      "loss": 1.2724,
       "step": 30
     },
     {
       "epoch": 1.97,
       "eval_accuracy": 0.6,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 0.9932,
-      "eval_samples_per_second": 60.409,
-      "eval_steps_per_second": 2.014,
+      "eval_loss": 1.0986911058425903,
+      "eval_runtime": 1.2181,
+      "eval_samples_per_second": 49.258,
+      "eval_steps_per_second": 1.642,
       "step": 38
     },
     {
       "epoch": 2.08,
       "learning_rate": 5.484764542936288e-05,
-      "loss": 4.939300736304528e+34,
+      "loss": 1.1316,
       "step": 40
     },
     {
       "epoch": 2.6,
       "learning_rate": 5.408587257617729e-05,
-      "loss": 5.064838948230818e+34,
+      "loss": 0.9711,
       "step": 50
     },
     {
       "epoch": 2.96,
-      "eval_accuracy": 0.7333333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 0.9664,
-      "eval_samples_per_second": 62.087,
-      "eval_steps_per_second": 2.07,
+      "eval_accuracy": 0.6666666666666666,
+      "eval_loss": 0.862390398979187,
+      "eval_runtime": 1.0302,
+      "eval_samples_per_second": 58.242,
+      "eval_steps_per_second": 1.941,
       "step": 57
     },
     {
       "epoch": 3.12,
       "learning_rate": 5.332409972299169e-05,
-      "loss": 5.610283252356208e+34,
+      "loss": 0.7924,
       "step": 60
     },
     {
       "epoch": 3.64,
       "learning_rate": 5.256232686980609e-05,
-      "loss": 5.61028206393377e+34,
+      "loss": 0.6349,
       "step": 70
     },
     {
       "epoch": 4.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0115,
-      "eval_samples_per_second": 59.32,
-      "eval_steps_per_second": 1.977,
+      "eval_accuracy": 0.7333333333333333,
+      "eval_loss": 0.7396789789199829,
+      "eval_runtime": 1.159,
+      "eval_samples_per_second": 51.769,
+      "eval_steps_per_second": 1.726,
       "step": 77
     },
     {
       "epoch": 4.16,
       "learning_rate": 5.18005540166205e-05,
-      "loss": 5.0691679750305965e+34,
+      "loss": 0.5199,
       "step": 80
     },
     {
       "epoch": 4.68,
       "learning_rate": 5.1038781163434903e-05,
-      "loss": 6.233647662032751e+34,
+      "loss": 0.4068,
       "step": 90
     },
     {
       "epoch": 4.99,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 0.9845,
-      "eval_samples_per_second": 60.944,
-      "eval_steps_per_second": 2.031,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.6979052424430847,
+      "eval_runtime": 1.165,
+      "eval_samples_per_second": 51.502,
+      "eval_steps_per_second": 1.717,
       "step": 96
     },
     {
       "epoch": 5.19,
       "learning_rate": 5.027700831024931e-05,
-      "loss": 5.476086590689547e+34,
+      "loss": 0.3676,
       "step": 100
     },
     {
       "epoch": 5.71,
       "learning_rate": 4.9515235457063714e-05,
-      "loss": 5.688204150188987e+34,
+      "loss": 0.2877,
       "step": 110
     },
     {
       "epoch": 5.97,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0324,
-      "eval_samples_per_second": 58.118,
-      "eval_steps_per_second": 1.937,
+      "eval_accuracy": 0.7833333333333333,
+      "eval_loss": 0.6270049810409546,
+      "eval_runtime": 1.0341,
+      "eval_samples_per_second": 58.021,
+      "eval_steps_per_second": 1.934,
       "step": 115
     },
     {
       "epoch": 6.23,
       "learning_rate": 4.8753462603878116e-05,
-      "loss": 5.125444135005291e+34,
+      "loss": 0.2292,
       "step": 120
     },
     {
       "epoch": 6.75,
       "learning_rate": 4.7991689750692524e-05,
-      "loss": 5.18171950269836e+34,
+      "loss": 0.2217,
       "step": 130
     },
     {
       "epoch": 6.96,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 0.9809,
-      "eval_samples_per_second": 61.167,
-      "eval_steps_per_second": 2.039,
+      "eval_accuracy": 0.8333333333333334,
+      "eval_loss": 0.6466542482376099,
+      "eval_runtime": 1.0399,
+      "eval_samples_per_second": 57.699,
+      "eval_steps_per_second": 1.923,
       "step": 134
     },
     {
       "epoch": 7.27,
       "learning_rate": 4.7229916897506926e-05,
-      "loss": 6.216331950974446e+34,
+      "loss": 0.2464,
       "step": 140
     },
     {
       "epoch": 7.79,
       "learning_rate": 4.6468144044321335e-05,
-      "loss": 5.298600453306717e+34,
+      "loss": 0.195,
       "step": 150
     },
     {
       "epoch": 8.0,
-      "eval_accuracy": 0.6666666666666666,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0194,
-      "eval_samples_per_second": 58.858,
-      "eval_steps_per_second": 1.962,
+      "eval_accuracy": 0.7833333333333333,
+      "eval_loss": 0.685775876045227,
+      "eval_runtime": 1.0519,
+      "eval_samples_per_second": 57.037,
+      "eval_steps_per_second": 1.901,
       "step": 154
     },
     {
       "epoch": 8.31,
       "learning_rate": 4.570637119113573e-05,
-      "loss": 5.761795228940361e+34,
+      "loss": 0.177,
       "step": 160
     },
     {
       "epoch": 8.83,
       "learning_rate": 4.494459833795014e-05,
-      "loss": 5.57132319958063e+34,
+      "loss": 0.1392,
       "step": 170
     },
     {
       "epoch": 8.99,
-      "eval_accuracy": 0.6833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0402,
-      "eval_samples_per_second": 57.682,
-      "eval_steps_per_second": 1.923,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.650473415851593,
+      "eval_runtime": 1.0115,
+      "eval_samples_per_second": 59.32,
+      "eval_steps_per_second": 1.977,
       "step": 173
     },
     {
       "epoch": 9.35,
       "learning_rate": 4.418282548476455e-05,
-      "loss": 5.203363844415632e+34,
+      "loss": 0.1695,
       "step": 180
     },
     {
       "epoch": 9.87,
       "learning_rate": 4.342105263157895e-05,
-      "loss": 5.493401905607039e+34,
+      "loss": 0.1534,
       "step": 190
     },
     {
       "epoch": 9.97,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 0.9879,
-      "eval_samples_per_second": 60.734,
-      "eval_steps_per_second": 2.024,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.6320337057113647,
+      "eval_runtime": 1.1254,
+      "eval_samples_per_second": 53.313,
+      "eval_steps_per_second": 1.777,
       "step": 192
     },
     {
       "epoch": 10.39,
       "learning_rate": 4.265927977839336e-05,
-      "loss": 5.7011904383067e+34,
+      "loss": 0.1318,
       "step": 200
     },
     {
       "epoch": 10.91,
       "learning_rate": 4.189750692520776e-05,
-      "loss": 5.142759846063597e+34,
+      "loss": 0.1136,
       "step": 210
     },
     {
       "epoch": 10.96,
-      "eval_accuracy": 0.7166666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0427,
-      "eval_samples_per_second": 57.542,
-      "eval_steps_per_second": 1.918,
+      "eval_accuracy": 0.7833333333333333,
+      "eval_loss": 0.8345922827720642,
+      "eval_runtime": 1.0135,
+      "eval_samples_per_second": 59.201,
+      "eval_steps_per_second": 1.973,
       "step": 211
     },
     {
       "epoch": 11.43,
       "learning_rate": 4.113573407202216e-05,
-      "loss": 5.883004810207683e+34,
+      "loss": 0.1195,
       "step": 220
     },
     {
       "epoch": 11.95,
       "learning_rate": 4.037396121883656e-05,
-      "loss": 5.181719898839173e+34,
+      "loss": 0.1025,
       "step": 230
     },
     {
       "epoch": 12.0,
-      "eval_accuracy": 0.7,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0312,
-      "eval_samples_per_second": 58.184,
-      "eval_steps_per_second": 1.939,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.6809566617012024,
+      "eval_runtime": 1.0794,
+      "eval_samples_per_second": 55.586,
+      "eval_steps_per_second": 1.853,
       "step": 231
     },
     {
       "epoch": 12.47,
       "learning_rate": 3.961218836565097e-05,
-      "loss": 5.281285138389224e+34,
+      "loss": 0.1025,
       "step": 240
     },
     {
       "epoch": 12.99,
       "learning_rate": 3.885041551246538e-05,
-      "loss": 5.766123859599328e+34,
+      "loss": 0.0894,
       "step": 250
     },
     {
       "epoch": 12.99,
-      "eval_accuracy": 0.7333333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.029,
-      "eval_samples_per_second": 58.308,
-      "eval_steps_per_second": 1.944,
+      "eval_accuracy": 0.7666666666666667,
+      "eval_loss": 0.8257845640182495,
+      "eval_runtime": 1.0081,
+      "eval_samples_per_second": 59.515,
+      "eval_steps_per_second": 1.984,
       "step": 250
     },
     {
       "epoch": 13.51,
       "learning_rate": 3.808864265927978e-05,
-      "loss": 5.805085100797342e+34,
+      "loss": 0.1308,
       "step": 260
     },
     {
       "epoch": 13.97,
       "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 0.9995,
-      "eval_samples_per_second": 60.031,
-      "eval_steps_per_second": 2.001,
+      "eval_loss": 0.9456142783164978,
+      "eval_runtime": 1.0373,
+      "eval_samples_per_second": 57.843,
+      "eval_steps_per_second": 1.928,
       "step": 269
     },
     {
       "epoch": 14.03,
       "learning_rate": 3.732686980609418e-05,
-      "loss": 5.047524821735762e+34,
+      "loss": 0.1002,
       "step": 270
     },
     {
       "epoch": 14.55,
       "learning_rate": 3.6565096952908585e-05,
-      "loss": 5.688204546329799e+34,
+      "loss": 0.0836,
       "step": 280
     },
     {
       "epoch": 14.96,
-      "eval_accuracy": 0.7333333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0235,
-      "eval_samples_per_second": 58.623,
-      "eval_steps_per_second": 1.954,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.9084436893463135,
+      "eval_runtime": 1.0482,
+      "eval_samples_per_second": 57.243,
+      "eval_steps_per_second": 1.908,
       "step": 288
     },
     {
       "epoch": 15.06,
       "learning_rate": 3.5803324099722994e-05,
-      "loss": 5.380850377939275e+34,
+      "loss": 0.0891,
       "step": 290
     },
     {
       "epoch": 15.58,
       "learning_rate": 3.5041551246537395e-05,
-      "loss": 5.610284440778645e+34,
+      "loss": 0.0813,
       "step": 300
     },
     {
       "epoch": 16.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0001,
-      "eval_samples_per_second": 59.996,
-      "eval_steps_per_second": 2.0,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.8688091039657593,
+      "eval_runtime": 1.0293,
+      "eval_samples_per_second": 58.294,
+      "eval_steps_per_second": 1.943,
       "step": 308
     },
     {
       "epoch": 16.1,
       "learning_rate": 3.4279778393351804e-05,
-      "loss": 5.333231479282514e+34,
+      "loss": 0.0928,
       "step": 310
     },
     {
       "epoch": 16.62,
       "learning_rate": 3.3518005540166206e-05,
-      "loss": 5.649243701272597e+34,
+      "loss": 0.1017,
       "step": 320
     },
     {
       "epoch": 16.99,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0006,
-      "eval_samples_per_second": 59.962,
-      "eval_steps_per_second": 1.999,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.8609302043914795,
+      "eval_runtime": 1.0686,
+      "eval_samples_per_second": 56.148,
+      "eval_steps_per_second": 1.872,
       "step": 327
     },
     {
       "epoch": 17.14,
       "learning_rate": 3.275623268698061e-05,
-      "loss": 5.337561298363918e+34,
+      "loss": 0.0703,
       "step": 330
     },
     {
       "epoch": 17.66,
       "learning_rate": 3.1994459833795016e-05,
-      "loss": 5.18171950269836e+34,
+      "loss": 0.076,
       "step": 340
     },
     {
       "epoch": 17.97,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0622,
-      "eval_samples_per_second": 56.487,
-      "eval_steps_per_second": 1.883,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.9015089273452759,
+      "eval_runtime": 1.0675,
+      "eval_samples_per_second": 56.207,
+      "eval_steps_per_second": 1.874,
       "step": 346
     },
     {
       "epoch": 18.18,
       "learning_rate": 3.123268698060942e-05,
-      "loss": 6.255293192172461e+34,
+      "loss": 0.0667,
       "step": 350
     },
     {
       "epoch": 18.7,
       "learning_rate": 3.0470914127423827e-05,
-      "loss": 4.831077443154918e+34,
+      "loss": 0.0726,
       "step": 360
     },
     {
       "epoch": 18.96,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0286,
-      "eval_samples_per_second": 58.329,
-      "eval_steps_per_second": 1.944,
+      "eval_accuracy": 0.7833333333333333,
+      "eval_loss": 0.9918167591094971,
+      "eval_runtime": 1.0851,
+      "eval_samples_per_second": 55.292,
+      "eval_steps_per_second": 1.843,
       "step": 365
     },
     {
       "epoch": 19.22,
       "learning_rate": 2.9709141274238225e-05,
-      "loss": 5.406824142597139e+34,
+      "loss": 0.0787,
       "step": 370
     },
     {
       "epoch": 19.74,
       "learning_rate": 2.894736842105263e-05,
-      "loss": 5.844045153572919e+34,
+      "loss": 0.0549,
       "step": 380
     },
     {
       "epoch": 20.0,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0606,
-      "eval_samples_per_second": 56.57,
-      "eval_steps_per_second": 1.886,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.9064477682113647,
+      "eval_runtime": 1.0254,
+      "eval_samples_per_second": 58.513,
+      "eval_steps_per_second": 1.95,
       "step": 385
     },
     {
       "epoch": 20.26,
       "learning_rate": 2.8185595567867035e-05,
-      "loss": 5.9133087900877645e+34,
+      "loss": 0.057,
       "step": 390
     },
     {
       "epoch": 20.78,
       "learning_rate": 2.742382271468144e-05,
-      "loss": 5.103799001006394e+34,
+      "loss": 0.0676,
       "step": 400
     },
     {
       "epoch": 20.99,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 0.998,
-      "eval_samples_per_second": 60.121,
-      "eval_steps_per_second": 2.004,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.881874680519104,
+      "eval_runtime": 1.3066,
+      "eval_samples_per_second": 45.919,
+      "eval_steps_per_second": 1.531,
       "step": 404
     },
     {
       "epoch": 21.3,
       "learning_rate": 2.6662049861495846e-05,
-      "loss": 5.610282856215395e+34,
+      "loss": 0.0518,
       "step": 410
     },
     {
       "epoch": 21.82,
       "learning_rate": 2.590027700831025e-05,
-      "loss": 5.220681140037188e+34,
+      "loss": 0.0717,
       "step": 420
     },
     {
       "epoch": 21.97,
-      "eval_accuracy": 0.7333333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.065,
-      "eval_samples_per_second": 56.338,
-      "eval_steps_per_second": 1.878,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.8607038259506226,
+      "eval_runtime": 1.0427,
+      "eval_samples_per_second": 57.543,
+      "eval_steps_per_second": 1.918,
       "step": 423
     },
     {
       "epoch": 22.34,
       "learning_rate": 2.5138504155124656e-05,
-      "loss": 5.06483934437163e+34,
+      "loss": 0.0593,
       "step": 430
     },
     {
       "epoch": 22.86,
       "learning_rate": 2.4376731301939058e-05,
-      "loss": 5.493402301747851e+34,
+      "loss": 0.0547,
       "step": 440
     },
     {
       "epoch": 22.96,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0022,
-      "eval_samples_per_second": 59.87,
-      "eval_steps_per_second": 1.996,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.8859201073646545,
+      "eval_runtime": 1.0387,
+      "eval_samples_per_second": 57.765,
+      "eval_steps_per_second": 1.926,
       "step": 442
     },
     {
       "epoch": 23.38,
       "learning_rate": 2.3614958448753463e-05,
-      "loss": 5.640585251532225e+34,
+      "loss": 0.0451,
       "step": 450
     },
     {
       "epoch": 23.9,
       "learning_rate": 2.2853185595567865e-05,
-      "loss": 5.883004810207683e+34,
+      "loss": 0.0466,
       "step": 460
     },
     {
       "epoch": 24.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0192,
-      "eval_samples_per_second": 58.872,
-      "eval_steps_per_second": 1.962,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.9327951073646545,
+      "eval_runtime": 1.1454,
+      "eval_samples_per_second": 52.385,
+      "eval_steps_per_second": 1.746,
       "step": 462
     },
     {
       "epoch": 24.42,
       "learning_rate": 2.2091412742382273e-05,
-      "loss": 5.406824142597139e+34,
+      "loss": 0.0535,
       "step": 470
     },
     {
       "epoch": 24.94,
       "learning_rate": 2.132963988919668e-05,
-      "loss": 5.298600453306717e+34,
+      "loss": 0.0715,
       "step": 480
     },
     {
       "epoch": 24.99,
       "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.039,
-      "eval_samples_per_second": 57.75,
-      "eval_steps_per_second": 1.925,
+      "eval_loss": 1.0178091526031494,
+      "eval_runtime": 1.0507,
+      "eval_samples_per_second": 57.103,
+      "eval_steps_per_second": 1.903,
       "step": 481
     },
     {
       "epoch": 25.45,
       "learning_rate": 2.056786703601108e-05,
-      "loss": 5.640584855391413e+34,
+      "loss": 0.0503,
       "step": 490
     },
     {
       "epoch": 25.97,
       "learning_rate": 1.9806094182825486e-05,
-      "loss": 5.454441852831462e+34,
+      "loss": 0.0446,
       "step": 500
     },
     {
       "epoch": 25.97,
       "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0158,
-      "eval_samples_per_second": 59.069,
-      "eval_steps_per_second": 1.969,
+      "eval_loss": 1.009350061416626,
+      "eval_runtime": 1.0259,
+      "eval_samples_per_second": 58.484,
+      "eval_steps_per_second": 1.949,
       "step": 500
     },
     {
       "epoch": 26.49,
       "learning_rate": 1.904432132963989e-05,
-      "loss": 5.164403791640056e+34,
+      "loss": 0.0468,
       "step": 510
     },
     {
       "epoch": 26.96,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0486,
-      "eval_samples_per_second": 57.219,
-      "eval_steps_per_second": 1.907,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.9175390005111694,
+      "eval_runtime": 1.0785,
+      "eval_samples_per_second": 55.634,
+      "eval_steps_per_second": 1.854,
       "step": 519
     },
     {
       "epoch": 27.01,
       "learning_rate": 1.8282548476454293e-05,
-      "loss": 5.865690683712629e+34,
+      "loss": 0.0523,
       "step": 520
     },
     {
       "epoch": 27.53,
       "learning_rate": 1.7520775623268698e-05,
-      "loss": 5.610284044637833e+34,
+      "loss": 0.0458,
       "step": 530
     },
     {
       "epoch": 28.0,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0074,
-      "eval_samples_per_second": 59.56,
-      "eval_steps_per_second": 1.985,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.8580291271209717,
+      "eval_runtime": 1.0307,
+      "eval_samples_per_second": 58.214,
+      "eval_steps_per_second": 1.94,
       "step": 539
     },
     {
       "epoch": 28.05,
       "learning_rate": 1.6759002770083103e-05,
-      "loss": 5.320244795023989e+34,
+      "loss": 0.0564,
       "step": 540
     },
     {
       "epoch": 28.57,
       "learning_rate": 1.5997229916897508e-05,
-      "loss": 5.999885364675227e+34,
+      "loss": 0.0392,
       "step": 550
     },
     {
       "epoch": 28.99,
       "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0083,
-      "eval_samples_per_second": 59.507,
-      "eval_steps_per_second": 1.984,
+      "eval_loss": 1.0588656663894653,
+      "eval_runtime": 1.0179,
+      "eval_samples_per_second": 58.945,
+      "eval_steps_per_second": 1.965,
       "step": 558
     },
     {
       "epoch": 29.09,
       "learning_rate": 1.5235457063711913e-05,
-      "loss": 4.870037495930495e+34,
+      "loss": 0.0302,
       "step": 560
     },
     {
       "epoch": 29.61,
       "learning_rate": 1.4473684210526315e-05,
-      "loss": 5.142759846063597e+34,
+      "loss": 0.0469,
       "step": 570
     },
     {
       "epoch": 29.97,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0313,
-      "eval_samples_per_second": 58.178,
+      "eval_accuracy": 0.8,
+      "eval_loss": 1.0904656648635864,
+      "eval_runtime": 1.0314,
+      "eval_samples_per_second": 58.173,
       "eval_steps_per_second": 1.939,
       "step": 577
     },
     {
       "epoch": 30.13,
       "learning_rate": 1.371191135734072e-05,
-      "loss": 5.545348642641142e+34,
+      "loss": 0.0365,
       "step": 580
     },
     {
       "epoch": 30.65,
       "learning_rate": 1.2950138504155125e-05,
-      "loss": 5.337562090645544e+34,
+      "loss": 0.0425,
       "step": 590
     },
     {
       "epoch": 30.96,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0057,
-      "eval_samples_per_second": 59.657,
-      "eval_steps_per_second": 1.989,
+      "eval_accuracy": 0.7833333333333333,
+      "eval_loss": 1.0077567100524902,
+      "eval_runtime": 1.0942,
+      "eval_samples_per_second": 54.834,
+      "eval_steps_per_second": 1.828,
       "step": 596
     },
     {
       "epoch": 31.17,
       "learning_rate": 1.2188365650969529e-05,
-      "loss": 6.099452188788528e+34,
+      "loss": 0.0503,
       "step": 600
     },
     {
       "epoch": 31.69,
       "learning_rate": 1.1426592797783932e-05,
-      "loss": 5.883004414066871e+34,
+      "loss": 0.0464,
       "step": 610
     },
     {
       "epoch": 32.0,
-      "eval_accuracy": 0.75,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0397,
-      "eval_samples_per_second": 57.709,
-      "eval_steps_per_second": 1.924,
+      "eval_accuracy": 0.7833333333333333,
+      "eval_loss": 1.0206407308578491,
+      "eval_runtime": 1.0927,
+      "eval_samples_per_second": 54.909,
+      "eval_steps_per_second": 1.83,
       "step": 616
     },
     {
       "epoch": 32.21,
       "learning_rate": 1.066481994459834e-05,
-      "loss": 5.437125745632344e+34,
+      "loss": 0.0315,
       "step": 620
     },
     {
       "epoch": 32.73,
       "learning_rate": 9.903047091412743e-06,
-      "loss": 5.259640400531139e+34,
+      "loss": 0.0336,
       "step": 630
     },
     {
       "epoch": 32.99,
-      "eval_accuracy": 0.7333333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0048,
-      "eval_samples_per_second": 59.711,
-      "eval_steps_per_second": 1.99,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.9653330445289612,
+      "eval_runtime": 1.0833,
+      "eval_samples_per_second": 55.384,
+      "eval_steps_per_second": 1.846,
       "step": 635
     },
     {
       "epoch": 33.25,
       "learning_rate": 9.141274238227146e-06,
-      "loss": 5.047524029454137e+34,
+      "loss": 0.0368,
       "step": 640
     },
     {
       "epoch": 33.77,
       "learning_rate": 8.379501385041551e-06,
-      "loss": 5.766123463458515e+34,
+      "loss": 0.0302,
       "step": 650
     },
     {
       "epoch": 33.97,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0194,
-      "eval_samples_per_second": 58.859,
-      "eval_steps_per_second": 1.962,
+      "eval_accuracy": 0.8,
+      "eval_loss": 0.9574002027511597,
+      "eval_runtime": 1.0375,
+      "eval_samples_per_second": 57.833,
+      "eval_steps_per_second": 1.928,
       "step": 654
     },
     {
       "epoch": 34.29,
       "learning_rate": 7.617728531855957e-06,
-      "loss": 5.934951943382599e+34,
+      "loss": 0.03,
       "step": 660
     },
     {
       "epoch": 34.81,
       "learning_rate": 6.85595567867036e-06,
-      "loss": 5.06483934437163e+34,
+      "loss": 0.0353,
       "step": 670
     },
     {
       "epoch": 34.96,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0082,
-      "eval_samples_per_second": 59.514,
-      "eval_steps_per_second": 1.984,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.962119996547699,
+      "eval_runtime": 1.0464,
+      "eval_samples_per_second": 57.339,
+      "eval_steps_per_second": 1.911,
       "step": 673
     },
     {
       "epoch": 35.32,
       "learning_rate": 6.0941828254847645e-06,
-      "loss": 5.437126141773157e+34,
+      "loss": 0.0372,
       "step": 680
     },
     {
       "epoch": 35.84,
       "learning_rate": 5.33240997229917e-06,
-      "loss": 5.688204546329799e+34,
+      "loss": 0.0344,
       "step": 690
     },
     {
       "epoch": 36.0,
-      "eval_accuracy": 0.7666666666666667,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.021,
-      "eval_samples_per_second": 58.765,
-      "eval_steps_per_second": 1.959,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.9792068600654602,
+      "eval_runtime": 1.0489,
+      "eval_samples_per_second": 57.204,
+      "eval_steps_per_second": 1.907,
       "step": 693
     },
     {
       "epoch": 36.36,
       "learning_rate": 4.570637119113573e-06,
-      "loss": 5.554007488522326e+34,
+      "loss": 0.036,
       "step": 700
     },
     {
       "epoch": 36.88,
       "learning_rate": 3.8088642659279783e-06,
-      "loss": 5.57132319958063e+34,
+      "loss": 0.0195,
       "step": 710
     },
     {
       "epoch": 36.99,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0038,
-      "eval_samples_per_second": 59.77,
-      "eval_steps_per_second": 1.992,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.9458684325218201,
+      "eval_runtime": 1.0129,
+      "eval_samples_per_second": 59.236,
+      "eval_steps_per_second": 1.975,
       "step": 712
     },
     {
       "epoch": 37.4,
       "learning_rate": 3.0470914127423822e-06,
-      "loss": 5.631927990214292e+34,
+      "loss": 0.0349,
       "step": 720
     },
     {
       "epoch": 37.92,
       "learning_rate": 2.2853185595567866e-06,
-      "loss": 5.142759449922784e+34,
+      "loss": 0.031,
       "step": 730
     },
     {
       "epoch": 37.97,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0138,
-      "eval_samples_per_second": 59.181,
-      "eval_steps_per_second": 1.973,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.9487714767456055,
+      "eval_runtime": 1.0492,
+      "eval_samples_per_second": 57.189,
+      "eval_steps_per_second": 1.906,
       "step": 731
     },
     {
       "epoch": 38.44,
       "learning_rate": 1.5235457063711911e-06,
-      "loss": 5.709848491906258e+34,
+      "loss": 0.0221,
       "step": 740
     },
     {
       "epoch": 38.96,
       "learning_rate": 7.617728531855956e-07,
-      "loss": 5.259640796671952e+34,
+      "loss": 0.0224,
       "step": 750
     },
     {
       "epoch": 38.96,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0051,
-      "eval_samples_per_second": 59.693,
-      "eval_steps_per_second": 1.99,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.943988561630249,
+      "eval_runtime": 1.0748,
+      "eval_samples_per_second": 55.822,
+      "eval_steps_per_second": 1.861,
       "step": 750
     },
     {
       "epoch": 39.48,
       "learning_rate": 0.0,
-      "loss": 5.350547586481632e+34,
+      "loss": 0.0309,
       "step": 760
     },
     {
       "epoch": 39.48,
-      "eval_accuracy": 0.7833333333333333,
-      "eval_loss": 4.986918644609257e+34,
-      "eval_runtime": 1.0125,
-      "eval_samples_per_second": 59.259,
-      "eval_steps_per_second": 1.975,
+      "eval_accuracy": 0.8166666666666667,
+      "eval_loss": 0.9447524547576904,
+      "eval_runtime": 1.0329,
+      "eval_samples_per_second": 58.09,
+      "eval_steps_per_second": 1.936,
       "step": 760
     },
     {
       "epoch": 39.48,
       "step": 760,
       "total_flos": 7.496244493905936e+18,
-      "train_loss": 5.490440419440805e+34,
-      "train_runtime": 1489.0102,
-      "train_samples_per_second": 65.816,
-      "train_steps_per_second": 0.51
+      "train_loss": 0.1845952101443943,
+      "train_runtime": 1526.0361,
+      "train_samples_per_second": 64.219,
+      "train_steps_per_second": 0.498
     }
   ],
   "logging_steps": 10,