End of training

all_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8709677419354839,
-    "eval_loss": 0.5119308829307556,
-    "eval_runtime": 12.5062,
-    "eval_samples_per_second": 12.394,
-    "eval_steps_per_second": 1.599
+    "eval_accuracy": 0.8903225806451613,
+    "eval_loss": 0.4209001660346985,
+    "eval_runtime": 15.6361,
+    "eval_samples_per_second": 9.913,
+    "eval_steps_per_second": 1.279
 }

runs/Oct15_17-32-31_b6767a947022/events.out.tfevents.1729013827.b6767a947022.7949.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:8560e201cadb4f1af1f532168408bb81279c36f2d398ad9a95b80eb5a62c9ceb
-size 411
+oid sha256:f2a04819cdbd36e0cc216b6a442ebe8a04ddbb170e38e6a3df0bdedc65384673
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
     "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8709677419354839,
-    "eval_loss": 0.5119308829307556,
-    "eval_runtime": 12.5062,
-    "eval_samples_per_second": 12.394,
-    "eval_steps_per_second": 1.599
+    "eval_accuracy": 0.8903225806451613,
+    "eval_loss": 0.4209001660346985,
+    "eval_runtime": 15.6361,
+    "eval_samples_per_second": 9.913,
+    "eval_steps_per_second": 1.279
 }

trainer_state.json

@@ -1,5 +1,5 @@
 {
-  "best_metric": 0.9142857142857143,
+  "best_metric": 0.8857142857142857,
   "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-148",
   "epoch": 3.22972972972973,
   "eval_steps": 500,
@@ -10,163 +10,163 @@
   "log_history": [
     {
       "epoch": 0.06756756756756757,
-      "grad_norm": 6.351539134979248,
+      "grad_norm": 6.64535665512085,
       "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.2741,
+      "loss": 2.3547,
       "step": 10
     },
     {
       "epoch": 0.13513513513513514,
-      "grad_norm": 7.739011287689209,
+      "grad_norm": 6.657472610473633,
       "learning_rate": 4.81203007518797e-05,
-      "loss": 2.2121,
+      "loss": 2.1896,
       "step": 20
     },
     {
       "epoch": 0.20270270270270271,
-      "grad_norm": 8.185063362121582,
+      "grad_norm": 8.04188346862793,
       "learning_rate": 4.43609022556391e-05,
-      "loss": 2.1008,
+      "loss": 2.0919,
       "step": 30
     },
     {
       "epoch": 0.25675675675675674,
-      "eval_accuracy": 0.4142857142857143,
-      "eval_loss": 1.834967017173767,
-      "eval_runtime": 6.5432,
-      "eval_samples_per_second": 10.698,
-      "eval_steps_per_second": 1.375,
+      "eval_accuracy": 0.5285714285714286,
+      "eval_loss": 1.8006081581115723,
+      "eval_runtime": 7.0865,
+      "eval_samples_per_second": 9.878,
+      "eval_steps_per_second": 1.27,
       "step": 38
     },
     {
       "epoch": 1.0135135135135136,
-      "grad_norm": 8.868829727172852,
+      "grad_norm": 10.020313262939453,
       "learning_rate": 4.0601503759398494e-05,
-      "loss": 1.952,
+      "loss": 1.9064,
       "step": 40
     },
     {
       "epoch": 1.0810810810810811,
-      "grad_norm": 10.57504940032959,
+      "grad_norm": 11.511276245117188,
       "learning_rate": 3.6842105263157895e-05,
-      "loss": 1.5486,
+      "loss": 1.5879,
       "step": 50
     },
     {
       "epoch": 1.1486486486486487,
-      "grad_norm": 9.441951751708984,
+      "grad_norm": 10.105522155761719,
       "learning_rate": 3.3082706766917295e-05,
-      "loss": 1.2984,
+      "loss": 1.2609,
       "step": 60
     },
     {
       "epoch": 1.2162162162162162,
-      "grad_norm": 7.836465358734131,
+      "grad_norm": 5.39172887802124,
       "learning_rate": 2.9323308270676693e-05,
-      "loss": 0.9279,
+      "loss": 0.9101,
       "step": 70
     },
     {
       "epoch": 1.2567567567567568,
-      "eval_accuracy": 0.7142857142857143,
-      "eval_loss": 1.0198957920074463,
-      "eval_runtime": 6.5052,
-      "eval_samples_per_second": 10.761,
-      "eval_steps_per_second": 1.384,
+      "eval_accuracy": 0.6857142857142857,
+      "eval_loss": 0.9759382009506226,
+      "eval_runtime": 6.7289,
+      "eval_samples_per_second": 10.403,
+      "eval_steps_per_second": 1.338,
       "step": 76
     },
     {
       "epoch": 2.027027027027027,
-      "grad_norm": 10.87349796295166,
+      "grad_norm": 6.035886764526367,
       "learning_rate": 2.556390977443609e-05,
-      "loss": 1.0519,
+      "loss": 0.9083,
       "step": 80
     },
     {
       "epoch": 2.0945945945945947,
-      "grad_norm": 9.65088176727295,
+      "grad_norm": 17.813278198242188,
       "learning_rate": 2.1804511278195487e-05,
-      "loss": 0.7748,
+      "loss": 0.6671,
       "step": 90
     },
     {
       "epoch": 2.1621621621621623,
-      "grad_norm": 12.413875579833984,
+      "grad_norm": 11.762858390808105,
       "learning_rate": 1.8045112781954888e-05,
-      "loss": 0.6615,
+      "loss": 0.5856,
       "step": 100
     },
     {
       "epoch": 2.22972972972973,
-      "grad_norm": 7.036044597625732,
+      "grad_norm": 4.943543434143066,
       "learning_rate": 1.4285714285714285e-05,
-      "loss": 0.5482,
+      "loss": 0.4855,
       "step": 110
     },
     {
       "epoch": 2.2567567567567566,
-      "eval_accuracy": 0.8,
-      "eval_loss": 0.6185019612312317,
-      "eval_runtime": 6.5402,
-      "eval_samples_per_second": 10.703,
-      "eval_steps_per_second": 1.376,
+      "eval_accuracy": 0.8285714285714286,
+      "eval_loss": 0.5223344564437866,
+      "eval_runtime": 6.5279,
+      "eval_samples_per_second": 10.723,
+      "eval_steps_per_second": 1.379,
       "step": 114
     },
     {
       "epoch": 3.0405405405405403,
-      "grad_norm": 3.553114175796509,
+      "grad_norm": 3.349546194076538,
       "learning_rate": 1.0526315789473684e-05,
-      "loss": 0.4077,
+      "loss": 0.3393,
       "step": 120
     },
     {
       "epoch": 3.108108108108108,
-      "grad_norm": 3.6690454483032227,
+      "grad_norm": 2.945603609085083,
       "learning_rate": 6.766917293233083e-06,
-      "loss": 0.3242,
+      "loss": 0.2937,
       "step": 130
     },
     {
       "epoch": 3.175675675675676,
-      "grad_norm": 13.059000015258789,
+      "grad_norm": 7.3917717933654785,
       "learning_rate": 3.007518796992481e-06,
-      "loss": 0.3476,
+      "loss": 0.2876,
       "step": 140
     },
     {
       "epoch": 3.22972972972973,
-      "eval_accuracy": 0.9142857142857143,
-      "eval_loss": 0.40149447321891785,
-      "eval_runtime": 6.6746,
-      "eval_samples_per_second": 10.487,
-      "eval_steps_per_second": 1.348,
+      "eval_accuracy": 0.8857142857142857,
+      "eval_loss": 0.40336811542510986,
+      "eval_runtime": 8.2936,
+      "eval_samples_per_second": 8.44,
+      "eval_steps_per_second": 1.085,
       "step": 148
     },
     {
       "epoch": 3.22972972972973,
       "step": 148,
       "total_flos": 1.460491890402263e+18,
-      "train_loss": 1.1264432153186283,
-      "train_runtime": 223.8174,
-      "train_samples_per_second": 5.29,
-      "train_steps_per_second": 0.661
+      "train_loss": 1.086421162695498,
+      "train_runtime": 248.8416,
+      "train_samples_per_second": 4.758,
+      "train_steps_per_second": 0.595
     },
     {
       "epoch": 3.22972972972973,
-      "eval_accuracy": 0.8709677419354839,
-      "eval_loss": 0.5119309425354004,
-      "eval_runtime": 13.3003,
-      "eval_samples_per_second": 11.654,
-      "eval_steps_per_second": 1.504,
+      "eval_accuracy": 0.8903225806451613,
+      "eval_loss": 0.4209001958370209,
+      "eval_runtime": 16.3349,
+      "eval_samples_per_second": 9.489,
+      "eval_steps_per_second": 1.224,
       "step": 148
     },
     {
       "epoch": 3.22972972972973,
-      "eval_accuracy": 0.8709677419354839,
-      "eval_loss": 0.5119308829307556,
-      "eval_runtime": 12.5062,
-      "eval_samples_per_second": 12.394,
-      "eval_steps_per_second": 1.599,
+      "eval_accuracy": 0.8903225806451613,
+      "eval_loss": 0.4209001660346985,
+      "eval_runtime": 15.6361,
+      "eval_samples_per_second": 9.913,
+      "eval_steps_per_second": 1.279,
       "step": 148
     }
   ],