End of training

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8903225806451613,
-    "eval_loss": 0.4209001660346985,
-    "eval_runtime": 15.6361,
-    "eval_samples_per_second": 9.913,
-    "eval_steps_per_second": 1.279
+    "epoch": 4.180952380952381,
+    "eval_accuracy": 0.8395061728395061,
+    "eval_loss": 0.38502195477485657,
+    "eval_runtime": 14.6842,
+    "eval_samples_per_second": 11.032,
+    "eval_steps_per_second": 1.43
 }

runs/Oct22_17-35-56_8e489873f98b/events.out.tfevents.1729619014.8e489873f98b.9878.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:47a36eda906cf555f668758ce5203e214667baabdec9779c85890d4a76f8a2e1
-size 411
+oid sha256:43ba3f9ec4e6ba556fd82059e569fc67ea6478a89e2bdf76bd106842ec7e0cde
+size 734

test_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 3.22972972972973,
-    "eval_accuracy": 0.8903225806451613,
-    "eval_loss": 0.4209001660346985,
-    "eval_runtime": 15.6361,
-    "eval_samples_per_second": 9.913,
-    "eval_steps_per_second": 1.279
+    "epoch": 4.180952380952381,
+    "eval_accuracy": 0.8395061728395061,
+    "eval_loss": 0.38502195477485657,
+    "eval_runtime": 14.6842,
+    "eval_samples_per_second": 11.032,
+    "eval_steps_per_second": 1.43
 }

trainer_state.json

@@ -1,177 +1,235 @@
 {
-  "best_metric": 0.8857142857142857,
-  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-148",
-  "epoch": 3.22972972972973,
+  "best_metric": 0.9066666666666666,
+  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-210",
+  "epoch": 4.180952380952381,
   "eval_steps": 500,
-  "global_step": 148,
+  "global_step": 210,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.06756756756756757,
-      "grad_norm": 6.64535665512085,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.3547,
+      "epoch": 0.047619047619047616,
+      "grad_norm": 7.849704265594482,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 2.3976,
       "step": 10
     },
     {
-      "epoch": 0.13513513513513514,
-      "grad_norm": 6.657472610473633,
-      "learning_rate": 4.81203007518797e-05,
-      "loss": 2.1896,
+      "epoch": 0.09523809523809523,
+      "grad_norm": 7.76956033706665,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 2.3357,
       "step": 20
     },
     {
-      "epoch": 0.20270270270270271,
-      "grad_norm": 8.04188346862793,
-      "learning_rate": 4.43609022556391e-05,
-      "loss": 2.0919,
+      "epoch": 0.14285714285714285,
+      "grad_norm": 15.013307571411133,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 2.282,
       "step": 30
     },
     {
-      "epoch": 0.25675675675675674,
-      "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": 0.19047619047619047,
+      "grad_norm": 7.262270450592041,
+      "learning_rate": 4.4973544973544974e-05,
+      "loss": 2.0368,
+      "step": 40
     },
     {
-      "epoch": 1.0135135135135136,
-      "grad_norm": 10.020313262939453,
-      "learning_rate": 4.0601503759398494e-05,
-      "loss": 1.9064,
-      "step": 40
+      "epoch": 0.20476190476190476,
+      "eval_accuracy": 0.29333333333333333,
+      "eval_loss": 2.0273361206054688,
+      "eval_runtime": 7.264,
+      "eval_samples_per_second": 10.325,
+      "eval_steps_per_second": 1.377,
+      "step": 43
     },
     {
-      "epoch": 1.0810810810810811,
-      "grad_norm": 11.511276245117188,
-      "learning_rate": 3.6842105263157895e-05,
-      "loss": 1.5879,
+      "epoch": 1.0333333333333334,
+      "grad_norm": 8.3132963180542,
+      "learning_rate": 4.232804232804233e-05,
+      "loss": 1.947,
       "step": 50
     },
     {
-      "epoch": 1.1486486486486487,
-      "grad_norm": 10.105522155761719,
-      "learning_rate": 3.3082706766917295e-05,
-      "loss": 1.2609,
+      "epoch": 1.0809523809523809,
+      "grad_norm": 14.42203426361084,
+      "learning_rate": 3.968253968253968e-05,
+      "loss": 1.594,
       "step": 60
     },
     {
-      "epoch": 1.2162162162162162,
-      "grad_norm": 5.39172887802124,
-      "learning_rate": 2.9323308270676693e-05,
-      "loss": 0.9101,
+      "epoch": 1.1285714285714286,
+      "grad_norm": 11.382922172546387,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 1.2509,
       "step": 70
     },
     {
-      "epoch": 1.2567567567567568,
-      "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": 1.1761904761904762,
+      "grad_norm": 9.775425910949707,
+      "learning_rate": 3.439153439153439e-05,
+      "loss": 1.1679,
+      "step": 80
     },
     {
-      "epoch": 2.027027027027027,
-      "grad_norm": 6.035886764526367,
-      "learning_rate": 2.556390977443609e-05,
-      "loss": 0.9083,
-      "step": 80
+      "epoch": 1.2047619047619047,
+      "eval_accuracy": 0.56,
+      "eval_loss": 1.0918534994125366,
+      "eval_runtime": 7.1505,
+      "eval_samples_per_second": 10.489,
+      "eval_steps_per_second": 1.398,
+      "step": 86
     },
     {
-      "epoch": 2.0945945945945947,
-      "grad_norm": 17.813278198242188,
-      "learning_rate": 2.1804511278195487e-05,
-      "loss": 0.6671,
+      "epoch": 2.019047619047619,
+      "grad_norm": 16.352764129638672,
+      "learning_rate": 3.1746031746031745e-05,
+      "loss": 1.0361,
       "step": 90
     },
     {
-      "epoch": 2.1621621621621623,
-      "grad_norm": 11.762858390808105,
-      "learning_rate": 1.8045112781954888e-05,
-      "loss": 0.5856,
+      "epoch": 2.066666666666667,
+      "grad_norm": 5.250826358795166,
+      "learning_rate": 2.91005291005291e-05,
+      "loss": 0.7379,
       "step": 100
     },
     {
-      "epoch": 2.22972972972973,
-      "grad_norm": 4.943543434143066,
-      "learning_rate": 1.4285714285714285e-05,
-      "loss": 0.4855,
+      "epoch": 2.1142857142857143,
+      "grad_norm": 11.715250968933105,
+      "learning_rate": 2.6455026455026456e-05,
+      "loss": 0.7289,
       "step": 110
     },
     {
-      "epoch": 2.2567567567567566,
-      "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": 2.1619047619047618,
+      "grad_norm": 16.230215072631836,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 0.5097,
+      "step": 120
     },
     {
-      "epoch": 3.0405405405405403,
-      "grad_norm": 3.349546194076538,
-      "learning_rate": 1.0526315789473684e-05,
-      "loss": 0.3393,
-      "step": 120
+      "epoch": 2.204761904761905,
+      "eval_accuracy": 0.8266666666666667,
+      "eval_loss": 0.5804122090339661,
+      "eval_runtime": 6.6096,
+      "eval_samples_per_second": 11.347,
+      "eval_steps_per_second": 1.513,
+      "step": 129
     },
     {
-      "epoch": 3.108108108108108,
-      "grad_norm": 2.945603609085083,
-      "learning_rate": 6.766917293233083e-06,
-      "loss": 0.2937,
+      "epoch": 3.0047619047619047,
+      "grad_norm": 18.1605224609375,
+      "learning_rate": 2.1164021164021164e-05,
+      "loss": 0.4609,
       "step": 130
     },
     {
-      "epoch": 3.175675675675676,
-      "grad_norm": 7.3917717933654785,
-      "learning_rate": 3.007518796992481e-06,
-      "loss": 0.2876,
+      "epoch": 3.052380952380952,
+      "grad_norm": 7.160711288452148,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 0.3043,
       "step": 140
     },
     {
-      "epoch": 3.22972972972973,
-      "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": 1.086421162695498,
-      "train_runtime": 248.8416,
-      "train_samples_per_second": 4.758,
-      "train_steps_per_second": 0.595
-    },
-    {
-      "epoch": 3.22972972972973,
-      "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": 0.8903225806451613,
-      "eval_loss": 0.4209001660346985,
-      "eval_runtime": 15.6361,
-      "eval_samples_per_second": 9.913,
-      "eval_steps_per_second": 1.279,
-      "step": 148
+      "epoch": 3.1,
+      "grad_norm": 14.182634353637695,
+      "learning_rate": 1.5873015873015872e-05,
+      "loss": 0.336,
+      "step": 150
+    },
+    {
+      "epoch": 3.1476190476190475,
+      "grad_norm": 2.916877031326294,
+      "learning_rate": 1.3227513227513228e-05,
+      "loss": 0.2992,
+      "step": 160
+    },
+    {
+      "epoch": 3.1952380952380954,
+      "grad_norm": 8.268482208251953,
+      "learning_rate": 1.0582010582010582e-05,
+      "loss": 0.2293,
+      "step": 170
+    },
+    {
+      "epoch": 3.204761904761905,
+      "eval_accuracy": 0.8266666666666667,
+      "eval_loss": 0.4854774475097656,
+      "eval_runtime": 6.5331,
+      "eval_samples_per_second": 11.48,
+      "eval_steps_per_second": 1.531,
+      "step": 172
+    },
+    {
+      "epoch": 4.038095238095238,
+      "grad_norm": 7.389148712158203,
+      "learning_rate": 7.936507936507936e-06,
+      "loss": 0.3101,
+      "step": 180
+    },
+    {
+      "epoch": 4.085714285714285,
+      "grad_norm": 4.5858259201049805,
+      "learning_rate": 5.291005291005291e-06,
+      "loss": 0.1873,
+      "step": 190
+    },
+    {
+      "epoch": 4.133333333333334,
+      "grad_norm": 2.300278425216675,
+      "learning_rate": 2.6455026455026455e-06,
+      "loss": 0.2008,
+      "step": 200
+    },
+    {
+      "epoch": 4.180952380952381,
+      "grad_norm": 2.982243537902832,
+      "learning_rate": 0.0,
+      "loss": 0.1882,
+      "step": 210
+    },
+    {
+      "epoch": 4.180952380952381,
+      "eval_accuracy": 0.9066666666666666,
+      "eval_loss": 0.3435481786727905,
+      "eval_runtime": 6.9489,
+      "eval_samples_per_second": 10.793,
+      "eval_steps_per_second": 1.439,
+      "step": 210
+    },
+    {
+      "epoch": 4.180952380952381,
+      "step": 210,
+      "total_flos": 2.0736178988651643e+18,
+      "train_loss": 0.9781298257055737,
+      "train_runtime": 354.3465,
+      "train_samples_per_second": 4.741,
+      "train_steps_per_second": 0.593
+    },
+    {
+      "epoch": 4.180952380952381,
+      "eval_accuracy": 0.8395061728395061,
+      "eval_loss": 0.3850219249725342,
+      "eval_runtime": 60.1785,
+      "eval_samples_per_second": 2.692,
+      "eval_steps_per_second": 0.349,
+      "step": 210
+    },
+    {
+      "epoch": 4.180952380952381,
+      "eval_accuracy": 0.8395061728395061,
+      "eval_loss": 0.38502195477485657,
+      "eval_runtime": 14.6842,
+      "eval_samples_per_second": 11.032,
+      "eval_steps_per_second": 1.43,
+      "step": 210
     }
   ],
   "logging_steps": 10,
-  "max_steps": 148,
+  "max_steps": 210,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9223372036854775807,
   "save_steps": 500,
@@ -187,7 +245,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.460491890402263e+18,
+  "total_flos": 2.0736178988651643e+18,
   "train_batch_size": 8,
   "trial_name": null,
   "trial_params": null