End of training

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.5,
-    "eval_accuracy": 0.7548387096774194,
-    "eval_loss": 0.8304252624511719,
-    "eval_runtime": 28.5162,
-    "eval_samples_per_second": 5.435,
-    "eval_steps_per_second": 1.368
+    "epoch": 1.0,
+    "eval_accuracy": 0.8129032258064516,
+    "eval_loss": 0.5109545588493347,
+    "eval_runtime": 27.7194,
+    "eval_samples_per_second": 5.592,
+    "eval_steps_per_second": 1.407
 }

runs/May22_12-56-09_940520f3922b/events.out.tfevents.1716382778.940520f3922b.34.3

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:002442083b589f464caeb8fdd4480c214698a1bdc75c3e36c9da95aceeef99e6
-size 405
+oid sha256:f0a0fbfa3a79d6dbf24fc976de7897320a990513c19b89ac9c024b5476f9d0b9
+size 722

test_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 1.5,
-    "eval_accuracy": 0.7548387096774194,
-    "eval_loss": 0.8304252624511719,
-    "eval_runtime": 28.5162,
-    "eval_samples_per_second": 5.435,
-    "eval_steps_per_second": 1.368
+    "epoch": 1.0,
+    "eval_accuracy": 0.8129032258064516,
+    "eval_loss": 0.5109545588493347,
+    "eval_runtime": 27.7194,
+    "eval_samples_per_second": 5.592,
+    "eval_steps_per_second": 1.407
 }

trainer_state.json

@@ -1,170 +1,105 @@
 {
-  "best_metric": 0.6428571428571429,
-  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-150",
-  "epoch": 1.5,
+  "best_metric": 0.7142857142857143,
+  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-75",
+  "epoch": 1.0,
   "eval_steps": 500,
-  "global_step": 150,
+  "global_step": 75,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.07,
-      "grad_norm": 9.588947296142578,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 2.3797,
+      "epoch": 0.13,
+      "grad_norm": 10.564898490905762,
+      "learning_rate": 4.850746268656717e-05,
+      "loss": 0.8833,
       "step": 10
     },
     {
-      "epoch": 0.13,
-      "grad_norm": 10.572150230407715,
-      "learning_rate": 4.814814814814815e-05,
-      "loss": 2.2698,
+      "epoch": 0.27,
+      "grad_norm": 20.67559051513672,
+      "learning_rate": 4.104477611940299e-05,
+      "loss": 0.9306,
       "step": 20
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 10.148682594299316,
-      "learning_rate": 4.4444444444444447e-05,
-      "loss": 2.1607,
+      "epoch": 0.4,
+      "grad_norm": 16.63642120361328,
+      "learning_rate": 3.358208955223881e-05,
+      "loss": 0.7969,
       "step": 30
     },
     {
-      "epoch": 0.27,
-      "grad_norm": 10.596107482910156,
-      "learning_rate": 4.074074074074074e-05,
-      "loss": 2.0383,
+      "epoch": 0.53,
+      "grad_norm": 3.232578992843628,
+      "learning_rate": 2.6119402985074626e-05,
+      "loss": 0.8249,
       "step": 40
     },
     {
-      "epoch": 0.33,
-      "grad_norm": 13.56712532043457,
-      "learning_rate": 3.7037037037037037e-05,
-      "loss": 2.0051,
+      "epoch": 0.67,
+      "grad_norm": 16.193866729736328,
+      "learning_rate": 1.865671641791045e-05,
+      "loss": 0.7562,
       "step": 50
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 10.223563194274902,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 1.8691,
+      "epoch": 0.8,
+      "grad_norm": 2.896273374557495,
+      "learning_rate": 1.119402985074627e-05,
+      "loss": 0.59,
       "step": 60
     },
     {
-      "epoch": 0.47,
-      "grad_norm": 15.075571060180664,
-      "learning_rate": 2.962962962962963e-05,
-      "loss": 1.8113,
+      "epoch": 0.93,
+      "grad_norm": 5.816062927246094,
+      "learning_rate": 3.7313432835820893e-06,
+      "loss": 0.4871,
       "step": 70
     },
     {
-      "epoch": 0.5,
-      "eval_accuracy": 0.38571428571428573,
-      "eval_loss": 1.5965501070022583,
-      "eval_runtime": 13.1326,
-      "eval_samples_per_second": 5.33,
-      "eval_steps_per_second": 1.371,
+      "epoch": 1.0,
+      "eval_accuracy": 0.7142857142857143,
+      "eval_loss": 0.7370076775550842,
+      "eval_runtime": 12.5539,
+      "eval_samples_per_second": 5.576,
+      "eval_steps_per_second": 1.434,
       "step": 75
     },
     {
-      "epoch": 1.03,
-      "grad_norm": 10.883992195129395,
-      "learning_rate": 2.5925925925925925e-05,
-      "loss": 1.4088,
-      "step": 80
-    },
-    {
-      "epoch": 1.1,
-      "grad_norm": 12.44357967376709,
-      "learning_rate": 2.2222222222222223e-05,
-      "loss": 1.2583,
-      "step": 90
-    },
-    {
-      "epoch": 1.17,
-      "grad_norm": 14.58968448638916,
-      "learning_rate": 1.8518518518518518e-05,
-      "loss": 1.1709,
-      "step": 100
-    },
-    {
-      "epoch": 1.23,
-      "grad_norm": 8.881256103515625,
-      "learning_rate": 1.4814814814814815e-05,
-      "loss": 0.9056,
-      "step": 110
-    },
-    {
-      "epoch": 1.3,
-      "grad_norm": 11.71822738647461,
-      "learning_rate": 1.1111111111111112e-05,
-      "loss": 0.9813,
-      "step": 120
-    },
-    {
-      "epoch": 1.37,
-      "grad_norm": 5.7398858070373535,
-      "learning_rate": 7.4074074074074075e-06,
-      "loss": 0.709,
-      "step": 130
-    },
-    {
-      "epoch": 1.43,
-      "grad_norm": 9.218167304992676,
-      "learning_rate": 3.7037037037037037e-06,
-      "loss": 0.6339,
-      "step": 140
-    },
-    {
-      "epoch": 1.5,
-      "grad_norm": 35.40492630004883,
-      "learning_rate": 0.0,
-      "loss": 0.895,
-      "step": 150
+      "epoch": 1.0,
+      "step": 75,
+      "total_flos": 3.73846047031296e+17,
+      "train_loss": 0.7330940500895182,
+      "train_runtime": 114.9093,
+      "train_samples_per_second": 2.611,
+      "train_steps_per_second": 0.653
     },
     {
-      "epoch": 1.5,
-      "eval_accuracy": 0.6428571428571429,
-      "eval_loss": 0.9439693093299866,
-      "eval_runtime": 13.1412,
-      "eval_samples_per_second": 5.327,
-      "eval_steps_per_second": 1.37,
-      "step": 150
-    },
-    {
-      "epoch": 1.5,
-      "step": 150,
-      "total_flos": 7.47692094062592e+17,
-      "train_loss": 1.4997930717468262,
-      "train_runtime": 266.4694,
-      "train_samples_per_second": 2.252,
-      "train_steps_per_second": 0.563
-    },
-    {
-      "epoch": 1.5,
-      "eval_accuracy": 0.7548387096774194,
-      "eval_loss": 0.8304252624511719,
-      "eval_runtime": 29.2907,
-      "eval_samples_per_second": 5.292,
-      "eval_steps_per_second": 1.331,
-      "step": 150
+      "epoch": 1.0,
+      "eval_accuracy": 0.8129032258064516,
+      "eval_loss": 0.5109546184539795,
+      "eval_runtime": 28.4314,
+      "eval_samples_per_second": 5.452,
+      "eval_steps_per_second": 1.372,
+      "step": 75
     },
     {
-      "epoch": 1.5,
-      "eval_accuracy": 0.7548387096774194,
-      "eval_loss": 0.8304252624511719,
-      "eval_runtime": 28.5162,
-      "eval_samples_per_second": 5.435,
-      "eval_steps_per_second": 1.368,
-      "step": 150
+      "epoch": 1.0,
+      "eval_accuracy": 0.8129032258064516,
+      "eval_loss": 0.5109545588493347,
+      "eval_runtime": 27.7194,
+      "eval_samples_per_second": 5.592,
+      "eval_steps_per_second": 1.407,
+      "step": 75
     }
   ],
   "logging_steps": 10,
-  "max_steps": 150,
+  "max_steps": 75,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 9223372036854775807,
   "save_steps": 500,
-  "total_flos": 7.47692094062592e+17,
+  "total_flos": 3.73846047031296e+17,
   "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null