End of training

README.md

@@ -18,7 +18,7 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [MCG-NJU/videomae-base](https://huggingface.co/MCG-NJU/videomae-base) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.2398
+- Loss: 0.2412
 - Accuracy: 0.9484
 
 ## Model description

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 7.101351351351352,
+    "eval_accuracy": 0.9483870967741935,
+    "eval_loss": 0.24116545915603638,
+    "eval_runtime": 27.3163,
+    "eval_samples_per_second": 5.674,
+    "eval_steps_per_second": 0.366
+}

runs/Aug20_13-15-14_8feefc1eac62/events.out.tfevents.1724161490.8feefc1eac62.34.1

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:f9ad8469aa53f04e9a48ddbe11fe0c4009de35aa491db295ce842a7e7aa4338c
-size 411
+oid sha256:c238f5c3cb3cf05433cdf9cf022671ff20d27fbf47cb32b098a489cb1b1a7bba
+size 734

test_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 7.101351351351352,
+    "eval_accuracy": 0.9483870967741935,
+    "eval_loss": 0.24116545915603638,
+    "eval_runtime": 27.3163,
+    "eval_samples_per_second": 5.674,
+    "eval_steps_per_second": 0.366
+}

trainer_state.json

@@ -0,0 +1,230 @@
+{
+  "best_metric": 0.9,
+  "best_model_checkpoint": "videomae-base-finetuned-ucf101-subset/checkpoint-133",
+  "epoch": 7.101351351351352,
+  "eval_steps": 500,
+  "global_step": 148,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.06756756756756757,
+      "grad_norm": 5.574437618255615,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 2.3597,
+      "step": 10
+    },
+    {
+      "epoch": 0.12837837837837837,
+      "eval_accuracy": 0.4714285714285714,
+      "eval_loss": 2.0514097213745117,
+      "eval_runtime": 11.8472,
+      "eval_samples_per_second": 5.909,
+      "eval_steps_per_second": 0.422,
+      "step": 19
+    },
+    {
+      "epoch": 1.0067567567567568,
+      "grad_norm": 5.533560752868652,
+      "learning_rate": 4.81203007518797e-05,
+      "loss": 2.2709,
+      "step": 20
+    },
+    {
+      "epoch": 1.0743243243243243,
+      "grad_norm": 5.730343818664551,
+      "learning_rate": 4.43609022556391e-05,
+      "loss": 1.9928,
+      "step": 30
+    },
+    {
+      "epoch": 1.1283783783783783,
+      "eval_accuracy": 0.5428571428571428,
+      "eval_loss": 1.4878058433532715,
+      "eval_runtime": 11.5879,
+      "eval_samples_per_second": 6.041,
+      "eval_steps_per_second": 0.431,
+      "step": 38
+    },
+    {
+      "epoch": 2.0135135135135136,
+      "grad_norm": 8.42638111114502,
+      "learning_rate": 4.0601503759398494e-05,
+      "loss": 1.5737,
+      "step": 40
+    },
+    {
+      "epoch": 2.081081081081081,
+      "grad_norm": 8.014702796936035,
+      "learning_rate": 3.6842105263157895e-05,
+      "loss": 1.1016,
+      "step": 50
+    },
+    {
+      "epoch": 2.1283783783783785,
+      "eval_accuracy": 0.7857142857142857,
+      "eval_loss": 0.7900044322013855,
+      "eval_runtime": 12.0411,
+      "eval_samples_per_second": 5.813,
+      "eval_steps_per_second": 0.415,
+      "step": 57
+    },
+    {
+      "epoch": 3.02027027027027,
+      "grad_norm": 5.1834025382995605,
+      "learning_rate": 3.3082706766917295e-05,
+      "loss": 0.742,
+      "step": 60
+    },
+    {
+      "epoch": 3.0878378378378377,
+      "grad_norm": 5.8055925369262695,
+      "learning_rate": 2.9323308270676693e-05,
+      "loss": 0.4919,
+      "step": 70
+    },
+    {
+      "epoch": 3.1283783783783785,
+      "eval_accuracy": 0.8285714285714286,
+      "eval_loss": 0.51225346326828,
+      "eval_runtime": 11.8089,
+      "eval_samples_per_second": 5.928,
+      "eval_steps_per_second": 0.423,
+      "step": 76
+    },
+    {
+      "epoch": 4.027027027027027,
+      "grad_norm": 2.2770769596099854,
+      "learning_rate": 2.556390977443609e-05,
+      "loss": 0.3665,
+      "step": 80
+    },
+    {
+      "epoch": 4.094594594594595,
+      "grad_norm": 7.455367565155029,
+      "learning_rate": 2.1804511278195487e-05,
+      "loss": 0.324,
+      "step": 90
+    },
+    {
+      "epoch": 4.128378378378378,
+      "eval_accuracy": 0.8714285714285714,
+      "eval_loss": 0.3622731864452362,
+      "eval_runtime": 11.8576,
+      "eval_samples_per_second": 5.903,
+      "eval_steps_per_second": 0.422,
+      "step": 95
+    },
+    {
+      "epoch": 5.033783783783784,
+      "grad_norm": 4.8652849197387695,
+      "learning_rate": 1.8045112781954888e-05,
+      "loss": 0.2438,
+      "step": 100
+    },
+    {
+      "epoch": 5.101351351351352,
+      "grad_norm": 7.068449020385742,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 0.167,
+      "step": 110
+    },
+    {
+      "epoch": 5.128378378378378,
+      "eval_accuracy": 0.8857142857142857,
+      "eval_loss": 0.2943699359893799,
+      "eval_runtime": 11.8441,
+      "eval_samples_per_second": 5.91,
+      "eval_steps_per_second": 0.422,
+      "step": 114
+    },
+    {
+      "epoch": 6.04054054054054,
+      "grad_norm": 2.7641441822052,
+      "learning_rate": 1.0526315789473684e-05,
+      "loss": 0.173,
+      "step": 120
+    },
+    {
+      "epoch": 6.108108108108108,
+      "grad_norm": 5.483329772949219,
+      "learning_rate": 6.766917293233083e-06,
+      "loss": 0.1024,
+      "step": 130
+    },
+    {
+      "epoch": 6.128378378378378,
+      "eval_accuracy": 0.9,
+      "eval_loss": 0.22581163048744202,
+      "eval_runtime": 11.8839,
+      "eval_samples_per_second": 5.89,
+      "eval_steps_per_second": 0.421,
+      "step": 133
+    },
+    {
+      "epoch": 7.047297297297297,
+      "grad_norm": 0.6104542016983032,
+      "learning_rate": 3.007518796992481e-06,
+      "loss": 0.0915,
+      "step": 140
+    },
+    {
+      "epoch": 7.101351351351352,
+      "eval_accuracy": 0.9,
+      "eval_loss": 0.22488869726657867,
+      "eval_runtime": 12.803,
+      "eval_samples_per_second": 5.467,
+      "eval_steps_per_second": 0.391,
+      "step": 148
+    },
+    {
+      "epoch": 7.101351351351352,
+      "step": 148,
+      "total_flos": 2.915999166844109e+18,
+      "train_loss": 0.8144391336151071,
+      "train_runtime": 1603.3969,
+      "train_samples_per_second": 1.477,
+      "train_steps_per_second": 0.092
+    },
+    {
+      "epoch": 7.101351351351352,
+      "eval_accuracy": 0.9483870967741935,
+      "eval_loss": 0.23980052769184113,
+      "eval_runtime": 27.6397,
+      "eval_samples_per_second": 5.608,
+      "eval_steps_per_second": 0.362,
+      "step": 148
+    },
+    {
+      "epoch": 7.101351351351352,
+      "eval_accuracy": 0.9483870967741935,
+      "eval_loss": 0.24116545915603638,
+      "eval_runtime": 27.3163,
+      "eval_samples_per_second": 5.674,
+      "eval_steps_per_second": 0.366,
+      "step": 148
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 148,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.915999166844109e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}