End of training

README.md

@@ -17,8 +17,8 @@ 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: 0.6328
-- Accuracy: 0.7692
+- Loss: 0.6847
+- Accuracy: 0.75
 
 ## Model description
 

all_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.2,
+    "eval_accuracy": 0.75,
+    "eval_loss": 0.6846519112586975,
+    "eval_runtime": 14.0555,
+    "eval_samples_per_second": 0.854,
+    "eval_steps_per_second": 0.427
+}

runs/Jul02_19-45-22_c6d9a0630dcb/events.out.tfevents.1719950183.c6d9a0630dcb.4133.5

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d32e9f043196da7e8b5d47cb6e62eb6fdbdeb1c7e2071a09d4136cdb086b849b
-size 405
+oid sha256:ab438901554fb6f68f8f2b184c3b5e8a747d69a21c4031db162e88e4ddd688d3
+size 722

test_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 4.2,
+    "eval_accuracy": 0.75,
+    "eval_loss": 0.6846519112586975,
+    "eval_runtime": 14.0555,
+    "eval_samples_per_second": 0.854,
+    "eval_steps_per_second": 0.427
+}

trainer_state.json

@@ -0,0 +1,189 @@
+{
+  "best_metric": 1.0,
+  "best_model_checkpoint": "videomae-base-finetuned-Risky-situations/checkpoint-25",
+  "epoch": 4.2,
+  "eval_steps": 500,
+  "global_step": 125,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.08,
+      "grad_norm": 8.516080856323242,
+      "learning_rate": 3.846153846153846e-05,
+      "loss": 0.5136,
+      "step": 10
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 40.52229690551758,
+      "learning_rate": 4.6875e-05,
+      "loss": 0.4618,
+      "step": 20
+    },
+    {
+      "epoch": 0.2,
+      "eval_accuracy": 1.0,
+      "eval_loss": 0.010277315974235535,
+      "eval_runtime": 11.9488,
+      "eval_samples_per_second": 1.004,
+      "eval_steps_per_second": 0.502,
+      "step": 25
+    },
+    {
+      "epoch": 1.04,
+      "grad_norm": 110.08763122558594,
+      "learning_rate": 4.2410714285714285e-05,
+      "loss": 0.0748,
+      "step": 30
+    },
+    {
+      "epoch": 1.12,
+      "grad_norm": 14.86829662322998,
+      "learning_rate": 3.794642857142857e-05,
+      "loss": 0.6214,
+      "step": 40
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 0.27806639671325684,
+      "learning_rate": 3.348214285714286e-05,
+      "loss": 0.3548,
+      "step": 50
+    },
+    {
+      "epoch": 1.2,
+      "eval_accuracy": 1.0,
+      "eval_loss": 0.002936301752924919,
+      "eval_runtime": 11.6674,
+      "eval_samples_per_second": 1.029,
+      "eval_steps_per_second": 0.514,
+      "step": 50
+    },
+    {
+      "epoch": 2.08,
+      "grad_norm": 0.032773926854133606,
+      "learning_rate": 2.9017857142857146e-05,
+      "loss": 0.2253,
+      "step": 60
+    },
+    {
+      "epoch": 2.16,
+      "grad_norm": 0.036098264157772064,
+      "learning_rate": 2.455357142857143e-05,
+      "loss": 0.2182,
+      "step": 70
+    },
+    {
+      "epoch": 2.2,
+      "eval_accuracy": 1.0,
+      "eval_loss": 0.0008991466020233929,
+      "eval_runtime": 11.8302,
+      "eval_samples_per_second": 1.014,
+      "eval_steps_per_second": 0.507,
+      "step": 75
+    },
+    {
+      "epoch": 3.04,
+      "grad_norm": 0.034782785922288895,
+      "learning_rate": 2.0089285714285717e-05,
+      "loss": 0.0011,
+      "step": 80
+    },
+    {
+      "epoch": 3.12,
+      "grad_norm": 0.4517887532711029,
+      "learning_rate": 1.5625e-05,
+      "loss": 0.3581,
+      "step": 90
+    },
+    {
+      "epoch": 3.2,
+      "grad_norm": 3.7396240234375,
+      "learning_rate": 1.1160714285714287e-05,
+      "loss": 0.2513,
+      "step": 100
+    },
+    {
+      "epoch": 3.2,
+      "eval_accuracy": 1.0,
+      "eval_loss": 0.003270552260801196,
+      "eval_runtime": 11.5723,
+      "eval_samples_per_second": 1.037,
+      "eval_steps_per_second": 0.518,
+      "step": 100
+    },
+    {
+      "epoch": 4.08,
+      "grad_norm": 357.18890380859375,
+      "learning_rate": 6.696428571428572e-06,
+      "loss": 0.2286,
+      "step": 110
+    },
+    {
+      "epoch": 4.16,
+      "grad_norm": 0.019339008256793022,
+      "learning_rate": 2.2321428571428573e-06,
+      "loss": 0.0021,
+      "step": 120
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 1.0,
+      "eval_loss": 0.0007905661477707326,
+      "eval_runtime": 14.2671,
+      "eval_samples_per_second": 0.841,
+      "eval_steps_per_second": 0.421,
+      "step": 125
+    },
+    {
+      "epoch": 4.2,
+      "step": 125,
+      "total_flos": 3.11516147367936e+17,
+      "train_loss": 0.26490905868075787,
+      "train_runtime": 483.7689,
+      "train_samples_per_second": 0.517,
+      "train_steps_per_second": 0.258
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.7692307692307693,
+      "eval_loss": 0.6328043341636658,
+      "eval_runtime": 16.8871,
+      "eval_samples_per_second": 0.77,
+      "eval_steps_per_second": 0.415,
+      "step": 125
+    },
+    {
+      "epoch": 4.2,
+      "eval_accuracy": 0.75,
+      "eval_loss": 0.6846519112586975,
+      "eval_runtime": 14.0555,
+      "eval_samples_per_second": 0.854,
+      "eval_steps_per_second": 0.427,
+      "step": 125
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 125,
+  "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": 3.11516147367936e+17,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}