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: 0.1957
-- Accuracy: 0.9429
+- Loss: 0.6407
+- Accuracy: 0.7286
 
 ## Model description
 
@@ -44,16 +44,14 @@ The following hyperparameters were used during training:
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
-- training_steps: 600
+- training_steps: 300
 
 ### Training results
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 1.2685        | 0.25  | 150  | 1.1467          | 0.5857   |
-| 1.2176        | 1.25  | 300  | 1.0170          | 0.6714   |
-| 0.2058        | 2.25  | 450  | 0.5327          | 0.8143   |
-| 0.2871        | 3.25  | 600  | 0.1957          | 0.9429   |
+| 1.1406        | 0.5   | 150  | 1.5088          | 0.4      |
+| 0.5076        | 1.5   | 300  | 0.6407          | 0.7286   |
 
 
 ### Framework versions