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 [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.0362
-- Accuracy: 0.6078
+- Loss: 0.8868
+- Accuracy: 0.6360
 
 ## Model description
 
@@ -46,17 +46,16 @@ 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
-- num_epochs: 5
+- num_epochs: 4
 
 ### Training results
 
-| Training Loss | Epoch  | Step | Validation Loss | Accuracy |
-|:-------------:|:------:|:----:|:---------------:|:--------:|
-| 1.4035        | 0.9767 | 21   | 1.2942          | 0.4735   |
-| 1.2173        | 2.0    | 43   | 1.1946          | 0.4382   |
-| 1.1276        | 2.9767 | 64   | 1.1574          | 0.5300   |
-| 1.0841        | 4.0    | 86   | 1.1034          | 0.5336   |
-| 1.0692        | 4.8837 | 105  | 1.0362          | 0.6078   |
+| Training Loss | Epoch | Step | Validation Loss | Accuracy |
+|:-------------:|:-----:|:----:|:---------------:|:--------:|
+| 1.4027        | 1.0   | 68   | 1.1856          | 0.5300   |
+| 1.0773        | 2.0   | 136  | 0.9565          | 0.6290   |
+| 0.9342        | 3.0   | 204  | 0.9063          | 0.6184   |
+| 0.9003        | 4.0   | 272  | 0.8868          | 0.6360   |
 
 
 ### Framework versions