Model save

README.md

@@ -17,8 +17,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/clip-vit-large-patch14](https://huggingface.co/openai/clip-vit-large-patch14) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 2.0961
-- Accuracy: 0.4377
+- Loss: 2.1829
+- Accuracy: 0.2367
 
 ## Model description
 
@@ -38,11 +38,11 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 5e-05
-- train_batch_size: 128
-- eval_batch_size: 128
+- train_batch_size: 32
+- eval_batch_size: 32
 - seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 512
+- gradient_accumulation_steps: 30
+- total_train_batch_size: 960
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
@@ -52,9 +52,9 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch  | Step | Validation Loss | Accuracy |
 |:-------------:|:------:|:----:|:---------------:|:--------:|
-| 2.2376        | 0.9953 | 105  | 2.2017          | 0.2982   |
-| 2.1641        | 2.0    | 211  | 2.1208          | 0.4042   |
-| 2.1453        | 2.9858 | 315  | 2.0961          | 0.4377   |
+| 2.2886        | 0.9953 | 56   | 2.2661          | 0.157    |
+| 2.2153        | 1.9905 | 112  | 2.2004          | 0.1945   |
+| 2.1981        | 2.9858 | 168  | 2.1829          | 0.2367   |
 
 
 ### Framework versions