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 [google/vit-base-patch16-224-in21k](https://huggingface.co/google/vit-base-patch16-224-in21k) on an unknown dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3936
-- Accuracy: 0.9092
+- Loss: 0.4665
+- Accuracy: 0.9105
 
 ## Model description
 
@@ -38,11 +38,11 @@ More information needed
 
 The following hyperparameters were used during training:
 - learning_rate: 0.0005
-- train_batch_size: 32
-- eval_batch_size: 32
+- train_batch_size: 64
+- eval_batch_size: 64
 - seed: 42
-- gradient_accumulation_steps: 4
-- total_train_batch_size: 128
+- gradient_accumulation_steps: 8
+- total_train_batch_size: 512
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_ratio: 0.1
@@ -53,10 +53,10 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Accuracy |
 |:-------------:|:-----:|:----:|:---------------:|:--------:|
-| 3.1138        | 1.0   | 525  | 1.0237          | 0.7924   |
-| 0.7639        | 2.0   | 1050 | 0.6481          | 0.85     |
-| 0.4929        | 3.0   | 1575 | 0.4907          | 0.8836   |
-| 0.3308        | 4.0   | 2100 | 0.3936          | 0.9092   |
+| 4.3875        | 1.0   | 131  | 2.1054          | 0.7873   |
+| 1.1975        | 2.0   | 262  | 0.8185          | 0.8627   |
+| 0.5774        | 2.99  | 393  | 0.5684          | 0.8918   |
+| 0.4082        | 3.99  | 524  | 0.4665          | 0.9105   |
 
 
 ### Framework versions

runs/Mar19_10-17-58_f3b04c87f885/events.out.tfevents.1710843479.f3b04c87f885.34.3

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:09a200995f5410ae7f9682e9499bf10220076e04259121e90e198f432ad8492b
-size 37800
+oid sha256:058d70029c79b17a15ace0828d7552803e1029f84a058097dd66d5552ca96ae1
+size 38154