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 [motheecreator/vit-Facial-Expression-Recognition](https://huggingface.co/motheecreator/vit-Facial-Expression-Recognition) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.4066
-- Accuracy: 0.8548
+- Loss: 0.7487
+- Accuracy: 0.8106
 
 ## Model description
 
@@ -43,31 +43,23 @@ The following hyperparameters were used during training:
 - seed: 42
 - gradient_accumulation_steps: 8
 - total_train_batch_size: 256
-- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
 - lr_scheduler_type: cosine
 - lr_scheduler_warmup_steps: 1000
-- num_epochs: 10
+- num_epochs: 3
 
 ### Training results
 
 | Training Loss | Epoch  | Step | Validation Loss | Accuracy |
 |:-------------:|:------:|:----:|:---------------:|:--------:|
-| 2.4389        | 0.9009 | 100  | 1.7820          | 0.6007   |
-| 0.7864        | 1.8018 | 200  | 0.6941          | 0.7625   |
-| 0.616         | 2.7027 | 300  | 0.5444          | 0.8104   |
-| 0.5426        | 3.6036 | 400  | 0.5069          | 0.8177   |
-| 0.4955        | 4.5045 | 500  | 0.4693          | 0.8292   |
-| 0.4483        | 5.4054 | 600  | 0.4387          | 0.8438   |
-| 0.4379        | 6.3063 | 700  | 0.4329          | 0.8495   |
-| 0.4073        | 7.2072 | 800  | 0.4256          | 0.8483   |
-| 0.3889        | 8.1081 | 900  | 0.4322          | 0.8393   |
-| 0.3765        | 9.0090 | 1000 | 0.4066          | 0.8548   |
-| 0.3101        | 9.9099 | 1100 | 0.4092          | 0.8545   |
+| 15.0942       | 0.8959 | 100  | 1.7638          | 0.5923   |
+| 9.8203        | 1.7962 | 200  | 1.1090          | 0.7259   |
+| 6.7293        | 2.6965 | 300  | 0.8104          | 0.8022   |
 
 
 ### Framework versions
 
-- Transformers 4.45.1
-- Pytorch 2.4.1+cu121
-- Datasets 3.0.1
-- Tokenizers 0.20.0
+- Transformers 4.46.1
+- Pytorch 2.5.1+cu124
+- Datasets 3.0.2
+- Tokenizers 0.20.1

all_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.8548114800225098,
-    "eval_loss": 0.4066202640533447,
-    "eval_runtime": 150.6221,
-    "eval_samples_per_second": 23.595,
-    "eval_steps_per_second": 0.744
+    "epoch": 2.992161254199328,
+    "eval_accuracy": 0.8105616093880972,
+    "eval_loss": 0.7486518621444702,
+    "eval_runtime": 132.5931,
+    "eval_samples_per_second": 26.992,
+    "eval_steps_per_second": 0.845
 }

eval_results.json

@@ -1,8 +1,8 @@
 {
-    "epoch": 10.0,
-    "eval_accuracy": 0.8548114800225098,
-    "eval_loss": 0.4066202640533447,
-    "eval_runtime": 150.6221,
-    "eval_samples_per_second": 23.595,
-    "eval_steps_per_second": 0.744
+    "epoch": 2.992161254199328,
+    "eval_accuracy": 0.8105616093880972,
+    "eval_loss": 0.7486518621444702,
+    "eval_runtime": 132.5931,
+    "eval_samples_per_second": 26.992,
+    "eval_steps_per_second": 0.845
 }