End of training

README.md

@@ -21,8 +21,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the Junhoee/STT_Korean_Dataset_80000 dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.3536
-- Cer: 32.7683
+- Loss: 0.4023
+- Cer: 12.9583
 
 ## Model description
 
@@ -41,10 +41,12 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 1e-05
-- train_batch_size: 16
+- learning_rate: 0.0002
+- train_batch_size: 8
 - eval_batch_size: 8
 - seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 16
 - 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: linear
 - lr_scheduler_warmup_steps: 500
@@ -55,10 +57,10 @@ The following hyperparameters were used during training:
 
 | Training Loss | Epoch | Step | Validation Loss | Cer     |
 |:-------------:|:-----:|:----:|:---------------:|:-------:|
-| 0.4502        | 0.25  | 1000 | 0.4243          | 74.9195 |
-| 0.3982        | 0.5   | 2000 | 0.3852          | 65.8768 |
-| 0.374         | 0.75  | 3000 | 0.3635          | 40.6152 |
-| 0.3294        | 1.0   | 4000 | 0.3536          | 32.7683 |
+| 0.9422        | 0.25  | 1000 | 0.9523          | 42.9308 |
+| 0.669         | 0.5   | 2000 | 0.6758          | 20.5051 |
+| 0.5142        | 0.75  | 3000 | 0.4989          | 17.0488 |
+| 0.3686        | 1.0   | 4000 | 0.4023          | 12.9583 |
 
 
 ### Framework versions