update model card README.md

README.md

@@ -16,8 +16,8 @@ should probably proofread and complete it, then remove this comment. -->
 
 This model is a fine-tuned version of [facebook/wav2vec2-base](https://huggingface.co/facebook/wav2vec2-base) on the None dataset.
 It achieves the following results on the evaluation set:
-- Loss: 0.8540
-- Wer: 1.0
+- Loss: 2.8796
+- Wer: 0.9833
 
 ## Model description
 
@@ -36,30 +36,31 @@ More information needed
 ### Training hyperparameters
 
 The following hyperparameters were used during training:
-- learning_rate: 0.0001
+- learning_rate: 1e-05
 - train_batch_size: 8
 - eval_batch_size: 8
 - seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 16
 - 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
+- lr_scheduler_warmup_steps: 500
+- training_steps: 4000
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Wer |
-|:-------------:|:-----:|:----:|:---------------:|:---:|
-| 2.6816        | 1.0   | 54   | 2.2369          | 1.0 |
-| 1.1823        | 2.0   | 108  | 1.0223          | 1.0 |
-| 1.0363        | 3.0   | 162  | 0.9690          | 1.0 |
-| 0.896         | 4.0   | 216  | 0.9106          | 1.0 |
-| 0.8138        | 5.0   | 270  | 0.8540          | 1.0 |
+| Training Loss | Epoch | Step | Validation Loss | Wer    |
+|:-------------:|:-----:|:----:|:---------------:|:------:|
+| 2.8346        | 200.0 | 1000 | 2.9858          | 1.0    |
+| 2.4286        | 400.0 | 2000 | 2.8819          | 0.9958 |
+| 2.133         | 600.0 | 3000 | 2.9096          | 0.9792 |
+| 2.0029        | 800.0 | 4000 | 2.8796          | 0.9833 |
 
 
 ### Framework versions
 
 - Transformers 4.25.0.dev0
 - Pytorch 1.12.1+cu113
-- Datasets 2.6.1
+- Datasets 2.7.0
 - Tokenizers 0.13.2