update model card README.md

README.md

@@ -14,10 +14,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # wav2vec2-commonvoice-hindi
 
-This model is a fine-tuned version of [Harveenchadha/vakyansh_hindi_base_pretrained](https://huggingface.co/Harveenchadha/vakyansh_hindi_base_pretrained) on the common_voice dataset.
+This model is a fine-tuned version of [theainerd/Wav2Vec2-large-xlsr-hindi](https://huggingface.co/theainerd/Wav2Vec2-large-xlsr-hindi) on the common_voice dataset.
 It achieves the following results on the evaluation set:
-- Loss: 3.5304
-- Wer: 1.0
+- Loss: 0.9825
+- Wer: 0.6763
 
 ## Model description
 
@@ -40,20 +40,19 @@ The following hyperparameters were used during training:
 - train_batch_size: 16
 - eval_batch_size: 8
 - seed: 42
+- gradient_accumulation_steps: 2
+- total_train_batch_size: 32
 - optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
 - lr_scheduler_type: linear
 - lr_scheduler_warmup_steps: 500
-- num_epochs: 50
+- num_epochs: 30
 - mixed_precision_training: Native AMP
 
 ### Training results
 
-| Training Loss | Epoch | Step | Validation Loss | Wer |
-|:-------------:|:-----:|:----:|:---------------:|:---:|
-| 6.8763        | 10.53 | 200  | 3.5392          | 1.0 |
-| 3.4778        | 21.05 | 400  | 3.5174          | 1.0 |
-| 3.5119        | 31.58 | 600  | 3.5362          | 1.0 |
-| 3.5133        | 42.11 | 800  | 3.5371          | 1.0 |
+| Training Loss | Epoch | Step | Validation Loss | Wer    |
+|:-------------:|:-----:|:----:|:---------------:|:------:|
+| No log        | 20.0  | 100  | 0.8801          | 0.6754 |
 
 
 ### Framework versions