update model card README.md
Browse files
README.md
CHANGED
|
@@ -14,9 +14,9 @@ should probably proofread and complete it, then remove this comment. -->
|
|
| 14 |
|
| 15 |
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the None dataset.
|
| 16 |
It achieves the following results on the evaluation set:
|
| 17 |
-
- Loss: 0.
|
| 18 |
-
- Wer: 0.
|
| 19 |
-
- Cer: 0.
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
@@ -43,28 +43,33 @@ The following hyperparameters were used during training:
|
|
| 43 |
- total_train_batch_size: 96
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
-
- num_epochs:
|
| 47 |
- mixed_precision_training: Native AMP
|
| 48 |
|
| 49 |
### Training results
|
| 50 |
|
| 51 |
-
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 52 |
-
|
| 53 |
-
| 2.8998 | 1.0 | 508
|
| 54 |
-
| 0.7426 | 2.0 | 1016 | 0.6532 | 0.5236 |
|
| 55 |
-
| 0.5694 | 3.0 | 1524 | 0.5495 | 0.4611 |
|
| 56 |
-
| 0.4966 | 4.0 | 2032 | 0.5361 | 0.4350 |
|
| 57 |
-
| 0.4555 | 5.0 | 2540 | 0.5335 | 0.4266 |
|
| 58 |
-
| 0.4203 | 6.0 | 3048 | 0.5040 | 0.4107 |
|
| 59 |
-
| 0.3951 | 7.0 | 3556 | 0.5213 | 0.4037 |
|
| 60 |
-
| 0.3675 | 8.0 | 4064 | 0.4949 | 0.3973 |
|
| 61 |
-
| 0.3555 | 9.0 | 4572 | 0.4968 | 0.3978 |
|
| 62 |
-
| 0.3408 | 10.0 | 5080 | 0.4827 | 0.3925 |
|
| 63 |
-
| 0.3286 | 11.0 | 5588 | 0.5011 | 0.3858 |
|
| 64 |
-
| 0.3156 | 12.0 | 6096 | 0.4871 | 0.3833 |
|
| 65 |
-
| 0.3047 | 13.0 | 6604 | 0.4835 | 0.3774 |
|
| 66 |
-
| 0.2929 | 14.0 | 7112 | 0.4844 | 0.3751 |
|
| 67 |
-
| 0.2912 | 15.0 | 7620
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 68 |
|
| 69 |
|
| 70 |
### Framework versions
|
|
|
|
| 14 |
|
| 15 |
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the None dataset.
|
| 16 |
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 0.4939
|
| 18 |
+
- Wer: 0.3698
|
| 19 |
+
- Cer: 0.1465
|
| 20 |
|
| 21 |
## Model description
|
| 22 |
|
|
|
|
| 43 |
- total_train_batch_size: 96
|
| 44 |
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 45 |
- lr_scheduler_type: linear
|
| 46 |
+
- num_epochs: 20
|
| 47 |
- mixed_precision_training: Native AMP
|
| 48 |
|
| 49 |
### Training results
|
| 50 |
|
| 51 |
+
| Training Loss | Epoch | Step | Cer | Validation Loss | Wer |
|
| 52 |
+
|:-------------:|:-----:|:-----:|:------:|:---------------:|:------:|
|
| 53 |
+
| 2.8998 | 1.0 | 508 | 0.2832 | 1.0261 | 0.7004 |
|
| 54 |
+
| 0.7426 | 2.0 | 1016 | 0.2026 | 0.6532 | 0.5236 |
|
| 55 |
+
| 0.5694 | 3.0 | 1524 | 0.1799 | 0.5495 | 0.4611 |
|
| 56 |
+
| 0.4966 | 4.0 | 2032 | 0.1729 | 0.5361 | 0.4350 |
|
| 57 |
+
| 0.4555 | 5.0 | 2540 | 0.1684 | 0.5335 | 0.4266 |
|
| 58 |
+
| 0.4203 | 6.0 | 3048 | 0.1641 | 0.5040 | 0.4107 |
|
| 59 |
+
| 0.3951 | 7.0 | 3556 | 0.1579 | 0.5213 | 0.4037 |
|
| 60 |
+
| 0.3675 | 8.0 | 4064 | 0.1563 | 0.4949 | 0.3973 |
|
| 61 |
+
| 0.3555 | 9.0 | 4572 | 0.1581 | 0.4968 | 0.3978 |
|
| 62 |
+
| 0.3408 | 10.0 | 5080 | 0.1561 | 0.4827 | 0.3925 |
|
| 63 |
+
| 0.3286 | 11.0 | 5588 | 0.1524 | 0.5011 | 0.3858 |
|
| 64 |
+
| 0.3156 | 12.0 | 6096 | 0.1524 | 0.4871 | 0.3833 |
|
| 65 |
+
| 0.3047 | 13.0 | 6604 | 0.1499 | 0.4835 | 0.3774 |
|
| 66 |
+
| 0.2929 | 14.0 | 7112 | 0.1489 | 0.4844 | 0.3751 |
|
| 67 |
+
| 0.2912 | 15.0 | 7620 | 0.4929 | 0.3763 | 0.1486 |
|
| 68 |
+
| 0.2969 | 16.0 | 8128 | 0.4990 | 0.3749 | 0.1481 |
|
| 69 |
+
| 0.2946 | 17.0 | 8636 | 0.4943 | 0.3735 | 0.1485 |
|
| 70 |
+
| 0.2851 | 18.0 | 9144 | 0.4893 | 0.3717 | 0.1477 |
|
| 71 |
+
| 0.279 | 19.0 | 9652 | 0.4977 | 0.3693 | 0.1464 |
|
| 72 |
+
| 0.2718 | 20.0 | 10160 | 0.4939 | 0.3698 | 0.1465 |
|
| 73 |
|
| 74 |
|
| 75 |
### Framework versions
|