update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,88 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- common_voice
|
| 7 |
+
model-index:
|
| 8 |
+
- name: wav2vec2-large-xls-r-300m-pt-colab
|
| 9 |
+
results: []
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 13 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 14 |
+
|
| 15 |
+
# wav2vec2-large-xls-r-300m-pt-colab
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-300m](https://huggingface.co/facebook/wav2vec2-xls-r-300m) on the common_voice dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 0.3637
|
| 20 |
+
- Wer: 0.2982
|
| 21 |
+
|
| 22 |
+
## Model description
|
| 23 |
+
|
| 24 |
+
More information needed
|
| 25 |
+
|
| 26 |
+
## Intended uses & limitations
|
| 27 |
+
|
| 28 |
+
More information needed
|
| 29 |
+
|
| 30 |
+
## Training and evaluation data
|
| 31 |
+
|
| 32 |
+
More information needed
|
| 33 |
+
|
| 34 |
+
## Training procedure
|
| 35 |
+
|
| 36 |
+
### Training hyperparameters
|
| 37 |
+
|
| 38 |
+
The following hyperparameters were used during training:
|
| 39 |
+
- learning_rate: 0.0003
|
| 40 |
+
- train_batch_size: 16
|
| 41 |
+
- eval_batch_size: 8
|
| 42 |
+
- seed: 42
|
| 43 |
+
- gradient_accumulation_steps: 2
|
| 44 |
+
- total_train_batch_size: 32
|
| 45 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 46 |
+
- lr_scheduler_type: linear
|
| 47 |
+
- lr_scheduler_warmup_steps: 500
|
| 48 |
+
- num_epochs: 30
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
| 55 |
+
| 4.591 | 1.15 | 400 | 0.9128 | 0.6517 |
|
| 56 |
+
| 0.5049 | 2.31 | 800 | 0.4596 | 0.4437 |
|
| 57 |
+
| 0.2871 | 3.46 | 1200 | 0.3964 | 0.3905 |
|
| 58 |
+
| 0.2077 | 4.61 | 1600 | 0.3958 | 0.3744 |
|
| 59 |
+
| 0.1695 | 5.76 | 2000 | 0.4040 | 0.3720 |
|
| 60 |
+
| 0.1478 | 6.92 | 2400 | 0.3866 | 0.3651 |
|
| 61 |
+
| 0.1282 | 8.07 | 2800 | 0.3987 | 0.3674 |
|
| 62 |
+
| 0.1134 | 9.22 | 3200 | 0.4128 | 0.3688 |
|
| 63 |
+
| 0.1048 | 10.37 | 3600 | 0.3928 | 0.3561 |
|
| 64 |
+
| 0.0938 | 11.53 | 4000 | 0.4048 | 0.3619 |
|
| 65 |
+
| 0.0848 | 12.68 | 4400 | 0.4229 | 0.3555 |
|
| 66 |
+
| 0.0798 | 13.83 | 4800 | 0.3974 | 0.3468 |
|
| 67 |
+
| 0.0688 | 14.98 | 5200 | 0.3870 | 0.3503 |
|
| 68 |
+
| 0.0658 | 16.14 | 5600 | 0.3875 | 0.3351 |
|
| 69 |
+
| 0.061 | 17.29 | 6000 | 0.4133 | 0.3417 |
|
| 70 |
+
| 0.0569 | 18.44 | 6400 | 0.3915 | 0.3414 |
|
| 71 |
+
| 0.0526 | 19.6 | 6800 | 0.3957 | 0.3231 |
|
| 72 |
+
| 0.0468 | 20.75 | 7200 | 0.4110 | 0.3301 |
|
| 73 |
+
| 0.0407 | 21.9 | 7600 | 0.3866 | 0.3186 |
|
| 74 |
+
| 0.0384 | 23.05 | 8000 | 0.3976 | 0.3193 |
|
| 75 |
+
| 0.0363 | 24.21 | 8400 | 0.3910 | 0.3177 |
|
| 76 |
+
| 0.0313 | 25.36 | 8800 | 0.3656 | 0.3109 |
|
| 77 |
+
| 0.0293 | 26.51 | 9200 | 0.3712 | 0.3092 |
|
| 78 |
+
| 0.0277 | 27.66 | 9600 | 0.3613 | 0.3054 |
|
| 79 |
+
| 0.0249 | 28.82 | 10000 | 0.3783 | 0.3015 |
|
| 80 |
+
| 0.0234 | 29.97 | 10400 | 0.3637 | 0.2982 |
|
| 81 |
+
|
| 82 |
+
|
| 83 |
+
### Framework versions
|
| 84 |
+
|
| 85 |
+
- Transformers 4.11.3
|
| 86 |
+
- Pytorch 1.10.0+cu102
|
| 87 |
+
- Datasets 1.13.3
|
| 88 |
+
- Tokenizers 0.10.3
|