Commit
•
5d8500a
1
Parent(s):
fce7839
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-sr-v4
|
| 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-sr-v4
|
| 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: 0.5570
|
| 20 |
+
- Wer: 0.3038
|
| 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: 800
|
| 48 |
+
- num_epochs: 200
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 55 |
+
| 8.2934 | 7.5 | 300 | 2.9777 | 0.9995 |
|
| 56 |
+
| 1.5049 | 15.0 | 600 | 0.5036 | 0.4806 |
|
| 57 |
+
| 0.3263 | 22.5 | 900 | 0.5822 | 0.4055 |
|
| 58 |
+
| 0.2008 | 30.0 | 1200 | 0.5609 | 0.4032 |
|
| 59 |
+
| 0.1543 | 37.5 | 1500 | 0.5203 | 0.3710 |
|
| 60 |
+
| 0.1158 | 45.0 | 1800 | 0.6458 | 0.3985 |
|
| 61 |
+
| 0.0997 | 52.5 | 2100 | 0.6227 | 0.4013 |
|
| 62 |
+
| 0.0834 | 60.0 | 2400 | 0.6048 | 0.3836 |
|
| 63 |
+
| 0.0665 | 67.5 | 2700 | 0.6197 | 0.3686 |
|
| 64 |
+
| 0.0602 | 75.0 | 3000 | 0.5418 | 0.3453 |
|
| 65 |
+
| 0.0524 | 82.5 | 3300 | 0.5310 | 0.3486 |
|
| 66 |
+
| 0.0445 | 90.0 | 3600 | 0.5599 | 0.3374 |
|
| 67 |
+
| 0.0406 | 97.5 | 3900 | 0.5958 | 0.3327 |
|
| 68 |
+
| 0.0358 | 105.0 | 4200 | 0.6017 | 0.3262 |
|
| 69 |
+
| 0.0302 | 112.5 | 4500 | 0.5613 | 0.3248 |
|
| 70 |
+
| 0.0285 | 120.0 | 4800 | 0.5659 | 0.3462 |
|
| 71 |
+
| 0.0213 | 127.5 | 5100 | 0.5568 | 0.3206 |
|
| 72 |
+
| 0.0215 | 135.0 | 5400 | 0.6524 | 0.3472 |
|
| 73 |
+
| 0.0162 | 142.5 | 5700 | 0.6223 | 0.3458 |
|
| 74 |
+
| 0.0137 | 150.0 | 6000 | 0.6625 | 0.3313 |
|
| 75 |
+
| 0.0114 | 157.5 | 6300 | 0.5739 | 0.3336 |
|
| 76 |
+
| 0.0101 | 165.0 | 6600 | 0.5906 | 0.3285 |
|
| 77 |
+
| 0.008 | 172.5 | 6900 | 0.5982 | 0.3112 |
|
| 78 |
+
| 0.0076 | 180.0 | 7200 | 0.5399 | 0.3094 |
|
| 79 |
+
| 0.0071 | 187.5 | 7500 | 0.5387 | 0.2991 |
|
| 80 |
+
| 0.0057 | 195.0 | 7800 | 0.5570 | 0.3038 |
|
| 81 |
+
|
| 82 |
+
|
| 83 |
+
### Framework versions
|
| 84 |
+
|
| 85 |
+
- Transformers 4.16.2
|
| 86 |
+
- Pytorch 1.10.0+cu111
|
| 87 |
+
- Datasets 1.18.2
|
| 88 |
+
- Tokenizers 0.11.0
|