Commit
·
3606b39
1
Parent(s):
c838104
update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
datasets:
|
| 6 |
+
- common_voice
|
| 7 |
+
model-index:
|
| 8 |
+
- name: ''
|
| 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 |
+
#
|
| 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: 0.1666
|
| 20 |
+
- Wer: 0.1718
|
| 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: 7.5e-05
|
| 40 |
+
- train_batch_size: 32
|
| 41 |
+
- eval_batch_size: 16
|
| 42 |
+
- seed: 42
|
| 43 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 44 |
+
- lr_scheduler_type: linear
|
| 45 |
+
- lr_scheduler_warmup_steps: 1000
|
| 46 |
+
- num_epochs: 50.0
|
| 47 |
+
- mixed_precision_training: Native AMP
|
| 48 |
+
|
| 49 |
+
### Training results
|
| 50 |
+
|
| 51 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 52 |
+
|:-------------:|:-----:|:----:|:---------------:|:------:|
|
| 53 |
+
| 3.489 | 2.56 | 400 | 3.3590 | 1.0 |
|
| 54 |
+
| 2.9903 | 5.13 | 800 | 2.9704 | 1.0001 |
|
| 55 |
+
| 1.6712 | 7.69 | 1200 | 0.6179 | 0.6566 |
|
| 56 |
+
| 1.2635 | 10.26 | 1600 | 0.3176 | 0.4531 |
|
| 57 |
+
| 1.0819 | 12.82 | 2000 | 0.2517 | 0.3508 |
|
| 58 |
+
| 1.0136 | 15.38 | 2400 | 0.2257 | 0.3124 |
|
| 59 |
+
| 0.9625 | 17.95 | 2800 | 0.1975 | 0.2311 |
|
| 60 |
+
| 0.901 | 20.51 | 3200 | 0.1986 | 0.2097 |
|
| 61 |
+
| 0.8842 | 23.08 | 3600 | 0.1904 | 0.2039 |
|
| 62 |
+
| 0.8542 | 25.64 | 4000 | 0.1847 | 0.1981 |
|
| 63 |
+
| 0.8244 | 28.21 | 4400 | 0.1805 | 0.1847 |
|
| 64 |
+
| 0.7689 | 30.77 | 4800 | 0.1736 | 0.1832 |
|
| 65 |
+
| 0.7825 | 33.33 | 5200 | 0.1698 | 0.1821 |
|
| 66 |
+
| 0.7817 | 35.9 | 5600 | 0.1758 | 0.1803 |
|
| 67 |
+
| 0.7488 | 38.46 | 6000 | 0.1663 | 0.1760 |
|
| 68 |
+
| 0.7171 | 41.03 | 6400 | 0.1636 | 0.1721 |
|
| 69 |
+
| 0.7222 | 43.59 | 6800 | 0.1663 | 0.1729 |
|
| 70 |
+
| 0.7156 | 46.15 | 7200 | 0.1633 | 0.1715 |
|
| 71 |
+
| 0.7121 | 48.72 | 7600 | 0.1666 | 0.1718 |
|
| 72 |
+
|
| 73 |
+
|
| 74 |
+
### Framework versions
|
| 75 |
+
|
| 76 |
+
- Transformers 4.17.0.dev0
|
| 77 |
+
- Pytorch 1.10.2+cu102
|
| 78 |
+
- Datasets 1.18.2.dev0
|
| 79 |
+
- Tokenizers 0.11.0
|