update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,89 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- automatic-speech-recognition
|
| 5 |
+
- geninhu/fpt-vi
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
model-index:
|
| 8 |
+
- name: xls-asr-vi-40h-1B
|
| 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 |
+
# xls-asr-vi-40h-1B
|
| 16 |
+
|
| 17 |
+
This model is a fine-tuned version of [facebook/wav2vec2-xls-r-1b](https://huggingface.co/facebook/wav2vec2-xls-r-1b) on the GENINHU/FPT-VI - NA dataset.
|
| 18 |
+
It achieves the following results on the evaluation set:
|
| 19 |
+
- Loss: 4.1691
|
| 20 |
+
- Wer: 0.4133
|
| 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: 5e-05
|
| 40 |
+
- train_batch_size: 16
|
| 41 |
+
- eval_batch_size: 16
|
| 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: 1500
|
| 48 |
+
- num_epochs: 50.0
|
| 49 |
+
- mixed_precision_training: Native AMP
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 54 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
| 55 |
+
| 4.6222 | 1.85 | 1500 | 5.9479 | 0.5474 |
|
| 56 |
+
| 1.1362 | 3.7 | 3000 | 7.9799 | 0.5094 |
|
| 57 |
+
| 0.7814 | 5.56 | 4500 | 5.0330 | 0.4724 |
|
| 58 |
+
| 0.6281 | 7.41 | 6000 | 2.3484 | 0.5020 |
|
| 59 |
+
| 0.5472 | 9.26 | 7500 | 2.2495 | 0.4793 |
|
| 60 |
+
| 0.4827 | 11.11 | 9000 | 1.1530 | 0.4768 |
|
| 61 |
+
| 0.4327 | 12.96 | 10500 | 1.6160 | 0.4646 |
|
| 62 |
+
| 0.3989 | 14.81 | 12000 | 3.2633 | 0.4703 |
|
| 63 |
+
| 0.3522 | 16.67 | 13500 | 2.2337 | 0.4708 |
|
| 64 |
+
| 0.3201 | 18.52 | 15000 | 3.6879 | 0.4565 |
|
| 65 |
+
| 0.2899 | 20.37 | 16500 | 5.4389 | 0.4599 |
|
| 66 |
+
| 0.2776 | 22.22 | 18000 | 3.5284 | 0.4537 |
|
| 67 |
+
| 0.2574 | 24.07 | 19500 | 2.1759 | 0.4649 |
|
| 68 |
+
| 0.2378 | 25.93 | 21000 | 3.3901 | 0.4448 |
|
| 69 |
+
| 0.217 | 27.78 | 22500 | 1.1632 | 0.4565 |
|
| 70 |
+
| 0.2115 | 29.63 | 24000 | 1.7441 | 0.4232 |
|
| 71 |
+
| 0.1959 | 31.48 | 25500 | 3.4992 | 0.4304 |
|
| 72 |
+
| 0.187 | 33.33 | 27000 | 3.6163 | 0.4369 |
|
| 73 |
+
| 0.1748 | 35.19 | 28500 | 3.6038 | 0.4467 |
|
| 74 |
+
| 0.17 | 37.04 | 30000 | 2.9708 | 0.4362 |
|
| 75 |
+
| 0.159 | 38.89 | 31500 | 3.2045 | 0.4279 |
|
| 76 |
+
| 0.153 | 40.74 | 33000 | 3.2427 | 0.4287 |
|
| 77 |
+
| 0.1463 | 42.59 | 34500 | 3.5439 | 0.4270 |
|
| 78 |
+
| 0.139 | 44.44 | 36000 | 3.9381 | 0.4150 |
|
| 79 |
+
| 0.1352 | 46.3 | 37500 | 4.1744 | 0.4092 |
|
| 80 |
+
| 0.1369 | 48.15 | 39000 | 4.2279 | 0.4154 |
|
| 81 |
+
| 0.1273 | 50.0 | 40500 | 4.1691 | 0.4133 |
|
| 82 |
+
|
| 83 |
+
|
| 84 |
+
### Framework versions
|
| 85 |
+
|
| 86 |
+
- Transformers 4.16.0.dev0
|
| 87 |
+
- Pytorch 1.10.1+cu102
|
| 88 |
+
- Datasets 1.17.1.dev0
|
| 89 |
+
- Tokenizers 0.11.0
|