update model card README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,87 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: cc-by-nc-4.0
|
| 3 |
+
tags:
|
| 4 |
+
- generated_from_trainer
|
| 5 |
+
model-index:
|
| 6 |
+
- name: v7-fine-tune-wav2vec2-Vietnamese-ARS-demo
|
| 7 |
+
results: []
|
| 8 |
+
---
|
| 9 |
+
|
| 10 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 11 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 12 |
+
|
| 13 |
+
# v7-fine-tune-wav2vec2-Vietnamese-ARS-demo
|
| 14 |
+
|
| 15 |
+
This model is a fine-tuned version of [nguyenvulebinh/wav2vec2-base-vietnamese-250h](https://huggingface.co/nguyenvulebinh/wav2vec2-base-vietnamese-250h) on the None dataset.
|
| 16 |
+
It achieves the following results on the evaluation set:
|
| 17 |
+
- Loss: 0.2806
|
| 18 |
+
- Wer: 0.2751
|
| 19 |
+
|
| 20 |
+
## Model description
|
| 21 |
+
|
| 22 |
+
More information needed
|
| 23 |
+
|
| 24 |
+
## Intended uses & limitations
|
| 25 |
+
|
| 26 |
+
More information needed
|
| 27 |
+
|
| 28 |
+
## Training and evaluation data
|
| 29 |
+
|
| 30 |
+
More information needed
|
| 31 |
+
|
| 32 |
+
## Training procedure
|
| 33 |
+
|
| 34 |
+
### Training hyperparameters
|
| 35 |
+
|
| 36 |
+
The following hyperparameters were used during training:
|
| 37 |
+
- learning_rate: 0.0001
|
| 38 |
+
- train_batch_size: 8
|
| 39 |
+
- eval_batch_size: 8
|
| 40 |
+
- seed: 42
|
| 41 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
| 42 |
+
- lr_scheduler_type: linear
|
| 43 |
+
- lr_scheduler_warmup_steps: 1000
|
| 44 |
+
- num_epochs: 10
|
| 45 |
+
- mixed_precision_training: Native AMP
|
| 46 |
+
|
| 47 |
+
### Training results
|
| 48 |
+
|
| 49 |
+
| Training Loss | Epoch | Step | Validation Loss | Wer |
|
| 50 |
+
|:-------------:|:-----:|:-----:|:---------------:|:------:|
|
| 51 |
+
| 10.9724 | 0.34 | 500 | 3.8863 | 1.0 |
|
| 52 |
+
| 3.4322 | 0.69 | 1000 | 2.6994 | 1.0953 |
|
| 53 |
+
| 2.431 | 1.03 | 1500 | 1.6603 | 1.0447 |
|
| 54 |
+
| 1.6436 | 1.37 | 2000 | 0.8842 | 0.7589 |
|
| 55 |
+
| 1.1161 | 1.71 | 2500 | 0.6005 | 0.5565 |
|
| 56 |
+
| 0.8657 | 2.06 | 3000 | 0.4911 | 0.4680 |
|
| 57 |
+
| 0.7653 | 2.4 | 3500 | 0.4501 | 0.4207 |
|
| 58 |
+
| 0.7013 | 2.74 | 4000 | 0.4149 | 0.3994 |
|
| 59 |
+
| 0.6306 | 3.09 | 4500 | 0.4065 | 0.3761 |
|
| 60 |
+
| 0.602 | 3.43 | 5000 | 0.3846 | 0.3682 |
|
| 61 |
+
| 0.5746 | 3.77 | 5500 | 0.3814 | 0.3486 |
|
| 62 |
+
| 0.5199 | 4.12 | 6000 | 0.3753 | 0.3354 |
|
| 63 |
+
| 0.4901 | 4.46 | 6500 | 0.3437 | 0.3228 |
|
| 64 |
+
| 0.4834 | 4.8 | 7000 | 0.3329 | 0.3175 |
|
| 65 |
+
| 0.4553 | 5.14 | 7500 | 0.3257 | 0.3135 |
|
| 66 |
+
| 0.4501 | 5.49 | 8000 | 0.3156 | 0.3077 |
|
| 67 |
+
| 0.4468 | 5.83 | 8500 | 0.3248 | 0.3072 |
|
| 68 |
+
| 0.4078 | 6.17 | 9000 | 0.3174 | 0.3033 |
|
| 69 |
+
| 0.4215 | 6.52 | 9500 | 0.3064 | 0.2959 |
|
| 70 |
+
| 0.4001 | 6.86 | 10000 | 0.3158 | 0.2936 |
|
| 71 |
+
| 0.3983 | 7.2 | 10500 | 0.3022 | 0.2890 |
|
| 72 |
+
| 0.3955 | 7.54 | 11000 | 0.2938 | 0.2890 |
|
| 73 |
+
| 0.381 | 7.89 | 11500 | 0.2893 | 0.2915 |
|
| 74 |
+
| 0.3484 | 8.23 | 12000 | 0.2881 | 0.2835 |
|
| 75 |
+
| 0.3492 | 8.57 | 12500 | 0.2899 | 0.2809 |
|
| 76 |
+
| 0.3825 | 8.92 | 13000 | 0.2857 | 0.2808 |
|
| 77 |
+
| 0.3356 | 9.26 | 13500 | 0.2841 | 0.2789 |
|
| 78 |
+
| 0.3361 | 9.6 | 14000 | 0.2789 | 0.2753 |
|
| 79 |
+
| 0.3353 | 9.95 | 14500 | 0.2806 | 0.2751 |
|
| 80 |
+
|
| 81 |
+
|
| 82 |
+
### Framework versions
|
| 83 |
+
|
| 84 |
+
- Transformers 4.17.0
|
| 85 |
+
- Pytorch 1.12.1+cu113
|
| 86 |
+
- Datasets 1.18.3
|
| 87 |
+
- Tokenizers 0.12.1
|