|
|
--- |
|
|
license: apache-2.0 |
|
|
base_model: facebook/wav2vec2-xls-r-300m |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
datasets: |
|
|
- common_voice_16_1 |
|
|
metrics: |
|
|
- wer |
|
|
model-index: |
|
|
- name: wav2vec2-large-xls-r-300m-tr-cv16.1-colab |
|
|
results: |
|
|
- task: |
|
|
name: Automatic Speech Recognition |
|
|
type: automatic-speech-recognition |
|
|
dataset: |
|
|
name: common_voice_16_1 |
|
|
type: common_voice_16_1 |
|
|
config: tr |
|
|
split: test |
|
|
args: tr |
|
|
metrics: |
|
|
- name: Wer |
|
|
type: wer |
|
|
value: 0.2775680437205623 |
|
|
--- |
|
|
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
|
|
# wav2vec2-large-xls-r-300m-tr-cv16.1-colab |
|
|
|
|
|
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_16_1 dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.2481 |
|
|
- Wer: 0.2776 |
|
|
|
|
|
## Model description |
|
|
|
|
|
More information needed |
|
|
|
|
|
## Intended uses & limitations |
|
|
|
|
|
More information needed |
|
|
|
|
|
## Training and evaluation data |
|
|
|
|
|
More information needed |
|
|
|
|
|
## Training procedure |
|
|
|
|
|
### Training hyperparameters |
|
|
|
|
|
The following hyperparameters were used during training: |
|
|
- learning_rate: 0.0003 |
|
|
- train_batch_size: 16 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 2 |
|
|
- total_train_batch_size: 32 |
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
|
- lr_scheduler_type: linear |
|
|
- lr_scheduler_warmup_steps: 500 |
|
|
- num_epochs: 10 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|
|:-------------:|:-----:|:-----:|:---------------:|:------:| |
|
|
| 5.5874 | 0.29 | 400 | 1.2182 | 0.9358 | |
|
|
| 0.8023 | 0.58 | 800 | 0.7425 | 0.7498 | |
|
|
| 0.5662 | 0.88 | 1200 | 0.5324 | 0.6233 | |
|
|
| 0.4553 | 1.17 | 1600 | 0.4375 | 0.5267 | |
|
|
| 0.4068 | 1.46 | 2000 | 0.4159 | 0.5051 | |
|
|
| 0.3797 | 1.75 | 2400 | 0.3861 | 0.4752 | |
|
|
| 0.3551 | 2.05 | 2800 | 0.3681 | 0.4484 | |
|
|
| 0.3059 | 2.34 | 3200 | 0.3491 | 0.4364 | |
|
|
| 0.297 | 2.63 | 3600 | 0.3437 | 0.4191 | |
|
|
| 0.292 | 2.92 | 4000 | 0.3261 | 0.4160 | |
|
|
| 0.2537 | 3.21 | 4400 | 0.3363 | 0.4105 | |
|
|
| 0.2448 | 3.51 | 4800 | 0.3527 | 0.4113 | |
|
|
| 0.2411 | 3.8 | 5200 | 0.3233 | 0.3975 | |
|
|
| 0.2324 | 4.09 | 5600 | 0.3038 | 0.3823 | |
|
|
| 0.213 | 4.38 | 6000 | 0.2982 | 0.3757 | |
|
|
| 0.2046 | 4.67 | 6400 | 0.2909 | 0.3591 | |
|
|
| 0.2064 | 4.97 | 6800 | 0.2914 | 0.3654 | |
|
|
| 0.1814 | 5.26 | 7200 | 0.2961 | 0.3567 | |
|
|
| 0.1774 | 5.55 | 7600 | 0.3105 | 0.3671 | |
|
|
| 0.1816 | 5.84 | 8000 | 0.2971 | 0.3524 | |
|
|
| 0.1621 | 6.14 | 8400 | 0.2837 | 0.3444 | |
|
|
| 0.1526 | 6.43 | 8800 | 0.2810 | 0.3371 | |
|
|
| 0.1492 | 6.72 | 9200 | 0.2696 | 0.3277 | |
|
|
| 0.1404 | 7.01 | 9600 | 0.2733 | 0.3200 | |
|
|
| 0.1276 | 7.3 | 10000 | 0.2672 | 0.3076 | |
|
|
| 0.1266 | 7.6 | 10400 | 0.2727 | 0.3126 | |
|
|
| 0.1259 | 7.89 | 10800 | 0.2516 | 0.3051 | |
|
|
| 0.1143 | 8.18 | 11200 | 0.2633 | 0.2963 | |
|
|
| 0.1098 | 8.47 | 11600 | 0.2592 | 0.2938 | |
|
|
| 0.1037 | 8.77 | 12000 | 0.2473 | 0.2914 | |
|
|
| 0.0995 | 9.06 | 12400 | 0.2566 | 0.2857 | |
|
|
| 0.0937 | 9.35 | 12800 | 0.2528 | 0.2812 | |
|
|
| 0.094 | 9.64 | 13200 | 0.2491 | 0.2799 | |
|
|
| 0.0927 | 9.93 | 13600 | 0.2481 | 0.2776 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.38.2 |
|
|
- Pytorch 2.2.1+cu121 |
|
|
- Datasets 2.18.0 |
|
|
- Tokenizers 0.15.2 |
|
|
|
