---
base_model: facebook/mms-1b-all
datasets:
- common_voice_17_0
library_name: transformers
license: cc-by-nc-4.0
metrics:
- wer
- bleu
tags:
- generated_from_trainer
model-index:
- name: wav2vec2-mms-1b-CV17.0-training_set_variations
  results:
  - task:
      type: automatic-speech-recognition
      name: Automatic Speech Recognition
    dataset:
      name: common_voice_17_0
      type: common_voice_17_0
      config: ta
      split: validation
      args: ta
    metrics:
    - type: wer
      value: 0.3597180870859695
      name: Wer
    - type: bleu
      value: 0.4226157099926465
      name: Bleu
---

<!-- 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-mms-1b-CV17.0-training_set_variations

This model is a fine-tuned version of [facebook/mms-1b-all](https://huggingface.co/facebook/mms-1b-all) on the common_voice_17_0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.2047
- Wer: 0.3597
- Cer: 0.0579
- Bleu: 0.4226

## 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: 0.001
- 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_ratio: 0.15
- training_steps: 2000
- mixed_precision_training: Native AMP

### Training results

| Training Loss | Epoch  | Step | Validation Loss | Wer    | Cer    | Bleu   |
|:-------------:|:------:|:----:|:---------------:|:------:|:------:|:------:|
| 6.3615        | 0.3906 | 100  | 0.2954          | 0.4162 | 0.0682 | 0.3508 |
| 0.2115        | 0.7812 | 200  | 0.2266          | 0.3822 | 0.0619 | 0.3888 |
| 0.1868        | 1.1719 | 300  | 0.2227          | 0.3755 | 0.0608 | 0.3981 |
| 0.1913        | 1.5625 | 400  | 0.2274          | 0.3912 | 0.0637 | 0.3779 |
| 0.1896        | 1.9531 | 500  | 0.2263          | 0.3858 | 0.0631 | 0.3867 |
| 0.1769        | 2.3438 | 600  | 0.2176          | 0.3785 | 0.0618 | 0.3942 |
| 0.1752        | 2.7344 | 700  | 0.2162          | 0.3816 | 0.0614 | 0.3887 |
| 0.1777        | 3.125  | 800  | 0.2098          | 0.3606 | 0.0582 | 0.4260 |
| 0.1747        | 3.5156 | 900  | 0.2078          | 0.3657 | 0.0585 | 0.4111 |
| 0.1672        | 3.9062 | 1000 | 0.2075          | 0.3770 | 0.0595 | 0.3920 |
| 0.1583        | 4.2969 | 1100 | 0.2060          | 0.3631 | 0.0580 | 0.4137 |
| 0.1713        | 4.6875 | 1200 | 0.2064          | 0.3664 | 0.0587 | 0.4118 |
| 0.1563        | 5.0781 | 1300 | 0.2047          | 0.3597 | 0.0579 | 0.4226 |


### Framework versions

- Transformers 4.44.2
- Pytorch 2.4.1+cu121
- Datasets 3.0.0
- Tokenizers 0.19.1