metadata
library_name: transformers
license: cc-by-nc-4.0
base_model: facebook/mms-1b-all
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: wav2vec2-large-mms-1b-dzo-colab-3
results: []
wav2vec2-large-mms-1b-dzo-colab-3
This model is a fine-tuned version of facebook/mms-1b-all on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.5800
- Wer: 0.4672
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: 0.005
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- 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 |
|---|---|---|---|---|
| 8.473 | 1.0 | 20 | 6.5541 | 1.0505 |
| 3.4553 | 2.0 | 40 | 3.1408 | 1.0 |
| 2.6261 | 3.0 | 60 | 2.0645 | 0.9461 |
| 1.1757 | 4.0 | 80 | 0.8704 | 0.6540 |
| 0.8657 | 5.0 | 100 | 0.6830 | 0.5446 |
| 0.7318 | 6.0 | 120 | 0.6215 | 0.5008 |
| 0.6546 | 7.0 | 140 | 0.5725 | 0.4705 |
| 0.5496 | 8.0 | 160 | 0.5756 | 0.4672 |
| 0.5476 | 9.0 | 180 | 0.5708 | 0.4705 |
| 0.5114 | 10.0 | 200 | 0.5800 | 0.4672 |
Framework versions
- Transformers 4.49.0.dev0
- Pytorch 2.5.1+cu121
- Datasets 3.2.0
- Tokenizers 0.21.0