metadata
license: cc-by-nc-4.0
base_model: facebook/mms-1b-all
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: shqip-mms-3
results: []
shqip-mms-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: 0.4195
- Wer: 0.3447
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: 32
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 100
- num_epochs: 2
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 1.5717 | 0.31 | 300 | 0.9548 | 0.7373 |
| 0.7823 | 0.63 | 600 | 0.7372 | 0.5886 |
| 0.6863 | 0.94 | 900 | 0.6122 | 0.5261 |
| 0.5908 | 1.25 | 1200 | 0.5215 | 0.4440 |
| 0.5237 | 1.57 | 1500 | 0.4693 | 0.3987 |
| 0.4662 | 1.88 | 1800 | 0.4195 | 0.3447 |
Framework versions
- Transformers 4.32.0.dev0
- Pytorch 2.1.0.dev20230810
- Datasets 2.14.3
- Tokenizers 0.13.3