|
|
--- |
|
|
library_name: transformers |
|
|
license: apache-2.0 |
|
|
base_model: parambharat/whisper-small-ta |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
model-index: |
|
|
- name: MTF-ta-en-translation |
|
|
results: [] |
|
|
--- |
|
|
|
|
|
<!-- 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. --> |
|
|
|
|
|
# MTF-ta-en-translation |
|
|
|
|
|
This model is a fine-tuned version of [parambharat/whisper-small-ta](https://huggingface.co/parambharat/whisper-small-ta) on an unknown dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.1324 |
|
|
- Bleu Score: 0.0299 |
|
|
|
|
|
## 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: 1e-05 |
|
|
- train_batch_size: 8 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 2 |
|
|
- total_train_batch_size: 16 |
|
|
- 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 |
|
|
- training_steps: 5000 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Bleu Score | |
|
|
|:-------------:|:-------:|:----:|:---------------:|:----------:| |
|
|
| 0.0939 | 2.9412 | 250 | 0.0864 | 0.0302 | |
|
|
| 0.0227 | 5.8824 | 500 | 0.0998 | 0.0301 | |
|
|
| 0.0048 | 8.8235 | 750 | 0.1083 | 0.0340 | |
|
|
| 0.001 | 11.7647 | 1000 | 0.1132 | 0.0312 | |
|
|
| 0.0005 | 14.7059 | 1250 | 0.1164 | 0.0308 | |
|
|
| 0.0003 | 17.6471 | 1500 | 0.1189 | 0.0322 | |
|
|
| 0.0002 | 20.5882 | 1750 | 0.1208 | 0.0311 | |
|
|
| 0.0002 | 23.5294 | 2000 | 0.1225 | 0.0307 | |
|
|
| 0.0002 | 26.4706 | 2250 | 0.1242 | 0.0334 | |
|
|
| 0.0001 | 29.4118 | 2500 | 0.1256 | 0.0321 | |
|
|
| 0.0001 | 32.3529 | 2750 | 0.1268 | 0.0327 | |
|
|
| 0.0001 | 35.2941 | 3000 | 0.1277 | 0.0324 | |
|
|
| 0.0001 | 38.2353 | 3250 | 0.1286 | 0.0311 | |
|
|
| 0.0001 | 41.1765 | 3500 | 0.1295 | 0.0309 | |
|
|
| 0.0001 | 44.1176 | 3750 | 0.1302 | 0.0311 | |
|
|
| 0.0001 | 47.0588 | 4000 | 0.1308 | 0.0310 | |
|
|
| 0.0001 | 50.0 | 4250 | 0.1314 | 0.0313 | |
|
|
| 0.0001 | 52.9412 | 4500 | 0.1320 | 0.0298 | |
|
|
| 0.0001 | 55.8824 | 4750 | 0.1323 | 0.0299 | |
|
|
| 0.0 | 58.8235 | 5000 | 0.1324 | 0.0299 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.48.3 |
|
|
- Pytorch 2.6.0+cu124 |
|
|
- Datasets 3.2.0 |
|
|
- Tokenizers 0.21.0 |
|
|
|
