metadata
language:
- bn
license: apache-2.0
base_model: bigscience/mt0-small
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Tahmid37/Tahmid37/mt0-small-text-to-ipa-v-final-1
results: []
Tahmid37/Tahmid37/mt0-small-text-to-ipa-v-final-1
This model is a fine-tuned version of bigscience/mt0-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0536
- Wer: 0.0358
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: 8
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 2000
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.1705 | 0.17 | 4000 | 0.0798 | 0.0763 |
| 0.1399 | 0.34 | 8000 | 0.0799 | 0.0728 |
| 0.1359 | 0.51 | 12000 | 0.0818 | 0.0778 |
| 0.1323 | 0.67 | 16000 | 0.0818 | 0.0750 |
| 0.1248 | 0.84 | 20000 | 0.0777 | 0.0683 |
| 0.1217 | 1.01 | 24000 | 0.0768 | 0.0675 |
| 0.0984 | 1.18 | 28000 | 0.0767 | 0.0691 |
| 0.0937 | 1.35 | 32000 | 0.0734 | 0.0617 |
| 0.0951 | 1.52 | 36000 | 0.0720 | 0.0590 |
| 0.0884 | 1.69 | 40000 | 0.0676 | 0.0513 |
| 0.0854 | 1.86 | 44000 | 0.0639 | 0.0491 |
| 0.0836 | 2.02 | 48000 | 0.0629 | 0.0453 |
| 0.0625 | 2.19 | 52000 | 0.0611 | 0.0443 |
| 0.0606 | 2.36 | 56000 | 0.0598 | 0.0418 |
| 0.0568 | 2.53 | 60000 | 0.0555 | 0.0383 |
| 0.0521 | 2.7 | 64000 | 0.0553 | 0.0375 |
| 0.0537 | 2.87 | 68000 | 0.0536 | 0.0358 |
Framework versions
- Transformers 4.34.1
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.14.1