|
|
--- |
|
|
language: |
|
|
- it |
|
|
license: apache-2.0 |
|
|
tags: |
|
|
- automatic-speech-recognition |
|
|
- common_voice |
|
|
- generated_from_trainer |
|
|
metrics: |
|
|
- wer |
|
|
model-index: |
|
|
- name: wav2vec2-common_voice-it_en |
|
|
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. --> |
|
|
|
|
|
# wav2vec2-common_voice-it_en |
|
|
|
|
|
This model is a fine-tuned version of [facebook/wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) on the COMMON_VOICE - IT dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.0432 |
|
|
- Wer: 0.0322 |
|
|
|
|
|
## 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: 0.0003 |
|
|
- train_batch_size: 7 |
|
|
- eval_batch_size: 8 |
|
|
- seed: 42 |
|
|
- gradient_accumulation_steps: 2 |
|
|
- total_train_batch_size: 14 |
|
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
|
- lr_scheduler_type: linear |
|
|
- lr_scheduler_warmup_steps: 500 |
|
|
- num_epochs: 15.0 |
|
|
- mixed_precision_training: Native AMP |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|
|:-------------:|:-----:|:-----:|:---------------:|:------:| |
|
|
| 0.4885 | 0.7 | 1200 | 0.2958 | 0.2618 | |
|
|
| 0.2986 | 1.4 | 2400 | 0.1802 | 0.1629 | |
|
|
| 0.2515 | 2.1 | 3600 | 0.1379 | 0.1317 | |
|
|
| 0.2013 | 2.8 | 4800 | 0.1208 | 0.1178 | |
|
|
| 0.1651 | 3.5 | 6000 | 0.1110 | 0.1159 | |
|
|
| 0.1559 | 4.2 | 7200 | 0.0923 | 0.0948 | |
|
|
| 0.1337 | 4.9 | 8400 | 0.0928 | 0.0931 | |
|
|
| 0.1162 | 5.6 | 9600 | 0.0753 | 0.0782 | |
|
|
| 0.1164 | 6.3 | 10800 | 0.0700 | 0.0714 | |
|
|
| 0.1057 | 7.0 | 12000 | 0.0630 | 0.0656 | |
|
|
| 0.0904 | 7.7 | 13200 | 0.0619 | 0.0624 | |
|
|
| 0.0807 | 8.4 | 14400 | 0.0609 | 0.0566 | |
|
|
| 0.0759 | 9.1 | 15600 | 0.0514 | 0.0490 | |
|
|
| 0.0657 | 9.8 | 16800 | 0.0504 | 0.0470 | |
|
|
| 0.0556 | 10.5 | 18000 | 0.0511 | 0.0431 | |
|
|
| 0.0534 | 11.2 | 19200 | 0.0484 | 0.0408 | |
|
|
| 0.0498 | 11.9 | 20400 | 0.0436 | 0.0383 | |
|
|
| 0.0441 | 12.6 | 21600 | 0.0458 | 0.0365 | |
|
|
| 0.0398 | 13.3 | 22800 | 0.0471 | 0.0354 | |
|
|
| 0.0379 | 14.0 | 24000 | 0.0402 | 0.0327 | |
|
|
| 0.0333 | 14.7 | 25200 | 0.0438 | 0.0326 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.30.0.dev0 |
|
|
- Pytorch 2.0.1+cu117 |
|
|
- Datasets 2.12.0 |
|
|
- Tokenizers 0.13.3 |
|
|
|
