whisper-large-v2-vi / README.md
geninhu's picture
Update README.md
d9d1e3e
---
language:
- vi
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice_11_0
metrics:
- wer
model-index:
- name: openai/whisper-large-v2
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_11_0 vi
type: mozilla-foundation/common_voice_11_0
config: vi
split: test
args: vi
metrics:
- name: Wer
type: wer
value: 17.26
---
<!-- 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. -->
# openai/whisper-medium
This model is a fine-tuned version of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2) on the common_voice_11_0 dataset.
It achieves the following results on the evaluation set:
- Loss: 0.4124
- Wer: 17.26
## 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: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 5000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:-----:|:----:|:---------------:|:-------:|
| 0.0625 | 62.0 | 1000 | 0.4124 | 17.26 |
| 0.0239 | 124.0 | 2000 | 0.6964 | 19.08 |
| 0.0145 | 187.0 | 3000 | 0.7282 | 18.0 |
| 0.0066 | 249.0 | 4000 | 0.7481 | 20.02 |
| 0.0027 | 312.0 | 5000 | 0.7599 | 19.14 |
### Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.0+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2