|
--- |
|
language: |
|
- zh |
|
license: apache-2.0 |
|
library_name: peft |
|
tags: |
|
- hf-asr-leaderboard |
|
- generated_from_trainer |
|
datasets: |
|
- mozilla-foundation/common_voice_15_0 |
|
base_model: openai/whisper-tiny |
|
model-index: |
|
- name: Whisper tiny TW - AlanDlink |
|
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. --> |
|
|
|
# Whisper tiny TW - AlanDlink |
|
|
|
This model is a fine-tuned version of [openai/whisper-tiny](https://huggingface.co/openai/whisper-tiny) on the Common Voice 15.0 dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.6078 |
|
|
|
## 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: 4 |
|
- eval_batch_size: 4 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 16 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 500 |
|
- training_steps: 4000 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | |
|
|:-------------:|:-----:|:----:|:---------------:| |
|
| 3.3802 | 0.67 | 500 | 3.3992 | |
|
| 2.1962 | 1.33 | 1000 | 2.1643 | |
|
| 1.4348 | 2.0 | 1500 | 1.4068 | |
|
| 0.7108 | 2.67 | 2000 | 0.6926 | |
|
| 0.6801 | 3.33 | 2500 | 0.6374 | |
|
| 0.6273 | 4.0 | 3000 | 0.6195 | |
|
| 0.6001 | 4.67 | 3500 | 0.6106 | |
|
| 0.6082 | 5.33 | 4000 | 0.6078 | |
|
|
|
|
|
### Framework versions |
|
|
|
- PEFT 0.7.1 |
|
- Transformers 4.36.2 |
|
- Pytorch 2.1.2+cu121 |
|
- Datasets 2.16.0 |
|
- Tokenizers 0.15.0 |
