metadata
license: apache-2.0
tags:
- generated_from_trainer
base_model: openai/whisper-large-v2
model-index:
- name: whisper-large-ver1
results: []
whisper-large-ver1
This model is a fine-tuned version of openai/whisper-large-v2 on an unknown dataset. It achieves the following results on the evaluation set:
- Cer: 10.8895
- Loss: 0.4810
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: 5e-06
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- 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 | Cer | Validation Loss |
|---|---|---|---|---|
| 0.024 | 5.6 | 1000 | 11.4526 | 0.3606 |
| 0.0038 | 11.2 | 2000 | 10.7559 | 0.4166 |
| 0.0009 | 16.81 | 3000 | 10.8609 | 0.4669 |
| 0.0007 | 22.45 | 4000 | 10.8895 | 0.4810 |
Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.0.0+cu118
- Datasets 2.18.0
- Tokenizers 0.15.2