metadata
library_name: transformers
license: apache-2.0
base_model: Baselhany/Graduation_Project_distillation_Whisper_base222
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Graduation_Project_distillation_Whisper_base222
results: []
Graduation_Project_distillation_Whisper_base222
This model is a fine-tuned version of Baselhany/Graduation_Project_distillation_Whisper_base222 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0046
- Wer: 0.0687
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: 0.0001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 5
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 4.7293 | 1.0 | 1174 | 0.0073 | 0.0842 |
| 3.2433 | 2.0 | 2348 | 0.0065 | 0.0817 |
| 2.1238 | 3.0 | 3522 | 0.0064 | 0.0734 |
| 1.6646 | 4.0 | 4696 | 0.0061 | 0.0751 |
| 1.4589 | 5.0 | 5870 | 0.0060 | 0.0706 |
Framework versions
- Transformers 4.51.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.1