./whisper-large-cit-synth-do0.15-wd0-lr1e-05-spelled
This model is a fine-tuned version of openai/whisper-large-v3 on the SF 200 synth 2000 dataset. It achieves the following results on the evaluation set:
- Loss: 0.3516
- Wer: 15.7077
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: 8
- seed: 42
- distributed_type: multi-GPU
- 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: 100
- training_steps: 300
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.2758 | 0.4040 | 50 | 0.2947 | 19.0194 |
| 0.1631 | 0.8081 | 100 | 0.2827 | 19.4450 |
| 0.0654 | 1.2121 | 150 | 0.2808 | 16.7253 |
| 0.0576 | 1.6162 | 200 | 0.2795 | 15.5597 |
| 0.045 | 2.0202 | 250 | 0.3022 | 15.5042 |
| 0.0163 | 2.4242 | 300 | 0.3516 | 15.7077 |
Framework versions
- Transformers 4.41.2
- Pytorch 1.13.1+cu117
- Datasets 2.19.2
- Tokenizers 0.19.1
- Downloads last month
- 2