# Load model directly
from transformers import AutoProcessor, AutoModelForSpeechSeq2Seq
processor = AutoProcessor.from_pretrained("tmtms/whisper_checkpoints6")
model = AutoModelForSpeechSeq2Seq.from_pretrained("tmtms/whisper_checkpoints6", device_map="auto")Quick Links
whisper_checkpoints6
This model was trained from scratch on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2877
- Wer: 23.6973
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: 24
- eval_batch_size: 8
- seed: 42
- 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
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.009 | 2.5510 | 1000 | 0.0302 | 0.0380 |
| 0.1874 | 6.3291 | 2000 | 0.2390 | 34.0874 |
| 0.111 | 9.4937 | 3000 | 0.2581 | 29.7236 |
| 0.0599 | 12.6582 | 4000 | 0.2877 | 23.6973 |
Framework versions
- Transformers 4.48.3
- Pytorch 2.5.1+cu124
- Datasets 3.3.2
- Tokenizers 0.21.0
- Downloads last month
- 8
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("automatic-speech-recognition", model="tmtms/whisper_checkpoints6")