Configuration Parsing
Warning:
In adapter_config.json: "peft.task_type" must be a string
whisper-large-v3-sandi-train-1
This model is a fine-tuned version of openai/whisper-large-v3 on the ntnu-smil/sandi2025-ds dataset. It achieves the following results on the evaluation set:
- Loss: 1.3156
- Wer: 160.7027
- Cer: 105.7464
- Decode Runtime: 615.9674
- Wer Runtime: 0.2787
- Cer Runtime: 0.5808
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: 7e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 32
- total_train_batch_size: 1024
- optimizer: Use adamw_torch with betas=(0.9,0.98) and epsilon=1e-06 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- training_steps: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer | Decode Runtime | Wer Runtime | Cer Runtime |
|---|---|---|---|---|---|---|---|---|
| 3.5332 | 1.05 | 4 | 1.9101 | 115.7476 | 319.7769 | 618.9739 | 0.3112 | 0.8226 |
| 1.7555 | 2.1 | 8 | 1.6223 | 223.6792 | 155.0881 | 662.9872 | 0.3751 | 0.7464 |
| 1.4625 | 3.15 | 12 | 1.4583 | 202.3767 | 117.3829 | 605.9002 | 0.3008 | 0.5968 |
| 2.6716 | 5.05 | 16 | 1.3557 | 171.9805 | 107.7457 | 621.8155 | 0.2918 | 0.5871 |
| 1.4119 | 6.1 | 20 | 1.3156 | 160.7027 | 105.7464 | 615.9674 | 0.2787 | 0.5808 |
Framework versions
- PEFT 0.15.1
- Transformers 4.50.3
- Pytorch 2.1.0+cu118
- Datasets 3.5.0
- Tokenizers 0.21.1
- Downloads last month
- 0
Inference Providers
NEW
This model isn't deployed by any Inference Provider.
🙋
Ask for provider support
Model tree for ntnu-smil/whisper-large-v3-sandi-train-1
Base model
openai/whisper-large-v3