metadata
library_name: peft
language:
- it
license: apache-2.0
base_model: openai/whisper-medium
tags:
- generated_from_trainer
datasets:
- b-brave-clean
metrics:
- wer
model-index:
- name: Whisper Medium
results:
- task:
type: automatic-speech-recognition
name: Automatic Speech Recognition
dataset:
name: b-brave-clean
type: b-brave-clean
config: default
split: test
args: default
metrics:
- type: wer
value: 40.97421203438395
name: Wer
Whisper Medium
This model is a fine-tuned version of openai/whisper-medium on the b-brave-clean dataset. It achieves the following results on the evaluation set:
- Loss: 0.6343
- Wer: 40.9742
- Cer: 28.2112
- Lr: 0.0000
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: 0.0003
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use 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_ratio: 0.3
- num_epochs: 8
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer | Cer | Lr |
|---|---|---|---|---|---|---|
| 1.3936 | 1.0 | 251 | 1.2047 | 77.7937 | 48.2532 | 0.0001 |
| 0.8092 | 2.0 | 502 | 0.8939 | 161.6046 | 108.6945 | 0.0002 |
| 0.5533 | 3.0 | 753 | 0.8686 | 145.1289 | 117.0475 | 0.0003 |
| 0.3427 | 4.0 | 1004 | 0.6722 | 50.7163 | 34.9094 | 0.0002 |
| 0.1878 | 5.0 | 1255 | 0.6712 | 74.9284 | 63.9611 | 0.0002 |
| 0.1152 | 6.0 | 1506 | 0.6282 | 43.6963 | 29.8923 | 0.0001 |
| 0.0547 | 7.0 | 1757 | 0.6345 | 60.3152 | 53.7956 | 0.0001 |
| 0.0263 | 8.0 | 2008 | 0.6343 | 40.9742 | 28.2112 | 0.0000 |
Framework versions
- PEFT 0.14.0
- Transformers 4.48.3
- Pytorch 2.2.0
- Datasets 3.2.0
- Tokenizers 0.21.0