metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- common_voice_11_0
metrics:
- wer
model-index:
- name: christoph-large-v2
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: common_voice_11_0
type: common_voice_11_0
config: sl
split: test
args: sl
metrics:
- name: Wer
type: wer
value: 40.46335421827189
christoph-large-v2
This model is a fine-tuned version of openai/whisper-large-v2 on the common_voice_11_0 dataset. It achieves the following results on the evaluation set:
- Loss: 0.3209
- Wer: 40.4634
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- 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: 500
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 0.0208 | 6.08 | 1000 | 0.2605 | 24.4645 |
| 0.0026 | 12.16 | 2000 | 0.2733 | 18.8110 |
| 0.0003 | 18.24 | 3000 | 0.3132 | 38.5400 |
| 0.0002 | 24.32 | 4000 | 0.3209 | 40.4634 |
Framework versions
- Transformers 4.28.0.dev0
- Pytorch 1.13.1+cu116
- Datasets 2.10.1
- Tokenizers 0.13.2