metadata
language:
- ka
license: apache-2.0
tags:
- whisper-event
- generated_from_trainer
datasets:
- mozilla-foundation/common_voice_11_0
metrics:
- wer
model-index:
- name: Whisper Small Georgian
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: mozilla-foundation/common_voice_11_0 kab
type: mozilla-foundation/common_voice_11_0
config: kab
split: test
args: kab
metrics:
- name: Wer
type: wer
value: 41.47529142173956
Whisper Small Georgian
This model is a fine-tuned version of openai/whisper-small on the mozilla-foundation/common_voice_11_0 kab dataset. It achieves the following results on the evaluation set:
- Loss: 0.5547
- Wer: 41.4753
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: 64
- eval_batch_size: 32
- seed: 42
- distributed_type: multi-GPU
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 400
- training_steps: 5000
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
0.4105 | 1.42 | 1000 | 0.5508 | 48.1244 |
0.258 | 2.84 | 2000 | 0.4905 | 42.0295 |
0.0998 | 4.27 | 3000 | 0.5220 | 41.6808 |
0.064 | 5.69 | 4000 | 0.5547 | 41.4753 |
0.0356 | 7.11 | 5000 | 0.5889 | 41.5737 |
Framework versions
- Transformers 4.26.0.dev0
- Pytorch 1.13.1+cu117
- Datasets 2.7.1.dev0
- Tokenizers 0.13.2