metadata
language:
- en
license: apache-2.0
base_model: openai/whisper-small
tags:
- generated_from_trainer
datasets:
- librispeech_asr
metrics:
- wer
model-index:
- name: Whisper-Small En-10m
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: librispeech
type: librispeech_asr
config: default
split: None
args: 'config: en, split: test-clean'
metrics:
- name: Wer
type: wer
value: 3.652744654395727
Whisper-Small En-10m
This model is a fine-tuned version of openai/whisper-small on the librispeech dataset. It achieves the following results on the evaluation set:
- Loss: 0.3711
- Wer: 3.6527
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: 5e-07
- train_batch_size: 4
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 300
- training_steps: 500
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
0.513 | 18.1818 | 100 | 0.7542 | 3.4448 |
0.2777 | 36.3636 | 200 | 0.6097 | 3.4693 |
0.0349 | 54.5455 | 300 | 0.3976 | 3.5732 |
0.0049 | 72.7273 | 400 | 0.3744 | 3.6324 |
0.0035 | 90.9091 | 500 | 0.3711 | 3.6527 |
Framework versions
- Transformers 4.41.0.dev0
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1