metadata
license: apache-2.0
base_model: openai/whisper-tiny
tags:
- generated_from_trainer
datasets:
- PolyAI/minds14
metrics:
- wer
model-index:
- name: michaelsh/whisper-tiny-minds-v5-numproc1
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: PolyAI/minds14
type: PolyAI/minds14
config: en-US
split: train
args: en-US
metrics:
- name: Wer
type: wer
value: 0.36609955891619406
michaelsh/whisper-tiny-minds-v5-numproc1
This model is a fine-tuned version of openai/whisper-tiny on the PolyAI/minds14 dataset. It achieves the following results on the evaluation set:
- Loss: 0.6160
- Wer Ortho: 0.3635
- Wer: 0.3661
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: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
---|---|---|---|---|---|
3.1128 | 1.0 | 29 | 1.5110 | 0.8028 | 0.7574 |
0.6583 | 2.0 | 58 | 0.5695 | 0.4347 | 0.4316 |
0.3271 | 3.0 | 87 | 0.5171 | 0.3945 | 0.3913 |
0.2003 | 4.0 | 116 | 0.5165 | 0.3912 | 0.3907 |
0.1189 | 5.0 | 145 | 0.5296 | 0.3819 | 0.3825 |
0.0623 | 6.0 | 174 | 0.5532 | 0.3747 | 0.3737 |
0.0326 | 7.0 | 203 | 0.5614 | 0.3865 | 0.3882 |
0.0149 | 8.0 | 232 | 0.6009 | 0.3628 | 0.3655 |
0.0093 | 9.0 | 261 | 0.6024 | 0.3707 | 0.3762 |
0.0038 | 10.0 | 290 | 0.6160 | 0.3635 | 0.3661 |
Framework versions
- Transformers 4.32.0.dev0
- Pytorch 2.0.0
- Datasets 2.1.0
- Tokenizers 0.13.3