metadata
license: apache-2.0
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: English-1
results: []
English-1
This model is a fine-tuned version of openai/whisper-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.9309
- Wer: 91.4530
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-05
- train_batch_size: 8
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Wer |
---|---|---|---|---|
No log | 1.0 | 22 | 9.1617 | 100.0 |
No log | 2.0 | 44 | 4.2719 | 100.0 |
No log | 3.0 | 66 | 2.9975 | 100.0 |
No log | 4.0 | 88 | 2.2389 | 99.5726 |
No log | 5.0 | 110 | 1.9309 | 91.4530 |
No log | 6.0 | 132 | 1.9309 | 91.4530 |
No log | 7.0 | 154 | 1.9309 | 91.4530 |
No log | 8.0 | 176 | 1.9309 | 91.4530 |
No log | 9.0 | 198 | 1.9309 | 91.4530 |
No log | 10.0 | 220 | 1.9309 | 91.4530 |
Framework versions
- Transformers 4.30.0.dev0
- Pytorch 2.1.0+cu118
- Datasets 2.15.0
- Tokenizers 0.13.3