metadata
license: apache-2.0
library_name: peft
tags:
- generated_from_trainer
base_model: openai/whisper-small
datasets:
- mozilla-foundation/common_voice_6_1
model-index:
- name: LoRA-Frisian-10m
results: []
LoRA-Frisian-10m
This model is a fine-tuned version of openai/whisper-small on the mozilla-foundation/common_voice_6_1 dataset. It achieves the following results on the evaluation set:
- Loss: 1.4050
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: 0.001
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 12
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| No log | 1.0 | 15 | 2.2933 |
| 1.9518 | 2.0 | 30 | 2.0861 |
| 1.9518 | 3.0 | 45 | 1.7375 |
| 1.1576 | 4.0 | 60 | 1.4904 |
| 0.3487 | 5.0 | 75 | 1.4023 |
| 0.3487 | 6.0 | 90 | 1.3944 |
| 0.1202 | 7.0 | 105 | 1.4065 |
| 0.1202 | 8.0 | 120 | 1.3962 |
| 0.039 | 9.0 | 135 | 1.3982 |
| 0.0143 | 10.0 | 150 | 1.4102 |
| 0.0143 | 11.0 | 165 | 1.4045 |
| 0.0097 | 12.0 | 180 | 1.4050 |
Framework versions
- PEFT 0.10.0
- Transformers 4.40.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1