metadata
library_name: transformers
language:
- dv
license: apache-2.0
base_model: openai/whisper-base
tags:
- generated_from_trainer
datasets:
- fsicoli/common_voice_15_0
metrics:
- wer
model-index:
- name: Whisper Base Dv - Nuwan
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: Common Voice 15
type: fsicoli/common_voice_15_0
config: dv
split: test
args: dv
metrics:
- name: Wer
type: wer
value: 110.55416318574214
Whisper Base Dv - Nuwan
This model is a fine-tuned version of openai/whisper-base on the Common Voice 15 dataset. It achieves the following results on the evaluation set:
- Loss: 0.9098
- Wer Ortho: 194.4305
- Wer: 110.5542
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-06
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: constant_with_warmup
- lr_scheduler_warmup_steps: 100
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer Ortho | Wer |
|---|---|---|---|---|---|
| 2.3429 | 2.3669 | 400 | 2.2946 | 202.8266 | 113.8993 |
| 1.2184 | 4.7337 | 800 | 1.2136 | 203.0430 | 112.1693 |
| 1.0668 | 7.1006 | 1200 | 1.0778 | 201.4657 | 111.3182 |
| 1.0197 | 9.4675 | 1600 | 1.0351 | 199.5743 | 111.5619 |
| 0.9854 | 11.8343 | 2000 | 1.0064 | 194.9470 | 110.4323 |
| 0.9512 | 14.2012 | 2400 | 0.9839 | 189.6776 | 109.2157 |
| 0.9335 | 16.5680 | 2800 | 0.9635 | 189.0215 | 109.1896 |
| 0.9074 | 18.9349 | 3200 | 0.9436 | 188.9866 | 109.4072 |
| 0.8867 | 21.3018 | 3600 | 0.9265 | 193.4394 | 110.2983 |
| 0.8529 | 23.6686 | 4000 | 0.9098 | 194.4305 | 110.5542 |
Framework versions
- Transformers 4.57.1
- Pytorch 2.9.0+cu126
- Datasets 3.6.0
- Tokenizers 0.22.1