metadata
library_name: transformers
language:
- hi
license: apache-2.0
base_model: openai/whisper-small
tags:
- hf-asr-leaderboard
- generated_from_trainer
datasets:
- aihub_adult_baseline
model-index:
- name: whisper-small-Yfreq_speed_pause
results: []
whisper-small-Yfreq_speed_pause
This model is a fine-tuned version of openai/whisper-small on the aihub old30 adult freq speed pause changed dataset. It achieves the following results on the evaluation set:
- Loss: 0.1902
- Cer: 5.1809
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 16
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 2
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Cer |
---|---|---|---|---|
0.461 | 0.1289 | 100 | 0.2808 | 6.4732 |
0.2699 | 0.2579 | 200 | 0.2328 | 6.2265 |
0.2456 | 0.3868 | 300 | 0.2341 | 6.6494 |
0.2159 | 0.5158 | 400 | 0.2159 | 6.0033 |
0.1958 | 0.6447 | 500 | 0.2121 | 5.9387 |
0.1784 | 0.7737 | 600 | 0.2066 | 5.7390 |
0.1765 | 0.9026 | 700 | 0.1997 | 5.5745 |
0.0805 | 1.0316 | 800 | 0.1957 | 5.5392 |
0.0774 | 1.1605 | 900 | 0.1987 | 5.4100 |
0.0728 | 1.2895 | 1000 | 0.1965 | 5.2749 |
0.0676 | 1.4184 | 1100 | 0.1952 | 5.4159 |
0.0629 | 1.5474 | 1200 | 0.1918 | 5.2632 |
0.0644 | 1.6763 | 1300 | 0.1916 | 5.2514 |
0.0606 | 1.8053 | 1400 | 0.1903 | 5.2984 |
0.0664 | 1.9342 | 1500 | 0.1902 | 5.1809 |
Framework versions
- Transformers 4.46.0.dev0
- Pytorch 2.4.1+cu121
- Datasets 3.0.1
- Tokenizers 0.20.1