|
--- |
|
license: apache-2.0 |
|
base_model: openai/whisper-large-v3 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- wer |
|
model-index: |
|
- name: whisper-large-v3-atco2-asr-atcosim |
|
results: [] |
|
--- |
|
|
|
<!-- This model card has been generated automatically according to the information the Trainer had access to. You |
|
should probably proofread and complete it, then remove this comment. --> |
|
|
|
# whisper-large-v3-atco2-asr-atcosim |
|
|
|
This model is a fine-tuned version of [openai/whisper-large-v3](https://huggingface.co./openai/whisper-large-v3) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.1039 |
|
- Wer: 22.2698 |
|
|
|
## 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: 8 |
|
- seed: 42 |
|
- distributed_type: multi-GPU |
|
- num_devices: 4 |
|
- total_train_batch_size: 64 |
|
- total_eval_batch_size: 32 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 250 |
|
- training_steps: 12644 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Wer | |
|
|:-------------:|:-----:|:-----:|:---------------:|:-------:| |
|
| 0.049 | 1.97 | 250 | 0.0613 | 41.3521 | |
|
| 0.0168 | 3.94 | 500 | 0.0656 | 25.3775 | |
|
| 0.0076 | 5.91 | 750 | 0.0703 | 16.7505 | |
|
| 0.0028 | 7.87 | 1000 | 0.0722 | 23.0540 | |
|
| 0.001 | 9.84 | 1250 | 0.0727 | 21.6365 | |
|
| 0.0008 | 11.81 | 1500 | 0.0728 | 24.0815 | |
|
| 0.0012 | 13.78 | 1750 | 0.0712 | 36.9653 | |
|
| 0.0025 | 15.75 | 2000 | 0.0701 | 21.1248 | |
|
| 0.0005 | 17.72 | 2250 | 0.0745 | 10.2458 | |
|
| 0.0006 | 19.69 | 2500 | 0.0781 | 26.3169 | |
|
| 0.0013 | 21.65 | 2750 | 0.0760 | 15.4127 | |
|
| 0.0073 | 23.62 | 3000 | 0.0790 | 85.4764 | |
|
| 0.0038 | 25.59 | 3250 | 0.0724 | 44.4682 | |
|
| 0.0003 | 27.56 | 3500 | 0.0772 | 37.4056 | |
|
| 0.0003 | 29.53 | 3750 | 0.0778 | 31.2238 | |
|
| 0.0 | 31.5 | 4000 | 0.0806 | 22.4040 | |
|
| 0.0 | 33.46 | 4250 | 0.0831 | 20.6886 | |
|
| 0.0 | 35.43 | 4500 | 0.0847 | 20.3322 | |
|
| 0.0 | 37.4 | 4750 | 0.0860 | 20.7935 | |
|
| 0.0 | 39.37 | 5000 | 0.0871 | 20.3657 | |
|
| 0.0 | 41.34 | 5250 | 0.0880 | 20.5293 | |
|
| 0.0 | 43.31 | 5500 | 0.0889 | 20.7977 | |
|
| 0.0 | 45.28 | 5750 | 0.0898 | 20.4957 | |
|
| 0.0 | 47.24 | 6000 | 0.0906 | 20.9612 | |
|
| 0.0 | 49.21 | 6250 | 0.0914 | 20.8564 | |
|
| 0.0 | 51.18 | 6500 | 0.0921 | 21.1919 | |
|
| 0.0 | 53.15 | 6750 | 0.0928 | 20.7809 | |
|
| 0.0 | 55.12 | 7000 | 0.0934 | 21.1793 | |
|
| 0.0 | 57.09 | 7250 | 0.0941 | 21.2087 | |
|
| 0.0 | 59.06 | 7500 | 0.0947 | 21.2255 | |
|
| 0.0 | 61.02 | 7750 | 0.0953 | 21.4142 | |
|
| 0.0 | 62.99 | 8000 | 0.0959 | 21.1961 | |
|
| 0.0 | 64.96 | 8250 | 0.0966 | 21.1080 | |
|
| 0.0 | 66.93 | 8500 | 0.0972 | 21.0955 | |
|
| 0.0 | 68.9 | 8750 | 0.0978 | 21.4226 | |
|
| 0.0 | 70.87 | 9000 | 0.0983 | 21.3681 | |
|
| 0.0 | 72.83 | 9250 | 0.0988 | 21.6532 | |
|
| 0.0 | 74.8 | 9500 | 0.0994 | 21.6155 | |
|
| 0.0 | 76.77 | 9750 | 0.0999 | 21.5107 | |
|
| 0.0 | 78.74 | 10000 | 0.1005 | 21.3974 | |
|
| 0.0 | 80.71 | 10250 | 0.1010 | 21.6407 | |
|
| 0.0 | 82.68 | 10500 | 0.1014 | 21.7120 | |
|
| 0.0 | 84.65 | 10750 | 0.1019 | 21.8755 | |
|
| 0.0 | 86.61 | 11000 | 0.1023 | 21.9510 | |
|
| 0.0 | 88.58 | 11250 | 0.1027 | 21.9636 | |
|
| 0.0 | 90.55 | 11500 | 0.1030 | 22.0223 | |
|
| 0.0 | 92.52 | 11750 | 0.1033 | 22.0265 | |
|
| 0.0 | 94.49 | 12000 | 0.1036 | 22.3536 | |
|
| 0.0 | 96.46 | 12250 | 0.1038 | 22.3956 | |
|
| 0.0 | 98.43 | 12500 | 0.1039 | 22.2698 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.35.0 |
|
- Pytorch 2.0.1+cu117 |
|
- Datasets 2.12.0 |
|
- Tokenizers 0.14.1 |
|
|
