File size: 3,939 Bytes
7a040dd aea07a0 c8712f6 7a040dd c8712f6 7a040dd c8712f6 7a040dd aea07a0 c8712f6 7a040dd aea07a0 c8712f6 aea07a0 7a040dd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 |
---
license: apache-2.0
library_name: peft
tags:
- generated_from_trainer
base_model: openai/whisper-small
datasets:
- librispeech
model-index:
- name: Whisper Small English 1h
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 Small English 1h
This model is a fine-tuned version of [openai/whisper-small](https://huggingface.co/openai/whisper-small) on the librispeech dataset.
It achieves the following results on the evaluation set:
- Loss: 0.3666
## 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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 50
- num_epochs: 50
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 1.8938 | 1.0 | 39 | 2.5946 |
| 1.4637 | 2.0 | 78 | 2.0459 |
| 1.3003 | 3.0 | 117 | 1.6739 |
| 0.9415 | 4.0 | 156 | 1.2729 |
| 0.8165 | 5.0 | 195 | 1.0158 |
| 0.6326 | 6.0 | 234 | 0.9033 |
| 0.5716 | 7.0 | 273 | 0.7272 |
| 0.4662 | 8.0 | 312 | 0.6731 |
| 0.4133 | 9.0 | 351 | 0.6433 |
| 0.4 | 10.0 | 390 | 0.6248 |
| 0.3862 | 11.0 | 429 | 0.6103 |
| 0.3901 | 12.0 | 468 | 0.5962 |
| 0.3661 | 13.0 | 507 | 0.5841 |
| 0.3609 | 14.0 | 546 | 0.5739 |
| 0.3439 | 15.0 | 585 | 0.5660 |
| 0.3391 | 16.0 | 624 | 0.5581 |
| 0.3231 | 17.0 | 663 | 0.5510 |
| 0.3097 | 18.0 | 702 | 0.5441 |
| 0.2994 | 19.0 | 741 | 0.5373 |
| 0.2991 | 20.0 | 780 | 0.5304 |
| 0.2972 | 21.0 | 819 | 0.5240 |
| 0.2898 | 22.0 | 858 | 0.5187 |
| 0.2809 | 23.0 | 897 | 0.5142 |
| 0.2845 | 24.0 | 936 | 0.5119 |
| 0.269 | 25.0 | 975 | 0.5074 |
| 0.2721 | 26.0 | 1014 | 0.5033 |
| 0.2633 | 27.0 | 1053 | 0.5006 |
| 0.2623 | 28.0 | 1092 | 0.4984 |
| 0.2492 | 29.0 | 1131 | 0.4931 |
| 0.25 | 30.0 | 1170 | 0.4861 |
| 0.2479 | 31.0 | 1209 | 0.4833 |
| 0.2416 | 32.0 | 1248 | 0.4777 |
| 0.2356 | 33.0 | 1287 | 0.4794 |
| 0.2281 | 34.0 | 1326 | 0.4663 |
| 0.2191 | 35.0 | 1365 | 0.4605 |
| 0.2218 | 36.0 | 1404 | 0.4600 |
| 0.2078 | 37.0 | 1443 | 0.4545 |
| 0.2122 | 38.0 | 1482 | 0.4470 |
| 0.2076 | 39.0 | 1521 | 0.4510 |
| 0.2004 | 40.0 | 1560 | 0.4326 |
| 0.2004 | 41.0 | 1599 | 0.4280 |
| 0.1901 | 42.0 | 1638 | 0.4342 |
| 0.1856 | 43.0 | 1677 | 0.4107 |
| 0.1802 | 44.0 | 1716 | 0.4060 |
| 0.1677 | 45.0 | 1755 | 0.4029 |
| 0.1658 | 46.0 | 1794 | 0.3922 |
| 0.1589 | 47.0 | 1833 | 0.3845 |
| 0.152 | 48.0 | 1872 | 0.3790 |
| 0.1493 | 49.0 | 1911 | 0.3691 |
| 0.1426 | 50.0 | 1950 | 0.3666 |
### Framework versions
- PEFT 0.10.0
- Transformers 4.40.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.0
- Tokenizers 0.19.1 |