File size: 4,029 Bytes
cf07ab8 |
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 |
---
library_name: peft
language:
- en
license: apache-2.0
base_model: openai/whisper-small.en
tags:
- generated_from_trainer
metrics:
- wer
model-index:
- name: Whisper Small NSC part 1,2,3 (2000 steps) - Jarrett Er
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 NSC part 1,2,3 (2000 steps) - Jarrett Er
This model is a fine-tuned version of [openai/whisper-small.en](https://huggingface.co/openai/whisper-small.en) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.9590
- Wer: 5.3473
## 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: 0.0001
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- training_steps: 2000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:------:|:----:|:---------------:|:--------:|
| 4.645 | 0.2008 | 50 | 4.1237 | 140.7387 |
| 2.1925 | 0.4016 | 100 | 1.7341 | 28.5557 |
| 1.611 | 0.6024 | 150 | 1.6017 | 22.2712 |
| 1.5577 | 0.8032 | 200 | 1.5606 | 21.5546 |
| 1.5236 | 1.0040 | 250 | 1.5283 | 24.0077 |
| 1.4578 | 1.2048 | 300 | 1.4763 | 20.7552 |
| 1.3667 | 1.4056 | 350 | 1.4251 | 19.3495 |
| 1.3611 | 1.6064 | 400 | 1.3817 | 18.3572 |
| 1.3913 | 1.8072 | 450 | 1.3487 | 18.1918 |
| 1.3194 | 2.0080 | 500 | 1.3160 | 16.9515 |
| 1.2678 | 2.2088 | 550 | 1.2973 | 15.9592 |
| 1.1661 | 2.4096 | 600 | 1.2656 | 14.7740 |
| 1.2014 | 2.6104 | 650 | 1.2339 | 14.3330 |
| 1.1511 | 2.8112 | 700 | 1.2054 | 14.0573 |
| 1.1239 | 3.0120 | 750 | 1.1740 | 13.0375 |
| 1.0758 | 3.2129 | 800 | 1.1534 | 12.7343 |
| 1.1137 | 3.4137 | 850 | 1.1401 | 11.7971 |
| 1.0582 | 3.6145 | 900 | 1.1203 | 10.8324 |
| 1.0749 | 3.8153 | 950 | 1.1061 | 10.5568 |
| 1.0305 | 4.0161 | 1000 | 1.0951 | 10.6119 |
| 1.0105 | 4.2169 | 1050 | 1.0773 | 9.7299 |
| 1.0031 | 4.4177 | 1100 | 1.0757 | 9.9228 |
| 1.0054 | 4.6185 | 1150 | 1.0621 | 8.8754 |
| 0.9836 | 4.8193 | 1200 | 1.0455 | 8.4895 |
| 0.9663 | 5.0201 | 1250 | 1.0369 | 7.8556 |
| 0.9473 | 5.2209 | 1300 | 1.0261 | 7.8556 |
| 0.9784 | 5.4217 | 1350 | 1.0184 | 7.7178 |
| 0.9462 | 5.6225 | 1400 | 1.0124 | 7.6626 |
| 0.9414 | 5.8233 | 1450 | 1.0069 | 7.2492 |
| 0.9262 | 6.0241 | 1500 | 1.0005 | 7.0562 |
| 0.9127 | 6.2249 | 1550 | 0.9946 | 6.8633 |
| 0.9324 | 6.4257 | 1600 | 0.9866 | 6.3120 |
| 0.9043 | 6.6265 | 1650 | 0.9793 | 6.3947 |
| 0.9223 | 6.8273 | 1700 | 0.9725 | 5.5678 |
| 0.9087 | 7.0281 | 1750 | 0.9697 | 5.4576 |
| 0.8943 | 7.2289 | 1800 | 0.9659 | 5.6781 |
| 0.8962 | 7.4297 | 1850 | 0.9634 | 5.5678 |
| 0.8992 | 7.6305 | 1900 | 0.9611 | 5.4851 |
| 0.9099 | 7.8313 | 1950 | 0.9601 | 5.4851 |
| 0.9182 | 8.0321 | 2000 | 0.9590 | 5.3473 |
### Framework versions
- PEFT 0.14.0
- Transformers 4.45.2
- Pytorch 2.5.1+cu124
- Datasets 3.2.1.dev0
- Tokenizers 0.20.3 |