File size: 3,854 Bytes
0102037 |
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 |
---
license: apache-2.0
base_model: openai/whisper-large-v2
tags:
- generated_from_trainer
datasets:
- ravnursson_asr
metrics:
- wer
model-index:
- name: whisper-large-v2-fo-100h-30k-steps
results:
- task:
name: Automatic Speech Recognition
type: automatic-speech-recognition
dataset:
name: ravnursson_asr
type: ravnursson_asr
config: ravnursson_asr
split: test
args: ravnursson_asr
metrics:
- name: Wer
type: wer
value: 4.9124219851016715
---
<!-- 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. -->
[<img src="https://raw.githubusercontent.com/wandb/assets/main/wandb-github-badge-28.svg" alt="Visualize in Weights & Biases" width="200" height="32"/>](https://wandb.ai/setur/huggingface/runs/og6v8hvi)
# whisper-large-v2-fo-100h-30k-steps
This model is a fine-tuned version of [openai/whisper-large-v2](https://huggingface.co/openai/whisper-large-v2) on the ravnursson_asr dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0837
- Wer: 4.9124
## 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
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- training_steps: 30000
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|:-------------:|:------:|:-----:|:---------------:|:-------:|
| 0.2336 | 0.2320 | 1000 | 0.2811 | 20.5154 |
| 0.1631 | 0.4640 | 2000 | 0.1950 | 15.0191 |
| 0.124 | 0.6961 | 3000 | 0.1548 | 12.6334 |
| 0.1234 | 0.9281 | 4000 | 0.1323 | 11.0077 |
| 0.0568 | 1.1601 | 5000 | 0.1257 | 10.2174 |
| 0.0493 | 1.3921 | 6000 | 0.1204 | 9.5380 |
| 0.0473 | 1.6241 | 7000 | 0.1123 | 9.2158 |
| 0.0489 | 1.8561 | 8000 | 0.1012 | 8.1588 |
| 0.0193 | 2.0882 | 9000 | 0.0983 | 7.7159 |
| 0.0242 | 2.3202 | 10000 | 0.0933 | 7.1522 |
| 0.0171 | 2.5522 | 11000 | 0.0939 | 7.2680 |
| 0.0277 | 2.7842 | 12000 | 0.0876 | 7.0364 |
| 0.0077 | 3.0162 | 13000 | 0.0890 | 6.2563 |
| 0.0102 | 3.2483 | 14000 | 0.0883 | 6.9609 |
| 0.0089 | 3.4803 | 15000 | 0.0871 | 6.2110 |
| 0.0119 | 3.7123 | 16000 | 0.0854 | 6.4425 |
| 0.0109 | 3.9443 | 17000 | 0.0839 | 5.7379 |
| 0.0026 | 4.1763 | 18000 | 0.0850 | 5.9946 |
| 0.0063 | 4.4084 | 19000 | 0.0878 | 5.9644 |
| 0.0039 | 4.6404 | 20000 | 0.0896 | 6.2966 |
| 0.0038 | 4.8724 | 21000 | 0.0842 | 5.9895 |
| 0.0028 | 5.1044 | 22000 | 0.0811 | 5.7026 |
| 0.0021 | 5.3364 | 23000 | 0.0828 | 5.2194 |
| 0.0009 | 5.5684 | 24000 | 0.0850 | 5.1792 |
| 0.0023 | 5.8005 | 25000 | 0.0826 | 5.1188 |
| 0.0005 | 6.0325 | 26000 | 0.0823 | 5.0936 |
| 0.0004 | 6.2645 | 27000 | 0.0818 | 4.9225 |
| 0.0017 | 6.4965 | 28000 | 0.0839 | 4.9980 |
| 0.0002 | 6.7285 | 29000 | 0.0843 | 4.9577 |
| 0.0004 | 6.9606 | 30000 | 0.0837 | 4.9124 |
### Framework versions
- Transformers 4.43.1
- Pytorch 2.3.1+cu121
- Datasets 2.20.0
- Tokenizers 0.19.1
|