Downloads · 30 days
4
17% of all-time downloads
w2v2-base-pretrained_lr5e-5_at0.8_da0.5
MelanieKoe/w2v2-base-pretrained_lr5e-5_at0.8_da0.5
w2v2-base-pretrained_lr5e-5_at0.8_da0.5 is a automatic speech recognition model from MelanieKoe. Use it when you need speech turned into text. It is set up for transformers. The card lists the license as apache-2.0.
should probably proofread and complete it, then remove this comment. --
All-time downloads
24
Public
Parameters
94.4M
12.1 GB on disk
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors378 MB · 100%
At a glance
- Task
- Automatic Speech Recognition
- Library
- transformers
- License
- apache-2.0
- Model type
- wav2vec2
- Access
- Public
- Created
- Apr 1, 2024
- Updated
- Apr 18, 2024
- SHA
- aeb60b62
Base models
Model card
From the Hugging Face model README
<!-- 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. -->
w2v2-base-pretrained_lr5e-5_at0.8_da0.5
This model is a fine-tuned version of facebook/wav2vec2-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.3736
- Wer: 0.1858
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: 5e-05
- train_batch_size: 32
- 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: 1000
- training_steps: 4000
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Wer |
|---|---|---|---|---|
| 20.81 | 10.87 | 250 | 3.8675 | 1.0 |
| 3.2992 | 21.74 | 500 | 3.1858 | 1.0 |
| 3.0813 | 32.61 | 750 | 3.0758 | 1.0 |
| 2.3718 | 43.48 | 1000 | 1.3480 | 0.8266 |
| 0.3075 | 54.35 | 1250 | 1.6498 | 0.2290 |
| 0.1081 | 65.22 | 1500 | 1.8213 | 0.2012 |
| 0.0732 | 76.09 | 1750 | 1.8933 | 0.1952 |
| 0.0517 | 86.96 | 2000 | 2.0154 | 0.2059 |
| 0.0386 | 97.83 | 2250 | 2.0444 | 0.1948 |
| 0.0323 | 108.7 | 2500 | 2.2603 | 0.2003 |
| 0.0272 | 119.57 | 2750 | 2.2578 | 0.1952 |
| 0.0234 | 130.43 | 3000 | 2.2854 | 0.1880 |
| 0.0203 | 141.3 | 3250 | 2.3553 | 0.1867 |
| 0.0181 | 152.17 | 3500 | 2.3723 | 0.1905 |
| 0.0165 | 163.04 | 3750 | 2.3793 | 0.1854 |
| 0.016 | 173.91 | 4000 | 2.3736 | 0.1858 |
Framework versions
- Transformers 4.35.0
- Pytorch 2.0.0
- Datasets 2.14.6
- Tokenizers 0.14.1
- Task
- Automatic Speech Recognition
- Library
- transformers
- Type
- wav2vec2
- License
- apache-2.0
- Created
- Apr 1, 2024
- Updated
- Apr 18, 2024