Downloads · 30 days
176
1% of all-time downloads
vit-base-blur
WT-MM/vit-base-blur
vit-base-blur is a image classification model from WT-MM. Use it when you need a label for an image. 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
16.7K
Public
Parameters
85.8M
5.5 GB on disk
Likes
7
Public
Repo makeup
Click a slice to open those files.
.bin343 MB · 50%
At a glance
- Task
- Image Classification
- Library
- transformers
- License
- apache-2.0
- Model type
- vit
- Access
- Public
- Created
- Jul 5, 2023
- Updated
- Jul 20, 2023
- SHA
- dc4baf4d
Datasets
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. -->
vit-base-blur
This model is a fine-tuned version of google/vit-base-patch16-224-in21k on the blurry images dataset. It achieves the following results on the evaluation set:
- Loss: 0.0008
- Accuracy: 1.0
Model description
Model trained for binary classification between 'noisy' (blurry) and clean images, where 'noisy' images are the result of unfinished/insufficient passes from an LDM for image generation
Intended uses & limitations
More information needed
Training and evaluation data
1000ish clean and blurry images using 30 and 10 steps respectively on SD2.1
Training procedure
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 0.0001
- 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
- num_epochs: 12
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.0082 | 1.02 | 100 | 0.0107 | 1.0 |
| 0.0079 | 2.04 | 200 | 0.0052 | 1.0 |
| 0.0029 | 3.06 | 300 | 0.0028 | 1.0 |
| 0.002 | 4.08 | 400 | 0.0020 | 1.0 |
| 0.0016 | 5.1 | 500 | 0.0015 | 1.0 |
| 0.0013 | 6.12 | 600 | 0.0013 | 1.0 |
| 0.0011 | 7.14 | 700 | 0.0011 | 1.0 |
| 0.001 | 8.16 | 800 | 0.0010 | 1.0 |
| 0.0009 | 9.18 | 900 | 0.0009 | 1.0 |
| 0.0008 | 10.2 | 1000 | 0.0008 | 1.0 |
| 0.0008 | 11.22 | 1100 | 0.0008 | 1.0 |
Framework versions
- Transformers 4.30.2
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3
- Task
- Image Classification
- Library
- transformers
- Type
- vit
- License
- apache-2.0
- Created
- Jul 5, 2023
- Updated
- Jul 20, 2023