Spaces:
Configuration error
Configuration error
File size: 1,785 Bytes
2ba5a69 4a1e31e 2ba5a69 |
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 |
# OutfitChanger: Dress in any outfit you want.
PS : It is not currently possible to wear both bottoms and tops at same time.
## For deploy the model on Google Colab :
- !git clone https://huggingface.co/spaces/altayavci/OutfitChanger/
(PS: if you want clone the repo from Github please use <https://git-lfs.com>)
- cd /content/OutfitChanger
- !pip install -q gradio
- !pip install -r requirements.txt
- !python3 app.py
## CONFIGS:
| Library | Version |
|---------------------|------------|
| ip-adapter | https://github.com/tencent-ailab/IP-Adapter |
| torch | 2.1.0+cu121 |
| numpy | 1.23.5 |
| diffusers | 0.25.0 |
| transformers | 4.35.2 |
| accelerate | 0.25.0 |
| scipy | 1.11.4 |
| safetensors | 0.4.1 |
| controlnet_aux | 0.0.7 |
| pillow | 9.4.0 |
| opencv-python | 4.8.0 |
| Gradio | 4.13.0 |
- MODELS: IP-Adapter, SDv.15: SG161222/Realistic_Vision_V4.0_noVAE, mattmdjaga/segformer_b2_clothes, lllyasviel/control_v11p_sd15_openpose
- If you have a high resolution outfit image, it is recommended to set OPENAI_CONSISTENCY_VAE = ENABLE in the .env file.
- If you want to use SG161222/Realistic_Vision_V5.1_noVAE, it is recommended that you do not change the default vae.
* Django flask fastapi ile api yazıp docker kaldırılmalı
* İmg2txt’in verdiği cevaba göre seedler manipüle edilmeli
* Segmentasyon için sam ya da owlvit kullanılmalı
* Brush ile birnevi kullanıcıdan mask alınmalı
* Sd15 yerine SDXL kullanılmalı
* Referans kıyafetlerdeki sürekliliği artırmak için consistencydecoder kullanılabilir
* Eller ve ayaklara biraz daha focuslanılmalı
|