search

levihsu / OOTDiffusion

Official implementation of OOTDiffusion: Outfitting Fusion based Latent Diffusion for Controllable Virtual Try-on
Other
5.45k stars 801 forks source link

how to train model with our own dataset #153

Open Jaynls opened 5 months ago

Jaynls commented 5 months ago

I hope this message finds you well. I wanted to take a moment to express my sincere gratitude for your incredible work on [name of the project]. Your dedication to making the code public and your responsiveness to issues and queries have been invaluable to me and countless others in the community. Your willingness to share your expertise and support fellow developers is truly commendable.

I've been exploring your project and found it immensely helpful in my learning journey. However, as someone relatively new to this field, I have a question that I hope you can help me with. I'm eager to understand how to train a model with our own dataset.

For instance, I have images of both clothing and people. Would these images alone be sufficient to train the model effectively, or would I need additional data such as OpenPose or Human Parsing of given image ? Furthermore, could you point me in the right direction to find scripts or resources for training or fine-tuning the model with our custom dataset?

Any guidance or advice you could provide on this matter would be greatly appreciated. Thank you once again for your contributions to the community and for your willingness to assist others like myself. Your work has truly made a difference.

levihsu commented 5 months ago

Our training script will be released later. Before that you could refer to some unofficial implementations like https://github.com/lyc0929/OOTDiffusion-train

timmyhk852 commented 5 months ago

Our training script will be released later. Before that you could refer to some unofficial implementations like https://github.com/lyc0929/OOTDiffusion-train

hi, does this mean we can use/upload our own human models and clothes? Can this be done using A1111 webui?