Open rupeshs opened 1 year ago
This is cool & a great idea man! How do I run it on Google Colab?
@powderblock Thanks, simply checkout my reporepo https://github.com/rupeshs/visual-chatgpt/tree/add-colab-support , you can directly open it in colab
Great work
@rupeshs : Indeed, great work, given how quickly you put this up.
That said, I couldn't get it to work.
First I faced this error:
ModuleNotFoundError: No module named 'ldm'
And after solving that by moving the nested ldm
directory to the root directory, I get this:
ModuleNotFoundError: No module named 'ControlNet'
Any thoughts?
Great work
@rsreetech Thank you
@ajohansson-git Thank you, I have updated colab notebook, it will fix ldm error, could you please check again?
The gradio app is loading and then responding that the API key is wrong (tested on many valid api keys). The key is certainly fine, since I am using it in multiple colabs. Any ideas?
The gradio app is loading and then responding that the API key is wrong (tested on many valid api keys). The key is certainly fine, since I am using it in multiple colabs. Any ideas?
Even I am facing the same issue.
Are you using openAI free account API key? We need to use paid account API key.
Are you using openAI free account API key? We need to use paid account API key.
I am using the paid one
@sangeethkumar1997 Creat a new API key then try. If you are using colab, please try my implementation it will work https://github.com/rupeshs/visual-chatgpt/tree/add-colab-support.
The gradio app is loading and then responding that the API key is wrong (tested on many valid api keys). The key is certainly fine, since I am using it in multiple colabs. Any ideas?
@riccitensor We need to use OpenAI paid API key, also create a new API key, and try.
Yes, I'm using the paid one. I've fixed the issue. For me, it was this:
Wrong: %env OPENAI_API_KEY="aaa" Right: %env OPENAI_API_KEY=aaa
@rupeshs Thanks for the colab.
my colab not show the pic, what's wrong?
@fanyhchn I have quick fix for that see the pull request. https://github.com/microsoft/visual-chatgpt/pull/172
@fanyhchn You can use my latest colab it will work, (supports xformers and fp16 optimizations faster image generation) https://github.com/rupeshs/visual-chatgpt/tree/colab-xformers-support
I tried to add colab support. Tested text to image,ImageCaptioning,BLIP VQA,image to canny tools. https://github.com/rupeshs/visual-chatgpt/tree/add-colab-support