Conversation
|
Updates: In current, the project are using Therefore, only |
|
Seems that text2img also does not support the negative prompt with diffusers v0.3.0. Not sure when they will make the next release. https://github.com/huggingface/diffusers/blob/v0.3.0/src/diffusers/pipelines/stable_diffusion/pipeline_stable_diffusion.py#L94 |
|
Hum, you are right, Although I can run the text2Img() + negative_prompt without any error. But seems the negative_prompt didn't work correctly, I will do some tests again. Thank you for your reply! |
8e3700e to
3ead225
Compare
5273b92 to
ab732e3
Compare
279e8a0 to
1f86504
Compare
89933a3 to
35b822d
Compare
|
Looks good to me. |

Hi, I am immensely enjoying your project.
However, I have heard that there is also a "Negative Prompt" that can be input.
I have checked the diffusers source #1 and diffusers source #2, and seems your project only need few change to implement it.
This request is my change and I have tested via colab.
Thanks again for your work!