mirror of
https://github.com/deepseek-ai/Janus.git
synced 2026-11-05 00:57:18 +00:00
[GH-ISSUE #144] txt+image as input generating image #110
Labels
No labels
pull-request
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
deepseek-ai/Janus#110
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Originally created by @Youngon on GitHub (Feb 6, 2025).
Original GitHub issue: https://github.com/deepseek-ai/Janus/issues/144
Could this model be used in such situation, user input a prompt along with a image, generate a result image.
for example:
prompt[generate a image suck like the input image] + image[IMG.png] -> generated image
@MqLeet commented on GitHub (Feb 8, 2025):
I also wonder whether Janus-Pro-7B has this ability!
@Youngon commented on GitHub (Feb 8, 2025):
I have changed the pipeline and can do "txt+img -> img" work, but the model doesn't work well, so looking up to the upgrade from janus.
@MqLeet commented on GitHub (Feb 8, 2025):
Hi @Youngon , could you please share your modified "txt+img -> img" pipeline and corresponding preprocess codes? Thank you!🤗
@Youngon commented on GitHub (Feb 8, 2025):
I checked my pipeline, if masking the image embeddings, it can generate image according to the prompt; while I add unstanding part, can successfully unstand the content of input image.
so the model doesn't have the abblity to do "txt + img -> img" task.
@Youngon commented on GitHub (Feb 8, 2025):
it does not work at all, the model has no abblity to use input image generating new image.
@MqLeet commented on GitHub (Feb 8, 2025):
alright, thanks for your time! 🥰