-
Notifications
You must be signed in to change notification settings - Fork 12
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add Stable Diffusion demo #100
base: main
Are you sure you want to change the base?
Conversation
tripy/examples/diffusion/model.py
Outdated
- Paper: https://arxiv.org/abs/1706.03762v7 | ||
""" | ||
|
||
if is_causal: # this path is not called in demoDiffusion |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Can we drop this if it's not called?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I have left it for now in case we want to add it as part of the API somewhere
tripy/examples/diffusion/model.py
Outdated
return tp.cast(tp.softmax((qk + attn_mask) if attn_mask is not None else qk, -1), query.dtype) @ value | ||
|
||
|
||
def sequential(input: tp.Tensor, ll: List[Callable[[tp.Tensor], tp.Tensor]]): |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I think it's worth making this part of the API, similar to torch.nn.Sequential
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is being done here: #321
0d3c2cd
to
c7c81bd
Compare
c7c81bd
to
5142a42
Compare
Signed-off-by: Akhil Goel <[email protected]>
Signed-off-by: Akhil Goel <[email protected]>
Root cause: Index for denoising timesteps were reversed while refactoring. Signed-off-by: Akhil Goel <[email protected]>
Signed-off-by: Akhil Goel <[email protected]>
Signed-off-by: Akhil Goel <[email protected]>
Signed-off-by: Akhil Goel <[email protected]>
Signed-off-by: Akhil Goel <[email protected]>
5142a42
to
d9dd478
Compare
Cleanup for Tripy inference, torch inference pipeline, and accuracy test are still WIP. README is currently minimal as well.