v1 #270
Merged
v1 #270
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Co-authored-by: Will Lin <[email protected]>
Co-authored-by: Peiyuan Zhang <[email protected]>
Signed-off-by: <> Co-authored-by: Will Lin <[email protected]> Co-authored-by: Ubuntu <ubuntu@awesome-gpu-name-8-inst-2tbsnfodvpomxv4tukw2dkfgyvz.c.nv-brev-20240723.internal> Co-authored-by: Ubuntu <ubuntu@awesome-gpu-name-9-inst-2tpydiudxfu1jg9xvpflm7oexie.c.nv-brev-20240723.internal>
…/FastVideo into rebased-refactor
Add wan dit
Co-authored-by: SolitaryThinker <[email protected]>
JerryZhou54
force-pushed
the
rebased-refactor
branch
from
c076f8a to
baa5b0e
Compare
Co-authored-by: Peiyuan Zhang <[email protected]>
Co-authored-by: Peiyuan Zhang <[email protected]> Co-authored-by: Will Lin <[email protected]>
zhisbug
approved these changes
Mar 29, 2025
| return torch.stack((o1, o2), dim=-1).flatten(-2) | ||
|
|
||
|
|
||
| @CustomOp.register("rotary_embedding") |
There was a problem hiding this comment.
a little bit surprised that we need to customize for rotary embedding class?
| uses_last_layer = feature_sample_layers[-1] in (len(hs_pool) - 1, -1) | ||
| if post_layer_norm is not None and uses_last_layer: | ||
| hs_pool[-1] = post_layer_norm(encoder_outputs) | ||
| return torch.cat(hs_pool, dim=-1) |
| from fastvideo.v1.inference_args import InferenceArgs | ||
| from fastvideo.v1.logger import init_logger | ||
|
|
||
| from ..composed_pipeline_base import ComposedPipelineBase |
There was a problem hiding this comment.
I saw a lot of relative imports throughout; can we avoid relative imports?
| def __init__(self, vae) -> None: | ||
| self.vae = vae | ||
|
|
||
| def forward( |
There was a problem hiding this comment.
one weird thing i noticed: you only use forward context in the text_encoding.py, not anywhere else? correct? if yes -- why?
There was a problem hiding this comment.
we also use it for DiTs (set in the denoising stage).
There was a problem hiding this comment.
input validation being a stage sounds strange, but i guess ok currently.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
[ ] Attn backend (PY)
[ ] Wan text encoder (PY)
[ ] wan vae (Wei)
[ ] wan pipeline (PY & Wei)
[ ] Merge wan dit code to refactor (Will)
[ ] Clean up code & loader directory（Will)
[ ] hunyuan text encoder (Will)