-
Notifications
You must be signed in to change notification settings - Fork 77
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 Movie Gen #783
base: master
Are you sure you want to change the base?
Add Movie Gen #783
Conversation
Movie gen
enable jit_level `O2` support
- preserve image / video orientation in transformations
# Conflicts: # mindone/utils/env.py
# Conflicts: # examples/moviegen/inference.py
# Conflicts: # examples/moviegen/README.md # examples/moviegen/configs/tae/train/mixed_256x256x16.yaml # examples/moviegen/configs/tae/train/mixed_256x256x32.yaml # examples/moviegen/docs/report.md # examples/moviegen/mg/models/tae/tae.py # examples/moviegen/scripts/args_train_tae.py # examples/moviegen/scripts/run/run_train_tae.sh # examples/moviegen/scripts/train_tae.py
# Conflicts: # examples/moviegen/docs/report.md
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.
there are 7 scripts in the top folder (train.py, train_tae.py, inferecne_text_enc.py...), which is unlike previous case using a unified train.py/infer.py as the only entrance. So I would suggest put them in scripts/
folder to keep the top code structure clean
Meta researchers found that scaling the training data, compute, and model parameters of a simple | ||
Transformer-based ([LLaMa3](https://arxiv.org/abs/2407.21783)) model trained with | ||
[Flow Matching](https://arxiv.org/abs/2210.02747) yields high quality generative models for video or audio. |
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 better to delete this paragraph to keep the README concise.
What does this PR do?
Fixes # (issue)
Adds # (feature)
Before submitting
What's New
. Here are thedocumentation guidelines
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.
@xxx