GitLab

* bugfix:adapt to  5B dit model, derive attention_head_dim from config[dim]

* [Fix] Wan2.2 Vae Encode refactor: drop args parameter and use self.cpu_offload

* [Fix] fix bug

* Fix bug

* Update run_wan_vace.sh

* feature: support qwen-image-edit(i2i)

* bugfix

* Add magcache

* fix

* lint

* fix

* update script

* Support cfg infer

* support cfg parallel

* update cfgsp/t2v config

* support magcache calibration

* calibration

* feature: support qwen-image t2i

* pass ci

---------
Co-authored-by: wushuo1 <wushuo1@sensetime.com>

- Updated `ServerConfig` to raise a RuntimeError when no free port is found, providing clearer guidance for configuration adjustments.
- Introduced chunked broadcasting and receiving methods in `DistributedManager` to handle large byte data more efficiently.
- Refactored `broadcast_task_data` and `receive_task_data` methods to utilize the new chunking methods for improved readability and performance.
- Enhanced error logging in `image_utils.py` by replacing print statements with logger warnings.
- Cleaned up the `main.py` file by removing unused signal handling code.