Onediff

- switch image size will not trigger re-compilation(i.e. no time cost);
- faster to save and load a graph;
- smaller static memory needed;
- much easier to switch graphs;

Note
1. After updating onediff to 0.12, oneflow also needs to be updated to the latest with https://github.com/siliconflow/onediff?tab=readme-ov-file#1-install-oneflow
2. If you are using save/load graph. Please use this new version to save the graph again.
3. For SVD(Stable Video Diffusion) onediff users, please use os.environ["VM_REBUILD_DYNAMIC_SHAPE"] = "1" to support DYNAMIC SHAPE with SVD. There has some re-compile cost, when enabling VM_REBUILD_DYNAMIC_SHAPE, so loading graph is needed. We will do more work on this.

OneDiff SDXL inference has been integrated into fal.ai playground
SDXL inference at the speed of thought: https://www.fal.ai/models/onediff-sdxl

And OneDiff ComfyUI nodes are in ComfyUI-Manager Now

- Release new performance data on SDXL.
- Support SD web UI SD and SDXL acceleration.
- Add SVD/DeepCache diffusers pipeline and optimize performance.

What's Changed
* Fix int8 lora by doombeaker in https://github.com/siliconflow/onediff/pull/431
* Fix deepcache for svd by lixiang007666 in https://github.com/siliconflow/onediff/pull/438
* Update onediff_comfy_nodes README by ccssu in https://github.com/siliconflow/onediff/pull/439
* Update README.md by eltociear in https://github.com/siliconflow/onediff/pull/440
* Fix diffusers 0 25 dev by ccssu in https://github.com/siliconflow/onediff/pull/444
* add_deep_cache_with_lora_workflow by clackhan in https://github.com/siliconflow/onediff/pull/442
* fix reamde typos by doombeaker in https://github.com/siliconflow/onediff/pull/455
* support mode parameter in PyTorch 2.1 by haoyang9804 in https://github.com/siliconflow/onediff/pull/458
* use fused_multi_head_attention_inference_v2 to implement CrossAttention by fpzh2011 in https://github.com/siliconflow/onediff/pull/459
* Add OneDiffCheckpointLoader by ccssu in https://github.com/siliconflow/onediff/pull/457
* Add onediffcheckpointloadersimple refine by ccssu in https://github.com/siliconflow/onediff/pull/460
* Fix SVD VAE compile for diffuser by lixiang007666 in https://github.com/siliconflow/onediff/pull/461
* Feature/svd support by chengzeyi in https://github.com/siliconflow/onediff/pull/454
* Update README.md for SVD by lixiang007666 in https://github.com/siliconflow/onediff/pull/462
* Revert "Fix SVD VAE compile for diffuser" by hjchen2 in https://github.com/siliconflow/onediff/pull/464
* Integrate deep cache diffusers pipeline by clackhan in https://github.com/siliconflow/onediff/pull/450
* support v1-5-pruned model in sd-webui by fpzh2011 in https://github.com/siliconflow/onediff/pull/463
* use object.__setattr__ in DualModule and DeployableModule to avoid unexpected key in self._modules by marigoold in https://github.com/siliconflow/onediff/pull/468
* fix an op convertion error by haoyang9804 in https://github.com/siliconflow/onediff/pull/469
* Quick fix vae load by strint in https://github.com/siliconflow/onediff/pull/470
* add perf chart by strint in https://github.com/siliconflow/onediff/pull/474
* Revert "use object.__setattr__ in DualModule and DeployableModule to avoid unexpected key in self._modules" by marigoold in https://github.com/siliconflow/onediff/pull/476
* update_onediff_comfy_nodes_README by ccssu in https://github.com/siliconflow/onediff/pull/477
* Sd Web Unet Model Quantization and Black Formatting by ccssu in https://github.com/siliconflow/onediff/pull/445
* Add a notebook showing how to save graph locally and load from it by kopyl in https://github.com/siliconflow/onediff/pull/475
* prefer cudnn convolution by chengzeyi in https://github.com/siliconflow/onediff/pull/480
* Fix lora parameter update by ccssu in https://github.com/siliconflow/onediff/pull/479
* Migrate to new CI deployments by jackalcooper in https://github.com/siliconflow/onediff/pull/482
* enable dual gemm glu by chengzeyi in https://github.com/siliconflow/onediff/pull/481
* Opt deep cache sdxl pipeline by clackhan in https://github.com/siliconflow/onediff/pull/485
* Change default env val by clackhan in https://github.com/siliconflow/onediff/pull/487
* add vae compile to sd-webui by fpzh2011 in https://github.com/siliconflow/onediff/pull/473
* fix read me of webui by fpzh2011 in https://github.com/siliconflow/onediff/pull/489
* Update README.md for connection by strint in https://github.com/siliconflow/onediff/pull/492

New Contributors
* eltociear made their first contribution in https://github.com/siliconflow/onediff/pull/440
* haoyang9804 made their first contribution in https://github.com/siliconflow/onediff/pull/458
* chengzeyi made their first contribution in https://github.com/siliconflow/onediff/pull/454
* kopyl made their first contribution in https://github.com/siliconflow/onediff/pull/475

**Full Changelog**: https://github.com/siliconflow/onediff/compare/0.11.3...0.11.4

What's Changed
* explain quantization feature by doombeaker in https://github.com/siliconflow/onediff/pull/402
* Update_onediff_comfy_nodes_readme by ccssu in https://github.com/siliconflow/onediff/pull/404
* Update README.md by strint in https://github.com/siliconflow/onediff/pull/405
* Add SVD compile node by lixiang007666 in https://github.com/siliconflow/onediff/pull/409
* Dev support controlnet in onediff comfy nodes by ccssu in https://github.com/siliconflow/onediff/pull/403
* remove rewrite_self_attention by hjchen2 in https://github.com/siliconflow/onediff/pull/412
* Fuse LoRA in examples/text_to_image_sdxl_lora.py by marigoold in https://github.com/siliconflow/onediff/pull/386
* all dynamic shape by strint in https://github.com/siliconflow/onediff/pull/411
* add image to image vae compile demo by strint in https://github.com/siliconflow/onediff/pull/406
* add comfyui quantization tool node by doombeaker in https://github.com/siliconflow/onediff/pull/413
* Run tests regularly by jackalcooper in https://github.com/siliconflow/onediff/pull/417
* add sd-webui extension by fpzh2011 in https://github.com/siliconflow/onediff/pull/401
* Compatibility with the latest ComfyUI update (commit b454a67) by ccssu in https://github.com/siliconflow/onediff/pull/418
* Update README.md by strint in https://github.com/siliconflow/onediff/pull/419
* add context manager `TensorInplaceAssign` by marigoold in https://github.com/siliconflow/onediff/pull/396
* add sd-webui performance data; fix refiner model bug by fpzh2011 in https://github.com/siliconflow/onediff/pull/420
* Add badge for regular tests by jackalcooper in https://github.com/siliconflow/onediff/pull/421
* add mock class for specific workflow by doombeaker in https://github.com/siliconflow/onediff/pull/416
* update benchmark scripts by hjchen2 in https://github.com/siliconflow/onediff/pull/425
* add oss model url by doombeaker in https://github.com/siliconflow/onediff/pull/429
* update benchmark readme by hjchen2 in https://github.com/siliconflow/onediff/pull/430
* dev_optimize_mock_torch by ccssu in https://github.com/siliconflow/onediff/pull/410
* Integrate deep cache by clackhan in https://github.com/siliconflow/onediff/pull/426


**Full Changelog**: https://github.com/siliconflow/onediff/compare/0.11.2...0.11.3

Speedup for ComfyUI, ControlNet, and SDXL Turbo.

- Add more examples.
- Some code refactoring and bug fixes.