From 27bdcce9b6daee866c4673f43216fdc15bf3da70 Mon Sep 17 00:00:00 2001 From: Xiaoyu Xu Date: Thu, 7 Dec 2023 11:49:25 +0800 Subject: [PATCH] Update README.md (#405) Co-authored-by: Shenghang Tsai --- README.md | 22 ++++++++++++++-------- 1 file changed, 14 insertions(+), 8 deletions(-) diff --git a/README.md b/README.md index 68aa4a73c..787993024 100644 --- a/README.md +++ b/README.md @@ -3,7 +3,7 @@ # OneDiff -A **drop-in acceleration lib** for diffusers and other **diffusion models**. +A **drop-in acceleration lib** for **ComfyUI**, **HF diffusers**, and other diffusion models. ## Performance of OneDiff Community Edition @@ -19,21 +19,27 @@ Updated on Nov 6, 2023. > **_NOTE:_** OneDiff Enterprise Edition delivers even higher performance and second-to-none deployment flexibility. ## Features -- Acceleration for SDXL, SD2.1 and so on +- Acceleration for popular libs + - [ComfyUI](https://github.com/Oneflow-Inc/onediff/tree/main/onediff_comfy_nodes) + - [HF diffusers 🤗](https://github.com/Oneflow-Inc/onediff/tree/main/examples) + - SD Web UI(On the way) +- Acceleration for state-of-the-art Models - [SDXL](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_sdxl.py) - - [SD2.1](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image.py) + - [SD 1.5/2.1](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image.py) - [LoRA](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_sdxl_lora.py) - [ControlNet](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_controlnet.py) - [LCM](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_lcm.py) and [LCM LoRA](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_lcm_lora_sdxl.py) - [SDXL Turbo](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_sdxl_turbo.py) -- Compile PyTorch models (like UNet) with one function - - [oneflow_compile](https://github.com/Oneflow-Inc/onediff/blob/a38c5ea475c07b4527981ec5723ccac083ed0a9c/examples/text_to_image_sdxl.py#L53) -- Enable multi-resolution input with the same compiled model - - [The Multi graph way](https://github.com/Oneflow-Inc/onediff/blob/a38c5ea475c07b4527981ec5723ccac083ed0a9c/examples/text_to_image_sdxl_save_load.py#L65) -- Save/load/change device of the compiled graph +- Drop-in acceleration + - [ComfyUI Nodes](https://github.com/Oneflow-Inc/onediff/tree/main/onediff_comfy_nodes) + - [Acceleration with oneflow_compile](https://github.com/Oneflow-Inc/onediff/blob/a38c5ea475c07b4527981ec5723ccac083ed0a9c/examples/text_to_image_sdxl.py#L53) +- [Multi-resolution input](https://github.com/Oneflow-Inc/onediff/blob/a38c5ea475c07b4527981ec5723ccac083ed0a9c/examples/text_to_image_sdxl_save_load.py#L65) +- Save/load the compiled result, deploy on different environments - [Save and Load](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_sdxl_save_load.py) - [Change device to do multi-process serving](https://github.com/Oneflow-Inc/onediff/blob/main/examples/text_to_image_sdxl_mp_load.py) + + ## Business inquiry on OneDiff Enterprise Edition If you need **unrestricted multiple resolution**, **quantization** support or any other more advanced features, please send an email to caishenghang@oneflow.org . Tell us about your **use case, deployment scale and requirements**!