Image edit · macOS · Windows · Linux

OmniGen2 EditScore

This program edits an image from an instruction, on your GPU. It uses OmniGen2 with the EditScore7B-v1.1 LoRA: OmniGen2 after more training with the EditScore-7B reward model. Download the program for your computer. Each file that you download contains the weights. The program does not connect to other computers.

macOS

Apple silicon · 16.3 GB

Download for macOS

Apple did not notarize this program. Unzip the file. Remove the quarantine flag one time. Then start the program:

cd omnigen2-editscore-macos-arm64
xattr -dr com.apple.quarantine .
./omnigen2-editscore

SHA-256 9940944c2a2b87213a1e93263eef8398afa3a2470aa6668a0d331caadcd1f818

Windows

x86-64 · NVIDIA · 19.0 GB

Download for Windows

Unzip the file. Start omnigen2-editscore.cmd.

SHA-256 2735d229f3a75ad7694008fe3641c9f5ab39686ed6f19d9d212c6322e47c6a54

Linux

x86-64 · NVIDIA · 20.0 GB

Download for Linux

mkdir omnigen2-editscore
tar xf omnigen2-editscore-linux-x86_64.tar -C omnigen2-editscore
cd omnigen2-editscore
./omnigen2-editscore

SHA-256 efa4aca07d0aa7612e9158c98bbecafeedfab65887d3ec3a5daee9ef59311296

Weights

In each desktop program · 15.9 GB

The folder w of each desktop program contains the weights: OmniGen2 from chibifire/omnigen2-base-df5dca8a in BF16, and the LoRA from OmniGen2-EditScore7B-v1.1. The program does not download files.

All checksums: SHA256SUMS. The files are in chibifire/omnigen2-editscore-desktop.

Edit an image

Give an image and an instruction

In the browser

Start the program with no option. The first start unpacks the Python environment into the folder env. Then the program loads the model and opens the edit page in your browser. Give an image and an instruction, then click Edit. The page and the edit stay on this computer.

From a terminal

Edit one image and save the result as a PNG file.

omnigen2-editscore --image in.png \
  --prompt "Change the background to a classroom." \
  --out edit.png

--steps N	Steps of the sampler (default 50)
--seed N	Seed of the noise (default 0)
--max-side N	Longest side of the result in pixels (default 1024)
--no-lora	Use OmniGen2 without the LoRA, to compare
--device	auto, cuda, mps or cpu

Example

Change the background to classroom

Edit by OmniGen2 with the EditScore7B-v1.1 LoRA

Linux package, NVIDIA RTX 3090, 50 steps, seed 0, 672 × 1024 pixels, approximately 2 minutes for each edit. The input image is from the OmniGen2 repository.

Status

Test results

Linux package on an NVIDIA RTX 3090 (Linux in WSL), with no connection to other computers. An edit was complete after 126 seconds (50 steps, 672 × 1024 pixels).
Windows package on an NVIDIA RTX 3090. We stopped all connections from the program to other computers. An edit from the terminal was complete after 125 seconds. We gave the image and the instruction on the edit page in Microsoft Edge, then clicked Edit. The page showed the edit after 125 seconds, and the browser loaded all 62 files of the page from this computer.
macOS on an Apple M2 Pro with 32 GB of memory. We used the program and the weights of the macOS package. The same edit was complete after 66 minutes. The computer used its disk for some of the data. We did not stop the connections on macOS.
The edits with the BF16 weights are almost the same as the edits with the FP32 weights of the previous version. Two edits with the same weights and the same seed also have small differences on this GPU. The differences of the BF16 edits are not larger.

Credits

Open source parts

OmniGen2 by VectorSpaceLab (Apache License 2.0) is the image model and its code. The weights are OmniGen2/OmniGen2 from chibifire/omnigen2-base-df5dca8a, stored in BF16. The program operated in BF16 before this change too.
OmniGen2-EditScore7B-v1.1 by VectorSpaceLab (Apache License 2.0) is the LoRA. EditScore (Apache License 2.0) is the reward model of its training.
Qwen2.5-VL-3B-Instruct by the Qwen team (Apache License 2.0) is the part of OmniGen2 that reads the image and the instruction.
PyTorch, Diffusers, Transformers and Gradio operate the model and the edit page. They have open source licenses. pixi-pack (BSD 3-Clause License) makes the Python environment.