|
|
--- |
|
|
language: |
|
|
- en |
|
|
base_model: |
|
|
- BAAI/Emu3-Stage1 |
|
|
--- |
|
|
# EARL - SFT think (S) (8B) |
|
|
|
|
|
**Model Size:** 8B parameters |
|
|
**Base Model:** [BAAI/Emu3-Stage1](https://huggingface.co/BAAI/Emu3-Stage1) |
|
|
**Dataset:** Simple Edit |
|
|
**Training Objective:** Supervised Fine-Tuning (SFT) with Chain-of-Thought reasoning |
|
|
|
|
|
This model is introduced in our paper: [EARL: The Promise of RL for Autoregressive Image Editing](https://arxiv.org/abs/2508.01119). |
|
|
|
|
|
## Overview |
|
|
|
|
|
EARL - SFT think (S) is a fine-tuned 8B vision-language model designed for autoregressive image editing. It extends the base Emu3 model with **chain-of-thought supervision**, enabling step-by-step reasoning to tackle complex editing tasks. Training leverages the Simple Edit dataset, focusing on editable instructions grounded in visual understanding. |
|
|
|
|
|
🔗 **Inference script and usage:** [GitHub Repository](https://github.com/saba96/EARL?tab=readme-ov-file) |
|
|
|
|
|
## Benchmark Results |
|
|
|
|
|
| Model | OmniEdit | EmuEdit | AURORA | MB | VisMin | I2EBench | **AVG** | |
|
|
|------------------|----------|---------|--------|------|--------|----------|---------| |
|
|
| **SFT (S)** | 5.73 | 3.66 | 3.58 | 3.19 | 3.57 | 3.59 | **3.88** | |
|
|
| **SFT think (S)** | 4.34 | 3.76 | 2.88 | 3.36 | 3.46 | 3.21 | **3.50** | |
|
|
|
|
|
> ⚠️ Despite integrating reasoning capabilities, the **SFT think** variant underperforms slightly compared to the standard **SFT** model in average benchmark scores. |
|
|
|
|
|
## Intended Use |
|
|
|
|
|
This model is suited for research and development in image editing tasks that benefit from interpretable reasoning, such as instructional or multi-step visual modifications. |
|
|
|
