VARGPT-v1.1: Improve Visual Autoregressive Large Unified Model via Iterative Instruction Tuning and Reinforcement Learning

This model was presented in the paper VARGPT-v1.1: Improve Visual Autoregressive Large Unified Model via Iterative Instruction Tuning and Reinforcement Learning. The project page can be found here: https://vargpt1-1.github.io/

VARGPT-v1.1 (7B+2B) modeling understanding and generation as two distinct paradigms within a unified model: predicting the next token for visual understanding and predicting the next scale for visual generation.

We provide the simple generation process for using our model. For more details, you could refer to Github: VARGPT-v1.1.

Visual Editing

  • You can perform batch evaluations, including evaluations of image generation and image editing:
cd VARGPT-family-training
bash run_scripts/run_eval_vargpt_v1_1_edit.sh

For more details, you could refer to Github: VARGPT-v1.1.

Downloads last month
64
Safetensors
Model size
10.9B params
Tensor type
BF16
·
I64
·
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support