Modelpublic

kimi000/silver-orbit-62

sourceHugging Faceotherupdated 29d agoView on Hugging Face

0likes14downloads

Model Card

FLUX.2 Klein Base 4B AlphaGRPO checkpoint

This is a complete native Diffusers Flux2KleinPipeline. The EMA LoRA is already merged into the transformer, so FAR and PEFT are not required for inference.

Source experiment: flux2_klein_base_4b_diffusionnft_alphagrpo_dvreward_alphagrpo20k_512px_20step_cfg4_16prompts_group14_7train_1dvreward_tp1_2node_nrt

Source checkpoint: step_600.pt

Training profile: 512px, 20 rollout steps, CFG 4, AlphaGRPO DVReward.

bash

python demo.py --prompt "A red cube beside a blue glass sphere."