How TDM‑R1 Boosts Few‑Step Image Generation: GenEval Jumps from 61% to 92% and Beats GPT‑4o

The TDM‑R1 framework introduces a two‑stage reinforcement learning pipeline that lets 4‑step diffusion models achieve a GenEval score of 92%, surpassing 80‑step baselines and GPT‑4o while also fixing instruction compliance, text rendering, and compositional generation issues.

GenEvalOCR improvementTDM-R1

0 likes · 15 min read

How TDM‑R1 Boosts Few‑Step Image Generation: GenEval Jumps from 61% to 92% and Beats GPT‑4o