A method called MTEdit proposes using packed trajectories instead of sequential rollouts to build high-quality multi-turn image editing training data, according to the research. The work addresses the difficulty of constructing such data, where conventional sequential rollout accumulates errors across turns. The approach targets better alignment with human-AI collaborative creation than single-turn editing.
Multi-turn image editing aims to modify an image according to a sequence of instructions, making it better aligned with human-AI collaborative creation than single-turn editing. However, constructing high-quality multi-turn training data remains challenging, as conventional sequential rollout accumu
EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.