HF engineers train a 9B model to build 3D rooms via Blender code

Hugging Face engineer Merve Noyann used TRL and OpenEnv to build a Blender RL environment and trained Qwen3.5-9B to write bpy code via tool calls, constructing low-poly 3D rooms from scratch using GRPO and GEPA, sharing rollouts and lessons learned.

2026-09-07 ~ 2026-09-08 · 3 related posts

• HF engineer trains Qwen3.5-9B with GRPO and GEPA to code Blender 3D rooms, GLM-5.3-Flash as judge — mervenoyann · 2026-09-07
• HF's merve shares GEPA agent-training rollouts built on TRL and OpenEnv — mervenoyann · 2026-09-07
• Training a 9B model with GRPO to build low-poly Blender rooms: lessons learned — TheMoonMidas · 2026-09-08