π0.5 finetuning on real manufacturing: 1 hour of clean data beat the previous 17
DominiqueCAPaul · x · 2026-09-24
A two-month ablation of π0.5 finetuning on a real manufacturing task, with all data and runs published. Final policy success rate: 98%.
Key findings:
- Pure scaling was the weakest lever: 5x more data only moved success from 63% to 76%.
- Diversity paid: 4h spread across five scenes beat 4h in the eval scene by 30pp.
- Quality crushed quantity: adding 1 clean, slower hour on top of 21h went 76% → 90% — that single hour added more than the previous 17.
- Quality alone: 1h clean data + 240 human-intervention rollouts (1.7h total) went 28% → 88%, beating the 21h model by 12 points with a twelfth of the data.
- Inference settings added more without retraining: 21h model 76% → 93%; 21h+1h model 90% → 98%.
Surprisingly unimportant: learning rate, batch size, relative vs. absolute joints, image augmentation.
More from Embodied
- Homebrew robot arm build log Day 7: geartrain CAD, wiring and gripper sizing — _Stocko_ · 2026-11-11
- Perceptron AI launches EgoHome, an open benchmark for sub-second egocentric hand understanding — rohanpaul_ai · 2026-11-11
- Runway open-sources Praxis-1, turning video pretraining into real robot control — runwayml · 2026-11-11
- Unitree G1 humanoid wearing VR goggles pilots an A1 drone in land-air robot demo — CyberRobooo · 2026-11-11
- CoRL Spotlight Meta-Ctrl predicts robot plan failures, Llama 3 8B beats GPT-5.5 on VirtualHome — guyvdb · 2026-11-11
- Luna, a jewelry-style AI wearable that speaks before you ask, launches on Kickstarter — nikola_mr64990 · 2026-11-11