Full-Weight VLA Fine-Tuning Without an fp32 Master Copy Silently Erases Most Updates
Practical warning: fine-tuning VLA weights in bf16 without an fp32 master copy rounds away AdamW updates below 1/512 of a weight; finds this recipe in 9 of 12 open VLA codebases and 79%+ of updates lost on LIBERO.实用警示:VLA 权重以 bf16 微调且不保留 fp32 主副本时,小于权重 1/512 的 AdamW 更新会被舍入抹掉;报告 12 个开源 VLA 代码库中 9 个的默认或文档配方如此训练,LIBERO 实验中至少 79% 的更新被抹除。
BibTeX
@misc{singh2026full,
title = {{Full-Weight VLA Fine-Tuning Without an fp32 Master Copy Silently Erases Most Updates}},
author = {Shubham Singh},
year = {2026},
eprint = {2610.31859},
archivePrefix = {arXiv},
primaryClass = {cs.RO},
url = {https://arxiv.org/abs/2610.31859},
}VLA & Robot Foundation ModelsVLA 与机器人基础模型Robot Learning: RL & Imitation机器人学习:强化学习与模仿学习fine-tuningtraining-recipemixed-precisionsim