Search results

Higher memory usage on GPU0 than other GPUs when finetuning univla-7b on Libero datasets

Status: Open.

#79 In OpenDriveLab/UniVLA;

· PuzhenYuan opened on Mar 31, 2026

ENV problem

Status: Open.

#78 In OpenDriveLab/UniVLA;

· DingdongD opened on Jan 24, 2026

stage 1 loss

Status: Open.

#77 In OpenDriveLab/UniVLA;

· waynechu1021 opened on Jan 18, 2026

Regarding deployment on a real robotic arm

Status: Open.

#76 In OpenDriveLab/UniVLA;

· zdq2002 opened on Jan 13, 2026

self.dino_encoder = torch.hub.load('facebookresearch/dinov2', 'dinov2_vitb14_reg')

Status: Open.

#75 In OpenDriveLab/UniVLA;

· 2561681244 opened on Jan 9, 2026

Question about calvin inference

Status: Open.

#73 In OpenDriveLab/UniVLA;

· Hitjss opened on Dec 24, 2025

Possible mislabeling in Fig. 8: LIBERO example in Group B appears to depict a pick-up action rather than a place action

Status: Open.

#69 In OpenDriveLab/UniVLA;

· senp98 opened on Nov 17, 2025

A question about the process of action in BridgedataV2

Status: Open.

#68 In OpenDriveLab/UniVLA;

· xianzeYao opened on Oct 28, 2025

Question about finetune code for qwbu/univla-7b-224-sft-simpler-bridge

Status: Open.

#67 In OpenDriveLab/UniVLA;

· minghaoguo20 opened on Oct 24, 2025

About exploiting wrist view camera

Status: Open.

#65 In OpenDriveLab/UniVLA;

· Eddiesyn opened on Sep 24, 2025

Finetuning on LIBERO using pretrained checkpoint with bridgev2-only dataset

Status: Open.

#64 In OpenDriveLab/UniVLA;

· feiyu12138 opened on Sep 11, 2025

Loss instability when resuming pretraining from converted HF weights

Status: Open.

#60 In OpenDriveLab/UniVLA;

· minghaoguo20 opened on Aug 20, 2025