Search results

A question about the frame and FPS settings in your paper.

Status: Open.

#17 In HumanMLLM/HumanOmniV2;

· wwxs123W opened on Dec 23, 2025

Why not show your model's results on OmniBench?

Status: Open.

#15 In HumanMLLM/HumanOmniV2;

· gengenkai opened on Sep 10, 2025

Request for Preprocessed qa_think.json Files for Daily-Omni and WorldSense

Status: Open.

#14 In HumanMLLM/HumanOmniV2;

· Ruize-Ma opened on Aug 17, 2025

Any code for running a demo？

Status: Open.

#13 In HumanMLLM/HumanOmniV2;

· King-king424 opened on Jul 31, 2025

How to reproduce Qwen2.5-Omni-7B-Thinker's 64.20% Result on IntentBench

Status: Open.

#12 In HumanMLLM/HumanOmniV2;

· shouzhoupeng opened on Jul 22, 2025

data_config/rollout_stage1.json？

Status: Open.

#11 In HumanMLLM/HumanOmniV2;

· shiyongde opened on Jul 18, 2025

Why don't we need to set the user part of the labels to -100 in sft.py?

Status: Open.

#10 In HumanMLLM/HumanOmniV2;

· wyclike opened on Jul 16, 2025

Working with VLLM

Status: Open.

#6 In HumanMLLM/HumanOmniV2;

· qq31415926 opened on Jul 12, 2025

table2

Status: Open.

#4 In HumanMLLM/HumanOmniV2;

· lijain opened on Jul 10, 2025

How much GPU resources are needed to reproduce this work？

Status: Open.

#3 In HumanMLLM/HumanOmniV2;

· vhzy opened on Jul 8, 2025

model is broken with vllm

Status: Open.

#2 In HumanMLLM/HumanOmniV2;

· AlexiaJM opened on Jul 2, 2025