Search results

can it support realtime inference?

Status: Open.

#22 In HumanMLLM/HumanOmni;

· ziyanxzy opened on Dec 15, 2025

能不能做个国产化支持 310p芯片加速推理

Status: Open.

#21 In HumanMLLM/HumanOmni;

· Zha-Miku opened on Sep 3, 2025

模型的指令跟随性比较差，输入video时几乎完全不听text的指令

Status: Open.

#20 In HumanMLLM/HumanOmni;

· Yuxuan2003 opened on Jul 1, 2025

fine-tuned

Status: Open.

#19 In HumanMLLM/HumanOmni;

· Mr-Millionairelive opened on May 20, 2025

推理时使用HumanOmni-Video的权重文件，只进行视觉推理出错

Status: Open.

#18 In HumanMLLM/HumanOmni;

· Shawnie-qi opened on May 18, 2025

Memory leak problem

Status: Open.

#17 In HumanMLLM/HumanOmni;

· WorkHaH opened on May 15, 2025

One unexpected output when running demo on chinese_video.mp4

Status: Open.

#16 In HumanMLLM/HumanOmni;

· YuanEZhou opened on Apr 27, 2025

LoRA finetuning fails

Status: Open.

#15 In HumanMLLM/HumanOmni;

· sefinch opened on Apr 24, 2025

Some weights of the model checkpoint were not used when initializing HumanOmniQwen2ForCausalLM

Status: Open.

#14 In HumanMLLM/HumanOmni;

· Lin-design13 opened on Apr 22, 2025

size mismatch for weight

Status: Open.

#13 In HumanMLLM/HumanOmni;

· cristiano-xw opened on Apr 9, 2025

微调所需硬件配置

Status: Open.

#12 In HumanMLLM/HumanOmni;

· WorkHaH opened on Apr 8, 2025

face feature in inference

Status: Open.

#11 In HumanMLLM/HumanOmni;

· sunwoo76 opened on Apr 7, 2025