VisionLLM

VisionLLM copied to clipboard

vllmv2 errors

4

``` ./install.sh: line 46: cd: apex: No such file or directory bash: compile.sh: No such file or directory ./install.sh: line 51: cd: visionllmv2/model/unipose/ops: No such file or directory python3: can't...

matbeedotcom

VisionLLM v2 finetune code？

7

When will the fine-tuning code be released? 什么时候开放微调代码？

shiyongde

How to get all object super link of $$H_{link}$$ in last-layer hidden states during testing ?

1

I have read [issue 11](https://github.com/OpenGVLab/VisionLLM/issues/11), but still have some questions. During testing, you utilize the forward function for testing https://github.com/OpenGVLab/VisionLLM/blob/028f8b3d8a8f7d41edef5c0046e8bc49e25e7144/VisionLLMv2/visionllmv2/eval/eval_det.py#L119 But as far as I know, the forward function cannot...

zc-zhao

Hi, I'm reading your paper VisionLLM: Large Language Model is also an Open-Ended Decoder for Vision-Centric Tasks, and there are some detail implementations I want to dive into. I only...

minhhb9x

Hey~We are the team of '自动驾驶之心'. Currently, we have accumulated a large number of fans in the fields of autonomous driving perception, positioning, planning and prediction, deployment and acceleration, hardware...

autodriving-heart

Hi @OpenGVLab 🤗 Niels here from the open-source team at Hugging Face. I discovered your work through Hugging Face's daily papers as yours got featured: https://huggingface.co/papers/2305.11175. The paper page lets...

NielsRogge

作者您好，很有兴读到了这么好的文章，请问能分享一下路由链接查询为8的权重吗？想用一下其中的路由查询链接的emb的权重，我们跳过了一阶段，直接二阶段开始训练8token随即初始化之后发现效果比用4token的权重效果差一些，因为想请问是否可以分享一下8token的权重，谢谢！

Lelege0

Fine-tuning VisionLLMv2

1

Hi authors~ Thanks for your good work! Could you provide some fine-tuning instructions and data preparation guidance?

guangqian-guo

作者，您好，请问您是否可以分享下DUT-OMRON/DUTS_coco_style.json 这个文件呢？

qduOliver