search

SY-Xuan / Pink

Pink: Unveiling the Power of Referential Comprehension for Multi-modal LLMs
72 stars 5 forks source link

Is the environment installation package missing? #3

Closed Jeremy-lf closed 4 months ago

Jeremy-lf commented 4 months ago

你好，我按照你给的环境安装流程进行了安装，但是执行stages1.sh无法运行，环境是否对呢？我把此处的llava改成pink，还是无法运行，感觉环境漏了很多东西，可以列举一下吗 image

Jeremy-lf commented 4 months ago

可以提供一个requirement.txt吗，很多包之间不兼容，自己一个一个安装太费劲了

SY-Xuan commented 4 months ago

Could you provide me with more information?

Jeremy-lf commented 4 months ago

Could you provide me with more information? image like this, i don't konw what package i should install, could you provide your environment for this project? thank you very much!

Jeremy-lf commented 4 months ago

image when i install protobuf, new problem happens

Jeremy-lf commented 4 months ago

image 你好，我现在想训练预训练，按照你的说明，下载了CC-3M 数据集，如1所示，然后配置了stage1.sh脚本，配置如2所示，然后也使用了llama2，在运行脚本的时候，不明白4这一步是在下载什么呢？然后程序挂了，可以帮忙看一下吗，十分感谢

Jeremy-lf commented 4 months ago

image 我的cuda版本

Jeremy-lf commented 4 months ago

image 你好，我现在想训练预训练，按照你的说明，下载了CC-3M 数据集，如1所示，然后配置了stage1.sh脚本，配置如2所示，然后也使用了llama2，在运行脚本的时候，不明白4这一步是在下载什么呢？然后程序挂了，可以帮忙看一下吗，十分感谢

image 找不到问题出在哪里，可以帮忙看一下吗😃😃😃

SY-Xuan commented 4 months ago

Could you give me full log. For now, it is the common output. I cannot give you any useful information. I also think your pytorch version is not correct.

Jeremy-lf commented 4 months ago

Thank your reply, it's problem about cuda and cudnn version not compatible, and it works normal