[Bug] 请问如何使用自己的数据集进行lmdeploy的awq量化？量化的校对数据集只支持文字问答数据集吗？

[lzk9508]

Checklist

• 1. I have searched related issues but cannot get the expected help.
• 2. The bug has not been fixed in the latest version.
• 3. Please note that if the bug-related issue you submitted lacks corresponding environment info and a minimal reproducible demo, it will be challenging for us to reproduce and resolve the issue, reducing the likelihood of receiving feedback.

Describe the bug

我需要量化的是一个多模态的internvl，感觉官方的校对数据集量化效果一般

Reproduction

awq量化

Environment

centos7,pytorch 2.1.0+cuda121,lmdeploy0.6.4

Error traceback

No response

[AllentDan]

May modify https://github.com/InternLM/lmdeploy/blob/v0.6.4/lmdeploy/lite/utils/calib_dataloader.py#L304 and https://github.com/InternLM/lmdeploy/blob/v0.6.4/lmdeploy/lite/apis/calibrate.py#L311 to create customized input tensors. Current codes only support token ids as inputs.