jyC23333

webUI上怎么进行多轮问答？

2

如题

### System Info I just set up a local tracing server, I change the port to 8005.  when I visit localhost:4173, It shows:  and the error is: ```...

[Usage] How can I change the language model into Qwen-7B?

1

### Describe the issue I want to change the LLM into Qwen，and I write a model file according to llava_llama.py: ```python # Copyright 2023 Haotian Liu # # Licensed under...

[Question] mismatch setting in training stage one.

1

### Question I find that the setting in ```pretrain.sh``` is mismatched with the pap As mentioned in paper, LLM should be freezed, but in the pretrain script, the LLM weights...

Are there some methods to visualize the attention of image encoder in clip?

你好，请问下如何进行SFT的代码会发布吗？

Bug at generate marks

9

### Search before asking - [X] I have searched the Multimodal Maestro [issues](https://github.com/roboflow/multimodal-maestro/issues) and found no similar bug report. ### Bug Traceback (most recent call last): File "/data/megvii/projects/Qwen-VL/scripts/test_maestro.py", line 7,...

Hi，阅读 miniCPM2.5 的技术报告，发现经历的训练阶段是比较多的，故有以下两点问题希望咨询下： 1. 2.6 与 2.5 都是按照技术报告中的三阶段进行训练的吗？ 2. 是否可以提供训练中各阶段的 loss 曲线作为参考？

Where to find multi-image training data?

1

Hi, I can only find a subset of the image from the dataset in [huggingface](https://huggingface.co/datasets/lmms-lab/LLaVA-OneVision-Data) And I don't find multi-image datas in the dataset. for example, tqa is a multi-image...