search

IuvenisSapiens / ComfyUI_Qwen2-VL-Instruct

The successful integration of Qwen2-VL-Instruct into the ComfyUI platform has enabled a smooth operation, supporting (but not limited to) text-based queries, video queries, single-image queries, and multi-image queries for generating captions or responses.
Apache License 2.0
67 stars 8 forks source link

can only concatenate str (not "list") to str #9

Open ALFAMAN69 opened 2 months ago

ALFAMAN69 commented 2 months ago

ComfyUI Error Report

Error Details

javacoffer2020 commented 2 months ago

The input and output of this node are lists, probably to match the video interpretation function. To interpret an image, you need to use the ‘Load Image Advanced’ node for the input, and then go through the ‘Multiple Paths Input’ node to select the corresponding serial number. The output is a text list. We haven't found any node that can convert this format (‘String List to String’ node doesn't work). So, there is only a manual solution. Connect a node for displaying texts to the output of the ‘Qwen2 VQA’ node, copy the contents of the node to a text input node, and then use this text input node to link to T5xxl.

ALFAMAN69 commented 2 months ago

Thank you very mach!

alexcong commented 2 months ago

The output of the Qwen VQA node is a tuple. Changing to a list will be much easier to work with. In nodes.py the last line return (result,) can be changed to return result and the output can be used by other nodes.

@IuvenisSapiens Can you change this? Thanks