Implementation:Open compass VLMEvalKit WeThinkVL

Field	Value
source	VLMEvalKit
domain	Vision, Model_Architecture

Overview

VLM adapter for the WeThink-VL model enabling benchmark evaluation in VLMEvalKit.

Description

WeThinkVL inherits from BaseModel and wraps the WeThink-VL model for use within the VLMEvalKit evaluation framework. It initializes the model and tokenizer/processor from a HuggingFace model path (default: model_path (required)) and provides the generate_inner method for inference. Inherits from both Qwen2VLPromptMixin and BaseModel, supporting chain-of-thought reasoning with answer tag extraction.

Usage

Register in vlmeval/config.py via supported_VLM and invoke through the standard evaluation pipeline.

Code Reference

Source: vlmeval/vlm/wethink_vl.py, Lines: L1-172
Import: from vlmeval.vlm.wethink_vl import WeThinkVL

Signature:

class WeThinkVL(BaseModel):
    INSTALL_REQ = False
    INTERLEAVE = True
    def __init__(self, model_path='model_path (required)', **kwargs): ...
    def generate_inner(self, message, dataset=None): ...

I/O Contract

Direction	Description
Inputs	message — list of dicts with type (text/image) and value; dataset — optional dataset name for custom prompting
Outputs	generate_inner() returns str (model response text)

Usage Examples

from vlmeval.vlm.wethink_vl import WeThinkVL
model = WeThinkVL(model_path='path/to/model')
response = model.generate_inner(message)

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment