Implementation:Open compass VLMEvalKit TeleMM2Thinking Wrapper

Field	Value
source	VLMEvalKit
domain	Vision, API_Integration

Overview

TeleMM2Thinking_Wrapper provides a VLMEvalKit API adapter for the TeleMM2 Thinking vision-language model from TeleAGI.

Description

TeleMM2Thinking_Wrapper inherits from BaseAPI and communicates with the TeleAGI streaming chat completions endpoint. It uses a custom prompt utility (TeleMM2Thinking_PromptUtil) for dataset-specific prompt construction with chain-of-thought reasoning and boxed answer formatting. Compared to the standard TeleMM2_Wrapper, this variant is specifically designed for the thinking-enhanced model. Authentication uses AUTHOR and APP_ID environment variables.

Usage

Use this adapter when evaluating TeleMM2 Thinking vision-language models with enhanced reasoning capabilities through the TeleAGI API platform.

Code Reference

Source: vlmeval/api/telemm_thinking.py, Lines: L1-294
Import: from vlmeval.api.telemm_thinking import TeleMM2Thinking_Wrapper

Signature:

class TeleMM2Thinking_Wrapper(BaseAPI):
    def __init__(self, model=None, retry=3, verbose=True, timeout=600,
                 authorization=None, app_id=None, system_prompt=None,
                 max_tokens=20480, **kwargs): ...
    def generate_inner(self, inputs, **kwargs): ...

I/O Contract

Direction	Description
Inputs	message — text/image/video content list; model-specific params via kwargs
Outputs	generate() returns str prediction; generate_inner() returns (int, str, str) tuple

Usage Examples

# Example instantiation
model = TeleMM2Thinking_Wrapper(model='telemm2-thinking')
response = model.generate(message)

Related Pages

Page Connections

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

Principle

Implementation

Heuristic

Environment