Custom Model#

LLM Model Evaluation#

Evaluation of large language models currently supports two methods: Native Evaluation and OpenCompass Evaluation.

Method 1: Using Native Evaluation#

See also

See Basic Usage Guide

Method 2: Using OpenCompass Evaluation#

See also

See OpenCompass Usage Guide

VLM Model Evaluation#

Currently, evaluation of multimodal models only supports the VLMEvalKit backend.

Using VLMEvalKit Backend#

See also

See VLMEvalKit Usage Guide