Search Results for author:
Found 3 papers, 3 papers with code
T-Eval: Evaluating the Tool Utilization Capability of Large Language Models Step by Step
Based on that, we further introduce T-Eval to evaluate the tool utilization capability step by step.
MultiModal-GPT: A Vision and Language Model for Dialogue with Humans
To further enhance the ability to chat with humans of the MultiModal-GPT, we utilize language-only instruction-following data to train the MultiModal-GPT jointly.
Deliberated Domain Bridging for Domain Adaptive Semantic Segmentation
In this work, we resort to data mixing to establish a deliberated domain bridging (DDB) for DASS, through which the joint distributions of source and target domains are aligned and interacted with each in the intermediate space.
Ranked #1 on
Domain Adaptation
on GTAV+Synscapes to Cityscapes
