Flan-T5 Large

Flan-T5 Large

Quick answer: Flan-T5 Large is Google's 780M-parameter instruction-tuned encoder-decoder model from the "Scaling Instruction-Finetuned Language Models" paper (Chung et al., 2022). It is open-weight (Apache 2.0) and used as a mid-size fine-tunable backbone in agent and grounding research, including as a MindAct backbone in the Mind2Web benchmark.

At a Glance

Historical significance

• Mid-size tier of the Flan-T5 family, balancing fine-tuning cost against capability
• Widely cited as a fine-tuning baseline in web-agent and grounding papers from 2022-2023

Modern context

• Long surpassed in raw capability by modern decoder-only LLMs
• Still used as a lightweight, fully open fine-tuning backbone in academic benchmarks and ablations