raiseLab

Setup

Model Selection

Frontier & Commercial Models (API)

Note the api code block not added yet

• GPT-4o
• Claude 3.5 Sonnet

Open-Weight Models

• Llama 3.1 (8B & 70B)
• Mistral 7B
• Gemma 2

The GENIE paper already gives us working datasets for four exact task types.

Links to dataset in links.text

• ARC-DA → your QA task baseline
• αNLG → your Commonsense Reasoning task
• WMT19/21 DE-EN → your Machine Translation task (Some dataset issue)
• XSUM → your Summarization task