Legacy DS to LLM: Use banking77 dataset, SFT, and GRPO
JoshPurtell · x · 2026-09-02
For data scientists with traditional ML backgrounds (e.g., XGBoost, statistics), a good bridge to LLMs is analyzing the classic classification task banking77 to build evals. The advice includes using SFT and GRPO to boost performance, referencing OpenAI Cookbooks, and experimenting with various models.
Related event: How Veteran ML Engineers Can Transition to LLMs(4 posts)→
More from Research
- Gemini-powered Spatial OS turns 8 photos into 3D digital twins — AI_Andrew · 2026-09-02
- Paper: AI Agents Push Humans Out of the Loop — mmitchell_ai · 2026-09-02
- Alignment Journal invites unpublished AI alignment preprints for submission — dhadfieldmenell · 2026-09-02
- Epoch Index suggests AI capabilities progress twice as fast with reasoning models — Jsevillamol · 2026-09-02
- Study Finds LLM Embeddings Predict Brain Language Response Better Than Simple Metrics — begusgasper · 2026-09-02
- Study Finds AI Interprets Bone Age X-Rays Better Than Clinicians — zakkohane · 2026-09-02