Lucas Moreau
Open-weight LLM fine-tuning and private deployment

Lucas Moreau

France
0 (0 Reviews)

About Me

LoRA and full fine-tunes, evaluation harnesses you can trust, and deployment on your own GPUs or VPC. You keep the weights and the data.

Skills

No skills added yet.

My Services

No services added yet.

Portfolio

A 7B model that beat GPT-4 on one narrow task
March 2026

A 7B model that beat GPT-4 on one narrow task

Curated 14k in-domain examples, fine-tuned Mistral-7B and shipped it on two A10s at a twentieth of the API cost.

Created
Mar 5, 2026