One LoRA, three media: Odin Lovis' Three.js Conf talk on image, video and 3D generation

OdinLovis · x · 2026-09-11

A retweeted shoutout to Odin Lovis' talk at Three.js Conf on using LoRAs to drive image, video, and 3D model generation — a cross-media approach to lightweight fine-tuning that attendees found fascinating. The talk is hosted via fal.ai.

Original post →

More from Multimodal

Multimodal channel →