cloneofsimo/lora
View on GitHub ↗ ★ 7,557 Jupyter Notebook
Using Low-rank adaptation to quickly fine-tune diffusion models.
6 mentions on Hacker News by 6 people: 1 story, 5 comments · first 2023-01-02 · latest 2024-03-08
20082026
-
…has an "extended" version nowadays that applies to the resnet part as well as the cross-attention: https://github.com/cloneofsimo/lora
comment by jph00 · 2024-03-08 · in Fine tune a 70B language model at home -
…You can clearly see that they've been using various model merges (this Lora merging techniques like https://github.com/cloneofsimo/lora ) to get two different models to combine and get the best of both. Many have done better hands and…
comment by chii · 2023-03-16 · in Midjourney v5 can do hands -
…and other improvements from the SD community really turn this into a super capable toolchain. [0] https://github.com/cloneofsimo/lora
comment by danielbln · 2023-03-02 · in Creating Isometric RPG Game Backgrounds -
Try the Lora fine-tuning method with few MB results: https://github.com/cloneofsimo/lora
comment by spyder · 2023-01-15 · in Show HN: Create your own video clips with Stable Diffusion -
…finetuning the weight deltas to achieve the look, faster than dreambooth, much smaller output. 1: https://github.com/cloneofsimo/lora
comment by ftufek · 2023-01-02 · in Remaking old computer graphics with AI image generation