Max_length in lora training

PixArt-alpha / PixArt-sigma

PixArt-Σ: Weak-to-Strong Training of Diffusion Transformer for 4K Text-to-Image Generation

https://pixart-alpha.github.io/PixArt-sigma-project/

GNU Affero General Public License v3.0

1.44k stars 68 forks source link

Max_length in lora training #67

Closed berryxue closed 2 months ago

berryxue commented 2 months ago

In train_pixart_lora_hf.py, Line 456, max_length=120. This means that when using lora to train sigma, can max_lenght only be 120, or can it be changed?

lawrence-cj commented 2 months ago

It can be changed