GitHub · Python
Jit-Roy/WeeLLM
WeeLLM runs large diffusion models with as little as 4 GB of VRAM, without any quantization. It dynamically determines how many layers can fit within the available VRAM and streams the text encoder and transformer layers to the GPU layer by layer, enabling inference on hardware with limited VRAM. It supports both safetensors and GGUF models.
- Stars
- 158
- Forks
- 13
- Stars gained in 24 hours
- 0
- Forks gained in 24 hours
- 0
Topics: diffusion, ernie-image, flux, flux2, flux2-klein, hidream, hidream-o1, ideogram-4, krea2, longcat
Latest data point: Nov 21, 2026 · 158 stars · 13 forks