Clybius / Learned-RoundingLinks
A repository of Python & PyTorch scripts which (currently) converts .safetensors models into scaled FP8 variants, utilizing gradient descent for optimal rounding.
☆27Updated 5 months ago
Alternatives and similar repositories for Learned-Rounding
Users that are interested in Learned-Rounding are comparing it to the libraries listed below
Sorting:
- ☆93Updated 9 months ago
- flux distillation and stuff☆124Updated 7 months ago
- Executable State Dict Recipes☆79Updated 2 weeks ago
- A UI made in Pyside6 to make training LoRA/LoCon and other LoRA type models in sd-scripts easy☆45Updated 2 weeks ago
- qwen-image and wan2.2 wan2.1 bridge☆174Updated last month
- ComfyUI extension for mixing model during sampling☆30Updated 4 months ago
- ComfyUI implementation of Long-CLIP☆163Updated 10 months ago
- The official code that integrates MagCache (Fast Video Generation with Magnitude-Aware Cache) with ComfyUI.☆267Updated 2 months ago
- Processes SafeTensors files for Stable Diffusion 1.5 (SD 1.5), Stable Diffusion XL (SDXL), and FLUX models. It extracts the UNet into a s…☆59Updated last year
- ComfyUI implemtation for NAG☆300Updated 3 months ago
- A repo to quantize diffusion models directly in ComfyUI☆105Updated 7 months ago
- A small set of unique adapters meant to bridge the dual_stream_shunt trained for guiding prompt embeddings and diffusion.☆14Updated 2 months ago
- gguf node for comfyui☆197Updated 2 weeks ago
- A powerful anti-burn allowing much higher CFG scales for latent diffusion models (for ComfyUI)☆228Updated last month
- Extension for A1111's Stable Diffusion Webui. Controls amount of detail.☆187Updated 3 months ago
- ComfyUI TCD implementation☆139Updated last month
- Executable Stable Diffusion merge recipes in comfyui☆95Updated 3 weeks ago
- Hot reloading for custom node developers☆118Updated last year
- Load and run SDNQ quantized models in ComfyUI with 50-75% VRAM savings!☆69Updated 3 weeks ago
- Attention Couple, NegPip (negative weights in prompts) for SDXL, FLUX and HunyuanVideo, more CFG++ and SMEA DY samplers, etc.☆234Updated this week
- Prodigy and Schedule-Free, together at last.☆84Updated 4 months ago
- 2-4x faster ComfyUI Image Upscaling using Tensorrt ⚡☆228Updated 3 weeks ago
- FSampler is a training‑free, sampler‑agnostic acceleration layer for diffusion sampling.☆121Updated 2 months ago
- LoRA Power-Merger for ComfyUI☆69Updated last month
- extension for Forge2 webui for Stable Diffusion; adds support for multi-prompting for Flux, SD3, sdXL; Shift for Flux, Sd3; override pred…☆37Updated 2 months ago
- Perturbed-Attention Guidance (PAG), Smoothed Energy Guidance (SEG), Sliding Window Guidance (SWG), PLADIS, Normalized Attention Guidance …☆292Updated 3 weeks ago
- A few nodes to mix sigmas and a custom scheduler that uses phi☆92Updated 5 months ago
- A repository of extra samplers, usable within ComfyUI for most nodes.☆99Updated last year
- Janky implementation of HiDiffusion for ComfyUI