aredden / flux-fp8-apiLinks
Flux diffusion model implementation using quantized fp8 matmul & remaining layers use faster half precision accumulate, which is ~2x faster on consumer devices.
☆274Updated 10 months ago
Alternatives and similar repositories for flux-fp8-api
Users that are interested in flux-fp8-api are comparing it to the libraries listed below
Sorting:
- End-to-end recipes for optimizing diffusion models with torchao and diffusers (inference and FP8 training).☆375Updated 2 months ago
- ☆284Updated 7 months ago
- ☆436Updated last year
- ⚡ Flash Diffusion ⚡: Accelerating Any Conditional Diffusion Model for Few Steps Image Generation (AAAI 2025 Oral)☆616Updated 5 months ago
- https://wavespeed.ai/ Context parallel attention that accelerates DiT model inference with dynamic caching☆356Updated last month
- ☆150Updated this week
- Generate long weighted prompt embeddings for Stable Diffusion☆133Updated 4 months ago
- ☆126Updated 5 months ago
- T-GATE: Temporally Gating Attention to Accelerate Diffusion Model for Free!☆405Updated 5 months ago
- faster parallel inference of mochi-1 video generation model☆124Updated 5 months ago
- SSD-1B, an open-source text-to-image model, outperforming previous versions by being 50% smaller and 60% faster than SDXL.☆177Updated last year
- Cog inference for flux models☆364Updated 3 weeks ago
- Faster generation with text-to-image diffusion models.☆224Updated last month
- [NeurIPS 2024] Boosting the performance of consistency models with PCM!☆489Updated 8 months ago
- Official Repository of the paper "Trajectory Consistency Distillation"☆350Updated last year
- Official code for "RB-Modulation: Training-Free Personalization of Diffusion Models using Stochastic Optimal Control"☆395Updated 5 months ago
- The best OSS video generation models☆134Updated 10 months ago
- [NeurIPS 2024] AsyncDiff: Parallelizing Diffusion Models by Asynchronous Denoising☆203Updated 6 months ago
- Various training scripts used to train bigasp☆98Updated last week
- IP Adapter Instruct☆208Updated last year
- AuraSR: GAN-based Super-Resolution for real-world☆508Updated 9 months ago
- [ICLR 2025] FasterCache: Training-Free Video Diffusion Model Acceleration with High Quality☆242Updated 7 months ago
- [ICML2025] An 8-step inversion and 8-step editing process works effectively with the FLUX-dev model. (3x speedup with results that are co…☆275Updated 3 months ago
- Qwen-Image text to image lora trainer☆166Updated this week
- See original repo here: https://github.com/google/RB-Modulation - ICLR 2025 (Oral)☆125Updated last year
- Text and image to video generation: Kandinsky 4.0 (2024)☆147Updated 8 months ago
- Accelerates Flux.1 image generation, just by using this node.☆137Updated 8 months ago
- InstantID-ROME: Improved Identity-Preserving Generation in Seconds 🔥☆231Updated last year
- A detailed diagram laying out the full Flux.1 [dev] architecture as shared by Black Forest Labs at https://github.com/black-forest-labs/f…☆75Updated 10 months ago
- Implicit Style-Content Separation using B-LoRA☆381Updated 9 months ago