aredden / torch-cublas-hgemmLinks
PyTorch half precision gemm lib w/ fused optional bias + optional relu/gelu
☆75Updated 10 months ago
Alternatives and similar repositories for torch-cublas-hgemm
Users that are interested in torch-cublas-hgemm are comparing it to the libraries listed below
Sorting:
- [WIP] Better (FP8) attention for Hopper☆33Updated 8 months ago
- ☆76Updated 10 months ago
- ☆17Updated 10 months ago
- (WIP) Parallel inference for black-forest-labs' FLUX model.☆18Updated 11 months ago
- 🎬 3.7× faster video generation E2E 🖼️ 1.6× faster image generation E2E ⚡ ColumnSparseAttn 9.3× vs FlashAttn‑3 💨 ColumnSparseGEMM 2.5× …☆87Updated last month
- Writing FLUX in Triton☆41Updated last year
- ☆49Updated last year
- Making Flux go brrr on GPUs.☆150Updated 3 months ago
- https://wavespeed.ai/ Context parallel attention that accelerates DiT model inference with dynamic caching☆385Updated 3 months ago
- RAM is all you need☆194Updated last week
- Faster generation with text-to-image diffusion models.☆228Updated 4 months ago
- High-throughput tensor loading for PyTorch☆105Updated this week
- Triton kernels for Flux☆22Updated 3 months ago
- [ACL 2023] The official implementation of "CAME: Confidence-guided Adaptive Memory Optimization"☆93Updated 7 months ago
- End-to-end recipes for optimizing diffusion models with torchao and diffusers (inference and FP8 training).☆381Updated 5 months ago
- Comparison of different stable diffusion implementations and optimizations☆39Updated last year
- Model Compression Toolbox for Large Language Models and Diffusion Models☆678Updated 2 months ago
- Flux diffusion model implementation using quantized fp8 matmul & remaining layers use faster half precision accumulate, which is ~2x fast…☆281Updated last year
- Model code for inferencing T5☆66Updated 7 months ago
- [NeurIPS 2025] Radial Attention: O(nlogn) Sparse Attention with Energy Decay for Long Video Generation☆531Updated last month
- [NeurIPS 2024] AsyncDiff: Parallelizing Diffusion Models by Asynchronous Denoising☆206Updated last month
- ☆137Updated last year
- ☆18Updated last year
- ☆49Updated 8 months ago
- ☆126Updated 7 months ago
- ☆55Updated this week
- A Unified Cache Acceleration Framework for 🤗Diffusers: Qwen-Image-Lightning, Qwen-Image, HunyuanImage, Wan, FLUX, etc.☆460Updated this week
- A parallelism VAE avoids OOM for high resolution image generation☆81Updated 2 months ago
- faster parallel inference of mochi-1 video generation model☆125Updated 8 months ago
- Prodigy and Schedule-Free, together at last.☆77Updated last month