ByteDance-Seed / cudaLLMLinks
☆106Updated last month
Alternatives and similar repositories for cudaLLM
Users that are interested in cudaLLM are comparing it to the libraries listed below
Sorting:
- ☆96Updated 4 months ago
- Quantized Attention on GPU☆44Updated 9 months ago
- ☆64Updated 4 months ago
- ☆50Updated 4 months ago
- DeeperGEMM: crazy optimized version☆70Updated 4 months ago
- Decoding Attention is specially optimized for MHA, MQA, GQA and MLA using CUDA core for the decoding stage of LLM inference.☆43Updated 3 months ago
- ☆82Updated 7 months ago
- ⚡️Write HGEMM from scratch using Tensor Cores with WMMA, MMA and CuTe API, Achieve Peak⚡️ Performance.☆112Updated 4 months ago
- Implement Flash Attention using Cute.☆95Updated 9 months ago
- PyTorch bindings for CUTLASS grouped GEMM.☆119Updated 3 months ago
- ☆94Updated 5 months ago
- 🤖FFPA: Extend FlashAttention-2 with Split-D, ~O(1) SRAM complexity for large headdim, 1.8x~3x↑🎉 vs SDPA EA.☆217Updated last month
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆100Updated this week
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.☆97Updated 2 months ago
- ☆72Updated 10 months ago
- A GPU-optimized system for efficient long-context LLMs decoding with low-bit KV cache.☆59Updated 2 weeks ago
- Code for paper: [ICLR2025 Oral] FlexPrefill: A Context-Aware Sparse Attention Mechanism for Efficient Long-Sequence Inference☆141Updated 4 months ago
- A Suite for Parallel Inference of Diffusion Transformers (DiTs) on multi-GPU Clusters☆48Updated last year
- An experimental communicating attention kernel based on DeepEP.☆34Updated last month
- ☆38Updated last month
- ☆126Updated 3 months ago
- Odysseus: Playground of LLM Sequence Parallelism☆77Updated last year
- We invite you to visit and follow our new repository at https://github.com/microsoft/TileFusion. TiledCUDA is a highly efficient kernel …☆185Updated 7 months ago
- ☆88Updated 10 months ago
- Framework to reduce autotune overhead to zero for well known deployments.☆82Updated this week
- QuTLASS: CUTLASS-Powered Quantized BLAS for Deep Learning☆92Updated last week
- Tritonbench is a collection of PyTorch custom operators with example inputs to measure their performance.☆223Updated this week
- Accelerate LLM preference tuning via prefix sharing with a single line of code☆43Updated 2 months ago
- [DAC'25] Official implement of "HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference"☆70Updated 3 months ago
- Standalone Flash Attention v2 kernel without libtorch dependency☆110Updated last year