lutnn / blink-mmLinks
☆15Updated last year
Alternatives and similar repositories for blink-mm
Users that are interested in blink-mm are comparing it to the libraries listed below
Sorting:
- ☆149Updated last year
- ☆62Updated last year
- ☆19Updated 9 months ago
- Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity☆216Updated last year
- PyTorch extension for emulating FP8 data formats on standard FP32 Xeon/GPU hardware.☆110Updated 7 months ago
- ☆39Updated 2 years ago
- ☆18Updated 4 years ago
- ☆48Updated last year
- SpInfer: Leveraging Low-Level Sparsity for Efficient Large Language Model Inference on GPUs☆50Updated 3 months ago
- PyTorch library for cost-effective, fast and easy serving of MoE models.☆210Updated 2 weeks ago
- ☆77Updated 5 months ago
- ☆37Updated last year
- QJL: 1-Bit Quantized JL transform for KV Cache Quantization with Zero Overhead☆26Updated 5 months ago
- LLM Inference analyzer for different hardware platforms☆80Updated last week
- A GPU-optimized system for efficient long-context LLMs decoding with low-bit KV cache.☆54Updated last week
- [ICLR'25] Fast Inference of MoE Models with CPU-GPU Orchestration☆218Updated 8 months ago
- FlashSparse significantly reduces the computation redundancy for unstructured sparsity (for SpMM and SDDMM) on Tensor Cores through a Swa…☆29Updated last week
- PIM-DL: Expanding the Applicability of Commodity DRAM-PIMs for Deep Learning via Algorithm-System Co-Optimization☆31Updated last year
- GPU operators for sparse tensor operations☆33Updated last year
- A lightweight design for computation-communication overlap.☆148Updated last month
- [ICLR 2025] TidalDecode: A Fast and Accurate LLM Decoding with Position Persistent Sparse Attention☆41Updated 3 months ago
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆112Updated last year
- Benchmark code for the "Online normalizer calculation for softmax" paper☆95Updated 6 years ago
- SparseTIR: Sparse Tensor Compiler for Deep Learning☆137Updated 2 years ago
- A Vectorized N:M Format for Unleashing the Power of Sparse Tensor Cores☆52Updated last year
- [DAC'25] Official implement of "HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference"☆59Updated last month
- A curated list for Efficient Large Language Models☆11Updated last year
- ArkVale: Efficient Generative LLM Inference with Recallable Key-Value Eviction (NIPS'24)☆41Updated 7 months ago
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.☆91Updated 3 weeks ago
- ☆9Updated last year