microsoft / RetrievalAttentionLinks
Scalable long-context LLM decoding that leverages sparsity—by treating the KV cache as a vector storage system.
☆46Updated last week
Alternatives and similar repositories for RetrievalAttention
Users that are interested in RetrievalAttention are comparing it to the libraries listed below
Sorting:
- ☆76Updated last month
- ☆59Updated last month
- ☆99Updated 6 months ago
- Stateful LLM Serving☆70Updated 2 months ago
- ☆62Updated 11 months ago
- A GPU-optimized system for efficient long-context LLMs decoding with low-bit KV cache.☆36Updated last month
- [NeurIPS 2024] Efficient LLM Scheduling by Learning to Rank☆48Updated 7 months ago
- [DAC'25] Official implement of "HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference"☆50Updated 2 weeks ago
- [SIGMOD 2025] PQCache: Product Quantization-based KVCache for Long Context LLM Inference☆52Updated 4 months ago
- [OSDI'24] Serving LLM-based Applications Efficiently with Semantic Variable☆158Updated 8 months ago
- ☆54Updated last year
- 16-fold memory access reduction with nearly no loss☆94Updated 2 months ago
- ☆49Updated 2 weeks ago
- SpInfer: Leveraging Low-Level Sparsity for Efficient Large Language Model Inference on GPUs☆47Updated 2 months ago
- [ICLR 2025] TidalDecode: A Fast and Accurate LLM Decoding with Position Persistent Sparse Attention☆39Updated last month
- PipeRAG: Fast Retrieval-Augmented Generation via Algorithm-System Co-design (KDD 2025)☆21Updated 11 months ago
- DeeperGEMM: crazy optimized version☆69Updated 3 weeks ago
- ☆83Updated 3 years ago
- [ICLR2025] Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decoding☆116Updated 6 months ago
- The Official Implementation of Ada-KV: Optimizing KV Cache Eviction by Adaptive Budget Allocation for Efficient LLM Inference☆76Updated 4 months ago
- ☆74Updated 4 years ago
- nnScaler: Compiling DNN models for Parallel Training☆113Updated last month
- Modular and structured prompt caching for low-latency LLM inference☆94Updated 6 months ago
- ArkVale: Efficient Generative LLM Inference with Recallable Key-Value Eviction (NIPS'24)☆40Updated 5 months ago
- High performance Transformer implementation in C++.☆124Updated 4 months ago
- Quantized Attention on GPU☆44Updated 6 months ago
- NEO is a LLM inference engine built to save the GPU memory crisis by CPU offloading☆35Updated 3 months ago
- A Suite for Parallel Inference of Diffusion Transformers (DiTs) on multi-GPU Clusters☆45Updated 10 months ago
- A lightweight design for computation-communication overlap.☆132Updated 3 weeks ago
- FlexFlow Serve: Low-Latency, High-Performance LLM Serving☆39Updated 3 weeks ago