NVIDIA / Star-Attention
Efficient LLM Inference over Long Sequences
β357Updated this week
Alternatives and similar repositories for Star-Attention:
Users that are interested in Star-Attention are comparing it to the libraries listed below
- LLM KV cache compression made easyβ397Updated this week
- [ICLR 2025] DuoAttention: Efficient Long-Context LLM Inference with Retrieval and Streaming Headsβ426Updated last week
- π Efficiently (pre)training foundation models with native PyTorch features, including FSDP for training and SDPA implementation of Flashβ¦β222Updated this week
- Fast Matrix Multiplications for Lookup Table-Quantized LLMsβ229Updated this week
- [NeurIPS 2024] KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantizationβ332Updated 6 months ago
- A family of compressed models obtained via pruning and knowledge distillationβ321Updated 3 months ago
- QServe: W4A8KV4 Quantization and System Co-design for Efficient LLM Servingβ496Updated this week
- [MLSys'24] Atom: Low-bit Quantization for Efficient and Accurate LLM Servingβ295Updated 7 months ago
- Fast low-bit matmul kernels in Tritonβ236Updated this week
- Applied AI experiments and examples for PyTorchβ225Updated this week
- Advanced Quantization Algorithm for LLMs/VLMs.β372Updated this week
- GEAR: An Efficient KV Cache Compression Recipefor Near-Lossless Generative Inference of LLMβ157Updated 7 months ago
- A throughput-oriented high-performance serving framework for LLMsβ737Updated 5 months ago
- π° Must-read papers on KV Cache Compression (constantly updating π€).β306Updated 2 weeks ago
- β220Updated 9 months ago
- Ring attention implementation with flash attentionβ674Updated 2 months ago
- Memory layers use a trainable key-value lookup mechanism to add extra parameters to a model without increasing FLOPs. Conceptually, sparsβ¦β297Updated 2 months ago
- VPTQ, A Flexible and Extreme low-bit quantization algorithmβ588Updated last week
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inferenceβ424Updated this week
- [ICLR'25] Fast Inference of MoE Models with CPU-GPU Orchestrationβ193Updated 3 months ago
- KV cache compression for high-throughput LLM inferenceβ115Updated 2 weeks ago
- β172Updated 4 months ago
- This repository contains the experimental PyTorch native float8 training UXβ221Updated 6 months ago
- [NeurIPS'24 Spotlight, ICLR'25] To speed up Long-context LLMs' inference, approximate and dynamic sparse calculate the attention, which rβ¦β917Updated last week
- [ICLR2025] Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decodingβ107Updated 2 months ago
- Code for Neurips24 paper: QuaRot, an end-to-end 4-bit inference of large language models.β342Updated 2 months ago
- [ICML 2024] CLLMs: Consistency Large Language Modelsβ372Updated 3 months ago
- A high-throughput and memory-efficient inference and serving engine for LLMsβ259Updated 4 months ago
- [ICML 2024] Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inferenceβ245Updated 2 months ago
- For releasing code related to compression methods for transformers, accompanying our publicationsβ408Updated last month