IST-DASLab / Sparse-Marlin
Boosting 4-bit inference kernels with 2:4 Sparsity
☆61Updated 3 months ago
Alternatives and similar repositories for Sparse-Marlin:
Users that are interested in Sparse-Marlin are comparing it to the libraries listed below
- Fast Matrix Multiplications for Lookup Table-Quantized LLMs☆199Updated this week
- ☆101Updated 2 months ago
- Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decoding☆100Updated 2 weeks ago
- Code for Palu: Compressing KV-Cache with Low-Rank Projection☆59Updated last month
- GEAR: An Efficient KV Cache Compression Recipefor Near-Lossless Generative Inference of LLM☆151Updated 5 months ago
- [MLSys'24] Atom: Low-bit Quantization for Efficient and Accurate LLM Serving☆281Updated 5 months ago
- Fast low-bit matmul kernels in Triton☆161Updated last week
- PyTorch bindings for CUTLASS grouped GEMM.☆56Updated last month
- Odysseus: Playground of LLM Sequence Parallelism☆61Updated 6 months ago
- [ICML 2024] Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference☆218Updated 3 weeks ago
- ☆94Updated 3 months ago
- Repository for Sparse Finetuning of LLMs via modified version of the MosaicML llmfoundry☆38Updated 11 months ago
- QUICK: Quantization-aware Interleaving and Conflict-free Kernel for efficient LLM inference☆113Updated 9 months ago
- KV cache compression for high-throughput LLM inference☆97Updated this week
- ☆201Updated 7 months ago
- An efficient GPU support for LLM inference with x-bit quantization (e.g. FP6,FP5).☆218Updated last month
- High-speed GEMV kernels, at most 2.7x speedup compared to pytorch baseline.☆91Updated 5 months ago
- ☆35Updated 9 months ago
- Implementation of Kangaroo: Lossless Self-Speculative Decoding via Double Early Exiting☆47Updated 5 months ago
- 16-fold memory access reduction with nearly no loss☆63Updated last month
- Triton-based implementation of Sparse Mixture of Experts.☆190Updated 2 weeks ago
- ☆51Updated 2 months ago
- A safetensors extension to efficiently store sparse quantized tensors on disk☆55Updated this week
- ShiftAddLLM: Accelerating Pretrained LLMs via Post-Training Multiplication-Less Reparameterization☆90Updated 2 months ago
- extensible collectives library in triton☆73Updated 2 months ago
- Official repository for LightSeq: Sequence Level Parallelism for Distributed Training of Long Context Transformers☆197Updated 3 months ago
- Ouroboros: Speculative Decoding with Large Model Enhanced Drafting (EMNLP 2024 main)☆79Updated 2 months ago
- Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity☆184Updated last year
- Materials for learning SGLang☆136Updated last week
- Cold Compress is a hackable, lightweight, and open-source toolkit for creating and benchmarking cache compression methods built on top of…☆92Updated 4 months ago