merrymercy / Awesome-Efficient-LLM
A curated list for Efficient Large Language Models
☆11Updated last year
Alternatives and similar repositories for Awesome-Efficient-LLM:
Users that are interested in Awesome-Efficient-LLM are comparing it to the libraries listed below
- An Attention Superoptimizer☆21Updated 2 months ago
- ☆19Updated 6 months ago
- FractalTensor is a programming framework that introduces a novel approach to organizing data in deep neural networks (DNNs) as a list of …☆24Updated 3 months ago
- An external memory allocator example for PyTorch.☆14Updated 3 years ago
- Tacker: Tensor-CUDA Core Kernel Fusion for Improving the GPU Utilization while Ensuring QoS☆21Updated last month
- [ICLR 2025] TidalDecode: A Fast and Accurate LLM Decoding with Position Persistent Sparse Attention☆31Updated 3 weeks ago
- ☆11Updated 3 years ago
- Quantized Attention on GPU☆45Updated 4 months ago
- ☆67Updated 2 months ago
- GPTQ inference TVM kernel☆38Updated 11 months ago
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.☆74Updated this week
- Stateful LLM Serving☆50Updated 2 weeks ago
- PipeInfer: Accelerating LLM Inference using Asynchronous Pipelined Speculation☆28Updated 4 months ago
- Memory Optimizations for Deep Learning (ICML 2023)☆62Updated last year
- DeeperGEMM: crazy optimized version☆63Updated 2 weeks ago
- ☆55Updated 9 months ago
- PyTorch compilation tutorial covering TorchScript, torch.fx, and Slapo☆18Updated 2 years ago
- A source-to-source compiler for optimizing CUDA dynamic parallelism by aggregating launches☆15Updated 5 years ago
- A minimal implementation of vllm.☆37Updated 8 months ago
- ☆26Updated last week
- GPU operators for sparse tensor operations☆31Updated last year
- Artifact for "Apparate: Rethinking Early Exits to Tame Latency-Throughput Tensions in ML Serving" [SOSP '24]☆22Updated 4 months ago
- Open deep learning compiler stack for cpu, gpu and specialized accelerators☆18Updated last week
- Debug print operator for cudagraph debugging☆10Updated 7 months ago
- ☆52Updated this week
- Official resporitory for "IPDPS' 24 QSync: Quantization-Minimized Synchronous Distributed Training Across Hybrid Devices".☆19Updated last year
- Framework to reduce autotune overhead to zero for well known deployments.☆63Updated last week
- A GPU-optimized system for efficient long-context LLMs decoding with low-bit KV cache.☆23Updated this week
- A practical way of learning Swizzle☆16Updated last month
- TensorRT LLM Benchmark Configuration☆13Updated 8 months ago