ModelTC / awesome-lm-systemLinks
Summary of system papers/frameworks/codes/tools on training or serving large model
☆57Updated 2 years ago
Alternatives and similar repositories for awesome-lm-system
Users that are interested in awesome-lm-system are comparing it to the libraries listed below
Sorting:
- ☆85Updated 9 months ago
- Odysseus: Playground of LLM Sequence Parallelism☆79Updated last year
- An easy-to-use package for implementing SmoothQuant for LLMs☆110Updated 9 months ago
- PyTorch bindings for CUTLASS grouped GEMM.☆141Updated 8 months ago
- GPTQ inference TVM kernel☆41Updated last year
- ☆117Updated 8 months ago
- Quantized Attention on GPU☆44Updated last year
- Performance of the C++ interface of flash attention and flash attention v2 in large language model (LLM) inference scenarios.☆44Updated 11 months ago
- ☆96Updated 10 months ago
- ☆105Updated last year
- Decoding Attention is specially optimized for MHA, MQA, GQA and MLA using CUDA core for the decoding stage of LLM inference.☆46Updated 7 months ago
- GPU operators for sparse tensor operations☆35Updated last year
- llama INT4 cuda inference with AWQ☆55Updated last year
- Flash-LLM: Enabling Cost-Effective and Highly-Efficient Large Generative Model Inference with Unstructured Sparsity☆233Updated 2 years ago
- A Suite for Parallel Inference of Diffusion Transformers (DiTs) on multi-GPU Clusters☆55Updated last year
- Framework to reduce autotune overhead to zero for well known deployments.☆95Updated 4 months ago
- [NeurIPS 2024] The official implementation of "Kangaroo: Lossless Self-Speculative Decoding for Accelerating LLMs via Double Early Exitin…☆65Updated last year
- [ICML 2023] SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models☆23Updated last year
- A simple calculation for LLM MFU.☆66Updated 4 months ago
- ☆84Updated last year
- ☆89Updated 3 years ago
- A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer☆96Updated 4 months ago
- ☆52Updated 8 months ago
- Standalone Flash Attention v2 kernel without libtorch dependency☆113Updated last year
- Keyformer proposes KV Cache reduction through key tokens identification and without the need for fine-tuning☆59Updated last year
- QQQ is an innovative and hardware-optimized W4A8 quantization solution for LLMs.☆154Updated 5 months ago
- ☆71Updated 10 months ago
- QAQ: Quality Adaptive Quantization for LLM KV Cache☆55Updated last year
- ☆65Updated 9 months ago
- [MLSys'24] Atom: Low-bit Quantization for Efficient and Accurate LLM Serving☆335Updated last year