sgl-project / genai-benchLinks
Genai-bench is a powerful benchmark tool designed for comprehensive token-level performance evaluation of large language model (LLM) serving systems.
☆250Updated 3 weeks ago
Alternatives and similar repositories for genai-bench
Users that are interested in genai-bench are comparing it to the libraries listed below
Sorting:
- Allow torch tensor memory to be released and resumed later☆195Updated 3 weeks ago
- ArcticInference: vLLM plugin for high-throughput, low-latency inference☆357Updated this week
- Perplexity GPU Kernels☆547Updated last month
- A low-latency & high-throughput serving engine for LLMs☆459Updated 2 months ago
- Dynamic Memory Management for Serving LLMs without PagedAttention☆453Updated 7 months ago
- Materials for learning SGLang☆703Updated 2 weeks ago
- torchcomms: a modern PyTorch communications API☆314Updated this week
- ByteCheckpoint: An Unified Checkpointing Library for LFMs☆256Updated 3 weeks ago
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆577Updated this week
- Toolchain built around the Megatron-LM for Distributed Training☆81Updated 3 weeks ago
- ☆336Updated last week
- LLM Serving Performance Evaluation Harness☆82Updated 10 months ago
- Efficient and easy multi-instance LLM serving☆519Updated 3 months ago
- JAX backend for SGL☆205Updated this week
- Accelerating MoE with IO and Tile-aware Optimizations☆469Updated last week
- Zero Bubble Pipeline Parallelism☆443Updated 7 months ago
- Virtualized Elastic KV Cache for Dynamic GPU Sharing and Beyond☆730Updated last month
- ☆96Updated 9 months ago
- Open Model Engine (OME) — Kubernetes operator for LLM serving, GPU scheduling, and model lifecycle management. Works with SGLang, vLLM, T…☆351Updated this week
- A minimal implementation of vllm.☆64Updated last year
- Offline optimization of your disaggregated Dynamo graph☆136Updated last week
- A high-performance and light-weight router for vLLM large scale deployment☆65Updated this week
- A high-performance RL training-inference weight synchronization framework, designed to enable second-level parameter updates from trainin…☆118Updated last week
- Stateful LLM Serving☆91Updated 9 months ago
- A unified library for building, evaluating, and storing speculative decoding algorithms for LLM inference in vLLM☆178Updated last week
- ☆122Updated last year
- A lightweight design for computation-communication overlap.☆206Updated last week
- [ICLR2025] Breaking Throughput-Latency Trade-off for Long Sequences with Speculative Decoding☆135Updated last year
- Perplexity open source garden for inference technology☆313Updated last week
- An early research stage expert-parallel load balancer for MoE models based on linear programming.☆476Updated last month