stepfun-ai / StepMeshLinks
☆298Updated last week
Alternatives and similar repositories for StepMesh
Users that are interested in StepMesh are comparing it to the libraries listed below
Sorting:
- A lightweight design for computation-communication overlap.☆177Updated 2 weeks ago
- nnScaler: Compiling DNN models for Parallel Training☆118Updated last week
- High performance Transformer implementation in C++.☆134Updated 8 months ago
- Pipeline Parallelism Emulation and Visualization☆67Updated 3 months ago
- Dynamic Memory Management for Serving LLMs without PagedAttention☆421Updated 4 months ago
- NVSHMEM‑Tutorial: Build a DeepEP‑like GPU Buffer☆130Updated 2 weeks ago
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆418Updated this week
- [USENIX ATC '24] Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Paral…☆65Updated last year
- Ultra and Unified CCL☆567Updated this week
- Allow torch tensor memory to be released and resumed later☆142Updated last week
- Zero Bubble Pipeline Parallelism☆427Updated 4 months ago
- DeepSeek-V3/R1 inference performance simulator☆170Updated 6 months ago
- A prefill & decode disaggregated LLM serving framework with shared GPU memory and fine-grained compute isolation.☆111Updated 4 months ago
- ☆123Updated 10 months ago
- Perplexity GPU Kernels☆476Updated last week
- A low-latency & high-throughput serving engine for LLMs☆422Updated 4 months ago
- ByteCheckpoint: An Unified Checkpointing Library for LFMs☆248Updated 2 months ago
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆412Updated this week
- kvcached: Elastic KV cache for dynamic GPU sharing and efficient multi-LLM inference.☆94Updated this week
- Stateful LLM Serving☆85Updated 6 months ago
- ☆98Updated last year
- ☆147Updated 6 months ago
- Efficient and easy multi-instance LLM serving☆491Updated last month
- Automated Parallelization System and Infrastructure for Multiple Ecosystems☆80Updated 10 months ago
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆324Updated 2 weeks ago
- ☆42Updated last year
- KV cache store for distributed LLM inference☆336Updated 3 weeks ago
- ☆95Updated 6 months ago
- FlexFlow Serve: Low-Latency, High-Performance LLM Serving☆61Updated 2 weeks ago
- ☆121Updated 9 months ago