stepfun-ai / StepMeshLinks
☆24Updated 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.☆154Updated last month
- nnScaler: Compiling DNN models for Parallel Training☆114Updated 3 weeks ago
- ☆109Updated 8 months ago
- High performance Transformer implementation in C++.☆128Updated 6 months ago
- Automated Parallelization System and Infrastructure for Multiple Ecosystems☆79Updated 8 months ago
- Pipeline Parallelism Emulation and Visualization☆54Updated last month
- Stateful LLM Serving☆77Updated 4 months ago
- PyTorch library for cost-effective, fast and easy serving of MoE models.☆213Updated 3 weeks ago
- Allow torch tensor memory to be released and resumed later☆91Updated 3 weeks ago
- DeeperGEMM: crazy optimized version☆70Updated 2 months ago
- ☆60Updated 3 months ago
- Dynamic Memory Management for Serving LLMs without PagedAttention☆405Updated 2 months ago
- ☆45Updated 6 months ago
- Ultra and Unified CCL☆440Updated this week
- [USENIX ATC '24] Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Paral…☆60Updated last year
- Sequence-level 1F1B schedule for LLMs.☆29Updated last month
- [OSDI'24] Serving LLM-based Applications Efficiently with Semantic Variable☆170Updated 10 months ago
- ByteCheckpoint: An Unified Checkpointing Library for LFMs☆226Updated 3 weeks ago
- ☆65Updated last year
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆253Updated this week
- ☆75Updated 4 years ago
- ☆96Updated 10 months ago
- ☆92Updated 4 months ago
- [ICML 2024] Serving LLMs on heterogeneous decentralized clusters.☆27Updated last year
- A low-latency & high-throughput serving engine for LLMs☆397Updated 2 months ago
- ☆101Updated 7 months ago
- DeepSeek-V3/R1 inference performance simulator☆156Updated 4 months ago
- ☆145Updated 4 months ago
- A prefill & decode disaggregated LLM serving framework with shared GPU memory and fine-grained compute isolation.☆103Updated 2 months ago
- FlexFlow Serve: Low-Latency, High-Performance LLM Serving☆48Updated this week