deepseek-ai / profile-dataLinks
Analyze computation-communication overlap in V3/R1.
☆1,103Updated 6 months ago
Alternatives and similar repositories for profile-data
Users that are interested in profile-data are comparing it to the libraries listed below
Sorting:
- Expert Parallelism Load Balancer☆1,272Updated 6 months ago
- A bidirectional pipeline parallelism algorithm for computation-communication overlap in V3/R1 training.☆2,863Updated 6 months ago
- DeepGEMM: clean and efficient FP8 GEMM kernels with fine-grained scaling☆5,752Updated this week
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,133Updated last month
- A PyTorch Native LLM Training Framework☆874Updated 3 weeks ago
- Mooncake is the serving platform for Kimi, a leading LLM service provided by Moonshot AI.☆4,039Updated this week
- Distributed Compiler based on Triton for Parallel Systems☆1,148Updated last week
- Domain-specific language designed to streamline the development of high-performance GPU/CPU/Accelerators kernels☆1,793Updated this week
- DeepEP: an efficient expert-parallel communication library☆8,564Updated this week
- Muon is Scalable for LLM Training☆1,318Updated 2 months ago
- MoBA: Mixture of Block Attention for Long-Context LLMs☆1,908Updated 6 months ago
- FlashInfer: Kernel Library for LLM Serving☆3,829Updated this week
- slime is an LLM post-training framework for RL Scaling.☆2,023Updated this week
- Materials for learning SGLang☆590Updated last month
- A throughput-oriented high-performance serving framework for LLMs☆893Updated 2 weeks ago
- 🐳 Efficient Triton implementations for "Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention"☆877Updated 6 months ago
- High-performance inference framework for large language models, focusing on efficiency, flexibility, and availability.☆1,291Updated this week
- DeepSeekMoE: Towards Ultimate Expert Specialization in Mixture-of-Experts Language Models☆1,801Updated last year
- Disaggregated serving system for Large Language Models (LLMs).☆695Updated 5 months ago
- Perplexity GPU Kernels☆476Updated 2 weeks ago
- Ring attention implementation with flash attention☆880Updated 3 weeks ago
- RTP-LLM: Alibaba's high-performance LLM inference engine for diverse applications.☆870Updated this week
- [NeurIPS'24 Spotlight, ICLR'25, ICML'25] To speed up Long-context LLMs' inference, approximate and dynamic sparse calculate the attention…☆1,133Updated this week
- Production-tested AI infrastructure tools for efficient AGI development and community-driven innovation☆7,922Updated 4 months ago
- NVIDIA Inference Xfer Library (NIXL)☆648Updated this week
- FlagGems is an operator library for large language models implemented in the Triton Language.☆690Updated this week
- Zero Bubble Pipeline Parallelism☆427Updated 4 months ago
- Train speculative decoding models effortlessly and port them smoothly to SGLang serving.☆412Updated this week
- FlagScale is a large model toolkit based on open-sourced projects.☆357Updated this week
- Transformers-compatible library for applying various compression algorithms to LLMs for optimized deployment with vLLM☆2,019Updated this week