deepseek-ai / LPLBLinks
An early research stage expert-parallel load balancer for MoE models based on linear programming.
☆495Updated 2 months ago
Alternatives and similar repositories for LPLB
Users that are interested in LPLB are comparing it to the libraries listed below
Sorting:
- Accelerating MoE with IO and Tile-aware Optimizations☆569Updated 2 weeks ago
- Tile-Based Runtime for Ultra-Low-Latency LLM Inference☆564Updated last week
- torchcomms: a modern PyTorch communications API☆327Updated this week
- Perplexity GPU Kernels☆554Updated 3 months ago
- Perplexity open source garden for inference technology☆359Updated last month
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆461Updated last month
- ArcticInference: vLLM plugin for high-throughput, low-latency inference☆384Updated this week
- JAX backend for SGL☆234Updated this week
- Block Diffusion for Ultra-Fast Speculative Decoding☆459Updated this week
- Autonomous GPU Kernel Generation via Deep Agents☆228Updated this week
- Genai-bench is a powerful benchmark tool designed for comprehensive token-level performance evaluation of large language model (LLM) serv…☆263Updated this week
- Miles is an enterprise-facing reinforcement learning framework for LLM and VLM post-training, forked from and co-evolving with slime.☆830Updated this week
- ByteCheckpoint: An Unified Checkpointing Library for LFMs☆268Updated last month
- Fast low-bit matmul kernels in Triton☆427Updated this week
- [ICLR2025 Spotlight] MagicPIG: LSH Sampling for Efficient LLM Generation☆248Updated last year
- Code for data-aware compression of DeepSeek models☆69Updated last month
- kernels, of the mega variety☆665Updated last week
- Efficient Long-context Language Model Training by Core Attention Disaggregation☆85Updated last week
- Checkpoint-engine is a simple middleware to update model weights in LLM inference engines☆902Updated this week
- Helpful kernel tutorials and examples for tile-based GPU programming☆630Updated this week
- KV cache compression for high-throughput LLM inference☆151Updated last year
- PyTorch library for cost-effective, fast and easy serving of MoE models.☆278Updated last week
- Dynamic Memory Management for Serving LLMs without PagedAttention☆457Updated 8 months ago
- Ship correct and fast LLM kernels to PyTorch☆140Updated 3 weeks ago
- Easy, Fast, and Scalable Multimodal AI☆106Updated last week
- High-performance distributed data shuffling (all-to-all) library for MoE training and inference☆111Updated last month
- ☆342Updated last week
- Our first fully AI generated deep learning system☆481Updated this week
- Open ABI and FFI for Machine Learning Systems☆313Updated last week
- TritonParse: A Compiler Tracer, Visualizer, and Reproducer for Triton Kernels☆194Updated this week