meta-pytorch / torchcommsLinks
torchcomms: a modern PyTorch communications API
☆320Updated this week
Alternatives and similar repositories for torchcomms
Users that are interested in torchcomms are comparing it to the libraries listed below
Sorting:
- NVIDIA NVSHMEM is a parallel programming interface for NVIDIA GPUs based on OpenSHMEM. NVSHMEM can significantly reduce multi-process com…☆443Updated 2 weeks ago
- Perplexity open source garden for inference technology☆332Updated 3 weeks ago
- Perplexity GPU Kernels☆552Updated 2 months ago
- An early research stage expert-parallel load balancer for MoE models based on linear programming.☆485Updated last month
- Helpful kernel tutorials and examples for tile-based GPU programming☆568Updated this week
- Genai-bench is a powerful benchmark tool designed for comprehensive token-level performance evaluation of large language model (LLM) serv…☆251Updated this week
- Accelerating MoE with IO and Tile-aware Optimizations☆542Updated this week
- ☆340Updated 2 weeks ago
- Allow torch tensor memory to be released and resumed later☆202Updated this week
- ☆154Updated last year
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆451Updated this week
- ByteCheckpoint: An Unified Checkpointing Library for LFMs☆261Updated last month
- A lightweight design for computation-communication overlap.☆209Updated 3 weeks ago
- A NCCL extension library, designed to efficiently offload GPU memory allocated by the NCCL communication library.☆80Updated last month
- High-performance distributed data shuffling (all-to-all) library for MoE training and inference☆106Updated 2 weeks ago
- Autonomous GPU Kernel Generation via Deep Agents☆211Updated this week
- Microsoft Collective Communication Library☆66Updated last year
- AMD RAD's multi-GPU Triton-based framework for seamless multi-GPU programming☆148Updated last week
- KV cache store for distributed LLM inference☆385Updated 2 months ago
- Dynamic Memory Management for Serving LLMs without PagedAttention☆454Updated 7 months ago
- JAX backend for SGL☆218Updated this week
- TritonParse: A Compiler Tracer, Visualizer, and Reproducer for Triton Kernels☆185Updated this week
- FlexFlow Serve: Low-Latency, High-Performance LLM Serving☆68Updated 4 months ago
- Automated Parallelization System and Infrastructure for Multiple Ecosystems☆82Updated last year
- ArcticInference: vLLM plugin for high-throughput, low-latency inference☆368Updated last week
- ☆81Updated 3 months ago
- DeepXTrace is a lightweight tool for precisely diagnosing slow ranks in DeepEP-based environments.☆86Updated 3 weeks ago
- Tritonbench is a collection of PyTorch custom operators with example inputs to measure their performance.☆310Updated this week
- NVIDIA Resiliency Extension is a python package for framework developers and users to implement fault-tolerant features. It improves the …☆247Updated last week
- ☆72Updated 11 months ago