sail-sg / zero-bubble-pipeline-parallelismLinks
Zero Bubble Pipeline Parallelism
☆396Updated 3 weeks ago
Alternatives and similar repositories for zero-bubble-pipeline-parallelism
Users that are interested in zero-bubble-pipeline-parallelism are comparing it to the libraries listed below
Sorting:
- Dynamic Memory Management for Serving LLMs without PagedAttention☆384Updated this week
- Distributed Triton for Parallel Systems☆775Updated this week
- PyTorch bindings for CUTLASS grouped GEMM.☆125Updated 5 months ago
- Applied AI experiments and examples for PyTorch☆271Updated last week
- USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference☆510Updated last week
- A collection of memory efficient attention operators implemented in the Triton language.☆271Updated last year
- Ring attention implementation with flash attention☆774Updated 2 weeks ago
- Disaggregated serving system for Large Language Models (LLMs).☆605Updated last month
- Official repository for DistFlashAttn: Distributed Memory-efficient Attention for Long-context LLMs Training☆208Updated 9 months ago
- ByteCheckpoint: An Unified Checkpointing Library for LFMs☆215Updated 2 months ago
- A low-latency & high-throughput serving engine for LLMs☆370Updated this week
- nnScaler: Compiling DNN models for Parallel Training☆113Updated last month
- Latency and Memory Analysis of Transformer Models for Training and Inference☆424Updated last month
- Perplexity GPU Kernels☆324Updated 2 weeks ago
- flash attention tutorial written in python, triton, cuda, cutlass☆368Updated 3 weeks ago
- Microsoft Automatic Mixed Precision Library☆602Updated 8 months ago
- ☆208Updated 10 months ago
- Materials for learning SGLang☆426Updated this week
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆365Updated this week
- 🚀 Collection of components for development, training, tuning, and inference of foundation models leveraging PyTorch native components.☆198Updated this week
- [MLSys'24] Atom: Low-bit Quantization for Efficient and Accurate LLM Serving☆305Updated 11 months ago
- [USENIX ATC '24] Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Paral…☆54Updated 10 months ago
- FlagScale is a large model toolkit based on open-sourced projects.☆280Updated last week
- NVIDIA Inference Xfer Library (NIXL)☆365Updated this week
- A baseline repository of Auto-Parallelism in Training Neural Networks☆142Updated 2 years ago
- [ICML 2024] Quest: Query-Aware Sparsity for Efficient Long-Context LLM Inference☆291Updated 6 months ago
- A lightweight design for computation-communication overlap.☆132Updated 3 weeks ago
- Efficient and easy multi-instance LLM serving☆423Updated this week
- [NeurIPS 2024] KVQuant: Towards 10 Million Context Length LLM Inference with KV Cache Quantization☆355Updated 9 months ago
- Cataloging released Triton kernels.☆226Updated 4 months ago