Sequence-level 1F1B schedule for LLMs.
☆37Aug 26, 2025Updated 9 months ago
Alternatives and similar repositories for Seq1F1B
Users that are interested in Seq1F1B are comparing it to the libraries listed below. We may earn a commission when you buy through links labeled 'Ad' on this page.
Sorting:
- Distributed IO-aware Attention algorithm☆24Sep 24, 2025Updated 8 months ago
- Official repository for the paper DynaPipe: Optimizing Multi-task Training through Dynamic Pipelines☆19Dec 8, 2023Updated 2 years ago
- Vocabulary Parallelism☆26Mar 10, 2025Updated last year
- [ACL 2025 main] FR-Spec: Frequency-Ranked Speculative Sampling☆54Jul 15, 2025Updated 10 months ago
- Zero Bubble Pipeline Parallelism☆456May 7, 2025Updated last year
- Proton VPN Special Offer - Get 70% off • AdSpecial partner offer. Trusted by over 100 million users worldwide. Tested, Approved and Recommended by Experts.
- Allow torch tensor memory to be released and resumed later☆245May 16, 2026Updated last week
- LLM training technologies developed by kwai☆72Jan 21, 2026Updated 4 months ago
- ☆47Sep 8, 2025Updated 8 months ago
- A benchmark suited especially for deep learning operators☆42Feb 13, 2023Updated 3 years ago
- ☆26Dec 5, 2022Updated 3 years ago
- Canvas: End-to-End Kernel Architecture Search in Neural Networks☆27Nov 18, 2024Updated last year
- CUDA SGEMM optimization note☆15Oct 31, 2023Updated 2 years ago
- Ring attention implementation with flash attention☆1,020Sep 10, 2025Updated 8 months ago
- Efficient Long-context Language Model Training by Core Attention Disaggregation☆103Apr 7, 2026Updated last month
- Managed Kubernetes at scale on DigitalOcean • AdDigitalOcean Kubernetes includes the control plane, bandwidth allowance, container registry, automatic updates, and more for free.
- ☆24Aug 15, 2023Updated 2 years ago
- Ouroboros: Speculative Decoding with Large Model Enhanced Drafting (EMNLP 2024 main)☆116Mar 20, 2025Updated last year
- A lightweight design for computation-communication overlap.☆234Jan 20, 2026Updated 4 months ago
- Utility scripts for PyTorch (e.g. Make Perfetto show some disappearing kernels, Memory profiler that understands more low-level allocatio…☆107Sep 11, 2025Updated 8 months ago
- ☆11Apr 10, 2024Updated 2 years ago
- Composable and Embeddable Communication Runtime for Distributed AI Services☆101May 19, 2026Updated last week
- ☆641Jan 14, 2026Updated 4 months ago
- ☆14Oct 3, 2024Updated last year
- ☆13Feb 22, 2023Updated 3 years ago
- Simple, predictable pricing with DigitalOcean hosting • AdAlways know what you'll pay with monthly caps and flat pricing. Enterprise-grade infrastructure trusted by 600k+ customers.
- ☆63Jul 21, 2024Updated last year
- [ICML 2023] SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models☆11Dec 13, 2023Updated 2 years ago
- 清华大学宿舍洗衣机空闲提醒小程序☆14Feb 4, 2021Updated 5 years ago
- gLLM: Global Balanced Pipeline Parallelism System for Distributed LLM Serving with Token Throttling☆59Updated this week
- [ICML 2024 Oral] Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs☆126Jul 4, 2025Updated 10 months ago
- A NCCL extension library, designed to efficiently offload GPU memory allocated by the NCCL communication library.☆108Dec 17, 2025Updated 5 months ago
- Best practices for training DeepSeek, Mixtral, Qwen and other MoE models using Megatron Core.☆192May 20, 2026Updated last week
- CPM.cu is a lightweight, high-performance CUDA implementation for LLMs, optimized for end-device inference and featuring cutting-edge tec…☆240Jan 14, 2026Updated 4 months ago
- ☆78May 4, 2021Updated 5 years ago
- Deploy to Railway using AI coding agents - Free Credits Offer • AdUse Claude Code, Codex, OpenCode, and more. Autonomous software development now has the infrastructure to match with Railway.
- ☆38Jan 15, 2021Updated 5 years ago
- [ICML 2024] Sparse Model Inversion: Efficient Inversion of Vision Transformers with Less Hallucination☆14Apr 29, 2025Updated last year
- Surrogate-based Hyperparameter Tuning System☆30Jun 29, 2023Updated 2 years ago
- A fast communication-overlapping library for tensor/expert parallelism on GPUs.☆1,309Aug 28, 2025Updated 8 months ago
- Dual-way gradient sparsification approach for async DNN training, based on PyTorch.☆10Dec 8, 2022Updated 3 years ago
- Bamboo is a system for running large pipeline-parallel DNNs affordably, reliably, and efficiently using spot instances.☆55Dec 11, 2022Updated 3 years ago
- ☆24May 9, 2025Updated last year