facebookresearch / fairringLinks
Fairring (FAIR + Herring) is a plug-in for PyTorch that provides a process group for distributed training that outperforms NCCL at large scales
☆65Updated 3 years ago
Alternatives and similar repositories for fairring
Users that are interested in fairring are comparing it to the libraries listed below
Sorting:
- System for automated integration of deep learning backends.☆47Updated 2 years ago
- ☆43Updated last year
- PipeTransformer: Automated Elastic Pipelining for Distributed Training of Large-scale Models. ICML 2021☆56Updated 3 years ago
- ☆144Updated 4 months ago
- Memory Optimizations for Deep Learning (ICML 2023)☆64Updated last year
- FTPipe and related pipeline model parallelism research.☆41Updated 2 years ago
- A Python library transfers PyTorch tensors between CPU and NVMe☆116Updated 6 months ago
- ☆91Updated 5 months ago
- Training neural networks in TensorFlow 2.0 with 5x less memory☆131Updated 3 years ago
- pytorch-profiler☆51Updated 2 years ago
- (NeurIPS 2022) Automatically finding good model-parallel strategies, especially for complex models and clusters.☆39Updated 2 years ago
- This repository contains the results and code for the MLPerf™ Training v1.0 benchmark.☆38Updated last year
- extensible collectives library in triton☆86Updated 2 months ago
- ☆74Updated 4 years ago
- Benchmark scripts for TVM☆74Updated 3 years ago
- An external memory allocator example for PyTorch.☆14Updated 3 years ago
- PET: Optimizing Tensor Programs with Partially Equivalent Transformations and Automated Corrections☆121Updated 3 years ago
- ☆50Updated last year
- A performant, memory-efficient checkpointing library for PyTorch applications, designed with large, complex distributed workloads in mind…☆157Updated 6 months ago
- A standalone GEMM kernel for fp16 activation and quantized weight, extracted from FasterTransformer☆92Updated 3 weeks ago
- End to End steps for adding custom ops in PyTorch.☆23Updated 4 years ago
- Benchmarks to capture important workloads.☆31Updated 4 months ago
- An Efficient Pipelined Data Parallel Approach for Training Large Model☆76Updated 4 years ago
- ☆105Updated 9 months ago
- Research and development for optimizing transformers☆128Updated 4 years ago
- Distributed preprocessing and data loading for language datasets☆39Updated last year
- TileFusion is an experimental C++ macro kernel template library that elevates the abstraction level in CUDA C for tile processing.☆90Updated 2 weeks ago
- A Python-embedded DSL that makes it easy to write fast, scalable ML kernels with minimal boilerplate.☆166Updated this week
- An Attention Superoptimizer☆21Updated 5 months ago
- Automated Parallelization System and Infrastructure for Multiple Ecosystems☆79Updated 7 months ago