facebookresearch / param
PArametrized Recommendation and Ai Model benchmark is a repository for development of numerous uBenchmarks as well as end to end nets for evaluation of training and inference platforms.
☆128Updated this week
Alternatives and similar repositories for param:
Users that are interested in param are comparing it to the libraries listed below
- Microsoft Collective Communication Library☆62Updated 2 months ago
- Synthesizer for optimal collective communication algorithms☆103Updated 10 months ago
- NCCL Profiling Kit☆127Updated 7 months ago
- ☆75Updated 2 years ago
- RDMA and SHARP plugins for nccl library☆176Updated 3 weeks ago
- NCCL Fast Socket is a transport layer plugin to improve NCCL collective communication performance on Google Cloud.☆115Updated last year
- Repository for MLCommons Chakra schema and tools☆84Updated 3 weeks ago
- An experimental parallel training platform☆54Updated 10 months ago
- Repository for MLCommons Chakra schema and tools☆39Updated last year
- TACCL: Guiding Collective Algorithm Synthesis using Communication Sketches☆69Updated last year
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆297Updated this week
- An Efficient Pipelined Data Parallel Approach for Training Large Model☆73Updated 4 years ago
- oneAPI Collective Communications Library (oneCCL)☆222Updated 3 weeks ago
- Microsoft Collective Communication Library☆333Updated last year
- This is a plugin which lets EC2 developers use libfabric as network provider while running NCCL applications.☆163Updated this week
- RCCL Performance Benchmark Tests☆59Updated last month
- Fine-grained GPU sharing primitives☆141Updated 4 years ago
- 🔮 Execution time predictions for deep neural network training iterations across different GPUs.☆59Updated 2 years ago
- Code for "Heterogenity-Aware Cluster Scheduling Policies for Deep Learning Workloads", which appeared at OSDI 2020☆126Updated 6 months ago
- An interference-aware scheduler for fine-grained GPU sharing☆122Updated 3 weeks ago
- ROCm Communication Collectives Library (RCCL)☆297Updated this week
- Thunder Research Group's Collective Communication Library☆33Updated 9 months ago
- Paella: Low-latency Model Serving with Virtualized GPU Scheduling☆58Updated 9 months ago
- (NeurIPS 2022) Automatically finding good model-parallel strategies, especially for complex models and clusters.☆37Updated 2 years ago
- Chimera: Efficiently Training Large-Scale Neural Networks with Bidirectional Pipelines.☆57Updated last year
- AlpaServe: Statistical Multiplexing with Model Parallelism for Deep Learning Serving (OSDI 23)☆81Updated last year
- Pytorch process group third-party plugin for UCC☆20Updated 10 months ago
- Magnum IO community repo☆84Updated 3 weeks ago
- A schedule language for large model training☆144Updated 8 months ago
- A resilient distributed training framework☆88Updated 10 months ago