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.
☆130Updated this week
Alternatives and similar repositories for param:
Users that are interested in param are comparing it to the libraries listed below
- NCCL Fast Socket is a transport layer plugin to improve NCCL collective communication performance on Google Cloud.☆116Updated last year
- NCCL Profiling Kit☆127Updated 8 months ago
- Synthesizer for optimal collective communication algorithms☆104Updated 11 months ago
- Microsoft Collective Communication Library☆60Updated 3 months ago
- RDMA and SHARP plugins for nccl library☆181Updated last month
- ☆75Updated 2 years ago
- MSCCL++: A GPU-driven communication stack for scalable AI applications☆308Updated this week
- Microsoft Collective Communication Library☆340Updated last year
- An Efficient Pipelined Data Parallel Approach for Training Large Model☆74Updated 4 years ago
- TACCL: Guiding Collective Algorithm Synthesis using Communication Sketches☆71Updated last year
- An experimental parallel training platform☆54Updated 11 months ago
- RCCL Performance Benchmark Tests☆59Updated this week
- Paella: Low-latency Model Serving with Virtualized GPU Scheduling☆58Updated 10 months ago
- Fine-grained GPU sharing primitives☆141Updated 5 years ago
- Magnum IO community repo☆84Updated last month
- Repository for MLCommons Chakra schema and tools☆89Updated 2 weeks ago
- FTPipe and related pipeline model parallelism research.☆41Updated last year
- A hierarchical collective communications library with portable optimizations☆32Updated 3 months ago
- This is a plugin which lets EC2 developers use libfabric as network provider while running NCCL applications.☆166Updated this week
- Repository for MLCommons Chakra schema and tools☆39Updated last year
- An interference-aware scheduler for fine-grained GPU sharing☆127Updated last month
- A schedule language for large model training☆145Updated 8 months ago
- Thunder Research Group's Collective Communication Library☆33Updated 10 months ago
- oneAPI Collective Communications Library (oneCCL)☆224Updated last week
- AlpaServe: Statistical Multiplexing with Model Parallelism for Deep Learning Serving (OSDI 23)☆82Updated last year
- Code for "Heterogenity-Aware Cluster Scheduling Policies for Deep Learning Workloads", which appeared at OSDI 2020☆126Updated 7 months ago
- Pytorch process group third-party plugin for UCC☆20Updated 10 months ago
- 🔮 Execution time predictions for deep neural network training iterations across different GPUs.☆60Updated 2 years ago
- Unified Collective Communication Library☆235Updated last week
- ☆36Updated 3 months ago