Statements (26)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
gptkb:software
|
| gptkbp:documentation |
https://pytorch.org/docs/stable/distributed.html
|
| https://www.w3.org/2000/01/rdf-schema#label |
torch.distributed package
|
| gptkbp:implementedIn |
torch.distributed.launch
torch.distributed.run |
| gptkbp:partOf |
gptkb:PyTorch
|
| gptkbp:provides |
distributed training
RPC framework distributed optimizer DistributedDataParallel ProcessGroup API collective communication primitives distributed autograd |
| gptkbp:supports |
gptkb:Gloo_backend
gptkb:NCCL_backend data parallelism model parallelism multi-GPU training MPI backend multi-node training TCP backend |
| gptkbp:usedBy |
gptkb:researchers
industry practitioners |
| gptkbp:usedFor |
scaling deep learning workloads
|
| gptkbp:bfsParent |
gptkb:PyTorch_Distributed
|
| gptkbp:bfsLayer |
6
|