Horovod 0.23.0-foss-2021a-CUDA-11.3.1-PyTorch-1.10.0
There is a newer install of Horovod
Horovod is a distributed training frameworkAccessing Horovod 0.23.0-foss-2021a-CUDA-11.3.1-PyTorch-1.10.0
To load the module for Horovod 0.23.0-foss-2021a-CUDA-11.3.1-PyTorch-1.10.0 please use this command on the BEAR systems (BlueBEAR and BEAR Cloud VMs):
📋
module load Horovod/0.23.0-foss-2021a-CUDA-11.3.1-PyTorch-1.10.0
BEAR Apps Version
Architectures
EL8-icelake (GPUs: NVIDIA A100, NVIDIA A30) — EL8-sapphirerapids
The listed architectures consist of two part: OS-CPU. The OS used is represented by EL and there are several different processor (CPU) types available on BlueBEAR. More information about the processor types on BlueBEAR is available on the BlueBEAR Job Submission page.
Extensions
- cloudpickle 2.0.0
- horovod 0.23.0
More Information
For more information visit the Horovod website.
Dependencies
This version of Horovod has a direct dependency on: CUDA/11.3.1 foss/2021a NCCL/2.10.3-GCCcore-10.3.0-CUDA-11.3.1 Python/3.9.5-GCCcore-10.3.0 PyTorch/1.10.0-foss-2021a-CUDA-11.3.1 PyYAML/5.4.1-GCCcore-10.3.0 UCX-CUDA/1.10.0-GCCcore-10.3.0-CUDA-11.3.1
Other Versions
These versions of Horovod are available on the BEAR systems (BlueBEAR and BEAR Cloud VMs). These will be retained in accordance with our Applications Support and Retention Policy.
Version | BEAR Apps Version |
---|---|
0.28.1-foss-2022b-CUDA-12.0.0-PyTorch-2.1.2 | 2022b |
0.28.0-foss-2022a-CUDA-11.7.0-PyTorch-2.0.1 | 2022a |
0.23.0-foss-2021a-CUDA-11.3.1-TensorFlow-2.6.0 | 2021a |
Last modified on 17th January 2022