DeepSpeed
add available memory check to accelerators
#4508
Merged

Commits
  • add available memory check to accelerator
    jeffra committed 2 years ago
  • catch case where nvmlInit fails
    jeffra committed 2 years ago
  • add pynvml to reqs
    jeffra committed 2 years ago
  • fix for cpu systems
    jeffra committed 2 years ago
  • Update accelerator/cuda_accelerator.py
    jeffra committed 2 years ago
  • simplify
    jeffra committed 2 years ago
Loading