DMTCP: Distributed MultiThreaded CheckPointing
7 daysStars47524h gain0Score0.0
Find trending repositories by name or description.
DMTCP: Distributed MultiThreaded CheckPointing
A performant, memory-efficient checkpointing library for PyTorch applications, designed with large, complex distributed workloads in mind.
The fastest automatic differentiation library for C++, with Python bindings. Forward, adjoint and higher-order derivatives, checkpointing, Eigen support and JIT. Reproducible benchmarks against CppAD, Adept and autodiff.
Pluggable data replication with checkpointing, batching, and integrity events