resume_from_checkpoint
full name: tenpy.resume_from_checkpoint
parent module:
tenpy
type: function
- tenpy.resume_from_checkpoint(*, filename=None, checkpoint_results=None, update_sim_params=None, simulation_class_kwargs=None)[source]
Resume a simulation run from a given checkpoint.
(All parameters have to be given as keyword arguments.)
- Parameters:
filename (None | str) – The filename of the checkpoint to be loaded. You can either specify the filename or the checkpoint_results.
checkpoint_results (None | dict) – Alternatively to filename the results of the simulation so far, i.e. directly the data dictionary saved at a simulation checkpoint.
update_sim_params (None | dict) – Allows to update specific
Simulation
parameters; ignored if None. Usesupdate_recursive()
to update values, such that the keys of update_sim_params can be recursive, e.g. algorithm_params/max_sweeps.simulation_class_kwargs (None | dict) – Further keyword arguments given to the simulation class, ignored if None.
- Returns:
The results from running the simulation, i.e., what
resume_run()
returned.- Return type:
results
Notes
The checkpoint_filename should be relative to the current working directory. If you use the
Simulation.directory
, the simulation class will attempt to change to that directory during initialization. Hence, either resume the simulation from the same directory where you originally started, or update theSimulation.directory
(and :cfg:option`Simulation.output_filename`) parameter with update_sim_params.