core.launchers.ray_on_slurm_launch

Classes

SPMDWorker
SPMDController	Represents an abstraction over things that run in a loop and can save/load state.

Functions

ray_entrypoint(runner_config)
ray_on_slurm_launch(config, log_dir)

Module Contents

class core.launchers.ray_on_slurm_launch.SPMDWorker(job_config: omegaconf.DictConfig, runner_config: omegaconf.DictConfig, worker_id: int, world_size: int, device: str, gp_size: int | None = None, master_addr: str | None = None, master_port: int | None = None)

runner_config

master_address

master_port

worker_id

device

gp_size

world_size

job_config

distributed_setup = False

_distributed_setup(worker_id: int, world_size: int, master_address: str, master_port: int, device: str, gp_size: int | None)

get_master_address_and_port()

run()

class core.launchers.ray_on_slurm_launch.SPMDController(job_config: omegaconf.DictConfig, runner_config: omegaconf.DictConfig)

Bases: fairchem.core.components.runner.Runner

Represents an abstraction over things that run in a loop and can save/load state.

ie: Trainers, Validators, Relaxation all fall in this category.

Note

When running with the fairchemv2 cli, the job_config and attribute is set at runtime to those given in the config file.

job_config

a managed attribute that gives access to the job config

Type:

DictConfig

job_config

runner_config

device

world_size

gp_group_size

ranks_per_node

num_nodes

workers

run()

save_state(checkpoint_location: str, is_preemption: bool = False) → bool

load_state(checkpoint_location: str | None) → None

core.launchers.ray_on_slurm_launch.ray_entrypoint(runner_config: omegaconf.DictConfig)

core.launchers.ray_on_slurm_launch.ray_on_slurm_launch(config: omegaconf.DictConfig, log_dir: str)