train_hyperbolic_antmaze.py

import faulthandler

import hydra

try:
    from torch.distributed.elastic.multiprocessing.errors import record
except Exception:
    def record(fn):
        return fn

from train_hyperbolic import run_hyperbolic_training


@hydra.main(
    version_base=None,
    config_path="./config/train",
    config_name="lewm_hyperbolic_antmaze",
)
def run(cfg):
    run_hyperbolic_training(cfg)


@record
def _main():
    faulthandler.enable(all_threads=True)
    run()


if __name__ == "__main__":
    _main()