Deep Koopman checkpoints — dhruv0208/deep-koopman-pendulum

This repository stores research checkpoints (.pt) for Deep Koopman models. Each file is a torch.save dict with state_dict, hyperparameters, system_name, seed, and epoch.

Files

• pendulum/deep_koopman_50_trained_post_trace.pt — family=deep_koopman, system=pendulum, epoch=100, seed=42

Load locally

import torch
ckpt = torch.load("<path/to/file.pt>", map_location="cpu", weights_only=False)
# ckpt["state_dict"], ckpt["hyperparameters"], ckpt["system_name"]