{ "config": { "run": { "name": "us_wildfire_size", "seed": 0, "dtype": "float32" }, "dataset": { "kind": "tabular", "name": "us_wildfire_size", "file": "us_wildfire_size.pt", "split": { "standardize": true } }, "network": { "name": "mlp", "params": { "width": 128, "depth": 3, "time_dim": 64, "dropout": 0.0, "use_norm": true } }, "models": { "ddpm_v": { "params": { "n_steps": 128, "sigma_max": 2.0, "sampler": "ddpm" } }, "dlpm_eps": { "params": { "n_steps": 128, "alpha": 1.8, "sampler": "native" } }, "tedm_origin": { "params": { "n_steps": 64, "nu": 2.2, "sigma_min": 0.005, "sigma_max": 5.0, "sigma_data": 1.0, "solver": "edm_stochastic_heun" } } }, "pilot": { "n_trials": 2, "train": { "n_epochs": 96, "early_stopping_patience": 32, "freq_logging": 16, "stats_freq_epochs": 1 }, "sweep": { "ddpm_v": { "model": { "sigma_max": [ 2.0, 5.0 ] }, "train": { "lr": [ 0.0002, 0.0005 ] } }, "dlpm_eps": { "train": { "lr": [ 0.0002, 0.0005 ] } }, "tedm_origin": { "train": { "lr": [ 0.0002, 0.0005 ] } } } }, "train": { "device": "cuda", "data_device": "cpu", "batch_size": 512, "n_epochs": 640, "early_stopping_patience": 100, "lr": 0.0005, "num_workers": 0, "use_adamw": true, "weight_decay": 1e-06, "grad_clip_norm": 10.0, "lr_schedule": "cosine", "warmup_steps": 100, "cosine_eta_min_ratio": 0.05, "freq_logging": 16, "stats_freq_epochs": 1, "log_grad_norm": true, "ckpt_freq_epochs": 64, "ckpt_keep_last": 4 }, "test": { "batch_size": 512, "loss_repeats": 2, "loss_samples": 2000, "n_samples": 32 }, "huggingface": { "namespace": "Diffusion-Research-Lab", "repository": "us-wildfire-size-1d-diffusion" } }, "model_name": "ddim_v", "model_params": { "n_steps": 128, "sigma_max": 2.0, "sampler": "ddim", "eta": 0.0, "sample_steps": null }, "sample_shape": [ 1 ], "data_scale": 0.03256954997777939 }