{ "fold_results": [ { "fold_idx": 0, "best_val_loss": 6.144314289093018, "epochs_trained": 25, "final_train_loss": 1.4692220211029052 }, { "fold_idx": 1, "best_val_loss": 8.569346030553183, "epochs_trained": 20, "final_train_loss": 1.5929443359375 }, { "fold_idx": 2, "best_val_loss": 3.7409281730651855, "epochs_trained": 22, "final_train_loss": 1.9401288827260335 }, { "fold_idx": 3, "best_val_loss": 3.47284197807312, "epochs_trained": 27, "final_train_loss": 1.8295514345169068 }, { "fold_idx": 4, "best_val_loss": 2.756531000137329, "epochs_trained": 19, "final_train_loss": 1.9399811571294612 } ], "summary": { "val_loss_mean": 4.936792294184367, "val_loss_std": 2.1438440638412697 }, "config": { "d_model": 256, "num_heads": 8, "n_attn_layers": 4, "fusion_strategy": "attention", "head_hidden_dim": 128, "dropout": 0.1, "use_mpnn": true, "lr": 0.0001, "weight_decay": 1e-05, "batch_size": 32, "epochs": 100, "patience": 15, "init_from_pretrain": "models/pretrain_delivery.pt", "freeze_backbone": false } }