{ "model": { "emb_dim": 300, "hidden": 512, "layers": 2, "dropout": 0.3 }, "batch_size": 64, "lr": [ 0.002, 0.0004 ], "sampling": { "temperature": 0.8, "top_p": 0.9, "rep_penalty": 1.2 }, "stage1": { "epochs": 29, "best_val_loss": 3.547162824700475, "samples": { "samples": 200, "four_lines": 1.0, "end_word_hit": 0.24 } }, "stage2": { "epochs": 30, "best_val_loss": 2.763802755120199, "samples": { "samples": 160, "four_lines": 1.0, "end_word_hit": 0.2734375 }, "samples_heldout": { "samples": 160, "four_lines": 1.0, "end_word_hit": 0.1875 } }, "test_loss": { "seen_traits": 2.7479417456012247, "heldout_traits": 3.7121890360397534 } }