| − | * 神经网络剪枝:临界保留率 p<nowiki><sub>c</sub></nowiki><blockquote>def find_critical_pruning(model, test_data): accuracies = [] for p in np.linspace(0.1, 1.0, 50): pruned_model = prune_model(model, p) acc = test(pruned_model, test_data) accuracies.append(acc) return np.gradient(accuracies).argmin() # 最大斜率点对应p_c</blockquote> | + | * 神经网络剪枝:临界保留率 p<nowiki><sub>c</sub></nowiki> |