增加 n_jobs 对 GridSearchCV 没有影响

我设置了简单的实验来检查运行 sklearn 时多核 CPU 的重要性GridSearchCV with KNeighborsClassifier。我得到的结果让我感到惊讶，我想知道我是否误解了多核的好处，或者也许我没有做得正确。

2-8 个作业之间的完成时间没有差异。怎么会 ？我注意到 CPU 性能选项卡上的差异。第一个单元运行时，CPU 使用率约为 13%，最后一个单元的 CPU 使用率逐渐增加到 100%。我原以为它会更快完成。也许不是线性更快，即 8 个作业会比 4 个作业快 2 倍，但要快一点。

我是这样设置的：

我使用的是jupyter-notebook，cell指的是jupyter-notebook cell。

我已经加载了 MNIST 并使用0.05测试尺寸为3000a 中的数字X_play.

from sklearn.datasets import fetch_mldata
from sklearn.model_selection import train_test_split

mnist = fetch_mldata('MNIST original')

X, y = mnist["data"], mnist['target']

X_train, X_test, y_train, y_test = X[:60000], X[60000:], y[:60000], y[60000:]
_, X_play, _, y_play = train_test_split(X_train, y_train, test_size=0.05, random_state=42, stratify=y_train, shuffle=True)

在下一个单元格中我进行了设置KNN and a GridSearchCV

from sklearn.neighbors import KNeighborsClassifier
from sklearn.model_selection import GridSearchCV

knn_clf = KNeighborsClassifier()
param_grid = [{'weights': ["uniform", "distance"], 'n_neighbors': [3, 4, 5]}]

然后我为 8 个 n_jobs 值创建了 8 个单元格。我的CPU是i7-4770，4核8线程。

grid_search = GridSearchCV(knn_clf, param_grid, cv=3, verbose=3, n_jobs=N_JOB_1_TO_8)
grid_search.fit(X_play, y_play)

Results

Parallel(n_jobs=1)]: Done  18 out of  18 | elapsed:  2.0min finished
Parallel(n_jobs=2)]: Done  18 out of  18 | elapsed:  1.4min finished
Parallel(n_jobs=3)]: Done  18 out of  18 | elapsed:  1.3min finished
Parallel(n_jobs=4)]: Done  18 out of  18 | elapsed:  1.3min finished
Parallel(n_jobs=5)]: Done  18 out of  18 | elapsed:  1.4min finished
Parallel(n_jobs=6)]: Done  18 out of  18 | elapsed:  1.4min finished
Parallel(n_jobs=7)]: Done  18 out of  18 | elapsed:  1.4min finished
Parallel(n_jobs=8)]: Done  18 out of  18 | elapsed:  1.4min finished

第二次测试

随机森林分类器的使用要好得多。测试尺寸为0.5, 30000 images.

from sklearn.ensemble import RandomForestClassifier

rf_clf = RandomForestClassifier()
param_grid = [{'n_estimators': [20, 30, 40, 50, 60], 'max_features': [100, 200, 300, 400, 500], 'criterion': ['gini', 'entropy']}]

Parallel(n_jobs=1)]: Done 150 out of 150 | elapsed: 110.9min finished
Parallel(n_jobs=2)]: Done 150 out of 150 | elapsed: 56.8min finished
Parallel(n_jobs=3)]: Done 150 out of 150 | elapsed: 39.3min finished
Parallel(n_jobs=4)]: Done 150 out of 150 | elapsed: 35.3min finished
Parallel(n_jobs=5)]: Done 150 out of 150 | elapsed: 36.0min finished
Parallel(n_jobs=6)]: Done 150 out of 150 | elapsed: 34.4min finished
Parallel(n_jobs=7)]: Done 150 out of 150 | elapsed: 32.1min finished
Parallel(n_jobs=8)]: Done 150 out of 150 | elapsed: 30.1min finished

以下是一些原因might成为此行为的原因

• With increasing no. of threads, there is an apparent overhead incurred for intializing and releasing each thread. I ran your code on my i7 7700HQ, I saw the following behaviour with each inceasing n_job
  • when n_job=1 and n_job=2每个线程的时间（GridSearchCV 评估每个模型以完全训练模型并测试模型的时间）为 2.9 秒（总时间约为 2 分钟）
  • when n_job=3，时间为3.4秒（总时间1.4分钟）
  • when n_job=4，时间为3.8秒（总时间58秒）
  • when n_job=5，时间为4.2秒（总时间51秒）
  • when n_job=6，时间为 4.2 秒（总时间约 49 秒）
  • when n_job=7，时间为 4.2 秒（总时间约 49 秒）
  • when n_job=8，时间为 4.2 秒（总时间约 49 秒）

• 现在如您所见，每个线程的时间增加了，但总体时间似乎减少了（尽管超出了n_job=4 the different was not exactly linear) and remained constained withn_jobs>=6` 这是因为初始化和释放线程会产生成本。看这个 github 问题 and 这个问题.

• 此外，可能还存在其他瓶颈，例如数据太大而无法同时广播到所有线程、线程对 RAM（或其他资源等）的抢占、数据如何推送到每个线程等。

• I suggest you to read about Ahmdal's Law which states that there is a theoretical bound on the speedup that can be achieved through parallelization which is given by the formula Image Source : Ahmdal's Law : Wikipedia

• 最后，这可能是由于数据大小和您用于训练的模型的复杂性所致。

Here is 一篇博文解释有关多线程的相同问题。