使用 AxSearch 运行 Tune 实验

在本教程中，我们将介绍 Ax，并运行一个简单的 Ray Tune 实验。Tune 的搜索算法与 Ax 集成，因此允许您无缝扩展 Ax 优化过程而不会牺牲性能。

Ax 是一个用于优化任何类型实验的平台，包括机器学习实验、A/B 测试和模拟。Ax 可以使用多臂强盗优化来优化离散配置（例如，A/B 测试的变体），并使用贝叶斯优化来优化连续/有序配置（例如，浮点/整数参数）。使用强化学习代理的 A/B 测试和模拟的结果通常表现出较高的噪声。Ax 支持在高噪声环境下表现优于传统贝叶斯优化的最先进算法。Ax 还支持常见于现实世界问题的多目标和约束优化（例如，在不增加数据使用的情况下提高加载时间）。Ax 属于“无导数”和“黑箱”优化的领域。

在这个例子中，我们最小化一个简单的目标，以简要演示如何通过 AxSearch 在 Ray Tune 中使用 AxSearch。需要记住的是，尽管强调机器学习实验，Ray Tune 优化任何隐式或显式目标。在此，我们假设已安装 ax-platform==0.2.4 库，并且 Python 版本 >= 3.7。要了解更多信息，请参考 Ax 网站。

点击下面查看我们这个示例所需的所有导入。 您也可以直接启动一个 Binder 实例来自己运行这个笔记本。 只需点击导航顶部的火箭图标。

Show code cell source Hide code cell source

import numpy as np
import time

import ray
from ray import train, tune
from ray.tune.search.ax import AxSearch

让我们首先定义一个经典的全局优化基准。 这里的形式是为了演示而明确定义的，然而它通常是一个黑箱。 我们故意暂停一段时间（0.02秒）来模拟一个长时间运行的机器学习实验。 这个设置假设我们正在运行多个实验的step，并尝试调整x超参数的6个维度。

def landscape(x):
    """
    哈特曼6D函数，包含6个局部最小值。
    它是开发全局优化算法的经典基准测试。
    """
    alpha = np.array([1.0, 1.2, 3.0, 3.2])
    A = np.array(
        [
            [10, 3, 17, 3.5, 1.7, 8],
            [0.05, 10, 17, 0.1, 8, 14],
            [3, 3.5, 1.7, 10, 17, 8],
            [17, 8, 0.05, 10, 0.1, 14],
        ]
    )
    P = 10 ** (-4) * np.array(
        [
            [1312, 1696, 5569, 124, 8283, 5886],
            [2329, 4135, 8307, 3736, 1004, 9991],
            [2348, 1451, 3522, 2883, 3047, 6650],
            [4047, 8828, 8732, 5743, 1091, 381],
        ]
    )
    y = 0.0
    for j, alpha_j in enumerate(alpha):
        t = 0
        for k in range(6):
            t += A[j, k] * ((x[k] - P[j, k]) ** 2)
        y -= alpha_j * np.exp(-t)
    return y

接下来，我们的 objective 函数接受一个 Tune config，在训练循环中评估我们实验的 landscape，并使用 train.report 将 landscape 报告回 Tune。

def objective(config):
    for i in range(config["iterations"]):
        x = np.array([config.get("x{}".format(i + 1)) for i in range(6)])
        train.report(
            {"timesteps_total": i, "landscape": landscape(x), "l2norm": np.sqrt((x ** 2).sum())}
        )
        time.sleep(0.02)

接下来我们定义一个搜索空间。关键的假设是最优的超参数位于这个空间内。然而，如果空间非常大，那么在短时间内可能很难找到这些超参数。

search_space = {
    "iterations":100,
    "x1": tune.uniform(0.0, 1.0),
    "x2": tune.uniform(0.0, 1.0),
    "x3": tune.uniform(0.0, 1.0),
    "x4": tune.uniform(0.0, 1.0),
    "x5": tune.uniform(0.0, 1.0),
    "x6": tune.uniform(0.0, 1.0)
}

现在我们定义来自 AxSearch 的搜索算法。如果您想要限制参数或甚至结果空间，可以通过以下方式轻松实现。

algo = AxSearch(
    parameter_constraints=["x1 + x2 <= 2.0"],
    outcome_constraints=["l2norm <= 1.25"],
)

我们还使用 ConcurrencyLimiter 将并发试验限制为4个。

algo = tune.search.ConcurrencyLimiter(algo, max_concurrent=4)

样本数量是将要尝试的超参数组合的数量。此调优运行设置为 1000 个样本。如果在您的机器上运行时间过长，可以减少此数量，或者可以通过在 train.RunConfig() 中设置 stop 参数轻松设置时间限制，具体如下所示。

num_samples = 100
stop_timesteps = 200

最后，我们进行实验以寻找提供的景观的全局最小值（该景观包含5个伪最小值）。作为度量的参数 "landscape" 通过 objective 函数的 session.report 提供。实验 "min" 通过 algo 在 search_space 中搜索，最小化 landscape 的 “mean_loss”，进行 num_samples 次，或在 "timesteps_total": stop_timesteps 时停止。前一句充分描述了我们旨在解决的搜索问题。考虑到这一点，请注意执行 tuner.fit() 的高效性。

tuner = tune.Tuner(
    objective,
    tune_config=tune.TuneConfig(
        metric="landscape",
        mode="min",
        search_alg=algo,
        num_samples=num_samples,
    ),
    run_config=train.RunConfig(
        name="ax",
        stop={"timesteps_total": stop_timesteps}
    ),
    param_space=search_space,
)
results = tuner.fit()

[INFO 07-22 15:04:18] ax.service.ax_client: Starting optimization with verbose logging. To disable logging, set the `verbose_logging` argument to `False`. Note that float values in the logs are rounded to 6 decimal points.
[INFO 07-22 15:04:18] ax.service.utils.instantiation: Created search space: SearchSpace(parameters=[FixedParameter(name='iterations', parameter_type=INT, value=100), RangeParameter(name='x1', parameter_type=FLOAT, range=[0.0, 1.0]), RangeParameter(name='x2', parameter_type=FLOAT, range=[0.0, 1.0]), RangeParameter(name='x3', parameter_type=FLOAT, range=[0.0, 1.0]), RangeParameter(name='x4', parameter_type=FLOAT, range=[0.0, 1.0]), RangeParameter(name='x5', parameter_type=FLOAT, range=[0.0, 1.0]), RangeParameter(name='x6', parameter_type=FLOAT, range=[0.0, 1.0])], parameter_constraints=[ParameterConstraint(1.0*x1 + 1.0*x2 <= 2.0)]).
[INFO 07-22 15:04:18] ax.modelbridge.dispatch_utils: Using Bayesian optimization since there are more ordered parameters than there are categories for the unordered categorical parameters.
[INFO 07-22 15:04:18] ax.modelbridge.dispatch_utils: Using Bayesian Optimization generation strategy: GenerationStrategy(name='Sobol+GPEI', steps=[Sobol for 12 trials, GPEI for subsequent trials]). Iterations after 12 will take longer to generate due to  model-fitting.
Detected sequential enforcement. Be sure to use a ConcurrencyLimiter.

== Status ==
Current time: 2022-07-22 15:04:35 (running for 00:00:16.56)
Memory usage on this node: 9.9/16.0 GiB
Using FIFO scheduling algorithm.
Resources requested: 0/16 CPUs, 0/0 GPUs, 0.0/5.13 GiB heap, 0.0/2.0 GiB objects
Current best trial: 34b7abda with landscape=-1.6624439263544026 and parameters={'iterations': 100, 'x1': 0.26526361983269453, 'x2': 0.9248840995132923, 'x3': 0.15171580761671066, 'x4': 0.43602637108415365, 'x5': 0.8573104059323668, 'x6': 0.08981018699705601}
Result logdir: /Users/kai/ray_results/ax
Number of trials: 10/10 (10 TERMINATED)

Trial name	status	loc	iterations	x1	x2	x3	x4	x5	x6	iter	total time (s)	ts	landscape	l2norm
objective_2dfbe86a	TERMINATED	127.0.0.1:44721	100	0.0558336	0.0896192	0.958956	0.234474	0.174516	0.970311	100	2.57372	99	-0.805233	1.39917
objective_2fa776c0	TERMINATED	127.0.0.1:44726	100	0.744772	0.754537	0.0950125	0.273877	0.0966829	0.368943	100	2.6361	99	-0.11286	1.16341
objective_2fabaa1a	TERMINATED	127.0.0.1:44727	100	0.405704	0.374626	0.935628	0.222185	0.787212	0.00812439	100	2.62393	99	-0.11348	1.35995
objective_2faee7c0	TERMINATED	127.0.0.1:44728	100	0.664728	0.207519	0.359514	0.704578	0.755882	0.812402	100	2.62069	99	-0.0119837	1.53035
objective_313d3d3a	TERMINATED	127.0.0.1:44747	100	0.0418746	0.992783	0.906027	0.594429	0.825393	0.646362	100	3.16233	99	-0.00677976	1.80573
objective_32c9acd8	TERMINATED	127.0.0.1:44726	100	0.126064	0.703408	0.344681	0.337363	0.401396	0.679202	100	3.12119	99	-0.904622	1.16864
objective_32cf8ca2	TERMINATED	127.0.0.1:44756	100	0.0910936	0.304138	0.869848	0.405435	0.567922	0.228608	100	2.70791	99	-0.146532	1.18178
objective_32d8dd20	TERMINATED	127.0.0.1:44758	100	0.603178	0.409057	0.729056	0.0825984	0.572948	0.508304	100	2.64158	99	-0.247223	1.28691
objective_34adf04a	TERMINATED	127.0.0.1:44768	100	0.454189	0.271772	0.530871	0.991841	0.691843	0.472366	100	2.70327	99	-0.0132915	1.49917
objective_34b7abda	TERMINATED	127.0.0.1:44771	100	0.265264	0.924884	0.151716	0.436026	0.85731	0.0898102	100	2.68521	99	-1.66244	1.37185

[INFO 07-22 15:04:19] ax.service.ax_client: Generated new trial 0 with parameters {'x1': 0.055834, 'x2': 0.089619, 'x3': 0.958956, 'x4': 0.234474, 'x5': 0.174516, 'x6': 0.970311, 'iterations': 100}.
[INFO 07-22 15:04:22] ax.service.ax_client: Generated new trial 1 with parameters {'x1': 0.744772, 'x2': 0.754537, 'x3': 0.095012, 'x4': 0.273877, 'x5': 0.096683, 'x6': 0.368943, 'iterations': 100}.
[INFO 07-22 15:04:22] ax.service.ax_client: Generated new trial 2 with parameters {'x1': 0.405704, 'x2': 0.374626, 'x3': 0.935628, 'x4': 0.222185, 'x5': 0.787212, 'x6': 0.008124, 'iterations': 100}.
[INFO 07-22 15:04:22] ax.service.ax_client: Generated new trial 3 with parameters {'x1': 0.664728, 'x2': 0.207519, 'x3': 0.359514, 'x4': 0.704578, 'x5': 0.755882, 'x6': 0.812402, 'iterations': 100}.

Result for objective_2dfbe86a:
  date: 2022-07-22_15-04-22
  done: false
  experiment_id: 4ef8a12ac94a4f4fa483ec18e347967f
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.3991721132671366
  landscape: -0.8052333562869153
  node_ip: 127.0.0.1
  pid: 44721
  time_since_restore: 0.00022912025451660156
  time_this_iter_s: 0.00022912025451660156
  time_total_s: 0.00022912025451660156
  timestamp: 1658498662
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 2dfbe86a
  warmup_time: 0.0035619735717773438
  

[INFO 07-22 15:04:24] ax.service.ax_client: Completed trial 0 with data: {'landscape': (-0.805233, None), 'l2norm': (1.399172, None)}.
[INFO 07-22 15:04:24] ax.service.ax_client: Generated new trial 4 with parameters {'x1': 0.041875, 'x2': 0.992783, 'x3': 0.906027, 'x4': 0.594429, 'x5': 0.825393, 'x6': 0.646362, 'iterations': 100}.

Result for objective_2faee7c0:
  date: 2022-07-22_15-04-24
  done: false
  experiment_id: 3699644e85ac439cb7c1a36ed0976307
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.530347488145437
  landscape: -0.011983676977099367
  node_ip: 127.0.0.1
  pid: 44728
  time_since_restore: 0.00022292137145996094
  time_this_iter_s: 0.00022292137145996094
  time_total_s: 0.00022292137145996094
  timestamp: 1658498664
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 2faee7c0
  warmup_time: 0.0027179718017578125
  
Result for objective_2fa776c0:
  date: 2022-07-22_15-04-24
  done: false
  experiment_id: c555bfed13ac43e5b8c8e9f6d4b9b2f7
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.1634068454629019
  landscape: -0.11285961764770336
  node_ip: 127.0.0.1
  pid: 44726
  time_since_restore: 0.000225067138671875
  time_this_iter_s: 0.000225067138671875
  time_total_s: 0.000225067138671875
  timestamp: 1658498664
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 2fa776c0
  warmup_time: 0.0026290416717529297
  
Result for objective_2dfbe86a:
  date: 2022-07-22_15-04-24
  done: true
  experiment_id: 4ef8a12ac94a4f4fa483ec18e347967f
  experiment_tag: 1_iterations=100,x1=0.0558,x2=0.0896,x3=0.9590,x4=0.2345,x5=0.1745,x6=0.9703
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.3991721132671366
  landscape: -0.8052333562869153
  node_ip: 127.0.0.1
  pid: 44721
  time_since_restore: 2.573719024658203
  time_this_iter_s: 0.0251619815826416
  time_total_s: 2.573719024658203
  timestamp: 1658498664
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 2dfbe86a
  warmup_time: 0.0035619735717773438
  
Result for objective_2fabaa1a:
  date: 2022-07-22_15-04-24
  done: false
  experiment_id: eb9287e4fe5f44c7868dc943e2642312
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.3599537840291782
  landscape: -0.11348012497414121
  node_ip: 127.0.0.1
  pid: 44727
  time_since_restore: 0.00022077560424804688
  time_this_iter_s: 0.00022077560424804688
  time_total_s: 0.00022077560424804688
  timestamp: 1658498664
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 2fabaa1a
  warmup_time: 0.0025510787963867188
  

[INFO 07-22 15:04:27] ax.service.ax_client: Completed trial 3 with data: {'landscape': (-0.011984, None), 'l2norm': (1.530347, None)}.
[INFO 07-22 15:04:27] ax.service.ax_client: Generated new trial 5 with parameters {'x1': 0.126064, 'x2': 0.703408, 'x3': 0.344681, 'x4': 0.337363, 'x5': 0.401396, 'x6': 0.679202, 'iterations': 100}.
[INFO 07-22 15:04:27] ax.service.ax_client: Completed trial 1 with data: {'landscape': (-0.11286, None), 'l2norm': (1.163407, None)}.
[INFO 07-22 15:04:27] ax.service.ax_client: Generated new trial 6 with parameters {'x1': 0.091094, 'x2': 0.304138, 'x3': 0.869848, 'x4': 0.405435, 'x5': 0.567922, 'x6': 0.228608, 'iterations': 100}.
[INFO 07-22 15:04:27] ax.service.ax_client: Completed trial 2 with data: {'landscape': (-0.11348, None), 'l2norm': (1.359954, None)}.
[INFO 07-22 15:04:27] ax.service.ax_client: Generated new trial 7 with parameters {'x1': 0.603178, 'x2': 0.409057, 'x3': 0.729056, 'x4': 0.082598, 'x5': 0.572948, 'x6': 0.508304, 'iterations': 100}.

Result for objective_313d3d3a:
  date: 2022-07-22_15-04-27
  done: false
  experiment_id: fa7afd557e154fbebe4f54d8eedb3573
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.805729990121368
  landscape: -0.006779757704679272
  node_ip: 127.0.0.1
  pid: 44747
  time_since_restore: 0.00021076202392578125
  time_this_iter_s: 0.00021076202392578125
  time_total_s: 0.00021076202392578125
  timestamp: 1658498667
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 313d3d3a
  warmup_time: 0.0029790401458740234
  
Result for objective_2faee7c0:
  date: 2022-07-22_15-04-27
  done: true
  experiment_id: 3699644e85ac439cb7c1a36ed0976307
  experiment_tag: 4_iterations=100,x1=0.6647,x2=0.2075,x3=0.3595,x4=0.7046,x5=0.7559,x6=0.8124
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.530347488145437
  landscape: -0.011983676977099367
  node_ip: 127.0.0.1
  pid: 44728
  time_since_restore: 2.6206929683685303
  time_this_iter_s: 0.027359962463378906
  time_total_s: 2.6206929683685303
  timestamp: 1658498667
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 2faee7c0
  warmup_time: 0.0027179718017578125
  
Result for objective_2fa776c0:
  date: 2022-07-22_15-04-27
  done: true
  experiment_id: c555bfed13ac43e5b8c8e9f6d4b9b2f7
  experiment_tag: 2_iterations=100,x1=0.7448,x2=0.7545,x3=0.0950,x4=0.2739,x5=0.0967,x6=0.3689
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.1634068454629019
  landscape: -0.11285961764770336
  node_ip: 127.0.0.1
  pid: 44726
  time_since_restore: 2.6361019611358643
  time_this_iter_s: 0.0264589786529541
  time_total_s: 2.6361019611358643
  timestamp: 1658498667
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 2fa776c0
  warmup_time: 0.0026290416717529297
  
Result for objective_32c9acd8:
  date: 2022-07-22_15-04-27
  done: false
  experiment_id: c555bfed13ac43e5b8c8e9f6d4b9b2f7
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.1686440476629836
  landscape: -0.9046216637367911
  node_ip: 127.0.0.1
  pid: 44726
  time_since_restore: 0.00020194053649902344
  time_this_iter_s: 0.00020194053649902344
  time_total_s: 0.00020194053649902344
  timestamp: 1658498667
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 32c9acd8
  warmup_time: 0.0026290416717529297
  
Result for objective_2fabaa1a:
  date: 2022-07-22_15-04-27
  done: true
  experiment_id: eb9287e4fe5f44c7868dc943e2642312
  experiment_tag: 3_iterations=100,x1=0.4057,x2=0.3746,x3=0.9356,x4=0.2222,x5=0.7872,x6=0.0081
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.3599537840291782
  landscape: -0.11348012497414121
  node_ip: 127.0.0.1
  pid: 44727
  time_since_restore: 2.623929977416992
  time_this_iter_s: 0.032716989517211914
  time_total_s: 2.623929977416992
  timestamp: 1658498667
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 2fabaa1a
  warmup_time: 0.0025510787963867188
  
Result for objective_32d8dd20:
  date: 2022-07-22_15-04-30
  done: false
  experiment_id: 171527593b0f4cbf941c0a03faaf0953
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.2869105702896437
  landscape: -0.24722262157458608
  node_ip: 127.0.0.1
  pid: 44758
  time_since_restore: 0.00021886825561523438
  time_this_iter_s: 0.00021886825561523438
  time_total_s: 0.00021886825561523438
  timestamp: 1658498670
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 32d8dd20
  warmup_time: 0.002732992172241211
  
Result for objective_32cf8ca2:
  date: 2022-07-22_15-04-29
  done: false
  experiment_id: 37610500f6df493aae4e7e46bb21bf09
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.1817810425508524
  landscape: -0.14653248187442922
  node_ip: 127.0.0.1
  pid: 44756
  time_since_restore: 0.00025081634521484375
  time_this_iter_s: 0.00025081634521484375
  time_total_s: 0.00025081634521484375
  timestamp: 1658498669
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 32cf8ca2
  warmup_time: 0.0032138824462890625
  

[INFO 07-22 15:04:30] ax.service.ax_client: Completed trial 4 with data: {'landscape': (-0.00678, None), 'l2norm': (1.80573, None)}.
[INFO 07-22 15:04:30] ax.service.ax_client: Generated new trial 8 with parameters {'x1': 0.454189, 'x2': 0.271772, 'x3': 0.530871, 'x4': 0.991841, 'x5': 0.691843, 'x6': 0.472366, 'iterations': 100}.
[INFO 07-22 15:04:30] ax.service.ax_client: Completed trial 5 with data: {'landscape': (-0.904622, None), 'l2norm': (1.168644, None)}.
[INFO 07-22 15:04:30] ax.service.ax_client: Generated new trial 9 with parameters {'x1': 0.265264, 'x2': 0.924884, 'x3': 0.151716, 'x4': 0.436026, 'x5': 0.85731, 'x6': 0.08981, 'iterations': 100}.

Result for objective_313d3d3a:
  date: 2022-07-22_15-04-30
  done: true
  experiment_id: fa7afd557e154fbebe4f54d8eedb3573
  experiment_tag: 5_iterations=100,x1=0.0419,x2=0.9928,x3=0.9060,x4=0.5944,x5=0.8254,x6=0.6464
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.805729990121368
  landscape: -0.006779757704679272
  node_ip: 127.0.0.1
  pid: 44747
  time_since_restore: 3.1623308658599854
  time_this_iter_s: 0.02911996841430664
  time_total_s: 3.1623308658599854
  timestamp: 1658498670
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 313d3d3a
  warmup_time: 0.0029790401458740234
  
Result for objective_32c9acd8:
  date: 2022-07-22_15-04-30
  done: true
  experiment_id: c555bfed13ac43e5b8c8e9f6d4b9b2f7
  experiment_tag: 6_iterations=100,x1=0.1261,x2=0.7034,x3=0.3447,x4=0.3374,x5=0.4014,x6=0.6792
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.1686440476629836
  landscape: -0.9046216637367911
  node_ip: 127.0.0.1
  pid: 44726
  time_since_restore: 3.1211891174316406
  time_this_iter_s: 0.02954697608947754
  time_total_s: 3.1211891174316406
  timestamp: 1658498670
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 32c9acd8
  warmup_time: 0.0026290416717529297
  

[INFO 07-22 15:04:32] ax.service.ax_client: Completed trial 7 with data: {'landscape': (-0.247223, None), 'l2norm': (1.286911, None)}.
[INFO 07-22 15:04:32] ax.service.ax_client: Completed trial 6 with data: {'landscape': (-0.146532, None), 'l2norm': (1.181781, None)}.

Result for objective_32d8dd20:
  date: 2022-07-22_15-04-32
  done: true
  experiment_id: 171527593b0f4cbf941c0a03faaf0953
  experiment_tag: 8_iterations=100,x1=0.6032,x2=0.4091,x3=0.7291,x4=0.0826,x5=0.5729,x6=0.5083
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.2869105702896437
  landscape: -0.24722262157458608
  node_ip: 127.0.0.1
  pid: 44758
  time_since_restore: 2.6415798664093018
  time_this_iter_s: 0.026781082153320312
  time_total_s: 2.6415798664093018
  timestamp: 1658498672
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 32d8dd20
  warmup_time: 0.002732992172241211
  
Result for objective_32cf8ca2:
  date: 2022-07-22_15-04-32
  done: true
  experiment_id: 37610500f6df493aae4e7e46bb21bf09
  experiment_tag: 7_iterations=100,x1=0.0911,x2=0.3041,x3=0.8698,x4=0.4054,x5=0.5679,x6=0.2286
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.1817810425508524
  landscape: -0.14653248187442922
  node_ip: 127.0.0.1
  pid: 44756
  time_since_restore: 2.707913875579834
  time_this_iter_s: 0.027456998825073242
  time_total_s: 2.707913875579834
  timestamp: 1658498672
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 32cf8ca2
  warmup_time: 0.0032138824462890625
  
Result for objective_34adf04a:
  date: 2022-07-22_15-04-33
  done: false
  experiment_id: 4f65c5b68f5c49d98fda388e37c83deb
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.4991655675380078
  landscape: -0.01329150870283869
  node_ip: 127.0.0.1
  pid: 44768
  time_since_restore: 0.00021600723266601562
  time_this_iter_s: 0.00021600723266601562
  time_total_s: 0.00021600723266601562
  timestamp: 1658498673
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 34adf04a
  warmup_time: 0.0027239322662353516
  
Result for objective_34b7abda:
  date: 2022-07-22_15-04-33
  done: false
  experiment_id: f135a2c40f5644ba9d2ae096a9dd10e0
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 1
  l2norm: 1.3718451333547932
  landscape: -1.6624439263544026
  node_ip: 127.0.0.1
  pid: 44771
  time_since_restore: 0.0002338886260986328
  time_this_iter_s: 0.0002338886260986328
  time_total_s: 0.0002338886260986328
  timestamp: 1658498673
  timesteps_since_restore: 0
  timesteps_total: 0
  training_iteration: 1
  trial_id: 34b7abda
  warmup_time: 0.002721071243286133
  

[INFO 07-22 15:04:35] ax.service.ax_client: Completed trial 8 with data: {'landscape': (-0.013292, None), 'l2norm': (1.499166, None)}.
[INFO 07-22 15:04:35] ax.service.ax_client: Completed trial 9 with data: {'landscape': (-1.662444, None), 'l2norm': (1.371845, None)}.

Result for objective_34adf04a:
  date: 2022-07-22_15-04-35
  done: true
  experiment_id: 4f65c5b68f5c49d98fda388e37c83deb
  experiment_tag: 9_iterations=100,x1=0.4542,x2=0.2718,x3=0.5309,x4=0.9918,x5=0.6918,x6=0.4724
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.4991655675380078
  landscape: -0.01329150870283869
  node_ip: 127.0.0.1
  pid: 44768
  time_since_restore: 2.7032668590545654
  time_this_iter_s: 0.029300928115844727
  time_total_s: 2.7032668590545654
  timestamp: 1658498675
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 34adf04a
  warmup_time: 0.0027239322662353516
  
Result for objective_34b7abda:
  date: 2022-07-22_15-04-35
  done: true
  experiment_id: f135a2c40f5644ba9d2ae096a9dd10e0
  experiment_tag: 10_iterations=100,x1=0.2653,x2=0.9249,x3=0.1517,x4=0.4360,x5=0.8573,x6=0.0898
  hostname: Kais-MacBook-Pro.local
  iterations_since_restore: 100
  l2norm: 1.3718451333547932
  landscape: -1.6624439263544026
  node_ip: 127.0.0.1
  pid: 44771
  time_since_restore: 2.6852078437805176
  time_this_iter_s: 0.029579877853393555
  time_total_s: 2.6852078437805176
  timestamp: 1658498675
  timesteps_since_restore: 0
  timesteps_total: 99
  training_iteration: 100
  trial_id: 34b7abda
  warmup_time: 0.002721071243286133
  

现在我们已经找到了最小化平均损失的超参数。

print("Best hyperparameters found were: ", results.get_best_result().config)

Best hyperparameters found were:  {'iterations': 100, 'x1': 0.26526361983269453, 'x2': 0.9248840995132923, 'x3': 0.15171580761671066, 'x4': 0.43602637108415365, 'x5': 0.8573104059323668, 'x6': 0.08981018699705601}