由于我们的NAS Bench-2012已扩展到NATS板凳,因此该仓库被弃用且未维护。请使用NATS板凳,该板凳的体系结构信息多于NAS-Bench-2010。
我们提出了一个具有固定搜索空间的算法 - 敏捷的NAS基准(NAS Bench-2012),该搜索空间几乎为任何最新的NAS算法提供了统一的基准测试。我们的搜索空间的设计灵感来自最受欢迎的基于单元格的搜索算法,该算法将单元格表示为有向的无环图。这里的每个边缘都与从预定义的操作集中选择的操作相关联。为了适用于所有NAS算法,NAS-Bench-201中定义的搜索空间包括4个节点和5个相关的操作选项,总共生成15,625个神经细胞候选者。
在此Markdown文件中,我们提供:
对于以下两件事,请使用Autodl-Project:
注意:请使用PyTorch >= 1.2.0
和Python >= 3.6.0
。
您只需键入pip install nas-bench-201
即可安装我们的API。请参阅此存储库中nas-bench-201
模块的源代码。
如果您有任何疑问或问题,请在此处发布或给我发电子邮件。
[弃用]可以从Google Drive或Baidu-Wangpan(代码:6U5D)下载NAS-Bench-201的旧基准文件。
[推荐] NAS-Bench-201的最新基准文件( NAS-Bench-201-v1_1-096897.pth
)可以从Google Drive下载。模型重量的文件太大(431G),我需要一些时间上传。请耐心等待,感谢您的理解。
您可以将其移动到所需的任何地方,并将其路径发送到我们的API以进行初始化。
NAS-Bench-201-v1_0-e61699.pth
(2.2G),其中e61699
是该文件的最后六位数字。它包含除每个试验训练的权重外的所有信息。NAS-Bench-201-v1_1-096897.pth
(4.7G),其中096897
是该文件的最后六位数字。与NAS-Bench-201-v1_0-e61699.pth
相比,它包含更多试验的信息,尤其是所有在所有数据集中受过12个时期训练的模型都是可避免的。我们建议使用NAS-Bench-201-v1_1-096897.pth
可以从Google Drive或Baidu-Wangpan下载NAS Bench-201中使用的培训和评估数据(代码:4FG7)。建议将这些数据放入$TORCH_HOME
( ~/.torch/
默认情况下)。如果您想亲自生成NAS BENCH-2010或类似的NAS数据集或培训模型,则需要这些数据。
在我们的测试代码中可以找到更多用法。
from nas_201_api import NASBench201API as API
api = API('$path_to_meta_nas_bench_file')
# Create an API without the verbose log
api = API('NAS-Bench-201-v1_1-096897.pth', verbose=False)
# The default path for benchmark file is '{:}/{:}'.format(os.environ['TORCH_HOME'], 'NAS-Bench-201-v1_1-096897.pth')
api = API(None)
len(api)
和每个体系结构api[i]
的架构数量: num = len(api)
for i, arch_str in enumerate(api):
print ('{:5d}/{:5d} : {:}'.format(i, len(api), arch_str))
# show all information for a specific architecture
api.show(1)
api.show(2)
# show the mean loss and accuracy of an architecture
info = api.query_meta_info_by_index(1) # This is an instance of `ArchResults`
res_metrics = info.get_metrics('cifar10', 'train') # This is a dict with metric names as keys
cost_metrics = info.get_comput_costs('cifar100') # This is a dict with metric names as keys, e.g., flops, params, latency
# get the detailed information
results = api.query_by_index(1, 'cifar100') # a dict of all trials for 1st net on cifar100, where the key is the seed
print ('There are {:} trials for this architecture [{:}] on cifar100'.format(len(results), api[1]))
for seed, result in results.items():
print ('Latency : {:}'.format(result.get_latency()))
print ('Train Info : {:}'.format(result.get_train()))
print ('Valid Info : {:}'.format(result.get_eval('x-valid')))
print ('Test Info : {:}'.format(result.get_eval('x-test')))
# for the metric after a specific epoch
print ('Train Info [10-th epoch] : {:}'.format(result.get_train(10)))
index = api.query_index_by_arch('|nor_conv_3x3~0|+|nor_conv_3x3~0|avg_pool_3x3~1|+|skip_connect~0|nor_conv_3x3~1|skip_connect~2|')
api.show(index)
此字符串|nor_conv_3x3~0|+|nor_conv_3x3~0|avg_pool_3x3~1|+|skip_connect~0|nor_conv_3x3~1|skip_connect~2|
方法:
node-0: the input tensor
node-1: conv-3x3( node-0 )
node-2: conv-3x3( node-0 ) + avg-pool-3x3( node-1 )
node-3: skip-connect( node-0 ) + conv-3x3( node-1 ) + skip-connect( node-2 )
config = api.get_net_config(123, 'cifar10') # obtain the network configuration for the 123-th architecture on the CIFAR-10 dataset
from models import get_cell_based_tiny_net # this module is in AutoDL-Projects/lib/models
network = get_cell_based_tiny_net(config) # create the network from configurration
print(network) # show the structure of this architecture
如果要加载此创建的网络的训练有素的权重,则需要使用api.get_net_param(123, ...)
获取权重,然后将其加载到网络中。
api.get_more_info(...)
可以在培训 /验证 /测试集上返回损失 /准确性 /时间,这非常有帮助。有关更多详细信息,请查看get_more_info函数中的注释。
有关其他用法,请参阅lib/nas_201_api/api.py
。我们在相应功能的评论中提供了一些用法信息。如果未提供您想要的内容,请随时开放讨论问题,我很乐意回答有关NAS Bench-201的任何问题。
在nas_201_api
中,我们定义了三个类: NASBench201API
, ArchResults
, ResultsCount
。
ResultsCount
维护特定试验的所有信息。一个人可以实例化结果并通过以下代码获取信息( 000157-FULL.pth
节省了157-架构的所有试验的所有信息):
from nas_201_api import ResultsCount
xdata = torch.load('000157-FULL.pth')
odata = xdata['full']['all_results'][('cifar10-valid', 777)]
result = ResultsCount.create_from_state_dict( odata )
print(result) # print it
print(result.get_train()) # print the final training loss/accuracy/[optional:time-cost-of-a-training-epoch]
print(result.get_train(11)) # print the training info of the 11-th epoch
print(result.get_eval('x-valid')) # print the final evaluation info on the validation set
print(result.get_eval('x-valid', 11)) # print the info on the validation set of the 11-th epoch
print(result.get_latency()) # print the evaluation latency [in batch]
result.get_net_param() # the trained parameters of this trial
arch_config = result.get_config(CellStructure.str2structure) # create the network with params
net_config = dict2config(arch_config, None)
network = get_cell_based_tiny_net(net_config)
network.load_state_dict(result.get_net_param())
ArchResults
保留了建筑所有试验的所有信息。请参阅以下用法:
from nas_201_api import ArchResults
xdata = torch.load('000157-FULL.pth')
archRes = ArchResults.create_from_state_dict(xdata['less']) # load trials trained with 12 epochs
archRes = ArchResults.create_from_state_dict(xdata['full']) # load trials trained with 200 epochs
print(archRes.arch_idx_str()) # print the index of this architecture
print(archRes.get_dataset_names()) # print the supported training data
print(archRes.get_compute_costs('cifar10-valid')) # print all computational info when training on cifar10-valid
print(archRes.get_metrics('cifar10-valid', 'x-valid', None, False)) # print the average loss/accuracy/time on all trials
print(archRes.get_metrics('cifar10-valid', 'x-valid', None, True)) # print loss/accuracy/time of a randomly selected trial
NASBench201API
是最高的API。请参阅以下用法:
from nas_201_api import NASBench201API as API
api = API('NAS-Bench-201-v1_1-096897.pth') # This will load all the information of NAS-Bench-201 except the trained weights
api = API('{:}/{:}'.format(os.environ['TORCH_HOME'], 'NAS-Bench-201-v1_1-096897.pth')) # The same as the above line while I usually save NAS-Bench-201-v1_1-096897.pth in ~/.torch/.
api.show(-1) # show info of all architectures
api.reload('{:}/{:}'.format(os.environ['TORCH_HOME'], 'NAS-BENCH-201-4-v1.0-archive'), 3) # This code will reload the information 3-th architecture with the trained weights
weights = api.get_net_param(3, 'cifar10', None) # Obtaining the weights of all trials for the 3-th architecture on cifar10. It will returns a dict, where the key is the seed and the value is the trained weights.
要获取培训和评估信息(请参阅此处的评论):
api.get_more_info(112, 'cifar10', None, hp='200', is_random=True)
# Query info of last training epoch for 112-th architecture
# using 200-epoch-hyper-parameter and randomly select a trial.
api.get_more_info(112, 'ImageNet16-120', None, hp='200', is_random=True)
如果您发现NAS-Bench-201有助于您的研究,请考虑引用它:
@inproceedings{dong2020nasbench201,
title = {NAS-Bench-201: Extending the Scope of Reproducible Neural Architecture Search},
author = {Dong, Xuanyi and Yang, Yi},
booktitle = {International Conference on Learning Representations (ICLR)},
url = {https://openreview.net/forum?id=HJxyZkBKDr},
year = {2020}
}