问题标签 [cntk]
For questions regarding programming in ECMAScript (JavaScript/JS) and its various dialects/implementations (excluding ActionScript). Note JavaScript is NOT the same as Java! Please include all relevant tags on your question; e.g., [node.js], [jquery], [json], [reactjs], [angular], [ember.js], [vue.js], [typescript], [svelte], etc.
docker - 你如何看待 microsoft/cntk docker 上的 Jupyter notebook？
我刚刚安装了 microsoft/cntk docker 映像，我想知道如何查看 Jupyter notebook 界面（以便在 /cntk/Tutorials 中执行 docker 映像中包含的教程）？
对于另一个 docker 映像，我只需执行以下命令：
docker run -d -p 8888:8888 -v /home/name/notebooks:/home/ds/notebooks dataquestio/python3-starter
然后我打开浏览器并转到 localhost:8888
谢谢
visual-studio-2015 - 了解 CNTK 2.0 Beta 8.0 源代码如何使用 Visual Studio 2015 在 Windows 10 上运行？
问题
我是一名学生，.NET 开发人员，我会感谢 CNTK 重量级贡献者和设计师的建议。我正在阅读文档和教程，实际的源代码令人印象深刻。
在 Visual Studio 2015 中使用的最佳 api 是什么，以便我可以进入 CNTK 中的功能，看看微软是如何编写多层感知器的？通常，对于 .NET 项目，我会编写一个我理解的示例应用程序，然后在调试器上浏览源代码。这是我学习的最好方法。我的问题是针对 Visual Studio 2015，但我愿意走出自己的舒适区，了解最佳实践是否是在不同的 IDE 甚至操作系统上执行此操作。
我计划按原样使用该工具，但我想尽可能熟悉该工具，因为我想开始在 Caffe、TensorFlow 等上使用它，因为我是一名 .NET 开发人员。
尝试
我已经阅读了有关 CTNK 的教程，发现自己经常使用大脑脚本，但我经常在命令行上调用它。我知道 Visual Studio 2015 中有一个 python api，我想知道这是否是使用 CNTK 访问后端的首选方法？
请不要对此投反对票。我已经阅读了提供的文档，我觉得在 GitHub 存储库上问这个问题是不合适的，这是我觉得问这个问题的唯一其他场所。如果有其他渠道或论坛可以向 CNTK 提出一般性问题，请告诉我，我会这样做，但如果没有，我感谢您的指导和对认真学习者的友好帮助。
python - Python DLL 加载失败：在 microsoft fastRCNN 脚本上找不到指定的模块
尝试运行 python 脚本时出现以下错误：
Traceback (most recent call last):
File "A1_GenerateInputROIs.py", line 5, in <module>
import PARAMETERS
File "C:\local\CNTK-2-0-beta9-0-Windows-64bit-CPU-Only\cntk\Examples\Image\Detection\FastRCNN\PARAMETERS.py", line 2, in <module>
from cntk_helpers import *
File "C:\local\CNTK-2-0-beta9-0-Windows-64bit-CPU-Only\cntk\Examples\Image\Detection\FastRCNN\cntk_helpers.py", line 7, in <module>
from fastRCNN.nms import nms as nmsPython
File "C:\local\CNTK-2-0-beta9-0-Windows-64bit-CPU-Only\cntk\Examples\Image\Detection\FastRCNN\fastRCNN\__init__.py", line 7, in <module>
from .imdb import imdb
File "C:\local\CNTK-2-0-beta9-0-Windows-64bit-CPU-Only\cntk\Examples\Image\Detection\FastRCNN\fastRCNN\imdb.py", line 19, in <module>
from .utils3_win64.cython_bbox import bbox_overlaps
ImportError: DLL load failed: The specified module could not be found.
我正在关注https://github.com/Microsoft/CNTK/wiki/Object-Detection-using-Fast-R-CNN上的教程。当我尝试在 ubuntu 系统上运行脚本时，我也遇到了类似的错误。
你认为问题出在哪里？
我正在探索的一个假设如下：
因为在教程中，它说：
本教程代码假设您使用的是 64 位版本的 Python 3.4，因为 utils_win64 下的 Fast R-CNN DLL 文件是为此版本预构建的
由于我使用的是 python 3.5 而不是 python 3.4，我开始认为这可能是原因。我会随时通知你我的发现。
谢谢
python - 'This' 函数不等同于（同构）从 CNTK 中的检查点恢复的函数
trainer.restore_from_checkpoint在 CNTK中调用时出现以下异常。
'This' 函数与从检查点恢复的函数不等价（同构）。
我的恢复代码在下面。这些与本文档中提到的创建训练器和保存方法trainer.dnn的结构相同。trainer.save_checkpoint("trainer.dnn")
python - 无法在 Visual Studio 2015 中导入 CNTK
每次我在我的python环境中导入cntk时，我都会收到“DLL加载失败：找不到指定的模块”
我的环境设置正确，我什至在编写代码时自动完成了cntk，但是每当我运行项目时，它总是在第一行失败。
我完全无法解决这个问题，因为我已经按照 CNTK 教程的每一步操作，并且这个问题的github 票仍然是开放的，这绝对没有帮助。
我的 python 路径是 C:\Anaconda3\envs\cntk-py35\Lib\site-packages;F:\cntkInstall\envs\cntk-py34\Lib\site-packages
出于绝望，我刚刚开始添加任何与路径相关的远程python，看看我是否可以让它工作。依然没有。请帮忙！
编辑：这是我得到的错误：
cntk - access trained parameter in CNTK
With a model like this, how can one access the trained parameters like weight and bias of each layer?
Thanks.
cntk - 同一个带有 dim [0, 0] 的矩阵已经在不同设备之间传输了 20 次
在大纪元期间在某种程度上，但是一旦大纪元结束，我就会看到这个关于复制空矩阵的警告。通常是什么导致此警告？
01/23/2017 13:06:49: Epoch[ 1 of 50]-Minibatch[691301-691400]: ce = 0.06757763 * 9404; errs = 1.595% * 9404; time = 14.3775s; samplesPerSecond = 654.1
01/23/2017 13:07:04: Epoch[ 1 of 50]-Minibatch[691401-691500]: ce = 0.08411693 * 9784; errs = 1.962% * 9784; time = 15.1554s; samplesPerSecond = 645.6
01/23/2017 13:07:18: Epoch[ 1 of 50]-Minibatch[691501-691600]: ce = 0.07443892 * 9847; errs = 1.696% * 9847; time = 14.1284s; samplesPerSecond = 697.0
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
01/23/2017 13:07:33: Epoch[ 1 of 50]-Minibatch[691601-691700]: ce = 0.07692308 * 9815; errs = 1.854% * 9815; time = 14.4867s; samplesPerSecond = 677.5
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
01/23/2017 13:07:48: Epoch[ 1 of 50]-Minibatch[691701-691800]: ce = 0.08028341 * 9809; errs = 1.906% * 9809; time = 14.7772s; samplesPerSecond = 663.8
01/23/2017 13:08:03: Epoch[ 1 of 50]-Minibatch[691801-691900]: ce = 0.09192892 * 10073; errs = 2.214% * 10073; time = 14.8481s; samplesPerSecond = 678.4
01/23/2017 13:08:17: Epoch[ 1 of 50]-Minibatch[691901-692000]: ce = 0.07414725 * 9616; errs = 1.841% * 9616; time = 14.9059s; samplesPerSecond = 645.1
01/23/2017 13:08:32: Finished Epoch[ 1 of 50]: [Training] ce = 0.08177092 * 67573150; errs = 1.962% * 67573150; totalSamplesSeen = 67573150; learningRatePerSample = 0.0020000001; epochTime=104968s
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
WARNING: The same matrix with dim [0, 0] has been transferred between different devices for 20 times.
python - 如何在使用 CNTK 训练期间直接访问梯度和修改权重（参数）？
我想计算梯度值（每个示例或小批量），并将权重直接修改为任何值（因此我可以用任何方法控制梯度下降，而不仅仅是提供的 sgd / 学习率计划）。我正在使用python接口。
python - CNTK 上的 word2vec CBOW 阅读器实现
我想用负采样实现 CBOW word2vec。我通读了 CNTK 的文档，但找不到可以将句子作为输入和输出2*k + 1单词的阅读器（当前单词和当前单词k左右的上下文单词）。另外，我想在python中实现这个。
是否有在 python 中创建自定义 cntk 文本阅读器的指南？
python - cntk 层中的线性激活函数？
在 CNTK 中，它有 relu、hardmax、softmax、sigmoid 和所有好东西，但我正在构建一个基于回归的算法，最后一层需要预测 2 个或更多回归输出。所以我需要 n 个节点，并且激活只是一次线性激活。我看到我可以将激活设置为无，这实际上是正确的吗？