MXNet | 易学教程

The NVIDIA driver on your system is too old

阅读更多关于 The NVIDIA driver on your system is too old

The NVIDIA driver on your system is too old https://www.zhihu.com/question/309583980 原因是我安装的pytorch版本是需要cudav9.2版的，而我电脑上安装的是v9.0版的，不兼容因此，解决方案是安装cudav9.2版，问题就解决了原因分析：刚刚重装的系统，其显卡驱动是由Windows 10自身提供的，可能版本比较旧，需要安装最新的驱动。解决：在NVIDIA官网上重新下载对应的显卡驱动并安装即可（注意：是驱动，并非CUDA）解决方法第一种方法是把显卡驱动更新一下。（1）nvidia-smi -L查看自己的显卡型号。（2）然后在 https://www.nvidia.cn/Download/index.aspx?lang=cn 中下载相应驱动。（3）安装驱动。第二种方法是安装cuda 7.5。（1）首先在 https://developer.nvidia.com/cuda-toolkit-archive 中根据机子环境，下载相应的cuda 7.5安装文件，我下的是sh文件。（2）然后安装cuda 7.5。（3）最后 sudo pip install mxnet-cu75==1.2.1 安装cuda 7.5的mxnet版本即可。来源： oschina 链接： https

解密阿里云大规模深度学习性能优化实践

阅读更多关于解密阿里云大规模深度学习性能优化实践

云栖号资讯：【点击查看更多行业资讯】在这里您可以找到不同行业的第一手的上云资讯，还在等什么，快来！作者 | 阿里云异构计算AI加速负责人游亮近日，斯坦福大学公布了最新的 DAWNBench 深度学习榜单，这是人工智能领域最权威的竞赛之一，是衡量深度学习优化策略、模型架构、软件框架、云和硬件等综合解决方案实力的标准之一。在图像识别（Image Classification on ImageNet）榜单中，阿里云包揽了训练时间、训练成本、推理延迟以及推理成本四项第一。 DAWNBench 官方显示，阿里云异构计算服务训练 ImageNet 128 万张图片仅需 2 分 38 秒，基于含光 800 的 AI 服务识别一张图片仅需 0.0739ms，同时在训练成本和推理成本上也实现世界纪录的突破。此次，阿里云创造四项纪录得益于阿里云自研加速框架 AIACC 及平头哥含光 800 芯片。其中，AIACC 是阿里云自主研发的飞天 AI 加速引擎，首次实现了统一加速 Tensorflow、PyTorch、MxNet 和 Caffe 等主流深度学习框架，在相同的硬件平台下，AIACC 能够显著提升人工智能训练与推理的性能。作为 AIACC 的研发负责人，我将在本文内与大家分享阿里云基于 AIACC 的大规模深度学习应用架构和性能优化实践。大规模分布式训练是未来趋势

How to load a trained MXnet model?

阅读更多关于 How to load a trained MXnet model?

问题 I have trained a network using MXnet, but am not sure how I can save and load the parameters for later use. First I define and train the network: dataIn = mx.sym.var('data') fc1 = mx.symbol.FullyConnected(data=dataIn, num_hidden=100) act1 = mx.sym.Activation(data=fc1, act_type="relu") fc2 = mx.symbol.FullyConnected(data=act1, num_hidden=50) act2 = mx.sym.Activation(data=fc2, act_type="relu") fc3 = mx.symbol.FullyConnected(data=act2, num_hidden=25) act3 = mx.sym.Activation(data=fc3, act_type=

How to load a trained MXnet model?

阅读更多关于 How to load a trained MXnet model?

Import MXNet file in Keras/Tensorflow

阅读更多关于 Import MXNet file in Keras/Tensorflow

问题 I am having trouble finding the answer to this. I have an MXNet file in the form of: model.json and model.params. What is the cleanest way to load the network into a Keras installation with TensorFlow backend? 回答1: Unfortunately, you cannot load native MXNet models into Keras. You can try to convert your model using MMdnn, but depending on complexity of your model it might not work. 来源： https://stackoverflow.com/questions/51732258/import-mxnet-file-in-keras-tensorflow

How does a module/optimizer choose which symbols to optimize?

阅读更多关于 How does a module/optimizer choose which symbols to optimize?

问题 I realize I can make my own module/optimizer to do this, but can existing mxnet modules be told to optimize only a subset of variables? Along those same lines, how does a module determine which symbols to optimize as it is? For example, unlike tensorflow in MXNet, both data and variables to be optimized are just "Variable" symbols, but somehow MXNet only affects the NDArrays for the actual variables and not data NDArrays. How does it check? Is there a naming convention it uses? If so, what is

Image Recognition with Scalar output using CNN MXnet in R

阅读更多关于 Image Recognition with Scalar output using CNN MXnet in R

问题 So I am trying to use image recognition using the mxnet package in R using a CNN to try and predict a scalar output (in my case wait time) based on the image. However, when I do this, I get the same resultant output (it predicts the same number which is probably just the average of all of the results). How do I get it to predict the scalar output correctly. Also, my image has already been pre-processed by greyscaling it and converting into the pixel format below. I am essentially using images

simple example of mxnet model parallelism

阅读更多关于 simple example of mxnet model parallelism

问题 The simple examples in the Guon tutorial for mxnet are very helpful to those of us who are just getting started with mxnet. As yet, there is not a simple example for model parallelism. I see the model parallelism example code for LSTM, but I am new to mxnet and it would help me (and perhaps others) to have a more streamlined example. So, I have created a model parallelism example by working off the regression example in the gluon tutorial, and by mixing in some code from mxnet.gluon.Trainer.

android-ndk: stoi / stof / stod / to_string is not a member of 'std'

阅读更多关于 android-ndk: stoi / stof / stod / to_string is not a member of 'std'

问题 I was compiling mxnet(v0.9.3) amalgamation with android NDK Standalone Toolchain on Ubuntu 14.04 64bit Desktop but met some errors. First error is: arm-linux-androideabi-g++ -std=c++11 -Wall -O3 -msse2 -Wno-unknown-pragmas -funroll-loops -Iinclude -fPIC -M -MT nnvm.o \ -I `pwd`/../ -I `pwd`/../include \ -D__MIN__=0 nnvm.cc > nnvm.d arm-linux-androideabi-g++: error: unrecognized command line option '-msse2' When I deleted '-msse2' option and run makefile again, it can compile more but later I

Mxnet - slow array copy to GPU

阅读更多关于 Mxnet - slow array copy to GPU

问题 My problem: How should I perform fast matrix multiplication in mxnet? My concrete problem: array copy to GPU is slow. What can be done about it? I create random arrays, copy them to the context, and then multiply. import mxnet as mx import mxnet.ndarray as nd from mxnet import profiler profiler.set_config(aggregate_stats=True) ctx = mx.cpu() # create arrays on CPU profiler.set_state('run') a = nd.random.uniform(-1, 1, shape=(10000, 10000), ctx=mx.cpu()) b = nd.random.uniform(-1, 1, shape=