https://github.com/xiandong79/mxnet-tutorial
mxnet-tutorial reorganised by myself.
https://github.com/xiandong79/mxnet-tutorial
mxnet mxnet-notebooks
Last synced: 3 months ago
JSON representation
mxnet-tutorial reorganised by myself.
- Host: GitHub
- URL: https://github.com/xiandong79/mxnet-tutorial
- Owner: xiandong79
- License: apache-2.0
- Created: 2017-04-29T16:29:08.000Z (about 8 years ago)
- Default Branch: master
- Last Pushed: 2017-05-05T04:28:11.000Z (about 8 years ago)
- Last Synced: 2025-01-31T15:34:39.873Z (5 months ago)
- Topics: mxnet, mxnet-notebooks
- Language: Jupyter Notebook
- Size: 1.91 MB
- Stars: 0
- Watchers: 1
- Forks: 2
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Python Notebooks for MXNet
This repo contains various notebooks ranging from basic usages of MXNet to
state-of-the-art deep learning applications.
## Outline
### Basic Concepts
#### Section-1
* NDArray: manipulating multi-dimensional array
* Symbol: symbolic expression for neural networks
* Module : intermediate-level and high-level interface for neural network training and inference.
* Loading data : feeding data into training/inference programs
* Mixed programming: developing training algorithms by using NDArray and Symbol together.
#### Section-2
* MNIST: basic use case of MXNet based on the [MNIST](http://yann.lecun.com/exdb/mnist/) dataset
* Optimizer: In gradient-base optimization algorithms, we update the parameters using the gradients in each iteration. We call this updating function as Optimizer.
* Image Data IO: how to prepare, load and train with image data in MXNet.
* Record IO: the python interface for reading and writing record io files.
### Tutorials
* MNIST: Recognize handwritten digits with multilayer perceptrons and convolutional neural networks
* Recognize image objects with pre-trained model on the full Imagenet dataset that containing more than 10M images and over 10K classes
* Char-LSTM: Generates Obama's speeches with character-level LSTM.
* Matrix Factorization: Recommend movies to users.
### How Tos
* Use a pretrained 50 layers' [Deep Residual Learning](https://arxiv.org/abs/1512.03385)(resnet) model for prediction and feature extraction
* Fine-tune the [Deep Residual Learning](https://arxiv.org/abs/1512.03385)(resnet) model.
* Use a pretrained [Inception-BatchNorm Network](https://arxiv.org/abs/1502.03167).
## How to use
The python notebooks are written in [Jupyter](http://jupyter.org/).
- **View** We can view the notebooks on either
[github](https://github.com/dmlc/mxnet-notebooks/blob/master/python/outline.ipynb)
or
[nbviewer](http://nbviewer.jupyter.org/github/dmlc/mxnet-notebooks/blob/master/python/outline.ipynb). But
note that the former may be failed to render a page, while the latter has
delays to view the recent changes.
- **Run** We can run and modify these notebooks if both [mxnet](http://mxnet.io/get_started/index.html#setup-and-installation) and [jupyter](http://jupyter.org/) are
installed. Here is an [example script](https://gist.github.com/mli/b64322f446b2043e3350ddcbfa5957be) to install all these packages on Ubuntu.
If you have a AWS account, here is an easier way to run the notebooks:
1. Launch a g2.2xlarge or p2.2xlarge instance by using AMI `ami-fe217de9` on N. Virginia (us-east-1). This AMI is built by using [this script](https://gist.github.com/mli/b64322f446b2043e3350ddcbfa5957be). Remember to open the TCP port 8888 in the security group.
2. Once launch is succeed, setup the following variable with proper value
```bash
export HOSTNAME=ec2-107-22-159-132.compute-1.amazonaws.com
export PERM=~/Downloads/my.pem
```
3. Now we should be able to ssh to the machine by
```bash
chmod 400 $PERM
ssh -i $PERM -L 8888:localhost:8888 ubuntu@HOSTNAME
```
Here we forward the EC2 machine's 8888 port into localhost.
4. Clone this repo on the EC2 machine and run jupyter
```bash
git clone https://github.com/dmlc/mxnet-notebooks
jupyter notebook
```
We can optional run `~/update_mxnet.sh` to update MXNet to the newest version.
5. Now we are able to view and edit the notebooks on the browser using the URL: http://localhost:8888/tree/mxnet-notebooks/python/outline.ipynb
## How to develop
Some general guidelines:
- A notebook covers a single concept or application
- Try to be as basic as possible. Put advanced usages at the end, and allow reader to skip it.
- Keep the cell outputs on the notebooks so that readers can see the results without running