Presentation 2

List

1.盛佩瑶

2.黎金宁

3.刘思柒

4.胥拿云

5.杨卓林

6.卢思迪

7.方博慧

8.陈欣昊

9.杨嘉成

Deadline

Please send the paper to TA's mail caodi0207@sjtu.edu.cn before 12/27 24:00.

Please send the PPT to TA's mail caodi0207@sjtu.edu.cn before 12/29 12:00.

DLCourse_AS

DeepLearning Course Assignments

Deadlines

Assignment1: 12/07 23:59(Please update collectSubmission.sh to avoid including the virtual env.)

Assignment2: 01/03 23:59

FullyConnectedNets.ipynb => BatchNormalization.ipynb => Dropout.ipynb => ConvolutionalNetworks.ipynb

Final Project:

A PDF is uploaded to the GitHub describing the task. The data and the script are on the FTP server and you can download the .zip file. (toy_read_py3.py is provided by 陈欣昊.)

The results based on the small training set must be provided, whereas results obtained based on the big training set are optional.

Final scores will be given according to the performances on the evaluation set (refers to the folder ./eval). However, the evaluation set shouldn't be used before the completion of training. That is to say, it is cheating if you use the evaluation set to train your model or choose the optimal model according to the performances on the evaluation set. We will give serious punishment if such behaviors are found.

Submission: (before 01/10)

The source codes of your project

The output file on the evaluation set (refers to toy_read.py)

A PDF report in English(including task definition, detailed approach, analysis and so on)

Score Structure:

The quality of your report(50%)

Results on the evaluation set(20%) + the rank in the class(10%)

Using several DL frameworks(TensorFlow, Pytorch, MXNet, Caffe and so on)(including comparisons of speed, performance in the report) (20% for 2 and a bouns 5% if you use 3 frameworks)

Introduction

DeepLearning Course Home Page: https://speechlab.sjtu.edu.cn/~kyu/node/10

These assignments are retrieved and modified from the assignments of CS231n. Some contents are deleted from their assignments and we also extended additional contents according to needs of our DeepLearning Course.

For each assignment, you should complete and hand in completed worksheet (including its outputs and any supporting code outside of the worksheet) with your assignment submission. If you encounter any difficulties, you may refer to the course notes of CS231n to get some inspiration.

Required tools: Python 2.7+, Jupyter Notebook.

Setup your workspace locally.

Step 1

[Option 1: Use Anaconda]: The preferred approach for installing all the assignment dependencies is to use Anaconda, which is a Python distribution that includes many of the most popular Python packages for science, math, engineering and data analysis. Once you install it you can skip all mentions of requirements and you’re ready to go directly to working on the assignment.

[Option 2: Manual install, virtual environment]: If you’d like to (instead of Anaconda) go with a more manual and risky installation route you will likely want to create a virtual environment for the project. If you choose not to use a virtual environment, it is up to you to make sure that all dependencies for the code are installed globally on your machine.

Let's take assignment1 for example. To set up a virtual environment, run the following:

cd assignment1
sudo pip install virtualenv      # This may already be installed
virtualenv .env                  # Create a virtual environment
source .env/bin/activate         # Activate the virtual environment
pip install -r requirements.txt  # Install dependencies
# Work on the assignment for a while ...
deactivate                       # Exit the virtual environment

if you meet the same situation in issue#4, you could try the following scripts before installing the requirements. (Thanks for Kipsora's contribution)

sudo apt install pkg-config
sudo apt install libncurses5-dev
sudo apt install libjpeg-dev
sudo apt install libfreetype6-dev
sudo apt install blt-dev

Step 2

[Download data]: Once you have the starter code, you will need to download the CIFAR-10 dataset. Run the following from the assignment1 directory:

cd cs231n/datasets
./get_datasets.sh

Step 3

[Start Jupyter Notebook]: After you have the CIFAR-10 data, you should start the Jupyter notebook server from the assignment1 directory. If you are unfamiliar with Jupyter, you can refer to Installing Jupyter Notebook and Running the Notebook.

for example, you may begin working on your assignment like this:

cd assignment1
jupyter notebook --port=5000
# Your browser will open a webpage for you to work on. If not, open 'localhost:5000' by yourselves.
# Click '*.ipynb' in the list displayed in the webpage, remember to keep the outputs of those code cells.

[NOTE]: If you are working in a virtual environment on OSX, you may encounter errors with matplotlib due to the issues described here. You can work around this issue by starting the IPython server using the start_ipython_osx.sh script from the assignment1 directory; the script assumes that your virtual environment is named .env.

Submit your work.

Once you are done working run the collectSubmission.sh script; this will produce a file called assignment1.zip. Rename it to StudentID-StudentName-AssignmentID.zip(for example, 12345-小明-as1.zip).

Upload your assignments to corresponding folder in http://202.120.38.125:9999. Public account: test/test.

Please refer to "上传说明.txt" on the website.

Kipsora/Deep-Learning-Course-2017