Python notebooks and experiments
python tutorial, staging for deep/machine learning course
1️⃣ Lesson 1 2️⃣ Lesson 2 3️⃣ Lesson 3 4️⃣ Lesson 4 5️⃣ lesson 5
Also created for coderbunker deep learning talk sessions
This course is the basic deep learning course that follows closely on Jeremy Howard's fantasic /free /life-changing fast.ai course.
The most of notebooks are just trails we left behind passing on their awesomeness.
Checklist before we start and a reading list
- Fast KMeans by batch with GPU, how to train a 60 minutes kmeans in 3 seconds, with example here
- A PyTorch training wrapper to simplify tracking
Follow the installation instructions
Run the jupyter notebook on anaconda3 environment
Usual pre-requisites for the learning.
python 3.6
numpy == '1.14.3'
pandas == '0.23.0'
tensorflow == '1.8.0'
keras == '2.2.0'
Other versions of above library will probably work.
Assuming your anaconda3 is at ~/anaconda3/
If you don't have any of these, try the following format in the command line:
~/anaconda3/bin/pip install keras==2.2.0
If you are on Mac:
~/anaconda3/bin/pip install torch torchvision
To install PyTorch. For other system, You'll have to visit their homepage to copy/paste the right command to install.
In some lines of code you might see
from forgebox.imports import *
You can do
pip install forgebox
If you want to be a contributor:
mail: b2ray2c@gmail.com
wechat: 417848506 remark:"python4ml"