In this post, I’ll briefly walk through a Jupyter notebook where I code CIFAR-10 Image classification task using Microsoft Azure. The entire notebook in complete detailed description can be found here.
Step I — Import the required libraries
We start by importing some python libraries for performing the classification task — Numpy, Matplotlib and AzureML core along with Pytorch.
import numpy as np
import matplotlib.pyplot as plt
from azureml.core import Workspace
# check core SDK version number
print("Azure ML SDK Version: ", azureml.core.VERSION)from azureml.train.dnn import PyTorch
Step II — Create the Workspace and the Experiment
In this post, I’ll talk about how I implemented the standard Q(λ) Learning Algorithm for the Mountain Car domain.
The code can be found here. It is written in C++ and I utilize the Eigen library for creating the vectors and matrices that store our policy parameters.
Compilation — You can type the following on a linux terminal to compile and run the code. Make sure to provide the path to the Eigen library while compiling.
g++ -std=c++11 -I ~/Documents/eigen qlambda_mountaincar.cpp -o qlambda_mountaincar.o
./qlambda_mountaincar.o <alpha> <epsilon> <lambda>
For this domain, the values I obtained after fine-tuning the hyper-parameters are α…
Searching for specificity without loss of generality