Resizing disk is needed. Default disk space is 8GB and this is totally not enough. I think 16GB is enough to run.
Download script in above link and run.
./resize.sh 16
Run disk resizing script in above link. And before running disk resizing, Cloud9 needs to set aws access key and secret key. You need to configure this things.
aws configure
git clone https://github.com/fastai/fastai
wget https://repo.continuum.io/miniconda/Miniconda3-latest-Linux-x86_64.sh
chmod a+x Miniconda3-latest-Linux-x86_64.sh
./Miniconda3-latest-Linux-x86_64.sh
Download dataset and unzip to "fastai/data/bulldozers/". Check below movie to download dataset. https://youtu.be/CzdWqFTmn0Y?t=939
unzip bulldoz.zip
Activate python3 virtual env. FastAI needs python3. I had a trouble because of python2 comparability.
conda create -n fastaienv python=3.6 anaconda
Add uninstalled packages. Cloud9 has default python version2.7. So that you need to specify python.
conda install -c anaconda bcolz
conda update -n base conda
conda install feather-format -c conda-forge
python3 -m pip install opencv-python
python3 -m pip install graphviz
python3 -m pip install sklearn_pandas
python3 -m pip install isoweek
python3 -m pip install pandas_summary
jupyter notebook --ip=0.0.0.0 --port=8080 --no-browser
Run Preview - Preview Running Application on Cloud9 then you can see one browser opened. And you can get accees link like below. Or you can check your instance id in your browser(example : https://ap-southeast-1.console.aws.amazon.com/cloud9/ide/[instance_id])
http://[your instance id].vfs.cloud9.ap-southeast-1.amazonaws.com/
Add above port.(Below link is an example)
http://[your instance id].vfs.cloud9.ap-southeast-1.amazonaws.com:8080/
In terminal, you can get logged access token. Insert the token to login.
http://[your instance id].vfs.cloud9.ap-southeast-1.amazonaws.com:8080/tree?token=[token]
DONE!
https://medium.com/@GuruAtWork/fast-ai-lesson-1-7fc38e978d37 https://github.com/fastai/fastai/blob/master/README.md#installation-issues