Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!
The default user is the root user, so you can install software as needed.
Google Colab notebook does not support terminal but it can be hacked to run JupyterLab.
You can …
Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!
Docker images are saved in /var/lib/docker.
You can link the directory to another place to save images elsewhere.
Another way is to change the configuration file of Docker.
For …
Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!
Point to the JDK directory if you can. Otherwise, point it to the JRE directory.
https://docs.oracle.com/cd/E19182-01/820-7851/inst_cli_jdk_javahome_t/
https://www.quora.com/What-is-the-use-of-JAVA_HOME
Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!
https://unix.stackexchange.com/questions/198590/what-is-a-bind-mount
Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!
User-based Filtering (Memory-based Filtering)
Item-based Filtering (Content-based Filtering)
Non-negative Matrix Factorization
Neural Matrix Factorization
Variational Autoencoder
Hybrid
New methods like VAE, AE, or Deep Collaborative outperform classical methods like NMF on …
Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!
GELU is the best activation function currently (at least in NLP).
,
where \(\Phi(x)\) is the cumulative distribution function of the standard normal distribution.