...
- files.txt stores the path of the input ROOT files;
- labels.txt stores the labels of the input ROOT files in case of classification problems;
model.py stores the definition of the custom ML model to use in the training phase, in the user’s favorite ML framework;
params.json stores the parameters on which MLaaS4HEP is based, e.g. number of events to use, chunk size, batch size, and redirector path for files located in remote storage;
preproc.json stores the definition of preprocessing operations to be applied to data.
...
If you don't want to use MLaaS4HEP in the Google Colab notebook but you want to use your resources, instead of installing all the dependencies you can use the MLaaS4HEP Docker image, i.e. felixfelicislp/mlaas:xrootd_pip. An example of the command to run is the following:
...
- The OAuth2-Proxy server allows to authenticate and authorize users to use the service. The authorization service we chose is https://cms-auth.web.cern.ch, so only CMS users can access this application.
- A MLaaS4HEP_server allows to submit MLaaS4HEP workflows, i.e. to produce trained ML models starting from ROOT files and ML algorithms defined by the user.
- xrootd proxy-cache server and X509 proxy renewer allows allow to create an X.509 proxy and renew it in order to make possible the access to remote ROOT files stored in grid sites.
- TFaaS takes the ML models trained and produced by the MLaaS4HEP server to make inference.
...