mlfmu - CLI interface¶

mlfmu is a command line tool to build FMUs from ONNX ML models.
Check the README and docs for more info.
You can also run `mlfmu <command> --help` for more info on a specific command.

mlfmu [-h] command ...

mlfmu options¶

-h, --help - show this help message and exit

mlfmu build¶

Build FMU from interface and model files

mlfmu build [-h] [-q | -v] [-l LOG] [-ll {DEBUG,INFO,WARNING,ERROR,CRITICAL}]
            [-i INTERFACE_FILE] [-m MODEL_FILE] [-f FMU_PATH]

mlfmu build options¶

-h, --help - show this help message and exit
-q, --quiet - console output will be quiet.
-v, --verbose - console output will be verbose.
-l LOG, --log LOG - name of log file. If specified, this will activate logging to file. If not, it does not log to file, only console. (default: None)
-ll LOG_LEVEL, --log-level LOG_LEVEL - log level applied to logging to file. Default: WARNING. (default: WARNING)
-i INTERFACE_FILE, --interface-file INTERFACE_FILE - JSON file describing the FMU following schema (default: None)
-m MODEL_FILE, --model-file MODEL_FILE - ONNX file containing the ML Model (default: None)
-f FMU_PATH, --fmu-path FMU_PATH - Path to where the built FMU should be saved (default: None)

mlfmu codegen¶

Generate FMU source code from interface and model files

mlfmu codegen [-h] [-q | -v] [-l LOG] [-ll {DEBUG,INFO,WARNING,ERROR,CRITICAL}]
              [--interface-file INTERFACE_FILE] [--model-file MODEL_FILE]
              [--fmu-source-path FMU_SOURCE_PATH]

mlfmu codegen options¶

-h, --help - show this help message and exit
-q, --quiet - console output will be quiet.
-v, --verbose - console output will be verbose.
-l LOG, --log LOG - name of log file. If specified, this will activate logging to file. If not, it does not log to file, only console. (default: None)
-ll LOG_LEVEL, --log-level LOG_LEVEL - log level applied to logging to file. Default: WARNING. (default: WARNING)
--interface-file INTERFACE_FILE - json file describing the FMU following schema (e.g. interface.json). (default: None)
--model-file MODEL_FILE - onnx file containing the ML Model (e.g. example.onnx). (default: None)
--fmu-source-path FMU_SOURCE_PATH - Path to where the generated FMU source code should be saved. Given path/to/folder the files can be found in path/to/folder/[FmuName] (default: None)

mlfmu compile¶

Build FMU from FMU source code

mlfmu compile [-h] [-q | -v] [-l LOG] [-ll {DEBUG,INFO,WARNING,ERROR,CRITICAL}]
              [--fmu-source-path FMU_SOURCE_PATH] [--fmu-path FMU_PATH]

mlfmu compile options¶

-h, --help - show this help message and exit
-q, --quiet - console output will be quiet.
-v, --verbose - console output will be verbose.
-l LOG, --log LOG - name of log file. If specified, this will activate logging to file. If not, it does not log to file, only console. (default: None)
-ll LOG_LEVEL, --log-level LOG_LEVEL - log level applied to logging to file. Default: WARNING. (default: WARNING)
--fmu-source-path FMU_SOURCE_PATH - Path to the folder where the FMU source code is located. The folder needs to have the same name as the FMU. E.g. path/to/folder/[FmuName] (default: None)
--fmu-path FMU_PATH - Path to where the built FMU should be saved. (default: None)

This tool utilizes cppfmu, source code is available at: https://github.com/viproma/cppfmu
_________________mlfmu___________________