General issue for updating the documentation of the aapb-eval repository (this).
Goals:
Provide sufficient and up to date information about how to run evaluations.
Explain any terminology related to the theme/app. e.g. what is "gentle"?
Terminology:
Each subdirectory in the repo contains a "theme" or a "task". The word "theme" has been used previously already. "Task" is another option to mean the same thing.
An example is "forced aligner".
Readmes:
requirements.txt to be created in env.
Correct running instructions (some are out of date)
Where to get needed input datas
details on the paths / input file types / invocation
example command of a run
The workflow should be broken into two steps:
"Running the apps" to get the appropriate mmifs needed. Having the golds correctly.
Actually "Running eval", which assumes mmifs and golds are already present. Staging them in the correct places, either local or online.
Because
General issue for updating the documentation of the aapb-eval repository (this).
Goals: Provide sufficient and up to date information about how to run evaluations. Explain any terminology related to the theme/app. e.g. what is "gentle"?
Terminology: Each subdirectory in the repo contains a "theme" or a "task". The word "theme" has been used previously already. "Task" is another option to mean the same thing. An example is "forced aligner".
Readmes: requirements.txt to be created in env. Correct running instructions (some are out of date) Where to get needed input datas details on the paths / input file types / invocation example command of a run
The workflow should be broken into two steps:
Done when
No response
Additional context
No response