This PR is not meant to be merged as-is, but to provide some feedback and bug fixes on specific areas. For our use case, we were interested in the paragraph chunking specifically, these changes were focused on that function.
Fixes:
fixes a ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all() exception while running chunking_text in quint.api.fast caused by incorrect casting of a dataframe (quint.api.fast L110)
Refactors:
quint.api.fastBody is now a dataclass subclassed from pydantic's BaseModel instead of an incorrectly implemented (?) fastapi.Body
Features:
Environment:
adds Pipfile for pipenv setup
Docs:
adds detailed instructions for quint setup within a pipenv
Others:
adds automated Linux deploy script (scripts/deploy.sh)
This PR is not meant to be merged as-is, but to provide some feedback and bug fixes on specific areas. For our use case, we were interested in the paragraph chunking specifically, these changes were focused on that function.
Fixes:
ValueError: The truth value of an array with more than one element is ambiguous. Use a.any() or a.all()
exception while runningchunking_text
inquint.api.fast
caused by incorrect casting of a dataframe (quint.api.fast L110)Refactors:
quint.api.fast
Body
is now a dataclass subclassed frompydantic
'sBaseModel
instead of an incorrectly implemented (?)fastapi.Body
Features:
Environment:
Pipfile
forpipenv
setupDocs:
pipenv
Others: