Building a Chatbot

Question

Hello there, happy Friday. I am trying to follow the python chatbot tutorial and I'm getting a LookupError on this line

year_docs = loader.loaddata( file=Path(f"./data/UBER/UBER{year}.html"), split_documents=False
)

... full screenshot attached. Any ideas?

Logan M · Answer

You've excluded the code and the end of the traceback (what loader is this, for example?)

gumjatang · Answer

my fault, here is the full traceback

gumjatang · Answer

i replaced the unstructredreader with the simpledirectoryreader and it seems to be working okay but still curious why the unstructuredreader is throwing the error

Logan M · Answer

hmmm, feels like an issue with nltk? (clicking the "view as scrollable element" button would give a more complete traceback)

Maybe try with a fresh venv? Or you might need to run nltk download for something. The full traceback would probably help indentify which nltk package is missing

gumjatang · Answer

thanks Logan!

gumjatang · Answer

sorry, new to python and resisting the urge to use the TS lib which im more familiar with

gumjatang · Answer

🙂

gumjatang · Answer

that fixed it. thanks again

Logan M · Answer

Great! (Usually this lib is included automatically with installing llama-index, kind of spooky!)

gumjatang · Answer

i think it tried (?) but i might have accidentally stopped the download when it inquired about cert signing

Find answers from the community

Building a Chatbot