vectordb = Chroma.from_documents(texts, embeddings) is giving error!

[Nisar-MLNLP]

[Sunil-1234]

Please help me with the same error.
Whenever I am trying for the first time it runs but for the second time it is throwing the above error.
Please help me fix this ASAP.

[Nandan-Jha]

Please help me with same error.
Everytime,I restart the notebook,it is throwing same error.

[Sunil-1234]

This code works for me second time again after restarting it throw same error
class MyDocument:
def init(self, text, metadata=None):
self.text = text
self.metadata = metadata
self.page_content = text # add this line to define page_content attribute
metadata = [{'source': str(i)} for i in range(len(texts))]
documents = [MyDocument(text, metadata=metadata[i]) for i, text in enumerate(texts)]
docsearch = Chroma.from_documents(documents, embeddings)

[tomviner]

The square bracket for builtin types syntax (list[str]) only works from Python 3.9 onwards.

This is fixed in chroma-core/chroma#160 just need to pip install 'chromadb>=0.3.5'

Or use at least Python 3.9

[tonmoyazad]

How to solve this?

While Creating the Chain Getting ValidationError: 1 validation error for LLMChain
prompt
none is not an allowed value (type=type_error.none.not_allowed)

but when I run this on google colab everything works fine.

Please advice what could be the issue here?

[Sujan-Roy]

You can solve the problem by import the following library

from langchain.vectorstores.chroma import Chroma

I face the same problem. I solve my problem by importing the above library.
I hope it will be helpful for you.

[barbmarques]

I continually get the following error, although I've run pip install chromadb on my Mac:

ImportError: Could not import chromadb python package. Please install it with pip install chromadb.

[godofecht]

Looks like they still haven't fixed it. Similar issue with from_texts