paper-qa

paper-qa copied to clipboard

This adds support for recognising the citations in the answer in the format of (Citation1, Citation2, Citation3) Currently, only Citation3 would be recognised in this citation.

jan-kubica

https://www.semanticscholar.org/me/research

aiyou9

'ChatOllama' object has no attribute 'model_name'

1

I am trying to use Paper QA with some models locally. When trying to create a `Docs()` object, I get an attribute error saying that the ChatOllama object has to...

BenjaminRosell

minimal example of how to use FAISS: storing, combining, reusing

1

hi! im looking for a solution where i can store my embeddings in an external DB, e.g. FAISS but it is not clear to me how to store and manipulate...

daliagachc

Pydantic throwing an error when following example code

2

I'm just trying to use a local ollama server and a sentence transformers embedding. But this seems to be failing I don't quite understand why. Any hints? Thanks! **Error:** Extra...

afonsoguerra

client.model_name doesn't work for langchain ChatAnthropic

1

When running the below: ``` docs = Docs(llm="langchain", client=ChatAnthropic(temperature=0, model_name="claude-3-haiku-20240307"), prompts=prompts) ``` Error: ``` | File "D:\_Ai\my\free\free\langserve\paper\packages\ask-paper\ask_paper\util.py", line 85, in ask_paper | _docs = Docs(llm="langchain", client=ChatAnthropic(temperature=0, model_name="claude-3-haiku-20240307"), prompts=prompts) | File...

yiouyou

adding support for claude, azure, cohere, replicate

1

Hi @whitead Noticed you're calling OpenAI w/ langchain. I'm working on litellm (simple library to standardize LLM API Calls - https://github.com/BerriAI/litellm) and was wondering if we could be helpful. Added...

krrishdholakia

Docs() generates exception due to langchain AIMessage() not supplying __len__() when I specify a langchain llm

1

I've just noticed that if I specify a langchain llm, eg: return Docs(llm = 'langchain', client = ChatOpenAI(), index_path = index_path) and then try to index a document I get...

maspotts

Hi, I managed to get a local llama version working, but I noticed it's not giving any citations in my test case (this only has 30-40 papers for now before...

alexanderchang1

html/xml tags

1

Do you strip html tags from the documents before embedding them into vectors? And if so do you support this stripping of xml tags also?

pedrocr83