increased context to 20, updateed prompt in app.py

changed temparure of gpt model to 1 in app.py

changed container name in app.py

compatible langchain verison

compatible langchain verison

compatible langchain verison

pin specific compatible versions because pip continuously downloading different versions trying to find a compatible combination

removed dockerfile to use hugging face default

ok

added more doc to retreive

move forward

Update app.py

Update app.py

add chatbot presentation and exemple input (#15)

welcome message not pushed on bot (#14)

quick fix in format_message helper function

clean up the sources to create direct access (#13)

change ouptu function of predict as it cannont handle langchain history (#12)

fix bug in predict (#11)

Really need to change the way the reference are outputted (#10)

Some refs are displayed twice, the bot is lacking general context regarding its overall aim (#9)

quick fix AIMessage

ai should start with a message provifing context and guardrail + the output of reference is ugly (#8)

Update app.py

change init model to use cpu

Update README.md

add info on the application startup to catch errors (#7)

avoid installing librairies for gpu (#6)

Update requirements.txt

Update requirements.txt

cahnage machine to cuda

Update requirements.txt

Update requirements.txt

Avoid PyTorch GPU build with full CUDA stack (#5)

Update requirements.txt

fixed embeddings loading

Nee to make app public to be used by third party c.a. github io (#4)

must include references when using retreiver (#3)

fixed bug retreiver name

removed magic methods

bmabla

fixed imports

skip import restrict usage

added reqs

load faiss from azure blb storage

load vector store from azure blob storage

load vector store from azure blob storage

Update requirements.txt

Update requirements.txt

Resolve merge: remove ENSCM_2013_CAZELLES.pdf