get latest code from orca-2 space

set repetition_penalty=1.12 for hftgi

refine summarize chain

added refine summary chain

Update llm_loader.py

fixed streaming for ctransformers

added ctransformers

fixed streaming issue

use ConversationChain + ConversationSummaryBufferMemory

fixed bug

clean up code

fixed bug on llama-2

gradio app: support chatting with llama-2

improve sync

refine TGI parmas

fixed bugs for tgi

added support for HF TGI

fixed bug on running openllm

added support for openllm

support Qwen chat model

added telegram bot

fixed bug

fixed bug

support concurrent access

refine llama-2 chat prompt

fixed minor typos

added USE_LLAMA_2_PROMPT_TEMPLATE

fixed bug: qa chain doesn't return any source doc

fixed bug

fixed bug for lc_serve

code complete

fixed bug in gradio app

completed gradio app for qa

WIP

fixed bug for llama-2 auth token handling

tested app_modules/llm_loader.py

refactor code

fxied UI bugs

get latest code from chat-with-pci-dss-v4

fixed bug on metadata url handling

show source document content in logs

add return at end of streaming

switch from Unstructured Loader to PyPDF as its results have page nubmer

added index for 12 AI books

Duplicate from inflaton/chat-with-pci-dss-v4