langchain-samples

A collection of langchain sample scripts

Samples using Ollama

• Install and run Ollama : https://ollama.com
• Create a local virtualenv, install python requirements and pull mistral llm :

python -m venv .venv
source .venv/bin/activate
pip3 install -r requirements
ollama pull mistral

langtest-basic-local

Run a python script with a simple question to mistral :

python3 ollama/langtest-basic-local.py

langtest-streamlit-simple

Ask questions to mistral using streamlit ui.

streamlit run ollama/langtest-streamlit-simple.py

langtest-conversationchain.py

Ask questions to llm using a ConversationChain.

streamlit run ollama/langtest-conversationchain.py

langtest-app-youtube

From the example of Nicholas Renotte : https://www.youtube.com/watch?v=MlK6SIjcjE8
Uses LLMChain, PromptTemplate, ConversationMemoryBuffer and WikipediaAPIWrapper to build an app generating a YouTube video title and script

streamlit run ollama/langtest-app-youtube.py

Samples using HuggingFace

Create a .env file with huggingface API Token generated on https://huggingface.co and source it

echo -e "export HUGGINGFACEHUB_API_TOKEN=\"<TOKEN>\"" >> .venv/bin/activate && source .venv/bin/activate

langtest-wikipedia-agent

Simple app to get information from wikipedia since how long do we know about a topic. Demonstrates the use of a wikipedia langchain agent/tool.

streamlit run huggingface/langtest-wikipedia-agent.py

Demonstrates the use of a wikipedia langchain agent/tool.

langtest-prompt-multi-inputs

Simple app to get the winner country of a world cup finals Demonstrates the use of Prompt and LLMChain with multiple inputs

streamlit run huggingface/langtest-prompt-multi-inputs.py