Gen-AI Developer Classroom notes 14/Jul/2025

Langchain contd

• Runnable interface and Refer Here for key methods
• Models in Langchain
  • Text Models
  • Chat Models
• For our purpose we use chat models & all chat models are derived from BaseChatModel, here we get additional capabilities rather than just invoke and stream we can
  
• Chat Models have a method which is bind tools which is for external tool executions, but ensure your model supports this.

SLMs

• We have many models which are SLMs which can be finetuned to perform a specific task which runs on your own servers.

• We have options such as

  • Microsoft phi
  • Google Gemma
  • Mistral
  • LLAMA
  • DeepSeek

• Running a SLM on your system (Ubuntu Linux)

  • WSL (Windows Subsystem for linux)
  • Virtual Machine

Steps for GCP Vm Configuration

• Create a firewall rull called as open-all as described in the class
• Now on your local system

ssh-keygen

• This create public key and private in ~/.ssh/id_ed25519.pub and ~/.ssh/id_ed25519
• Create a vm and copy the ssh public key in ssh-keys section of gcp
• Now lets install ollama on ubuntu
• Ollama pull <model> to get and run the model in the background & communicate with model via command line ollmama run <model>