LangChain Tutorial from Deeplearning.ai

Retrieval Augmented Generation (RAG)

• Extract: Documents come in all sorts of file formats (.doc, .pdf, etc.) and contain all sorts of data formats (text, tables, images, movies). These must be extracted and put into a format that can be processed by the next stages.
• Chunking: Text data is broken into smaller chunks – a process inventively named ‘chunking’.
• Embedding: Converting a chunk into a ‘dense vector’ that represents the meaning of the text.
• Loading: Adding the embedding and original data to a database.
• Database
  • The database is going to provide storage for the embedding and data. Often these are vector databases due to the embedding, but graph databases and traditional databases are also used.
• Query
  • Embedding: The query is converted to a dense vector using the same embedding model.
  • Retrieval: The stored and query vectors represent meaning, so retrieval is the process of finding the k entries in the database that are ‘closest’ to the query vector. Lots of details here!
  • k results are provided to an LLM which uses them to form an ‘augmented’ response.

LLM

Ollama

Ollama Models

Ollama Github

Download Ollama

> ollama run llama3
> ollama pull nomic-embed-text

LangChain Chat with Your Data

• Document Loading
• Document Splitting
• Vectorestores and Embedding
• Retrieval
• Question Answering
• Chat

LangChain for LLM Application Development

LangChain: Memory

• ConversationBufferMemory
• ConversationBufferWindowMemory
• ConversationTokenBufferMemory
• ConversationSummaryMemory

LangChain: Chains

• LLMChain
• Sequential Chains
  • SimpleSequentialChain
  • SequentialChain
• Router Chain

LangChain: Evaluation

• Example generation
• Manual evaluation (and debuging)
• LLM-assisted evaluation
• LangChain evaluation platform

Apps

• Find Parking Spots