- using .env to save some sensitive informations
- here are the variables you need to define in .env.
OLLAMA_API=http://localhost:11434
OLLAMA_MODEL=mistral:latest
MONGO_URI=mongodb://root:root@localhost
PG_URI=postgresql://yourname:password@localhost:5433/db_name
LLAMA_CLOUD_API_KEY=
OPENAI_API_KEY=
- you need to create a .env file mannually under the source code directory.
- the recommendions version of python is 3.11, otherwise it might has some incompatible issues.
- env.py is the basic settings for using local ollama
- 01.foundations.ipynb convered the knowledges of the book chapters from 1 to 5 , including:
- reader,
- splitter,
- phaser,
- meta-data extractor,
- pipepine data ingestion.
- indexing
- storage context
- doc & nodes & index persistence.
- 02.retriever-query.ipynb convered the 6th and 7th chapters.
- retrievers
- VectorStoreIndex retrivers
- DocumentSummaryIndex Retrivers
- TreeIndex Retrivers
- KnowledgeGraphIndex Retrivers
- buiding more advanced retrival machanisms
- native retrieval method
- implementing metadata filters
- using selector for more advanced decesion logic
- transforming and rewriting queries
- more specific sub-queries
- Understanding the concepts of dense and sparse retrieval
- dense retrival
- sparse retrival
- using postprocessor to rerank, transform and fiter nodes
- how postprocessor works
- SimilarlyPostProcessor
- KeywordNodePostProcessor
- PrevNextNodePostProcessor
- LongContextRecorder
- PIINodePostProcessor
- NERPIINodePostProcessor
- MetadataReplacementPostProcessor
- SentenceEmbeddingPostProcessor
- Time-based post processors
- Re-ranking post processors
- final thoughts about ndoe post processors
- understanding response synthesizers
- implementing output parsing techniques
- extracting structured outputs using output parsers
- extracting structured outputs using Pydantic Programs
- building and using query engine
- exploring different methods to building query engines
- advanced uses of the QueryEngine interface.
- retrievers
- 03.chat-agent.ipynb introduces how to build a chat system working with agent.
- chat
- discovering chatengine
- understanding the different chat modes
- agent
- building tools and Tookspec classes for our agent
- understanding reasoning loops
- OpenAIAgent
- ReActAgent
- how do we interact with agents
- enhancing our agents with the help of utility tools
- using the LLMCompiler agent for more advanced scenarios
- using the low-level-agent protocal API
- chat
- 04.customizing-deploying.ipynb
- 05.propmpt-engineering.ipynb