Langchain Express

Explore Neo4J RAG LLM pipeline using IBM BAM https://bam.res.ibm.com/ with custom chains and agents.

Development server

Run npm run watch:deploy for a dev server. The application will automatically reload if you change any of the source files.

To interact with Langchain Express Server, head over to https://github.com/playground/mesh-genai and clone the repo. Run ng serve for a dev server. Hit this url http://localhost:4200/ to bring up the UI.

Navigate to Langchain JS tab:

Settings:     Provide Langchain Express URL(ex: http://localhost:3000).
Loader:       Upload documents to be embedded and store in vector db (.pdf, .txt, .json, .csv)
Try Me:
  Ask Me:     Ask questions related to contents that have been uploaded by specifying the collection name
  Ask Web:    Ask questions by crawling the web with the provided url using retrieval chains/tools
  Ask Agent:  Coming soon  

Vector Storage - ChromaDB

docker run -d --rm --name chromadb -p 8000:8000 -v ./chroma:/chroma/chroma -e IS_PERSISTENT=TRUE -e ANONYMIZED_TELEMETRY=TRUE chromadb/chroma:latest

Known issue

To get around this error for the time being

Error [ERR_REQUIRE_ESM]: require() of ES Module /home/playground/langchain-express/node_modules/@xenova/transformers/src/transformers.js from /home/playground/langchain-express/node_modules/@langchain/community/dist/embeddings/hf_transformers.cjs not supported.

replace line #4 and #82 in /node_modules/@langchain/community/dist/embeddings/hf_transformers.cjs with the following to dynamically import “@xenova/transformers”

Line #4

const transformers_1 = (async() => {
    return await import('@xenova/transformers')  
  })();

Line #82

const pipe = await (this.pipelinePromise ??= (await import("@xenova/transformers")).pipeline("feature-extraction", this.modelName));

TODO

Explore Neo4J Knowledge Graph and custom Agents