Fixes #63: Implement Customizable Model Selection (Updated Implementation)

[ahmad2b]

Hey @bracesproul! Coming back with a fresh implementation in a new PR (previous: #107 ). I made a clean start by incorporating all the requested changes while avoiding conflicts with main branch updates.

Implemented Changes:

• Switched to initChatModel from langchain for model handling as advised
• Model selection now flows through config.configurable
• All our AI models are neatly organized in constants.ts
• Added support for all specified models:
  • OpenAI: gpt-4o-mini
  • Anthropic: claude-3-haiku-20240307
  • Fireworks: accounts/fireworks/models/llama-v3p1-70b-instruct and accounts/fireworks/models/llama-v3p1-8b-instruct

Targeted Updates:

• Selectively replaced gpt-4o-mini instances while preserving specified nodes (& tool nodes):
  • updateArtifact
  • updateHighlightedText
  • reflectionGraph
  • generateArtifacts
  • rewriteArtifacts
• New threads start fresh with default model
• Model switches create new threads (clean slate policy!)
• Old threads remember their model preferences (they're sentimental)

I created this new PR for a cleaner implementation path, avoiding merge conflicts with recent main branch updates. The core functionality remains as discussed, just with a fresher codebase!

Ready for round 2 of review! Hit me with any feedback - my code and I are ready to adapt 💪

Fix #63

[vercel[bot]]

@ahmad2b is attempting to deploy a commit to the LangChain Team on Vercel.

A member of the Team first needs to authorize it.

[ahmad2b]

Thank you @bracesproul for the approval! I'm glad you liked the PR.

You can find me on Twitter at @mahmad2b. I'd love to be credited during the announcement! 🎉 super excited to see this go live!