Skip to content

Langchain Agent finetuning using 7B - LLAMA 2 , on hotpotQA (Retroformer framework)

Notifications You must be signed in to change notification settings

algopapi/RetroformAgent

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

6 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

RETROFORMER: RETROSPECTIVE LARGE LANGUAGE

AGENTS WITH POLICY GRADIENT OPTIMIZATION - Implementation

  • The retroformer model is a longchat-7b-32k model using the fastchat API.

  • Currently implementting the Finetune pipeline.

All credit goes to the original authors of the paper: https://arxiv.org/pdf/2308.02151.pdf

Im just implementing this for the fun of it. (and because i want to see if i can somehow use it in an agent swam)

Stack:

  • Actor agents: Langchain
  • Agent architecture: MRKL (ReAct Zero Shot)
  • Actor model: gpt-4
  • Retro model: longchat-7b-32k
  • Environment: HotpotQA

To start the locally hosted server:

python3 -m fastchat.serve.controller python3 -m fastchat.serve.model_worker --model-names "gpt-3.5-turbo,text-davinci-003,text-embedding-ada-002" --model-path lmsys/vicuna-7b-v1.5 python3 -m fastchat.serve.openai_api_server --host localhost --port 8000

About

Langchain Agent finetuning using 7B - LLAMA 2 , on hotpotQA (Retroformer framework)

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages