No project description provided
Project description
Domain Adapted Language Modeling Toolkit
This repository primarily contains code for fine-tuning a fully differential Retrieval Augmented Generation (RAG-end2end) architecture. For the first time in the literature, we modified the initial RAG-end2end model (TACL paper, HuggingFace implementation) to work with decoder-only language models like Llma, Falcon, or GPT. We also incorporated the in-batch negative concept alongside the RAG's marginalization to make the entire process efficient.
-
Inside the Training folder, you'll find two codes to train the RAG-end2end and Retriever with contrastive learning.
-
All evaluations related to the Retriever and the Generator are located in the Evaluation folder.
-
Additionally, we have data processing codes and synthetic data generation code inside the Datasets folder.
Project Setup
Create your virtual environment and install. We suggest pyenv
python -m venv .venv && source .venv/bin/activate
pip install invoke && pyenv rehash
inv install
Train Retriever Only
Train Retriever and Generator Jointly
Arcee Domain Pretrained Models - DPT (Coming Soon)
- Arcee-DPT-PubMed-7b
- Arcee-DPT-Patent-7b
- Arcee-DPT-SEC-7b
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.