Replication Package for Compliance Checking

Content description

.
├── Code
├── Data
├── Paper
└── Checkpoint

Code: implementations of all the elements discussed in the program.
- We first use the below LLMs for CausalLM task.
  - Mistral
  - Zephyr
  - Mixtral
- We then use the below LLMs and traditional ML algorithms for Bi-Encoders and Embeddings
  - BERT
  - SVM and Decision Tree
- Finally we have the gradio.py which demonstrates the demo based on the uplodade fine-tuned Mixtral model checkpoint
Data: contains DPA training and test datasets as well as a smaller filtered dataset to test with higher inference speed.
Paper: contains the list of papers reviewed throughout the program.
Checkpoint: The link for fine-tuned Mixtral model
Our Demo presentation is provided at: bit.ly/3SUBDnK

Name		Name	Last commit message	Last commit date
Latest commit History 28 Commits
Checkpoint		Checkpoint
Code		Code
Data		Data
Paper		Paper
Demo.png		Demo.png
README.md		README.md