🤗 Transformers: the model-definition framework for state-of-the-art machine learning models in text, vision, audio, and multimodal models, for both inference and training.

The official Meta Llama 3 GitHub site

Open source annotation tool for machine learning practitioners.

Python bindings for llama.cpp

Python client for Baidu Yun (Personal Cloud Storage) 百度云/百度网盘Python客户端

A PyTorch native platform for training generative AI models

A more memory-efficient rewrite of the HF transformers implementation of Llama for use with quantized weights.

Code repo for "WebArena: A Realistic Web Environment for Building Autonomous Agents"

Tranco: An improved top websites ranking

Fast computation of Krippendorff's alpha agreement measure in Python.

A tokenizer and sentence splitter for German and English web and social media texts.

Inference code for facebook LLaMA models with Wrapyfi support

This repository holds the code for PolicyLint and PoliCheck, which identifies internal contradictions within privacy policies and analyzes data flow to privacy policy consistency.

Inference code for LLaMA models

PoliGraph: Automated Privacy Policy Analysis using Knowledge Graphs

Reproducing state-of-the-art results

Artifacts of the paper "Arcanum: Detecting and Evaluating the Privacy Risks of Browser Extensions on Web Pages and Web Content" in USENIX Security Symposium 2024

Toolchain to retrieve and parse privacy policies from websites as described in our paper "Unifying Privacy Policy Detection" by Henry Hosseini, Martin Degeling, Christine Utz, and Thomas Hupperich.…

Source code of PurPliance analysis tool.

insight data engineering fellow project

A simple and efficient llama3 local service deployment solution that supports real-time streaming response and is optimized for common Chinese character garbled characters.

Classifier and Feature Extraction scripts used for the CookieBlock extension.

This tool converts HTML representations of privacy policies to plaintext. Full details of the approach can be found in Appendix A of our PolicyLint paper in USENIX Security 2019.

A script tool which recut the original llama3_70B_instruct model into 2 or 4 shards, so that one can run the model efficiently on a `2x80GB` or `4x40GB` GPUs environments.

Implemented Locality Sensitive Hashing (LSH) algorithm using Jaccard Similarity to find similar movies and measure similarity, precision and recall.