Span-level grounding verification for RAG, code, and tool-grounded AI outputs.

AdaSeq: An All-in-One Library for Developing State-of-the-Art Sequence Understanding Models

A Simple but Powerful SOTA NER Model | Official Code For Label Supervised LLaMA Finetuning

Transformer-based models implemented in tensorflow 2.x(using keras).

Code and data form the paper BERT Got a Date: Introducing Transformers to Temporal Tagging

Unofficial (Golang) Go bindings for the Hugging Face Inference API

Local-first privacy guard: anonymize your documents before sharing with LLMs.

A collection of datasets for Ukrainian language

Lightweight self-hosted span annotation tool

b站 AI日日新 不定期更新使用Python框架完成机器学习、深度学习、数据科学任务

A deep research study introducing the Gene Drift Hypothesis: a framework explaining how tokenomics mutate across market cycles. Analyzes evolutionary forces, selective pressures, behavioral traits, and economic genes that rise, fall, or mutate through bull/bear phases, shaping token species over time.

A research-grade exploration of the Tokenomics Ecological Framework, analyzing how tokens behave as predator, prey, parasite, and symbiotic species. Examines ecosystem interactions, evolutionary pressures, species population cycles, and the dynamics of economic predation, mutation, drift, and long-term survival across market cycles.

Extracting terms from text using XLM-R for token and sequence classification

A research-grade framework for extracting, classifying, and analyzing the “genetic” behavior of smart contract tokens. Identifies economic traits, supply mutations, fee patterns, permission risks, upgradeability vectors, and scam species using a structured gene taxonomy with risk scoring, HTML reports, and token comparison tools.

Token classification using Phobert Models for Vietnamese

CNER: Concept and Named Entity Recognition

bullet: A Zero-Shot / Few-Shot Learning, LLM Based, text classification framework

Labeled Russian text token-by-token for training models for NER task based samples got from parsing different resources and generated by ChatGPT.

The MERIT Dataset is a fully synthetic, labeled dataset created for training and benchmarking LLMs on Visually Rich Document Understanding tasks. It is also designed to help detect biases and improve interpretability in LLMs, where we are actively working. This repository is actively maintained, and new features are continuously being added.

Applied Deep Learning 深度學習之應用 by Vivian Chen 陳縕儂 at NTU CSIE