🦙 Integrating LLMs into structured NLP pipelines

本地完整部署ASR(K2)-NLP(Rasa,Spacy)-LLM(Chatglm2)-TTS(Vits)

Converting unstructured text to knowledge graph using spacy LLM and ne04j.

This is the repository for my 2024 Tap Institute Course on spaCy with LLMs

Experiments with LLMs and LatinCy using spacy-llm

FLaNK-python-ExtractCompanyName-processor, Python, NiFi 2, NLP, SpaCy, LLM, Gen AI

🦙🪐 fusion of spacy's supervised learning or rule-based components; spacy-llms engaged: text processing, entity extraction & summaries

Elegant prompt versioning and LLM model configuration with spacy-llm

Learn rule-based models from examples using LLM-powered synthesis. Replace expensive LLM calls with fast, deterministic, inspectable regex, code, or spaCy rules.

Example project showing how to connect spacy-llm with Open WebGUI.

No description available

No description available

No description available

Our project leverages a sophisticated hybrid NLP architecture that brings together Retrieval-Augmented Generation (RAG), Large Language Models (LLMs), and spaCy's advanced linguistic capabilities.

Using Jertehs corpuses to train spacy with tranformer and LLM

Scrapy, WebScaping, XML Parsing, ChatGPT API usage, BERT, SpaCY, ATLAS, LLM's in NLP and Visual Similarity Detection( CLIP, FROMAGE and RESNET)

Comparative NER study: Regex vs spaCy vs DSPy LLM. Measures precision, recall, F1, cost, and latency across synthetically generated records with Streamlit dashboard.

NLP learning (when LLMs weren't popular and we learned by doing 🤖). Tokenization, NER, lemmatization, audio processing & pipelines with spaCy and NLTK.

A high-performance AI security middleware that detects and masks PII using Microsoft Presidio and SpaCy. Features a secure rehydration vault to enable safe enterprise LLM integration.

LLMs often ignore instructions to avoid smart quotes, EM/EN dashes, and other symbols. This macOS menu bar app combines spaCy NLP for context-aware processing with a rule-based system to scrub typographic characters from LLM (or any other) output.

🔒 Proprietary AI Recruitment Suite by Shakeel Qureshi | Automates: sourcing → engagement → screening → retention using LLM/NLP (spaCy, custom models) | TalentSync™, EngageIQ™, FitScore™, TalentRisk™ | Contact: shakeel.techrecruiter@gmail.com

Offline-first legal anonymization tool for GDPR compliant text and file scrubbing. Designed for lawyers and legal professionals with explainable, reproducible results. Supports layered detection using patterns, spaCy, Candle LLMs, and Microsoft Presidio.

AI-based legal contract generator using SpaCy NER, rule-based NLP, and Jinja2 templates—no LLMs needed. Uses the CaseHOLD dataset to guide legal context extraction. Generates clean, structured contracts from natural language input via a simple Flask/Streamlit interface.

NLP app for book content analysis: text samples, word frequencies, wordcloud, entity recognition (story characters). Local LLM Q&A about the book content. Experimental: character personality a extraction and chat with them. Here for front-end: https://bookglancer.streamlit.app/ #nlt #spacy

Mask named entities in your text before sending it to an LLM, then restore them afterwards — all locally in your browser session. This Streamlit app uses spaCy to detect entities, lets you review and tweak replacements, and encode/decode text for privacy.

The Knowledge Graph System is a comprehensive AI-powered pipeline for extracting entities and relationships from unstructured documents (PDF, TXT, DOCX) to build structured knowledge graphs. It leverages natural language processing (NLP) techniques, including SpaCy for rule-based extraction and OpenAI's large language models (LLMs) for query

This project leverages Natural Language Processing (NLP) with spaCy and a Large Language Model (LLM) through the Gemini API to automate the generation of comprehensive test cases from user stories. The generated test cases encompass functional, security, performance, accessibility (WCAG 2.1 compliance), and cross-browser testing scenarios.

A showcase of my work involving natural language processing (USYD SOFT4446). Includes my use of Python and Python libraries PyTorch, spaCy, keras, transformers and TensorFlow to create and make use of Neural Networks (NN), Recurrent Neural Networks (RNN), Named Entity Recognition (NER), Sentiment Analysis, Transformers and LLM Integration.

An NLP/LLM-powered TV-Series-Analysis System for understanding story elements through building character relationships networks and classifying themes of any TV show. Build on techs like python, pytorch, gradio, scrapy, pandas, numpy, matplotlib, nltk, networkx, spacy's en_core_web_trf & hugging face's facebook/bart-large-mnli transformer-models

Easy to do complex stuff using spacy-llm