minilm

Large-scale Self-supervised Pre-training Across Tasks, Languages, and Modalities

Unattended Lightweight Text Classifiers with LLM Embeddings

Faster, smaller BERT models in just a few lines of code.

Ash — offline survival assistant for iOS. Gemma 4 E2B/E4B fully on-device (text · image · voice) with RAG-grounded answers over 56 emergency-response packs. Built for the Kaggle Gemma 4 Good Hackathon.

Lightweight cross-lingual coreference resolution with spaCy using ONNX Runtime inference of transformer models.

Twitter Sentiment Analysis using all-MiniLM-L6-v2 embeddings.

Biomedical RAG Tool for Gene - Disease Association

Extract skills from job postings & resumes — trained gazetteer+MiniLM+MLP pipeline (73% F1), identical in Python, JS, Ruby, Rust

CPU-friendly text classifier: resume | job_post | other — routes browser captures in AI Career OS

Prototype of a customizable local AI assistant demonstrating modular architecture, secure workspace design, and retrieval-based conversations.

Domain and task adaptive pretrained model (~33M params) built on MiniLM for climate-related NLP tasks.

This project utilises Natural Language Processing to generate questions, create exams and evaluate answers without using an LLM. It uses semantic chunking, retrieval-augmented question/answer generation, and a grading method built on Natural Language Inference and sentence embeddings, grounded in the original source text.

End-to-end multi-agent research system with 6 specialized agents (Planner, Search, Scraper, Retriever, Writer, Evaluator) to separate task planning, live source discovery, grounded synthesis, and output evaluation.

This repository serves as the RAG Project MiniLM Model from Microsoft named as sentence-transformers/all-MiniLM-L6-v2 as the embedding model for my resume , and utilizing OpenAI to answer questions based on the Resume

This project is a Retrieval-Augmented Generation (RAG) pipeline built over a sampled subset of the arXiv academic paper metadata. It was developed as a submission for the TechHub CodeSprint Challenge 2026.

Dashboard Streamlit de scoring crédit explicable + veille NLP comparative BERT vs MiniLM pour la classification de produits e-commerce.

Entrelinhas: leitor PDF Flutter com tradução offline, significados sobre as palavras e comparação contextual A/B com MiniLM.

Industry-grade POC assessing evidence for mapping legacy plant maintenance records to SAP S/4HANA PM catalog codes.

comprehensive solutions for Adobe's Document Intelligence Hackathon 2025, encompassing two distinct challenges focused on advanced PDF processing and persona-driven content analysis. Both implementations adhere to stringent performance requirements including sub-60-second execution times and containerized deployment within 1GB resource constraints

Fully local and open-source AI study companion for lecture PDFs - with slide summarization, smart Q&A, and flashcard creation using LangChain and Hugging Face Transformers.