hmnshudhmn24

A fine-tuned T5-small model trained on the CNN/DailyMail dataset for news summarization. This model converts long news articles into concise, readable summaries. - Base model: t5-small - Dataset: cnndailymail v3.0.0 - Task: Summarization - Language: English - Framework: PyTorch | Input | Output | |-------|---------| | "AI is transforming industries across the world..." | "AI is changing how industries operate globally." | 🏷️ Tags `t5` `summarization` `nlp` `transformers` `cnndailymail`

OMNI-VIS-ASSIST is an advanced multimodal instruction-following AI assistant that interprets both images and text prompts to generate detailed, structured, and insightful explanations. 🚀 Features - Understands visual + textual input - Performs image captioning, chart summarization, and visual reasoning - Converts image content into Markdown, tables, and Mermaid flowcharts - Works with large multimodal models (Qwen3-VL, BLIP, etc.) or fallback captioners

🌀 FLAN-DREAMFORGE-XL Generative World Builder & Story Universe Creator > “From a single sentence, entire worlds are born.” 🌍 Overview FLAN-DREAMFORGE-XL is a creative text-generation model that constructs fictional universes, complete with: - 🏔️ Locations - 🧬 Species - 📖 Histories - 🧙 Cultures and Myths It fine-tunes the Google FLAN-T5-XL architecture to generate cohesive, lore-rich stories from natural language prompts. ⚙️ Features ✅ Generates worlds, civilizations, and backstories ✅ Context-aware multi-entity relationships ✅ Adjustable creativity & coherence balance ✅ Streamlit-based visual UI (`app.py`) ✅ FastAPI-ready endpoint (`api.py`) 🧩 Example Output: > “The planet Solareth, blanketed in shimmering dunes, thrives beneath twin suns. Its glass spires channel solar winds into energy streams, while the Seraphians—translucent beings of sand and spirit—guard the ancient archives buried below.” 🧠 Tech Stack - 🧩 Transformers (FLAN-T5-XL) - 🐍 Python 3.10+ - 🌐 Streamlit / FastAPI - 🧠 PyTorch - 📄 YAML Config System 🧪 Demo Notebook 👉 Check out the Colab demo: `notebooks/demodreamforge.ipynb` 💡 Citation If you use this model in your research or creative projects, please cite:

DOCINTEL extracts structured insights from scanned PDFs and images using naver-clova-ix/donut-base (Donut). It supports OCR fallback, entity extraction, and document summarization via Donut on page images. > ⚠️ Install system dependencies: `poppler` and `tesseract` for pdf2image and pytesseract respectively. 3. Upload a PDF and call endpoints (see examples/democommands.txt). Files - `ocrextractor.py` — PDF→images→OCR pipeline - `pdfloader.py` — extract embedded text from PDFs - `entitytagger.py` — regex-based entity extraction - `summarizedoc.py` — DONUT-based summarizer for page images - `app.py` — FastAPI server with upload/summary endpoints Notes - Donut requires vision-encoder-decoder inference which may need GPU for speed. - For text-only PDFs consider using `extracttextfrompdf` then a text summarizer instead of Donut. - This repo is a prototype/demo. Validate on your data before production use.

AutoDocGen 🧠 — AI Code Documentation & Test Generator AutoDocGen is an advanced model built on CodeT5 that automatically generates documentation, comments, and unit tests for source code files. --- 🚀 Features - Auto-generate docstrings for Python functions - Create unit tests from given functions - Trainable and extendable on custom datasets --- 📦 Files - `traincodet5docgen.py` — training script - `inference.py` — run doc generation on your code - `cli.py` — command-line interface - `configs/trainconfig.json` — training parameters - `dataexamples/` — contains example dataset and code - `tests/` — unit test folder --- 🧠 Model Description This model fine-tunes CodeT5-small from Hugging Face Transformers on a dataset of Python code and natural language descriptions. --- ⚖️ License Licensed under the Apache License 2.0.

🧠 CODE-EXPLAIN-VIZ: AI-Powered Code Understanding & Visualization CODE-EXPLAIN-VIZ is an advanced AI-driven code comprehension and visualization system designed to explain source code in natural language and generate visual flow diagrams automatically. Built on top of CodeT5, this model reads raw Python code and produces a detailed explanation, step-by-step logic breakdown, and a Mermaid flowchart that represents the control flow of the program. 3. Copy the Mermaid flowchart text printed by CLI into a Mermaid live editor (https://mermaid.live) or render with mermaid-cli to see the visual flowchart. What you get - `short` one-line explanation - `detailed` explanation (multi-line) - `mermaid` flowchart text describing control flow - `unittests` template (pytest) How it works - A sequence-to-sequence model (CodeT5) generates natural language explanations from code. - `vizgenerator.py` parses the function AST and produces a reliable mermaid flowchart. - Combining both yields both human-friendly narrative and precise structural view. Train / Fine-tune Use `traindocgen.py` with a JSONL dataset (each line: `{"code": "...", "doc": "..."}`).

🧠 CHRONOGUARD: Temporal Anomaly Detection & Forecasting Model CHRONOGUARD is an advanced hybrid deep learning model for time-series anomaly detection, trend forecasting, and temporal risk visualization. It combines Temporal Convolutional Networks (TCN), Bidirectional LSTMs, and Attention Mechanisms to learn both short-term fluctuations and long-term dependencies from sequential data. - 🧩 Multimodal Input Support — numeric, categorical, and contextual data - 🔍 Attention-based Anomaly Detection — identifies irregular temporal patterns in real time - 📈 Forecast Generation — predicts next-step or multi-step sequences - 🧠 Explainability via Heatmaps — attention and saliency visualizations for model interpretability - 💾 Lightweight, Scalable Architecture — works on CPU/GPU and deploys easily to Hugging Face Spaces or Streamlit

🧠 AutoCodeFix-AI: Transformer-Powered Automatic Code Repair AutoCodeFix-AI is an advanced yet lightweight AI system that automatically detects, explains, and fixes code errors in Python and C-style languages using transformer-based models such as StarCoder2 and CodeT5+. It provides: - ⚙️ A FastAPI backend for scalable inference - 🎛️ A Gradio web demo for live interaction and testing - 📚 Notebooks for model finetuning and evaluation ✅ Automated code error detection and correction ✅ Natural-language explanations of fixes ✅ Transformer-based fine-tuning (StarCoder2 / CodeT5+) ✅ Ready-to-use REST API (`FastAPI`) ✅ Interactive Hugging Face demo via `Gradio` ✅ Modular and lightweight (~40 MB project) | Model | Task | Source | |-------|------|--------| | 🪐 StarCoder2 | Code completion & repair | Hugging Face | | ⚙️ CodeT5+ | Code understanding & generation | Hugging Face | 1. Data Exploration: `notebooks/01dataexploration.ipynb` 2. Preprocessing: `notebooks/02preprocessing.ipynb` 3. Model Finetuning: `notebooks/03modelfinetuning.ipynb` 4. Evaluation: `notebooks/04evaluation.ipynb` 5. Inference Demo: `notebooks/05inferencedemo.ipynb`