-
model-insight-rag Public
A modular agentic Retrieval-Augmented Generation (RAG) system for managing AI model scripts, metadata, and generated images. Supports advanced querying, schema-based validation, image search, model…
-
-
vecraft-vector-database Public
VecraftDB is an embeddable Python vector database with MVCC transactions, WAL durability, HNSW similarity search, JSON/metadata filtering, mmap+SQLite storage, Prometheus metrics, and a FastAPI ser…
Python MIT License UpdatedMay 31, 2025 -
A PyTorch implementation of a diffusion model for generating Fashion-MNIST-like digit images. The repository demonstrates both unconditional and conditional image generation, clearly showing how di…
-
A PyTorch implementation of a latent diffusion model for generating cat/dog-like digit images. The repository demonstrates both unconditional and conditional image generation, clearly showing how d…
Python UpdatedMay 18, 2025 -
A class-conditional image generation framework that combines a Variational Autoencoder (VAE) and a Denoising Diffusion Probabilistic Model (DDPM), tailored for the Oxford 102 Flowers dataset. The a…
-
Advanced generative model combining VAE-GAN and conditional diffusion to create high-quality CelebA face images based on the "Smiling" attribute. Features state-of-the-art architectures with attent…
Python UpdatedMay 18, 2025 -
seq2seq-math-models Public
A collection of Transformer and RNN-based Seq2Seq models for arithmetic operations (addition, subtraction, multiplication, and division). Includes scripts, training data generation, and evaluation …
Python UpdatedMay 5, 2025 -
A PyTorch implementation of a diffusion model for generating CIFAR10 (cat/dog)-like digit images. The repository demonstrates both unconditional and conditional image generation, clearly showing ho…
-
mnist-generative-diffusion Public
A PyTorch implementation of a diffusion model for generating MNIST-like digit images. The repository demonstrates both unconditional and conditional image generation, clearly showing how diffusion …
{{ message }}