Pinned Loading

1. Attention-Visualizer-Transformer-Attention-Weights Attention-Visualizer-Transformer-Attention-Weights Public

   Interactively visualize BERT and GPT-2 transformer attention weights per layer and head.

   JavaScript

2. Custom-Transformer-from-Scratch Custom-Transformer-from-Scratch Public

   A transformer **sequence-to-sequence model** that predicts the next token given all previous tokens.

   Python

3. AdaptiveRAG AdaptiveRAG Public

   A production-grade RAG using Modular, Self-RAG & Agentic patterns with hybrid dense+sparse retrieval, cross-encoder reranking, and self-critique loops • Used RRF across BM25 + sentence-transformers…

   Python

4. DDIM_Image_Generation DDIM_Image_Generation Public

   A Denoising Diffusion Implicit Model (DDIM) trained from scratch on 30,000 faces from the CelebA-HQ dataset. Built entirely in PyTorch without any pretrained components. What it does: Generate — sa…

   Python

5. HR-Assistant-Fine-tuned HR-Assistant-Fine-tuned Public

   Llama 3.1 8B fine-tuned on the public `PranavTM/LeavePolicy` dataset (TechMojo internal HR policies)

   Python

6. ThinkMCP-Agentic-Tools ThinkMCP-Agentic-Tools Public

   MCP server with adaptive thinking — Claude reasons between every tool call. 13 tools: research, memory, reasoning & actions. Works with Claude Desktop, Cursor, and VS Code out of the box.

   Python