Goainglys

Native Go ML Projects Collection

A comprehensive collection of machine learning and deep learning implementations written in pure Go with no external dependencies.

| Project | Description | Status | |---------|-------------|--------| | Transformer | Complete Transformer architecture (Vaswani et al., 2017) | ✅ Complete | | Vector Database | HNSW-based vector similarity search | ✅ Complete | | ASR Library | Automatic Speech Recognition with MFCC & LSTM | ✅ Complete | | RAG Evaluation | Comprehensive RAG/LLM evaluation toolkit | ✅ Complete | | Training Dashboard | Real-time ML training metrics visualization | ✅ Complete | | Model Registry | Model versioning and management system | ✅ Complete | | Agents Framework | AI agent framework with native transformers | ✅ Complete | | Fine-tuning | Neural network fine-tuning with LoRA support | ✅ Complete | | GPT/BERT | Native Go GPT and BERT transformer models | ✅ Complete | | MCP Marketplace | MCP apps marketplace with REST API and web UI | ✅ Complete | | LRS-Agents-NBX | Learning Rate Scheduler agents with NBX integration | ✅ Complete |

/workspace/
├── tensor/              # Core tensor operations library
│   ├── tensor.go       # Tensor data structure
│   └── ops.go          # Mathematical operations
├── transformer/         # Transformer architecture implementation
│   ├── layer.go        # Multi-head attention, FFN, positional encoding
│   ├── model.go        # Encoder/Decoder, embeddings, full model
│   ├── backward.go     # Backpropagation through layers
│   └── train.go        # Training utilities
├── vector_db/          # Native vector database
│   ├── main.go         # VectorDB with CRUD operations
│   ├── hnsw.go         # HNSW index for ANN search
│   ├── types.go        # Type defi