choicezyc-zym/ai-learning-portfolio

AI Learning Portfolio documents an author’s AI learning path and concrete projects, from traditional ML to a FastAPI-backed multi-step agent app. It’s aimed at learners and engineers who want runnable examples of preprocessing and classification, CNN training, local RAG, agent tool-calling, and backend/frontend AI app patterns. Key projects: - Student Pass Prediction: tabular ML with scikit-learn. - MNIST CNN: PyTorch CNN for digit classification. - Mini RAG: local retrieval (SentenceTransformers), cosine similarity, Ollama/Qwen2.5 generation. - Mini Agent: JSON tool-call planning, calculator and file-reader tools. - RAG Agent: RAG + multi-step agent, FastAPI backend, Streamlit frontend. Core features: semantic embeddings, local LLM integration, structured JSON tool schemas, validation and retry logic, execution history, JSONL logging, API endpoints, request tracking, and automated evaluation/tests.

AI-related keywords found in this repo's description, topics, or README summary — grouped by category. Each badge links to the corresponding ranking detail page.