Q-LorA fine-tuning on Mistral-7B for Spine related queries. Dataset acquistion: OpenAI Whisper, GPT-4 Turbo.
Identify speakers from King Corpus using 20ms frames. Architectures used: GRU, LSTM, CNN
Autonomous driving data collection and real-time ride monitoring. Prototype built using RPi & OpenCV YOLO.
