AWS Cert - AI Practitioner Exam
AWS Certification Prep AI Practitioner ML Process: Training Data (labeled or not; structured or not) => ML algorithm => Model Steps: 1) Get data, 2) ML (supervised, un supervised , reinforced) 3) Inferencing (batch or real-time) Deep Learning Neural networks take input run through hidden layer and output the answer. Good for vision and audio. Gen AI Steps: 1) Get unlabeled data, 2) Pretrain into FM, 3) Adapt into task output Tokens are like words in a sentence. Types of FMs: 1) LLM, SAGE MAKER SageMaker = ML workflow mgmt that offers auto training, tuning, and deployment of ML and solves data prep and model training tasks . Uses are: predictive analytics, image recognition, voice AI with NLP, fraud detection, recommendation systems, and time-series forecasting. Security via IAM. Does its own logging. CloudWatch can monitor metrics and set alarms. Integrates with S3 for storing d...