Explore a comprehensive conference talk on productionizing AI and ML algorithms in cloud environments. Delve into the development phase of machine learning algorithms, focusing on natural language processing tasks and concepts. Learn about Hugging Face, a popular platform for NLP models, and discover MLflow for managing the machine learning lifecycle. Examine practical use cases, including text classification, and understand the benefits of using MLflow's tracking API for model development. Gain insights into model training frameworks, fine-tuning NLP models with Hugging Face, and effective model evaluation techniques. Explore PyTorch serving for ML models and learn strategies for productionizing algorithms in AWS cloud. This 25-minute presentation by Deepak Karunanidhi at Conf42 Cloud Native 2024 offers valuable knowledge for data scientists and machine learning engineers looking to deploy AI and ML solutions in cloud environments.
Productionize AI and ML Algorithms in Cloud Environments
Overview
Syllabus
intro
preamble
development phase of ml algorithm
natural language processing tasks
natural language processing : the concept
hugging face
use case 1 - example - text classification
mlflow
ml model experiments
model development without mlflow
mlflow tracking api: simple & pythonic!
model training framework
hugging face fine tuning nlp...
model evaluation
pytorch serving - ml models
productionizing in aws cloud
thank you
Taught by
Conf42
