Emakia Training Program 2024 -2025
Collaborations and Training
We are thrilled to offer specialized AI training to students from both the Montclair State University NLP Lab and Pan-Atlantic University (PAU). Our program provides a robust system for validating labeled data and model outputs, incorporating advanced tools like Large Language Models (LLMs) and Retrieval-Augmented Generation (RAG). Training sessions for Emakia participants will be conducted biweekly, with Montclair NLP Lab sessions scheduled every three weeks, ensuring comprehensive support and skill development across institutions.
We are excited to offer AI training to students from Montclair University NLP Lab and Pan-Atlantic University (PAU).
Anna Feldman
Professor, Montclair State University
Dr. Anna Feldman co-leads the NLP Lab at Montclair State University,
integrating linguistics and computing to advance NLP and
computational linguistics.
Omowumi (Molly) Ogunyemi
Applied Philosophy (Anthropology & Ethics)
Omowumi Ogunyemi, with expertise from PAU, contributes essential ethical
insights to AI and NLP, emphasizing human flourishing.
Our program provides a comprehensive system for validating labeled data and model outputs using advanced tools like LLM and RAG. Emakia training sessions will be held every 2 weeks, and Montclair NLP Lab sessions will occur every 3 weeks.
Training Topics:
-
Evaluating Labels with LangChain OpenAI and Lexicons
-
Evaluating Labels with RAG and the Evaluate Library
-
Using Facebook’s RoBERTa Hate Speech Model for Label Evaluation
-
Evaluating Labels with OpenAI’s Moderation API and IBM’s MAX-Toxic-Comment-Classifier
-
Using Meta-LLaMA/LLaMA-2 for Label Evaluation
-
Determining Best Model Combinations and Classifying Outcomes
-
Google Vertex AI Text Classifier with Node.js
-
Deploying and Running Model Classifiers on Collected Data
-
Evaluating Model Predictions Based on the Above Steps
-
Our open-source code is available at Emakia GitHub, and our dataset can be found at Emakia Dataset on Kaggle.
Training Program October 2023 to March 2024
The program had ran from October 2023 to March 2024, aimed at delivering significant AI programming experience to participants Lucie Tronczyk, Carmen Villalobos, Mie Haga, and Sikieng Sok. The main goal is to provide hands-on training in developing a system equipped with an AI text classifier. AI and machine learning exposure, the program provides these junior women engineers access to the Google Cloud platform, including tools like Vertex AI, Sentiment Analysis, and Big Query, to enhance their AI learning experience.
Emakia is thrilled to highlight the exceptional contributions of Lucile Tronczyk, Mie Haga, Carmen Villalobos, and Sikieng Sok, who have successfully completed AI training, focusing on the Vertex AI text classifier. To explore our groundbreaking work, visit our source code at https://lnkd.in/gpWaHW9x and access our dataset at https://lnkd.in/g2YDjaDd. Their remarkable project was showcased at the Women Who Code San Francisco event, titled "Transforming Online Spaces: The Power of AI in Battling Social Media Harassment with Google Cloud." Watch their insightful presentation here: https://lnkd.in/gc8RJRtR.
