wesslen Goto Github PK

followers: 188.0 following: 261.0 repos: 89.0 gists: 82.0

Name: Ryan Wesslen

Type: User

Bio: Data Scientist | ML Engineer | HCI/Vis researcher

Twitter: ryanwesslen

Location: Charlotte, NC

Blog: http://wesslen.netlify.app

Hi there 👋

These are some of my favorite repos:

┣━━ 🤖 llm/ml/nlp demos
┃   ┣━━ Modal examples: axolotl LLM fine tuning and applications (WIP)
┃   ┣━━ Fireworks Colab Fine Tuning
┃   ┣━━ HF SFT QLoRA with Mistral7B and Colab T4
┃   ┣━━ Getting Started with Prompt Engineering with OpenAI API
┃   ┣━━ Bayesian beagle: Quarto blog of weekly LLM-generated summaries of Arxiv preprints
┃   ┣━━ Prodigy NER fine-tuning med7 for synthetic doctors' notes
┃   ┣━━ Prodigy LLM prompt tournament for tweet generation by topic
┃   ┣━━ LLM summarization evaluation with Prodigy
┃   ┣━━ Address extraction NER model with LLM synthetic data and Prodigy
┃   ┗━━ spaCy-Prodigy workflow for NER Citation model on eCFR Banking Regulation
┃   
┣━━ 🛠️ deployment demos
┃   ┣━━ Deploying a Huggingface model to AWS SageMaker with Snowflake connector
┃   ┣━━ Deploy Prodigy via Streamlit app on HF Spaces
┃   ┣━━ Docker templates for Prodigy
┃   ┣━━ Prodigy Docker Compose
┃   ┣━━ Prodigy OpenShift deployment
┃   ┗━━ FastAPI spaCy deployment on HF Spaces
┃   
┣━━ 📄 papers
┃   ┣━━ VitaLITy Vis (Vis 2021)
┃   ┣━━ Myopic Loss Aversion (Vis 2021)
┃   ┣━━ Bayesian Correlation Judgement (Vis 2020)
┃   ┣━━ Verifi (ICWSM 2018)
┃   ┣━━ Twitter Bumper Stickers (ICWSM 2018)
┃   ┗━━ Anchoring Effect in Visual Analytics (VAST 2017)
┃   
┗━━ 💬 teaching
    ┣━━ DSBA6122: Advanced Business Analytics (Summer 2024) Notebooks
    ┣━━ DSBA6010|STAT7027: Bayesian Statistics (Spring 2022)
    ┣━━ DSBA5122: Visual Analytics (Fall 2019)
    ┗━━ DSBA5122: Visual Analytics (Spring 2019)

Ryan Wesslen's Projects

2021-analytics-frontiers-uncertainty

arxiv-frontpage

Ryan's personal frontpage app

bayesian-beagle

Bayesian beagle: Quarto blog of weekly LLM-generated summaries of Arxiv preprints

bayesian-correlation-judgement-vis-2020

Code and supplemental materials for Karduni et al., 2020 IEEE Vis. "A Bayesian cognition approach for belief updating of correlation judgement through uncertainty visualizations"