Data Science & ML Agent Skills

Expert guidance for data analysis, visualization, and Jupyter Notebook development with pandas, matplotlib, seaborn, and numpy.

Use when the user wants to create a dataset, generate synthetic data, or build a data generation pipeline.

Run evaluations for Hugging Face Hub models using inspect-ai and lighteval on local hardware.

Trains and fine-tunes vision models for object detection, image classification and SAM segmentation using Hugging Face Transformers on…

Fine-tune any HuggingFace CV, VLM or LLM model on local NVIDIA GPUs inside an NGC PyTorch container.

Track and visualize ML training experiments with Trackio, including metric logging, alerts and dashboard visualization.

Train and fine-tune transformer language models using TRL (Transformers Reinforcement Learning), with SFT, DPO, GRPO, KTO, RLOO and…

Write a marimo notebook in a Python file in the right format.

Run Megatron-LM and Megatron Bridge training with mock or real data, covering correlation testing, available recipes and multi-GPU examples.

Run bounded, evidence-driven training research through W&B Launch: smoke-test real jobs, execute serial trials, compare metrics and…

Official NVIDIA-authored guidance for cuDF GPU DataFrames, pandas acceleration, dask-cuDF, ETL, joins, groupby, CSV/Parquet I/O and…

Rigorous A/B test statistical analysis — significance calculations, sample ratio mismatch checks and test design validation.

Train or fine-tune language and vision models using TRL or Unsloth with Hugging Face Jobs infrastructure.

Temporal pattern detection and forecasting — trends, seasonality, anomaly detection and simple forecasting models for planning.