← All drillsA Developer's Toolkit for the 2020s
What you'll be able to do- ✓Train a small neural network in PyTorch and read the four steps of a training loop
- ✓Log a training run to TensorBoard and read the numbers back from the event files
- ✓Ask SQL questions of an experiment log with DuckDB, straight from a CSV or Parquet file
- ✓Compare DuckDB with pandas on a large Parquet file, timing the whole job and not just the group-by
- ✓Train a reinforcement-learning agent on a Gymnasium environment with Ray RLlib
A short tour of four tools a working AI developer meets in the 2020s: PyTorch to train a model, TensorBoard to watch a run, DuckDB to query what you logged, and Ray RLlib to train an agent. One small experiment runs through all three lessons.
⌁ Training loops, run monitoring, in-process analytical SQL and scalable reinforcement learning are everyday parts of an applied AI team's toolchain. PyTorch was created at Meta, TensorBoard at Google, Ray RLlib at Anyscale and the UC Berkeley RISELab, and DuckDB by DuckDB Labs.
Start this internshipCreate an account to unlock the 4 sections, the workbench, and askThili.
BeginSections
1. Why this course: from the theory you teach to the tools teams use now
🔒 locked2. Train It and Watch It: PyTorch and TensorBoard
🔒 locked3. Ask Questions of What You Logged: DuckDB
🔒 locked4. Train an Agent Through a Framework: Gymnasium and Ray RLlib
🔒 lockedDig deeper
📄PyTorch: An Imperative Style, High-Performance Deep Learning Library (Paszke et al., 2019)
paper📄RLlib: Abstractions for Distributed Reinforcement Learning (Liang et al., 2017)
paper🔗PyTorch documentation: torch.utils.tensorboard
article🔗Ray RLlib documentation: algorithms
article🔗DuckDB documentation
article