PyTorch
Also written as Torch
The framework most teams use to build and train machine-learning models in Python — the tool underneath most modern AI work, including the models behind LLMs.
Think of it like
The workshop and the power tools. It doesn't tell you what to build; it's what you build with.
Junior or senior?
Junior sounds like
Followed a tutorial and fine-tuned an existing model.
Senior sounds like
Has debugged training that silently produced garbage, and can explain what they changed.
Ask them
“Tell me about a model you trained yourself. What went wrong the first time?”
Sounds like real experience
Names a specific problem they hit — the model not learning, running out of GPU memory, training being far too slow — and what they changed to fix it.
Probe further if
Only describes running code someone else wrote, or lists PyTorch among many frameworks with no story about actually using it.