Follow the thread
The shape, dtype and meaning of numerical data matter at every boundary.
Use this tag to connect basic tensor vocabulary with the way data moves through an application. The articles cover rectangular arrays, batches, token representations and the contracts that describe them. Read the shape guide first if an axis is unclear; choose an integration guide when the question is where tensor creation should happen.
A useful reading exercise is to annotate one example end to end. Name every axis, record the type of each value and identify the component that is allowed to transform it. Then inspect what the next component expects. This makes a mismatch concrete before it becomes a vague model-quality problem.
4 articles tagged Tensors
Trace text from vocabulary tokens to embedding vectors and shaped tensors. Understand how masks, pooling and model versions affect API design, memory estimates and reliable similarity search. Keep representation size separate from the meaning of its values.
7 min readChoose the right Azure endpoint for an inference workload, define numerical data contracts and keep preprocessing, model versions, batch recovery and rollout decisions traceable across the full pipeline. Cover both interactive requests and larger asynchronous classification jobs.
7 min readFollow the transformations between a prompt and a useful response. Understand token IDs, attention masks, model-output shapes, decoding choices, and the metadata that makes an LLM integration easier to debug and evaluate.
7 min readA tensor contract needs more than an array of numbers. Learn how to name axes, validate numerical types, preserve feature meaning, and batch compatible examples without hiding errors at the API boundary.
7 min read
TensorAPI.com
