FullStack AI Hub

Learning outcomes

Separate learning from generation
Identify inference-only controls
Explain parameter updates

Mental model

Training changes model parameters from examples and an objective. Inference keeps parameters fixed and computes outputs from the current input, optional tools, and decoding controls.

Training data

Loss and gradients

Parameter updates

Frozen model

Inference requests

Theory

During training, a forward pass produces predictions, a loss measures error, backpropagation computes gradients, and an optimizer updates parameters. During autoregressive inference, the model repeatedly predicts a next-token distribution, selects a token, appends it, and continues. Fine-tuning is training; prompting, retrieval, caching, and tool use primarily change inference context or workflow.

Alternatives and trade-offs

Prompting changes instructions, RAG supplies current knowledge, tools perform actions, and fine-tuning changes learned behavior. Choose based on the gap rather than treating every quality problem as a training problem.

Failure modes and misconceptions

Temperature cannot teach durable knowledge, RAG does not update model weights, and a successful training loss does not establish production usefulness.

Knowledge check

Reflect before revealing the guide

Which system changes can be deployed without modifying model parameters?

Decision scenario

A model knows the policy but returns the wrong JSON shape. Strengthen schemas and validation before considering fine-tuning; the issue is an interface contract, not missing knowledge.

Relationships

prerequisite

Transformer Architecture

Architecture knowledge clarifies learning and generation computation.