Six lessons covered batch over a list, LLM helpers, retry on bad output, self-consistency, cost-aware batching, and a final integration. Eight questions, 80% pass.
| Lesson | Concept |
|---|---|
| 22 | Batch — one LLM call per item, output list aligned with input |
| 23 | Helpers — ask, ask_json, classify wrap Agent.run_sync |
| 24 | Retry on bad output — validate, re-call with feedback, cap attempts |
| 25 | Self-consistency — call N times, vote |
| 26 | Cost-aware batching — result.usage().total_tokens across the batch |
| 27 | Final integration — composition |
Closeout next: graduation post-likert.
Create a free account to get started. Paid plans unlock all tracks.
Six lessons covered batch over a list, LLM helpers, retry on bad output, self-consistency, cost-aware batching, and a final integration. Eight questions, 80% pass.
| Lesson | Concept |
|---|---|
| 22 | Batch — one LLM call per item, output list aligned with input |
| 23 | Helpers — ask, ask_json, classify wrap Agent.run_sync |
| 24 | Retry on bad output — validate, re-call with feedback, cap attempts |
| 25 | Self-consistency — call N times, vote |
| 26 | Cost-aware batching — result.usage().total_tokens across the batch |
| 27 | Final integration — composition |
Closeout next: graduation post-likert.