READER EDITION / ISSUE 002 / Revised 2026-11-20
Creative Agentic Computing
CodeFarmer / Issue 002
YOUR OWN AI WORKSHOP

Compare the builds.
Inspect the difference.

By Alvin

Each row is one generated function from our 128GB M5 Max. Checks counts four test examples; elapsed seconds includes model loading.

MODEL / BRIEFCHECKSELAPSED s
Qwen3.5 4B / notebook4/434.20
Qwen3.5 4B / triage4/417.07
Llama3.2 3B / notebook1/49.70
Llama3.2 3B / triage1/410.87

Qwen passed both sets. Llama passed only the empty-list examples: it dropped unconfirmed notebook notes and lost source labels from triage.

Give the field a clear job

if 'confirmed' not in note or note['confirmed']:
    report.append({'source_id': source_id,
                   'quote': quote})

Our notebook fixture expects every note. Llama treated confirmed as a selection rule. The next page adds an explicit inclusion sentence and follows the resulting change.

Decode rate measures reply generation after prompt processing: Qwen produced about 28–29 tokens/s; Llama 28–34. Elapsed times also include loading. We left the operating system’s file cache in place.