From Spreadsheet Labels to a Product Classification Adapter

A complete small project from CSV export to registry release. The useful result is a method that preserves evidence, exposes constraints, and makes the next decision reviewable. This note describes an engineering workflow rather than a claim about a particular organization or production event.

Establish the boundary

Inspect labels, blanks, duplicates, and category drift in a fictional CSV export. Keep the relevant source version, configuration, and stable identifier together while this work is performed. That gives a reviewer a direct path from an observed outcome to the input and rule that produced it. Treat automatic checks and human judgment as separate controls, then record both in the build note.

Use a fixed holdout or review sample for this step. Include ordinary records, rare boundary cases, and at least one input designed to fail safely. A useful change improves the intended condition without removing evidence about a known limitation.

Map unstable labels to durable category IDs

Map unstable labels to durable category IDs. Keep the relevant source version, configuration, and stable identifier together while this work is performed. That gives a reviewer a direct path from an observed outcome to the input and rule that produced it. Treat automatic checks and human judgment as separate controls, then record both in the build note.

Generate train and validation JSONL while preserving rare classes

Generate train and validation JSONL while preserving rare classes. Keep the relevant source version, configuration, and stable identifier together while this work is performed. That gives a reviewer a direct path from an observed outcome to the input and rule that produced it. Treat automatic checks and human judgment as separate controls, then record both in the build note.

Train a compact Qwen adapter with conservative settings

Train a compact Qwen adapter with conservative settings. Keep the relevant source version, configuration, and stable identifier together while this work is performed. That gives a reviewer a direct path from an observed outcome to the input and rule that produced it. Treat automatic checks and human judgment as separate controls, then record both in the build note.

Evaluate classification accuracy, invalid IDs, and rare-class recall

Evaluate classification accuracy, invalid IDs, and rare-class recall. Keep the relevant source version, configuration, and stable identifier together while this work is performed. That gives a reviewer a direct path from an observed outcome to the input and rule that produced it. Treat automatic checks and human judgment as separate controls, then record both in the build note.

The artifact with taxonomy version and limitations

Publish the artifact with taxonomy version and limitations. Keep the relevant source version, configuration, and stable identifier together while this work is performed. That gives a reviewer a direct path from an observed outcome to the input and rule that produced it. Treat automatic checks and human judgment as separate controls, then record both in the build note.

Practical check

Working Python CSV-to-JSONL converter and one training command. The example below is intentionally small enough to run before a larger build or publication workflow.

import csv
import json
from pathlib import Path

with Path("labels.csv").open(encoding="utf-8", newline="") as handle:
    rows = list(csv.DictReader(handle))

with Path("train.jsonl").open("w", encoding="utf-8") as handle:
    for row in rows:
        handle.write(json.dumps(row, ensure_ascii=False) + "\n")

print(f"wrote {len(rows)} records")

Evidence for the next decision

Keep the input digest, outcome, and reviewer note with the resulting artifact. Assuming spreadsheet labels are already a coherent taxonomy. When evidence is incomplete, preserve the failing example, define the missing validation, and defer publication until results can be compared fairly.

Make the process idempotent. Repeating the same step on unchanged input should not append metadata, discard extra detail, or silently change ownership. Idempotence makes retries safe and exposes hidden state when outputs differ.

Review results by failure category rather than a single blended number. A small number of high-risk failures can outweigh an improvement on common cases, particularly when output feeds a release workflow.

Keep accepted and rejected examples together. A concise rejected sample with its expected outcome is a durable regression test and prevents the same edge condition from returning as an unexplained surprise.

Use a separate holdout set for release decisions. Training records can guide implementation, but they cannot demonstrate correct behaviour on new wording, new sources, or changed input order.