liqiang
b119135836
Build latest book artifacts / build (push) Canceled after 0s
dependency resolution / resolve (3.11) (push) Canceled after 0s
dependency resolution / resolve (3.13) (push) Canceled after 0s
deploy-pages / build (push) Canceled after 0s
deploy-pages / deploy (push) Canceled after 0s
i18n consistency check / check (push) Canceled after 0s
provider adoption tests / test (chapter2/context-compression) (push) Canceled after 0s
provider adoption tests / test (chapter2/prompt-injection) (push) Canceled after 0s
provider adoption tests / test (chapter2/system-hint) (push) Canceled after 0s
provider adoption tests / test (chapter3/log-sanitization) (push) Canceled after 0s
web-search-agent tests / test (push) Canceled after 0s
web-search-agent tests / agentbook (push) Canceled after 0s
51 lines
2.3 KiB
Python
51 lines
2.3 KiB
Python
import json
|
|
from pathlib import Path
|
|
|
|
ROOT = Path(__file__).parents[1]
|
|
|
|
def test_splits_are_nonempty_and_structured():
|
|
for name in ("train", "eval", "boundary"):
|
|
rows = [json.loads(x) for x in (ROOT / "data" / f"{name}.jsonl").read_text(encoding="utf-8").splitlines() if x.strip()]
|
|
assert rows
|
|
assert all({"id", "kind", "language", "article_type", "source", "prompt", "target"} <= set(row) for row in rows)
|
|
|
|
def read(name):
|
|
return [json.loads(x) for x in (ROOT / "data" / f"{name}.jsonl").read_text(encoding="utf-8").splitlines() if x.strip()]
|
|
|
|
def test_campaign_sizes_and_id_disjointness():
|
|
train, evaluation, boundary = map(read, ("train", "eval", "boundary"))
|
|
assert (len(train), len(evaluation), len(boundary)) == (1024, 256, 256)
|
|
ids = [x["id"] for x in train + evaluation + boundary]
|
|
assert len(ids) == len(set(ids))
|
|
|
|
def test_eval_ids_do_not_overlap_train_ids():
|
|
train = {json.loads(x)["id"] for x in (ROOT / "data" / "train.jsonl").read_text(encoding="utf-8").splitlines() if x.strip()}
|
|
eval_ids = {json.loads(x)["id"] for x in (ROOT / "data" / "eval.jsonl").read_text(encoding="utf-8").splitlines() if x.strip()}
|
|
assert train.isdisjoint(eval_ids)
|
|
|
|
def test_sources_are_unique_across_all_splits():
|
|
rows = read("train") + read("eval") + read("boundary")
|
|
sources = [x["source"] for x in rows]
|
|
assert len(sources) == len(set(sources))
|
|
|
|
def test_tool_targets_are_valid_json_and_direct_targets_are_exact():
|
|
for row in read("train") + read("eval") + read("boundary"):
|
|
if row["kind"] == "tool_json":
|
|
parsed = json.loads(row["target"])
|
|
assert parsed["old_string"] == row["source"]
|
|
else:
|
|
assert row["target"] == row["source"]
|
|
|
|
def test_special_character_cases_are_present():
|
|
rows = read("train")
|
|
assert any("\n" in x["source"] for x in rows)
|
|
assert any("\\n" in x["source"] for x in rows)
|
|
assert any("\u200b" in x["source"] for x in rows)
|
|
assert any("中" in x["source"] for x in rows)
|
|
|
|
def test_tokenizer_audit_covers_multiple_open_tokenizers():
|
|
report = json.loads((ROOT / "validation" / "tokenizer_audit.json").read_text(encoding="utf-8"))
|
|
assert report["probe_count"] == 512
|
|
assert len(report["tokenizers"]) >= 3
|
|
assert all("roundtrip_rate" in value and "mean_tokens" in value for value in report["tokenizers"].values())
|