apunkt/SurfSense
1
0
Fork 0
mirror of https://github.com/MODSetter/SurfSense.git synced 2026-07-16 23:01:06 +02:00
Code Issues Projects Releases Packages Wiki Activity

refactor: streamline TikTok and Instagram scraping logic by removing search_queries and enhancing documentation for clarity

Browse source
This commit is contained in:
DESKTOP-RTLN3BA\$punk 2026-07-13 17:11:25 -07:00
parent e8b3692b54
commit 2b018c4474
111 changed files with 1800 additions and 1580 deletions
Download patch file Download diff file Expand all files Collapse all files

28
surfsense_evals/scripts/summarise_parser_compare_run.py
View file

@ -23,7 +23,9 @@ ARTIFACT = RUN_DIR / "run_artifact.json"
def main() -> None:
rows = [json.loads(line) for line in RAW.read_text(encoding="utf-8").splitlines() if line.strip()]
rows = [
json.loads(line) for line in RAW.read_text(encoding="utf-8").splitlines() if line.strip()
]
print(f"raw rows: {len(rows)}")
by_qid: dict[str, list[dict]] = defaultdict(list)
@ -31,11 +33,19 @@ def main() -> None:
by_qid[row["qid"]].append(row)
print(f"unique questions: {len(by_qid)}")
arm_metrics: dict[str, dict] = defaultdict(lambda: {
"n": 0, "n_correct": 0, "n_failed": 0, "n_empty": 0,
"costs": [], "in_tokens": [], "out_tokens": [], "latency_ms": [],
"by_format": defaultdict(lambda: {"n": 0, "correct": 0}),
})
arm_metrics: dict[str, dict] = defaultdict(
lambda: {
"n": 0,
"n_correct": 0,
"n_failed": 0,
"n_empty": 0,
"costs": [],
"in_tokens": [],
"out_tokens": [],
"latency_ms": [],
"by_format": defaultdict(lambda: {"n": 0, "correct": 0}),
}
)
for row in rows:
arm = row["arm"]
@ -70,7 +80,9 @@ def main() -> None:
print()
print("=" * 100)
print(f"{'arm':<25} {'n':>4} {'acc%':>6} {'F1%':>6} {'fail':>5} {'$ mean':>10} {'$ median':>10} {'in tok mean':>12} {'out tok mean':>12} {'p50 ms':>8}")
print(
f"{'arm':<25} {'n':>4} {'acc%':>6} {'F1%':>6} {'fail':>5} {'$ mean':>10} {'$ median':>10} {'in tok mean':>12} {'out tok mean':>12} {'p50 ms':>8}"
)
print("=" * 100)
art = json.loads(ARTIFACT.read_text(encoding="utf-8"))
per_arm_art = art["metrics"]["per_arm"]
@ -110,7 +122,7 @@ def main() -> None:
print("Aggregated cost (from run_artifact.json):")
for arm, row in per_arm_art.items():
print(
f" {arm:<25} acc={row['accuracy']*100:5.1f}% "
f" {arm:<25} acc={row['accuracy'] * 100:5.1f}% "
f" $/Q LLM={row['llm_cost_per_q']:.4f} "
f" preprocess total=${row['preprocess_cost_total']:.2f} "
f" $/Q total={row['total_cost_per_q']:.4f}"