Files
waggle-os/benchmarks/probes/judge-swap-validation/deepseek-mt2048-probe.py
Oleg Maslov 0c3e2ead3b
Some checks failed
Installer Smoke / installer-smoke (push) Has been cancelled
moving
2026-09-02 10:10:29 +02:00

275 lines
9.7 KiB
Python

"""
§1.3h Option C follow-up — DeepSeek max_tokens=2048 verification
=================================================================
PM-ratified bounded probe: re-run DeepSeek on the same 7 splits from
§1.3h with max_tokens=2048 (up from 1024). Purpose: classify the 2/7
parse failures from §1.3h as truncation_fixable vs structural.
Scope:
- Single candidate (DeepSeek only)
- Same 7-split sample (split-cases-sample.jsonl SHA 6df4ed0f...)
- One parameter delta: max_tokens 1024 -> 2048
- No new sample selection, no new Opus/GPT calls
- Artefacts additive in benchmarks/probes/judge-swap-validation/
Scope guards (identical to §1.3h):
- Manifest v5 anchor fc16925 immutable
- Parent HEAD = ae0d312 (§1.3h anchor)
- §11 frozen paths untouched
- No runner modification, no v6 emission
Budget: $0.50 cap, 15 min wall-clock cap.
"""
from __future__ import annotations
import json
import sys
import time
import urllib.error
import urllib.request
from datetime import datetime, timezone
from pathlib import Path
try:
sys.stdout.reconfigure(encoding="utf-8", errors="replace")
sys.stderr.reconfigure(encoding="utf-8", errors="replace")
except Exception:
pass
PROBE_DIR = Path("D:/Projects/waggle-os/benchmarks/probes/judge-swap-validation")
SAMPLE_PATH = PROBE_DIR / "split-cases-sample.jsonl"
OUT_PATH = PROBE_DIR / "deepseek-split-responses-v2-mt2048.jsonl"
# Verbatim judge prompt (same as §1.3g and §1.3h probes)
JUDGE_PROMPT_TEMPLATE = "\n".join([
"You are evaluating whether an LLM's answer is correct against ground truth.",
"",
"## Question",
"{question}",
"",
"## Ground-truth answer",
"{ground_truth}",
"",
"## Ground-truth supporting context (excerpt shown to the model)",
"{context}",
"",
"## Model's answer",
"{model_answer}",
"",
"## Your task",
"",
"Step 1: Determine if the model's answer is correct.",
"- \"correct\" means the model's answer contains all required facts from ground truth, with no additional incorrect claims.",
"- Minor phrasing differences, synonyms, or alternative but equivalent formulations are acceptable.",
"- Extra detail is acceptable ONLY if it is factually correct.",
"",
"Step 2: If incorrect, assign exactly one failure mode using this decision tree:",
"",
"1. Does the model explicitly refuse or say it does not know? -> F1 (ABSTAIN)",
"2. Does the model answer a DIFFERENT question than was asked (coherent but off-topic)? -> F5 (OFF-TOPIC)",
"3. Does the model rely on entities, names, dates, or claims that do NOT appear in the ground-truth context (fabrication)? -> F4 (HALLUCINATED)",
"4. Does the model correctly state SOME required facts but miss others, without stating any incorrect facts? -> F2 (PARTIAL)",
"5. Otherwise (model states facts derived from the context but gets them wrong): -> F3 (INCORRECT)",
"",
"Step 3: Return JSON only, no prose, in this exact schema:",
"",
"{{",
" \"verdict\": \"correct\" | \"incorrect\",",
" \"failure_mode\": null | \"F1\" | \"F2\" | \"F3\" | \"F4\" | \"F5\",",
" \"rationale\": \"one sentence explaining the verdict\"",
"}}",
"",
"If verdict is \"correct\", failure_mode MUST be null.",
"If verdict is \"incorrect\", failure_mode MUST be one of F1-F5.",
])
def ts() -> str:
return datetime.now(timezone.utc).isoformat()
def logmsg(msg: str) -> None:
print(f"{ts()} {msg}", flush=True)
def load_env() -> dict[str, str]:
env_path = Path("D:/Projects/waggle-os/.env")
out: dict[str, str] = {}
for line in env_path.read_text(encoding="utf-8").splitlines():
line = line.strip()
if not line or line.startswith("#") or "=" not in line:
continue
k, _, v = line.partition("=")
out[k.strip()] = v.strip().strip('"').strip("'")
return out
def extract_json_body(raw: str) -> dict | None:
import re
if not raw:
return None
trimmed = raw.strip()
if trimmed.startswith("```"):
m = re.match(r"^```(?:json)?\s*\n?(.*?)```\s*$", trimmed, re.DOTALL)
if m:
trimmed = m.group(1).strip()
try:
return json.loads(trimmed)
except Exception:
pass
first = trimmed.find("{")
last = trimmed.rfind("}")
if first != -1 and last != -1 and last > first:
try:
return json.loads(trimmed[first:last + 1])
except Exception:
return None
return None
def parse_verdict(raw: str) -> tuple[str | None, str | None, str | None]:
body = extract_json_body(raw)
if not isinstance(body, dict):
return (None, None, None)
v = body.get("verdict")
fm = body.get("failure_mode")
rat = body.get("rationale")
if v not in ("correct", "incorrect"):
return (None, None, None)
if fm is not None and fm not in ("F1", "F2", "F3", "F4", "F5"):
fm = None
return (v, fm, rat if isinstance(rat, str) else None)
def http_post_json(url: str, headers: dict, body: dict, timeout_s: int = 60) -> tuple[int, dict | str]:
req = urllib.request.Request(
url, data=json.dumps(body).encode("utf-8"), method="POST",
headers={"Content-Type": "application/json", **headers},
)
try:
with urllib.request.urlopen(req, timeout=timeout_s) as resp:
raw = resp.read().decode("utf-8", errors="replace")
try:
return resp.status, json.loads(raw)
except Exception:
return resp.status, raw
except urllib.error.HTTPError as e:
try:
return e.code, e.read().decode("utf-8", errors="replace")[:2000]
except Exception:
return e.code, ""
except Exception as e:
return 0, f"{type(e).__name__}: {e}"
def call_deepseek_mt2048(prompt: str, api_key: str, max_attempts: int = 3) -> dict:
url = "https://api.deepseek.com/v1/chat/completions"
headers = {"Authorization": f"Bearer {api_key}"}
body = {
"model": "deepseek-v4-pro",
"messages": [{"role": "user", "content": prompt}],
"temperature": 0.0,
"max_tokens": 2048, # ← THE SINGLE DELTA
}
started = time.time()
retries = 0
last_err = None
for attempt in range(max_attempts):
status, resp = http_post_json(url, headers, body, timeout_s=60)
if status == 200 and isinstance(resp, dict):
choices = resp.get("choices") or []
if choices:
msg = choices[0].get("message") or {}
content = msg.get("content") or msg.get("reasoning_content") or ""
usage = resp.get("usage", {})
return {
"raw_text": content,
"status": 200,
"error": None,
"retries": retries,
"latency_ms": int((time.time() - started) * 1000),
"prompt_tokens": usage.get("prompt_tokens"),
"completion_tokens": usage.get("completion_tokens"),
}
last_err = f"status={status} resp={str(resp)[:400]}"
retries += 1
if attempt < max_attempts - 1:
time.sleep(2 ** attempt)
return {
"raw_text": "",
"status": 0,
"error": last_err,
"retries": retries,
"latency_ms": int((time.time() - started) * 1000),
"prompt_tokens": None,
"completion_tokens": None,
}
def main() -> int:
logmsg("[mt2048] DeepSeek max_tokens bump verification START")
env = load_env()
key = env.get("DEEPSEEK_API_KEY", "").strip()
if not key:
logmsg("[mt2048] FATAL DEEPSEEK_API_KEY missing")
return 2
sample = []
with SAMPLE_PATH.open("r", encoding="utf-8") as f:
for line in f:
line = line.strip()
if line:
sample.append(json.loads(line))
logmsg(f"[mt2048] loaded {len(sample)} split instances from {SAMPLE_PATH.name}")
rows = []
for i, s in enumerate(sample):
prompt = JUDGE_PROMPT_TEMPLATE.format(
question=s.get("question") or "",
ground_truth=s.get("ground_truth") or "",
context=s.get("context") or "",
model_answer=s.get("model_answer") or "",
)
resp = call_deepseek_mt2048(prompt, key)
verdict, fm, rat = parse_verdict(resp["raw_text"])
rows.append({
"instance_id": s["instance_id"],
"cell": s["cell"],
"provider": "deepseek",
"model_id": "deepseek-v4-pro",
"routing": "direct",
"max_tokens": 2048,
"http_status": resp["status"],
"error": resp.get("error"),
"retries": resp["retries"],
"latency_ms": resp["latency_ms"],
"prompt_tokens": resp.get("prompt_tokens"),
"completion_tokens": resp.get("completion_tokens"),
"raw_text": resp["raw_text"],
"parsed_verdict": verdict,
"parsed_failure_mode": fm,
"parsed_rationale": rat,
"opus_verdict_ref": s.get("opus_verdict"),
"gpt_verdict_ref": s.get("gpt_verdict"),
})
logmsg(
f"[mt2048] {i+1}/{len(sample)} {s['instance_id']:30} cell={s['cell']:14} "
f"status={resp['status']} verdict={verdict} retries={resp['retries']} "
f"completion_tokens={resp.get('completion_tokens')}"
)
OUT_PATH.parent.mkdir(parents=True, exist_ok=True)
with OUT_PATH.open("w", encoding="utf-8") as f:
for r in rows:
f.write(json.dumps(r, ensure_ascii=False) + "\n")
parsed = sum(1 for r in rows if r.get("parsed_verdict") is not None)
logmsg(f"[mt2048] wrote {len(rows)} rows, parsed_ok={parsed}")
logmsg("[mt2048] END")
return 0
if __name__ == "__main__":
sys.exit(main())