File size: 5,198 Bytes
9936912
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
#!/usr/bin/env python3
"""Validate ControlAI DAPT splits, provenance, deduplication, and token counts."""

from __future__ import annotations

import argparse
import hashlib
import json
import re
import sys
from pathlib import Path

from transformers import AutoTokenizer


def normalized_hash(text: str) -> str:
    text = re.sub(r"\s+", " ", text).casefold().strip()
    return hashlib.sha256(text.encode("utf-8")).hexdigest()


def load(path: Path) -> tuple[list[dict], list[str]]:
    rows = []
    errors = []
    if not path.exists():
        return rows, [f"missing file: {path}"]
    with path.open(encoding="utf-8") as stream:
        for line_number, line in enumerate(stream, start=1):
            try:
                row = json.loads(line)
            except json.JSONDecodeError as exc:
                errors.append(f"{path}:{line_number}: invalid JSON ({exc.msg})")
                continue
            row["_location"] = f"{path}:{line_number}"
            rows.append(row)
    return rows, errors


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("dataset", type=Path, nargs="?", default=Path("data/training/dapt_v1"))
    parser.add_argument("--tokenizer", default="Qwen/Qwen3-4B-Instruct-2507")
    parser.add_argument("--allow-download", action="store_true")
    args = parser.parse_args()

    train, errors = load(args.dataset / "train.jsonl")
    valid, valid_errors = load(args.dataset / "valid.jsonl")
    errors.extend(valid_errors)
    summary_path = args.dataset / "summary.json"
    summary = json.loads(summary_path.read_text()) if summary_path.exists() else {}

    seen_hashes = {}
    documents = {"train": set(), "valid": set()}
    source_counts = {"train": {}, "valid": {}}
    for split, rows in (("train", train), ("valid", valid)):
        for row in rows:
            location = row["_location"]
            for field in (
                "text",
                "token_count",
                "document_id",
                "source_id",
                "source_pool",
                "text_sha256",
                "simhash64",
            ):
                if row.get(field) in (None, ""):
                    errors.append(f"{location}: missing {field}")
            text = row.get("text")
            if not isinstance(text, str) or not text.strip():
                continue
            digest = normalized_hash(text)
            if digest != row.get("text_sha256"):
                errors.append(f"{location}: text_sha256 mismatch")
            if digest in seen_hashes:
                errors.append(
                    f"{location}: exact duplicate also present at {seen_hashes[digest]}"
                )
            else:
                seen_hashes[digest] = location
            if row.get("source_pool") == "mathworks_r2026a":
                errors.append(f"{location}: proprietary MathWorks text is disallowed")
            if row.get("document_id"):
                documents[split].add(f"{row.get('source_pool')}::{row['document_id']}")
            pool = row.get("source_pool", "unknown")
            source_counts[split][pool] = source_counts[split].get(pool, 0) + 1

    leaked_documents = documents["train"] & documents["valid"]
    if leaked_documents:
        errors.append(
            f"document split leakage: {len(leaked_documents)} document ids occur in both splits"
        )

    expected = {
        "train_rows": len(train),
        "valid_rows": len(valid),
        "train_tokens": sum(row.get("token_count", 0) for row in train),
        "valid_tokens": sum(row.get("token_count", 0) for row in valid),
    }
    for key, value in expected.items():
        if summary.get(key) != value:
            errors.append(f"summary {key}={summary.get(key)!r}, actual={value}")

    tokenizer = AutoTokenizer.from_pretrained(
        args.tokenizer, local_files_only=not args.allow_download
    )
    # Full token recount is deliberate: a stale count changes training estimates.
    for row in train + valid:
        actual = len(tokenizer.encode(row["text"], add_special_tokens=False))
        if actual != row.get("token_count"):
            errors.append(
                f"{row['_location']}: token_count={row.get('token_count')}, actual={actual}"
            )

    print(f"train rows: {len(train):,}")
    print(f"validation rows: {len(valid):,}")
    print(f"unique normalized texts: {len(seen_hashes):,}")
    print(f"train documents: {len(documents['train']):,}")
    print(f"validation documents: {len(documents['valid']):,}")
    print(f"train tokens: {expected['train_tokens']:,}")
    print(f"validation tokens: {expected['valid_tokens']:,}")
    print("source rows:")
    for split in ("train", "valid"):
        print(f"  {split}: {dict(sorted(source_counts[split].items()))}")

    if errors:
        print("validation failed:", file=sys.stderr)
        for error in errors[:100]:
            print(f"- {error}", file=sys.stderr)
        if len(errors) > 100:
            print(f"- ... {len(errors) - 100} additional errors", file=sys.stderr)
        return 1
    print("validation passed")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())