File size: 8,212 Bytes
ceac1e4 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 | #!/usr/bin/env python3
"""Scaffold a canonical ICML 2026 reproduction Trackio logbook."""
from __future__ import annotations
import argparse
import json
import re
import sys
def repro_slug_from_title(title: str, *, max_len: int = 96) -> str:
clean = re.sub(r"^Reproduction:\s*", "", title.strip(), flags=re.I).strip()
slug = re.sub(r"[^a-zA-Z0-9]+", "-", clean.lower()).strip("-") or "page"
base = f"repro-{slug}"
if len(base) <= max_len:
return base
return base[:max_len].rstrip("-")
def main() -> int:
parser = argparse.ArgumentParser(
description="Create the canonical ICML reproduction logbook template."
)
parser.add_argument("--title", required=True, help="Paper title")
parser.add_argument("--orid", required=True, help="OpenReview forum id")
parser.add_argument("--arxiv", help="arXiv id (for HF paper link)")
parser.add_argument("--openreview-url", help="Full OpenReview paper URL")
parser.add_argument(
"--hf-indexed",
action="store_true",
help="Paper is indexed on huggingface.co/papers",
)
parser.add_argument(
"--claims-json",
help='JSON array of claim strings, e.g. \'["headline accuracy"]\'',
)
parser.add_argument(
"--username",
help="HF username for metadata.space_id (username/repro-<slug>)",
)
args = parser.parse_args()
try:
from trackio import logbook as lb
except ImportError:
print(
"trackio is required. Install with: uv pip install --upgrade trackio",
file=sys.stderr,
)
return 1
if hasattr(lb, "scaffold_icml_logbook"):
claims = None
if args.claims_json:
try:
claims = json.loads(args.claims_json)
except json.JSONDecodeError as exc:
print(f"Invalid --claims-json: {exc}", file=sys.stderr)
return 1
if not isinstance(claims, list):
print("--claims-json must be a JSON array.", file=sys.stderr)
return 1
try:
info = lb.scaffold_icml_logbook(
title=args.title,
orid=args.orid,
claims=claims,
arxiv_id=args.arxiv,
openreview_url=args.openreview_url,
hf_indexed=args.hf_indexed,
username=args.username,
)
except lb.LogbookError as exc:
print(str(exc), file=sys.stderr)
return 1
_print_next_steps(info)
return 0
claims = None
if args.claims_json:
try:
claims = json.loads(args.claims_json)
except json.JSONDecodeError as exc:
print(f"Invalid --claims-json: {exc}", file=sys.stderr)
return 1
if not isinstance(claims, list):
print("--claims-json must be a JSON array.", file=sys.stderr)
return 1
paper_title = re.sub(r"^Reproduction:\s*", "", args.title.strip(), flags=re.I).strip()
logbook_title = f"Reproduction: {paper_title}"
slug = repro_slug_from_title(paper_title)
space_id = f"{args.username}/{slug}" if args.username else None
orid = args.orid.strip()
if not orid:
print("OpenReview id (--orid) is required.", file=sys.stderr)
return 1
if lb.find_project_dir() and (
lb.logbook_root(lb.find_project_dir()) / "pages" / "index.md"
).exists():
print("A logbook already exists in this directory.", file=sys.stderr)
return 1
try:
proj = lb.create_logbook(title=logbook_title, space_id=space_id)
paper_link = (
f"[HF paper page](https://huggingface.co/papers/{args.arxiv.strip()})"
if args.hf_indexed and args.arxiv
else f"[OpenReview paper]({(args.openreview_url or f'https://openreview.net/forum?id={orid}').strip()})"
)
claim_specs = []
for i, claim in enumerate(claims or [], start=1):
claim_text = str(claim).strip()
if not claim_text:
continue
if not re.match(r"^Claim\s+\d+\s*:", claim_text, re.I):
claim_text = f"Claim {i}: {claim_text}"
claim_specs.append((claim_text, lb.ensure_page(proj, claim_text)))
exec_slug = lb.ensure_page(proj, "Executive summary")
concl_slug = lb.ensure_page(proj, "Conclusion")
index_lines = [
f"# {logbook_title}",
"",
paper_link,
"",
lb.TOC_HEADING,
"",
lb.TOC_HEADER,
lb.TOC_SEP,
f"| [Executive summary](#/{exec_slug}) |",
]
for claim_title, claim_slug in claim_specs:
index_lines.append(f"| [{claim_title}](#/{claim_slug}) |")
index_lines += [f"| [Conclusion](#/{concl_slug}) |", ""]
(lb._pages_dir(proj) / "index.md").write_text(
"\n".join(index_lines), encoding="utf-8"
)
metadata = lb.read_metadata(proj)
metadata["tags"] = ["icml2026-repro", f"paper-{orid}"]
if args.arxiv:
metadata["paper"] = {"arxiv_id": args.arxiv.strip()}
if space_id:
metadata["space_id"] = space_id
lb.write_metadata(proj, metadata)
summary_body = (
"Write a 3–5 sentence outcome-first summary here.\n\n"
"## Scope & cost\n\n"
"| Item | Value |\n"
"| --- | --- |\n"
"| GPU / compute | |\n"
"| Wall time | |\n"
"| Feasibility | |\n"
)
lb.add_markdown_cell(proj, exec_slug, summary_body, title="Executive summary")
summary_id = lb.last_cell_id(proj, page=exec_slug)
if summary_id:
lb.set_cell_pinned(proj, summary_id, pinned=True, page=exec_slug)
lb.add_figure_cell(
proj,
exec_slug,
html=(
"<p>Build a reproduction poster with "
'<a href="https://github.com/Chenruishuo/posterly">Chenruishuo/posterly</a> '
"and replace this cell with <code>poster_embed.html</code>.</p>"
),
title="Reproduction poster (poster_embed.html)",
)
poster_id = lb.last_cell_id(proj, page=exec_slug)
if poster_id:
lb.set_cell_pinned(proj, poster_id, pinned=True, page=exec_slug)
lb.add_markdown_cell(
proj,
concl_slug,
(
"Add a reproduction bundle artifact cell here after running:\n\n"
"```bash\n"
f'trackio.log_artifact("./repro_{slug[6:]}/", name="repro-bundle", type="dataset")\n'
f"trackio logbook cell artifact {slug}/repro-bundle:v0 "
'--page "Conclusion" --title "Reproduction bundle" --type dataset\n'
"```"
),
title="Reproduction bundle",
)
for claim_title, claim_slug in claim_specs:
lb.add_markdown_cell(
proj,
claim_slug,
f"Document setup, runs, and results for **{claim_title}**.",
title=claim_title,
)
lb.write_site_files(proj)
except lb.LogbookError as exc:
print(str(exc), file=sys.stderr)
return 1
_print_next_steps(
{
"title": logbook_title,
"slug": slug,
"space_id": space_id or f"<username>/{slug}",
}
)
return 0
def _print_next_steps(info: dict) -> None:
print(f"Scaffolded logbook: {info['title']}")
print(f"Publish slug: {info['slug']}")
print(f"Publish target: {info['space_id']}")
print(
"\nNext steps:\n"
" 1. Reproduce each claim (log commands, Hub assets, results)\n"
" 2. Fill Executive summary + poster_embed.html (Chenruishuo/posterly)\n"
" 3. Add reproduction bundle artifact on Conclusion\n"
" 4. curl -sL …/validate_icml_logbook.py | python3 - --space "
f"{info['space_id']}\n"
f" 5. trackio logbook publish {info['space_id']}"
)
if __name__ == "__main__":
raise SystemExit(main())
|