File size: 22,618 Bytes
f039f41
 
 
 
 
 
 
 
 
a642242
 
f039f41
 
 
 
cc249f3
f039f41
 
a642242
f039f41
 
 
 
a642242
f039f41
 
 
9ddbe3c
 
 
f039f41
9ddbe3c
f039f41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc249f3
 
 
3bbb0bf
cc249f3
 
3bbb0bf
 
cc249f3
 
3bbb0bf
 
 
 
cc249f3
 
f039f41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9ddbe3c
f039f41
 
 
 
 
 
 
 
 
 
 
 
cc249f3
f039f41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9ddbe3c
f039f41
 
 
 
 
 
 
 
 
 
 
 
a642242
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc249f3
 
 
 
 
 
 
 
a642242
 
cc249f3
 
f039f41
 
 
9ddbe3c
f039f41
cc249f3
a642242
 
 
f039f41
 
 
 
 
 
 
cc249f3
 
 
f039f41
cc249f3
 
f039f41
 
 
 
 
 
 
 
 
 
 
9ddbe3c
 
 
 
 
 
a642242
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9ddbe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc249f3
 
3bbb0bf
9ddbe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
cc249f3
 
 
 
9ddbe3c
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f039f41
 
cc249f3
 
 
 
 
 
 
3bbb0bf
cc249f3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f039f41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
9ddbe3c
f039f41
 
335e669
 
 
 
 
f039f41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
a642242
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f039f41
 
 
 
 
 
9ddbe3c
f039f41
 
a642242
 
 
 
 
 
 
 
 
 
 
 
9ddbe3c
 
 
 
 
 
 
 
 
 
 
 
cc249f3
 
 
3bbb0bf
9ddbe3c
 
cc249f3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
f039f41
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
from __future__ import annotations

import argparse
import json
import uuid
from dataclasses import replace
from pathlib import Path
from typing import Any

from .agentic import run_agentic_cases, scenario_set, scenarios
from .agentic_report import summarize_agentic_trials
from .api import OpenAIClient
from .boundary import probe_context_boundary
from .context import build_context_case, expand_context_config
from .holdout import derive_holdout_config
from .model_card import render_benchmark_block, update_model_card, validate_post_publish_run
from .provenance import make_run_manifest, write_manifest
from .quality import generate_quality_cases
from .release import sanitize_manifest, sanitize_result_row, scan_public_tree
from .runner import api_key_from_env, run_cases
from .summary import summarize_results
from .tokenizer import TransformersTokenizer
from .tools import generate_tool_cases
from .util import read_json, read_jsonl, write_json, write_jsonl
from .vision import generate_vision_cases


def _variants(path: Path | None, label: str | None = None) -> list[dict[str, Any]]:
    if path is not None and label is not None:
        raise ValueError("--label and --variants are mutually exclusive")
    if path is None:
        return [{"label": label or "default", "request_overrides": {}}]
    value = read_json(path)
    if not isinstance(value, list) or not value:
        raise ValueError("Variant file must contain a non-empty JSON array")
    return value


def _client(args: argparse.Namespace) -> OpenAIClient:
    return OpenAIClient(
        base_url=args.base_url,
        api_key=api_key_from_env(args.api_key_env),
        timeout_s=args.timeout,
    )


def plan_context(args: argparse.Namespace) -> None:
    config = read_json(args.config)
    specs = expand_context_config(config)
    value = {
        "schema_version": "1.0",
        "suite_id": config["suite_id"],
        "cases": len(specs),
        "total_requested_prompt_tokens": sum(item.target_tokens for item in specs),
        "by_matrix": {},
        "specs": [item.__dict__ | {"case_id": item.case_id} for item in specs],
    }
    for spec in specs:
        value["by_matrix"][spec.matrix] = value["by_matrix"].get(spec.matrix, 0) + 1
    write_json(args.output, value)


def _tokenizer(args: argparse.Namespace) -> TransformersTokenizer:
    return TransformersTokenizer(
        args.tokenizer,
        enable_thinking=args.thinking == "on",
        reasoning_effort=args.reasoning_effort,
        preserve_thinking=True,
    )


def _reasoning_request_overrides(args: argparse.Namespace) -> dict[str, Any]:
    value: dict[str, Any] = {"thinking": {"enabled": args.thinking == "on"}}
    if args.reasoning_effort:
        value["reasoning_effort"] = args.reasoning_effort
    return value


def _record_evaluated_artifact(manifest: dict[str, Any], args: argparse.Namespace) -> None:
    repo_id = getattr(args, "hf_repo", None)
    revision = getattr(args, "hf_revision", None)
    benchmark_revision = getattr(args, "benchmark_revision", None)
    if bool(repo_id) != bool(revision):
        raise ValueError("--hf-repo and --hf-revision must be supplied together")
    if repo_id and not benchmark_revision:
        raise ValueError("--benchmark-revision is required for a post-publication run")
    if repo_id:
        manifest["evaluated_artifact"] = {"repo_id": repo_id, "revision": revision}
        manifest["benchmark_source"] = {
            "repo_id": "Shiftedx/shiftedx-bench",
            "revision": benchmark_revision,
        }


def _validate_context_variant_profile(args: argparse.Namespace, variants: list[dict[str, Any]]) -> None:
    declared = {
        bool((variant.get("request_overrides", {}).get("thinking") or {}).get("enabled"))
        for variant in variants
        if "thinking" in (variant.get("request_overrides") or {})
    }
    if len(declared) > 1:
        raise ValueError("All variants in one exact-token context run must use the same thinking mode")
    if declared and declared != {args.thinking == "on"}:
        raise ValueError("--thinking must match the variant file so local and server templates are identical")


def inspect_context(args: argparse.Namespace) -> None:
    config = read_json(args.config)
    specs = expand_context_config(config)
    selected = next((item for item in specs if item.case_id == args.case_id), None)
    if selected is None:
        raise ValueError(f"Unknown case identifier: {args.case_id}")
    tokenizer = _tokenizer(args)
    case = build_context_case(selected, tokenizer)
    value = case.to_dict(include_expected=True)
    if not args.include_prompt:
        value["messages"] = [{"role": item["role"], "content": "<omitted>"} for item in case.messages]
    write_json(args.output, value)


def run_context(args: argparse.Namespace) -> None:
    config = read_json(args.config)
    specs = expand_context_config(config)
    if args.matrix:
        specs = [item for item in specs if item.matrix == args.matrix]
    if args.limit is not None:
        specs = specs[: args.limit]
    tokenizer = _tokenizer(args)
    variants = _variants(args.variants, args.label)
    _validate_context_variant_profile(args, variants)
    run_id = str(uuid.uuid4())
    manifest = make_run_manifest(
        run_id=run_id,
        suite_id=config["suite_id"],
        model=args.model,
        config_path=args.config,
        tokenizer_fingerprint=tokenizer.fingerprint,
        variants=variants,
    )
    manifest["planned_cases"] = len(specs)
    manifest["requested_prompt_tokens"] = sum(item.target_tokens for item in specs)
    _record_evaluated_artifact(manifest, args)
    write_manifest(args.output.with_suffix(".manifest.json"), manifest)
    reasoning_overrides = _reasoning_request_overrides(args)
    cases = (
        replace(build_context_case(spec, tokenizer), request_overrides=reasoning_overrides)
        for spec in specs
    )
    run_cases(
        cases,
        client=_client(args),
        model=args.model,
        output_path=args.output,
        variants=variants,
        run_id=run_id,
        stream=args.stream,
    )


def run_suite(args: argparse.Namespace) -> None:
    config = read_json(args.config)
    variants = _variants(args.variants, args.label)
    if args.suite == "quality":
        cases = generate_quality_cases(config["quality"]["seeds"], config["quality"].get("families"))
        suite_id = "shiftedx-quality-v1"
    elif args.suite == "tools":
        cases = generate_tool_cases(config["tools"]["seeds"])
        suite_id = "shiftedx-tools-v1"
    elif args.suite == "vision":
        cases = generate_vision_cases(args.output.parent / "vision-fixtures", config["vision"]["seeds"])
        suite_id = "shiftedx-vision-v1"
    elif args.suite == "agentic":
        if len(variants) != 1:
            raise ValueError("Agentic runner currently accepts one request variant per invocation")
        request_overrides = dict(variants[0].get("request_overrides") or {})
        thinking = getattr(args, "thinking", None)
        reasoning_effort = getattr(args, "reasoning_effort", None)
        if thinking is not None:
            request_overrides["thinking"] = {"enabled": thinking == "on"}
        if reasoning_effort is not None:
            request_overrides["reasoning_effort"] = reasoning_effort
        variants = [{**variants[0], "request_overrides": request_overrides}]
        agentic_set = getattr(args, "agentic_set", "core")
        selected = scenario_set(agentic_set)
        case_id = getattr(args, "case_id", None)
        if case_id is not None:
            selected = [item for item in selected if item.case_id == case_id]
            if not selected:
                raise ValueError(f"Unknown case identifier for {agentic_set}: {case_id}")
        if args.limit is not None:
            selected = selected[: args.limit]
        run_id = str(uuid.uuid4())
        manifest = make_run_manifest(
            run_id=run_id, suite_id="shiftedx-agentic-v1", model=args.model,
            config_path=args.config, tokenizer_fingerprint=None, variants=variants,
        )
        manifest["planned_cases"] = len(selected)
        manifest["agentic_control_profile"] = getattr(args, "agentic_control_profile", "baseline")
        manifest["agentic_set"] = agentic_set
        _record_evaluated_artifact(manifest, args)
        write_manifest(args.output.with_suffix(".manifest.json"), manifest)
        run_agentic_cases(
            client=_client(args), model=args.model, output_path=args.output,
            request_overrides=variants[0].get("request_overrides") or {},
            variant_label=str(variants[0]["label"]),
            limit=args.limit,
            run_id=run_id,
            control_profile=getattr(args, "agentic_control_profile", "baseline"),
            agentic_set=agentic_set,
            case_id=case_id,
        )
        return
    else:
        raise ValueError(args.suite)
    if args.limit is not None:
        cases = cases[: args.limit]
    run_id = str(uuid.uuid4())
    manifest = make_run_manifest(
        run_id=run_id, suite_id=suite_id, model=args.model, config_path=args.config,
        tokenizer_fingerprint=None, variants=variants,
    )
    _record_evaluated_artifact(manifest, args)
    write_manifest(args.output.with_suffix(".manifest.json"), manifest)
    run_cases(
        cases,
        client=_client(args), model=args.model, output_path=args.output,
        variants=variants, run_id=run_id, stream=args.stream,
    )


def summarize(args: argparse.Namespace) -> None:
    rows = []
    for path in args.inputs:
        rows.extend(read_jsonl(path))
    write_json(
        args.output,
        summarize_results(rows, args.effective_threshold, baseline_variant=args.baseline),
    )


def summarize_agentic(args: argparse.Namespace) -> None:
    rows = []
    for path in args.inputs:
        rows.extend(read_jsonl(path))
    write_json(args.output, summarize_agentic_trials(rows))


def export_public_results(args: argparse.Namespace) -> None:
    rows = []
    for path in args.inputs:
        rows.extend(read_jsonl(path))
    write_jsonl(args.output, (sanitize_result_row(row) for row in rows))
    if args.manifest is not None:
        if args.manifest_output is None:
            raise ValueError("--manifest-output is required with --manifest")
        write_json(args.manifest_output, sanitize_manifest(read_json(args.manifest)))
    elif args.manifest_output is not None:
        raise ValueError("--manifest is required with --manifest-output")


def _bundled_config(name: str) -> Path:
    path = Path(__file__).resolve().parents[2] / "configs" / name
    if not path.exists():
        raise FileNotFoundError(
            f"Bundled config not found at {path}; pass an explicit config path from the source repository"
        )
    return path


def run_quant_gate(args: argparse.Namespace) -> None:
    args.output_dir.mkdir(parents=True, exist_ok=True)
    context_config = args.context_config or _bundled_config("context-quant-gate-v1.json")
    suite_config = args.suite_config or _bundled_config("suites-quant-gate-v1.json")
    common = {
        "base_url": args.base_url,
        "model": args.model,
        "hf_repo": args.hf_repo,
        "hf_revision": args.hf_revision,
        "benchmark_revision": args.benchmark_revision,
        "api_key_env": args.api_key_env,
        "timeout": args.timeout,
        "variants": None,
        "label": args.label,
        "stream": True,
    }
    run_context(
        argparse.Namespace(
            **common,
            output=args.output_dir / "context.jsonl",
            config=context_config,
            tokenizer=args.tokenizer,
            thinking=args.thinking,
            reasoning_effort=args.reasoning_effort,
            matrix=None,
            limit=None,
        )
    )
    suite_settings = read_json(suite_config)
    outputs = [args.output_dir / "context.jsonl"]
    suites = ["quality", "tools", "agentic"]
    if args.profile == "vision":
        suites.append("vision")
    for suite in suites:
        output = args.output_dir / f"{suite}.jsonl"
        run_suite(
            argparse.Namespace(
                **common,
                output=output,
                config=suite_config,
                suite=suite,
                limit=suite_settings.get("agentic_limit") if suite == "agentic" else None,
            )
        )
        outputs.append(output)
    rows = []
    for path in outputs:
        rows.extend(read_jsonl(path))
    write_json(
        args.output_dir / "summary.json",
        summarize_results(rows, baseline_variant=args.label),
    )


def render_model_card(args: argparse.Namespace) -> None:
    rows, benchmark_version = validate_post_publish_run(
        args.run_dir,
        variant=args.variant,
        profile=args.profile,
        model_repo=args.model_repo,
        model_revision=args.model_revision,
        benchmark_revision=args.benchmark_revision,
    )
    block = render_benchmark_block(
        rows,
        variant=args.variant,
        model_repo=args.model_repo,
        model_revision=args.model_revision,
        benchmark_revision=args.benchmark_revision,
        benchmark_version=benchmark_version,
        host=args.host,
        runtime=args.runtime,
        kv_cache=args.kv_cache,
        mtp_depth=args.mtp_depth,
        max_window_status=args.max_window_status,
    )
    args.output.parent.mkdir(parents=True, exist_ok=True)
    args.output.write_text(update_model_card(args.readme.read_text(encoding="utf-8"), block), encoding="utf-8")


def validate_release(args: argparse.Namespace) -> None:
    result = scan_public_tree(args.root)
    if args.output:
        write_json(args.output, result)
    else:
        print(json.dumps(result, indent=2, sort_keys=True))
    if not result["ok"]:
        raise SystemExit(1)


def make_holdout(args: argparse.Namespace) -> None:
    master_key = api_key_from_env(args.master_key_env)
    if master_key is None:
        raise RuntimeError("Holdout master key is required")
    config = derive_holdout_config(read_json(args.template), master_key, args.release_id)
    write_json(args.output, config)


def probe_boundary(args: argparse.Namespace) -> None:
    tokenizer = _tokenizer(args)
    result = probe_context_boundary(
        client=_client(args), tokenizer=tokenizer, model=args.model,
        context_window=args.context_window, reserved_output_tokens=args.reserved_output_tokens,
        include_positive=args.include_positive,
    )
    write_json(args.output, result)
    if not result["passed"]:
        raise SystemExit(1)


def add_runtime_arguments(parser: argparse.ArgumentParser) -> None:
    parser.add_argument("--base-url", required=True, help="OpenAI-compatible base URL ending in /v1")
    parser.add_argument("--model", required=True)
    parser.add_argument("--output", required=True, type=Path)
    parser.add_argument("--api-key-env", help="Environment-variable name containing the API key")
    parser.add_argument("--timeout", type=float, default=900)
    parser.add_argument("--variants", type=Path)
    parser.add_argument("--label", help="Stable candidate label stored in every result row")
    parser.add_argument("--limit", type=int)
    parser.add_argument("--stream", action="store_true")
    parser.add_argument("--hf-repo", help="Published Hugging Face repo evaluated by this run")
    parser.add_argument("--hf-revision", help="Full immutable Hugging Face revision evaluated by this run")
    parser.add_argument(
        "--benchmark-revision", help="Full immutable Shiftedx Bench revision used by this run"
    )


def add_tokenizer_profile_arguments(parser: argparse.ArgumentParser) -> None:
    parser.add_argument("--tokenizer", required=True, type=Path)
    parser.add_argument("--thinking", choices=["on", "off"], default="on")
    parser.add_argument("--reasoning-effort", default="medium")


def make_parser() -> argparse.ArgumentParser:
    parser = argparse.ArgumentParser(prog="shiftedx-bench")
    subparsers = parser.add_subparsers(dest="command", required=True)

    command = subparsers.add_parser("plan-context")
    command.add_argument("--config", required=True, type=Path)
    command.add_argument("--output", required=True, type=Path)
    command.set_defaults(func=plan_context)

    command = subparsers.add_parser("inspect-context")
    command.add_argument("--config", required=True, type=Path)
    add_tokenizer_profile_arguments(command)
    command.add_argument("--case-id", required=True)
    command.add_argument("--output", required=True, type=Path)
    command.add_argument("--include-prompt", action="store_true")
    command.set_defaults(func=inspect_context)

    command = subparsers.add_parser("run-context")
    add_runtime_arguments(command)
    command.add_argument("--config", required=True, type=Path)
    add_tokenizer_profile_arguments(command)
    command.add_argument("--matrix")
    command.set_defaults(func=run_context)

    command = subparsers.add_parser("run-suite")
    add_runtime_arguments(command)
    command.add_argument("--config", required=True, type=Path)
    command.add_argument("--suite", required=True, choices=["quality", "tools", "agentic", "vision"])
    command.add_argument(
        "--agentic-control-profile",
        choices=["baseline", "shiftedx-harness-v1"],
        default="baseline",
        help="Optional Shiftedx Agent Harness profile for the agentic suite only",
    )
    command.add_argument(
        "--thinking", choices=["on", "off"],
        help="Explicit reasoning mode for agentic requests; omit to use the endpoint default",
    )
    command.add_argument(
        "--reasoning-effort",
        help="Explicit reasoning effort for agentic requests; omit to use the endpoint default",
    )
    command.add_argument(
        "--agentic-set", choices=["core", "expanded", "repo"], default="core",
        help="Agentic scenario set; core remains the lightweight quant-gate default",
    )
    command.add_argument("--case-id", help="Run one named agentic case from the selected set")
    command.set_defaults(func=run_suite)

    command = subparsers.add_parser("summarize")
    command.add_argument("inputs", nargs="+", type=Path)
    command.add_argument("--output", required=True, type=Path)
    command.add_argument("--effective-threshold", type=float, default=0.90)
    command.add_argument("--baseline", help="Explicit baseline variant for paired comparison")
    command.set_defaults(func=summarize)

    command = subparsers.add_parser("summarize-agentic")
    command.add_argument("inputs", nargs="+", type=Path)
    command.add_argument("--output", required=True, type=Path)
    command.set_defaults(func=summarize_agentic)

    command = subparsers.add_parser("export-public-results")
    command.add_argument("inputs", nargs="+", type=Path)
    command.add_argument("--output", required=True, type=Path)
    command.add_argument("--manifest", type=Path)
    command.add_argument("--manifest-output", type=Path)
    command.set_defaults(func=export_public_results)

    command = subparsers.add_parser("run-quant-gate")
    command.add_argument("--base-url", required=True, help="OpenAI-compatible base URL ending in /v1")
    command.add_argument("--model", required=True)
    command.add_argument("--tokenizer", required=True, type=Path)
    command.add_argument("--label", required=True)
    command.add_argument("--output-dir", required=True, type=Path)
    command.add_argument("--context-config", type=Path)
    command.add_argument("--suite-config", type=Path)
    command.add_argument("--api-key-env")
    command.add_argument("--timeout", type=float, default=1800)
    command.add_argument("--thinking", choices=["on", "off"], default="on")
    command.add_argument("--reasoning-effort", default="medium")
    command.add_argument("--profile", choices=["vision", "text"], default="vision")
    command.add_argument("--hf-repo", help="Published Hugging Face repo evaluated by this run")
    command.add_argument("--hf-revision", help="Full immutable Hugging Face revision evaluated by this run")
    command.add_argument("--benchmark-revision", help="Full immutable Shiftedx Bench revision used by this run")
    command.set_defaults(func=run_quant_gate)

    command = subparsers.add_parser("render-model-card")
    command.add_argument("--run-dir", required=True, type=Path)
    command.add_argument("--readme", required=True, type=Path)
    command.add_argument("--output", required=True, type=Path)
    command.add_argument("--variant", required=True)
    command.add_argument("--profile", choices=["vision", "text"], default="vision")
    command.add_argument("--model-repo", required=True)
    command.add_argument("--model-revision", required=True)
    command.add_argument("--benchmark-revision", required=True)
    command.add_argument("--host", required=True)
    command.add_argument("--runtime", required=True)
    command.add_argument("--kv-cache", choices=["off", "q8", "q4"], required=True)
    command.add_argument("--mtp-depth", default="not-applicable")
    command.add_argument(
        "--max-window-status", choices=["not-run", "host-limited", "passed"], default="not-run"
    )
    command.set_defaults(func=render_model_card)

    command = subparsers.add_parser("validate-release")
    command.add_argument("--root", required=True, type=Path)
    command.add_argument("--output", type=Path)
    command.set_defaults(func=validate_release)

    command = subparsers.add_parser("make-holdout")
    command.add_argument("--template", required=True, type=Path)
    command.add_argument("--release-id", required=True)
    command.add_argument("--master-key-env", required=True)
    command.add_argument("--output", required=True, type=Path)
    command.set_defaults(func=make_holdout)

    command = subparsers.add_parser("probe-boundary")
    command.add_argument("--base-url", required=True)
    command.add_argument("--model", required=True)
    add_tokenizer_profile_arguments(command)
    command.add_argument("--context-window", type=int, default=262144)
    command.add_argument("--reserved-output-tokens", type=int, default=2048)
    command.add_argument("--include-positive", action="store_true")
    command.add_argument("--api-key-env")
    command.add_argument("--timeout", type=float, default=900)
    command.add_argument("--output", required=True, type=Path)
    command.set_defaults(func=probe_boundary)
    return parser


def main() -> None:
    args = make_parser().parse_args()
    args.func(args)


if __name__ == "__main__":
    main()