fix: cover patched PyTorch weight-load versions

[mldangelo-oai]

Summary

• add regression coverage for vulnerable PyTorch prerelease/dev versions before weight deserialization
• add benign guards for final, local, post, and patch releases that should still use safe weights_only loading
• update the root changelog

Validation

• PYTHONPATH=/private/tmp/modelaudit-c121 PROMPTFOO_DISABLE_TELEMETRY=1 /Users/mdangelo/code/modelaudit/.venv/bin/pytest tests/scanners/test_weight_distribution_scanner.py::TestWeightDistributionScanner::test_blocks_torch_load_for_vulnerable_pytorch_prereleases tests/scanners/test_weight_distribution_scanner.py::TestWeightDistributionScanner::test_allows_torch_load_for_stable_patched_pytorch tests/scanners/test_weight_distribution_scanner.py::TestWeightDistributionScanner::test_blocks_torch_load_when_pytorch_version_unknown -q
• /Users/mdangelo/code/modelaudit/.venv/bin/ruff check modelaudit/scanners/weight_distribution_scanner.py tests/scanners/test_weight_distribution_scanner.py
• /Users/mdangelo/code/modelaudit/.venv/bin/ruff format --check modelaudit/scanners/weight_distribution_scanner.py tests/scanners/test_weight_distribution_scanner.py
• PYTHONPATH=/private/tmp/modelaudit-c121 /Users/mdangelo/code/modelaudit/.venv/bin/mypy modelaudit/scanners/weight_distribution_scanner.py tests/scanners/test_weight_distribution_scanner.py
• git diff --check

[mldangelo-oai]

@codex review

[chatgpt-codex-connector]

Codex Review: Didn't find any major issues. Another round soon, please!

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[github-actions]

Workflow run and artifacts

Performance Benchmarks

Compared 12 shared benchmarks with a regression threshold of 15%.
Status: 0 regressions, 0 improved, 12 stable, 0 new, 0 missing.
Aggregate shared-benchmark median: 769.61ms -> 774.33ms (+0.6%).

Workload	Benchmark	Target	Size	Files	Baseline	Current	Change	Status
clean-training-checkpoint	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_clean_training_checkpoint	safe_large	278.2 KiB	1	19.67ms	19.97ms	+1.5%	stable
warm-cache-rescan	tests/benchmarks/test_scan_benchmarks.py::test_scan_warm_cached_repository_rescan	release-candidate	547.3 KiB	32	65.56ms	66.38ms	+1.2%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_hex]	nested_hex	130 B	1	428.1us	423.2us	-1.1%	stable
single-checkpoint-preflight	tests/benchmarks/test_scan_benchmarks.py::test_scan_single_checkpoint_before_load	single_checkpoint.pkl	183.0 KiB	1	41.25ms	40.79ms	-1.1%	stable
direct-malicious-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_direct_malicious_upload	malicious_reduce	52 B	1	1.47ms	1.45ms	-1.1%	stable
duplicate-heavy-registry	tests/benchmarks/test_scan_benchmarks.py::test_scan_duplicate_registry_snapshot	registry-snapshot	915.2 KiB	13	218.26ms	220.62ms	+1.1%	stable
mixed-model-repository	tests/benchmarks/test_scan_benchmarks.py::test_scan_release_candidate_repository	release-candidate	547.3 KiB	32	292.24ms	294.49ms	+0.8%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_base64]	nested_base64	98 B	1	422.2us	419.3us	-0.7%	stable
chunked-upload-stream	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_chunked_upload_stream	chunked_stream	278.2 KiB	1	22.95ms	22.80ms	-0.6%	stable
nested-payload-review	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_nested_payload_review[nested_raw]	nested_raw	78 B	1	410.7us	412.4us	+0.4%	stable
suspicious-pickle-intake	tests/benchmarks/test_scan_benchmarks.py::test_scan_suspicious_pickle_intake	suspicious-intake	183.8 KiB	4	105.40ms	105.04ms	-0.3%	stable
padded-multi-stream-upload	tests/benchmarks/test_picklescan_benchmarks.py::test_picklescan_padded_multi_stream_upload	multi_stream_padded	4.1 KiB	1	1.54ms	1.53ms	-0.1%	stable

[ianw-oai]

Looks good.