Skip to content

Package simulator with portable Qwen calibrations#14

Merged
kiddyboots216 merged 4 commits into
apanda-devfrom
feature/training-engine-sim
Jul 10, 2026
Merged

Package simulator with portable Qwen calibrations#14
kiddyboots216 merged 4 commits into
apanda-devfrom
feature/training-engine-sim

Conversation

@kiddyboots216

@kiddyboots216 kiddyboots216 commented Jul 10, 2026

Copy link
Copy Markdown
Contributor

Summary

  • move the CPU-only training simulator into the installable xorl.sim package and add stable CLI entry points
  • port the portable analytical, timing, support, memory, feasibility, and scenario-planning core
  • add sanitized Qwen3-235B, Qwen3.5-397B, and Qwen3.6-35B calibration packs with versioned manifests and frozen goldens
  • add correctness-aware kernel-variant comparison and a consolidated pack/analytical validator
  • exclude workload manifests, cluster launchers, scheduler configuration, and storage configuration

Validation

  • pytest -q tests/experiments/test_training_sim.py (26 passed)
  • ruff check src/xorl/sim tests/experiments/test_training_sim.py
  • ruff format --check src/xorl/sim tests/experiments/test_training_sim.py
  • python -m xorl.sim.validate (3 packs, 222 checks, 0 failed)
  • built the wheel and validated xorl-sim-packs, xorl-sim-validate, xorl-sim-predict, and xorl-sim-feasibility from a clean Python 3.12 virtual environment
  • audited the shipped simulator for infrastructure manifests, launch scripts, scheduler settings, storage settings, and internal absolute paths

Base status

  • based directly on apanda-dev at 6c77c025
  • simulator-focused checks pass; repository-wide lint is currently blocked by pre-existing apanda-dev shebang-mode, ruff, format, and codespell failures outside this PR diff

@broly-code-security-scanner

broly-code-security-scanner Bot commented Jul 10, 2026

Copy link
Copy Markdown

Broly Security Scan

Note

Baseline snapshot is missing for this repo. Broly is running in PR-only fallback mode until the first scheduled baseline completes. This does not block the PR.

Note

Summary

1 actionable finding(s) in this PR

  • 🔵 1 low
Severity Scanner Issue Location Dismiss Verdict
🔵 LOW SAST Path traversal via the path subcommand allows
reading arbitrary filesystem paths.
src/xorl/sim/calibration_packs.py:53 d5 🔺 TRUE_POSITIVE · Confidence: MEDIUM

Dismiss false positives

Each dismissable row has a Dismiss key (d1, d2, …) in the table above. Reply to this comment (or post on the PR) with /broly dismiss d2: your reason to mark a false positive, or /broly undismiss d2 to reverse it. Broly records every dismissal in the section below this comment (updated in place) and suppresses the finding on the next scan.

Key Finding
d5 🔵 LOW · Path traversal via the path subcommand allows reading arbitrary filesystem ... · src/xorl/sim/calibration_packs.py:53

Broly — SAST (zai-org/GLM-5.2) · Secrets · SCA · GH Actions (zizmor) · Containers · SBOM · Powered by Together AI

@github-advanced-security

Copy link
Copy Markdown

You are seeing this message because GitHub Code Scanning has recently been set up for this repository, or this pull request contains the workflow file for the Code Scanning tool.

What Enabling Code Scanning Means:

  • The 'Security' tab will display more code scanning analysis results (e.g., for the default branch).
  • Depending on your configuration and choice of analysis tool, future pull requests will be annotated with code scanning analysis results.
  • You will be able to see the analysis results for the pull request's branch on this overview once the scans have completed and the checks have passed.

For more information about GitHub Code Scanning, check out the documentation.

Comment thread src/xorl/sim/calibration_packs.py Fixed
Comment thread src/xorl/sim/model_metadata.py Fixed
Comment thread src/xorl/sim/model_metadata.py Fixed
@kiddyboots216 kiddyboots216 changed the base branch from main to apanda-dev July 10, 2026 03:21
@kiddyboots216 kiddyboots216 force-pushed the feature/training-engine-sim branch from 464fb29 to 435178c Compare July 10, 2026 03:21
Comment thread src/xorl/sim/calibration_packs.py Fixed
Comment on lines +53 to +64
def resolve_calibration_pack(value: str | Path) -> Path:
"""Resolve a filesystem path or ``builtin:<name>`` calibration-pack reference."""

raw = str(value)
name = raw.removeprefix("builtin:")
available_names = list_calibration_packs()
if raw.startswith("builtin:") or (not Path(raw).exists() and name in available_names):
if name not in available_names:
available = ", ".join(available_names) or "none"
raise ValueError(f"unknown built-in calibration pack {name!r}; available: {available}")
return _PACK_ROOT / name
return Path(raw)
@kiddyboots216 kiddyboots216 merged commit 435178c into apanda-dev Jul 10, 2026
1 of 3 checks passed
@kiddyboots216 kiddyboots216 deleted the feature/training-engine-sim branch July 10, 2026 10:35
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants