AGENT SKILL

proxy_recovery_evaluator

Evaluate whether an SR3 soft-mode proxy recovery exercised conditioning, diffusion loss, optimizer update, and iterative refinement evidence.

License 许可
Apache-2.0
最近更新
2026-09-03
Stars
★ 351
Forks
27

01 前言 Front Matter

name
proxy_recovery_evaluator
description
Evaluate whether an SR3 soft-mode proxy recovery exercised conditioning, diffusion loss, optimizer update, and iterative refinement evidence.

02 资源文件 2

03 SKILL.md 正文

Proxy Recovery Evaluator

Use this skill at the end of an SR3 reduced recovery to assemble metrics and mechanism checks. Do not use it to override failing training traces or missing source-boundary evidence.

Inputs

  • Pair metadata from the conditional protocol.
  • Training trace from the diffusion objective.
  • Sampler trajectory from iterative refinement.
  • Runtime blocker information.

Outputs

  • Numeric metrics such as loss decrease and final distance improvement.
  • Mechanism booleans required for soft-mode proxy acceptance.

Workflow

  1. Confirm that the pair is explicitly marked proxy when no full dataset exists.
  2. Confirm loss decreased and parameters changed.
  3. Confirm the reverse trajectory has multiple refinement steps.
  4. Keep full-runtime booleans false when full packages/models are unavailable.
  5. Emit mechanism checks for recovery validation and analysis.

Validation

Run python tests/test_evaluator.py or validate the skill tree with --run-tests.

Limitations

Passing proxy checks supports mechanism-faithful reduced recovery only; it is not a claim of paper-level image metrics.

04 同仓库更多 Skills

VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-triton-opt

Make a single PassNet Triton kernel fast AND numerically faithful: performance model (when a replacement can win at all), block/grid/warp tuning, autotune policy, launch overhead, and per-op numeric recipes to pass the dtype baseline tolerances. Use when a pass matches and is correct but speedup below expected, or when correctness fails by small numeric margins.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-solve

END-TO-END playbook for solving one PassNet sample: analyze the computation graphs, decide the optimization strategy (what to fuse, what to replace, what to leave alone), drive the iteration loop, and maximize the sample score. This is the ENTRY skill — invoke it first for any PassNet optimization task; it tells you when to use passnet-pattern-fusion, passnet-triton-opt and passnet-feedback.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-skill

PassNet GPU kernel optimization via compiler passes. Design and implement Triton-based optimization passes, create pass files under ./pass_dir/, self-evaluate with pass_evaluator, and iterate to maximize GPU speedup.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-pattern-fusion

Author PassNet pass files: write patterns that actually MATCH the FX graph, pick fusion regions (which ops to absorb into one Triton kernel), and structure multi-pass submissions with the mandatory shared-dispatch architecture. Use when creating or fixing pattern()/replacement_args()/replacement_func() files, when a pass "failed to match", or when deciding how to fuse multiple kernels/ops.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-orchestrate

Round/batch planning and decision support for PassNet graph optimization. Use for multi-sample triage, worker allocation, eval-budget policy, and post-eval keep/revert decisions. For solving one concrete sample end to end, use passnet-solve as the entry skill; consult this skill only when you need broader planning or a specific decision gate.

★ 351
VectorSpaceLab
VectorSpaceLab/AREX-Skill

passnet-feedback

Fast, deterministic feedback for PassNet work: pre-flight pattern/match verification WITHOUT burning a GPU evaluation, per-node bottleneck analysis of a sample's graphs, and eval-log parsing into per-variant status + estimated score + failure classification. Use BEFORE every GPU evaluation (check_pattern), at the START of a sample (analyze_graph), and AFTER every evaluation (parse_eval_log).

★ 351