lexiflow_recovery_evaluation
Evaluate whether a recovery run exercised LexiFlow mechanisms and achieved a declared proxy or benchmark target with auditable metrics.
- License 许可
- Apache-2.0
- 最近更新
- 2026-09-03
- Stars
- ★ 351
- Forks
- 27
01 前言 Front Matter
- name
- lexiflow_recovery_evaluation
- description
- Evaluate whether a recovery run exercised LexiFlow mechanisms and achieved a declared proxy or benchmark target with auditable metrics.
02 资源文件 2
03 SKILL.md 正文
LexiFlow Recovery Evaluation
Use this skill when implementing or auditing the LexiFlow paper mechanism for bounded recovery or reusable optimization tooling. Do not use it as a generic scalarized multi-objective optimizer; it assumes priority-ordered minimization objectives, optional goals, and non-negative tolerances.
Inputs
- LexiFlow run trace
- Baseline run trace
- Module invocation records
- Declared paper or proxy target
Outputs
- Numeric metrics including lexi_success_rate and second-objective gain
- Mechanism-check booleans
- Validation-ready recovery summary
Workflow
- Confirm objective values are ordered by priority and are minimization values.
- Preserve the paper mechanism described in the module document rather than replacing it with weighted scalarization.
- Run the companion script or import its pure functions for deterministic behavior.
- Save command outputs, trace files, and metrics when the skill is used in a recovery experiment.
Validation
Run python scripts/lexiflow_recovery_evaluation.py --self-test when available, then run the bundled test command through the Distiller skill-tree validator.
Limitations
This generated skill is source-repository independent and derived from the paper text. Full HPO benchmark reproduction still requires real model training stacks and datasets; the included scripts are intended for deterministic mechanism validation and bounded proxy recovery.
04 同仓库更多 Skills
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.
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.
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.
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.
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.
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).