Skip to content

kali20gakki/msAgent / skills/quantizer/quantization-accuracy-tuning-orchestrator

quantization-accuracy-tuning-orchestrator

Verified

End-to-end automated model quantization and accuracy tuning workflow. Use when user asks for automated model quantization and accuracy tuning, e.g. "自动量化", "量化调优", "一键量化", "精度调优", etc. 当用户给出多卡(≥2 卡)卡号时,量化阶段通过 EP 并行适配(msmodelslim-ep-parallel-adaptation)保证多卡 EP 并行,调优主流程仍在本 Skill。

scripts/accuracy_lookup.py

View on GitHub
# -------------------------------------------------------------------------
# This file is part of the MindStudio project.
# Copyright (c) 2026 Huawei Technologies Co.,Ltd.
#
# MindStudio is licensed under Mulan PSL v2.
# You can use this software according to the terms and conditions of the Mulan PSL v2.
# You may obtain a copy of Mulan PSL v2 at:
#
#          http://license.coscl.org.cn/MulanPSL2
#
# THIS SOFTWARE IS PROVIDED ON AN "AS IS" BASIS, WITHOUT WARRANTIES OF ANY KIND,
# EITHER EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO NON-INFRINGEMENT,
# MERCHANTABILITY OR FIT FOR A PARTICULAR PURPOSE.
# See the Mulan PSL v2 for more details.
# -------------------------------------------------------------------------

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""Lookup accuracy cache (replaces MCP accuracy_lookup)."""

from __future__ import annotations

import argparse
import sys
from pathlib import Path
from typing import Any, Dict

# ---------------------------------------------------------------------------
# Bootstrap: add shared library to sys.path before any cross-skill imports
# ---------------------------------------------------------------------------
_common_dir = Path(__file__).resolve().parents[2] / "msmodelslim-tools-common" / "scripts"
if str(_common_dir) not in sys.path:
    sys.path.insert(0, str(_common_dir))

from script_utils import emit_result, ensure_msmodelslim


def _load_practice(practice_path: str):
    from msmodelslim.core.practice import PracticeConfig
    from msmodelslim.utils.security import yaml_safe_load

    content = yaml_safe_load(practice_path)
    return PracticeConfig.model_validate(content)


def _load_evaluate_config(evaluate_config_path: str):
    from msmodelslim.infra.service_oriented_evaluate_service import ServiceOrientedEvaluateServiceConfig
    from msmodelslim.utils.security import yaml_safe_load

    config_dict = yaml_safe_load(evaluate_config_path)
    return ServiceOrientedEvaluateServiceConfig.model_validate(config_dict)


def accuracy_lookup(save_path: str, evaluate_config_path: str, practice_path: str) -> Dict[str, Any]:
    from msmodelslim.infra.yaml_practice_accuracy_manager import YamlTuningAccuracyManager

    try:
        history_path = str(Path(save_path) / "history")
        accuracy = YamlTuningAccuracyManager().load_accuracy(history_path)
        evaluate_config = _load_evaluate_config(evaluate_config_path)
        practice_obj = _load_practice(practice_path)
        evaluate_result = accuracy.get_accuracy(practice_obj, evaluate_config)
        if evaluate_result is None:
            return {"ok": True, "cache_hit": False}
        return {
            "ok": True,
            "cache_hit": True,
            "evaluate_result": evaluate_result.model_dump(),
        }
    except Exception as exc:
        return {"ok": False, "error": str(exc)}


def main() -> int:
    parser = argparse.ArgumentParser(description="Lookup accuracy cache for practice + evaluate config.")
    parser.add_argument("--save-path", required=True)
    parser.add_argument("--evaluate-config-path", required=True)
    parser.add_argument("--practice-path", required=True)
    args = parser.parse_args()

    ensure_msmodelslim()
    return emit_result(
        accuracy_lookup(args.save_path, args.evaluate_config_path, args.practice_path)
    )


if __name__ == "__main__":
    sys.exit(main())
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall