Ascend/msagent / skills/accuracy/compare-result-analyzer
compare-result-analyzer
Unverified分析 msProbe compare 精度比对结果文件(CSV/XLSX),定位整网精度问题与数值差异。支持统计量模式,输出分析报告。支持辅助信息协助定位:辅助信息(历史结论/已知根因/已排除算子/嫌疑区域等)建议在请求中附带(不依赖交互工具)。若请求中未附带且运行环境无 AskUserQuestion,则跳过辅助信息收集、照常完成分析,报告中省略辅助信息章节并注明原因。
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 29.5 KB |
| assets/multi_card_report_template.md | markdown | 8.8 KB |
| assets/report_template.md | markdown | 25.9 KB |
| assets/summary_report_template.md | markdown | 4.7 KB |
| assets/verify_report_template.md | markdown | 5.7 KB |
| references/aux_info.md | markdown | 5.4 KB |
| references/constraints.md | markdown | 42.9 KB |
| references/multi_card_rules.md | markdown | 5.5 KB |
| references/operator_types.md | markdown | 9.7 KB |
| references/thresholds.md | markdown | 5.2 KB |
| references/verify_op.md | markdown | 11.4 KB |
| scripts/_common.py | script | 7.5 KB |
| scripts/_data_io.py | script | 13.0 KB |
| scripts/_noise_filter.py | script | 15.9 KB |
| scripts/_pooling.py | script | 10.2 KB |
| scripts/_propagation.py | script | 31.5 KB |
| scripts/_verify_core.py | script | 31.7 KB |
| scripts/_verify_engine.py | script | 34.0 KB |
| scripts/analyze_stat.py | script | 44.4 KB |
| scripts/cascade_threshold.py | script | 29.8 KB |
| scripts/verify_op.py | script | 3.2 KB |
scripts/verify_op.py
View on GitHub#!/usr/bin/env python3
"""单算子精度验证框架 — NPU vs CPU — CLI入口
用法:
python verify_op.py <compare.csv> --op-list "op1,op2" -o verify.json
"""
import argparse
import json
import os
import sys
from typing import Optional, Dict, List
from _verify_core import parse_csv, group_ops, get_operator_fn, _parse_op_list_item
from _verify_engine import verify_operator, VerifyResult, print_results, output_json
def main():
parser = argparse.ArgumentParser(description="单算子精度验证框架 — NPU vs CPU")
parser.add_argument("csv", nargs="?", help="msProbe compare CSV")
parser.add_argument("--op-list", help="逗号分隔算子列表")
parser.add_argument("--output", "-o", help="输出 JSON 路径")
args = parser.parse_args()
if not args.csv:
parser.print_help()
sys.exit(1)
print(f" 📂 解析 CSV: {args.csv}")
tensors = parse_csv(args.csv)
if not tensors:
print(" ⚠ 未发现 API 算子")
if args.output:
_out = args.output or os.path.join(
os.path.dirname(os.path.abspath(args.csv)),
'.compare_result_analyzer',
f'{os.path.splitext(os.path.basename(args.csv))[0]}_verify.json')
os.makedirs(os.path.dirname(_out), exist_ok=True)
output_json([], _out, atol=1e-4, rtol=1e-3)
sys.exit(0)
groups = group_ops(tensors)
print(f" 📊 {len(tensors)} tensors, {len(groups)} 算子实例")
op_directions: Dict[str, Optional[str]] = {}
if args.op_list:
raw_items = [n.strip() for n in args.op_list.split(",") if n.strip()]
for item in raw_items:
key, d = _parse_op_list_item(item)
op_directions[key] = d
groups = {k: v for k, v in groups.items() if k in op_directions}
if not groups:
print(f" ❌ 未找到: {args.op_list}")
sys.exit(1)
print(f" 🔍 过滤后: {len(groups)} 算子实例")
all_results = []
for key, group in groups.items():
try:
results = verify_operator(group, 1e-4, 1e-3,
direction=op_directions.get(key))
except Exception as e:
results = [VerifyResult(
op_name=group.op_name, instance=group.instance,
direction=op_directions.get(key, "forward"),
tensor_name="N/A", shape_match=False,
max_diff=0, l2norm_diff=0, mean_diff=0,
max_rel_err=0, mean_rel_err=0, passed=False,
error=f"验证异常: {str(e)}", construct_l2norm_err=0)]
all_results.extend(results)
print_results(all_results)
if args.output:
output_json(all_results, args.output, atol=1e-4, rtol=1e-3)
else:
d = os.path.join(os.path.dirname(os.path.abspath(args.csv)),
'.compare_result_analyzer')
os.makedirs(d, exist_ok=True)
output_json(all_results,
os.path.join(d, f'{os.path.splitext(os.path.basename(args.csv))[0]}_verify.json'),
atol=1e-4, rtol=1e-3)
failed = sum(1 for r in all_results if not r.passed)
sys.exit(1 if failed > 0 else 0)
if __name__ == "__main__":
main()
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| deprecation-and-migrationaddyosmani/agent-skillsManages deprecation and migration. Use when removing old systems, APIs, or features. Use when migrating users from one implementation to another. Use when migrating a database schema in production, such as renaming or dropping a column without downtime (expand/contract). Use when deciding whether to maintain or sunset existing code. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| performance-optimizationaddyosmani/agent-skillsOptimizes application performance across frontend, backend, queries, and databases. Use when performance requirements exist, when you suspect performance regressions, when Core Web Vitals or load times need improvement, when N+1 query patterns need fixing, or when profiling reveals bottlenecks. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| Agent Cost Reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| Agent Cost Reportthedotmack/claude-memBelievable agent cost report for any period, default the last 7 full days PT, not counting today. Measured tokens from Claude Code transcripts priced at OpenRouter list prices (ESTIMATED), measured provider spend when a sanctioned source exists, note-taker cost separate, Timing-style HTML/PDF plus report.json, line-items.csv, evidence.json. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |