Closed
[Documentation|文档反馈]: aclnnDequantSwigluQuantV2 quantScale/quantOffset 未声明 static+groupIndex 下的形状([1] 被拒 561002) #7559
liuhanwen8Add on 24 days agoClosed at 16 days ago
24 days ago Added label: documentation
yuning_chen
24 days ago Comments:
24 days ago Comments:
24 days ago Assign east_yang as responsible person
yuning_chen
24 days ago Comments:
24 days ago Comments:
您好,该问题已收到,当前问题 @east_yang 正在处理中


east_yang
19 days ago Comments:
19 days ago Comments:
/unassign @east_yang


19 days ago Removed responsible person east_yang
east_yang
19 days ago Comments:
19 days ago Comments:
/assign @fengying555682


19 days ago Assign fengying555682 as responsible person
east_yang
19 days ago Comments:
19 days ago Comments:
/assign @guijianwei


CANN-robot
19 days ago Comments:
19 days ago Comments:
18 days ago Assign guijianwei as responsible person
18 days ago Removed responsible person fengying555682
guijianwei
18 days ago Comments:
18 days ago Comments:
/assign @jzj007


CANN-robot
18 days ago Comments:
18 days ago Comments:
18 days ago Assign jzj007 as responsible person
18 days ago Removed responsible person guijianwei
17 days ago Associated pull request
16 days ago Close issue
16 days ago Added label: resolved
现象
《aclnnDequantSwigluQuantV2》的接口说明里
quantScaleOptional只写了 static 模式下quantScale的形状为[1],没有声明「static +groupIndexOptional非空」这一组合的形状要求;quantOffsetOptional行同样只说「数据类型与shape同quantScale」。按资料字面构造的[1]在该组合下被接口拒绝(561002),调用方从资料里找不到正确形状。quantScaleOptional的形状说明原文(逐字):同一格里紧随其后的三句分别给出:dynamic 时可为
[H]、[2H]或[groupNum, H];groupIndexOptional为空指针且activateDim为尾轴时[H];groupIndexOptional不为空指针且activateDim为尾轴时[groupNum, H]。后两句没有限定只对 dynamic 生效,但都排在 dynamic 之后。即:唯一提到 static 的句子给的是
[1];提到groupIndexOptional非空的两句都跟在 dynamic 的语境之后,读者无法判断它们对 static 是否生效。实测(910B2 / CANN 9.2.0,
quantMode="static"、activateDim=-1、x=INT32、groupIndex=[4,4],H=64、groupNum=2):[1])[1][1][2,64][2,64][2,64][1][1][2,64]无
groupIndex时[1]正常返回 0,说明[1]本身不是非法形状,缺的是「分组场景要[groupNum, H]」这条声明。期望
资料应显式写出「static +
groupIndexOptional非空」时quantScale/quantOffset的形状为[groupNum, H],因为实现要求如此(上表 4 组 A/B:只有[2,64]+[2,64]返回 0),而现有文字在这一组合下给的是[1]。判定依据是仓内资料本身:
quantScaleOptional的形状说明里所有含static的句子,没有任何一句同时给出groupNum维度。下文复现脚本对这一条做文本断言,找不到即判缺陷。环境
Ascend-cann-toolkit 9.2.0)cann/ops-nn,master0e188786quant/dequant_swiglu_quant/docs/aclnnDequantSwigluQuantV2.md复现步骤
cann/ops-nn的 master(0e188786)。check_doc.py。python3 check_doc.py --repo-root <ops-nn 仓根>。[DEFECT] static 分支的形状句未覆盖 groupIndex 场景,退出码 1。aclnnDequantSwigluQuantV2GetWorkspaceSize,只有[2,64]+[2,64]返回 0。复现脚本
import os import re import sys def main(): repo = sys.argv[sys.argv.index("--repo-root") + 1] if "--repo-root" in sys.argv else "." path = os.path.join(repo, "quant/dequant_swiglu_quant/docs/aclnnDequantSwigluQuantV2.md") if not os.path.isfile(path): print("[SKIP] 找不到资料:%s" % path) return 2 txt = open(path, encoding="utf-8").read() head = "quantScaleOptional(aclTensor*)" tail = "quantOffsetOptional(aclTensor*)" if head not in txt or tail not in txt: print("[SKIP] 资料结构已变,找不到 quantScaleOptional 表格行") return 2 cell = txt.split(head, 1)[1].split(tail, 1)[0] cell = re.sub(r"<[^>]+>", "", cell) sents = [s.strip() for s in re.split(r"[。;\n]", cell) if "static" in s] if not sents: print("[SKIP] 该单元格没有 static 分支说明") return 2 bad = [s for s in sents if "groupNum" not in s] if bad: print("[DEFECT] static 分支的形状句未覆盖 groupIndex 场景:%s" % bad[0][:160]) return 1 print("[OK] static 分支已覆盖 groupNum") return 0 sys.exit(main())实测运行输出:
上表 5 组实测对应的 C 直调参数(
x=INT32[8,128]、weightScale=[2,128]、activationScale=[8]、groupIndex=[4,4]、y=INT8[8,64]、scaleOut=[8]、quantMode="static"、dstType=2、roundMode="rint"、activateDim=-1、swigluMode=0、clampLimit=7.0、gluAlpha=1.702、gluBias=1.0),逐组只改quantScale/quantOffset形状。影响
静态量化 + MoE 分组(
groupIndexOptional非空)是按资料写代码的必然路径,这一组合的形状没有声明,调用方只能靠试错;失败码561002只说「参数非法」,不指向具体张量。本轮 1000 条用例的构造阶段因此失败 173 条(17%),按[groupNum,H]修正后降到 46 条(剩余为另一条 dtype 组合问题)。建议
在
quantScaleOptional的形状说明补一行,按「quantMode × groupIndex」给形状:[1][groupNum, H][H][groupNum, H]并在
quantOffsetOptional行保留「形状同 quantScale」。