Skip to content
版 本

Cast

产 品 支 持 情 况

产 品

是 否 支 持

Ascend 950PR/Ascend 950DT

Atlas A3 训 练 系 列 产 品/Atlas A3 推 理 系 列 产 品

Atlas A2 训 练 系 列 产 品/Atlas A2 推 理 系 列 产 品

Atlas 200I/500 A2 推 理 产 品

Atlas 推 理 系 列 产 品AI Core

Atlas 推 理 系 列 产 品Vector Core

x

Atlas 训 练 系 列 产 品

Kirin X90

Kirin 9030

功 能 说 明

根 据 源 操 作 数 和 目 的 操 作 数tensor的 数 据 类 型 进 行 精 度 转 换。

函 数 原 型

  • tensor前n个 数 据 计 算。

    C++
    template <typename T, typename U>
    __aicore__ inline void Cast(const LocalTensor<T>& dst, const LocalTensor<U>& src, const RoundMode& roundMode, const uint32_t count)
    
  • tensor高 维 切 分 计 算。

    • mask逐bit模 式。

      C++
      template <typename T, typename U, bool isSetMask = true>
      __aicore__ inline void Cast(const LocalTensor<T>& dst, const LocalTensor<U>& src, const RoundMode& roundMode, const uint64_t mask[], const uint8_t repeatTime, const UnaryRepeatParams& repeatParams)
      
    • mask连 续 模 式。

      C++
      template <typename T, typename U, bool isSetMask = true>
      __aicore__ inline void Cast(const LocalTensor<T>& dst, const LocalTensor<U>& src, const RoundMode& roundMode, const uint64_t mask, const uint8_t repeatTime, const UnaryRepeatParams& repeatParams)
      

参 数 说 明

表 1 模 板 参 数 说 明

参 数 名描 述
T输 出Tensor的 数 据 类 型。
U输 入Tensor的 数 据 类 型。
isSetMask是 否 在 接 口 内 部 设 置mask,默 认 为true。
•true,表 示 在 接 口 内 部 设 置mask。
•false,表 示 在 接 口 外 部 设 置mask,开 发 者 需 要 使 用SetVectorMask接 口 设 置mask值。这 种 模 式 下,本 接 口 入 参 中 的mask值 必 须 设 置 为 占 位 符MASK_PLACEHOLDER。

表 2 参 数 说 明

参 数 名输 入/输 出描 述
dst输 出目 的 操 作 数。
类 型 为LocalTensor,支 持 的TPosition为VECIN/VECCALC/VECOUT。
src输 入源 操 作 数。
类 型 为LocalTensor,支 持 的TPosition为VECIN/VECCALC/VECOUT。
roundMode输 入精 度 转 换 处 理 模 式,类 型 是RoundMode。
RoundMode为 枚 举 类 型,用 以 控 制 精 度 转 换 处 理 模 式,可 参 考精 度 转 换 规 则。RoundMode取 值 说 明 请 参 考RoundMode取 值 说 明
count输 入参 与 计 算 的 元 素 个 数。
注:参 数 取 值 范 围 和 操 作 数 的 数 据 类 型 有 关,数 据 类 型 不 同,能 够 处 理 的 元 素 个 数 最 大 值 不 同,最 大 处 理 的 数 据 量 不 能 超 过UB大 小 限 制。
mask/mask[]输 入mask用 于 控 制 每 次 迭 代 内 参 与 计 算 的 元 素。
注 意:数 据 类 型 转 换 的mask会 按 照 输 入 和 输 出 类 型 中sizeof(dtype)较 大 的 来 筛 选。
设 置 详 见掩 码 操 作
repeatTime输 入重 复 迭 代 次 数。
矢 量 计 算 单 元,每 次 读 取 连 续 的256Bytes数 据 进 行 计 算,为 完 成 对 输 入 数 据 的 处 理,必 须 通 过 多 次 迭 代(repeat)才 能 完 成 所 有 数 据 的 读 取 与 计 算。repeatTime表 示 迭 代 的 次 数。
关 于 该 参 数 的 具 体 描 述 请 参 考高 维 切 分
repeatParams输 入控 制 操 作 数 地 址 步 长 的 参 数。UnaryRepeatParams类 型,包 含 操 作 数 相 邻 迭 代 间 相 同DataBlock的 地 址 步 长,操 作 数 同 一 迭 代 内 不 同DataBlock的 地 址 步 长 等 参 数。
相 邻 迭 代 间 的 地 址 步 长 参 数 说 明 请 参 考repeatStride;同 一 迭 代 内DataBlock的 地 址 步 长 参 数 说 明 请 参 考dataBlockStride

RoundMode取 值 说 明

枚 举 类 型 具 体 定 义 为:

C++
enum class RoundMode {
  CAST_NONE = 0,  // 在 转 换 有 精 度 损 失 时 表 示CAST_RINT模 式,不 涉 及 精 度 损 失 时 表 示 不 舍 入
  CAST_RINT,      // rint,四 舍 六 入 五 成 双 舍 入
  CAST_FLOOR,     // floor,向 负 无 穷 舍 入
  CAST_CEIL,      // ceil,向 正 无 穷 舍 入
  CAST_ROUND,     // round,四 舍 五 入 舍 入
  CAST_TRUNC,     // trunc,向 零 舍 入
  CAST_ODD,       // Von Neumann rounding,最 近 邻 奇 数 舍 入
  CAST_HYBRID,    // hybrid,目 前 特 指 输 出 结 果 是hifloat8_t数 据 时,会 用 到 的 一 种 随 机 舍 入
};

:仅 在 支 持hifloat8_t数 据 类 型 的 硬 件 平 台,才 支 持CAST_HYBRID转 换 模 式。

数 据 类 型

不 同 数 据 类 型 转 换 支 持 的 舍 入 模 式 不 同,可 参 考精 度 转 换 规 则

Ascend 950PR/Ascend 950DT支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
int4b_tint16_tCAST_NONE
int4b_thalfCAST_NONE
int4b_tbfloat16_tCAST_NONE
int8_tint16_tCAST_NONE
int8_thalfCAST_NONE
int8_tint32_tCAST_NONE
uint8_tuint16_tCAST_NONE
uint8_thalfCAST_NONE
uint8_tuint32_tCAST_NONE
fp4x2_e2m1_tbfloat16_tCAST_NONE
fp4x2_e1m2_tbfloat16_tCAST_NONE
hifloat8_thalfCAST_NONE
hifloat8_tfloatCAST_NONE
fp8_e5m2_tfloatCAST_NONE
fp8_e4m3fn_tfloatCAST_NONE
int16_tint4b_tCAST_NONE
int16_tuint8_tCAST_NONE
int16_thalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
int16_tint32_tCAST_NONE
int16_tuint32_tCAST_NONE
int16_tfloatCAST_NONE
uint16_tuint8_tCAST_NONE
uint16_tuint32_tCAST_NONE
halfint4b_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfuint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfhifloat8_tCAST_ROUND/CAST_HYBRID
halfint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfbfloat16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halffloatCAST_NONE
bfloat16_tfp4x2_e2m1_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
bfloat16_tfp4x2_e1m2_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
bfloat16_thalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
bfloat16_tint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
bfloat16_tfloatCAST_NONE
int32_tuint8_tCAST_NONE
int32_tint16_tCAST_NONE
int32_tuint16_tCAST_NONE
int32_thalfroundMode不 生 效,与SetDeqScale(half scale)接 口 配 合 使 用。
int32_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
int32_tint64_tCAST_NONE
uint32_tuint8_tCAST_NONE
uint32_tint16_tCAST_NONE
uint32_tuint16_tCAST_NONE
floathifloat8_tCAST_ROUND/CAST_HYBRID
floatfp8_e5m2_tCAST_RINT
floatfp8_e4m3fn_tCAST_RINT
floatint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floathalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_ODD/CAST_NONE
floatbfloat16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint64_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
complex32complex64CAST_NONE
int64_tint32_tCAST_NONE
int64_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
int64_tdoubleCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
doublebfloat16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
doubleint32_tCAST_TRUNC
doublefloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
doubleint64_tCAST_TRUNC
complex64complex32CAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_ODD/CAST_NONE
complex64complex64CAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC

Atlas A3 训 练 系 列 产 品/Atlas A3 推 理 系 列 产 品 支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
int4b_thalfCAST_NONE
int8_thalfCAST_NONE
uint8_thalfCAST_NONE
int16_thalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
int16_tfloatCAST_NONE
halfint4b_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfuint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halffloatCAST_NONE
bfloat16_tint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
bfloat16_tfloatCAST_NONE
int32_tint16_tCAST_NONE
int32_thalfroundMode不 生 效,与SetDeqScale(half scale)接 口 配 合 使 用。
int32_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
int32_tint64_tCAST_NONE
floatint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floathalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_ODD/CAST_NONE
floatbfloat16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint64_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
int64_tint32_tCAST_NONE
int64_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC

Atlas A2 训 练 系 列 产 品/Atlas A2 推 理 系 列 产 品 支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
int4b_thalfCAST_NONE
int8_thalfCAST_NONE
uint8_thalfCAST_NONE
int16_thalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
int16_tfloatCAST_NONE
halfint4b_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfuint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halffloatCAST_NONE
bfloat16_tint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
bfloat16_tfloatCAST_NONE
int32_tint16_tCAST_NONE
int32_thalfroundMode不 生 效,与SetDeqScale(half scale)接 口 配 合 使 用。
int32_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
int32_tint64_tCAST_NONE
floatint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floathalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_ODD/CAST_NONE
floatbfloat16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint64_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
int64_tint32_tCAST_NONE
int64_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC

Atlas 200I/500 A2 推 理 产 品 支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
int8_thalfCAST_NONE
uint8_thalfCAST_NONE
int16_thalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
int16_tfloatCAST_NONE
halfint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfuint8_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halfint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halffloatCAST_NONE
bfloat16_tint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
bfloat16_tfloatCAST_NONE
int32_tint16_tCAST_NONE
int32_thalfCAST_NONE
int32_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
int32_tint64_tCAST_NONE
floatint16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floathalfCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_ODD/CAST_NONE
floatbfloat16_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
floatint64_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
int64_tint32_tCAST_NONE
int64_tfloatCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC

Atlas 推 理 系 列 产 品AI Core支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
int8_thalfCAST_NONE
uint8_thalfCAST_NONE
int16_thalfCAST_NONE
halfint4b_tCAST_NONE
halfint8_tCAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfuint8_tCAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfint16_tCAST_RINT
halfint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halffloatCAST_NONE
int32_tint16_tCAST_NONE
int32_thalfroundMode不 生 效,与SetDeqScale(half scale)接 口 配 合 使 用。
int32_tfloatCAST_NONE
floathalfCAST_ODD/CAST_NONE
floatint64_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC

Atlas 训 练 系 列 产 品 支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
int8_thalfCAST_NONE
uint8_thalfCAST_NONE
int16_thalfCAST_NONE
halfint8_tCAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfuint8_tCAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC/CAST_NONE
halfint32_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC
halffloatCAST_NONE
int32_tfloatCAST_NONE
floathalfCAST_ODD/CAST_NONE
floatint64_tCAST_RINT/CAST_FLOOR/CAST_CEIL/CAST_ROUND/CAST_TRUNC

Kirin X90支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
halfuint8_tCAST_CEIL
halfint32_tCAST_FLOOR
int32_tfloatCAST_CEIL
floatbfloat16_tCAST_ROUND
bfloat16_tfloatCAST_NONE

Kirin 9030支 持 的 数 据 类 型 组 合 情 况

src数 据 类 型dst数 据 类 型支 持 的roundMode
halfuint8_tCAST_CEIL
halfint32_tCAST_FLOOR
int32_tfloatCAST_CEIL
floatbfloat16_tCAST_ROUND
bfloat16_tfloatCAST_NONE

返 回 值 说 明

无。

约 束 说 明

  • 操 作 数 地 址 对 齐 要 求 请 参 见通 用 地 址 对 齐 约 束

  • 操 作 数 地 址 重 叠 约 束 请 参 考通 用 地 址 重 叠 约 束。特 别 地,对 于 长 度 较 小 的 数 据 类 型 转 换 为 长 度 较 大 的 数 据 类 型 时,地 址 重 叠 可 能 会 导 致 结 果 错 误。

  • 当 源 操 作 数 和 目 的 操 作 数 位 数 不 同 时,计 算 输 入 参 数 以 数 据 类 型 的 字 节 较 大 的 为 准。例 如,源 操 作 数 为half类 型,目 的 操 作 数 为int32_t类 型 时,每 次 迭 代 最 多 操 作64个 元 素,为 保 证 输 出 和 输 入 是 连 续 的,dstRepStride应 设 置 为8,srcRepStride应 设 置 为4。

  • 当dst或src为int4b_t时,由 于 一 个int4b_t只 占 半 个 字 节,故 申 请Tensor空 间 时,只 需 申 请 相 同 数 量 的int8_t数 据 空 间 的 一 半。host侧 目 前 暂 不 支 持int4b_t,故 在 申 请int4b_t类 型 的tensor时,应 先 申 请 一 个 类 型 为int8_t的tensor,再 用Reinterpretcast接 口 转 化 为int4b_t类 型 的tensor,接 着 调 用Cast指 令。

  • 当dst或src为int4b_t时,tensor高 维 切 分 计 算 接 口 的 连 续 模 式 的mask与tensor前n个 数 据 计 算 接 口 的count必 须 为 偶 数;对 于tensor高 维 切 分 计 算 接 口 的 逐bit模 式,对 应 同 一 字 节 的 相 邻 两 个 比 特 位 的 数 值 必 须 一 致,即0-1位 数 值 一 致,2-3位 数 值 一 致,4-5位 数 值 一 致,以 此 类 推。

  • int32_t到half数 据 类 型 的 转 换,设 置 舍 入 模 式 无 效,需 要 与SetDeqScale接 口 配 合 使 用。

  • 针 对Ascend 950PR/Ascend 950DT,complex32/complex64/double数 据 类 型 仅 支 持tensor前n个 数 据 计 算 接 口。

调 用 示 例

本 样 例 中 只 展 示Compute流 程 中 的 部 分 代 码。本 样 例 的srcLocal为half类 型,dstLocal为int32_t类 型,计 算mask时 以int32_t为 准。

如 果 您 需 要 运 行 样 例 代 码,完 整 的 调 用 样 例 请 参 考Cast样 例

  • tensor高 维 切 分 计 算 样 例-mask连 续 模 式。

    C++
    uint64_t mask = 256 / sizeof(int32_t);
    // repeatTime = 8, 64 elements one repeat, 512 elements total
    // dstBlkStride, srcBlkStride = 1, no gap between blocks in one repeat
    // dstRepStride = 8, srcRepStride = 4, no gap between repeats
    AscendC::Cast(dstLocal, srcLocal, AscendC::RoundMode::CAST_CEIL, mask, 8, { 1, 1, 8, 4 });
    
  • tensor高 维 切 分 计 算 样 例-mask逐bit模 式。

    C++
    uint64_t mask[2] = { 0, UINT64_MAX };
    // repeatTime = 8, 64 elements one repeat, 512 elements total
    // dstBlkStride, srcBlkStride = 1, no gap between blocks in one repeat
    // dstRepStride = 8, srcRepStride = 4, no gap between repeats
    AscendC::Cast(dstLocal, srcLocal, AscendC::RoundMode::CAST_CEIL, mask, 8, { 1, 1, 8, 4 });
    
  • tensor前n个 数 据 计 算 样 例。

    C++
    uint32_t count = 512; // 参 与 计 算 的 元 素 个 数
    AscendC::Cast(dstLocal, srcLocal, AscendC::RoundMode::CAST_CEIL, count);
    

根 据 不 同 的RoundMode取 值,输 出 结 果 会 有 差 异,下 面 样 例 以RoundMode::CAST_CEIL(向 正 无 穷 舍 入)为 例。

结 果 示 例 如 下:

输 入 数 据(srcLocal): 
[1.4, 1.5, 1.6, 2.4, 2.5, 2.6, ... 2.6]
输 出 数 据(dstLocal): 
[2, 2, 2, 3, 3, 3, ... 3]

当RoundMode为RoundMode::CAST_NONE(half转int32_t有 精 度 损 失,此 时 同CAST_RINT模 式)或RoundMode::CAST_RINT(四 舍 六 入 五 成 双 舍 入)时,结 果 示 例 如 下:

输 入 数 据(srcLocal): 
[1.4, 1.5, 1.6, 2.4, 2.5, 2.6, ... 2.6]
输 出 数 据(dstLocal): 
[1, 2, 2, 2, 2, 3, ... 3]

当RoundMode为RoundMode::CAST_FLOOR(向 负 无 穷 舍 入)时,结 果 示 例 如 下:

输 入 数 据(srcLocal): 
[1.4, 1.5, 1.6, 2.4, 2.5, 2.6, ... 2.6]
输 出 数 据(dstLocal): 
[1, 1, 1, 2, 2, 2, ... 2]

当RoundMode为RoundMode::CAST_ROUND(四 舍 五 入 舍 入)时,结 果 示 例 如 下:

输 入 数 据(srcLocal): 
[1.4, 1.5, 1.6, 2.4, 2.5, 2.6, ... 2.6]
输 出 数 据(dstLocal): 
[1, 2, 2, 2, 3, 3, ... 3]

当RoundMode为RoundMode::CAST_TRUNC(向 零 舍 入)时,结 果 示 例 如 下:

输 入 数 据(srcLocal): 
[1.4, 1.5, 1.6, 2.4, 2.5, 2.6, ... 2.6]
输 出 数 据(dstLocal): 
[1, 1, 1, 2, 2, 2, ... 2]
  • 当Cast涉 及int4b_t时,调 用 示 例 如 下:

    dstLocal为int8_t类 型,srcLocal为half类 型。

    C++
    inBufferSize_ = srcSize;  // src buffer size
    outBufferSize_ = srcSize / 2;   //dst buffer size
    uint64_t mask = 128;
    AscendC::LocalTensor<half> srcLocal;
    srcLocal.SetSize(inBufferSize_);
    AscendC::LocalTensor<int8_t> dstLocal;
    dstLocal.SetSize(outBufferSize_);
    AscendC::LocalTensor<AscendC::int4b_t> dstLocalTmp = dstLocal.ReinterpretCast<AscendC::int4b_t>();
    // repeatTime = 1, 128 elements one repeat, 128 elements total
    // dstBlkStride, srcBlkStride = 1, no gap between blocks in one repeat
    // dstRepStride = 2, srcRepStride = 8, no gap between repeats
    AscendC::Cast<AscendC::int4b_t, half>(dstLocalTmp, srcLocal, AscendC::RoundMode::CAST_CEIL, mask, 1, {1, 1, 2, 8});
    

更 多 样 例

您 可 以 参 考 以 下 样 例,了 解 如 何 使 用Cast指 令 的tensor高 维 切 分 计 算 接 口,进 行 更 灵 活 的 操 作、实 现 更 高 级 的 功 能。

通 过tensor高 维 切 分 计 算 接 口 中 的mask连 续 模 式,实 现 数 据 非 连 续 计 算。

uint64_t mask = 32;  // 每 个 迭 代 内 只 计 算 前32个 数
AscendC::Cast(dstLocal, srcLocal, AscendC::RoundMode::CAST_CEIL, mask, 8, { 1, 1, 8, 4 });

结 果 示 例 如 下,输 出 数 据 中 每 个 迭 代 的 前32个 数 以 外 的 数 据 是 未 初 始 化 数 据:

输 入 数 据(srcLocal): 
[37.4     7.11   53.5    19.44   22.66   43.     43.16    5.316  74.2
 15.7    87.75   86.94   92.56   25.45   36.06   94.6    73.6    30.48
 48.16   12.55   27.81   14.67    6.58   48.38   67.5    57.5    63.3
 85.2     3.654  68.7    52.53   16.38   13.945  63.84   87.2    82.5
 85.7    27.78   15.41   41.66   31.38   14.65   88.25    0.0332 43.06
 46.88   15.57   87.1    53.16   33.5    91.06   36.5    55.34   60.53
  3.238  23.92   97.5    91.1    78.44   54.47   82.     53.8    72.1
 25.06   32.12   15.88   33.38   36.7    33.3    84.4    19.25    1.743
 46.16   22.06    4.582  71.1    15.94   22.23   53.47   17.05   48.56
 94.44   77.4    90.2    46.56   92.4     9.45   68.44   35.7    31.62
 68.1    63.7    77.     92.06   20.45   27.67   93.4    22.39   17.22
 73.06    7.12   25.34   36.34   13.54   38.12   24.56   86.56   69.7
 68.3    30.38   68.4    86.1    54.44   70.     55.3    48.6    59.03
 64.44   15.45   66.5    92.7    60.7    52.22   47.     99.75   41.94
 43.06   89.5    36.9    62.5     1.306  48.06    9.37   62.25   20.61
 43.8    69.25   27.22   71.44   52.75   11.82   80.6    63.44   53.22
 85.44   25.25    2.309  26.88   84.5    29.83    9.93   81.9    97.75
 75.75   97.7    72.     19.86   26.62   88.7    74.06    9.24   42.5
 14.      39.44   98.56   66.94   89.     57.12   39.     11.57   19.05
 86.56   32.66   19.25   99.3    95.6    58.7    79.6    37.38   65.
 75.7     8.586  77.7     2.68   75.7    77.56   39.1    39.72   64.06
 98.44   30.27   31.9    94.4    85.94    4.965   2.758  92.4    49.53
 50.75    5.7    19.69   87.6    20.08   88.8    87.4    63.6    68.3
 78.9    45.66   10.01   35.25   71.9    37.38   39.7    43.47   11.67
 64.3    35.62   74.3    59.3    28.69   29.56   23.14   36.22    4.88
 70.5    25.05   72.6    71.6    32.28   34.66   80.     96.1    98.7
 12.91   95.4    61.97   87.94   19.1    40.47   89.6    84.     29.72
 17.8    81.44   23.25   33.03   18.67   78.     49.62   63.1    72.75
 77.25    3.74   38.9    17.92   76.     25.62   34.53   84.     32.03
 57.3     9.21    6.836  68.9    35.78   96.75   56.3    96.1    23.45
 78.75   94.25   12.44   56.7    24.55   25.11   90.7    50.94   78.4
  3.576  21.81   53.28   26.2    43.1     7.742  13.4    86.44   86.9
 13.93   16.48   91.06   42.3    95.5    66.8    40.6    98.06   71.9
 67.6    55.9    82.44   93.75   41.53   23.62   40.12   40.53   80.7
 80.25   96.3    51.38   93.6    91.3    32.84   88.     69.7    63.16
 41.75   43.22   43.22   31.73   84.9    91.6    80.     53.34   27.12
 76.6    97.25   44.5    30.28   74.3    76.06   40.     41.28   37.72
 99.56   18.73   16.45   92.75   79.1    40.3    68.     23.98   88.7
 86.6    24.97   59.6    28.25   82.94   46.12   60.12   34.53   79.7
 11.086  20.25   44.88   39.97   42.12   62.7    30.66   42.56   16.69
 85.2    90.8    78.75   26.16   18.14   94.06   40.3    20.16   38.
 12.99   95.44   76.25   26.03   76.     30.06   27.25   84.56   30.45
 66.1    83.25    3.732  39.1    54.22   82.8    43.22   53.03   11.66
 88.1     6.83   66.8    44.4     7.5    24.77   74.4    35.9    79.75
 41.62   37.06   60.12   57.9    96.94   84.25   39.88   22.55   72.7
 58.9    44.75   90.4    46.34   71.3    16.4    26.12   21.45   10.27
 2.      41.53   39.03   80.25    2.11    7.88   72.2    27.83   88.1
 67.56   10.72   52.84   91.2    97.6    51.44   74.7     3.527  79.25
 11.3    19.16   39.53    3.469  98.7    45.72   40.16   47.1    71.8
 11.81   52.97   71.44   37.7    26.81   46.22   26.94    4.805  12.18
 70.4    51.4    24.2    83.9     9.62   12.445  57.6    85.8    55.12
 88.25   32.38   62.88    1.903  47.72   35.9    48.94   86.06   32.44
  1.219  35.56   49.78   49.97   24.45   94.5    99.94   44.72    3.404
 83.6    23.14   76.7    91.7    24.33   20.62   24.72    4.55   88.94
 87.44   95.75   41.56   13.77   34.6    95.94   77.1    24.28   70.06
 10.06   11.38   88.8    57.22   94.56   35.     79.8    58.22   44.06
 26.9    16.25   99.94   51.1    42.38   84.25    0.9604 48.1   ]

输 出 数 据(dstLocal): 
[        38          8         54         20         23         43
         44          6         75         16         88         87
         93         26         37         95         74         31
         49         13         28         15          7         49
         68         58         64         86          4         69
         53         17 1879993057 1827499998 1823960025 1570990114
 1828150463 1811639312 1794470101 1754296176 1888841335 1715628997
 1839753994 1850888497 1889364175 1891068936 1823369913 1769105534
 1815638091 1808559970 1601662785 1739089473 1863146361 1694785989
 1597138938 1836478181 1888774249 1637707434 1877372650 1796304934
 1887530885 1839295471 1707240971 1873242695         33         16
         34         37         34         85         20          2
         47         23          5         72         16         23
         54         18         49         95         78         91
         47         93         10         69         36         32
         69         64         77         93         21         28
 1753837732 1488743807 1711632378 1799581711 1818783215 1891790695
 1837723802 1752132873 1727950918 1760390205 1866887130 1824876865
 1807839436 1890544910 1889755550 1787129270 1502702106 1841065201
 1820156583 1779396288 1760521448 1844604520 1831039103 1843491014
 1891199259 1839493317 1801349958 1577807434 1811377215 1879404734
 1826057367 1837853054         37         63          2         49
         10         63         21         44         70         28
         72         53         12         81         64         54
         86         26          3         27         85         30
         10         82         98         76         98         72
         20         27         89         75 1890086927 1826517134
 1814783944 1824156809 1875733079 1842114682 1845456975 1830120794
 1787980861 1807380585 1535469972 1883860884 1889167601 1747872128
 1888317235 1720937006 1836806331 1654152236 1695309475 1892773593
 1840737395 1868392748 1833724316 1600153936 1869310159 1883467778
 1892641857 1776248953 1833201514 1886743848 1745972258 1860657622
         95         86          5          3         93         50
         51          6         20         88         21         89
         88         64         69         79         46         11
         36         72         38         40         44         12
         65         36         75         60         29         30
         24         37 1733652589 1756325317 1685744372 1780772214
 1660252348 1629973784 1847815925 1828941229 1683778661 1519480967
 1762160488 1844801381 1832742021 1891724641 1761701480 1695312651
 1429433841 1774275423 1828349211 1779786303 1835953259 1784896595
 1858432988 1413442268 1893363867 1886679050 1872588913 1473866635
 1793158916 1762946052 1719627087 1893231666         76         26
         35         84         33         58         10          7
         69         36         97         57         97         24
         79         95         13         57         25         26
         91         51         79          4         22         54
         27         44          8         14         87         87
 1208960410 1208567888 1215973275 1214859418 1210992732 1208305714
 1165379704 1215252623 1197033625 1212172318 1217415148 1211058280
 1206798479 1215645776 1209223143 1217611809 1212500082 1215383615
 1208567769 1200179260 1216694386 1218070398 1195526187 1211516213
 1213089850 1213941743 1216825148 1212565573 1216694248 1217546087
 1200178778 1215973524         92         80         54         28
         77         98         45         31         75         77
         40         42         38        100         19         17
         93         80         41         68         24         89
         87         25         60         29         83         47
         61         35         80         12 1189759014 1218070662
 1211057953 1216825400 1215121414 1214596952 1216169420 1210075102
 1209157633 1213941279 1195984973 1211648118 1201686666 1212041360
 1216103688 1212500024 1173112887 1194608648 1216825427 1209747582
 1207191587 1214859224 1203980407 1215711379 1213155353 1203259396
 1214859350 1211779156 1217218713 1202473003 1216628529 1196771437
         44         54         12         89          7         67
         45          8         25         75         36         80
         42         38         61         58         97         85
         40         23         73         59         45         91
         47         72         17         27         22         11
         91         42 1211516990 1198213215 1203848735 1217349746
 1212303398 1217808150 1215514752 1209878647 1214138433 1215711277
 1212041273 1215383541 1214728294 1197754500 1169574019 1208371214
 1214269569 1216301092 1216563283 1213548677 1217873970 1203128459
 1209812695 1218136029 1194805359 1204439186 1218005120 1213941626
 1217153046 1208109091 1215055928 1215318166          5         13
         71         52         25         84         10         13
         58         86         56         89         33         63
          2         48         36         49         87         33
          2         36         50         50         25         95
        100         45          4         84         24         77
 1202210969 1213679767 1209288847 1217480263 1184319410 1214072674
 1188382819 1217283870 1200769107 1217939416 1199327294 1213351841
 1206667407 1217153163 1215580283 1214138474 1206798167 1194477696
 1193690499 1214072706 1216825421 1216693888 1217611496 1198540949
 1199654414 1206405188 1214203847 1165183076 1213745302 1208830102
 1209944118 1215121459]

免 责 声 明:本 站 内 容 由 asc-devkit 仓 master 分 支 自 动 编 译 生 成,属 于 持 续 开 发 版 本,可 能 存 在 缺 陷,仅 供 预 览 与 参 考。如 需 稳 定 及 商 用 资 料,请 查 阅 官 方 昇 腾 社 区