Home
last modified time | relevance | path

Searched refs:float32 (Results 1 – 22 of 22) sorted by relevance

/ohos5.0/docs/zh-cn/application-dev/ai/mindspore/
H A Dmindspore-lite-converter-guidelines.md89 | --fp16 | 否 | 设定在模型序列化时是否需要将float32数据格式的权重存储为float16数据格式。<br/>默认值为off…
139 input_dtypes=float32;float32
143 output_dtypes=float32
H A Dmindspore-guidelines-based-js.md119 // 处理readBuffer，转换成float32格式，并进行标准化处理
H A Dmindspore-guidelines-based-native.md103 // 处理readBuffer，转换成float32格式，并进行标准化处理
/ohos5.0/docs/zh-cn/device-dev/reference/hdi-apis/nnrt/
H A D_model_config_v10.md20 | boolean [enableFloat16](#enablefloat16) | float32浮点模型是否以float16浮点运行 |
35 float32浮点模型是否以float16浮点运行
H A D_model_config_v20.md20 | boolean [enableFloat16](#enablefloat16) | float32浮点模型是否以float16浮点运行 |
36 float32浮点模型是否以float16浮点运行
/ohos5.0/docs/en/application-dev/ai/mindspore/
H A Dmindspore-lite-converter-guidelines.md90 | --fp16 | No | Whether to store the weights of float32 data as float16…
140 input_dtypes=float32;float32
144 output_dtypes=float32
H A Dmindspore-guidelines-based-js.md119 // Convert readBuffer to the float32 format, and standardize the image.
H A Dmindspore-guidelines-based-native.md103 // Convert readBuffer to the float32 format, and standardize the image.
/ohos5.0/foundation/graphic/graphic_3d/lume/LumeRender/api/render/device/
H A Dpipeline_state_desc.h655 constexpr ClearColorValue(float r, float g, float b, float a) : float32 { r, g, b, a } {}; in ClearColorValue()
659 float float32[4]; member
/ohos5.0/foundation/ai/neural_network_runtime/
H A Dneural-network-runtime-guidelines.md291 // 添加Add算子的第一个输入张量，类型为float32，张量形状为[1, 2, 2, 3]
311 // 添加Add算子的第二个输入张量，类型为float32，张量形状为[1, 2, 2, 3]
355 // 设置Add算子的输出张量，类型为float32，张量形状为[1, 2, 2, 3]
/ohos5.0/docs/zh-cn/application-dev/ai/nnrt/
H A Dneural-network-runtime-guidelines.md282 // 添加Add算子的第一个输入张量，类型为float32，张量形状为[1, 2, 2, 3]
302 // 添加Add算子的第二个输入张量，类型为float32，张量形状为[1, 2, 2, 3]
346 // 设置Add算子的输出张量，类型为float32，张量形状为[1, 2, 2, 3]
/ohos5.0/docs/en/application-dev/ai/nnrt/
H A Dneural-network-runtime-guidelines.md282 …// Add the first input tensor of the float32 type for the Add operator. The tensor shape is [1, 2,…
302 …// Add the second input tensor of the float32 type for the Add operator. The tensor shape is [1, 2…
346 …// Add the output tensor of the float32 type for the Add operator. The tensor shape is [1, 2, 2, 3…
/ohos5.0/foundation/graphic/graphic_3d/lume/LumeRender/src/nodecontext/
H A Drender_node_parser_util.cpp241 FromJson(*pos, context.data.clearValue.color.float32); in FromJson()
/ohos5.0/docs/zh-cn/application-dev/reference/apis-neural-network-runtime-kit/
H A D_neural_network_runtime.md499 | OH_NN_FLOAT32 | 张量数据类型为float32 |
1147 浮点模型默认使用float32精度计算。如果在支持float16精度的硬件上调用该接口，float32浮点数精度的模型将以float16的精度执行计算， 可减少内存占用和执行时间。
1160 | enableFloat16 | Float16低精度计算标志位。设置为true时，执行Float16推理；设置为false时，执行float32推理。 |
H A D_neural_nework_runtime.md499 | OH_NN_FLOAT32 | 张量数据类型为float32 |
1147 浮点模型默认使用float32精度计算。如果在支持float16精度的硬件上调用该接口，float32浮点数精度的模型将以float16的精度执行计算， 可减少内存占用和执行时间。
1160 | enableFloat16 | Float16低精度计算标志位。设置为true时，执行Float16推理；设置为false时，执行float32推理。 |
/ohos5.0/docs/zh-cn/application-dev/reference/apis-mindspore-lite-kit/
H A Djs-apis-mindSporeLite.md682 | O2 | 2 | 将网络转换为float16, 保持批量归一化层和损失函数为float32。 |
H A D_mind_spore.md599 | OH_AI_KO2 | 将网络转换为float16, 保持批量归一化层和损失函数为float32。 |
/ohos5.0/docs/en/application-dev/reference/apis-mindspore-lite-kit/
H A Djs-apis-mindSporeLite.md682 …loat16 and keeps the precision type of the batch normalization layer and loss function as float32.|
H A D_mind_spore.md599 …loat16 and keeps the precision type of the batch normalization layer and loss function as float32.|
/ohos5.0/foundation/graphic/graphic_3d/lume/LumeRender/src/gles/
H A Drender_backend_gles.cpp1276 glClearBufferfv(GL_COLOR, static_cast<GLint>(idx), ref.clearValue.color.float32); in HandleColorAttachments()
/ohos5.0/docs/en/application-dev/reference/apis-neural-network-runtime-kit/
H A D_neural_network_runtime.md499 | OH_NN_FLOAT32 | float32 type.|
1147 …int model uses float32 for computing. If this API is called on a device that supports float16, flo…
1160 …, float16 inference is performed. If this parameter is set to **false**, float32 inference is perf…
/ohos5.0/docs/zh-cn/application-dev/reference/common/
H A D_j_s_v_m.md739 | JSVM_FLOAT32_ARRAY | float32类型。 |