set_dl_model_layer_paramT_set_dl_model_layer_paramSetDlModelLayerParamSetDlModelLayerParamset_dl_model_layer_param (算子)
名称
set_dl_model_layer_paramT_set_dl_model_layer_paramSetDlModelLayerParamSetDlModelLayerParamset_dl_model_layer_param — 设置给定层的参数值。
签名
void SetDlModelLayerParam(const HTuple& DLModelHandle, const HTuple& LayerName, const HTuple& ParamName, const HTuple& ParamValue)
static void HDlLayer::SetDlModelLayerParam(const HDlModel& DLModelHandle, const HString& LayerName, const HString& ParamName, const HTuple& ParamValue)
static void HDlLayer::SetDlModelLayerParam(const HDlModel& DLModelHandle, const HString& LayerName, const HString& ParamName, const HString& ParamValue)
static void HDlLayer::SetDlModelLayerParam(const HDlModel& DLModelHandle, const char* LayerName, const char* ParamName, const char* ParamValue)
static void HDlLayer::SetDlModelLayerParam(const HDlModel& DLModelHandle, const wchar_t* LayerName, const wchar_t* ParamName, const wchar_t* ParamValue)
(
Windows only)
描述
The operator set_dl_model_layer_paramset_dl_model_layer_paramSetDlModelLayerParamSetDlModelLayerParamSetDlModelLayerParamset_dl_model_layer_param sets the value
ParamValueParamValueParamValueParamValueparamValueparam_value of the parameter ParamNameParamNameParamNameParamNameparamNameparam_name for a layer.
The layer is referred to by its name LayerNameLayerNameLayerNameLayerNamelayerNamelayer_name in the model
DLModelHandleDLModelHandleDLModelHandleDLModelHandleDLModelHandledlmodel_handle. You can retrieve the layer names using
get_dl_model_paramget_dl_model_paramGetDlModelParamGetDlModelParamGetDlModelParamget_dl_model_param with its option 'layer_names'"layer_names""layer_names""layer_names""layer_names""layer_names" or
'summary'"summary""summary""summary""summary""summary".
Which generic and layer-specific parameters can be set is described
in the entries of the operators (create_dl_layer_*), which
are used for creating the layer.
执行信息
- 多线程类型：可重入（与非独占算子并行运行）。
- 多线程作用域：全局（可从任何线程调用）。
- 未采用并行化处理。
参数
DLModelHandleDLModelHandleDLModelHandleDLModelHandleDLModelHandledlmodel_handle (输入控制) dl_model → HDlModel, HTupleHHandleHTupleHtuple (handle) (IntPtr) (HHandle) (handle)
Deep learning model.
LayerNameLayerNameLayerNameLayerNamelayerNamelayer_name (输入控制) string → HTuplestrHTupleHtuple (string) (string) (HString) (char*)
Name of the output layer.
ParamNameParamNameParamNameParamNameparamNameparam_name (输入控制) attribute.name → HTuplestrHTupleHtuple (string) (string) (HString) (char*)
Name of the set parameter.
默认值：
[]
值列表：
'bias_filler'"bias_filler""bias_filler""bias_filler""bias_filler""bias_filler", 'bias_filler_const_val'"bias_filler_const_val""bias_filler_const_val""bias_filler_const_val""bias_filler_const_val""bias_filler_const_val", 'bias_filler_variance_norm'"bias_filler_variance_norm""bias_filler_variance_norm""bias_filler_variance_norm""bias_filler_variance_norm""bias_filler_variance_norm", 'is_inference_output'"is_inference_output""is_inference_output""is_inference_output""is_inference_output""is_inference_output", 'leaky_relu_alpha'"leaky_relu_alpha""leaky_relu_alpha""leaky_relu_alpha""leaky_relu_alpha""leaky_relu_alpha", 'learning_rate_multiplier'"learning_rate_multiplier""learning_rate_multiplier""learning_rate_multiplier""learning_rate_multiplier""learning_rate_multiplier", 'learning_rate_multiplier_bias'"learning_rate_multiplier_bias""learning_rate_multiplier_bias""learning_rate_multiplier_bias""learning_rate_multiplier_bias""learning_rate_multiplier_bias", 'name'"name""name""name""name""name", 'upper_bound'"upper_bound""upper_bound""upper_bound""upper_bound""upper_bound", 'weight_filler'"weight_filler""weight_filler""weight_filler""weight_filler""weight_filler", 'weight_filler_const_val'"weight_filler_const_val""weight_filler_const_val""weight_filler_const_val""weight_filler_const_val""weight_filler_const_val", 'weight_filler_variance_norm'"weight_filler_variance_norm""weight_filler_variance_norm""weight_filler_variance_norm""weight_filler_variance_norm""weight_filler_variance_norm"
ParamValueParamValueParamValueParamValueparamValueparam_value (输入控制) attribute.value(-array) → HTupleMaybeSequence[Union[int, float, str]]HTupleHtuple (string / integer / real) (string / int / long / double) (HString / Hlong / double) (char* / Hlong / double)
Value of the set parameter.
默认值：
[]
值列表：
'const'"const""const""const""const""const", 'false'"false""false""false""false""false", 'msra'"msra""msra""msra""msra""msra", 'norm_average'"norm_average""norm_average""norm_average""norm_average""norm_average", 'norm_in'"norm_in""norm_in""norm_in""norm_in""norm_in", 'norm_out'"norm_out""norm_out""norm_out""norm_out""norm_out", 'true'"true""true""true""true""true", 'xavier'"xavier""xavier""xavier""xavier""xavier"
另见
set_dl_model_paramset_dl_model_paramSetDlModelParamSetDlModelParamSetDlModelParamset_dl_model_param,
get_dl_model_paramget_dl_model_paramGetDlModelParamGetDlModelParamGetDlModelParamget_dl_model_param,
get_dl_model_layer_paramget_dl_model_layer_paramGetDlModelLayerParamGetDlModelLayerParamGetDlModelLayerParamget_dl_model_layer_param
模块
基础。此算子采用动态许可机制（详见《安装指南》）。所需模块取决于算子的具体使用场景：
三维计量学、光学字符识别/光学字符验证、深度学习训练