Conozca lo último en aprendizaje automático, IA generativa y más en el
Simposio WiML 2023.
Organiza tus páginas con colecciones
Guarda y categoriza el contenido según tus preferencias.
tensorflow :: operaciones :: ResourceSparseApplyFtrlV2
#include <training_ops.h>
Actualice las entradas relevantes en '* var' de acuerdo con el esquema Ftrl-proximal.
Resumen
Es decir, para las filas para las que hemos graduado, actualizamos var, acum y linear de la siguiente manera: grad_with_shrinkage = grad + 2 * l2_shrinkage * var acum_new = acum + grad_with_shrinkage * grad_with_shrinkage linear + = grad_with_shrinkage + (acum_new ^ (- lr_power) - acum ^ (-lr_potencia)) / lr * var cuadrática = 1.0 / (acumula_nuevo ^ (lr_potencia) * lr) + 2 * l2 var = (signo (lineal) * l1 - lineal) / cuadrático si | lineal | > l1 más 0.0 acum = acum_nuevo
Argumentos:
- alcance: un objeto de alcance
- var: debe ser de una variable ().
- acum: debe ser de una variable ().
- lineal: debe ser de una variable ().
- grad: El gradiente.
- índices: Un vector de índices en la primera dimensión de var y acum.
- lr: factor de escala. Debe ser un escalar.
- l1: Regularización L1. Debe ser un escalar.
- l2: Regularización de la contracción L2. Debe ser un escalar.
- lr_power: factor de escala. Debe ser un escalar.
Atributos opcionales (consulte Attrs
):
- use_locking: Si es
True
, la actualización de los tensores var y acumuladores estará protegida por un bloqueo; de lo contrario, el comportamiento no está definido, pero puede mostrar menos contención.
Devoluciones:
Constructores y Destructores |
---|
ResourceSparseApplyFtrlV2 (const :: tensorflow::Scope & scope, :: tensorflow::Input var, :: tensorflow::Input accum, :: tensorflow::Input linear, :: tensorflow::Input grad, :: tensorflow::Input indices, :: tensorflow::Input lr, :: tensorflow::Input l1, :: tensorflow::Input l2, :: tensorflow::Input l2_shrinkage, :: tensorflow::Input lr_power)
|
ResourceSparseApplyFtrlV2 (const :: tensorflow::Scope & scope, :: tensorflow::Input var, :: tensorflow::Input accum, :: tensorflow::Input linear, :: tensorflow::Input grad, :: tensorflow::Input indices, :: tensorflow::Input lr, :: tensorflow::Input l1, :: tensorflow::Input l2, :: tensorflow::Input l2_shrinkage, :: tensorflow::Input lr_power, const ResourceSparseApplyFtrlV2::Attrs & attrs) |
Atributos públicos
Funciones publicas
ResourceSparseApplyFtrlV2
ResourceSparseApplyFtrlV2(
const ::tensorflow::Scope & scope,
::tensorflow::Input var,
::tensorflow::Input accum,
::tensorflow::Input linear,
::tensorflow::Input grad,
::tensorflow::Input indices,
::tensorflow::Input lr,
::tensorflow::Input l1,
::tensorflow::Input l2,
::tensorflow::Input l2_shrinkage,
::tensorflow::Input lr_power
)
ResourceSparseApplyFtrlV2
ResourceSparseApplyFtrlV2(
const ::tensorflow::Scope & scope,
::tensorflow::Input var,
::tensorflow::Input accum,
::tensorflow::Input linear,
::tensorflow::Input grad,
::tensorflow::Input indices,
::tensorflow::Input lr,
::tensorflow::Input l1,
::tensorflow::Input l2,
::tensorflow::Input l2_shrinkage,
::tensorflow::Input lr_power,
const ResourceSparseApplyFtrlV2::Attrs & attrs
)
operador :: tensorflow :: Operación
operator::tensorflow::Operation() const
Funciones estáticas públicas
UseLocking
Attrs UseLocking(
bool x
)
Salvo que se indique lo contrario, el contenido de esta página está sujeto a la licencia Atribución 4.0 de Creative Commons, y los ejemplos de código están sujetos a la licencia Apache 2.0. Para obtener más información, consulta las políticas del sitio de Google Developers. Java es una marca registrada de Oracle o sus afiliados.
Última actualización: 2020-04-20 (UTC)
[null,null,["Última actualización: 2020-04-20 (UTC)"],[],[],null,["# tensorflow::ops::ResourceSparseApplyFtrlV2 Class Reference\n\ntensorflow::ops::ResourceSparseApplyFtrlV2\n==========================================\n\n`#include \u003ctraining_ops.h\u003e`\n\nUpdate relevant entries in '\\*var' according to the Ftrl-proximal scheme.\n\nSummary\n-------\n\nThat is for rows we have grad for, we update var, accum and linear as follows: grad_with_shrinkage = grad + 2 \\* l2_shrinkage \\* var accum_new = accum + grad_with_shrinkage \\* grad_with_shrinkage linear += grad_with_shrinkage + (accum_new\\^(-lr_power) - accum\\^(-lr_power)) / lr \\* var quadratic = 1.0 / (accum_new\\^(lr_power) \\* lr) + 2 \\* l2 var = (sign(linear) \\* l1 - linear) / quadratic if \\|linear\\| \\\u003e l1 else 0.0 accum = accum_new\n\nArguments:\n\n- scope: A [Scope](/versions/r1.15/api_docs/cc/class/tensorflow/scope#classtensorflow_1_1_scope) object\n- var: Should be from a Variable().\n- accum: Should be from a Variable().\n- linear: Should be from a Variable().\n- grad: The gradient.\n- indices: A vector of indices into the first dimension of var and accum.\n- lr: Scaling factor. Must be a scalar.\n- l1: L1 regularization. Must be a scalar.\n- l2: L2 shrinkage regulariation. Must be a scalar.\n- lr_power: Scaling factor. Must be a scalar.\n\n\u003cbr /\u003e\n\nOptional attributes (see [Attrs](/versions/r1.15/api_docs/cc/struct/tensorflow/ops/resource-sparse-apply-ftrl-v2/attrs#structtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1_1_attrs)):\n\n- use_locking: If `True`, updating of the var and accum tensors will be protected by a lock; otherwise the behavior is undefined, but may exhibit less contention.\n\n\u003cbr /\u003e\n\nReturns:\n\n- the created [Operation](/versions/r1.15/api_docs/cc/class/tensorflow/operation#classtensorflow_1_1_operation)\n\n\u003cbr /\u003e\n\n| ### Constructors and Destructors ||\n|---|---|\n| [ResourceSparseApplyFtrlV2](#classtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1a77306a45450ada78aafa759f7b197723)`(const ::`[tensorflow::Scope](/versions/r1.15/api_docs/cc/class/tensorflow/scope#classtensorflow_1_1_scope)` & scope, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` var, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` accum, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` linear, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` grad, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` indices, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` lr, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` l1, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` l2, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` l2_shrinkage, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` lr_power)` ||\n| [ResourceSparseApplyFtrlV2](#classtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1a76e64a98f17ed91ba2513c27b045eb40)`(const ::`[tensorflow::Scope](/versions/r1.15/api_docs/cc/class/tensorflow/scope#classtensorflow_1_1_scope)` & scope, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` var, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` accum, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` linear, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` grad, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` indices, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` lr, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` l1, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` l2, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` l2_shrinkage, ::`[tensorflow::Input](/versions/r1.15/api_docs/cc/class/tensorflow/input#classtensorflow_1_1_input)` lr_power, const `[ResourceSparseApplyFtrlV2::Attrs](/versions/r1.15/api_docs/cc/struct/tensorflow/ops/resource-sparse-apply-ftrl-v2/attrs#structtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1_1_attrs)` & attrs)` ||\n\n| ### Public attributes ||\n|-----------------------------------------------------------------------------------------------------------|---------------------------------------------------------------------------------------------------|\n| [operation](#classtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1a392e0892ef99d41bc86397a57df9a98f) | [Operation](/versions/r1.15/api_docs/cc/class/tensorflow/operation#classtensorflow_1_1_operation) |\n\n| ### Public functions ||\n|--------------------------------------------------------------------------------------------------------------------------------------------|---------|\n| [operator::tensorflow::Operation](#classtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1afba48dacb7460dfeb6303226504eec7e)`() const ` | ` ` ` ` |\n\n| ### Public static functions ||\n|----------------------------------------------------------------------------------------------------------------------|--------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| [UseLocking](#classtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1a5a50996a22963c9d267ae9f2b76fa63c)`(bool x)` | [Attrs](/versions/r1.15/api_docs/cc/struct/tensorflow/ops/resource-sparse-apply-ftrl-v2/attrs#structtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2_1_1_attrs) |\n\n| ### Structs ||\n|--------------------------------------------------------------------------------------------------------------------------------------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|\n| [tensorflow::ops::ResourceSparseApplyFtrlV2::Attrs](/versions/r1.15/api_docs/cc/struct/tensorflow/ops/resource-sparse-apply-ftrl-v2/attrs) | Optional attribute setters for [ResourceSparseApplyFtrlV2](/versions/r1.15/api_docs/cc/class/tensorflow/ops/resource-sparse-apply-ftrl-v2#classtensorflow_1_1ops_1_1_resource_sparse_apply_ftrl_v2). |\n\nPublic attributes\n-----------------\n\n### operation\n\n```text\nOperation operation\n``` \n\nPublic functions\n----------------\n\n### ResourceSparseApplyFtrlV2\n\n```gdscript\n ResourceSparseApplyFtrlV2(\n const ::tensorflow::Scope & scope,\n ::tensorflow::Input var,\n ::tensorflow::Input accum,\n ::tensorflow::Input linear,\n ::tensorflow::Input grad,\n ::tensorflow::Input indices,\n ::tensorflow::Input lr,\n ::tensorflow::Input l1,\n ::tensorflow::Input l2,\n ::tensorflow::Input l2_shrinkage,\n ::tensorflow::Input lr_power\n)\n``` \n\n### ResourceSparseApplyFtrlV2\n\n```gdscript\n ResourceSparseApplyFtrlV2(\n const ::tensorflow::Scope & scope,\n ::tensorflow::Input var,\n ::tensorflow::Input accum,\n ::tensorflow::Input linear,\n ::tensorflow::Input grad,\n ::tensorflow::Input indices,\n ::tensorflow::Input lr,\n ::tensorflow::Input l1,\n ::tensorflow::Input l2,\n ::tensorflow::Input l2_shrinkage,\n ::tensorflow::Input lr_power,\n const ResourceSparseApplyFtrlV2::Attrs & attrs\n)\n``` \n\n### operator::tensorflow::Operation\n\n```gdscript\n operator::tensorflow::Operation() const \n``` \n\nPublic static functions\n-----------------------\n\n### UseLocking\n\n```text\nAttrs UseLocking(\n bool x\n)\n```"]]