万万没想到
chang zherui
5 years ago
1 changed files with 9 additions and 9 deletions
Unified View
Diff Options
+ 9
- 9
mindspore/ops/operations/nn_ops.py
View File
| @@ -1219,14 +1219,14 @@ class ApplyMomentum(PrimitiveWithInfer): | |||||
| gradient_scale (float): The scale of the gradient. Default: 1.0. | gradient_scale (float): The scale of the gradient. Default: 1.0. | ||||
| Inputs: | Inputs: | ||||
| - **variable** (Tensor) - Weights to be update. | |||||
| - **variable** (Tensor) - Weights to be updated. | |||||
| - **accumulation** (Tensor) - Accumulated gradient value by moment weight. | - **accumulation** (Tensor) - Accumulated gradient value by moment weight. | ||||
| - **learning_rate** (float) - Learning rate. | - **learning_rate** (float) - Learning rate. | ||||
| - **gradient** (Tensor) - Gradients. | - **gradient** (Tensor) - Gradients. | ||||
| - **momentum** (float) - Momentum. | - **momentum** (float) - Momentum. | ||||
| Outputs: | Outputs: | ||||
| Tensor, parameters to be update. | |||||
| Tensor, parameters to be updated. | |||||
| Examples: | Examples: | ||||
| >>> net = ResNet50() | >>> net = ResNet50() | ||||
| @@ -1318,15 +1318,15 @@ class SGD(PrimitiveWithInfer): | |||||
| nesterov (bool): Enable Nesterov momentum. Default: False. | nesterov (bool): Enable Nesterov momentum. Default: False. | ||||
| Inputs: | Inputs: | ||||
| - **parameters** (Tensor) - Parameters to be update. | |||||
| - **parameters** (Tensor) - Parameters to be updated. | |||||
| - **gradient** (Tensor) - Gradients. | - **gradient** (Tensor) - Gradients. | ||||
| - **learning_rate** (Tensor) - Learning rate. e.g. Tensor(0.1, mindspore.float32). | - **learning_rate** (Tensor) - Learning rate. e.g. Tensor(0.1, mindspore.float32). | ||||
| - **accum** (Tensor) - Accum(velocity) to be update. | |||||
| - **accum** (Tensor) - Accum(velocity) to be updated. | |||||
| - **momentum** (Tensor) - Momentum. e.g. Tensor(0.1, mindspore.float32). | - **momentum** (Tensor) - Momentum. e.g. Tensor(0.1, mindspore.float32). | ||||
| - **stat** (Tensor) - States to be updated with the same shape as gradient. | - **stat** (Tensor) - States to be updated with the same shape as gradient. | ||||
| Outputs: | Outputs: | ||||
| Tensor, parameters to be update. | |||||
| Tensor, parameters to be updated. | |||||
| """ | """ | ||||
| @prim_attr_register | @prim_attr_register | ||||
| @@ -2141,7 +2141,7 @@ class Adam(PrimitiveWithInfer): | |||||
| If False, updates the gradients without using NAG. Default: False. | If False, updates the gradients without using NAG. Default: False. | ||||
| Inputs: | Inputs: | ||||
| - **var** (Tensor) - Weights to be update. | |||||
| - **var** (Tensor) - Weights to be updated. | |||||
| - **m** (Tensor) - The 1st moment vector in the updating formula. | - **m** (Tensor) - The 1st moment vector in the updating formula. | ||||
| - **v** (Tensor) - the 2nd moment vector in the updating formula. | - **v** (Tensor) - the 2nd moment vector in the updating formula. | ||||
| - **beta1_power** (float) - :math:`beta_1^t` in the updating formula. | - **beta1_power** (float) - :math:`beta_1^t` in the updating formula. | ||||
| @@ -2251,8 +2251,8 @@ class SparseApplyAdagrad(PrimitiveWithInfer): | |||||
| use_locking (bool): If True, updating of the var and accum tensors will be protected. Default: False. | use_locking (bool): If True, updating of the var and accum tensors will be protected. Default: False. | ||||
| Inputs: | Inputs: | ||||
| - **var** (Tensor) - Variable to be update. The type must be float32. | |||||
| - **accum** (Tensor) - Accum to be update. The shape must be the same as `var`'s shape, | |||||
| - **var** (Tensor) - Variable to be updated. The type must be float32. | |||||
| - **accum** (Tensor) - Accum to be updated. The shape must be the same as `var`'s shape, | |||||
| the type must be float32. | the type must be float32. | ||||
| - **grad** (Tensor) - Gradient. The shape must be the same as `var`'s shape | - **grad** (Tensor) - Gradient. The shape must be the same as `var`'s shape | ||||
| except first dimension, the type must be float32. | except first dimension, the type must be float32. | ||||
| @@ -2372,7 +2372,7 @@ class LARSUpdate(PrimitiveWithInfer): | |||||
| use_clip (bool): Whether to use clip operation for calculating the local learning rate. Default: False. | use_clip (bool): Whether to use clip operation for calculating the local learning rate. Default: False. | ||||
| Inputs: | Inputs: | ||||
| - **weight** (Tensor) - The weight to be update. | |||||
| - **weight** (Tensor) - The weight to be updated. | |||||
| - **gradient** (Tensor) - The gradient of weight, which has the same shape and dtype with weight. | - **gradient** (Tensor) - The gradient of weight, which has the same shape and dtype with weight. | ||||
| - **norm_weight** (Tensor) - A scalar tensor, representing the square sum of weight. | - **norm_weight** (Tensor) - A scalar tensor, representing the square sum of weight. | ||||
| - **norm_gradient** (Tensor) - A scalar tensor, representing the square sum of gradient. | - **norm_gradient** (Tensor) - A scalar tensor, representing the square sum of gradient. | ||||