# Copyright (c) 2018 PaddlePaddle Authors. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. from __future__ import print_function import copy import six import warnings import functools from . import layers from . import framework from . import core from . import name_scope from .dygraph import base as imperative_base __all__ = [ 'set_gradient_clip', 'ErrorClipByValue', 'GradientClipByValue', 'GradientClipByNorm', 'GradientClipByGlobalNorm' ] class BaseErrorClipAttr(object): def __str__(self): raise NotImplementedError() def _append_clip_op(self, block, grad_name): raise NotImplementedError() class ErrorClipByValue(BaseErrorClipAttr): """ Clips tensor values to the range [min, max]. Given a tensor ``t`` (see Examples below), this operation clips its value \ to ``min`` and ``max`` inplace. - Any values less than min are set to min. - Any values greater than max are set to max. Args: max (float): The maximum value to clip by. min (float, optional): The minimum value to clip by. if not set by user, \ will be set to ``-max`` by framework. Examples: .. code-block:: python import paddle.fluid as fluid BATCH_SIZE = 128 CLIP_MAX = 2e-6 CLIP_MIN = -1e-6 prog = fluid.framework.Program() with fluid.program_guard(main_program=prog): image = fluid.layers.data( name='x', shape=[784], dtype='float32') hidden1 = fluid.layers.fc(input=image, size=128, act='relu') hidden2 = fluid.layers.fc(input=hidden1, size=64, act='relu') predict = fluid.layers.fc( input=hidden2, size=10, act='softmax') label = fluid.layers.data(name='y', shape=[1], dtype='int64') cost = fluid.layers.cross_entropy(input=predict, label=label) avg_cost = fluid.layers.mean(cost) prog_clip = prog.clone() prog_clip.block(0).var(hidden1.name)._set_error_clip( fluid.clip.ErrorClipByValue( max=CLIP_MAX, min=CLIP_MIN) """ def __init__(self, max, min=None): max = float(max) if min is None: min = -max else: min = float(min) self.max = max self.min = min def __str__(self): return "ByValue, min=%f, max=%f" % (self.min, self.max) def _append_clip_op(self, block, grad_name): clip_op_desc = block.desc.append_op() clip_op_desc.set_type("clip") clip_op_desc.set_input("X", [grad_name]) clip_op_desc.set_output("Out", [grad_name]) clip_op_desc._set_attr("min", self.min) clip_op_desc._set_attr("max", self.max) def error_clip_callback(block, context): # the context is a grad_to_var map grad_to_var = context op_desc = block.desc.op(block.desc.op_size() - 1) for grad_n in [n for n in op_desc.output_arg_names() if n in grad_to_var]: fwd_var = block._var_recursive(grad_to_var[grad_n]) error_clip = getattr(fwd_var, "error_clip", None) if not (error_clip is None or isinstance(error_clip, BaseErrorClipAttr)): raise TypeError( "Variable's error_clip should be an instance of BaseErrorClipAttr or None." ) if error_clip is not None: error_clip._append_clip_op(block, grad_n) class GradientClipBase(object): def __init__(self, need_clip=None): if need_clip is not None and not callable(need_clip): raise TypeError( "The type of need_clip must be funciton, and it can filter out " "parameter that does't need gradient clip. This function must return " "True or False, and True means that clipping is required. Please refer to " "API documention of GradientClipByGlobalNorm / GradientClipByNorm " "/GradientClipByValue.") self._need_clip_func = need_clip def __str__(self): raise NotImplementedError() @imperative_base.no_grad def _dygraph_clip(self, params_grads): raise NotImplementedError def _static_clip(self, params_grads): raise NotImplementedError def __call__(self, params_grads): assert len( params_grads ) > 0, "The number of trainable parameters should be greater than 0." if framework.in_dygraph_mode(): return self._dygraph_clip(params_grads) else: for p, g in params_grads: if getattr(p, 'gradient_clip_attr', None) is not None: warnings.warn( "'set_gradient_clip' will be ineffective, because you have " "pass 'grad_clip' into 'minimize'. So, 'set_gradient_clip' " "is redundant and you can remove it.") break return self._static_clip(params_grads) def _process_context(self, context, param, grad): raise NotImplementedError() def _create_operators(self, param, grad): raise NotImplementedError() class GradientClipByValue(GradientClipBase): """ Limit the value of multi-dimensional Tensor :math:`X` to the range [min, max]. - Any values less than min are set to ``min``. - Any values greater than max are set to ``max``. The multi-dimensional Tensor :math:`X` is not passed from this class, but the gradients of all parameters in ``Program`` . If ``need_clip`` is not None, then only part of gradients can be selected for gradient clipping. Gradient clip will takes effect after being set in ``optimizer.minimize(grad_clip)`` , see the document ``optimizer`` (for example: :ref:`api_fluid_optimizer_SGDOptimizer`). Args: max (float): The maximum value to clip by. min (float, optional): The minimum value to clip by. if not set by user, it will be set to ``-max`` automatically. In this case, ``max`` must be greater than 0. need_clip (function, optional): Type: function. This function accepts a ``Parameter`` and returns ``bool`` (True: the gradient of this ``Parameter`` need to be clipped, False: not need). Default: None, and gradients of all parameters in the network will be clipped. Examples: .. code-block:: python # use for Static mode import paddle import paddle.fluid as fluid import numpy as np main_prog = fluid.Program() startup_prog = fluid.Program() with fluid.program_guard( main_program=main_prog, startup_program=startup_prog): image = fluid.data( name='x', shape=[-1, 2], dtype='float32') predict = fluid.layers.fc(input=image, size=3, act='relu') # Trainable parameters: fc_0.w.0, fc_0.b.0 loss = fluid.layers.mean(predict) # Clip all parameters in network: clip = fluid.clip.GradientClipByValue(min=-1, max=1) # Clip a part of parameters in network: (e.g. fc_0.w_0) # pass a function(fileter_func) to need_clip, and fileter_func receive a Parameter, and return bool # def fileter_func(Parameter): # # It can be easily filtered by Parameter.name (name can be set in fluid.ParamAttr, and the default name is fc_0.w_0, fc_0.b_0) # return Parameter.name=="fc_0.w_0" # clip = fluid.clip.GradientClipByValue(min=-1, max=1, need_clip=fileter_func) sgd_optimizer = fluid.optimizer.SGDOptimizer(learning_rate=0.1) sgd_optimizer.minimize(loss, grad_clip=clip) place = fluid.CPUPlace() exe = fluid.Executor(place) x = np.random.uniform(-100, 100, (10, 2)).astype('float32') exe.run(startup_prog) out = exe.run(main_prog, feed={'x': x}, fetch_list=loss) # use for Dygraph mode import paddle import paddle.fluid as fluid with fluid.dygraph.guard(): linear = fluid.dygraph.Linear(10, 10) # Trainable parameters:: linear_0.w.0, linear_0.b.0 inputs = fluid.layers.uniform_random([32, 10]).astype('float32') out = linear(fluid.dygraph.to_variable(inputs)) loss = fluid.layers.reduce_mean(out) loss.backward() # Clip all parameters in network: clip = fluid.clip.GradientClipByValue(min=-1, max=1) # Clip a part of parameters in network: (e.g. linear_0.w_0) # pass a function(fileter_func) to need_clip, and fileter_func receive a ParamBase, and return bool # def fileter_func(ParamBase): # # It can be easily filtered by ParamBase.name(name can be set in fluid.ParamAttr, and the default name is linear_0.w_0, linear_0.b_0) # return ParamBase.name == "linear_0.w_0" # # Note: linear.weight and linear.bias can return the weight and bias of dygraph.Linear, respectively, and can be used to filter # return ParamBase.name == linear.weight.name # clip = fluid.clip.GradientClipByValue(min=-1, max=1, need_clip=fileter_func) sgd_optimizer = fluid.optimizer.SGD( learning_rate=0.1, parameter_list=linear.parameters()) sgd_optimizer.minimize(loss, grad_clip=clip) """ def __init__(self, max, min=None, need_clip=None): super(GradientClipByValue, self).__init__(need_clip) if min is None: assert (max > 0.0) min = -max self.max = float(max) self.min = float(min) def __str__(self): return "Gradient Clip By Value, min = %f, max=%f" % (self.min, self.max) @imperative_base.no_grad def _dygraph_clip(self, params_grads): params_and_grads = [] for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func(p): params_and_grads.append((p, g)) continue new_grad = layers.clip(x=g, min=self.min, max=self.max) params_and_grads.append((p, new_grad)) return params_and_grads def _static_clip(self, params_grads): params_and_grads = [] with framework.name_scope('gradient_clip'): for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func( p): params_and_grads.append((p, g)) continue with p.block.program._optimized_guard([p, g]): new_grad = layers.clip(x=g, min=self.min, max=self.max) params_and_grads.append((p, new_grad)) _correct_clip_op_role_var(params_and_grads) return params_and_grads def _process_context(self, context, param, grad): pass def _create_operators(self, param, grad): new_grad = layers.clip(x=grad, min=self.min, max=self.max) return param, new_grad class GradientClipByNorm(GradientClipBase): """ Limit the l2 norm of multi-dimensional Tensor :math:`X` to ``clip_norm`` . - If the l2 norm of :math:`X` is greater than ``clip_norm`` , :math:`X` will be compressed by a ratio. - If the l2 norm of :math:`X` is less than or equal to ``clip_norm`` , nothing will be done. The multidimensional Tensor :math:`X` is not passed from this class, but the gradients of all parameters in ``Program`` . If ``need_clip`` is not None, then only part of gradients can be selected for gradient clipping. Gradient clip will takes effect after being set in ``optimizer.minimize(grad_clip)`` , see the document ``optimizer`` (for example: :ref:`api_fluid_optimizer_SGDOptimizer`). The clipping formula is: .. math:: Out = \\left \{ \\begin{aligned} & X & & if (norm(X) \\leq clip\_norm) \\\\ & \\frac{clip\_norm*X}{norm(X)} & & if (norm(X) > clip\_norm) \\\\ \\end{aligned} \\right. where :math:`norm(X)` represents the L2 norm of :math:`X`. .. math:: norm(X) = ( \\sum_{i=1}^{n}|x\_i|^2)^{ \\frac{1}{2}} Args: clip_norm(float): The maximum norm value. need_clip (function, optional): Type: function. This function accepts a ``Parameter`` and returns ``bool`` (True: the gradient of this ``Parameter`` need to be clipped, False: not need). Default: None, and gradients of all parameters in the network will be clipped. Examples: .. code-block:: python # use for Static mode import paddle import paddle.fluid as fluid import numpy as np main_prog = fluid.Program() startup_prog = fluid.Program() with fluid.program_guard( main_program=main_prog, startup_program=startup_prog): image = fluid.data( name='x', shape=[-1, 2], dtype='float32') predict = fluid.layers.fc(input=image, size=3, act='relu') # Trainable parameters: fc_0.w.0, fc_0.b.0 loss = fluid.layers.mean(predict) # Clip all parameters in network: clip = fluid.clip.GradientClipByNorm(clip_norm=1.0) # Clip a part of parameters in network: (e.g. linear_0.w_0) # pass a function(fileter_func) to need_clip, and fileter_func receive a Parameter, and return bool # def fileter_func(Parameter): # # It can be easily filtered by Parameter.name (name can be set in fluid.ParamAttr, and the default name is fc_0.w_0, fc_0.b_0) # return Parameter.name=="fc_0.w_0" # clip = fluid.clip.GradientClipByNorm(clip_norm=1.0, need_clip=fileter_func) sgd_optimizer = fluid.optimizer.SGDOptimizer(learning_rate=0.1) sgd_optimizer.minimize(loss, grad_clip=clip) place = fluid.CPUPlace() exe = fluid.Executor(place) x = np.random.uniform(-100, 100, (10, 2)).astype('float32') exe.run(startup_prog) out = exe.run(main_prog, feed={'x': x}, fetch_list=loss) # use for Dygraph mode import paddle import paddle.fluid as fluid with fluid.dygraph.guard(): linear = fluid.dygraph.Linear(10, 10) # Trainable: linear_0.w.0, linear_0.b.0 inputs = fluid.layers.uniform_random([32, 10]).astype('float32') out = linear(fluid.dygraph.to_variable(inputs)) loss = fluid.layers.reduce_mean(out) loss.backward() # Clip all parameters in network: clip = fluid.clip.GradientClipByNorm(clip_norm=1.0) # Clip a part of parameters in network: (e.g. linear_0.w_0) # pass a function(fileter_func) to need_clip, and fileter_func receive a ParamBase, and return bool # def fileter_func(ParamBase): # # It can be easily filtered by ParamBase.name(name can be set in fluid.ParamAttr, and the default name is linear_0.w_0, linear_0.b_0) # return ParamBase.name == "linear_0.w_0" # # Note: linear.weight and linear.bias can return the weight and bias of dygraph.Linear, respectively, and can be used to filter # return ParamBase.name == linear.weight.name # clip = fluid.clip.GradientClipByNorm(clip_norm=1.0, need_clip=fileter_func) sgd_optimizer = fluid.optimizer.SGD( learning_rate=0.1, parameter_list=linear.parameters()) sgd_optimizer.minimize(loss, grad_clip=clip) """ def __init__(self, clip_norm, need_clip=None): super(GradientClipByNorm, self).__init__(need_clip) self.clip_norm = float(clip_norm) def __str__(self): return "Gradient Clip By Norm, clip_norm=%f" % self.clip_norm @imperative_base.no_grad def _dygraph_clip(self, params_grads): params_and_grads = [] for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func(p): params_and_grads.append((p, g)) continue new_grad = layers.clip_by_norm(x=g, max_norm=self.clip_norm) params_and_grads.append((p, new_grad)) return params_and_grads def _static_clip(self, params_grads): params_and_grads = [] with framework.name_scope('gradient_clip'): for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func( p): params_and_grads.append((p, g)) continue with p.block.program._optimized_guard([p, g]): new_grad = layers.clip_by_norm(x=g, max_norm=self.clip_norm) params_and_grads.append((p, new_grad)) _correct_clip_op_role_var(params_and_grads) return params_and_grads def _process_context(self, context, param, grad): pass def _create_operators(self, param, grad): new_grad = layers.clip_by_norm(x=grad, max_norm=self.clip_norm) return param, new_grad class GradientClipByGlobalNorm(GradientClipBase): """ Given a list of Tensor :math:`t\_list` , calculate the global norm for the elements of all tensors in :math:`t\_list` , and limit it to ``clip_norm`` . - If the global norm is greater than ``clip_norm`` , all elements of :math:`t\_list` will be compressed by a ratio. - If the global norm is less than or equal to ``clip_norm`` , nothing will be done. The list of Tensor :math:`t\_list` is not passed from this class, but the gradients of all parameters in ``Program`` . If ``need_clip`` is not None, then only part of gradients can be selected for gradient clipping. Gradient clip will takes effect after being set in ``optimizer.minimize(grad_clip)`` , see the document ``optimizer`` (for example: :ref:`api_fluid_optimizer_SGDOptimizer`). The clipping formula is: .. math:: t\_list[i] = t\_list[i] * \\frac{clip\_norm}{\max(global\_norm, clip\_norm)} where: .. math:: global\_norm = \sqrt{\sum_{i=0}^{N-1}(l2norm(t\_list[i]))^2} Args: clip_norm (float): The maximum norm value. group_name (str, optional): The group name for this clip. Default value is ``default_group`` need_clip (function, optional): Type: function. This function accepts a ``Parameter`` and returns ``bool`` (True: the gradient of this ``Parameter`` need to be clipped, False: not need). Default: None, and gradients of all parameters in the network will be clipped. Examples: .. code-block:: python # use for Static mode import paddle import paddle.fluid as fluid import numpy as np main_prog = fluid.Program() startup_prog = fluid.Program() with fluid.program_guard( main_program=main_prog, startup_program=startup_prog): image = fluid.data( name='x', shape=[-1, 2], dtype='float32') predict = fluid.layers.fc(input=image, size=3, act='relu') # Trainable parameters: fc_0.w.0, fc_0.b.0 loss = fluid.layers.mean(predict) # Clip all parameters in network: clip = fluid.clip.GradientClipByGlobalNorm(clip_norm=1.0) # Clip a part of parameters in network: (e.g. fc_0.w_0) # pass a function(fileter_func) to need_clip, and fileter_func receive a ParamBase, and return bool # def fileter_func(Parameter): # # It can be easily filtered by Parameter.name (name can be set in fluid.ParamAttr, and the default name is fc_0.w_0, fc_0.b_0) # return Parameter.name=="fc_0.w_0" # clip = fluid.clip.GradientClipByGlobalNorm(clip_norm=1.0, need_clip=fileter_func) sgd_optimizer = fluid.optimizer.SGDOptimizer(learning_rate=0.1) sgd_optimizer.minimize(loss, grad_clip=clip) place = fluid.CPUPlace() exe = fluid.Executor(place) x = np.random.uniform(-100, 100, (10, 2)).astype('float32') exe.run(startup_prog) out = exe.run(main_prog, feed={'x': x}, fetch_list=loss) # use for Dygraph mode import paddle import paddle.fluid as fluid with fluid.dygraph.guard(): linear = fluid.dygraph.Linear(10, 10) # Trainable: linear_0.w.0, linear_0.b.0 inputs = fluid.layers.uniform_random([32, 10]).astype('float32') out = linear(fluid.dygraph.to_variable(inputs)) loss = fluid.layers.reduce_mean(out) loss.backward() # Clip all parameters in network: clip = fluid.clip.GradientClipByGlobalNorm(clip_norm=1.0) # Clip a part of parameters in network: (e.g. linear_0.w_0) # pass a function(fileter_func) to need_clip, and fileter_func receive a ParamBase, and return bool # def fileter_func(ParamBase): # # It can be easily filtered by ParamBase.name(name can be set in fluid.ParamAttr, and the default name is linear_0.w_0, linear_0.b_0) # return ParamBase.name == "linear_0.w_0" # # Note: linear.weight and linear.bias can return the weight and bias of dygraph.Linear, respectively, and can be used to filter # return ParamBase.name == linear.weight.name # clip = fluid.clip.GradientClipByGlobalNorm(clip_norm=1.0, need_clip=fileter_func) sgd_optimizer = fluid.optimizer.SGD( learning_rate=0.1, parameter_list=linear.parameters()) sgd_optimizer.minimize(loss, grad_clip=clip) """ def __init__(self, clip_norm, group_name="default_group", need_clip=None): super(GradientClipByGlobalNorm, self).__init__(need_clip) self.clip_norm = float(clip_norm) self.group_name = group_name def __str__(self): return "Gradient Clip By GlobalNorm, global_norm=%f" % (self.clip_norm) @imperative_base.no_grad def _dygraph_clip(self, params_grads): params_and_grads = [] sum_square_list = [] for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func(p): continue merge_grad = g if g.type == core.VarDesc.VarType.SELECTED_ROWS: merge_grad = layers.merge_selected_rows(g) merge_grad = layers.get_tensor_from_selected_rows(merge_grad) square = layers.square(merge_grad) sum_square = layers.reduce_sum(square) sum_square_list.append(sum_square) # all parameters have been filterd out if len(sum_square_list) == 0: return params_grads global_norm_var = layers.concat(sum_square_list) global_norm_var = layers.reduce_sum(global_norm_var) global_norm_var = layers.sqrt(global_norm_var) max_global_norm = layers.fill_constant( shape=[1], dtype='float32', value=self.clip_norm) clip_var = layers.elementwise_div( x=max_global_norm, y=layers.elementwise_max( x=global_norm_var, y=max_global_norm)) for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func(p): params_and_grads.append((p, g)) continue new_grad = layers.elementwise_mul(x=g, y=clip_var) params_and_grads.append((p, new_grad)) return params_and_grads def _static_clip(self, params_grads): params_and_grads = [] sum_square_list = [] with framework.name_scope('gradient_clip'): for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func( p): continue merge_grad = g with p.block.program._optimized_guard([p, g]): if g.type == core.VarDesc.VarType.SELECTED_ROWS: merge_grad = layers.merge_selected_rows(g) merge_grad = layers.get_tensor_from_selected_rows( merge_grad) square = layers.square(merge_grad) sum_square = layers.reduce_sum(input=square) sum_square_list.append(sum_square) # all parameters have been filterd out if len(sum_square_list) == 0: return params_grads with p.block.program._optimized_guard([p, g]): global_norm_var = layers.sums(sum_square_list) global_norm_var = layers.sqrt(x=global_norm_var) max_global_norm = layers.fill_constant( shape=[1], dtype="float32", value=self.clip_norm) scale_var = layers.elementwise_div( x=max_global_norm, y=layers.elementwise_max( x=max_global_norm, y=global_norm_var)) for p, g in params_grads: if g is None: continue if self._need_clip_func is not None and not self._need_clip_func( p): params_and_grads.append((p, g)) continue with p.block.program._optimized_guard([p, g]): new_grad = layers.elementwise_mul(x=g, y=scale_var) params_and_grads.append((p, new_grad)) _correct_clip_op_role_var(params_and_grads) return params_and_grads def _process_context(self, context, param, grad): if self.group_name not in context: context[self.group_name] = [] context[self.group_name + "_clip_value"] = self.clip_norm context[self.group_name + "_clip"] = layers.fill_constant( shape=[1], dtype="float32", value=self.clip_norm) else: if not self.clip_norm == context[self.group_name + "_clip_value"]: raise ValueError( "All parameters' 'clip_norm' of a same group should be the same" ) merge_grad = grad if grad.type == core.VarDesc.VarType.SELECTED_ROWS: merge_grad = layers.merge_selected_rows(grad) merge_grad = layers.get_tensor_from_selected_rows(merge_grad) square = layers.square(merge_grad) local_norm_var = layers.reduce_sum(input=square) context[self.group_name].append(local_norm_var) self.context = context def _create_operators(self, param, grad): group_scale_name = self.group_name + "_scale" if group_scale_name not in self.context: group_norm_var = layers.sums(input=self.context[self.group_name]) group_norm_var = layers.sqrt(x=group_norm_var) clip_var = self.context[self.group_name + "_clip"] group_scale_var = layers.elementwise_div( x=clip_var, y=layers.elementwise_max( x=clip_var, y=group_norm_var)) assert group_scale_var.shape == (1, ) self.context[group_scale_name] = group_scale_var new_grad = layers.elementwise_mul( x=grad, y=self.context[group_scale_name]) return param, new_grad @framework.dygraph_not_support def set_gradient_clip(clip, param_list=None, program=None): """ Warning: This API must be used after building network, and before ``minimize`` , and it may be removed in future releases, so it is not recommended. It is recommended to use ``minimize(loss, grad_clip=clip)`` to clip gradient. There are three clipping strategies: :ref:`api_fluid_clip_GradientClipByGlobalNorm` , :ref:`api_fluid_clip_GradientClipByNorm` , :ref:`api_fluid_clip_GradientClipByValue` . To specify parameters that require gradient clip. Args: grad_clip (GradientClipBase, optional): Gradient cliping strategy, it's an instance of some derived class of ``GradientClipBase`` . There are three cliping strategies ( :ref:`api_fluid_clip_GradientClipByGlobalNorm` , :ref:`api_fluid_clip_GradientClipByNorm` , :ref:`api_fluid_clip_GradientClipByValue` ). Default value: None, and there is no gradient clipping. param_list (list(Variable), optional): Parameters that require gradient clip. It can be a list of parameter or a list of parameter's name. Default None, meaning that all parameters in the program will be included. program (Program, optional): The program where parameters are located. Default None, meaning that using :ref:`api_fluid_default_main_program` . Returns: None Examples: .. code-block:: python import paddle.fluid as fluid def network(): image = fluid.data(name='image', shape=[ None, 28], dtype='float32') param_attr1 = fluid.ParamAttr("fc1_param") fc1 = fluid.layers.fc(image, size=10, param_attr=param_attr1) param_attr2 = fluid.ParamAttr("fc2_param") fc2 = fluid.layers.fc(fc1, size=10, param_attr=param_attr2) loss = fluid.layers.reduce_mean(fc2) return loss # network 1: clip all parameter gradient with fluid.program_guard(fluid.Program(), fluid.Program()): loss = network() fluid.clip.set_gradient_clip( fluid.clip.GradientClipByGlobalNorm(clip_norm=2.0)) sgd = fluid.optimizer.SGD(learning_rate=1e-3) sgd.minimize(loss) # network 2: clip parameter gradient by name with fluid.program_guard(fluid.Program(), fluid.Program()): loss = network() fluid.clip.set_gradient_clip( fluid.clip.GradientClipByValue(min=-1.0, max=1.0), param_list=["fc1_param", "fc2_param"]) sgd = fluid.optimizer.SGD(learning_rate=1e-3) sgd.minimize(loss) # network 3: clip parameter gradient by value with fluid.program_guard(fluid.Program(), fluid.Program()): loss = network() param_var1 = fluid.default_main_program().global_block().var("fc1_param") param_var2 = fluid.default_main_program().global_block().var("fc2_param") fluid.clip.set_gradient_clip( fluid.clip.GradientClipByValue(min=-1.0, max=1.0), param_list=[param_var1, param_var2]) sgd = fluid.optimizer.SGD(learning_rate=1e-3) sgd.minimize(loss) # network 4: use 'set_gradient_clip' and 'minimize(grad_clip=clip)' together with fluid.program_guard(fluid.Program(), fluid.Program()): loss = network() clip1 = fluid.clip.GradientClipByValue(min=-1.0, max=1.0) clip2 = fluid.clip.GradientClipByNorm(clip_norm=1.0) # Set the gradient clipping strategy: clip1 fluid.clip.set_gradient_clip(clip1) # Set the gradient clipping strategy: clip2 sgd = fluid.optimizer.SGD(learning_rate=1e-3) sgd.minimize(loss, grad_clip=clip2) # 'set_gradient_clip' will not take effect when setting has a conflict, # and the gradient clipping strategy will be 'clip2' """ warnings.warn("Caution! 'set_gradient_clip' is not recommended " "and may be deprecated in future! " "We recommend a new strategy: clip gradient by " "'optimizer.minimize(loss, grad_clip=clip)'. " "This method can reduce the mistakes, please " "see documention of 'optimzier.minimize'.") if not isinstance(clip, GradientClipBase): raise TypeError( "'clip' should be an instance of GradientClipBase's derived class") if program is None: program = framework.default_main_program() for op in program.block(0).ops: if 'op_namescope' in op.all_attrs() and "optimizer" in op.attr( "op_namescope"): warnings.warn( "'minimize' has been invoked before, this will make 'set_gradient_clip' " "be ineffective! Please invoke 'set_gradient_clip' before 'minimize'." ) break if param_list is None: param_list = program.block(0).all_parameters() if all(isinstance(elem, six.string_types) for elem in param_list): param_list = [program.block(0).var(elem) for elem in param_list] if not all(isinstance(elem, framework.Parameter) for elem in param_list): raise TypeError( "'param_list' should be a list of Parameter or basestring(parameter's name)." ) for param in param_list: param.gradient_clip_attr = copy.deepcopy(clip) def append_gradient_clip_ops(param_grads): context = dict() for p, g in param_grads: if g is None: continue with p.block.program._optimized_guard( [p, g]), framework.name_scope('gradient_clip_@CLIP'): clip_attr = getattr(p, 'gradient_clip_attr', None) if clip_attr is None: return param_grads if not isinstance(clip_attr, GradientClipBase): raise TypeError( "clip attribute should be an instance of GradientClipBase") clip_attr._process_context(context=context, param=p, grad=g) res = [] for p, g in param_grads: if g is None: continue with p.block.program._optimized_guard( [p, g]), framework.name_scope('graident_clip_@CLIP'): param, new_grad = clip_attr._create_operators(param=p, grad=g) res.append([param, new_grad]) _correct_clip_op_role_var(res) return res # change wrong mapping relation between param & grad in clip op def _correct_clip_op_role_var(params_grads): for param, grad in params_grads: if grad is None: continue for op in param.block.program.global_block().ops: if 'op_namescope' in op.all_attrs() and "gradient_clip" in op.attr( "op_namescope"): if op.attr('op_role_var'): param_name = op.attr('op_role_var')[0] index = 0 for i in range(len(params_grads)): if params_grads[i][0].name == param_name: index = i correct_p_g = [param_name, params_grads[index][1].name] op._set_attr('op_role_var', correct_p_g) ClipByValue = GradientClipByValue ClipByNorm = GradientClipByNorm ClipByGlobalNorm = GradientClipByGlobalNorm