RecomputeOptimizer

api_attr

declarative programming (static graph)

class paddle.fluid.optimizer.RecomputeOptimizer(optimizer)[source]

Recompute Optimizer Wrapper

Normally, a training step contains three sub-steps: first, run forward Operators to calculate the loss; second, run backward Operators to calculate gradient of the parameters; third, apply optimization method to update the value of the parameters.

In the forward computation process, all variables that are needed by backward computation process will be kept in memory, which occupy a great amount of memory when the network becomes very deep.

Recompute split the network to k segments. In each segment, It will recompute the forward Operators, before running backward operators. It is very helpful for saving memory.

The Variables that separate a network to segments are called as checkpoints, and users should set it manually. The usage is very simple:

Parameters

optimizer (Optimizer) – The optimizer that is applied to parameters.

Examples

import paddle.fluid as fluid
import numpy as np
def gen_data():
    return {"x": np.random.random(size=(32, 32)).astype('float32'),
    "y": np.random.randint(2, size=(32, 1)).astype('int64')}
def mlp(input_x, input_y, hid_dim=128, label_dim=2):
    print(input_x)
    fc_1 = fluid.layers.fc(input=input_x, size=hid_dim)
    prediction = fluid.layers.fc(input=[fc_1], size=label_dim, act='softmax')
    cost = fluid.layers.cross_entropy(input=prediction, label=input_y)
    sum_cost = fluid.layers.reduce_mean(cost)
    return sum_cost, fc_1, prediction
input_x = fluid.layers.data(name="x", shape=[32], dtype='float32')
input_y = fluid.layers.data(name="y", shape=[1], dtype='int64')
cost, fc_1, pred = mlp(input_x, input_y)

sgd = fluid.optimizer.Adam(learning_rate=0.01)
sgd = fluid.optimizer.RecomputeOptimizer(sgd)
sgd._set_checkpoints([fc_1, pred])
sgd.minimize(cost)

print("Finished optimize")
place = fluid.CPUPlace()
exe = fluid.Executor(place)
exe.run(fluid.default_startup_program())
step = 10

for i in range(step):
    cost_val = exe.run(feed=gen_data(),
           program=fluid.default_main_program(),
           fetch_list=[cost.name])
    print("step=%d cost=%f" % (i, cost_val[0]))
minimize(loss, startup_program=None, parameter_list=None, no_grad_set=None)

Add operations to minimize loss by updating parameter_list.

Parameters

  • loss (Variable) – A Variable containing the value to minimize.

  • startup_program (Program, optional) – Program for initializing parameters in parameter_list. The default value is None, at this time default_startup_program will be used.

  • parameter_list (list, optional) – List of Variable or Variable.name to update to minimize loss. The default value is None, at this time all parameters will be updated.

  • no_grad_set (set, optional) – Set of Variable or Variable.name that don’t need to be updated. The default value is None.

Returns

tuple (optimize_ops, params_grads), A list of operators appended by minimize and a list of (param, grad) variable pairs, param is Parameter, grad is the gradient value corresponding to the parameter. The returned tuple can be passed to fetch_list in Executor.run() to indicate program pruning. If so, the program will be pruned by feed and fetch_list before run, see details in Executor.

Return type

tuple

Examples

Please refer to the example of current Optimizer.

clear_gradients()

Clear the gradients of all optimized parameters for model.

Returns

None

Examples

import paddle.fluid as fluid
import numpy as np

with fluid.dygraph.guard():
    value = np.arange(26).reshape(2, 13).astype("float32")
    a = fluid.dygraph.to_variable(value)
    linear = fluid.Linear(13, 5, dtype="float32")
    # This can be any optimizer supported by dygraph.
    adam = fluid.optimizer.Adam(learning_rate = 0.01,
                                parameter_list = linear.parameters())
    out = linear(a)
    out.backward()
    adam.minimize(out)
    adam.clear_gradients()
current_step_lr()

Note

This API is ONLY available in Dygraph mode

Get current step learning rate. The return value is all the same When LearningRateDecay is not used, otherwise return the step learning rate.

Returns

The learning rate of the current step.

Return type

float

Examples

import paddle.fluid as fluid
import numpy as np

# example1: LearningRateDecay is not used, return value is all the same
with fluid.dygraph.guard():
    emb = fluid.dygraph.Embedding([10, 10])
    adam = fluid.optimizer.Adam(0.001, parameter_list = emb.parameters())
    lr = adam.current_step_lr()
    print(lr) # 0.001

# example2: PiecewiseDecay is used, return the step learning rate
with fluid.dygraph.guard():
    inp = np.random.uniform(-0.1, 0.1, [10, 10]).astype("float32")
    linear = fluid.dygraph.nn.Linear(10, 10)
    inp = fluid.dygraph.to_variable(inp)
    out = linear(inp)
    loss = fluid.layers.reduce_mean(out)

    bd = [2, 4, 6, 8]
    value = [0.2, 0.4, 0.6, 0.8, 1.0]
    adam = fluid.optimizer.Adam(fluid.dygraph.PiecewiseDecay(bd, value, 0),
                           parameter_list=linear.parameters())

    # first step: learning rate is 0.2
    np.allclose(adam.current_step_lr(), 0.2, rtol=1e-06, atol=0.0) # True

    # learning rate for different steps
    ret = [0.2, 0.2, 0.4, 0.4, 0.6, 0.6, 0.8, 0.8, 1.0, 1.0, 1.0, 1.0]
    for i in range(12):
        adam.minimize(loss)
        lr = adam.current_step_lr()
        np.allclose(lr, ret[i], rtol=1e-06, atol=0.0) # True
set_dict(state_dict)

Load optimizer state dict. For Adam optimizer, contains beta1, beta2, momentum etc. If LearningRateDecay have been used, global_step will be changed.

Parameters

state_dict (dict) – Dict contains all the Variable needed by optimizer

Returns

None

Examples

with fluid.dygraph.guard():
    emb = fluid.dygraph.Embedding([10, 10])

    state_dict = emb.state_dict()
    fluid.save_dygraph(state_dict, "paddle_dy")

    adam = fluid.optimizer.Adam(learning_rate=fluid.layers.noam_decay( 100, 10000),
                                parameter_list=emb.parameters())
    state_dict = adam.state_dict()
    fluid.save_dygraph(state_dict, "paddle_dy")

    para_state_dict, opti_state_dict = fluid.load_dygraph( "paddle_dy")

    adam.set_dict(opti_state_dict)
state_dict()

Get state dict information from optimizer. It contain all the variable used by optimizer. For Adam optimizer, contains beta1, beta2, momentum etc. If LearningRateDecay have been used, global_step will be include in state dict. If the optimizer never be called(minimize function), the state_dict is empty.

Args: None :returns: dict contains all the variable used by optimizer :rtype: state_dict(dict)

Examples

import paddle.fluid as fluid

with fluid.dygraph.guard():
    emb = fluid.dygraph.Embedding([10, 10])

    adam = fluid.optimizer.Adam(0.001, parameter_list=emb.parameters())
    state_dict = adam.state_dict()