CNNSupervisedTrainer.ftl 7.17 KB
Newer Older
Nicola Gatto's avatar
Nicola Gatto committed
1 2 3 4 5 6 7 8
import mxnet as mx
import logging
import numpy as np
import time
import os
import shutil
from mxnet import gluon, autograd, nd

9
class ${tc.fileNameWithoutEnding}:
Nicola Gatto's avatar
Nicola Gatto committed
10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
    def __init__(self, data_loader, net_constructor, net=None):
        self._data_loader = data_loader
        self._net_creator = net_constructor
        self._net = net

    def train(self, batch_size=64,
              num_epoch=10,
              eval_metric='acc',
              optimizer='adam',
              optimizer_params=(('learning_rate', 0.001),),
              load_checkpoint=True,
              context='gpu',
              checkpoint_period=5,
              normalize=True):
        if context == 'gpu':
            mx_context = mx.gpu()
        elif context == 'cpu':
            mx_context = mx.cpu()
        else:
            logging.error("Context argument is '" + context + "'. Only 'cpu' and 'gpu are valid arguments'.")

        if 'weight_decay' in optimizer_params:
            optimizer_params['wd'] = optimizer_params['weight_decay']
            del optimizer_params['weight_decay']
        if 'learning_rate_decay' in optimizer_params:
            min_learning_rate = 1e-08
            if 'learning_rate_minimum' in optimizer_params:
                min_learning_rate = optimizer_params['learning_rate_minimum']
                del optimizer_params['learning_rate_minimum']
            optimizer_params['lr_scheduler'] = mx.lr_scheduler.FactorScheduler(
                                                   optimizer_params['step_size'],
                                                   factor=optimizer_params['learning_rate_decay'],
                                                   stop_factor_lr=min_learning_rate)
            del optimizer_params['step_size']
            del optimizer_params['learning_rate_decay']


        train_iter, test_iter, data_mean, data_std = self._data_loader.load_data(batch_size)
        if self._net is None:
            if normalize:
                self._net_creator.construct(
51
                    context=mx_context, data_mean=data_mean, data_std=data_std)
Nicola Gatto's avatar
Nicola Gatto committed
52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71
            else:
                self._net_creator.construct(context=mx_context)

        begin_epoch = 0
        if load_checkpoint:
            begin_epoch = self._net_creator.load(mx_context)
        else:
            if os.path.isdir(self._net_creator._model_dir_):
                shutil.rmtree(self._net_creator._model_dir_)

        self._net = self._net_creator.net

        try:
            os.makedirs(self._net_creator._model_dir_)
        except OSError:
            if not os.path.isdir(self._net_creator._model_dir_):
                raise

        trainer = mx.gluon.Trainer(self._net.collect_params(), optimizer, optimizer_params)

72 73 74 75 76 77 78 79 80 81 82 83
        loss_functions = {}

        for output_name, last_layer in self._net.last_layers.items():
            if last_layer == 'softmax':
                loss_functions[output_name] = mx.gluon.loss.SoftmaxCrossEntropyLoss()
            elif last_layer == 'sigmoid':
                loss_functions[output_name] = mx.gluon.loss.SigmoidBinaryCrossEntropyLoss()
            elif last_layer == 'linear':
                loss_functions[output_name] = mx.gluon.loss.L2Loss()
            else:
                loss_functions[output_name] = mx.gluon.loss.L2Loss()
                logging.warning("Invalid last layer, defaulting to L2 loss")
Nicola Gatto's avatar
Nicola Gatto committed
84 85 86 87 88 89 90

        speed_period = 50
        tic = None

        for epoch in range(begin_epoch, begin_epoch + num_epoch):
            train_iter.reset()
            for batch_i, batch in enumerate(train_iter):
91
                <#list tc.architectureInputs as input_name>
92 93 94 95
                ${input_name}_data = batch.data[${input_name?index}].as_in_context(mx_context)
                </#list>
                <#list tc.architectureOutputs as output_name>
                ${output_name}_label = batch.label[${output_name?index}].as_in_context(mx_context)
96 97
                </#list>

Nicola Gatto's avatar
Nicola Gatto committed
98
                with autograd.record():
99 100 101
                    ${tc.join(tc.architectureOutputs, ", ", "", "_output")} = self._net(${tc.join(tc.architectureInputs, ", ", "", "_data")})

                    loss = <#list tc.architectureOutputs as output_name>loss_functions['${output_name}'](${output_name}_output, ${output_name}_label)<#sep> + </#list>
Nicola Gatto's avatar
Nicola Gatto committed
102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123

                loss.backward()
                trainer.step(batch_size)

                if tic is None:
                    tic = time.time()
                else:
                    if batch_i % speed_period == 0:
                        try:
                            speed = speed_period * batch_size / (time.time() - tic)
                        except ZeroDivisionError:
                            speed = float("inf")

                        logging.info("Epoch[%d] Batch[%d] Speed: %.2f samples/sec" % (epoch, batch_i, speed))

                        tic = time.time()

            tic = None

            train_iter.reset()
            metric = mx.metric.create(eval_metric)
            for batch_i, batch in enumerate(train_iter):
124
                <#list tc.architectureInputs as input_name>
125
                ${input_name}_data = batch.data[${input_name?index}].as_in_context(mx_context)
126 127
                </#list>

128 129 130 131 132 133 134 135 136 137 138
                labels = [
                    <#list tc.architectureOutputs as output_name>batch.label[${output_name?index}].as_in_context(mx_context)<#sep>, </#list>
                ]

                ${tc.join(tc.architectureOutputs, ", ", "", "_output")} = self._net(${tc.join(tc.architectureInputs, ", ", "", "_data")})

                predictions = [
                    <#list tc.architectureOutputs as output_name>mx.nd.argmax(${output_name}_output, axis=1)<#sep>, </#list>
                ]

                metric.update(preds=predictions, labels=labels)
Nicola Gatto's avatar
Nicola Gatto committed
139 140 141 142 143
            train_metric_score = metric.get()[1]

            test_iter.reset()
            metric = mx.metric.create(eval_metric)
            for batch_i, batch in enumerate(test_iter):
144
                <#list tc.architectureInputs as input_name>
145
                ${input_name}_data = batch.data[${input_name?index}].as_in_context(mx_context)
146 147
                </#list>

148 149 150 151 152 153 154 155 156 157 158
                labels = [
                    <#list tc.architectureOutputs as output_name>batch.label[${output_name?index}].as_in_context(mx_context)<#sep>, </#list>
                ]

                ${tc.join(tc.architectureOutputs, ", ", "", "_output")} = self._net(${tc.join(tc.architectureInputs, ", ", "", "_data")})

                predictions = [
                    <#list tc.architectureOutputs as output_name>mx.nd.argmax(${output_name}_output, axis=1)<#sep>, </#list>
                ]

                metric.update(preds=predictions, labels=labels)
Nicola Gatto's avatar
Nicola Gatto committed
159 160 161 162 163 164 165 166 167 168 169 170
            test_metric_score = metric.get()[1]

            logging.info("Epoch[%d] Train: %f, Test: %f" % (epoch, train_metric_score, test_metric_score))

            if (epoch - begin_epoch) % checkpoint_period == 0:
                self._net.save_parameters(self.parameter_path() + '-' + str(epoch).zfill(4) + '.params')

        self._net.save_parameters(self.parameter_path() + '-' + str(num_epoch + begin_epoch).zfill(4) + '.params')
        self._net.export(self.parameter_path() + '_newest', epoch=0)

    def parameter_path(self):
        return self._net_creator._model_dir_ + self._net_creator._model_prefix_