pytorch
diff --git a/‎test/test_utils.py‎
Lines changed: 0 additions & 160 deletions b/‎test/test_utils.py‎
Lines changed: 0 additions & 160 deletions
diff --git a/‎torch/utils/trainer/__init__.py‎
Lines changed: 0 additions & 2 deletions b/‎torch/utils/trainer/__init__.py‎
Lines changed: 0 additions & 2 deletions
diff --git a/‎torch/utils/trainer/plugins/__init__.py‎
Lines changed: 0 additions & 5 deletions b/‎torch/utils/trainer/plugins/__init__.py‎
Lines changed: 0 additions & 5 deletions
diff --git a/‎torch/utils/trainer/plugins/accuracy.py‎
Lines changed: 0 additions & 19 deletions b/‎torch/utils/trainer/plugins/accuracy.py‎
Lines changed: 0 additions & 19 deletions
diff --git a/‎torch/utils/trainer/plugins/logger.py‎
Lines changed: 0 additions & 83 deletions b/‎torch/utils/trainer/plugins/logger.py‎
Lines changed: 0 additions & 83 deletions
diff --git a/‎torch/utils/trainer/plugins/loss.py‎
Lines changed: 0 additions & 8 deletions b/‎torch/utils/trainer/plugins/loss.py‎
Lines changed: 0 additions & 8 deletions
diff --git a/‎torch/utils/trainer/plugins/monitor.py‎
Lines changed: 0 additions & 57 deletions b/‎torch/utils/trainer/plugins/monitor.py‎
Lines changed: 0 additions & 57 deletions
diff --git a/‎torch/utils/trainer/plugins/plugin.py‎
Lines changed: 0 additions & 10 deletions b/‎torch/utils/trainer/plugins/plugin.py‎
Lines changed: 0 additions & 10 deletions
@@ -14,9 +14,6 @@
 import torch.cuda
 import warnings
 from torch.utils.checkpoint import checkpoint, checkpoint_sequential
-from torch.utils.trainer import Trainer
-from torch.utils.trainer.plugins import *
-from torch.utils.trainer.plugins.plugin import Plugin
 from torch.autograd._functions.utils import prepare_onnx_paddings
 from torch.autograd._functions.utils import check_onnx_broadcast
 from common import IS_WINDOWS, IS_PPC, skipIfRocm
@@ -26,85 +23,6 @@
 from common import TestCase, run_tests, download_file
 
 
-class SimplePlugin(Plugin):
-
-    def __init__(self, interval):
-        super(SimplePlugin, self).__init__(interval)
-        self.trainer = None
-        self.num_iteration = 0
-        self.num_epoch = 0
-        self.num_batch = 0
-        self.num_update = 0
-
-    def register(self, trainer):
-        self.trainer = trainer
-
-    def iteration(self, *args):
-        self.iteration_args = args
-        self.num_iteration += 1
-
-    def epoch(self, *args):
-        self.epoch_args = args
-        self.num_epoch += 1
-
-    def batch(self, *args):
-        self.batch_args = args
-        self.num_batch += 1
-
-    def update(self, *args):
-        self.update_args = args
-        self.num_update += 1
-
-
-class ModelMock(object):
-
-    def __init__(self):
-        self.num_calls = 0
-        self.output = torch.ones(1, 1, requires_grad=True)
-
-    def __call__(self, i):
-        self.num_calls += 1
-        return self.output * 2
-
-
-class CriterionMock(object):
-
-    def __init__(self):
-        self.num_calls = 0
-
-    def __call__(self, out, target):
-        self.num_calls += 1
-        return out
-
-
-class OptimizerMock(object):
-    max_evals = 5
-    min_evals = 1
-
-    def __init__(self):
-        self.num_steps = 0
-        self.num_evals = 0
-
-    def step(self, closure):
-        for i in range(random.randint(self.min_evals, self.max_evals)):
-            loss = closure()
-            self.num_evals += 1
-        self.num_steps += 1
-
-    def zero_grad(self):
-        pass
-
-
-class DatasetMock(object):
-
-    def __iter__(self):
-        for i in range(10):
-            yield torch.randn(2, 10), torch.randperm(10)[:2]
-
-    def __len__(self):
-        return 10
-
-
 class RandomDatasetMock(object):
 
     def __getitem__(self, index):
@@ -279,84 +197,6 @@ def test_multi_drop(self):
         self.assertEqual(len(list(dataiter)), 1)
 
 
-class TestTrainer(TestCase):
-
-    intervals = [
-        [(1, 'iteration')],
-        [(1, 'epoch')],
-        [(1, 'batch')],
-        [(1, 'update')],
-        [(5, 'iteration')],
-        [(5, 'epoch')],
-        [(5, 'batch')],
-        [(5, 'update')],
-        [(1, 'iteration'), (1, 'epoch')],
-        [(5, 'update'), (1, 'iteration')],
-        [(2, 'epoch'), (1, 'batch')],
-    ]
-
-    def setUp(self):
-        self.optimizer = OptimizerMock()
-        self.trainer = Trainer(ModelMock(), CriterionMock(),
-                               self.optimizer, DatasetMock())
-        self.num_epochs = 3
-        self.dataset_size = len(self.trainer.dataset)
-        self.num_iters = self.num_epochs * self.dataset_size
-
-    def test_register_plugin(self):
-        for interval in self.intervals:
-            simple_plugin = SimplePlugin(interval)
-            self.trainer.register_plugin(simple_plugin)
-            self.assertEqual(simple_plugin.trainer, self.trainer)
-
-    def test_optimizer_step(self):
-        self.trainer.run(epochs=1)
-        self.assertEqual(self.trainer.optimizer.num_steps, 10)
-
-    def test_plugin_interval(self):
-        for interval in self.intervals:
-            self.setUp()
-            simple_plugin = SimplePlugin(interval)
-            self.trainer.register_plugin(simple_plugin)
-            self.trainer.run(epochs=self.num_epochs)
-            units = {
-                ('iteration', self.num_iters),
-                ('epoch', self.num_epochs),
-                ('batch', self.num_iters),
-                ('update', self.num_iters)
-            }
-            for unit, num_triggers in units:
-                call_every = None
-                for i, i_unit in interval:
-                    if i_unit == unit:
-                        call_every = i
-                        break
-                if call_every:
-                    expected_num_calls = math.floor(num_triggers / call_every)
-                else:
-                    expected_num_calls = 0
-                num_calls = getattr(simple_plugin, 'num_' + unit)
-                self.assertEqual(num_calls, expected_num_calls, 0)
-
-    def test_model_called(self):
-        self.trainer.run(epochs=self.num_epochs)
-        num_model_calls = self.trainer.model.num_calls
-        num_crit_calls = self.trainer.criterion.num_calls
-        self.assertEqual(num_model_calls, num_crit_calls)
-        for num_calls in [num_model_calls, num_crit_calls]:
-            lower_bound = OptimizerMock.min_evals * self.num_iters
-            upper_bound = OptimizerMock.max_evals * self.num_iters
-            self.assertEqual(num_calls, self.trainer.optimizer.num_evals)
-            self.assertLessEqual(lower_bound, num_calls)
-            self.assertLessEqual(num_calls, upper_bound)
-
-    def test_model_gradient(self):
-        self.trainer.run(epochs=self.num_epochs)
-        output_var = self.trainer.model.output
-        expected_grad = torch.ones(1, 1) * 2 * self.optimizer.num_evals
-        self.assertEqual(output_var.grad.data, expected_grad)
-
-
 test_dir = os.path.abspath(os.path.dirname(str(__file__)))