conflict in trainer solved

fastNLP/core/losses.py

@@ -1,23 +1,29 @@
 import torch
+import torch.nn.functional as F
 
+from fastNLP.core.utils import CheckError
+from fastNLP.core.utils import CheckRes
 from fastNLP.core.utils import _get_arg_list
 from fastNLP.core.utils import _map_args
 from fastNLP.core.utils import get_func_signature
 from fastNLP.core.utils import _build_args
+from fastNLP.core.utils import _check_function_or_method
 
 
 class LossBase(object):
     def __init__(self):
         # key: name in target function; value: name in output function
         self.param_map = {}
+        self._checked = False
 
     def get_loss(self, *args, **kwargs):
         raise NotImplementedError
 
-    def __call__(self, output_dict, target_dict):
+    def __call__(self, output_dict, target_dict, force_check=False):
         """
         :param output_dict: A dict from forward function of the network.
         :param target_dict: A dict from DataSet.batch_y.
+        :param force_check: Boolean. Force to check the mapping functions when it is running.
         :return:
         """
         args, defaults, defaults_val, varargs, kwargs = _get_arg_list(self.get_loss)
@@ -27,50 +33,94 @@ class LossBase(object):
             )
 
         param_map = self.param_map
-        for keys in args:
-            if keys not in param_map:
-                param_map.update({keys: keys})
-        for keys in defaults:
-            if keys not in param_map:
-                param_map.update({keys: keys})
+        if args is None:
+            raise RuntimeError(
+                f"There is not any param in function{get_func_signature(self.get_loss)}"
+            )
+        self._checked = self._checked and not force_check
+        if not self._checked:
+            for keys in args:
+                if keys not in param_map:
+                    param_map.update({keys: keys})
+            if defaults is not None:
+                for keys in defaults:
+                    if keys not in param_map:
+                        param_map.update({keys: keys})
+            self.param_map = param_map
         # param map: key= name in get_loss function, value= name in param dict
-        reversed_param_map = {val: key for key, val in param_map}
+        reversed_param_map = {val: key for key, val in param_map.items()}
         # reversed param map: key= name in param dict, value= name in get_loss function
 
+        duplicated = []
+        missing = []
+        if not self._checked:
+            for keys, val in output_dict.items():
+                if keys in target_dict.keys():
+                    duplicated.append(keys)
+
         param_val_dict = {}
         for keys, val in output_dict.items():
-            if keys not in target_dict.keys():
-                param_val_dict.update({keys: val})
-            else:
-                raise RuntimeError("conflict Error in output dict and target dict with name {}".format(keys))
+            param_val_dict.update({keys: val})
         for keys, val in target_dict.items():
-            if keys not in output_dict.keys():
-                param_val_dict.update({keys: val})
-            else:
-                raise RuntimeError("conflict Error in output dict and target dict with name {}".format(keys))
+            param_val_dict.update({keys: val})
 
-        for keys in args:
-            if param_map[keys] not in param_val_dict.keys():
-                raise RuntimeError(f"missing param {keys} in function {get_func_signature(self.get_loss)}")
+        if not self._checked:
+            for keys in args:
+                if param_map[keys] not in param_val_dict.keys():
+                    missing.append(keys)
+
+        if len(duplicated) > 0 or len(missing) > 0:
+            raise CheckError(
+                CheckRes(missing=missing, unused=[], duplicated=duplicated, required=[], all_needed=[]),
+                func_signature=get_func_signature(self.get_loss)
+            )
+
+        self._checked = True
 
         param_map_val = _map_args(reversed_param_map, **param_val_dict)
-        param_value = _build_args(**param_map_val)
+        param_value = _build_args(self.get_loss, **param_map_val)
 
         loss = self.get_loss(**param_value)
 
         if not (isinstance(loss, torch.Tensor) and len(loss.size()) == 0):
             if not isinstance(loss, torch.Tensor):
-                raise RuntimeError("loss ERROR: loss except a torch.Tensor but get {}".format(type(loss)))
-            raise RuntimeError("loss ERROR: len(loss.size()) except 0 but got {}".format(len(loss.size())))
+                raise RuntimeError(f"loss ERROR: loss except a torch.Tensor but get {type(loss)}")
+            raise RuntimeError(f"loss ERROR: the size of loss except torch.Size([]) but got {loss.size}")
 
         return loss
 
 
 class NewLoss(LossBase):
     def __init__(self, func, key_map=None, **kwargs):
-        super(NewLoss).__init__()
-        if not callable(func):
-            raise RuntimeError("")
+        super(NewLoss, self).__init__()
+        _check_function_or_method(func)
+        if key_map is not None:
+            if not isinstance(key_map, dict):
+                raise RuntimeError(f"Loss error: key_map except a {type({})} but got a {type(key_map)}")
+            self.param_map = key_map
+        if len(kwargs) > 0:
+            for key, val in kwargs.items():
+                self.param_map.update({key: val})
+
+        self.get_loss = func
+
+
+class L1Loss(LossBase):
+    def __init__(self):
+        super(L1Loss, self).__init__()
+        self.get_loss = F.l1_loss
+
+
+class BCELoss(LossBase):
+    def __init__(self):
+        super(BCELoss, self).__init__()
+        self.get_loss = F.binary_cross_entropy
+
+
+class NLLLoss(LossBase):
+    def __init__(self):
+        super(NLLLoss, self).__init__()
+        self.get_loss = F.nll_loss
 
 
 class LossInForward(LossBase):

fastNLP/core/optimizer.py

@@ -2,61 +2,28 @@ import torch
 
 
 class Optimizer(object):
-    """Wrapper of optimizer from framework
+    def __init__(self, model_params, **kwargs):
+        if model_params is not None and not isinstance(model_params, torch.Tensor):
+            raise RuntimeError("model parameters should be torch.Tensor, rather than {}".format(type(model_params)))
+        self.model_params = model_params
+        self.settings = kwargs
 
-            1. Adam: lr (float), weight_decay (float)
-            2. AdaGrad
-            3. RMSProp
-            4. SGD: lr (float), momentum (float)
 
-    """
+class SGD(Optimizer):
+    def __init__(self, model_params=None, lr=0.001, momentum=0.9):
+        super(SGD, self).__init__(model_params, lr=lr, momentum=momentum)
 
-    def __init__(self, optimizer_name, **kwargs):
-        """
-        :param optimizer_name: str, the name of the optimizer
-        :param kwargs: the arguments
-
-        """
-        self.optim_name = optimizer_name
-        self.kwargs = kwargs
-
-    @property
-    def name(self):
-        """The name of the optimizer.
-
-        :return: str
-        """
-        return self.optim_name
+    def construct_from_pytorch(self, model_params):
+        if self.model_params is None:
+            self.model_params = model_params
+        return torch.optim.SGD(self.model_params, **self.settings)
 
-    @property
-    def params(self):
-        """The arguments used to create the optimizer.
 
-        :return: dict of (str, *)
-        """
-        return self.kwargs
+class Adam(Optimizer):
+    def __init__(self, model_params=None, lr=0.001, weight_decay=0.8):
+        super(Adam, self).__init__(model_params, lr=lr, weight_decay=weight_decay)
 
     def construct_from_pytorch(self, model_params):
-        """Construct a optimizer from framework over given model parameters."""
-
-        if self.optim_name in ["SGD", "sgd"]:
-            if "lr" in self.kwargs:
-                if "momentum" not in self.kwargs:
-                    self.kwargs["momentum"] = 0
-                optimizer = torch.optim.SGD(model_params, lr=self.kwargs["lr"], momentum=self.kwargs["momentum"])
-            else:
-                raise ValueError("requires learning rate for SGD optimizer")
-
-        elif self.optim_name in ["adam", "Adam"]:
-            if "lr" in self.kwargs:
-                if "weight_decay" not in self.kwargs:
-                    self.kwargs["weight_decay"] = 0
-                optimizer = torch.optim.Adam(model_params, lr=self.kwargs["lr"],
-                                             weight_decay=self.kwargs["weight_decay"])
-            else:
-                raise ValueError("requires learning rate for Adam optimizer")
-
-        else:
-            raise NotImplementedError
-
-        return optimizer
+        if self.model_params is None:
+            self.model_params = model_params
+        return torch.optim.Adam(self.model_params, **self.settings)

fastNLP/core/trainer.py

@@ -1,20 +1,22 @@
-import itertools
 import os
 import time
 import warnings
-from collections import defaultdict
 from datetime import datetime
 from datetime import timedelta
 
 import torch
-from torch import nn
 from tensorboardX import SummaryWriter
+from torch import nn
 
 from fastNLP.core.batch import Batch
-from fastNLP.core.optimizer import Optimizer
+from fastNLP.core.dataset import DataSet
+from fastNLP.core.losses import _prepare_losser
+from fastNLP.core.metrics import _prepare_metrics
+from fastNLP.core.optimizer import Adam
 from fastNLP.core.sampler import RandomSampler
 from fastNLP.core.sampler import SequentialSampler
 from fastNLP.core.tester import Tester
+from fastNLP.core.utils import CheckError
 from fastNLP.core.utils import _build_args
 from fastNLP.core.utils import _check_arg_dict_list
 from fastNLP.core.utils import _move_dict_value_to_device
@@ -30,9 +32,12 @@ class Trainer(object):
     """Main Training Loop
 
     """
-    def __init__(self, train_data, model, losser=None, metrics=None, n_epochs=3, batch_size=32, print_every=-1, validate_every=-1,
+
+    def __init__(self, train_data, model, losser=None, metrics=None, n_epochs=3, batch_size=32, print_every=-1,
+                 validate_every=-1,
                  dev_data=None, use_cuda=False, save_path="./save",
-                 optimizer=Optimizer("Adam", lr=0.01, weight_decay=0), check_code_level=0,
+                 optimizer=Adam(lr=0.01, weight_decay=0), check_code_level=0,
+                 metric_key=None,
                  **kwargs):
         super(Trainer, self).__init__()
 
@@ -49,6 +54,13 @@ class Trainer(object):
 
         # prepare evaluate
         metrics = _prepare_metrics(metrics)
+
+        # parse metric_key
+        # increase_better is True. It means the exp result gets better if the indicator increases.
+        # It is true by default.
+        self.increase_better = False if metric_key[0] == "-" else True
+        self.metric_key = metric_key[1:] if metric_key[0] == "+" or metric_key[0] == "-" else metric_key
+
         # prepare loss
         losser = _prepare_losser(losser)
 
@@ -67,12 +79,10 @@ class Trainer(object):
         self.save_path = save_path
         self.print_every = int(print_every)
         self.validate_every = int(validate_every)
-        self._best_accuracy = 0
+        self.best_metric_indicator = None
 
         self._model_device = model.parameters().__next__().device
 
-        # TODO self._best_accuracy不能表现出当前的metric多种的情况
-
         if isinstance(optimizer, torch.optim.Optimizer):
             self.optimizer = optimizer
         else:
@@ -102,7 +112,7 @@ class Trainer(object):
             if torch.cuda.is_available() and self.use_cuda:
                 self.model = self.model.cuda()
 
-            self.mode(self.model, is_test=False)
+            self._mode(self.model, is_test=False)
 
             start = time.time()
             self.start_time = str(datetime.now().strftime('%Y-%m-%d-%H-%M-%S'))
@@ -112,7 +122,9 @@ class Trainer(object):
                     def __getattr__(self, item):
                         def pass_func(*args, **kwargs):
                             pass
+
                         return pass_func
+
                 self._summary_writer = psudoSW()
             else:
                 path = os.path.join(self.save_path, 'tensorboard_logs_{}'.format(self.start_time))
@@ -121,19 +133,20 @@ class Trainer(object):
             epoch = 1
             while epoch <= self.n_epochs:
 
-                data_iterator = Batch(self.train_data, batch_size=self.batch_size, sampler=RandomSampler(), as_numpy=False)
+                data_iterator = Batch(self.train_data, batch_size=self.batch_size, sampler=RandomSampler(),
+                                      as_numpy=False)
 
-                self._train_epoch(data_iterator, self.model, epoch, self.dev_data, start)
+                self._train_epoch(data_iterator, self.model, epoch, start)
 
                 # validate_every override validation at end of epochs
                 if self.dev_data and self.validate_every <= 0:
-                    self.do_validation()
+                    self._do_validation()
                 epoch += 1
         finally:
             self._summary_writer.close()
             del self._summary_writer
 
-    def _train_epoch(self, data_iterator, model, epoch, dev_data, start, **kwargs):
+    def _train_epoch(self, data_iterator, model, epoch, start):
         """Training process in one epoch.
 
             kwargs should contain:
@@ -144,10 +157,10 @@ class Trainer(object):
         for batch_x, batch_y in data_iterator:
             # TODO 这里可能会遇到问题，万一用户在model内部修改了prediction的device就会有问题
             _move_dict_value_to_device(self._model_device, batch_x, batch_y)
-            prediction = self.data_forward(model, batch_x)
-            loss = self.get_loss(prediction, batch_y)
-            self.grad_backward(loss)
-            self.update()
+            prediction = self._data_forward(model, batch_x)
+            loss = self._compute_loss(prediction, batch_y)
+            self._grad_backward(loss)
+            self._update()
             self._summary_writer.add_scalar("loss", loss.item(), global_step=self.step)
             for name, param in self.model.named_parameters():
                 if param.requires_grad:
@@ -162,18 +175,19 @@ class Trainer(object):
                 print(print_output)
 
             if self.validate_every > 0 and self.step % self.validate_every == 0:
-                self.do_validation()
+                self._do_validation()
 
             self.step += 1
 
-    def do_validation(self):
+    def _do_validation(self):
         res = self.tester.test()
         for name, num in res.items():
             self._summary_writer.add_scalar("valid_{}".format(name), num, global_step=self.step)
-        if self.save_path is not None and self.best_eval_result(res):
-            self.save_model(self.model, 'best_model_' + self.start_time)
+        if self.save_path is not None and self._better_eval_result(res):
+            self._save_model(self.model,
+                            "best_" + "_".join([self.model.__class__.__name__, self.metric_key, self.start_time]))
 
-    def mode(self, model, is_test=False):
+    def _mode(self, model, is_test=False):
         """Train mode or Test mode. This is for PyTorch currently.
 
         :param model: a PyTorch model
@@ -185,20 +199,20 @@ class Trainer(object):
         else:
             model.train()
 
-    def update(self):
+    def _update(self):
         """Perform weight update on a model.
 
         """
         self.optimizer.step()
 
-    def data_forward(self, network, x):
+    def _data_forward(self, network, x):
         x = _build_args(network.forward, **x)
         y = network(**x)
         if not isinstance(y, dict):
             raise TypeError(f"The return value of {get_func_signature(network.forward)} should be dict, got {type(y)}.")
         return y
 
-    def grad_backward(self, loss):
+    def _grad_backward(self, loss):
         """Compute gradient with link rules.
 
         :param loss: a scalar where back-prop starts
@@ -208,7 +222,7 @@ class Trainer(object):
         self.model.zero_grad()
         loss.backward()
 
-    def get_loss(self, predict, truth):
+    def _compute_loss(self, predict, truth):
         """Compute loss given prediction and ground truth.
 
         :param predict: prediction dict, produced by model.forward
@@ -217,34 +231,59 @@ class Trainer(object):
         """
         return self.losser(predict, truth)
 
-    def save_model(self, model, model_name, only_param=False):
+    def _save_model(self, model, model_name, only_param=False):
         model_name = os.path.join(self.save_path, model_name)
         if only_param:
             torch.save(model.state_dict(), model_name)
         else:
             torch.save(model, model_name)
 
-    def best_eval_result(self, metrics):
+    def _better_eval_result(self, metrics):
         """Check if the current epoch yields better validation results.
 
-        :return: bool, True means current results on dev set is the best.
+        :return bool value: True means current results on dev set is the best.
         """
         if isinstance(metrics, tuple):
             loss, metrics = metrics
 
         if isinstance(metrics, dict):
             if len(metrics) == 1:
-                accuracy = list(metrics.values())[0]
+                # only single metric, just use it
+                metric_dict = list(metrics.values())[0]
+                metrics_name = list(metrics.keys())[0]
             else:
-                accuracy = metrics[self.eval_sort_key]
-        else:
-            accuracy = metrics
-
-        if accuracy > self._best_accuracy:
-            self._best_accuracy = accuracy
-            return True
-        else:
-            return False
+                metrics_name = self.metrics[0].__class__.__name__
+                if metrics_name not in metrics:
+                    raise RuntimeError(f"{metrics_name} is chosen to do validation, but got {metrics}")
+                metric_dict = metrics[metrics_name]
+
+            if len(metric_dict) == 1:
+                indicator_val, indicator = list(metric_dict.values())[0], list(metric_dict.keys())[0]
+            elif len(metric_dict) > 1 and self.metric_key is None:
+                raise RuntimeError(
+                    f"Got multiple metric keys: {metric_dict}, but metric_key is not set. Which one to use?")
+            else:
+                # metric_key is set
+                if self.metric_key not in metric_dict:
+                    raise RuntimeError(f"matric key {self.metric_key} not found in {metric_dict}")
+                indicator_val = metric_dict[self.metric_key]
+
+            is_better = True
+            if self.best_metric_indicator is None:
+                # first-time validation
+                self.best_metric_indicator = indicator_val
+            else:
+                if self.increase_better is True:
+                    if indicator_val > self.best_metric_indicator:
+                        self.best_metric_indicator = indicator_val
+                    else:
+                        is_better = False
+                else:
+                    if indicator_val < self.best_metric_indicator:
+                        self.best_metric_indicator = indicator_val
+                    else:
+                        is_better = False
+            return is_better
 
 
 DEFAULT_CHECK_BATCH_SIZE = 2
@@ -285,18 +324,17 @@ def _check_code(dataset, model, losser, metrics, batch_size=DEFAULT_CHECK_BATCH_
                         f"should be torch.size([])")
             loss.backward()
         except CheckError as e:
-            _check_loss_evaluate(prev_func=model.forward, func=e.func_signature,
+            pre_func_signature = get_func_signature(model.forward)
+            _check_loss_evaluate(prev_func_signature=pre_func_signature, func_signature=e.func_signature,
                                  check_res=e.check_res, output=output, batch_y=batch_y,
                                  check_level=check_level)
         model.zero_grad()
-        if batch_count+1>=DEFAULT_CHECK_NUM_BATCH:
+        if batch_count + 1 >= DEFAULT_CHECK_NUM_BATCH:
             break
 
     if dev_data is not None:
-        tester = Tester(data=dataset[:batch_size*DEFAULT_CHECK_NUM_BATCH], model=model, metrics=metrics,
+        tester = Tester(data=dataset[:batch_size * DEFAULT_CHECK_NUM_BATCH], model=model, metrics=metrics,
                         batch_size=batch_size, verbose=-1)
         tester.test()
 
 
-
-

test/core/test_loss.py

@@ -2,6 +2,7 @@ import math
 import unittest
 
 import torch as tc
+import torch.nn.functional as F
 
 import fastNLP.core.losses as loss
 
@@ -13,7 +14,11 @@ class TestLoss(unittest.TestCase):
 
 		print (".----------------------------------")
 
-		loss_func = loss.Loss("nll")
+		# loss_func = loss.Loss("nll")
+		print(callable(tc.nn.NLLLoss))
+		loss_func = loss.NewLoss(F.nll_loss)
+
+		nll_loss = loss.NLLLoss()
 
 		#pdb.set_trace()
 
@@ -35,16 +40,18 @@ class TestLoss(unittest.TestCase):
 
 
 		y = tc.log(y)
-		los = loss_func(y , gy)
+		los = loss_func({'input': y}, {'target': gy})
+		losses = nll_loss({'input': y}, {'target': gy})
 
 		r = -math.log(.3) - math.log(.3) - math.log(.1)
 		r /= 3
 		print ("loss = %f" % (los))
 		print ("r = %f" % (r))
+		print ("nll_loss = %f" % (losses))
 
 		self.assertEqual(int(los * 1000), int(r * 1000))
 
-	def test_case_2(self):
+	def _test_case_2(self):
 		#验证squash()的正确性
 		print ("----------------------------------")
 
@@ -74,7 +81,8 @@ class TestLoss(unittest.TestCase):
 		#pdb.set_trace()
 
 		y = tc.log(y)
-		los = loss_func(y , gy)
+		#los = loss_func({'input': y}, {'target': gy})
+		los = loss_func(y, gy)
 		print ("loss = %f" % (los))
 
 		r = -log(.3) - log(.3) - log(.1) - log(.3) - log(.7) - log(.1)
@@ -89,7 +97,8 @@ class TestLoss(unittest.TestCase):
 
 		log = math.log
 
-		loss_func = loss.Loss("nll")
+		#loss_func = loss.Loss("nll")
+		loss_func = loss.NLLLoss()
 
 		#pdb.set_trace()
 
@@ -117,7 +126,7 @@ class TestLoss(unittest.TestCase):
 
 		yy = tc.nn.utils.rnn.pack_padded_sequence(y , lens , batch_first = True).data
 		gyy = tc.nn.utils.rnn.pack_padded_sequence(gy , lens , batch_first = True).data
-		los = loss_func(yy , gyy)
+		los = loss_func({'input': yy}, {'target': gyy})
 		print ("loss = %f" % (los))
 
 
@@ -303,5 +312,58 @@ class TestLoss(unittest.TestCase):
 		print ("r = %f" % (r))
 		self.assertEqual(int(los * 1000), int(r * 1000))
 
+	def test_case_8(self):
+		def func(a, b):
+			import torch.nn.functional as F
+			return F.cross_entropy(a, b)
+
+		def func2(a, truth):
+			return func(a, truth)
+
+		def func3(predict, truth):
+			return func(predict, truth)
+
+		def func4(a, b, c=2):
+			return (a + b) * c
+
+		def func6(a, b, **kwargs):
+			c = kwargs['c']
+			return (a + b) * c
+
+		import torch
+		from fastNLP.core.losses import LossBase, NewLoss
+
+		get_loss = NewLoss(func, {'a': 'predict', 'b': 'truth'})
+		predict = torch.randn(5, 3)
+		truth = torch.LongTensor([1, 0, 1, 2, 1])
+		loss1 = get_loss({'predict': predict}, {'truth': truth})
+		get_loss_2 = NewLoss(func2, {'a': 'predict'})
+		loss2 = get_loss_2({'predict': predict}, {'truth': truth})
+		get_loss_3 = NewLoss(func3)
+		loss3 = get_loss_3({'predict': predict}, {'truth': truth})
+		print(loss1, loss2, loss3)
+		assert loss1 == loss2 and loss1 == loss3
+
+		get_loss_4 = NewLoss(func4)
+		loss4 = get_loss_4({'a': 1, 'b': 3}, {})
+		print(loss4)
+		assert loss4 == (1 + 3) * 2
+
+		get_loss_5 = NewLoss(func4)
+		loss5 = get_loss_5({'a': 1, 'b': 3}, {'c': 4})
+		print(loss5)
+		assert loss5 == (1 + 3) * 4
+
+		get_loss_6 = NewLoss(func6)
+		loss6 = get_loss_6({'a': 1, 'b': 3}, {'c': 4})
+		print(loss6)
+		assert loss6 == (1 + 3) * 4
+
+		get_loss_7 = NewLoss(func6, c='cc')
+		loss7 = get_loss_7({'a': 1, 'b': 3}, {'cc': 4})
+		print(loss7)
+		assert loss7 == (1 + 3) * 4
+
+
 if __name__ == "__main__":
 	unittest.main()

test/core/test_optimizer.py

@@ -0,0 +1,21 @@
+import unittest
+
+import torch
+
+from fastNLP.core.optimizer import SGD
+
+
+class TestOptim(unittest.TestCase):
+    def test_case(self):
+        optim = SGD(torch.LongTensor(10))
+        print(optim.__dict__)
+
+        optim_2 = SGD(lr=0.001)
+        print(optim_2.__dict__)
+
+        optim_2 = SGD(lr=0.002, momentum=0.989)
+        print(optim_2.__dict__)
+
+    def test_case_2(self):
+        with self.assertRaises(RuntimeError):
+            _ = SGD(0.001)

test/core/test_trainer.py

@@ -4,3 +4,4 @@ import unittest
 class TestTrainer(unittest.TestCase):
     def test_case_1(self):
         pass
+