Fix data: imports

CHANGELOG.md

@@ -8,11 +8,12 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Added
 
--
+- Reinforcement learning tutorials ([#1205](https://github.com/catalyst-team/catalyst/pull/1205))
+- customization demo ([#1207](https://github.com/catalyst-team/catalyst/pull/1207))
 
 ### Changed
 
--
+- tests moved to `tests` folder ([#1208](https://github.com/catalyst-team/catalyst/pull/1208))
 
 ### Removed
 
@@ -21,6 +22,7 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 ### Fixed
 
 - customizing what happens in `train()` notebook ([#1203](https://github.com/catalyst-team/catalyst/pull/1203))
+- transforms imports under catalyst.data ([#1211](https://github.com/catalyst-team/catalyst/pull/1211))
 
 
 ## [21.04.2] - 2021-04-30

README.md

@@ -64,7 +64,7 @@ import os
 from torch import nn, optim
 from torch.utils.data import DataLoader
 from catalyst import dl, utils
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 from catalyst.contrib.datasets import MNIST
 
 model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
@@ -206,7 +206,7 @@ from torch import nn, optim
 from torch.nn import functional as F
 from torch.utils.data import DataLoader
 from catalyst import dl, metrics
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 from catalyst.contrib.datasets import MNIST
 
 model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
@@ -241,7 +241,7 @@ class CustomRunner(dl.Runner):
         logits = self.model(x)
         # compute the loss
         loss = F.cross_entropy(logits, y)
-        # compute other metrics of interest
+        # compute the metrics
         accuracy01, accuracy03 = metrics.accuracy(logits, y, topk=(1, 3))
         # log metrics
         self.batch_metrics.update(
@@ -611,7 +611,7 @@ import os
 from torch import nn, optim
 from torch.utils.data import DataLoader
 from catalyst import dl
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 from catalyst.contrib.datasets import MNIST
 
 model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
@@ -668,7 +668,7 @@ import torch
 from torch import nn
 from torch.utils.data import DataLoader
 from catalyst import dl
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 from catalyst.contrib.datasets import MNIST
 from catalyst.contrib.nn import IoULoss
 
@@ -733,7 +733,7 @@ from torch import nn, optim
 from torch.nn import functional as F
 from torch.utils.data import DataLoader
 from catalyst import dl
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 from catalyst.contrib.datasets import MNIST
 
 # [!] teacher model should be already pretrained
@@ -895,7 +895,7 @@ from torch.utils.data import DataLoader
 from catalyst import dl
 from catalyst.contrib.datasets import MNIST
 from catalyst.contrib.nn.modules import Flatten, GlobalMaxPool2d, Lambda
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 
 latent_dim = 128
 generator = nn.Sequential(
@@ -1038,7 +1038,7 @@ from torch.nn import functional as F
 from torch.utils.data import DataLoader
 from catalyst import dl, metrics
 from catalyst.contrib.datasets import MNIST
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 
 LOG_SCALE_MAX = 2
 LOG_SCALE_MIN = -10
@@ -1167,7 +1167,7 @@ from torch import nn, optim
 from torch.utils.data import DataLoader
 from catalyst import dl, utils
 from catalyst.contrib.datasets import MNIST
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 
 
 class CustomRunner(dl.IRunner):
@@ -1277,7 +1277,7 @@ from torch import nn, optim
 from torch.utils.data import DataLoader
 from catalyst import dl, utils
 from catalyst.contrib.datasets import MNIST
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 
 
 class CustomRunner(dl.IRunner):
@@ -1393,7 +1393,7 @@ import torch
 from torch import nn
 from torch.utils.data import DataLoader
 from catalyst import dl
-from catalyst.data.transforms import ToTensor
+from catalyst.data import ToTensor
 from catalyst.contrib.datasets import MNIST
 
 
@@ -1491,9 +1491,9 @@ best practices for your deep learning research and development.
 - [20.04](https://catalyst-team.github.io/catalyst/v20.04/index.html), [20.04.1](https://catalyst-team.github.io/catalyst/v20.04.1/index.html), [20.04.2](https://catalyst-team.github.io/catalyst/v20.04.2/index.html)
 
 ### Notebooks
-- Introduction tutorial "[Customizing what happens in `train`](./examples/notebooks/customizing_what_happens_in_train.ipynb)" [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/customizing_what_happens_in_train.ipynb)
-- Demo with [customization examples](./examples/notebooks/customization_tutorial.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/customization_tutorial.ipynb)
-- [Reinforcement Learning with Catalyst](./examples/notebooks/reinforcement_learning.ipynb) [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/reinforcement_learning.ipynb)
+- [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/customizing_what_happens_in_train.ipynb) Introduction tutorial "[Customizing what happens in `train`](./examples/notebooks/customizing_what_happens_in_train.ipynb)"
+- [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/customization_tutorial.ipynb) Demo with [customization examples](./examples/notebooks/customization_tutorial.ipynb)
+- [![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/reinforcement_learning.ipynb) [Reinforcement Learning with Catalyst](./examples/notebooks/reinforcement_learning.ipynb)
 
 ### Notable Blog Posts
 - [Catalyst 2021–Accelerated PyTorch 2.0](https://medium.com/catalyst-team/catalyst-2021-accelerated-pytorch-2-0-850e9b575cb6?source=friends_link&sk=865d3c472cfb10379864656fedcfe762)

catalyst/callbacks/metrics/segmentation.py

@@ -28,7 +28,7 @@ class IOUCallback(BatchMetricCallback):
         from torch import nn
         from torch.utils.data import DataLoader
         from catalyst import dl
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
         from catalyst.contrib.nn import IoULoss
 
@@ -138,7 +138,7 @@ class DiceCallback(BatchMetricCallback):
         from torch import nn
         from torch.utils.data import DataLoader
         from catalyst import dl
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
         from catalyst.contrib.nn import IoULoss
 
@@ -252,7 +252,7 @@ class TrevskyCallback(BatchMetricCallback):
         from torch import nn
         from torch.utils.data import DataLoader
         from catalyst import dl
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
         from catalyst.contrib.nn import IoULoss
 

catalyst/callbacks/onnx.py

@@ -37,7 +37,7 @@ class OnnxCallback(Callback):
             from torch.utils.data import DataLoader
 
             from catalyst import dl
-            from catalyst.data.transforms import ToTensor
+            from catalyst.data import ToTensor
             from catalyst.contrib.datasets import MNIST
             from catalyst.contrib.nn.modules import Flatten
 

catalyst/callbacks/quantization.py

@@ -33,7 +33,7 @@ class QuantizationCallback(Callback):
             from torch.utils.data import DataLoader
 
             from catalyst import dl
-            from catalyst.data.transforms import ToTensor
+            from catalyst.data import ToTensor
             from catalyst.contrib.datasets import MNIST
             from catalyst.contrib.nn.modules import Flatten
 

catalyst/callbacks/tracing.py

@@ -32,7 +32,7 @@ class TracingCallback(Callback):
             from torch.utils.data import DataLoader
 
             from catalyst import dl
-            from catalyst.data.transforms import ToTensor
+            from catalyst.data import ToTensor
             from catalyst.contrib.datasets import MNIST
             from catalyst.contrib.nn.modules import Flatten
 
@@ -94,7 +94,7 @@ def __init__(
                 from torch.utils.data import DataLoader
 
                 from catalyst import dl
-                from catalyst.data.transforms import ToTensor
+                from catalyst.data import ToTensor
                 from catalyst.contrib.datasets import MNIST
                 from catalyst.contrib.nn.modules import Flatten
 

catalyst/core/runner.py

@@ -86,7 +86,7 @@ class IRunner(ICallback, ILogger, ABC):
         from torch.utils.data import DataLoader
         from catalyst import dl, utils
         from catalyst.contrib.datasets import MNIST
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
 
 
         class CustomRunner(dl.IRunner):

catalyst/data/__init__.py

@@ -30,4 +30,6 @@
     HardClusterSampler,
 )
 
+from catalyst.data.transforms import Compose, Normalize, ToTensor, to_tensor, normalize
+
 from catalyst.contrib.data import *

catalyst/metrics/_additive.py

@@ -35,7 +35,7 @@ class AdditiveValueMetric(IMetric):
         from torch.nn import functional as F
         from torch.utils.data import DataLoader
         from catalyst import dl, metrics
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
 
         model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))

catalyst/metrics/_segmentation.py

@@ -236,7 +236,7 @@ class IOUMetric(RegionBasedMetric):
         from torch import nn
         from torch.utils.data import DataLoader
         from catalyst import dl
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
         from catalyst.contrib.nn import IoULoss
 
@@ -365,7 +365,7 @@ class DiceMetric(RegionBasedMetric):
         from torch import nn
         from torch.utils.data import DataLoader
         from catalyst import dl
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
         from catalyst.contrib.nn import IoULoss
 
@@ -498,7 +498,7 @@ class TrevskyMetric(RegionBasedMetric):
         from torch import nn
         from torch.utils.data import DataLoader
         from catalyst import dl
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
         from catalyst.contrib.nn import IoULoss
 

catalyst/runners/runner.py

@@ -84,7 +84,7 @@ class Runner(IRunner):
         from torch.nn import functional as F
         from torch.utils.data import DataLoader
         from catalyst import dl, metrics
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
 
         model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
@@ -407,7 +407,7 @@ def train(
             from torch.nn import functional as F
             from torch.utils.data import DataLoader
             from catalyst import dl, metrics
-            from catalyst.data.transforms import ToTensor
+            from catalyst.data import ToTensor
             from catalyst.contrib.datasets import MNIST
 
             model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
@@ -577,7 +577,7 @@ def predict_loader(
             from torch.nn import functional as F
             from torch.utils.data import DataLoader
             from catalyst import dl, metrics
-            from catalyst.data.transforms import ToTensor
+            from catalyst.data import ToTensor
             from catalyst.contrib.datasets import MNIST
 
             model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
@@ -673,7 +673,7 @@ def on_loader_end(self, runner):
     def evaluate_loader(
         self,
         loader: DataLoader,
-        callbacks: "Union[List[Callback], OrderedDict[str, Callback]]",
+        callbacks: "Union[List[Callback], OrderedDict[str, Callback]]" = None,
         model: Optional[Model] = None,
         seed: int = 42,
         verbose: bool = False,
@@ -746,7 +746,7 @@ class SupervisedRunner(ISupervisedRunner, Runner):
         from torch import nn, optim
         from torch.utils.data import DataLoader
         from catalyst import dl, utils
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
 
         model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))

catalyst/runners/supervised.py

@@ -49,7 +49,7 @@ class ISupervisedRunner(IRunner):
         from torch import nn, optim
         from torch.utils.data import DataLoader
         from catalyst import dl, utils
-        from catalyst.data.transforms import ToTensor
+        from catalyst.data import ToTensor
         from catalyst.contrib.datasets import MNIST
 
         model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))

docs/faq/ddp.rst

@@ -13,7 +13,7 @@ Suppose you have the following pipeline with MNIST Classification:
     from torch import nn, optim
     from torch.utils.data import DataLoader
     from catalyst import dl
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST
 
     model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))

docs/faq/finetuning.rst

@@ -19,7 +19,7 @@ it's quite easy to create such complex pipeline with a few line of code:
     from torch.utils.data import DataLoader
     from catalyst import dl, utils
     from catalyst.contrib.datasets import MNIST
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
 
 
     class CustomRunner(dl.IRunner):
@@ -144,7 +144,7 @@ Due to multiprocessing setup during distrubuted training, the multistage runs lo
     from torch.utils.data import DataLoader, DistributedSampler
     from catalyst import dl, utils
     from catalyst.contrib.datasets import MNIST
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
 
 
     class CustomRunner(dl.IRunner):

docs/faq/inference.rst

@@ -13,7 +13,7 @@ Suppose you have the following classification pipeline:
     from torch.nn import functional as F
     from torch.utils.data import DataLoader
     from catalyst import dl, metrics, utils
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST
 
     model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))

docs/faq/logging.rst

@@ -33,7 +33,7 @@ You could log any new metric in a straightforward way:
     from torch.nn import functional as F
     from torch.utils.data import DataLoader
     from catalyst import dl, metrics
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST
 
     model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))

docs/faq/multi_components.rst

@@ -14,7 +14,7 @@ Suppose you have the following classification pipeline (in pure PyTorch):
     from torch.nn import functional as F
     from torch.utils.data import DataLoader
     from catalyst import dl, metrics, utils
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST
 
     model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
@@ -96,7 +96,7 @@ Multi-model example:
     from torch.nn import functional as F
     from torch.utils.data import DataLoader
     from catalyst import dl, metrics, utils
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST
 
     # <--- multi-model setup --->
@@ -191,7 +191,7 @@ Multi-optimizer example:
     from torch.nn import functional as F
     from torch.utils.data import DataLoader
     from catalyst import dl, metrics, utils
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST
 
     # <--- multi-model/optimizer setup --->
@@ -290,7 +290,7 @@ Multi-criterion example:
     from torch.nn import functional as F
     from torch.utils.data import DataLoader
     from catalyst import dl, metrics, utils
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST
 
     model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))

docs/faq/optuna.rst

@@ -14,7 +14,7 @@ You can easily use Optuna for hyperparameters optimization:
     from torch import nn
     from torch.utils.data import DataLoader
     from catalyst import dl
-    from catalyst.data.transforms import ToTensor
+    from catalyst.data import ToTensor
     from catalyst.contrib.datasets import MNIST