pre-release

catalyst-team · Jul 29, 2021 · 7d7cb68 · 7d7cb68
1 parent c1cdec1
commit 7d7cb68
Show file tree

Hide file tree

Showing 7 changed files with 70 additions and 75 deletions.
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -8,6 +8,25 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ### Added
 
+-
+
+### Changed
+
+-
+
+### Removed
+
+-
+
+### Fixed
+
+-
+
+
+## [21.07] - 2021-07-29
+
+### Added
+
 - added `pre-commit` hook to run codestyle checker on commit ([#1257](https://github.com/catalyst-team/catalyst/pull/1257))
 - `on publish` github action for docker and docs added ([#1260](https://github.com/catalyst-team/catalyst/pull/1260))
 - MixupCallback and `utils.mixup_batch` ([#1241](https://github.com/catalyst-team/catalyst/pull/1241))

diff --git a/CONTRIBUTING.md b/CONTRIBUTING.md
@@ -109,7 +109,7 @@ We are using the Github CI for our test cases validation:
 - [unit tests](https://github.com/catalyst-team/catalyst/blob/master/.github/workflows/dl_cpu.yml#L113)
 - [integrations tests](https://github.com/catalyst-team/catalyst/blob/master/.github/workflows/dl_cpu.yml#L114#L117)
 
-We also have a [colab minimal CI/CD](https://colab.research.google.com/drive/1JCGTVvWlrIsLXMPRRRSWiAstSLic4nbA) as an independent step-by-step handmade tests option.
+We also have a [colab minimal CI/CD](https://colab.research.google.com/github/catalyst-team/catalyst/blob/master/examples/notebooks/colab_ci_cd.ipynb) as an independent step-by-step handmade tests option.
 Please use it as a collaborative platform, if you have any issues during the PR.
 
 ### Codestyle

diff --git a/README.md b/README.md
@@ -71,13 +71,11 @@ model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
 criterion = nn.CrossEntropyLoss()
 optimizer = optim.Adam(model.parameters(), lr=0.02)
 
+train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
 loaders = {
-    "train": DataLoader(
-        MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-    ),
-    "valid": DataLoader(
-        MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-    ),
+    "train": DataLoader(train_data, batch_size=32),
+    "valid": DataLoader(valid_data, batch_size=32),
 }
 
 runner = dl.SupervisedRunner(
@@ -220,13 +218,11 @@ from catalyst.contrib.datasets import MNIST
 model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
 optimizer = optim.Adam(model.parameters(), lr=0.02)
 
+train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
 loaders = {
-    "train": DataLoader(
-        MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-    ),
-    "valid": DataLoader(
-        MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-    ),
+    "train": DataLoader(train_data, batch_size=32),
+    "valid": DataLoader(valid_data, batch_size=32),
 }
 
 class CustomRunner(dl.Runner):
@@ -626,13 +622,11 @@ model = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
 criterion = nn.CrossEntropyLoss()
 optimizer = optim.Adam(model.parameters(), lr=0.02)
 
+train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
 loaders = {
-    "train": DataLoader(
-        MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-    ),
-    "valid": DataLoader(
-        MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-    ),
+    "train": DataLoader(train_data, batch_size=32),
+    "valid": DataLoader(valid_data, batch_size=32),
 }
 
 runner = dl.SupervisedRunner()
@@ -688,13 +682,11 @@ model = nn.Sequential(
 criterion = IoULoss()
 optimizer = torch.optim.Adam(model.parameters(), lr=0.02)
 
+train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
 loaders = {
-    "train": DataLoader(
-        MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-    ),
-    "valid": DataLoader(
-        MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-    ),
+    "train": DataLoader(train_data, batch_size=32),
+    "valid": DataLoader(valid_data, batch_size=32),
 }
 
 class CustomRunner(dl.SupervisedRunner):
@@ -750,13 +742,11 @@ student = nn.Sequential(nn.Flatten(), nn.Linear(28 * 28, 10))
 criterion = {"cls": nn.CrossEntropyLoss(), "kl": nn.KLDivLoss(reduction="batchmean")}
 optimizer = optim.Adam(student.parameters(), lr=0.02)
 
+train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
 loaders = {
-    "train": DataLoader(
-        MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-    ),
-    "valid": DataLoader(
-        MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-    ),
+    "train": DataLoader(train_data, batch_size=32),
+    "valid": DataLoader(valid_data, batch_size=32),
 }
 
 class DistilRunner(dl.Runner):
@@ -934,11 +924,8 @@ optimizer = {
     "generator": torch.optim.Adam(generator.parameters(), lr=0.0003, betas=(0.5, 0.999)),
     "discriminator": torch.optim.Adam(discriminator.parameters(), lr=0.0003, betas=(0.5, 0.999)),
 }
-loaders = {
-    "train": DataLoader(
-        MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-    )
-}
+train_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
+loaders = {"train": DataLoader(train_data, batch_size=32)}
 
 class CustomRunner(dl.Runner):
     def predict_batch(self, batch):
@@ -1099,13 +1086,11 @@ class CustomRunner(dl.IRunner):
         return 3
 
     def get_loaders(self, stage: str):
+        train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+        valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
         loaders = {
-            "train": DataLoader(
-                MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-            ),
-            "valid": DataLoader(
-                MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-            ),
+            "train": DataLoader(train_data, batch_size=32),
+            "valid": DataLoader(valid_data, batch_size=32),
         }
         return loaders
 
@@ -1202,13 +1187,11 @@ class CustomRunner(dl.IRunner):
         return 3
 
     def get_loaders(self, stage: str):
+        train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+        valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
         loaders = {
-            "train": DataLoader(
-                MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-            ),
-            "valid": DataLoader(
-                MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-            ),
+            "train": DataLoader(train_data, batch_size=32),
+            "valid": DataLoader(valid_data, batch_size=32),
         }
         return loaders
 
@@ -1311,13 +1294,11 @@ class CustomRunner(dl.IRunner):
         return 3
 
     def get_loaders(self, stage: str):
+        train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+        valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
         loaders = {
-            "train": DataLoader(
-                MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-            ),
-            "valid": DataLoader(
-                MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-            ),
+            "train": DataLoader(train_data, batch_size=32),
+            "valid": DataLoader(valid_data, batch_size=32),
         }
         return loaders
 
@@ -1409,13 +1390,11 @@ def objective(trial):
     lr = trial.suggest_loguniform("lr", 1e-3, 1e-1)
     num_hidden = int(trial.suggest_loguniform("num_hidden", 32, 128))
 
+    train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+    valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
     loaders = {
-        "train": DataLoader(
-            MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-        ),
-        "valid": DataLoader(
-            MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-        ),
+        "train": DataLoader(train_data, batch_size=32),
+        "valid": DataLoader(valid_data, batch_size=32),
     }
     model = nn.Sequential(
         nn.Flatten(), nn.Linear(784, num_hidden), nn.ReLU(), nn.Linear(num_hidden, 10)
@@ -1582,6 +1561,7 @@ best practices for your deep learning research and development.
 
 ### Documentation
 - [master](https://catalyst-team.github.io/catalyst/)
+- [21.07](https://catalyst-team.github.io/catalyst/v21.07/index.html)
 - [21.06](https://catalyst-team.github.io/catalyst/v21.06/index.html)
 - [21.05](https://catalyst-team.github.io/catalyst/v21.05/index.html) ([Catalyst — A PyTorch Framework for Accelerated Deep Learning R&D](https://medium.com/pytorch/catalyst-a-pytorch-framework-for-accelerated-deep-learning-r-d-ad9621e4ca88?source=friends_link&sk=885b4409aecab505db0a63b06f19dcef))
 - [21.04/21.04.1](https://catalyst-team.github.io/catalyst/v21.04/index.html), [21.04.2](https://catalyst-team.github.io/catalyst/v21.04.2/index.html)

diff --git a/catalyst/__version__.py b/catalyst/__version__.py
@@ -1 +1 @@
-__version__ = "21.06"
+__version__ = "21.07"
diff --git a/docs/index.rst b/docs/index.rst
@@ -46,13 +46,11 @@ Getting started
     criterion = nn.CrossEntropyLoss()
     optimizer = optim.Adam(model.parameters(), lr=0.02)
 
+    train_data = MNIST(os.getcwd(), train=True, download=True, transform=ToTensor())
+    valid_data = MNIST(os.getcwd(), train=False, download=True, transform=ToTensor())
     loaders = {
-        "train": DataLoader(
-            MNIST(os.getcwd(), train=True, download=True, transform=ToTensor()), batch_size=32
-        ),
-        "valid": DataLoader(
-            MNIST(os.getcwd(), train=False, download=True, transform=ToTensor()), batch_size=32
-        ),
+        "train": DataLoader(train_data, batch_size=32),
+        "valid": DataLoader(valid_data, batch_size=32),
     }
 
     runner = dl.SupervisedRunner(

diff --git a/examples/engines/README.md b/examples/engines/README.md
@@ -2,12 +2,12 @@
 
 Let's check different
 DataParallel and DistributedDataParallel multi-GPU setups with Catalyst Engines. 
-> *Please use `pip install git+https://github.com/catalyst-team/catalyst@master --upgrade` before the `v21.06` release.*
 
 
 ## PyTorch
 ```bash
 pip install catalyst
+CUDA_VISIBLE_DEVICES="0" python multi_gpu.py --engine=de
 CUDA_VISIBLE_DEVICES="0,1" python multi_gpu.py --engine=dp
 CUDA_VISIBLE_DEVICES="0,1" python multi_gpu.py --engine=ddp
 ```

diff --git a/examples/engines/multi_gpu.py b/examples/engines/multi_gpu.py
@@ -13,6 +13,7 @@
 from catalyst.data import transforms
 
 E2E = {
+    "de": dl.DeviceEngine,
     "dp": dl.DataParallelEngine,
     "ddp": dl.DistributedDataParallelEngine,
 }
@@ -97,14 +98,11 @@ def get_loaders(self, stage: str):
         transform = transforms.Compose(
             [transforms.ToTensor(), transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]
         )
+        train_data = CIFAR10(os.getcwd(), train=True, download=True, transform=transform)
+        valid_data = CIFAR10(os.getcwd(), train=False, download=True, transform=transform)
         return {
-            "train": DataLoader(
-                CIFAR10(os.getcwd(), train=True, download=True, transform=transform), batch_size=32
-            ),
-            "valid": DataLoader(
-                CIFAR10(os.getcwd(), train=False, download=True, transform=transform),
-                batch_size=32,
-            ),
+            "train": DataLoader(train_data, batch_size=32),
+            "valid": DataLoader(valid_data, batch_size=32),
         }
 
     def get_model(self, stage: str):