[pre-commit.ci] pre-commit suggestions

.pre-commit-config.yaml

@@ -9,7 +9,7 @@ ci:
 
 repos:
   - repo: https://github.com/pre-commit/pre-commit-hooks
-    rev: v4.5.0
+    rev: v4.6.0
     hooks:
       - id: end-of-file-fixer
       - id: trailing-whitespace
@@ -23,7 +23,7 @@ repos:
       - id: detect-private-key
 
   - repo: https://github.com/codespell-project/codespell
-    rev: v2.2.6
+    rev: v2.3.0
     hooks:
       - id: codespell
         additional_dependencies: [tomli]
@@ -37,7 +37,7 @@ repos:
         args: ["--in-place"]
 
   - repo: https://github.com/pre-commit/mirrors-prettier
-    rev: v3.1.0
+    rev: v4.0.0-alpha.8
     hooks:
       - id: prettier
         files: \.(json|yml|yaml|toml)
@@ -54,7 +54,7 @@ repos:
           - mdformat_frontmatter
 
   - repo: https://github.com/astral-sh/ruff-pre-commit
-    rev: v0.3.5
+    rev: v0.5.0
     hooks:
       # try to fix what is possible
       - id: ruff

course_UvA-DL/05-transformers-and-MH-attention/Transformers_MHAttention.py

@@ -87,7 +87,7 @@
         os.makedirs(file_path.rsplit("/", 1)[0], exist_ok=True)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:
@@ -796,7 +796,7 @@ def _create_model(self):
             num_heads=self.hparams.num_heads,
             dropout=self.hparams.dropout,
         )
-        # Output classifier per sequence lement
+        # Output classifier per sequence element
         self.output_net = nn.Sequential(
             nn.Linear(self.hparams.model_dim, self.hparams.model_dim),
             nn.LayerNorm(self.hparams.model_dim),
@@ -948,8 +948,8 @@ def _calculate_loss(self, batch, mode="train"):
         acc = (preds.argmax(dim=-1) == labels).float().mean()
 
         # Logging
-        self.log("%s_loss" % mode, loss)
-        self.log("%s_acc" % mode, acc)
+        self.log(f"{mode}_loss", loss)
+        self.log(f"{mode}_acc", acc)
         return loss, acc
 
     def training_step(self, batch, batch_idx):
@@ -1419,8 +1419,8 @@ def _calculate_loss(self, batch, mode="train"):
         preds = preds.squeeze(dim=-1)  # Shape: [Batch_size, set_size]
         loss = F.cross_entropy(preds, labels)  # Softmax/CE over set dimension
         acc = (preds.argmax(dim=-1) == labels).float().mean()
-        self.log("%s_loss" % mode, loss)
-        self.log("%s_acc" % mode, acc, on_step=False, on_epoch=True)
+        self.log(f"{mode}_loss", loss)
+        self.log(f"{mode}_acc", acc, on_step=False, on_epoch=True)
         return loss, acc
 
     def training_step(self, batch, batch_idx):

course_UvA-DL/06-graph-neural-networks/GNN_overview.py

@@ -61,7 +61,7 @@
         os.makedirs(file_path.rsplit("/", 1)[0], exist_ok=True)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:
@@ -616,7 +616,7 @@ def forward(self, data, mode="train"):
         elif mode == "test":
             mask = data.test_mask
         else:
-            assert False, "Unknown forward mode: %s" % mode
+            assert False, f"Unknown forward mode: {mode}"
 
         loss = self.loss_module(x[mask], data.y[mask])
         acc = (x[mask].argmax(dim=-1) == data.y[mask]).sum().float() / mask.sum()
@@ -671,7 +671,7 @@ def train_node_classifier(model_name, dataset, **model_kwargs):
     trainer.logger._default_hp_metric = None  # Optional logging argument that we don't need
 
     # Check whether pretrained model exists. If yes, load it and skip training
-    pretrained_filename = os.path.join(CHECKPOINT_PATH, "NodeLevel%s.ckpt" % model_name)
+    pretrained_filename = os.path.join(CHECKPOINT_PATH, f"NodeLevel{model_name}.ckpt")
     if os.path.isfile(pretrained_filename):
         print("Found pretrained model, loading...")
         model = NodeLevelGNN.load_from_checkpoint(pretrained_filename)
@@ -790,7 +790,7 @@ def print_results(result_dict):
 # %%
 print("Data object:", tu_dataset.data)
 print("Length:", len(tu_dataset))
-print("Average label: %4.2f" % (tu_dataset.data.y.float().mean().item()))
+print(f"Average label: {tu_dataset.data.y.float().mean().item():4.2f}")
 
 # %% [markdown]
 # The first line shows how the dataset stores different graphs.
@@ -957,7 +957,7 @@ def train_graph_classifier(model_name, **model_kwargs):
     trainer.logger._default_hp_metric = None
 
     # Check whether pretrained model exists. If yes, load it and skip training
-    pretrained_filename = os.path.join(CHECKPOINT_PATH, "GraphLevel%s.ckpt" % model_name)
+    pretrained_filename = os.path.join(CHECKPOINT_PATH, f"GraphLevel{model_name}.ckpt")
     if os.path.isfile(pretrained_filename):
         print("Found pretrained model, loading...")
         model = GraphLevelGNN.load_from_checkpoint(pretrained_filename)

course_UvA-DL/07-deep-energy-based-generative-models/Deep_Energy_Models.py

@@ -68,7 +68,7 @@
         os.makedirs(file_path.rsplit("/", 1)[0], exist_ok=True)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:
@@ -770,7 +770,7 @@ def train_model(**kwargs):
     rand_imgs = torch.rand((128,) + model.hparams.img_shape).to(model.device)
     rand_imgs = rand_imgs * 2 - 1.0
     rand_out = model.cnn(rand_imgs).mean()
-    print("Average score for random images: %4.2f" % (rand_out.item()))
+    print(f"Average score for random images: {rand_out.item():4.2f}")
 
 # %% [markdown]
 # As we hoped, the model assigns very low probability to those noisy images.
@@ -781,7 +781,7 @@ def train_model(**kwargs):
     train_imgs, _ = next(iter(train_loader))
     train_imgs = train_imgs.to(model.device)
     train_out = model.cnn(train_imgs).mean()
-    print("Average score for training images: %4.2f" % (train_out.item()))
+    print(f"Average score for training images: {train_out.item():4.2f}")
 
 # %% [markdown]
 # The scores are close to 0 because of the regularization objective that was added to the training.
@@ -803,8 +803,8 @@ def compare_images(img1, img2):
     plt.xticks([(img1.shape[2] + 2) * (0.5 + j) for j in range(2)], labels=["Original image", "Transformed image"])
     plt.yticks([])
     plt.show()
-    print("Score original image: %4.2f" % score1)
-    print("Score transformed image: %4.2f" % score2)
+    print(f"Score original image: {score1:4.2f}")
+    print(f"Score transformed image: {score2:4.2f}")
 
 
 # %% [markdown]

course_UvA-DL/08-deep-autoencoders/Deep_Autoencoders.py

@@ -64,7 +64,7 @@
     file_path = os.path.join(CHECKPOINT_PATH, file_name)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:

course_UvA-DL/09-normalizing-flows/NF_image_modeling.py

@@ -68,7 +68,7 @@
     file_path = os.path.join(CHECKPOINT_PATH, file_name)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:
@@ -518,7 +518,7 @@ def visualize_dequantization(quants, prior=None):
         plt.plot([inp[indices[0][-1]]] * 2, [0, prob[indices[0][-1]]], color=color)
         x_ticks.append(inp[indices[0][0]])
     x_ticks.append(inp.max())
-    plt.xticks(x_ticks, ["%.1f" % x for x in x_ticks])
+    plt.xticks(x_ticks, [f"{x:.1f}" for x in x_ticks])
     plt.plot(inp, prob, color=(0.0, 0.0, 0.0))
     # Set final plot properties
     plt.ylim(0, prob.max() * 1.1)
@@ -1199,8 +1199,8 @@ def print_num_params(model):
 table = [
     [
         key,
-        "%4.3f bpd" % flow_dict[key]["result"]["val"][0]["test_bpd"],
-        "%4.3f bpd" % flow_dict[key]["result"]["test"][0]["test_bpd"],
+        "{:4.3f} bpd".format(flow_dict[key]["result"]["val"][0]["test_bpd"]),
+        "{:4.3f} bpd".format(flow_dict[key]["result"]["test"][0]["test_bpd"]),
         "%2.0f ms" % (1000 * flow_dict[key]["result"]["time"]),
         "%2.0f ms" % (1000 * flow_dict[key]["result"].get("samp_time", 0)),
         "{:,}".format(sum(np.prod(p.shape) for p in flow_dict[key]["model"].parameters())),

course_UvA-DL/10-autoregressive-image-modeling/Autoregressive_Image_Modeling.py

@@ -92,7 +92,7 @@
     file_path = os.path.join(CHECKPOINT_PATH, file_name)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:

course_UvA-DL/11-vision-transformer/Vision_Transformer.py

@@ -69,7 +69,7 @@
         os.makedirs(file_path.rsplit("/", 1)[0], exist_ok=True)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:
@@ -353,8 +353,8 @@ def _calculate_loss(self, batch, mode="train"):
         loss = F.cross_entropy(preds, labels)
         acc = (preds.argmax(dim=-1) == labels).float().mean()
 
-        self.log("%s_loss" % mode, loss)
-        self.log("%s_acc" % mode, acc)
+        self.log(f"{mode}_loss", loss)
+        self.log(f"{mode}_acc", acc)
         return loss
 
     def training_step(self, batch, batch_idx):
@@ -396,7 +396,7 @@ def train_model(**kwargs):
     # Check whether pretrained model exists. If yes, load it and skip training
     pretrained_filename = os.path.join(CHECKPOINT_PATH, "ViT.ckpt")
     if os.path.isfile(pretrained_filename):
-        print("Found pretrained model at %s, loading..." % pretrained_filename)
+        print(f"Found pretrained model at {pretrained_filename}, loading...")
         # Automatically loads the model with the saved hyperparameters
         model = ViT.load_from_checkpoint(pretrained_filename)
     else:

course_UvA-DL/12-meta-learning/Meta_Learning.py

@@ -92,7 +92,7 @@
         os.makedirs(file_path.rsplit("/", 1)[0], exist_ok=True)
     if not os.path.isfile(file_path):
         file_url = base_url + file_name
-        print("Downloading %s..." % file_url)
+        print(f"Downloading {file_url}...")
         try:
             urllib.request.urlretrieve(file_url, file_path)
         except HTTPError as e:
@@ -525,8 +525,8 @@ def calculate_loss(self, batch, mode):
         preds, labels, acc = self.classify_feats(prototypes, classes, query_feats, query_targets)
         loss = F.cross_entropy(preds, labels)
 
-        self.log("%s_loss" % mode, loss)
-        self.log("%s_acc" % mode, acc)
+        self.log(f"{mode}_loss", loss)
+        self.log(f"{mode}_acc", acc)
         return loss
 
     def training_step(self, batch, batch_idx):
@@ -573,7 +573,7 @@ def train_model(model_class, train_loader, val_loader, **kwargs):
     # Check whether pretrained model exists. If yes, load it and skip training
     pretrained_filename = os.path.join(CHECKPOINT_PATH, model_class.__name__ + ".ckpt")
     if os.path.isfile(pretrained_filename):
-        print("Found pretrained model at %s, loading..." % pretrained_filename)
+        print(f"Found pretrained model at {pretrained_filename}, loading...")
         # Automatically loads the model with the saved hyperparameters
         model = model_class.load_from_checkpoint(pretrained_filename)
     else:
@@ -947,8 +947,8 @@ def outer_loop(self, batch, mode="train"):
             opt.step()
             opt.zero_grad()
 
-        self.log("%s_loss" % mode, sum(losses) / len(losses))
-        self.log("%s_acc" % mode, sum(accuracies) / len(accuracies))
+        self.log(f"{mode}_loss", sum(losses) / len(losses))
+        self.log(f"{mode}_acc", sum(accuracies) / len(accuracies))
 
     def training_step(self, batch, batch_idx):
         self.outer_loop(batch, mode="train")

course_UvA-DL/13-contrastive-learning/SimCLR.py

@@ -760,7 +760,7 @@ def train_resnet(batch_size, max_epochs=100, **kwargs):
     # Check whether pretrained model exists. If yes, load it and skip training
     pretrained_filename = os.path.join(CHECKPOINT_PATH, "ResNet.ckpt")
     if os.path.isfile(pretrained_filename):
-        print("Found pretrained model at %s, loading..." % pretrained_filename)
+        print(f"Found pretrained model at {pretrained_filename}, loading...")
         model = ResNet.load_from_checkpoint(pretrained_filename)
     else:
         L.seed_everything(42)  # To be reproducible

flash_tutorials/image_classification/image_classification.py

@@ -1,5 +1,5 @@
 # %% [markdown]
-# In this tutorial, we'll go over the basics of lightning Flash by finetuning/predictin with an ImageClassifier on [Hymenoptera Dataset](https://www.kaggle.com/ajayrana/hymenoptera-data) containing ants and bees images.
+# In this tutorial, we'll go over the basics of lightning Flash by finetuning/prediction with an ImageClassifier on [Hymenoptera Dataset](https://www.kaggle.com/ajayrana/hymenoptera-data) containing ants and bees images.
 #
 # # Finetuning
 #