updates

Author: Programmer-RD-AI

ML/__init__.py

@@ -27,7 +27,9 @@
 from sklearn.metrics import *
 from torch.hub import *
 import torchtext.functional as F
+import warnings
 
+warnings.filterwarnings("ignore")
 print(torch.__version__, torchvision.__version__, torchtext.__version__)
 os.environ["CUDA_LAUNCH_BLOCKING"] = "1"
 # os.environ["WANDB_SILENT"] = "true"

ML/dataset/main_loaders.py

@@ -29,7 +29,7 @@ def __init__(
         self.batch_size = batch_size
         self.get_batches()
 
-    def get_batches(self):
+    def get_batches(self) -> None:
         X = self.X_train if self.train else self.X_test
         y = self.y_train if self.train else self.y_test
         X_batches = []

ML/dataset/valid_loaders.py

@@ -5,7 +5,10 @@
 class Valid_Loader(Loader):
     def __init__(self, *args) -> None:
         super().__init__(*args)
+        self.data["id"].dropna(inplace=True)
         self.X = self.data["text"].to_numpy()
+        self.ids = self.data["id"].to_numpy()
+        print(len(self.X), len(self.ids))
 
     def __getitem__(self, index) -> np.array:
-        return self.transform(self.X[index])
+        return (self.ids[index], [self.transform(self.X[index])])

ML/helper_functions/__init__.py

@@ -1,4 +1,4 @@
 from ML.helper_functions.load_data import *
 from ML.helper_functions.test import *
 from ML.helper_functions.train import *
-from ML.helper_functions.transformer import *
+from ML.helper_functions.transformations.transformer import *

ML/helper_functions/alert.py

@@ -0,0 +1,6 @@
+from ML import *
+
+
+class Alert:
+    def __init__(self) -> None:
+        pass

ML/helper_functions/clearcache.py

@@ -2,8 +2,8 @@
 
 
 class ClearCache:
-    def __enter__(self):
+    def __enter__(self) -> None:
         torch.cuda.empty_cache()
 
-    def __exit__(self, *args):
+    def __exit__(self, *args) -> None:
         torch.cuda.empty_cache()

ML/helper_functions/load_data.py

@@ -49,7 +49,7 @@ def ld(self) -> Tuple[DataLoader, DataLoader, DataLoader]:
             num_workers=round(os.cpu_count() / 2),
         )
         self.valid_data_loader = DataLoader(
-            self.dataset_valid(self.valid_path, None),
+            self.dataset_valid(self.valid_path, self.main_transform),
             batch_size=None,
             shuffle=False,
             num_workers=round(os.cpu_count() / 2),

ML/helper_functions/test.py

@@ -16,7 +16,7 @@ def __init__(
         self.model = model
         self.name = name
 
-    def test(self):
+    def test(self) -> Dict:
         a_tot = 0
         p_tot = 0
         r_tot = 0
@@ -29,7 +29,7 @@ def test(self):
                 X = F.to_tensor(X, padding_value=1).to("cuda")
                 y = torch.tensor(y).to("cuda")
                 preds = torch.argmax(torch.softmax(self.model(X), dim=1), dim=1)
-                loss = self.criteria(preds, y.view(-1, 1).squeeze(1))
+                loss = self.criterion(preds.float(), y.view(-1, 1).squeeze(1).float())
                 results = classification_report(
                     preds.cpu(), y.view(-1, 1).squeeze(1).cpu(), output_dict=True
                 )
@@ -43,10 +43,25 @@ def test(self):
                 f1_tot += f1score
                 l_tot += loss.item()
                 n += 1
+        print(loss.item(), l_tot, l_tot / n)
         return {
             f"{self.name} precision": p_tot / n,
             f"{self.name} recall": r_tot / n,
             f"{self.name} f1-score": f1_tot / n,
             f"{self.name} accuracy": a_tot / n,
             f"{self.name} loss": l_tot / n,
         }
+
+    def make_predictions(self, run_name: str, epoch: int) -> pd.DataFrame:
+        ids = []
+        target = []
+        for i, X in enumerate(self.valid_dataloader):
+            X = F.to_tensor(X, padding_value=1).to("cuda")
+            pred = torch.argmax(torch.softmax(self.model(X), dim=1), dim=1).squeeze().cpu().item()
+            ids.append(i)
+            target.append(pred)
+        if run_name not in os.listdir("./ML/predictions/"):
+            os.mkdir(f"./ML/predictions/{run_name}")
+        df = pd.DataFrame({"id": ids, "target": target})
+        df.to_csv(f"./ML/predictions/{run_name}/{epoch}.csv", index=False)
+        return df

ML/helper_functions/train.py

@@ -26,12 +26,12 @@ def __init__(
         self.optimizer = optimizer
         self.lr_schedular = lr_schedular
 
-    def train(self, run_name):
+    def train(self, run_name: str) -> None:
         print(torchinfo.summary(self.model))
         wandb.init(project=PROJECT_NAME, name=run_name, config=self.config)
         wandb.watch(self.model, log="all")
         iterator = tqdm(range(self.epochs))
-        for _ in iterator:
+        for epoch in iterator:
             torch.cuda.empty_cache()
             for i, (X, y) in enumerate(self.train_dataloader):
                 y = y[0]
@@ -61,6 +61,13 @@ def train(self, run_name):
                     "Train",
                 ).test()
             )
+            Test(
+                self.train_dataloader,
+                self.valid_dataloader,
+                self.criterion,
+                self.model,
+                "Train",
+            ).make_predictions(run_name, epoch)
             iterator.set_description(f"Testing Done")
             self.model.train()
         wandb.save()

ML/helper_functions/transformations/__init__.py

@@ -0,0 +1,2 @@
+from ML.helper_functions.transformations.randomize import *
+from ML.helper_functions.transformations.transformer import *