diff --git a/bnpm/ca2p_preprocessing.py b/bnpm/ca2p_preprocessing.py
index 45815dd..6823d90 100644
--- a/bnpm/ca2p_preprocessing.py
+++ b/bnpm/ca2p_preprocessing.py
@@ -2,7 +2,7 @@
 import scipy.signal
 import matplotlib.pyplot as plt
 import torch
-from tqdm import tqdm
+from tqdm.auto import tqdm
 
 import time
 import gc
diff --git a/bnpm/decomposition.py b/bnpm/decomposition.py
index b381208..8d5f591 100644
--- a/bnpm/decomposition.py
+++ b/bnpm/decomposition.py
@@ -762,11 +762,12 @@ def fit(self, X):
             X.moveaxis(self.batch_dimension, 0),
             torch.arange(X.shape[self.batch_dimension]),
         )
-        self.kwargs_dataloader.pop('batch_size', None)
+        kwargs_dataloader_tmp = copy.deepcopy(self.kwargs_dataloader)
+        kwargs_dataloader_tmp.pop('batch_size', None)
         dataloader = torch.utils.data.DataLoader(
             dataset, 
             batch_size=self.batch_size, 
-            **self.kwargs_dataloader,
+            **kwargs_dataloader_tmp,
         )
 
         kwargs_tmp = copy.deepcopy(self.kwargs_CP_NN_HALS)
diff --git a/bnpm/file_helpers.py b/bnpm/file_helpers.py
index ffcaef7..cab2391 100644
--- a/bnpm/file_helpers.py
+++ b/bnpm/file_helpers.py
@@ -4,7 +4,7 @@
 from pathlib import Path
 import zipfile
 
-from tqdm import tqdm
+from tqdm.auto import tqdm
 
 from . import path_helpers
 
@@ -92,7 +92,7 @@ def prepare_path(path, mkdir=False, exist_ok=True):
     return str(path_obj)
 
 ### Custom functions for preparing paths for saving and loading files and directories
-def prepare_filepath_for_saving(filepath, mkdir=False, allow_overwrite=True):
+def prepare_filepath_for_saving(filepath, mkdir=False, allow_overwrite=False):
     return prepare_path(filepath, mkdir=mkdir, exist_ok=allow_overwrite)
 def prepare_filepath_for_loading(filepath, must_exist=True):
     path = prepare_path(filepath, mkdir=False, exist_ok=must_exist)
@@ -116,7 +116,7 @@ def pickle_save(
     mode='wb', 
     zipCompress=False, 
     mkdir=False, 
-    allow_overwrite=True,
+    allow_overwrite=False,
     library='pickle',
     **kwargs_zipfile,
 ):
@@ -221,7 +221,7 @@ def pickle_load(
             return pickle.load(f)
 
 
-def json_save(obj, filepath, indent=4, mode='w', mkdir=False, allow_overwrite=True):
+def json_save(obj, filepath, indent=4, mode='w', mkdir=False, allow_overwrite=False):
     """
     Saves an object to a json file.
     Uses json.dump.
@@ -267,7 +267,7 @@ def json_load(filepath, mode='r'):
         return json.load(f)
 
 
-def yaml_save(obj, filepath, indent=4, mode='w', mkdir=False, allow_overwrite=True):
+def yaml_save(obj, filepath, indent=4, mode='w', mkdir=False, allow_overwrite=False):
     """
     Saves an object to a yaml file.
     Uses yaml.dump.
@@ -363,7 +363,7 @@ def matlab_save(
     obj,
     filepath,
     mkdir=False, 
-    allow_overwrite=True,
+    allow_overwrite=False,
     clean_string=True,
     list_to_objArray=True,
     none_to_nan=True,
@@ -470,7 +470,7 @@ def download_file(
     hash_type='MD5', 
     hash_hex=None,
     mkdir=False,
-    allow_overwrite=True,
+    allow_overwrite=False,
     write_mode='wb',
     verbose=True,
     chunk_size=1024,
diff --git a/bnpm/neural_networks.py b/bnpm/neural_networks.py
index d80be38..807cc0a 100644
--- a/bnpm/neural_networks.py
+++ b/bnpm/neural_networks.py
@@ -2,7 +2,7 @@
 
 import torch
 import numpy as np
-from tqdm import tqdm
+from tqdm.auto import tqdm
 
 class RegressionRNN(torch.nn.Module):
     """
diff --git a/bnpm/parallel_helpers.py b/bnpm/parallel_helpers.py
index b089475..297436f 100644
--- a/bnpm/parallel_helpers.py
+++ b/bnpm/parallel_helpers.py
@@ -3,7 +3,7 @@
 import multiprocessing as mp
 from functools import partial
 import numpy as np
-from tqdm import tqdm
+from tqdm.auto import tqdm
 
 class ParallelExecutionError(Exception):
     """
diff --git a/bnpm/server.py b/bnpm/server.py
index 30d32be..7638140 100644
--- a/bnpm/server.py
+++ b/bnpm/server.py
@@ -1261,7 +1261,7 @@ def make_rsync_command(
 Implemented by casperdcl here: https://github.com/tqdm/tqdm/issues/311#issuecomment-387066847
 """
 try:
-    from tqdm import tqdm
+    from tqdm.auto import tqdm
 except ImportError:
     class _TqdmWrap(object):
         # tqdm not installed - construct and return dummy/basic versions
diff --git a/bnpm/similarity.py b/bnpm/similarity.py
index b5cebb8..f2348b5 100644
--- a/bnpm/similarity.py
+++ b/bnpm/similarity.py
@@ -7,7 +7,7 @@
 import scipy.optimize
 from numba import njit, prange, jit
 import torch
-from tqdm import tqdm
+from tqdm.auto import tqdm
 
 from . import indexing, torch_helpers
 
diff --git a/bnpm/torch_helpers.py b/bnpm/torch_helpers.py
index b7ae02d..d6da368 100644
--- a/bnpm/torch_helpers.py
+++ b/bnpm/torch_helpers.py
@@ -10,7 +10,7 @@
 import torch
 from torch.utils.data import Dataset
 import numpy as np
-from tqdm import tqdm
+from tqdm.auto import tqdm
 
 from . import indexing
 from . import misc
@@ -523,7 +523,89 @@ def __getitem__(
                 idx (int):
                     The index of the requested sample.
         """
-        return self.X[idx], idx
+        return self.X[idx]
+    
+
+class Dataset_numpy(Dataset):
+    """
+    Creates a PyTorch dataset from a numpy array. 
+    RH 2024
+
+    Args:
+        X (np.ndarray):
+            The data from which to create the dataset.
+        axis (int):
+            The dimension along which to sample the data.
+        device (str): 
+            The device where the tensors will be stored. 
+        dtype (torch.dtype): 
+            The data type to use for the tensor.
+
+    Attributes:
+        X (np.ndarray or np.memmap):
+            The data from the numpy file.
+        n_samples (int):
+            The number of samples in the dataset.
+
+    Returns:
+        (torch.utils.data.Dataset):
+            A PyTorch dataset.
+    """
+    def __init__(
+        self,
+        X: Union[np.ndarray, np.memmap],
+        axis: int = 0,
+        device: str = 'cpu',
+        dtype: torch.dtype = torch.float32,
+    ):
+        """
+        Initializes the Dataset_NumpyFile with the provided parameters.
+        """
+        assert isinstance(X, (np.ndarray, np.memmap)), 'X must be a numpy array or memmap.'
+        self.X = X
+        self.n_samples = self.X.shape[axis]
+        self.is_memmap = isinstance(self.X, np.memmap) 
+        self.axis = axis
+        self.device = device
+        self.dtype = dtype
+
+    def __len__(self) -> int:
+        """
+        Returns the number of samples in the dataset.
+
+        Returns:
+            (int):
+                n_samples (int):
+                    The number of samples in the dataset.
+        """
+        return self.n_samples
+    
+    def __getitem__(
+        self,
+        idx: int,
+    ) -> Tuple[torch.Tensor, int]:
+        """
+        Returns a single sample and its index from the dataset.
+
+        Args:
+            idx (int):
+                The index of the sample to return.
+
+        Returns:
+            sample (torch.Tensor):
+                The requested sample from the dataset.
+        """
+        arr = np.take(self.X, idx, axis=self.axis)
+        if self.is_memmap:
+            arr = np.array(arr)
+        return torch.as_tensor(arr, dtype=self.dtype, device=self.device)
+    
+    def close(self):
+        """
+        Closes the numpy file.
+        """
+        if self.is_memmap:
+            self.X.close()
 
 
 class BatchRandomSampler(torch.utils.data.Sampler):
diff --git a/bnpm/video.py b/bnpm/video.py
index 6c55903..52e6a9c 100644
--- a/bnpm/video.py
+++ b/bnpm/video.py
@@ -7,7 +7,7 @@
 import torchvision
 import numpy as np
 import cv2
-from tqdm import tqdm
+from tqdm.auto import tqdm
 
 
 ###############################################################################