diff --git a/enhancer/data/dataset.py b/enhancer/data/dataset.py
index d2b7526..02a1d3b 100644
--- a/enhancer/data/dataset.py
+++ b/enhancer/data/dataset.py
@@ -5,7 +5,6 @@ from typing import Optional
 
 import pytorch_lightning as pl
 import torch.nn.functional as F
-from sklearn.model_selection import train_test_split
 from torch.utils.data import DataLoader, Dataset, IterableDataset
 
 from enhancer.data.fileprocessor import Fileprocessor
@@ -54,7 +53,7 @@ class TaskDataset(pl.LightningDataModule):
         name: str,
         root_dir: str,
         files: Files,
-        valid_size: float = 0.20,
+        valid_minutes: float = 0.20,
         duration: float = 1.0,
         sampling_rate: int = 48000,
         matching_function=None,
@@ -73,10 +72,10 @@ class TaskDataset(pl.LightningDataModule):
         if num_workers is None:
             num_workers = multiprocessing.cpu_count() // 2
         self.num_workers = num_workers
-        if valid_size > 0.0:
-            self.valid_size = valid_size
+        if valid_minutes > 0.0:
+            self.valid_minutes = valid_minutes
         else:
-            raise ValueError("valid_size must be greater than 0")
+            raise ValueError("valid_minutes must be greater than 0")
 
     def setup(self, stage: Optional[str] = None):
         """
@@ -91,8 +90,8 @@ class TaskDataset(pl.LightningDataModule):
                 self.name, train_clean, train_noisy, self.matching_function
             )
             train_data = fp.prepare_matching_dict()
-            self.train_data, self.val_data = train_test_split(
-                train_data, test_size=0.20, shuffle=True, random_state=42
+            self.train_data, self.val_data = self.train_valid_split(
+                train_data, valid_minutes=self.valid_minutes, random_state=42
             )
 
             self._validation = self.prepare_mapstype(self.val_data)
@@ -105,6 +104,28 @@ class TaskDataset(pl.LightningDataModule):
             test_data = fp.prepare_matching_dict()
             self._test = self.prepare_mapstype(test_data)
 
+    def train_valid_split(
+        self, data, valid_minutes: float = 20, random_state: int = 42
+    ):
+
+        valid_minutes *= 60
+        valid_min_now = 0.0
+        valid_indices = []
+        random_indices = list(range(0, len(data)))
+        rng = create_unique_rng(random_state)
+        rng.shuffle(random_indices)
+        i = 0
+        while valid_min_now <= valid_minutes:
+            valid_indices.append(random_indices[i])
+            valid_min_now += data[random_indices[i]]["duration"]
+            i += 1
+
+        train_data = [
+            item for i, item in enumerate(data) if i not in valid_indices
+        ]
+        valid_data = [item for i, item in enumerate(data) if i in valid_indices]
+        return train_data, valid_data
+
     def prepare_mapstype(self, data):
 
         metadata = []
@@ -172,7 +193,7 @@ class EnhancerDataset(TaskDataset):
         name: str,
         root_dir: str,
         files: Files,
-        valid_size=0.2,
+        valid_minutes=5.0,
         duration=1.0,
         sampling_rate=48000,
         matching_function=None,
@@ -184,7 +205,7 @@ class EnhancerDataset(TaskDataset):
             name=name,
             root_dir=root_dir,
             files=files,
-            valid_size=valid_size,
+            valid_minutes=valid_minutes,
             sampling_rate=sampling_rate,
             duration=duration,
             matching_function=matching_function,