diff --git a/references/classification/train.py b/references/classification/train.py
index 9b1994bad57..7addad30350 100644
--- a/references/classification/train.py
+++ b/references/classification/train.py
@@ -57,7 +57,7 @@ def evaluate(model, criterion, data_loader, device, print_freq=100, log_suffix="
     header = f"Test: {log_suffix}"
 
     num_processed_samples = 0
-    with torch.no_grad():
+    with torch.inference_mode():
         for image, target in metric_logger.log_every(data_loader, print_freq, header):
             image = image.to(device, non_blocking=True)
             target = target.to(device, non_blocking=True)
diff --git a/references/classification/train_quantization.py b/references/classification/train_quantization.py
index 5bf64aea721..ae4e81b0133 100644
--- a/references/classification/train_quantization.py
+++ b/references/classification/train_quantization.py
@@ -112,7 +112,7 @@ def main(args):
         print("Starting training for epoch", epoch)
         train_one_epoch(model, criterion, optimizer, data_loader, device, epoch, args.print_freq)
         lr_scheduler.step()
-        with torch.no_grad():
+        with torch.inference_mode():
             if epoch >= args.num_observer_update_epochs:
                 print("Disabling observer for subseq epochs, epoch = ", epoch)
                 model.apply(torch.quantization.disable_observer)
diff --git a/references/classification/utils.py b/references/classification/utils.py
index c186a60fc1e..9e043582d13 100644
--- a/references/classification/utils.py
+++ b/references/classification/utils.py
@@ -181,7 +181,7 @@ def update_parameters(self, model):
 
 def accuracy(output, target, topk=(1,)):
     """Computes the accuracy over the k top predictions for the specified values of k"""
-    with torch.no_grad():
+    with torch.inference_mode():
         maxk = max(topk)
         batch_size = target.size(0)
         if target.ndim == 2:
diff --git a/references/detection/engine.py b/references/detection/engine.py
index 2ca7df808ef..f07894dd99f 100644
--- a/references/detection/engine.py
+++ b/references/detection/engine.py
@@ -68,7 +68,7 @@ def _get_iou_types(model):
     return iou_types
 
 
-@torch.no_grad()
+@torch.inference_mode()
 def evaluate(model, data_loader, device):
     n_threads = torch.get_num_threads()
     # FIXME remove this and make paste_masks_in_image run on the GPU
diff --git a/references/detection/utils.py b/references/detection/utils.py
index c708ca05413..c3413ba828a 100644
--- a/references/detection/utils.py
+++ b/references/detection/utils.py
@@ -95,7 +95,7 @@ def reduce_dict(input_dict, average=True):
     world_size = get_world_size()
     if world_size < 2:
         return input_dict
-    with torch.no_grad():
+    with torch.inference_mode():
         names = []
         values = []
         # sort the keys so that they are consistent across processes
diff --git a/references/segmentation/train.py b/references/segmentation/train.py
index 3a41f86ba87..73c49b6c543 100644
--- a/references/segmentation/train.py
+++ b/references/segmentation/train.py
@@ -49,7 +49,7 @@ def evaluate(model, data_loader, device, num_classes):
     confmat = utils.ConfusionMatrix(num_classes)
     metric_logger = utils.MetricLogger(delimiter="  ")
     header = "Test:"
-    with torch.no_grad():
+    with torch.inference_mode():
         for image, target in metric_logger.log_every(data_loader, 100, header):
             image, target = image.to(device), target.to(device)
             output = model(image)
diff --git a/references/segmentation/utils.py b/references/segmentation/utils.py
index 2bb5451289a..f2d43340f8e 100644
--- a/references/segmentation/utils.py
+++ b/references/segmentation/utils.py
@@ -76,7 +76,7 @@ def update(self, a, b):
         n = self.num_classes
         if self.mat is None:
             self.mat = torch.zeros((n, n), dtype=torch.int64, device=a.device)
-        with torch.no_grad():
+        with torch.inference_mode():
             k = (a >= 0) & (a < n)
             inds = n * a[k].to(torch.int64) + b[k]
             self.mat += torch.bincount(inds, minlength=n ** 2).reshape(n, n)
diff --git a/references/similarity/train.py b/references/similarity/train.py
index c8f041acdad..4eab5e27188 100644
--- a/references/similarity/train.py
+++ b/references/similarity/train.py
@@ -51,7 +51,7 @@ def find_best_threshold(dists, targets, device):
     return best_thresh, accuracy
 
 
-@torch.no_grad()
+@torch.inference_mode()
 def evaluate(model, loader, device):
     model.eval()
     embeds, labels = [], []
diff --git a/references/video_classification/train.py b/references/video_classification/train.py
index f944cff7794..86c0daf738f 100644
--- a/references/video_classification/train.py
+++ b/references/video_classification/train.py
@@ -52,7 +52,7 @@ def evaluate(model, criterion, data_loader, device):
     model.eval()
     metric_logger = utils.MetricLogger(delimiter="  ")
     header = "Test:"
-    with torch.no_grad():
+    with torch.inference_mode():
         for video, target in metric_logger.log_every(data_loader, 100, header):
             video = video.to(device, non_blocking=True)
             target = target.to(device, non_blocking=True)
diff --git a/references/video_classification/utils.py b/references/video_classification/utils.py
index 956c4f85239..31ba2ff489c 100644
--- a/references/video_classification/utils.py
+++ b/references/video_classification/utils.py
@@ -159,7 +159,7 @@ def log_every(self, iterable, print_freq, header=None):
 
 def accuracy(output, target, topk=(1,)):
     """Computes the accuracy over the k top predictions for the specified values of k"""
-    with torch.no_grad():
+    with torch.inference_mode():
         maxk = max(topk)
         batch_size = target.size(0)