From bfdae2cd990c3eed8956f3430c6a8c81f9f20304 Mon Sep 17 00:00:00 2001
From: Paul-Edouard Sarlin <15985472+sarlinpe@users.noreply.github.com>
Date: Tue, 23 Jan 2024 13:46:43 +0100
Subject: [PATCH] Drop last in distributed mode (#55)

---
 gluefactory/datasets/base_dataset.py | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/gluefactory/datasets/base_dataset.py b/gluefactory/datasets/base_dataset.py
index ef622cbc..b3114c99 100644
--- a/gluefactory/datasets/base_dataset.py
+++ b/gluefactory/datasets/base_dataset.py
@@ -161,9 +161,12 @@ def get_data_loader(self, split, shuffle=None, pinned=False, distributed=False):
         except omegaconf.MissingMandatoryValue:
             batch_size = self.conf.batch_size
         num_workers = self.conf.get("num_workers", batch_size)
+        drop_last = True if split == "train" else False
         if distributed:
             shuffle = False
-            sampler = torch.utils.data.distributed.DistributedSampler(dataset)
+            sampler = torch.utils.data.distributed.DistributedSampler(
+                dataset, drop_last=drop_last
+            )
         else:
             sampler = None
             if shuffle is None:
@@ -178,7 +181,7 @@ def get_data_loader(self, split, shuffle=None, pinned=False, distributed=False):
             num_workers=num_workers,
             worker_init_fn=worker_init_fn,
             prefetch_factor=self.conf.prefetch_factor,
-            drop_last=True if split == "train" else False,
+            drop_last=drop_last,
         )
 
     def get_overfit_loader(self, split):