fix to ensure that returned tensors after the tokenization is Long (#7039)

* fix to ensure that returned tensors after the tokenization is Long

* fix to ensure that returned tensors after the tokenization is Long
Co-authored-by: Ashwin Geet Dsa <adsa@grvingt-6.nancy.grid5000.fr>

src/transformers/data/data_collator.py

@@ -149,7 +149,7 @@ class DataCollatorForLanguageModeling:
     ) -> torch.Tensor:
         # In order to accept both lists of lists and lists of Tensors
         if isinstance(examples[0], (list, tuple)):
-            examples = [torch.Tensor(e) for e in examples]
+            examples = [torch.tensor(e, dtype=torch.long) for e in examples]
         length_of_first = examples[0].size(0)
         are_tensors_same_length = all(x.size(0) == length_of_first for x in examples)
         if are_tensors_same_length: