Improve truncation_side (#14947)

* Enabling `truncation_side` for Slow and Fast tokenizer.
Co-Authored-by: Niels Rogge <48327001+NielsRogge@users.noreply.github.com>

* Disable failing tests.

* Layout xlm.

* assert -> assertEqual.
Co-authored-by: Niels Rogge <48327001+NielsRogge@users.noreply.github.com>

src/transformers/tokenization_utils_base.py

@@ -1437,6 +1437,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
     # to make sure `tokenizer.pad(...)` works correctly
     model_input_names: List[str] = ["input_ids", "token_type_ids", "attention_mask"]
     padding_side: str = "right"
+    truncation_side: str = "right"
     slow_tokenizer_class = None
 
     def __init__(self, **kwargs):
@@ -1514,7 +1515,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
         return (
             f"{'PreTrainedTokenizerFast' if self.is_fast else 'PreTrainedTokenizer'}(name_or_path='{self.name_or_path}', "
             f"vocab_size={self.vocab_size}, model_max_len={self.model_max_length}, is_fast={self.is_fast}, "
-            f"padding_side='{self.padding_side}', special_tokens={self.special_tokens_map_extended})"
+            f"padding_side='{self.padding_side}', truncation_side='{self.truncation_side}', special_tokens={self.special_tokens_map_extended})"
         )
 
     def get_vocab(self) -> Dict[str, int]:
@@ -3041,8 +3042,15 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
         ):
             if len(ids) > num_tokens_to_remove:
                 window_len = min(len(ids), stride + num_tokens_to_remove)
-                overflowing_tokens = ids[-window_len:]
-                ids = ids[:-num_tokens_to_remove]
+                if self.truncation_side == "left":
+                    overflowing_tokens = ids[:window_len]
+                    ids = ids[num_tokens_to_remove:]
+                elif self.truncation_side == "right":
+                    overflowing_tokens = ids[-window_len:]
+                    ids = ids[:-num_tokens_to_remove]
+                else:
+                    raise ValueError(f"invalid truncation strategy: {self.truncation_side}, use 'left' or 'right'.")
+
             else:
                 error_msg = (
                     f"We need to remove {num_tokens_to_remove} to truncate the input "
@@ -3063,14 +3071,30 @@ class PreTrainedTokenizerBase(SpecialTokensMixin, PushToHubMixin):
             )
             for _ in range(num_tokens_to_remove):
                 if pair_ids is None or len(ids) > len(pair_ids):
-                    ids = ids[:-1]
+                    if self.truncation_side == "right":
+                        ids = ids[:-1]
+                    elif self.truncation_side == "left":
+                        ids = ids[1:]
+                    else:
+                        raise ValueError("invalid truncation strategy:" + str(self.truncation_side))
                 else:
-                    pair_ids = pair_ids[:-1]
+                    if self.truncation_side == "right":
+                        pair_ids = pair_ids[:-1]
+                    elif self.truncation_side == "left":
+                        pair_ids = pair_ids[1:]
+                    else:
+                        raise ValueError("invalid truncation strategy:" + str(self.truncation_side))
         elif truncation_strategy == TruncationStrategy.ONLY_SECOND and pair_ids is not None:
             if len(pair_ids) > num_tokens_to_remove:
                 window_len = min(len(pair_ids), stride + num_tokens_to_remove)
-                overflowing_tokens = pair_ids[-window_len:]
-                pair_ids = pair_ids[:-num_tokens_to_remove]
+                if self.truncation_side == "right":
+                    overflowing_tokens = pair_ids[-window_len:]
+                    pair_ids = pair_ids[:-num_tokens_to_remove]
+                elif self.truncation_side == "left":
+                    overflowing_tokens = pair_ids[:window_len]
+                    pair_ids = pair_ids[num_tokens_to_remove:]
+                else:
+                    raise ValueError("invalid truncation strategy:" + str(self.truncation_side))
             else:
                 logger.error(
                     f"We need to remove {num_tokens_to_remove} to truncate the input "

src/transformers/tokenization_utils_fast.py

@@ -356,6 +356,7 @@ class PreTrainedTokenizerFast(PreTrainedTokenizerBase):
                 "max_length": max_length,
                 "stride": stride,
                 "strategy": truncation_strategy.value,
+                "direction": self.truncation_side,
             }
 
             # _truncation might contain more keys that the target `transformers`

tests/test_tokenization_auto.py

@@ -209,6 +209,7 @@ class AutoTokenizerTest(unittest.TestCase):
         self.assertEqual(tokenizer.vocab_size, 30000)
         self.assertEqual(tokenizer.unk_token, "[UNK]")
         self.assertEqual(tokenizer.padding_side, "right")
+        self.assertEqual(tokenizer.truncation_side, "right")
 
     def test_auto_tokenizer_from_local_folder(self):
         tokenizer = AutoTokenizer.from_pretrained(SMALL_MODEL_IDENTIFIER)

tests/test_tokenization_common.py

@@ -1398,6 +1398,64 @@ class TokenizerTesterMixin:
                 assert sequence_length == padded_sequence_left_length
                 assert encoded_sequence == padded_sequence_left
 
+    def test_right_and_left_truncation(self):
+        tokenizers = self.get_tokenizers(do_lower_case=False)
+        for tokenizer in tokenizers:
+            with self.subTest(f"{tokenizer.__class__.__name__}"):
+                sequence = "This is a test sequence"
+
+                # RIGHT PADDING - Check that it correctly pads when a maximum length is specified along with the padding flag set to True
+                truncation_size = 3
+                tokenizer.truncation_side = "right"
+                encoded_sequence = tokenizer.encode(sequence, add_special_tokens=False)
+                sequence_length = len(encoded_sequence)
+                # Remove EOS/BOS tokens
+                truncated_sequence = tokenizer.encode(
+                    sequence, max_length=sequence_length - truncation_size, truncation=True, add_special_tokens=False
+                )
+                truncated_sequence_length = len(truncated_sequence)
+                self.assertEqual(sequence_length, truncated_sequence_length + truncation_size)
+                self.assertEqual(encoded_sequence[:-truncation_size], truncated_sequence)
+
+                # LEFT PADDING - Check that it correctly pads when a maximum length is specified along with the truncation flag set to True
+                tokenizer.truncation_side = "left"
+                sequence_length = len(encoded_sequence)
+                truncated_sequence = tokenizer.encode(
+                    sequence, max_length=sequence_length - truncation_size, truncation=True, add_special_tokens=False
+                )
+                truncated_sequence_length = len(truncated_sequence)
+                self.assertEqual(sequence_length, truncated_sequence_length + truncation_size)
+                self.assertEqual(encoded_sequence[truncation_size:], truncated_sequence)
+
+                # RIGHT & LEFT PADDING - Check that nothing is done for 'longest' and 'no_truncation'
+                sequence_length = len(encoded_sequence)
+
+                tokenizer.truncation_side = "right"
+                truncated_sequence_right = tokenizer.encode(sequence, truncation=True, add_special_tokens=False)
+                truncated_sequence_right_length = len(truncated_sequence_right)
+                self.assertEqual(sequence_length, truncated_sequence_right_length)
+                self.assertEqual(encoded_sequence, truncated_sequence_right)
+
+                tokenizer.truncation_side = "left"
+                truncated_sequence_left = tokenizer.encode(
+                    sequence, truncation="longest_first", add_special_tokens=False
+                )
+                truncated_sequence_left_length = len(truncated_sequence_left)
+                self.assertEqual(sequence_length, truncated_sequence_left_length)
+                self.assertEqual(encoded_sequence, truncated_sequence_left)
+
+                tokenizer.truncation_side = "right"
+                truncated_sequence_right = tokenizer.encode(sequence, add_special_tokens=False)
+                truncated_sequence_right_length = len(truncated_sequence_right)
+                self.assertEqual(sequence_length, truncated_sequence_right_length)
+                self.assertEqual(encoded_sequence, truncated_sequence_right)
+
+                tokenizer.truncation_side = "left"
+                truncated_sequence_left = tokenizer.encode(sequence, truncation=False, add_special_tokens=False)
+                truncated_sequence_left_length = len(truncated_sequence_left)
+                self.assertEqual(sequence_length, truncated_sequence_left_length)
+                self.assertEqual(encoded_sequence, truncated_sequence_left)
+
     def test_padding_to_max_length(self):
         """We keep this test for backward compatibility but it should be remove when `pad_to_max_length` is deprecated."""
         tokenizers = self.get_tokenizers(do_lower_case=False)

tests/test_tokenization_layoutlmv2.py

@@ -371,6 +371,10 @@ class LayoutLMv2TokenizationTest(TokenizerTesterMixin, unittest.TestCase):
                 decoded = tokenizer.decode(encoded, spaces_between_special_tokens=self.space_between_special_tokens)
                 self.assertIn(decoded, [output, output.lower()])
 
+    @unittest.skip("Not implemented")
+    def test_right_and_left_truncation(self):
+        pass
+
     def test_encode_plus_with_padding(self):
         tokenizers = self.get_tokenizers(do_lower_case=False)
         for tokenizer in tokenizers:

tests/test_tokenization_layoutxlm.py

@@ -939,6 +939,10 @@ class LayoutXLMTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
 
                 shutil.rmtree(tmpdirname)
 
+    @unittest.skip("Not implemented")
+    def test_right_and_left_truncation(self):
+        pass
+
     def test_right_and_left_padding(self):
         tokenizers = self.get_tokenizers(do_lower_case=False)
         for tokenizer in tokenizers:

tests/test_tokenization_tapas.py

@@ -904,6 +904,10 @@ class TapasTokenizationTest(TokenizerTesterMixin, unittest.TestCase):
 
                 shutil.rmtree(tmpdirname)
 
+    @unittest.skip("Not implemented")
+    def test_right_and_left_truncation(self):
+        pass
+
     def test_right_and_left_padding(self):
         tokenizers = self.get_tokenizers(do_lower_case=False)
         for tokenizer in tokenizers: