Enhance DataCollatorForLanguageModeling with Configurable Token Replacement Probabilities (#35251)

* DataCollatorForLanguageModeling class was updated with new parameters that provides more control over the token masking and relacing

* DataCollatorForLanguageModeling class was updated with new parameters that provides more control over the token masking and relacing

* Addressed review comments, modified the docstring and made a test for the DataCollatorForLanguageModeling

src/transformers/data/data_collator.py

@@ -691,11 +691,17 @@ class DataCollatorForLanguageModeling(DataCollatorMixin):
             tokens and the value to predict for the masked token.
         mlm_probability (`float`, *optional*, defaults to 0.15):
             The probability with which to (randomly) mask tokens in the input, when `mlm` is set to `True`.
+        mask_replace_prob (`float`, *optional*, defaults to 0.8):
+            The probability with which masked tokens are replaced by the tokenizer's mask token (e.g., `[MASK]`).
+            Defaults to 0.8, meaning 80% of the masked tokens will be replaced with `[MASK]`.
+            Only works when `mlm` is set to `True`.
+        random_replace_prob (`float`, *optional*, defaults to 0.1):
+            The probability with which masked tokens are replaced by random tokens from the tokenizer's vocabulary.
+            Defaults to 0.1, meaning 10% of the masked tokens will be replaced with random tokens. The remaining
+            masked tokens (1 - mask_replace_prob - random_replace_prob) are left unchanged.
+            Only works when `mlm` is set to `True`.
         pad_to_multiple_of (`int`, *optional*):
-            If set will pad the sequence to a multiple of the provided value.
-
-            This is especially useful to enable the use of Tensor Cores on NVIDIA hardware with compute capability >=
-            7.0 (Volta).
+            If set, will pad the sequence to a multiple of the provided value.
         return_tensors (`str`):
             The type of Tensor to return. Allowable values are "np", "pt" and "tf".
 
@@ -705,11 +711,36 @@ class DataCollatorForLanguageModeling(DataCollatorMixin):
     BatchEncoding, with the `"special_tokens_mask"` key, as returned by a [`PreTrainedTokenizer`] or a
     [`PreTrainedTokenizerFast`] with the argument `return_special_tokens_mask=True`.
 
-    </Tip>"""
+    <Example Options and Expectations>
+
+    1. Default Behavior:
+        - `mask_replace_prob=0.8`, `random_replace_prob=0.1`.
+        - Expect 80% of masked tokens replaced with `[MASK]`, 10% replaced with random tokens, and 10% left unchanged.
+
+    2. All masked tokens replaced by `[MASK]`:
+        - `mask_replace_prob=1.0`, `random_replace_prob=0.0`.
+        - Expect all masked tokens to be replaced with `[MASK]`. No tokens are left unchanged or replaced with random tokens.
+
+    3. No `[MASK]` replacement, only random tokens:
+        - `mask_replace_prob=0.0`, `random_replace_prob=1.0`.
+        - Expect all masked tokens to be replaced with random tokens. No `[MASK]` replacements or unchanged tokens.
+
+    4. Balanced replacement:
+        - `mask_replace_prob=0.5`, `random_replace_prob=0.4`.
+        - Expect 50% of masked tokens replaced with `[MASK]`, 40% replaced with random tokens, and 10% left unchanged.
+
+    Note:
+        The sum of `mask_replace_prob` and `random_replace_prob` must not exceed 1. If their sum is less than 1, the
+        remaining proportion will consist of masked tokens left unchanged.
+
+    </Tip>
+    """
 
     tokenizer: PreTrainedTokenizerBase
     mlm: bool = True
     mlm_probability: float = 0.15
+    mask_replace_prob: float = 0.8
+    random_replace_prob: float = 0.1
     pad_to_multiple_of: Optional[int] = None
     tf_experimental_compile: bool = False
     return_tensors: str = "pt"
@@ -720,6 +751,15 @@ class DataCollatorForLanguageModeling(DataCollatorMixin):
                 "This tokenizer does not have a mask token which is necessary for masked language modeling. "
                 "You should pass `mlm=False` to train on causal language modeling instead."
             )
+        if self.mlm_probability < 0 or self.mlm_probability > 1:
+            raise ValueError("mlm_probability should be between 0 and 1.")
+        if self.mask_replace_prob + self.random_replace_prob > 1:
+            raise ValueError("The sum of mask_replace_prob and random_replace_prob should not exceed 1")
+        if self.mask_replace_prob < 0 or self.mask_replace_prob > 1:
+            raise ValueError("mask_replace_prob should be between 0 and 1.")
+        if self.random_replace_prob < 0 or self.random_replace_prob > 1:
+            raise ValueError("random_replace_prob should be between 0 and 1.")
+
         if self.tf_experimental_compile:
             import tensorflow as tf
 
@@ -749,18 +789,28 @@ class DataCollatorForLanguageModeling(DataCollatorMixin):
         # Replace unmasked indices with -100 in the labels since we only compute loss on masked tokens
         labels = tf.where(masked_indices, inputs, -100)
 
-        # 80% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
-        indices_replaced = self.tf_bernoulli(input_shape, 0.8) & masked_indices
+        # mask_replace_prob% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
+        indices_replaced = self.tf_bernoulli(input_shape, self.mask_replace_prob) & masked_indices
 
         inputs = tf.where(indices_replaced, mask_token_id, inputs)
 
-        # 10% of the time, we replace masked input tokens with random word
-        indices_random = self.tf_bernoulli(input_shape, 0.5) & masked_indices & ~indices_replaced
+        if self.mask_replace_prob == 1 or self.random_replace_prob == 0:
+            return inputs, labels
+
+        remaining_prob = 1 - self.mask_replace_prob
+        # scaling the random_replace_prob to the remaining probability for example if
+        # mask_replace_prob = 0.8 and random_replace_prob = 0.1,
+        # then random_replace_prob_scaled = 0.1 / 0.2 = 0.5
+        random_replace_prob_scaled = self.random_replace_prob / remaining_prob
+        # random_replace_prob% of the time, we replace masked input tokens with random word
+        indices_random = (
+            self.tf_bernoulli(input_shape, random_replace_prob_scaled) & masked_indices & ~indices_replaced
+        )
         random_words = tf.random.uniform(input_shape, maxval=vocab_size, dtype=inputs.dtype)
 
         inputs = tf.where(indices_random, random_words, inputs)
 
-        # The rest of the time (10% of the time) we keep the masked input tokens unchanged
+        # The rest of the time ((1-random_replace_prob-mask_replace_prob)% of the time) we keep the masked input tokens unchanged
         return inputs, labels
 
     def tf_call(self, examples: List[Union[List[int], Any, Dict[str, Any]]]) -> Dict[str, Any]:
@@ -849,16 +899,29 @@ class DataCollatorForLanguageModeling(DataCollatorMixin):
         masked_indices = torch.bernoulli(probability_matrix).bool()
         labels[~masked_indices] = -100  # We only compute loss on masked tokens
 
-        # 80% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
-        indices_replaced = torch.bernoulli(torch.full(labels.shape, 0.8)).bool() & masked_indices
+        # mask_replace_prob% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
+        indices_replaced = torch.bernoulli(torch.full(labels.shape, self.mask_replace_prob)).bool() & masked_indices
         inputs[indices_replaced] = self.tokenizer.convert_tokens_to_ids(self.tokenizer.mask_token)
 
-        # 10% of the time, we replace masked input tokens with random word
-        indices_random = torch.bernoulli(torch.full(labels.shape, 0.5)).bool() & masked_indices & ~indices_replaced
+        if self.mask_replace_prob == 1 or self.random_replace_prob == 0:
+            return inputs, labels
+
+        remaining_prob = 1 - self.mask_replace_prob
+        # scaling the random_replace_prob to the remaining probability for example if
+        # mask_replace_prob = 0.8 and random_replace_prob = 0.1,
+        # then random_replace_prob_scaled = 0.1 / 0.2 = 0.5
+        random_replace_prob_scaled = self.random_replace_prob / remaining_prob
+
+        # random_replace_prob% of the time, we replace masked input tokens with random word
+        indices_random = (
+            torch.bernoulli(torch.full(labels.shape, random_replace_prob_scaled)).bool()
+            & masked_indices
+            & ~indices_replaced
+        )
         random_words = torch.randint(len(self.tokenizer), labels.shape, dtype=torch.long)
         inputs[indices_random] = random_words[indices_random]
 
-        # The rest of the time (10% of the time) we keep the masked input tokens unchanged
+        # The rest of the time ((1-random_replace_prob-mask_replace_prob)% of the time) we keep the masked input tokens unchanged
         return inputs, labels
 
     def numpy_call(self, examples: List[Union[List[int], Any, Dict[str, Any]]]) -> Dict[str, Any]:
@@ -905,14 +968,24 @@ class DataCollatorForLanguageModeling(DataCollatorMixin):
         masked_indices = np.random.binomial(1, probability_matrix, size=probability_matrix.shape).astype(bool)
         labels[~masked_indices] = -100  # We only compute loss on masked tokens
 
-        # 80% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
-        indices_replaced = np.random.binomial(1, 0.8, size=labels.shape).astype(bool) & masked_indices
+        # mask_replace_prob% of the time, we replace masked input tokens with tokenizer.mask_token ([MASK])
+        indices_replaced = (
+            np.random.binomial(1, self.mask_replace_prob, size=labels.shape).astype(bool) & masked_indices
+        )
         inputs[indices_replaced] = self.tokenizer.mask_token_id
 
-        # 10% of the time, we replace masked input tokens with random word
-        # indices_random = torch.bernoulli(torch.full(labels.shape, 0.5)).bool() & masked_indices & ~indices_replaced
+        if self.mask_replace_prob == 1 or self.random_replace_prob == 0:
+            return inputs, labels
+
+        remaining_prob = 1 - self.mask_replace_prob
+        # scaling the random_replace_prob to the remaining probability for example if
+        # mask_replace_prob = 0.8 and random_replace_prob = 0.1,
+        # then random_replace_prob_scaled = 0.1 / 0.2 = 0.5
+        random_replace_prob_scaled = self.random_replace_prob / remaining_prob
         indices_random = (
-            np.random.binomial(1, 0.5, size=labels.shape).astype(bool) & masked_indices & ~indices_replaced
+            np.random.binomial(1, random_replace_prob_scaled, size=labels.shape).astype(bool)
+            & masked_indices
+            & ~indices_replaced
         )
         random_words = np.random.randint(
             low=0, high=len(self.tokenizer), size=np.count_nonzero(indices_random), dtype=np.int64

tests/trainer/test_data_collator.py

@@ -1020,6 +1020,52 @@ class TFDataCollatorIntegrationTest(unittest.TestCase):
         self.assertTrue(tf.reduce_any(masked_tokens))
         # self.assertTrue(all(x == -100 for x in batch["labels"].numpy()[~masked_tokens.numpy()].tolist()))
 
+    def test_probability_sum_error(self):
+        """Test that the sum of mask_replace_prob and random_replace_prob exceeding 1 raises an error."""
+        tokenizer = BertTokenizer(self.vocab_file)
+        with self.assertRaises(ValueError):
+            DataCollatorForLanguageModeling(tokenizer=tokenizer, mask_replace_prob=0.9, random_replace_prob=0.2)
+
+    def test_all_mask_replacement(self):
+        """Test behavior when mask_replace_prob=1."""
+        tokenizer = BertTokenizer(self.vocab_file)
+
+        # pytorch call
+        collator = DataCollatorForLanguageModeling(
+            tokenizer=tokenizer, mask_replace_prob=1, random_replace_prob=0, return_tensors="pt"
+        )
+
+        inputs = torch.tensor([0, 1, 2, 3, 4, 5])
+        features = [{"input_ids": inputs} for _ in range(8)]
+        batch = collator(features)
+
+        # confirm that every token is either the original token or [MASK]
+        self.assertTrue(torch.all((batch["input_ids"] == inputs) | (batch["input_ids"] == tokenizer.mask_token_id)))
+
+        # tf call
+        collator = DataCollatorForLanguageModeling(
+            tokenizer=tokenizer, mask_replace_prob=1, random_replace_prob=0, return_tensors="tf"
+        )
+        inputs = tf.constant([0, 1, 2, 3, 4, 5])
+        features = [{"input_ids": inputs} for _ in range(8)]
+        batch = collator(features)
+
+        # confirm that every token is either the original token or [MASK]
+        self.assertTrue(
+            tf.reduce_all((batch["input_ids"] == inputs) | (batch["input_ids"] == tokenizer.mask_token_id))
+        )
+
+        # numpy call
+        collator = DataCollatorForLanguageModeling(
+            tokenizer=tokenizer, mask_replace_prob=1, random_replace_prob=0, return_tensors="np"
+        )
+        inputs = np.array([0, 1, 2, 3, 4, 5])
+        features = [{"input_ids": inputs} for _ in range(8)]
+        batch = collator(features)
+
+        # confirm that every token is either the original token or [MASK]
+        self.assertTrue(np.all((batch["input_ids"] == inputs) | (batch["input_ids"] == tokenizer.mask_token_id)))
+
     def test_data_collator_for_language_modeling(self):
         no_pad_features = [{"input_ids": list(range(10))}, {"input_ids": list(range(10))}]
         pad_features = [{"input_ids": list(range(5))}, {"input_ids": list(range(10))}]