[integration] Add support for Transformers v4.46.0 (#3026)

sentence_transformers/trainer.py

@@ -7,9 +7,11 @@
 from typing import TYPE_CHECKING, Any, Callable
 
 import torch
+from packaging.version import parse as parse_version
 from torch import nn
 from torch.utils.data import BatchSampler, ConcatDataset, DataLoader, SubsetRandomSampler
 from transformers import EvalPrediction, PreTrainedTokenizerBase, Trainer, TrainerCallback
+from transformers import __version__ as transformers_version
 from transformers.data.data_collator import DataCollator
 from transformers.integrations import WandbCallback
 from transformers.trainer import TRAINING_ARGS_NAME
@@ -202,19 +204,24 @@ def __init__(
             train_dataset = DatasetDict(train_dataset)
         if isinstance(eval_dataset, dict) and not isinstance(eval_dataset, DatasetDict):
             eval_dataset = DatasetDict(eval_dataset)
-        super().__init__(
-            model=None if self.model_init else model,
-            args=args,
-            data_collator=data_collator,
-            train_dataset=train_dataset,
-            eval_dataset=eval_dataset,
-            tokenizer=tokenizer,
-            model_init=model_init,
-            compute_metrics=compute_metrics,
-            callbacks=callbacks,
-            optimizers=optimizers,
-            preprocess_logits_for_metrics=preprocess_logits_for_metrics,
-        )
+        super_kwargs = {
+            "model": None if self.model_init else model,
+            "args": args,
+            "data_collator": data_collator,
+            "train_dataset": train_dataset,
+            "eval_dataset": eval_dataset,
+            "model_init": model_init,
+            "compute_metrics": compute_metrics,
+            "callbacks": callbacks,
+            "optimizers": optimizers,
+            "preprocess_logits_for_metrics": preprocess_logits_for_metrics,
+        }
+        # Transformers v4.46.0 changed the `tokenizer` argument to a more general `processing_class` argument
+        if parse_version(transformers_version) >= parse_version("4.46.0"):
+            super_kwargs["processing_class"] = tokenizer
+        else:
+            super_kwargs["tokenizer"] = tokenizer
+        super().__init__(**super_kwargs)
         # Every Sentence Transformer model can always return a loss, so we set this to True
         # to avoid having to specify it in the data collator or model's forward
         self.can_return_loss = True
@@ -311,6 +318,7 @@ def compute_loss(
         model: SentenceTransformer,
         inputs: dict[str, torch.Tensor | Any],
         return_outputs: bool = False,
+        num_items_in_batch=None,
     ) -> torch.Tensor | tuple[torch.Tensor, dict[str, Any]]:
         """
         Computes the loss for the SentenceTransformer model.
@@ -325,6 +333,7 @@ def compute_loss(
             model (SentenceTransformer): The SentenceTransformer model.
             inputs (Dict[str, Union[torch.Tensor, Any]]): The input data for the model.
             return_outputs (bool, optional): Whether to return the outputs along with the loss. Defaults to False.
+            num_items_in_batch (int, optional): The number of items in the batch. Defaults to None. Unused, but required by the transformers Trainer.
 
         Returns:
             Union[torch.Tensor, Tuple[torch.Tensor, Dict[str, Any]]]: The computed loss. If `return_outputs` is True, returns a tuple of loss and outputs. Otherwise, returns only the loss.