Switch `return_dict` to `True` by default. (#8530)

* Use the CI to identify failing tests

* Remove from all examples and tests

* More default switch

* Fixes

* More test fixes

* More fixes

* Last fixes hopefully

* Use the CI to identify failing tests

* Remove from all examples and tests

* More default switch

* Fixes

* More test fixes

* More fixes

* Last fixes hopefully

* Run on the real suite

* Fix slow tests

docs/source/model_doc/bertgeneration.rst

@@ -40,7 +40,7 @@ Usage:
   labels = tokenizer('This is a short summary', return_tensors="pt").input_ids
 
   # train...
-  loss = bert2bert(input_ids=input_ids, decoder_input_ids=labels, labels=labels, return_dict=True).loss
+  loss = bert2bert(input_ids=input_ids, decoder_input_ids=labels, labels=labels).loss
   loss.backward()
 
 

docs/source/model_doc/t5.rst

@@ -64,7 +64,7 @@ token. T5 can be trained / fine-tuned both in a supervised and unsupervised fash
   input_ids = tokenizer('The <extra_id_0> walks in <extra_id_1> park', return_tensors='pt').input_ids
   labels = tokenizer('<extra_id_0> cute dog <extra_id_1> the <extra_id_2>', return_tensors='pt').input_ids
   # the forward function automatically creates the correct decoder_input_ids
-  loss = model(input_ids=input_ids, labels=labels, return_dict=True).loss
+  loss = model(input_ids=input_ids, labels=labels).loss
 
 - Supervised training
 
@@ -77,7 +77,7 @@ token. T5 can be trained / fine-tuned both in a supervised and unsupervised fash
   input_ids = tokenizer('translate English to German: The house is wonderful.', return_tensors='pt').input_ids
   labels = tokenizer('Das Haus ist wunderbar.', return_tensors='pt').input_ids
   # the forward function automatically creates the correct decoder_input_ids
-  loss = model(input_ids=input_ids, labels=labels, return_dict=True).loss
+  loss = model(input_ids=input_ids, labels=labels).loss
 
 
 T5Config

docs/source/task_summary.rst

@@ -89,7 +89,7 @@ each other. The process is the following:
     >>> import torch
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased-finetuned-mrpc")
-    >>> model = AutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc", return_dict=True)
+    >>> model = AutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc")
 
     >>> classes = ["not paraphrase", "is paraphrase"]
 
@@ -122,7 +122,7 @@ each other. The process is the following:
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased-finetuned-mrpc")
-    >>> model = TFAutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc", return_dict=True)
+    >>> model = TFAutoModelForSequenceClassification.from_pretrained("bert-base-cased-finetuned-mrpc")
 
     >>> classes = ["not paraphrase", "is paraphrase"]
 
@@ -211,7 +211,7 @@ Here is an example of question answering using a model and a tokenizer. The proc
     >>> import torch
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
-    >>> model = AutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad", return_dict=True)
+    >>> model = AutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
 
     >>> text = r"""
     ... 🤗 Transformers (formerly known as pytorch-transformers and pytorch-pretrained-bert) provides general-purpose
@@ -253,7 +253,7 @@ Here is an example of question answering using a model and a tokenizer. The proc
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
-    >>> model = TFAutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad", return_dict=True)
+    >>> model = TFAutoModelForQuestionAnswering.from_pretrained("bert-large-uncased-whole-word-masking-finetuned-squad")
 
     >>> text = r"""
     ... 🤗 Transformers (formerly known as pytorch-transformers and pytorch-pretrained-bert) provides general-purpose
@@ -373,7 +373,7 @@ Here is an example of doing masked language modeling using a model and a tokeniz
     >>> import torch
 
     >>> tokenizer = AutoTokenizer.from_pretrained("distilbert-base-cased")
-    >>> model = AutoModelWithLMHead.from_pretrained("distilbert-base-cased", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("distilbert-base-cased")
 
     >>> sequence = f"Distilled models are smaller than the models they mimic. Using them instead of the large versions would help {tokenizer.mask_token} our carbon footprint."
 
@@ -389,7 +389,7 @@ Here is an example of doing masked language modeling using a model and a tokeniz
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("distilbert-base-cased")
-    >>> model = TFAutoModelWithLMHead.from_pretrained("distilbert-base-cased", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("distilbert-base-cased")
 
     >>> sequence = f"Distilled models are smaller than the models they mimic. Using them instead of the large versions would help {tokenizer.mask_token} our carbon footprint."
 
@@ -437,7 +437,7 @@ of tokens.
     >>> from torch.nn import functional as F
 
     >>> tokenizer = AutoTokenizer.from_pretrained("gpt2")
-    >>> model = AutoModelWithLMHead.from_pretrained("gpt2", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("gpt2")
 
     >>> sequence = f"Hugging Face is based in DUMBO, New York City, and "
 
@@ -461,7 +461,7 @@ of tokens.
     >>> import tensorflow as tf
 
     >>> tokenizer = AutoTokenizer.from_pretrained("gpt2")
-    >>> model = TFAutoModelWithLMHead.from_pretrained("gpt2", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("gpt2")
 
     >>> sequence = f"Hugging Face is based in DUMBO, New York City, and "
 
@@ -520,7 +520,7 @@ Here is an example of text generation using ``XLNet`` and its tokenizer.
     >>> ## PYTORCH CODE
     >>> from transformers import AutoModelWithLMHead, AutoTokenizer
 
-    >>> model = AutoModelWithLMHead.from_pretrained("xlnet-base-cased", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("xlnet-base-cased")
     >>> tokenizer = AutoTokenizer.from_pretrained("xlnet-base-cased")
 
     >>> # Padding text helps XLNet with short prompts - proposed by Aman Rusia in https://github.com/rusiaaman/XLNet-gen#methodology
@@ -545,7 +545,7 @@ Here is an example of text generation using ``XLNet`` and its tokenizer.
     >>> ## TENSORFLOW CODE
     >>> from transformers import TFAutoModelWithLMHead, AutoTokenizer
 
-    >>> model = TFAutoModelWithLMHead.from_pretrained("xlnet-base-cased", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("xlnet-base-cased")
     >>> tokenizer = AutoTokenizer.from_pretrained("xlnet-base-cased")
 
     >>> # Padding text helps XLNet with short prompts - proposed by Aman Rusia in https://github.com/rusiaaman/XLNet-gen#methodology
@@ -664,7 +664,7 @@ Here is an example of doing named entity recognition, using a model and a tokeni
     >>> from transformers import AutoModelForTokenClassification, AutoTokenizer
     >>> import torch
 
-    >>> model = AutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english", return_dict=True)
+    >>> model = AutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english")
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
 
     >>> label_list = [
@@ -692,7 +692,7 @@ Here is an example of doing named entity recognition, using a model and a tokeni
     >>> from transformers import TFAutoModelForTokenClassification, AutoTokenizer
     >>> import tensorflow as tf
 
-    >>> model = TFAutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english", return_dict=True)
+    >>> model = TFAutoModelForTokenClassification.from_pretrained("dbmdz/bert-large-cased-finetuned-conll03-english")
     >>> tokenizer = AutoTokenizer.from_pretrained("bert-base-cased")
 
     >>> label_list = [
@@ -790,7 +790,7 @@ CNN / Daily Mail), it yields very good results.
     >>> ## PYTORCH CODE
     >>> from transformers import AutoModelWithLMHead, AutoTokenizer
 
-    >>> model = AutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> # T5 uses a max_length of 512 so we cut the article to 512 tokens.
@@ -799,7 +799,7 @@ CNN / Daily Mail), it yields very good results.
     >>> ## TENSORFLOW CODE
     >>> from transformers import TFAutoModelWithLMHead, AutoTokenizer
 
-    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> # T5 uses a max_length of 512 so we cut the article to 512 tokens.
@@ -843,7 +843,7 @@ Here is an example of doing translation using a model and a tokenizer. The proce
     >>> ## PYTORCH CODE
     >>> from transformers import AutoModelWithLMHead, AutoTokenizer
 
-    >>> model = AutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = AutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> inputs = tokenizer.encode("translate English to German: Hugging Face is a technology company based in New York and Paris", return_tensors="pt")
@@ -851,7 +851,7 @@ Here is an example of doing translation using a model and a tokenizer. The proce
     >>> ## TENSORFLOW CODE
     >>> from transformers import TFAutoModelWithLMHead, AutoTokenizer
 
-    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base", return_dict=True)
+    >>> model = TFAutoModelWithLMHead.from_pretrained("t5-base")
     >>> tokenizer = AutoTokenizer.from_pretrained("t5-base")
 
     >>> inputs = tokenizer.encode("translate English to German: Hugging Face is a technology company based in New York and Paris", return_tensors="tf")

docs/source/training.rst

@@ -39,7 +39,7 @@ head on top of the encoder with an output size of 2. Models are initialized in `
 .. code-block:: python
 
     from transformers import BertForSequenceClassification
-    model = BertForSequenceClassification.from_pretrained('bert-base-uncased', return_dict=True)
+    model = BertForSequenceClassification.from_pretrained('bert-base-uncased')
     model.train()
 
 This is useful because it allows us to make use of the pre-trained BERT encoder and easily train it on whatever

examples/lxmert/demo.ipynb

@@ -210,7 +210,6 @@
     "        visual_feats=features,\n",
     "        visual_pos=normalized_boxes,\n",
     "        token_type_ids=inputs.token_type_ids,\n",
-    "        return_dict=True,\n",
     "        output_attentions=False,\n",
     "    )\n",
     "    output_vqa = lxmert_vqa(\n",
@@ -219,7 +218,6 @@
     "        visual_feats=features,\n",
     "        visual_pos=normalized_boxes,\n",
     "        token_type_ids=inputs.token_type_ids,\n",
-    "        return_dict=True,\n",
     "        output_attentions=False,\n",
     "    )\n",
     "    # get prediction\n",

examples/question-answering/run_squad.py

@@ -321,7 +321,7 @@ def evaluate(args, model, tokenizer, prefix=""):
             eval_feature = features[feature_index.item()]
             unique_id = int(eval_feature.unique_id)
 
-            output = [to_list(output[i]) for output in outputs]
+            output = [to_list(output[i]) for output in outputs.to_tuple()]
 
             # Some models (XLNet, XLM) use 5 arguments for their predictions, while the other "simpler"
             # models only use two.

examples/rag/eval_rag.py

@@ -95,7 +95,7 @@ def evaluate_batch_retrieval(args, rag_model, questions):
         truncation=True,
     )["input_ids"].to(args.device)
 
-    question_enc_outputs = rag_model.rag.question_encoder(retriever_input_ids, return_dict=True)
+    question_enc_outputs = rag_model.rag.question_encoder(retriever_input_ids)
     question_enc_pool_output = question_enc_outputs.pooler_output
 
     result = rag_model.retriever(

examples/rag/finetune.py

@@ -204,7 +204,6 @@ class GenerativeQAModule(BaseTransformer):
             decoder_input_ids=decoder_input_ids,
             use_cache=False,
             labels=lm_labels,
-            return_dict=True,
             **rag_kwargs,
         )
 

examples/rag/use_own_knowledge_dataset.py

@@ -47,7 +47,7 @@ def embed(documents: dict, ctx_encoder: DPRContextEncoder, ctx_tokenizer: DPRCon
     input_ids = ctx_tokenizer(
         documents["title"], documents["text"], truncation=True, padding="longest", return_tensors="pt"
     )["input_ids"]
-    embeddings = ctx_encoder(input_ids.to(device=device), return_dict=True).pooler_output
+    embeddings = ctx_encoder(input_ids.to(device=device)).pooler_output
     return {"embeddings": embeddings.detach().cpu().numpy()}
 
 

examples/seq2seq/distillation.py

@@ -153,7 +153,6 @@ class SummarizationDistiller(SummarizationModule):
             output_hidden_states=self.do_calc_hidden_loss,
             output_attentions=False,
             use_cache=False,
-            return_dict=True,
         )
         lm_logits = student_outputs.logits
 
@@ -179,7 +178,6 @@ class SummarizationDistiller(SummarizationModule):
                 input_ids,
                 attention_mask=src_mask,
                 output_hidden_states=self.do_calc_hidden_loss,
-                return_dict=True,
             )
             if self.different_base_models:
                 teacher_enc_outputs = all_teacher_encoder_outputs.last_hidden_state
@@ -199,7 +197,6 @@ class SummarizationDistiller(SummarizationModule):
             decoder_input_ids=decoder_input_ids,
             output_hidden_states=self.do_calc_hidden_loss,
             use_cache=False,  # since we are not passing labels, never let this default to True
-            return_dict=True,
         )
         dec_mask = decoder_input_ids.ne(pad_token_id)
         loss_ce = self.calc_ce_loss(dec_mask, lm_logits, teacher_outputs.logits)

examples/seq2seq/test_seq2seq_examples.py

@@ -185,7 +185,7 @@ class TestSummarizationDistiller(TestCasePlus):
 
     @require_torch_non_multi_gpu_but_fix_me
     def test_loss_fn(self):
-        model = AutoModelForSeq2SeqLM.from_pretrained(BART_TINY, return_dict=True)
+        model = AutoModelForSeq2SeqLM.from_pretrained(BART_TINY)
         input_ids, mask = model.dummy_inputs["input_ids"], model.dummy_inputs["attention_mask"]
         target_ids = torch.tensor([[0, 4, 8, 2], [0, 8, 2, 1]], dtype=torch.long, device=model.device)
         decoder_input_ids = target_ids[:, :-1].contiguous()  # Why this line?

model_cards/microsoft/prophetnet-large-uncased/README.md

@@ -23,7 +23,7 @@ target_str = "us rejects charges against its ambassador in bolivia"
 input_ids = tokenizer(input_str, return_tensors="pt").input_ids
 labels = tokenizer(target_str, return_tensors="pt").input_ids
 
-loss = model(input_ids, labels=labels, return_dict=True).loss
+loss = model(input_ids, labels=labels).loss
 ```
 
 ### Citation

model_cards/microsoft/xprophetnet-large-wiki100-cased/README.md

@@ -26,7 +26,7 @@ target_str = "us rejects charges against its ambassador in bolivia"
 input_ids = tokenizer(input_str, return_tensors="pt").input_ids
 labels = tokenizer(target_str, return_tensors="pt").input_ids
 
-loss = model(input_ids, labels=labels, return_dict=True).loss
+loss = model(input_ids, labels=labels).loss
 ```
 
 Note that since this model is a multi-lingual model it can be fine-tuned on all kinds of other languages.

model_cards/mrm8488/codebert-base-finetuned-detect-insecure-code/README.md

@@ -45,7 +45,7 @@ from transformers import AutoTokenizer, AutoModelForSequenceClassification
 import torch
 import numpy as np
 tokenizer = AutoTokenizer.from_pretrained('mrm8488/codebert-base-finetuned-detect-insecure-code')
-model = AutoModelForSequenceClassification.from_pretrained('mrm8488/codebert-base-finetuned-detect-insecure-code', return_dict=True)
+model = AutoModelForSequenceClassification.from_pretrained('mrm8488/codebert-base-finetuned-detect-insecure-code')
 
 inputs = tokenizer("your code here", return_tensors="pt", truncation=True, padding='max_length')
 labels = torch.tensor([1]).unsqueeze(0)  # Batch size 1

model_cards/sentence-transformers/LaBSE/README.md

@@ -13,7 +13,7 @@ sentences = ["Hello World", "Hallo Welt"]
 encoded_input = tokenizer(sentences, padding=True, truncation=True, max_length=64, return_tensors='pt')
 
 with torch.no_grad():
-    model_output = model(**encoded_input, return_dict=True)
+    model_output = model(**encoded_input)
 
 embeddings = model_output.pooler_output
 embeddings = torch.nn.functional.normalize(embeddings)

scripts/fsmt/fsmt-make-super-tiny-model.py

@@ -59,7 +59,7 @@ print(f"num of params {tiny_model.num_parameters()}")
 
 # Test
 batch = tokenizer.prepare_seq2seq_batch(["Making tiny model"])
-outputs = tiny_model(**batch, return_dict=True)
+outputs = tiny_model(**batch)
 
 print("test output:", len(outputs.logits[0]))
 

scripts/fsmt/fsmt-make-tiny-model.py

@@ -30,7 +30,7 @@ print(f"num of params {tiny_model.num_parameters()}")
 
 # Test
 batch = tokenizer.prepare_seq2seq_batch(["Making tiny model"])
-outputs = tiny_model(**batch, return_dict=True)
+outputs = tiny_model(**batch)
 
 print("test output:", len(outputs.logits[0]))
 

src/transformers/configuration_utils.py

@@ -55,7 +55,7 @@ class PretrainedConfig(object):
             Whether or not the model should returns all attentions.
         use_cache (:obj:`bool`, `optional`, defaults to :obj:`True`):
             Whether or not the model should return the last key/values attentions (not used by all models).
-        return_dict (:obj:`bool`, `optional`, defaults to :obj:`False`):
+        return_dict (:obj:`bool`, `optional`, defaults to :obj:`True`):
             Whether or not the model should return a :class:`~transformers.file_utils.ModelOutput` instead of a plain
             tuple.
         is_encoder_decoder (:obj:`bool`, `optional`, defaults to :obj:`False`):
@@ -163,7 +163,7 @@ class PretrainedConfig(object):
 
     def __init__(self, **kwargs):
         # Attributes with defaults
-        self.return_dict = kwargs.pop("return_dict", False)
+        self.return_dict = kwargs.pop("return_dict", True)
         self.output_hidden_states = kwargs.pop("output_hidden_states", False)
         self.output_attentions = kwargs.pop("output_attentions", False)
         self.use_cache = kwargs.pop("use_cache", True)  # Not used by all models

src/transformers/file_utils.py

@@ -559,7 +559,7 @@ PT_TOKEN_CLASSIFICATION_SAMPLE = r"""
         >>> import torch
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
         >>> labels = torch.tensor([1] * inputs["input_ids"].size(1)).unsqueeze(0)  # Batch size 1
@@ -576,7 +576,7 @@ PT_QUESTION_ANSWERING_SAMPLE = r"""
         >>> import torch
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
         >>> inputs = tokenizer(question, text, return_tensors='pt')
@@ -596,7 +596,7 @@ PT_SEQUENCE_CLASSIFICATION_SAMPLE = r"""
         >>> import torch
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
         >>> labels = torch.tensor([1]).unsqueeze(0)  # Batch size 1
@@ -612,7 +612,7 @@ PT_MASKED_LM_SAMPLE = r"""
         >>> import torch
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("The capital of France is {mask}.", return_tensors="pt")
         >>> labels = tokenizer("The capital of France is Paris.", return_tensors="pt")["input_ids"]
@@ -629,7 +629,7 @@ PT_BASE_MODEL_SAMPLE = r"""
         >>> import torch
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
         >>> outputs = model(**inputs)
@@ -644,7 +644,7 @@ PT_MULTIPLE_CHOICE_SAMPLE = r"""
         >>> import torch
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced."
         >>> choice0 = "It is eaten with a fork and a knife."
@@ -666,7 +666,7 @@ PT_CAUSAL_LM_SAMPLE = r"""
         >>> from transformers import {tokenizer_class}, {model_class}
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint})
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="pt")
         >>> outputs = model(**inputs, labels=inputs["input_ids"])
@@ -681,7 +681,7 @@ TF_TOKEN_CLASSIFICATION_SAMPLE = r"""
         >>> import tensorflow as tf
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="tf")
         >>> input_ids = inputs["input_ids"]
@@ -699,7 +699,7 @@ TF_QUESTION_ANSWERING_SAMPLE = r"""
         >>> import tensorflow as tf
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> question, text = "Who was Jim Henson?", "Jim Henson was a nice puppet"
         >>> input_dict = tokenizer(question, text, return_tensors='tf')
@@ -718,7 +718,7 @@ TF_SEQUENCE_CLASSIFICATION_SAMPLE = r"""
         >>> import tensorflow as tf
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="tf")
         >>> inputs["labels"] = tf.reshape(tf.constant(1), (-1, 1)) # Batch size 1
@@ -735,7 +735,7 @@ TF_MASKED_LM_SAMPLE = r"""
         >>> import tensorflow as tf
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("The capital of France is {mask}.", return_tensors="tf")
         >>> inputs["labels"] = tokenizer("The capital of France is Paris.", return_tensors="tf")["input_ids"]
@@ -752,7 +752,7 @@ TF_BASE_MODEL_SAMPLE = r"""
         >>> import tensorflow as tf
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="tf")
         >>> outputs = model(inputs)
@@ -767,7 +767,7 @@ TF_MULTIPLE_CHOICE_SAMPLE = r"""
         >>> import tensorflow as tf
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> prompt = "In Italy, pizza served in formal settings, such as at a restaurant, is presented unsliced."
         >>> choice0 = "It is eaten with a fork and a knife."
@@ -788,7 +788,7 @@ TF_CAUSAL_LM_SAMPLE = r"""
         >>> import tensorflow as tf
 
         >>> tokenizer = {tokenizer_class}.from_pretrained('{checkpoint}')
-        >>> model = {model_class}.from_pretrained('{checkpoint}', return_dict=True)
+        >>> model = {model_class}.from_pretrained('{checkpoint}')
 
         >>> inputs = tokenizer("Hello, my dog is cute", return_tensors="tf")
         >>> outputs = model(inputs)

src/transformers/modeling_albert.py

@@ -416,7 +416,7 @@ class AlbertTransformer(nn.Module):
         head_mask=None,
         output_attentions=False,
         output_hidden_states=False,
-        return_dict=False,
+        return_dict=True,
     ):
         hidden_states = self.embedding_hidden_mapping_in(hidden_states)
 
@@ -764,7 +764,7 @@ class AlbertForPreTraining(AlbertPreTrainedModel):
             >>> import torch
 
             >>> tokenizer = AlbertTokenizer.from_pretrained('albert-base-v2')
-            >>> model = AlbertForPreTraining.from_pretrained('albert-base-v2', return_dict=True)
+            >>> model = AlbertForPreTraining.from_pretrained('albert-base-v2')
 
             >>> input_ids = torch.tensor(tokenizer.encode("Hello, my dog is cute", add_special_tokens=True)).unsqueeze(0)  # Batch size 1
             >>> outputs = model(input_ids)