Tokenizers: ability to load from model subfolder (#8586)

* <small>tiny typo</small>

* Tokenizers: ability to load from model subfolder

* use subfolder for local files as well

* Uniformize model shortcut name => model id

* from s3 => from huggingface.co
Co-authored-by: Quentin Lhoest <lhoest.q@gmail.com>

src/transformers/models/auto/modeling_tf_auto.py

@@ -403,10 +403,9 @@ TF_AUTO_MODEL_PRETRAINED_DOCSTRING = r"""
             pretrained_model_name_or_path:
                 Can be either:
 
-                    - A string with the `shortcut name` of a pretrained model to load from cache or download, e.g.,
-                      ``bert-base-uncased``.
-                    - A string with the `identifier name` of a pretrained model that was user-uploaded to our S3, e.g.,
-                      ``dbmdz/bert-base-german-cased``.
+                    - A string, the `model id` of a pretrained model hosted inside a model repo on huggingface.co.
+                      Valid model ids can be located at the root-level, like ``bert-base-uncased``, or namespaced under
+                      a user or organization name, like ``dbmdz/bert-base-german-cased``.
                     - A path to a `directory` containing model weights saved using
                       :func:`~transformers.PreTrainedModel.save_pretrained`, e.g., ``./my_model_directory/``.
                     - A path or url to a `PyTorch state_dict save file` (e.g, ``./pt_model/pytorch_model.bin``). In
@@ -420,8 +419,8 @@ TF_AUTO_MODEL_PRETRAINED_DOCSTRING = r"""
                 Configuration for the model to use instead of an automatically loaded configuration. Configuration can
                 be automatically loaded when:
 
-                    - The model is a model provided by the library (loaded with the `shortcut name` string of a
-                      pretrained model).
+                    - The model is a model provided by the library (loaded with the `model id` string of a pretrained
+                      model).
                     - The model was saved using :meth:`~transformers.PreTrainedModel.save_pretrained` and is reloaded
                       by suppyling the save directory.
                     - The model is loaded by suppyling a local directory as ``pretrained_model_name_or_path`` and a
@@ -507,7 +506,7 @@ class TFAutoModel(object):
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModel
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = TFAutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModel.from_config(config)
         """
@@ -533,7 +532,7 @@ class TFAutoModel(object):
 
             >>> from transformers import AutoConfig, AutoModel
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModel.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -601,7 +600,7 @@ class TFAutoModelForPreTraining(object):
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForPreTraining
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelForPreTraining.from_config(config)
         """
@@ -627,7 +626,7 @@ class TFAutoModelForPreTraining(object):
 
             >>> from transformers import AutoConfig, TFAutoModelForPreTraining
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForPreTraining.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -701,7 +700,7 @@ class TFAutoModelWithLMHead(object):
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelWithLMHead
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelWithLMHead.from_config(config)
         """
@@ -733,7 +732,7 @@ class TFAutoModelWithLMHead(object):
 
             >>> from transformers import AutoConfig, TFAutoModelWithLMHead
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelWithLMHead.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -808,7 +807,7 @@ class TFAutoModelForCausalLM:
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForCausalLM
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('gpt2')
             >>> model = TFAutoModelForCausalLM.from_config(config)
         """
@@ -834,7 +833,7 @@ class TFAutoModelForCausalLM:
 
             >>> from transformers import AutoConfig, TFAutoModelForCausalLM
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForCausalLM.from_pretrained('gpt2')
 
             >>> # Update configuration during loading
@@ -902,7 +901,7 @@ class TFAutoModelForMaskedLM:
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForMaskedLM
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelForMaskedLM.from_config(config)
         """
@@ -928,7 +927,7 @@ class TFAutoModelForMaskedLM:
 
             >>> from transformers import AutoConfig, TFAutoModelForMaskedLM
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForMaskedLM.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -996,7 +995,7 @@ class TFAutoModelForSeq2SeqLM:
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForSeq2SeqLM
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('t5')
             >>> model = TFAutoModelForSeq2SeqLM.from_config(config)
         """
@@ -1024,7 +1023,7 @@ class TFAutoModelForSeq2SeqLM:
 
             >>> from transformers import AutoConfig, TFAutoModelForSeq2SeqLM
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForSeq2SeqLM.from_pretrained('t5-base')
 
             >>> # Update configuration during loading
@@ -1094,7 +1093,7 @@ class TFAutoModelForSequenceClassification(object):
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForSequenceClassification
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelForSequenceClassification.from_config(config)
         """
@@ -1122,7 +1121,7 @@ class TFAutoModelForSequenceClassification(object):
 
             >>> from transformers import AutoConfig, TFAutoModelForSequenceClassification
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForSequenceClassification.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -1191,7 +1190,7 @@ class TFAutoModelForQuestionAnswering(object):
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForQuestionAnswering
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelForQuestionAnswering.from_config(config)
         """
@@ -1219,7 +1218,7 @@ class TFAutoModelForQuestionAnswering(object):
 
             >>> from transformers import AutoConfig, TFAutoModelForQuestionAnswering
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForQuestionAnswering.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -1288,7 +1287,7 @@ class TFAutoModelForTokenClassification:
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForTokenClassification
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelForTokenClassification.from_config(config)
         """
@@ -1316,7 +1315,7 @@ class TFAutoModelForTokenClassification:
 
             >>> from transformers import AutoConfig, TFAutoModelForTokenClassification
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForTokenClassification.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -1386,7 +1385,7 @@ class TFAutoModelForMultipleChoice:
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForMultipleChoice
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelForMultipleChoice.from_config(config)
         """
@@ -1414,7 +1413,7 @@ class TFAutoModelForMultipleChoice:
 
             >>> from transformers import AutoConfig, TFAutoModelForMultipleChoice
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForMultipleChoice.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading
@@ -1484,7 +1483,7 @@ class TFAutoModelForNextSentencePrediction:
         Examples::
 
             >>> from transformers import AutoConfig, TFAutoModelForNextSentencePrediction
-            >>> # Download configuration from S3 and cache.
+            >>> # Download configuration from huggingface.co and cache.
             >>> config = AutoConfig.from_pretrained('bert-base-uncased')
             >>> model = TFAutoModelForNextSentencePrediction.from_config(config)
         """
@@ -1512,7 +1511,7 @@ class TFAutoModelForNextSentencePrediction:
 
             >>> from transformers import AutoConfig, TFAutoModelForNextSentencePrediction
 
-            >>> # Download model and configuration from S3 and cache.
+            >>> # Download model and configuration from huggingface.co and cache.
             >>> model = TFAutoModelForNextSentencePrediction.from_pretrained('bert-base-uncased')
 
             >>> # Update configuration during loading

src/transformers/models/auto/tokenization_auto.py

@@ -250,10 +250,9 @@ class AutoTokenizer:
             pretrained_model_name_or_path (:obj:`str`):
                 Can be either:
 
-                    - A string with the `shortcut name` of a predefined tokenizer to load from cache or download, e.g.,
-                      ``bert-base-uncased``.
-                    - A string with the `identifier name` of a predefined tokenizer that was user-uploaded to our S3,
-                      e.g., ``dbmdz/bert-base-german-cased``.
+                    - A string, the `model id` of a predefined tokenizer hosted inside a model repo on huggingface.co.
+                      Valid model ids can be located at the root-level, like ``bert-base-uncased``, or namespaced under
+                      a user or organization name, like ``dbmdz/bert-base-german-cased``.
                     - A path to a `directory` containing vocabulary files required by the tokenizer, for instance saved
                       using the :func:`~transformers.PreTrainedTokenizer.save_pretrained` method, e.g.,
                       ``./my_model_directory/``.
@@ -280,6 +279,9 @@ class AutoTokenizer:
                 The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a
                 git-based system for storing models and other artifacts on huggingface.co, so ``revision`` can be any
                 identifier allowed by git.
+            subfolder (:obj:`str`, `optional`):
+                In case the relevant files are located inside a subfolder of the model repo on huggingface.co (e.g. for
+                facebook/rag-token-base), specify it here.
             use_fast (:obj:`bool`, `optional`, defaults to :obj:`True`):
                 Whether or not to try to load the fast version of the tokenizer.
             kwargs (additional keyword arguments, `optional`):
@@ -291,10 +293,10 @@ class AutoTokenizer:
 
             >>> from transformers import AutoTokenizer
 
-            >>> # Download vocabulary from S3 and cache.
+            >>> # Download vocabulary from huggingface.co and cache.
             >>> tokenizer = AutoTokenizer.from_pretrained('bert-base-uncased')
 
-            >>> # Download vocabulary from S3 (user-uploaded) and cache.
+            >>> # Download vocabulary from huggingface.co (user-uploaded) and cache.
             >>> tokenizer = AutoTokenizer.from_pretrained('dbmdz/bert-base-german-cased')
 
             >>> # If vocabulary files are in a directory (e.g. tokenizer was saved using `save_pretrained('./test/saved_model/')`)

src/transformers/models/encoder_decoder/modeling_encoder_decoder.py

@@ -214,10 +214,9 @@ class EncoderDecoderModel(PreTrainedModel):
             encoder_pretrained_model_name_or_path (:obj: `str`, `optional`):
                 Information necessary to initiate the encoder. Can be either:
 
-                    - A string with the `shortcut name` of a pretrained model to load from cache or download, e.g.,
-                      ``bert-base-uncased``.
-                    - A string with the `identifier name` of a pretrained model that was user-uploaded to our S3, e.g.,
-                      ``dbmdz/bert-base-german-cased``.
+                    - A string, the `model id` of a pretrained model hosted inside a model repo on huggingface.co.
+                      Valid model ids can be located at the root-level, like ``bert-base-uncased``, or namespaced under
+                      a user or organization name, like ``dbmdz/bert-base-german-cased``.
                     - A path to a `directory` containing model weights saved using
                       :func:`~transformers.PreTrainedModel.save_pretrained`, e.g., ``./my_model_directory/``.
                     - A path or url to a `tensorflow index checkpoint file` (e.g, ``./tf_model/model.ckpt.index``). In
@@ -228,10 +227,9 @@ class EncoderDecoderModel(PreTrainedModel):
             decoder_pretrained_model_name_or_path (:obj: `str`, `optional`, defaults to `None`):
                 Information necessary to initiate the decoder. Can be either:
 
-                    - A string with the `shortcut name` of a pretrained model to load from cache or download, e.g.,
-                      ``bert-base-uncased``.
-                    - A string with the `identifier name` of a pretrained model that was user-uploaded to our S3, e.g.,
-                      ``dbmdz/bert-base-german-cased``.
+                    - A string, the `model id` of a pretrained model hosted inside a model repo on huggingface.co.
+                      Valid model ids can be located at the root-level, like ``bert-base-uncased``, or namespaced under
+                      a user or organization name, like ``dbmdz/bert-base-german-cased``.
                     - A path to a `directory` containing model weights saved using
                       :func:`~transformers.PreTrainedModel.save_pretrained`, e.g., ``./my_model_directory/``.
                     - A path or url to a `tensorflow index checkpoint file` (e.g, ``./tf_model/model.ckpt.index``). In

src/transformers/models/lxmert/tokenization_lxmert.py

@@ -24,7 +24,7 @@ VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt"}
 
 ####################################################
 # Mapping from the keyword arguments names of Tokenizer `__init__`
-# to pretrained vocabulary URL for all the model shortcut names.
+# to pretrained vocabulary URL for all the model ids.
 ####################################################
 PRETRAINED_VOCAB_FILES_MAP = {
     "vocab_file": {
@@ -33,13 +33,13 @@ PRETRAINED_VOCAB_FILES_MAP = {
 }
 
 ####################################################
-# Mapping from model shortcut names to max length of inputs
+# Mapping from model ids to max length of inputs
 ####################################################
 PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
     "unc-nlp/lxmert-base-uncased": 512,
 }
 ####################################################
-# Mapping from model shortcut names to a dictionary of additional
+# Mapping from model ids to a dictionary of additional
 # keyword arguments for Tokenizer `__init__`.
 # To be used for checkpoint specific configurations.
 ####################################################

src/transformers/models/lxmert/tokenization_lxmert_fast.py

@@ -25,7 +25,7 @@ VOCAB_FILES_NAMES = {"vocab_file": "vocab.txt", "tokenizer_file": "tokenizer.jso
 
 ####################################################
 # Mapping from the keyword arguments names of Tokenizer `__init__`
-# to pretrained vocabulary URL for all the model shortcut names.
+# to pretrained vocabulary URL for all the model ids.
 ####################################################
 PRETRAINED_VOCAB_FILES_MAP = {
     "vocab_file": {
@@ -37,13 +37,13 @@ PRETRAINED_VOCAB_FILES_MAP = {
 }
 
 ####################################################
-# Mapping from model shortcut names to max length of inputs
+# Mapping from model ids to max length of inputs
 ####################################################
 PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
     "unc-nlp/lxmert-base-uncased": 512,
 }
 ####################################################
-# Mapping from model shortcut names to a dictionary of additional
+# Mapping from model ids to a dictionary of additional
 # keyword arguments for Tokenizer `__init__`.
 # To be used for checkpoint specific configurations.
 ####################################################

src/transformers/models/rag/modeling_rag.py

@@ -238,10 +238,9 @@ class RagPreTrainedModel(PreTrainedModel):
             question_encoder_pretrained_model_name_or_path (:obj: `str`, `optional`, defaults to `None`):
                 Information necessary to initiate the question encoder. Can be either:
 
-                    - A string with the `shortcut name` of a pretrained model to load from cache or download, e.g.,
-                      ``bert-base-uncased``.
-                    - A string with the `identifier name` of a pretrained model that was user-uploaded to our S3, e.g.,
-                      ``dbmdz/bert-base-german-cased``.
+                    - A string, the `model id` of a pretrained model hosted inside a model repo on huggingface.co.
+                      Valid model ids can be located at the root-level, like ``bert-base-uncased``, or namespaced under
+                      a user or organization name, like ``dbmdz/bert-base-german-cased``.
                     - A path to a `directory` containing model weights saved using
                       :func:`~transformers.PreTrainedModel.save_pretrained`, e.g., ``./my_model_directory/``.
                     - A path or url to a `tensorflow index checkpoint file` (e.g, ``./tf_model/model.ckpt.index``). In
@@ -252,10 +251,9 @@ class RagPreTrainedModel(PreTrainedModel):
             generator_pretrained_model_name_or_path (:obj: `str`, `optional`, defaults to `None`):
                 Information necessary to initiate the generator. Can be either:
 
-                    - A string with the `shortcut name` of a pretrained model to load from cache or download, e.g.,
-                      ``bert-base-uncased``.
-                    - A string with the `identifier name` of a pretrained model that was user-uploaded to our S3, e.g.,
-                      ``dbmdz/bert-base-german-cased``.
+                    - A string, the `model id` of a pretrained model hosted inside a model repo on huggingface.co.
+                      Valid model ids can be located at the root-level, like ``bert-base-uncased``, or namespaced under
+                      a user or organization name, like ``dbmdz/bert-base-german-cased``.
                     - A path to a `directory` containing model weights saved using
                       :func:`~transformers.PreTrainedModel.save_pretrained`, e.g., ``./my_model_directory/``.
                     - A path or url to a `tensorflow index checkpoint file` (e.g, ``./tf_model/model.ckpt.index``). In

src/transformers/models/rag/tokenization_rag.py

@@ -49,10 +49,12 @@ class RagTokenizer:
         if config is None:
             config = RagConfig.from_pretrained(pretrained_model_name_or_path)
 
-        question_encoder_path = os.path.join(pretrained_model_name_or_path, "question_encoder_tokenizer")
-        generator_path = os.path.join(pretrained_model_name_or_path, "generator_tokenizer")
-        question_encoder = AutoTokenizer.from_pretrained(question_encoder_path, config=config.question_encoder)
-        generator = AutoTokenizer.from_pretrained(generator_path, config=config.generator)
+        question_encoder = AutoTokenizer.from_pretrained(
+            pretrained_model_name_or_path, config=config.question_encoder, subfolder="question_encoder_tokenizer"
+        )
+        generator = AutoTokenizer.from_pretrained(
+            pretrained_model_name_or_path, config=config.generator, subfolder="generator_tokenizer"
+        )
         return cls(question_encoder=question_encoder, generator=generator)
 
     def __call__(self, *args, **kwargs):

src/transformers/models/reformer/tokenization_reformer.py

@@ -38,7 +38,7 @@ VOCAB_FILES_NAMES = {"vocab_file": "spiece.model"}
 
 ####################################################
 # Mapping from the keyword arguments names of Tokenizer `__init__`
-# to pretrained vocabulary URL for all the model shortcut names.
+# to pretrained vocabulary URL for all the model ids.
 ####################################################
 PRETRAINED_VOCAB_FILES_MAP = {
     "vocab_file": {
@@ -47,7 +47,7 @@ PRETRAINED_VOCAB_FILES_MAP = {
 }
 
 ####################################################
-# Mapping from model shortcut names to max length of inputs
+# Mapping from model ids to max length of inputs
 ####################################################
 PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
     "google/reformer-crime-and-punishment": 524288,

src/transformers/models/reformer/tokenization_reformer_fast.py

@@ -43,7 +43,7 @@ VOCAB_FILES_NAMES = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.
 
 ####################################################
 # Mapping from the keyword arguments names of Tokenizer `__init__`
-# to pretrained vocabulary URL for all the model shortcut names.
+# to pretrained vocabulary URL for all the model ids.
 ####################################################
 PRETRAINED_VOCAB_FILES_MAP = {
     "vocab_file": {
@@ -55,7 +55,7 @@ PRETRAINED_VOCAB_FILES_MAP = {
 }
 
 ####################################################
-# Mapping from model shortcut names to max length of inputs
+# Mapping from model ids to max length of inputs
 ####################################################
 PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
     "google/reformer-crime-and-punishment": 524288,

src/transformers/models/t5/modeling_t5.py

@@ -50,7 +50,7 @@ _CONFIG_FOR_DOC = "T5Config"
 _TOKENIZER_FOR_DOC = "T5Tokenizer"
 
 ####################################################
-# This dict contains shortcut names and associated url
+# This dict contains ids and associated url
 # for the pretrained weights provided with the models
 ####################################################
 T5_PRETRAINED_MODEL_ARCHIVE_LIST = [

src/transformers/models/t5/tokenization_t5.py

@@ -39,7 +39,7 @@ VOCAB_FILES_NAMES = {"vocab_file": "spiece.model"}
 
 ####################################################
 # Mapping from the keyword arguments names of Tokenizer `__init__`
-# to pretrained vocabulary URL for all the model shortcut names.
+# to pretrained vocabulary URL for all the model ids.
 ####################################################
 PRETRAINED_VOCAB_FILES_MAP = {
     "vocab_file": {
@@ -52,7 +52,7 @@ PRETRAINED_VOCAB_FILES_MAP = {
 }
 
 ####################################################
-# Mapping from model shortcut names to max length of inputs
+# Mapping from model ids to max length of inputs
 ####################################################
 PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
     "t5-small": 512,

src/transformers/models/t5/tokenization_t5_fast.py

@@ -42,7 +42,7 @@ VOCAB_FILES_NAMES = {"vocab_file": "spiece.model", "tokenizer_file": "tokenizer.
 
 ####################################################
 # Mapping from the keyword arguments names of Tokenizer `__init__`
-# to pretrained vocabulary URL for all the model shortcut names.
+# to pretrained vocabulary URL for all the model ids.
 ####################################################
 PRETRAINED_VOCAB_FILES_MAP = {
     "vocab_file": {
@@ -62,7 +62,7 @@ PRETRAINED_VOCAB_FILES_MAP = {
 }
 
 ####################################################
-# Mapping from model shortcut names to max length of inputs
+# Mapping from model ids to max length of inputs
 ####################################################
 PRETRAINED_POSITIONAL_EMBEDDINGS_SIZES = {
     "t5-small": 512,

src/transformers/tokenization_utils_base.py

@@ -1615,10 +1615,9 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
             pretrained_model_name_or_path (:obj:`str`):
                 Can be either:
 
-                - A string with the `shortcut name` of a predefined tokenizer to load from cache or download, e.g.,
-                  ``bert-base-uncased``.
-                - A string with the `identifier name` of a predefined tokenizer that was user-uploaded to our S3, e.g.,
-                  ``dbmdz/bert-base-german-cased``.
+                - A string, the `model id` of a predefined tokenizer hosted inside a model repo on huggingface.co.
+                  Valid model ids can be located at the root-level, like ``bert-base-uncased``, or namespaced under a
+                  user or organization name, like ``dbmdz/bert-base-german-cased``.
                 - A path to a `directory` containing vocabulary files required by the tokenizer, for instance saved
                   using the :meth:`~transformers.tokenization_utils_base.PreTrainedTokenizerBase.save_pretrained`
                   method, e.g., ``./my_model_directory/``.
@@ -1641,6 +1640,9 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
                 The specific model version to use. It can be a branch name, a tag name, or a commit id, since we use a
                 git-based system for storing models and other artifacts on huggingface.co, so ``revision`` can be any
                 identifier allowed by git.
+            subfolder (:obj:`str`, `optional`):
+                In case the relevant files are located inside a subfolder of the model repo on huggingface.co (e.g. for
+                facebook/rag-token-base), specify it here.
             inputs (additional positional arguments, `optional`):
                 Will be passed along to the Tokenizer ``__init__`` method.
             kwargs (additional keyword arguments, `optional`):
@@ -1651,10 +1653,10 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
         Examples::
 
             # We can't instantiate directly the base class `PreTrainedTokenizerBase` so let's show our examples on a derived class: BertTokenizer
-            # Download vocabulary from S3 and cache.
+            # Download vocabulary from huggingface.co and cache.
             tokenizer = BertTokenizer.from_pretrained('bert-base-uncased')
 
-            # Download vocabulary from S3 (user-uploaded) and cache.
+            # Download vocabulary from huggingface.co (user-uploaded) and cache.
             tokenizer = BertTokenizer.from_pretrained('dbmdz/bert-base-german-cased')
 
             # If vocabulary files are in a directory (e.g. tokenizer was saved using `save_pretrained('./test/saved_model/')`)
@@ -1676,6 +1678,7 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
         proxies = kwargs.pop("proxies", None)
         local_files_only = kwargs.pop("local_files_only", False)
         revision = kwargs.pop("revision", None)
+        subfolder = kwargs.pop("subfolder", None)
 
         s3_models = list(cls.max_model_input_sizes.keys())
         vocab_files = {}
@@ -1722,13 +1725,20 @@ class PreTrainedTokenizerBase(SpecialTokensMixin):
                 # Look for the tokenizer files
                 for file_id, file_name in {**cls.vocab_files_names, **additional_files_names}.items():
                     if os.path.isdir(pretrained_model_name_or_path):
+                        if subfolder is not None:
+                            full_file_name = os.path.join(pretrained_model_name_or_path, subfolder, file_name)
+                        else:
                             full_file_name = os.path.join(pretrained_model_name_or_path, file_name)
                         if not os.path.exists(full_file_name):
                             logger.info("Didn't find file {}. We won't load it.".format(full_file_name))
                             full_file_name = None
                     else:
                         full_file_name = hf_bucket_url(
-                            pretrained_model_name_or_path, filename=file_name, revision=revision, mirror=None
+                            pretrained_model_name_or_path,
+                            filename=file_name,
+                            subfolder=subfolder,
+                            revision=revision,
+                            mirror=None,
                         )
 
                     vocab_files[file_id] = full_file_name

templates/adding_a_new_example_script/{{cookiecutter.directory_name}}/run_{{cookiecutter.example_shortcut}}.py

@@ -75,7 +75,7 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}
     )
     use_fast_tokenizer: bool = field(
         default=True,
@@ -98,7 +98,7 @@ class ModelArguments:
         default=None, metadata={"help": "Pretrained tokenizer name or path if not the same as model_name"}
     )
     cache_dir: Optional[str] = field(
-        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from s3"}
+        default=None, metadata={"help": "Where do you want to store the pretrained models downloaded from huggingface.co"}
     )
     use_fast_tokenizer: bool = field(
         default=True,