Adapt repository creation to latest hf_hub (#21158)

* Adapt repository creation to latest hf_hub

* Update all examples

* Fix other tests, add Flax examples

* Address review comments

examples/flax/image-captioning/run_image_captioning_flax.py

@@ -45,7 +45,7 @@ from flax import jax_utils, traverse_util
 from flax.jax_utils import unreplicate
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard, shard_prng_key
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AutoFeatureExtractor,
     AutoTokenizer,
@@ -430,7 +430,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Get the datasets: you can either provide your own CSV/JSON training and evaluation files (see below)
     # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/

examples/flax/language-modeling/run_bart_dlm_flax.py

@@ -45,7 +45,7 @@ from flax import jax_utils, traverse_util
 from flax.jax_utils import pad_shard_unpad
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     FLAX_MODEL_FOR_MASKED_LM_MAPPING,
@@ -502,7 +502,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)
     # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/

examples/flax/language-modeling/run_clm_flax.py

@@ -46,7 +46,7 @@ from flax import jax_utils, traverse_util
 from flax.jax_utils import pad_shard_unpad, unreplicate
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard, shard_prng_key
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     FLAX_MODEL_FOR_CAUSAL_LM_MAPPING,
@@ -376,7 +376,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     #  Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)
     # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/

examples/flax/language-modeling/run_mlm_flax.py

@@ -46,7 +46,7 @@ from flax import jax_utils, traverse_util
 from flax.jax_utils import pad_shard_unpad
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     FLAX_MODEL_FOR_MASKED_LM_MAPPING,
@@ -416,7 +416,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)
     # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/

examples/flax/language-modeling/run_t5_mlm_flax.py

@@ -45,7 +45,7 @@ from flax import jax_utils, traverse_util
 from flax.jax_utils import pad_shard_unpad
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     FLAX_MODEL_FOR_MASKED_LM_MAPPING,
@@ -542,7 +542,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)
     # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/

examples/flax/question-answering/run_qa.py

@@ -44,7 +44,7 @@ from flax import struct, traverse_util
 from flax.jax_utils import pad_shard_unpad, replicate, unreplicate
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AutoConfig,
     AutoTokenizer,
@@ -467,7 +467,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # region Load Data
     # Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)

examples/flax/summarization/run_summarization_flax.py

@@ -46,7 +46,7 @@ from flax import jax_utils, traverse_util
 from flax.jax_utils import pad_shard_unpad, unreplicate
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard, shard_prng_key
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     FLAX_MODEL_FOR_SEQ_TO_SEQ_CAUSAL_LM_MAPPING,
@@ -450,7 +450,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Get the datasets: you can either provide your own CSV/JSON training and evaluation files (see below)
     # or just provide the name of one of the public datasets available on the hub at https://huggingface.co/datasets/

examples/flax/text-classification/run_flax_glue.py

@@ -39,7 +39,7 @@ from flax import struct, traverse_util
 from flax.jax_utils import pad_shard_unpad, replicate, unreplicate
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AutoConfig,
     AutoTokenizer,
@@ -350,7 +350,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Get the datasets: you can either provide your own CSV/JSON training and evaluation files (see below)
     # or specify a GLUE benchmark task (the dataset will be downloaded automatically from the datasets Hub).

examples/flax/token-classification/run_flax_ner.py

@@ -41,7 +41,7 @@ from flax import struct, traverse_util
 from flax.jax_utils import pad_shard_unpad, replicate, unreplicate
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AutoConfig,
     AutoTokenizer,
@@ -406,7 +406,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Get the datasets: you can either provide your own CSV/JSON/TXT training and evaluation files (see below)
     # or just provide the name of one of the public datasets for token classification task available on the hub at https://huggingface.co/datasets/

examples/flax/vision/run_image_classification.py

@@ -43,7 +43,7 @@ from flax import jax_utils
 from flax.jax_utils import pad_shard_unpad, unreplicate
 from flax.training import train_state
 from flax.training.common_utils import get_metrics, onehot, shard, shard_prng_key
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     FLAX_MODEL_FOR_IMAGE_CLASSIFICATION_MAPPING,
@@ -298,7 +298,8 @@ def main():
             )
         else:
             repo_name = training_args.hub_model_id
-        repo = Repository(training_args.output_dir, clone_from=repo_name)
+        create_repo(repo_name, exist_ok=True, token=training_args.hub_token)
+        repo = Repository(training_args.output_dir, clone_from=repo_name, token=training_args.hub_token)
 
     # Initialize datasets and pre-processing transforms
     # We use torchvision here for faster pre-processing

examples/pytorch/image-classification/run_image_classification_no_trainer.py

@@ -40,7 +40,7 @@ import transformers
 from accelerate import Accelerator
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AutoConfig,
     AutoFeatureExtractor,
@@ -246,7 +246,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/language-modeling/run_clm_no_trainer.py

@@ -41,7 +41,7 @@ import transformers
 from accelerate import Accelerator, DistributedType
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     MODEL_MAPPING,
@@ -282,7 +282,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/language-modeling/run_mlm_no_trainer.py

@@ -41,7 +41,7 @@ import transformers
 from accelerate import Accelerator, DistributedType
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     MODEL_MAPPING,
@@ -291,7 +291,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/multiple-choice/run_swag_no_trainer.py

@@ -40,7 +40,7 @@ import transformers
 from accelerate import Accelerator
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     MODEL_MAPPING,
@@ -317,7 +317,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/question-answering/run_qa_beam_search_no_trainer.py

@@ -38,7 +38,7 @@ import transformers
 from accelerate import Accelerator
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AdamW,
     DataCollatorWithPadding,
@@ -332,7 +332,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/question-answering/run_qa_no_trainer.py

@@ -38,7 +38,7 @@ import transformers
 from accelerate import Accelerator
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     MODEL_MAPPING,
@@ -370,7 +370,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/semantic-segmentation/run_semantic_segmentation_no_trainer.py

@@ -36,7 +36,7 @@ import transformers
 from accelerate import Accelerator
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository, hf_hub_download
+from huggingface_hub import Repository, create_repo, hf_hub_download
 from transformers import (
     AutoConfig,
     AutoFeatureExtractor,
@@ -354,7 +354,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/speech-pretraining/run_wav2vec2_pretraining_no_trainer.py

@@ -31,7 +31,7 @@ from tqdm.auto import tqdm
 import transformers
 from accelerate import Accelerator
 from accelerate.logging import get_logger
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AdamW,
     SchedulerType,
@@ -422,7 +422,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
         elif args.output_dir is not None:
             os.makedirs(args.output_dir, exist_ok=True)
     accelerator.wait_for_everyone()

examples/pytorch/summarization/run_summarization_no_trainer.py

@@ -40,7 +40,7 @@ from accelerate import Accelerator
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
 from filelock import FileLock
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     CONFIG_MAPPING,
     MODEL_MAPPING,
@@ -373,7 +373,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore:

examples/pytorch/text-classification/run_glue_no_trainer.py

@@ -32,7 +32,7 @@ import transformers
 from accelerate import Accelerator
 from accelerate.logging import get_logger
 from accelerate.utils import set_seed
-from huggingface_hub import Repository
+from huggingface_hub import Repository, create_repo
 from transformers import (
     AutoConfig,
     AutoModelForSequenceClassification,
@@ -244,7 +244,8 @@ def main():
                 repo_name = get_full_repo_name(Path(args.output_dir).name, token=args.hub_token)
             else:
                 repo_name = args.hub_model_id
-            repo = Repository(args.output_dir, clone_from=repo_name)
+            create_repo(repo_name, exist_ok=True, token=args.hub_token)
+            repo = Repository(args.output_dir, clone_from=repo_name, token=args.hub_token)
 
             with open(os.path.join(args.output_dir, ".gitignore"), "w+") as gitignore:
                 if "step_*" not in gitignore: