SUMIN/HuggingFace_transformer
1
0
Fork 0
Code Issues Pull Requests Actions 7 Packages Projects Releases Wiki Activity

Allow trust_remote_code in example scripts (#25248)

Browse Source 
* pytorch examples

* pytorch mim no trainer

* cookiecutter

* flax examples

* missed line in pytorch run_glue

* tensorflow examples

* tensorflow run_clip

* tensorflow run_mlm

* tensorflow run_ner

* tensorflow run_clm

* pytorch example from_configs

* pytorch no trainer examples

* Revert "tensorflow run_clip"

This reverts commit 261f86ac1f1c9e05dd3fd0291e1a1f8e573781d5.

* fix: duplicated argument
This commit is contained in:
Jackmin801
2023-08-07 22:32:25 +08:00
committed by GitHub
parent 65001cb1c8
commit 145109382a
49 changed files with 790 additions and 65 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
examples/tensorflow/image-classification/run_image_classification.py
View File

@@ -173,6 +173,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
ignore_mismatched_sizes: bool = field(
default=False,
metadata={"help": "Will enable to load a pretrained model whose head dimensions are different."},
@@ -323,12 +333,14 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
image_processor = AutoImageProcessor.from_pretrained(
model_args.image_processor_name or model_args.model_name_or_path,
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# If we don't have a validation split, split off a percentage of train as validation.
@@ -449,6 +461,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
ignore_mismatched_sizes=model_args.ignore_mismatched_sizes,
)
num_replicas = training_args.strategy.num_replicas_in_sync

37
examples/tensorflow/language-modeling/run_clm.py
View File

@@ -128,6 +128,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
def __post_init__(self):
if self.config_overrides is not None and (self.config_name is not None or self.model_name_or_path is not None):
@@ -366,17 +376,26 @@ def main():
# In distributed training, the .from_pretrained methods guarantee that only one local process can concurrently
# download model & vocab.
if model_args.config_name:
config = AutoConfig.from_pretrained(model_args.config_name)
config = AutoConfig.from_pretrained(
model_args.config_name,
trust_remote_code=model_args.trust_remote_code,
)
elif model_args.model_name_or_path:
config = AutoConfig.from_pretrained(model_args.model_name_or_path)
config = AutoConfig.from_pretrained(
model_args.model_name_or_path, trust_remote_code=model_args.trust_remote_code
)
else:
config = CONFIG_MAPPING[model_args.model_type]()
logger.warning("You are instantiating a new config instance from scratch.")
if model_args.tokenizer_name:
tokenizer = AutoTokenizer.from_pretrained(model_args.tokenizer_name)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name, trust_remote_code=model_args.trust_remote_code
)
elif model_args.model_name_or_path:
tokenizer = AutoTokenizer.from_pretrained(model_args.model_name_or_path)
tokenizer = AutoTokenizer.from_pretrained(
model_args.model_name_or_path, trust_remote_code=model_args.trust_remote_code
)
else:
raise ValueError(
"You are instantiating a new tokenizer from scratch. This is not supported by this script."
@@ -479,12 +498,16 @@ def main():
with training_args.strategy.scope():
# region Prepare model
if checkpoint is not None:
model = TFAutoModelForCausalLM.from_pretrained(checkpoint, config=config)
model = TFAutoModelForCausalLM.from_pretrained(
checkpoint, config=config, trust_remote_code=model_args.trust_remote_code
)
elif model_args.model_name_or_path:
model = TFAutoModelForCausalLM.from_pretrained(model_args.model_name_or_path, config=config)
model = TFAutoModelForCausalLM.from_pretrained(
model_args.model_name_or_path, config=config, trust_remote_code=model_args.trust_remote_code
)
else:
logger.info("Training new model from scratch")
model = TFAutoModelForCausalLM.from_config(config)
model = TFAutoModelForCausalLM.from_config(config, trust_remote_code=model_args.trust_remote_code)
# We resize the embeddings only when necessary to avoid index errors. If you are creating a model from scratch
# on a small vocab and want a smaller embedding size, remove this test.

36
examples/tensorflow/language-modeling/run_mlm.py
View File

@@ -126,6 +126,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
def __post_init__(self):
if self.config_overrides is not None and (self.config_name is not None or self.model_name_or_path is not None):
@@ -348,19 +358,25 @@ def main():
# In distributed training, the .from_pretrained methods guarantee that only one local process can concurrently
# download model & vocab.
if checkpoint is not None:
config = AutoConfig.from_pretrained(checkpoint)
config = AutoConfig.from_pretrained(checkpoint, trust_remote_code=model_args.trust_remote_code)
elif model_args.config_name:
config = AutoConfig.from_pretrained(model_args.config_name)
config = AutoConfig.from_pretrained(model_args.config_name, trust_remote_code=model_args.trust_remote_code)
elif model_args.model_name_or_path:
config = AutoConfig.from_pretrained(model_args.model_name_or_path)
config = AutoConfig.from_pretrained(
model_args.model_name_or_path, trust_remote_code=model_args.trust_remote_code
)
else:
config = CONFIG_MAPPING[model_args.model_type]()
logger.warning("You are instantiating a new config instance from scratch.")
if model_args.tokenizer_name:
tokenizer = AutoTokenizer.from_pretrained(model_args.tokenizer_name)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name, trust_remote_code=model_args.trust_remote_code
)
elif model_args.model_name_or_path:
tokenizer = AutoTokenizer.from_pretrained(model_args.model_name_or_path)
tokenizer = AutoTokenizer.from_pretrained(
model_args.model_name_or_path, trust_remote_code=model_args.trust_remote_code
)
else:
raise ValueError(
"You are instantiating a new tokenizer from scratch. This is not supported by this script."
@@ -495,12 +511,16 @@ def main():
with training_args.strategy.scope():
# region Prepare model
if checkpoint is not None:
model = TFAutoModelForMaskedLM.from_pretrained(checkpoint, config=config)
model = TFAutoModelForMaskedLM.from_pretrained(
checkpoint, config=config, trust_remote_code=model_args.trust_remote_code
)
elif model_args.model_name_or_path:
model = TFAutoModelForMaskedLM.from_pretrained(model_args.model_name_or_path, config=config)
model = TFAutoModelForMaskedLM.from_pretrained(
model_args.model_name_or_path, config=config, trust_remote_code=model_args.trust_remote_code
)
else:
logger.info("Training new model from scratch")
model = TFAutoModelForMaskedLM.from_config(config)
model = TFAutoModelForMaskedLM.from_config(config, trust_remote_code=model_args.trust_remote_code)
# We resize the embeddings only when necessary to avoid index errors. If you are creating a model from scratch
# on a small vocab and want a smaller embedding size, remove this test.

13
examples/tensorflow/multiple-choice/run_swag.py
View File

@@ -162,6 +162,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
@dataclass
@@ -349,6 +359,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
@@ -356,6 +367,7 @@ def main():
use_fast=model_args.use_fast_tokenizer,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# endregion
@@ -442,6 +454,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
num_replicas = training_args.strategy.num_replicas_in_sync

13
examples/tensorflow/question-answering/run_qa.py
View File

@@ -93,6 +93,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
@dataclass
@@ -352,6 +362,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
@@ -359,6 +370,7 @@ def main():
use_fast=True,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# endregion
@@ -639,6 +651,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
if training_args.do_train:
training_dataset = model.prepare_tf_dataset(

13
examples/tensorflow/summarization/run_summarization.py
View File

@@ -115,6 +115,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
@dataclass
@@ -402,6 +412,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
@@ -409,6 +420,7 @@ def main():
use_fast=model_args.use_fast_tokenizer,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
prefix = data_args.source_prefix if data_args.source_prefix is not None else ""
@@ -527,6 +539,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# We resize the embeddings only when necessary to avoid index errors. If you are creating a model from scratch

13
examples/tensorflow/text-classification/run_glue.py
View File

@@ -180,6 +180,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
# endregion
@@ -298,6 +308,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
@@ -305,6 +316,7 @@ def main():
use_fast=model_args.use_fast_tokenizer,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# endregion
@@ -388,6 +400,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# endregion

14
examples/tensorflow/text-classification/run_text_classification.py
View File

@@ -186,6 +186,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
# endregion
@@ -315,6 +325,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
else:
config = AutoConfig.from_pretrained(
@@ -322,12 +333,14 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# endregion
@@ -416,6 +429,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# endregion

40
examples/tensorflow/token-classification/run_ner.py
View File

@@ -91,6 +91,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
@dataclass
@@ -304,9 +314,17 @@ def main():
# In distributed training, the .from_pretrained methods guarantee that only one local process can concurrently
# download model & vocab.
if model_args.config_name:
config = AutoConfig.from_pretrained(model_args.config_name, num_labels=num_labels)
config = AutoConfig.from_pretrained(
model_args.config_name,
num_labels=num_labels,
trust_remote_code=model_args.trust_remote_code,
)
elif model_args.model_name_or_path:
config = AutoConfig.from_pretrained(model_args.model_name_or_path, num_labels=num_labels)
config = AutoConfig.from_pretrained(
model_args.model_name_or_path,
num_labels=num_labels,
trust_remote_code=model_args.trust_remote_code,
)
else:
config = CONFIG_MAPPING[model_args.model_type]()
logger.warning("You are instantiating a new config instance from scratch.")
@@ -319,9 +337,18 @@ def main():
)
if config.model_type in {"gpt2", "roberta"}:
tokenizer = AutoTokenizer.from_pretrained(tokenizer_name_or_path, use_fast=True, add_prefix_space=True)
tokenizer = AutoTokenizer.from_pretrained(
tokenizer_name_or_path,
use_fast=True,
add_prefix_space=True,
trust_remote_code=model_args.trust_remote_code,
)
else:
tokenizer = AutoTokenizer.from_pretrained(tokenizer_name_or_path, use_fast=True)
tokenizer = AutoTokenizer.from_pretrained(
tokenizer_name_or_path,
use_fast=True,
trust_remote_code=model_args.trust_remote_code,
)
# endregion
# region Preprocessing the raw datasets
@@ -392,10 +419,13 @@ def main():
model = TFAutoModelForTokenClassification.from_pretrained(
model_args.model_name_or_path,
config=config,
trust_remote_code=model_args.trust_remote_code,
)
else:
logger.info("Training new model from scratch")
model = TFAutoModelForTokenClassification.from_config(config)
model = TFAutoModelForTokenClassification.from_config(
config, trust_remote_code=model_args.trust_remote_code
)
# We resize the embeddings only when necessary to avoid index errors. If you are creating a model from scratch
# on a small vocab and want a smaller embedding size, remove this test.

13
examples/tensorflow/translation/run_translation.py
View File

@@ -109,6 +109,16 @@ class ModelArguments:
"help": "The `use_auth_token` argument is deprecated and will be removed in v4.34. Please use `token`."
},
)
trust_remote_code: bool = field(
default=False,
metadata={
"help": (
"Whether or not to allow for custom models defined on the Hub in their own modeling files. This option"
"should only be set to `True` for repositories you trust and in which you have read the code, as it will"
"execute code present on the Hub on your local machine."
)
},
)
@dataclass
@@ -366,6 +376,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
tokenizer = AutoTokenizer.from_pretrained(
model_args.tokenizer_name if model_args.tokenizer_name else model_args.model_name_or_path,
@@ -373,6 +384,7 @@ def main():
use_fast=model_args.use_fast_tokenizer,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
prefix = data_args.source_prefix if data_args.source_prefix is not None else ""
@@ -480,6 +492,7 @@ def main():
cache_dir=model_args.cache_dir,
revision=model_args.model_revision,
token=model_args.token,
trust_remote_code=model_args.trust_remote_code,
)
# We resize the embeddings only when necessary to avoid index errors. If you are creating a model from scratch