Skip to content

[Pipelines] Add revision tag to all default pipelines #17667

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
patrickvonplaten merged 10 commits into from
Jun 30, 2022
Merged

[Pipelines] Add revision tag to all default pipelines #17667

Show file tree
Hide file tree
Changes from 2 commits
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
d54a4cb
trigger test failure
patrickvonplaten Jun 10, 2022
ca6a820
upload revision poc
patrickvonplaten Jun 10, 2022
3623fed
Update src/transformers/pipelines/base.py
patrickvonplaten Jun 21, 2022
38d686d
up
patrickvonplaten Jun 24, 2022
ab5324e
add test
patrickvonplaten Jun 30, 2022
618aee5
Merge branch 'main' of https://github.com/huggingface/transformers in…
patrickvonplaten Jun 30, 2022
2fcd9b6
Merge branch 'revision_tags_for_default_pipeline' of https://github.c…
patrickvonplaten Jun 30, 2022
24205b8
correct some stuff
patrickvonplaten Jun 30, 2022
4e4ff87
Update src/transformers/pipelines/__init__.py
patrickvonplaten Jun 30, 2022
f8ef027
correct require flag
patrickvonplaten Jun 30, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
11 changes: 7 additions & 4 deletions src/transformers/pipelines/__init__.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,7 +41,7 @@
Pipeline,
PipelineDataFormat,
PipelineException,
get_default_model,
get_default_model_and_revision,
infer_framework_load_model,
)
from .conversational import Conversation, ConversationalPipeline
Expand Down Expand Up @@ -127,7 +127,7 @@
"impl": AudioClassificationPipeline,
"tf": (),
"pt": (AutoModelForAudioClassification,) if is_torch_available() else (),
"default": {"model": {"pt": "superb/wav2vec2-base-superb-ks"}},
"default": {"model": {"pt": ("superb/wav2vec2-base-superb-ks", "372e048")}},
"type": "audio",
},
"automatic-speech-recognition": {
Expand Down Expand Up @@ -528,8 +528,11 @@ def pipeline(
# Use default model/config/tokenizer for the task if no model is provided
if model is None:
# At that point framework might still be undetermined
model = get_default_model(targeted_task, framework, task_options)
logger.warning(f"No model was supplied, defaulted to {model} (https://huggingface.co/{model})")
model, default_revision = get_default_model_and_revision(targeted_task, framework, task_options)
revision = revision if revision is not None else default_revision
logger.warning(
f"No model was supplied, defaulted to {model} and revision {revision} (https://huggingface.co/{model})"
Comment thread
patrickvonplaten marked this conversation as resolved.
Outdated
)

# Retrieve use_auth_token and add it to model_kwargs to be used in .from_pretrained
model_kwargs["use_auth_token"] = model_kwargs.get("use_auth_token", use_auth_token)
Expand Down
2 changes: 1 addition & 1 deletion src/transformers/pipelines/base.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -341,7 +341,7 @@ def get_framework(model, revision: Optional[str] = None):
return framework


def get_default_model(targeted_task: Dict, framework: Optional[str], task_options: Optional[Any]) -> str:
def get_default_model_and_revision(targeted_task: Dict, framework: Optional[str], task_options: Optional[Any]) -> str:
Comment thread
patrickvonplaten marked this conversation as resolved.
Outdated
"""
Select a default model to use for a given task. Defaults to pytorch if ambiguous.

Expand Down