Skip to content

fix: add pytorch 1.8.1 for huggingface #2642

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
merged 19 commits into from
Sep 23, 2021
Merged

fix: add pytorch 1.8.1 for huggingface #2642

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
19 commits
Select commit Hold shift + click to select a range
f7e91ef
added pytorch 1.8.1 for supporting huggingface
Sep 14, 2021
a81cfbf
fix: add pytorch 1.8.1 for huggingface
Sep 15, 2021
bca9414
Merge branch 'master' into huggingface_cu111
ahsan-z-khan Sep 15, 2021
e95e0ce
fix: add pytorch 1.8.1 for huggingface
Sep 15, 2021
be2c875
refactored code for flake
Sep 15, 2021
5ea64cc
refactored code for docstyle
Sep 15, 2021
6c134b9
fix: add alias for pytorch 1.8
Sep 16, 2021
31c20ab
Merge branch 'master' into huggingface_cu111
jeniyat Sep 16, 2021
d4cc967
Merge branch 'master' into huggingface_cu111
jeniyat Sep 20, 2021
f8f5483
Merge branch 'master' into huggingface_cu111
ahsan-z-khan Sep 20, 2021
015888c
update to master
Sep 20, 2021
90097d3
add version alias for pytorch1.8 in huggingface.json
Sep 20, 2021
613fa6a
add version alias for pytorch1.8 in huggingface.json
Sep 20, 2021
28f5ac1
removed empty file, corrected grammar
Sep 21, 2021
194025a
removed empty line in test_estimator
Sep 21, 2021
86b70be
removed empty line in test_estimator
Sep 21, 2021
40db98e
removed empty line in test_estimator
Sep 21, 2021
89c1145
Merge branch 'master' into huggingface_cu111
shreyapandit Sep 21, 2021
ca1e896
Merge branch 'master' into huggingface_cu111
ahsan-z-khan Sep 22, 2021
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
80 changes: 75 additions & 5 deletions src/sagemaker/image_uri_config/huggingface.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -147,6 +147,7 @@
"version_aliases": {
"pytorch1.6": "pytorch1.6.0",
"pytorch1.7": "pytorch1.7.1",
"pytorch1.8": "pytorch1.8.1",
"tensorflow2.4": "tensorflow2.4.1"
},
"pytorch1.6.0": {
Expand Down Expand Up @@ -178,7 +179,8 @@
"us-west-1": "763104351884",
"us-west-2": "763104351884"
},
"repository": "huggingface-pytorch-training"
"repository": "huggingface-pytorch-training",
"container_version": {"gpu":"cu110-ubuntu18.04"}
},
"pytorch1.7.1": {
"py_versions": ["py36"],
Expand Down Expand Up @@ -209,7 +211,40 @@
"us-west-1": "763104351884",
"us-west-2": "763104351884"
},
"repository": "huggingface-pytorch-training"
"repository": "huggingface-pytorch-training",
"container_version": {"gpu":"cu110-ubuntu18.04"}
},
"pytorch1.8.1": {
"py_versions": ["py36"],
"registries": {
"af-south-1": "626614931356",
"ap-east-1": "871362719292",
"ap-northeast-1": "763104351884",
"ap-northeast-2": "763104351884",
"ap-northeast-3": "364406365360",
"ap-south-1": "763104351884",
"ap-southeast-1": "763104351884",
"ap-southeast-2": "763104351884",
"ca-central-1": "763104351884",
"cn-north-1": "727897471807",
"cn-northwest-1": "727897471807",
"eu-central-1": "763104351884",
"eu-north-1": "763104351884",
"eu-west-1": "763104351884",
"eu-west-2": "763104351884",
"eu-west-3": "763104351884",
"eu-south-1": "692866216735",
"me-south-1": "217643126080",
"sa-east-1": "763104351884",
"us-east-1": "763104351884",
"us-east-2": "763104351884",
"us-gov-west-1": "442386744353",
"us-iso-east-1": "886529160074",
"us-west-1": "763104351884",
"us-west-2": "763104351884"
},
"repository": "huggingface-pytorch-training",
"container_version": {"gpu":"cu111-ubuntu18.04"}
},
"tensorflow2.4.1": {
"py_versions": ["py37"],
Expand Down Expand Up @@ -240,7 +275,8 @@
"us-west-1": "763104351884",
"us-west-2": "763104351884"
},
"repository": "huggingface-tensorflow-training"
"repository": "huggingface-tensorflow-training",
"container_version": {"gpu":"cu110-ubuntu18.04"}
}
}
}
Expand Down Expand Up @@ -286,7 +322,40 @@
"us-west-1": "763104351884",
"us-west-2": "763104351884"
},
"repository": "huggingface-pytorch-inference"
"repository": "huggingface-pytorch-inference",
"container_version": {"gpu":"cu110-ubuntu18.04", "cpu":"ubuntu18.04" }
},
"pytorch1.8.1": {
"py_versions": ["py36"],
"registries": {
"af-south-1": "626614931356",
"ap-east-1": "871362719292",
"ap-northeast-1": "763104351884",
"ap-northeast-2": "763104351884",
"ap-northeast-3": "364406365360",
"ap-south-1": "763104351884",
"ap-southeast-1": "763104351884",
"ap-southeast-2": "763104351884",
"ca-central-1": "763104351884",
"cn-north-1": "727897471807",
"cn-northwest-1": "727897471807",
"eu-central-1": "763104351884",
"eu-north-1": "763104351884",
"eu-west-1": "763104351884",
"eu-west-2": "763104351884",
"eu-west-3": "763104351884",
"eu-south-1": "692866216735",
"me-south-1": "217643126080",
"sa-east-1": "763104351884",
"us-east-1": "763104351884",
"us-east-2": "763104351884",
"us-gov-west-1": "442386744353",
"us-iso-east-1": "886529160074",
"us-west-1": "763104351884",
"us-west-2": "763104351884"
},
"repository": "huggingface-pytorch-inference",
"container_version": {"gpu":"cu111-ubuntu18.04", "cpu":"ubuntu18.04" }
},
"tensorflow2.4.1": {
"py_versions": ["py37"],
Expand Down Expand Up @@ -317,7 +386,8 @@
"us-west-1": "763104351884",
"us-west-2": "763104351884"
},
"repository": "huggingface-tensorflow-inference"
"repository": "huggingface-tensorflow-inference",
"container_version": {"gpu":"cu110-ubuntu18.04", "cpu":"ubuntu18.04" }
}
}
}
Expand Down
20 changes: 18 additions & 2 deletions src/sagemaker/image_uris.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,6 +41,9 @@ def retrieve(
):
"""Retrieves the ECR URI for the Docker image matching the given arguments.

Ideally this function should not be called directly, rather it should be called from the
fit() function inside framework estimator.

Args:
framework (str): The name of the framework or algorithm.
region (str): The AWS region.
Expand All @@ -56,7 +59,11 @@ def retrieve(
image_scope (str): The image type, i.e. what it is used for.
Valid values: "training", "inference", "eia". If ``accelerator_type`` is set,
``image_scope`` is ignored.
container_version (str): the version of docker image
container_version (str): the version of docker image.
Ideally the value of parameter should be created inside the framework.
For custom use, see the list of supported container versions:
https://github.com/aws/deep-learning-containers/blob/master/available_images.md
(default: None).
distribution (dict): A dictionary with information on how to run distributed training
(default: None).

Expand All @@ -66,10 +73,12 @@ def retrieve(
Raises:
ValueError: If the combination of arguments specified is not supported.
"""

config = _config_for_framework_and_scope(framework, image_scope, accelerator_type)
original_version = version
version = _validate_version_and_set_if_needed(version, config, framework)
version_config = config["versions"][_version_for_config(version, config)]

if framework == HUGGING_FACE_FRAMEWORK:
if version_config.get("version_aliases"):
full_base_framework_version = version_config["version_aliases"].get(
Expand All @@ -81,7 +90,6 @@ def retrieve(

py_version = _validate_py_version_and_set_if_needed(py_version, version_config, framework)
version_config = version_config.get(py_version) or version_config

registry = _registry_from_region(region, version_config["registries"])
hostname = utils._botocore_resolver().construct_endpoint("ecr", region)["hostname"]

Expand All @@ -91,11 +99,16 @@ def retrieve(
instance_type, config.get("processors") or version_config.get("processors")
)

# if container version is available in .json file, utilize that
if version_config.get("container_version"):
container_version = version_config["container_version"][processor]

if framework == HUGGING_FACE_FRAMEWORK:
pt_or_tf_version = (
re.compile("^(pytorch|tensorflow)(.*)$").match(base_framework_version).group(2)
)
tag_prefix = f"{pt_or_tf_version}-transformers{original_version}"

else:
tag_prefix = version_config.get("tag_prefix", version)

Expand All @@ -105,6 +118,7 @@ def retrieve(
py_version,
container_version,
)

if _should_auto_select_container_version(instance_type, distribution):
container_versions = {
"tensorflow-2.3-gpu-py37": "cu110-ubuntu18.04-v3",
Expand All @@ -120,7 +134,9 @@ def retrieve(
"pytorch-1.6-gpu-py3": "cu110-ubuntu18.04-v3",
"pytorch-1.6.0-gpu-py3": "cu110-ubuntu18.04",
}

key = "-".join([framework, tag])

if key in container_versions:
tag = "-".join([tag, container_versions[key]])

Expand Down
2 changes: 1 addition & 1 deletion tests/conftest.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -400,7 +400,7 @@ def _huggingface_base_fm_version(huggingface_version, base_fw, fixture_prefix):
if len(original_version.split(".")) == 2:
base_fw_version = ".".join(base_fw_version.split(".")[:-1])
versions.append(base_fw_version)
return versions
return sorted(versions, reverse=True)


def _generate_huggingface_base_fw_latest_versions(
Expand Down