Freigeben über


Databricks Runtime 12.0 für Machine Learning (nicht unterstützt)

Databricks Runtime 12.0 für Machine Learning bietet eine sofort einsatzbereite Umgebung für maschinelles Lernen und Data Science auf Basis von Databricks Runtime 12.0 (nicht unterstützt). Databricks Runtime ML enthält viele beliebte Machine Learning-Bibliotheken, einschließlich TensorFlow, PyTorch und XGBoost. Databricks Runtime ML enthält AutoML, ein Tool zum automatischen Trainieren von Machine Learning-Pipelines. Databricks Runtime ML unterstützt auch verteiltes Deep Learning-Training mit Horovod.

Weitere Informationen, einschließlich Anweisungen zum Erstellen eines Databricks Runtime ML-Clusters, finden Sie unter KI und Machine Learning in Databricks.

Neue Features und Verbesserungen

Databricks Runtime 12.0 ML basiert auf Databricks Runtime 12.0. Informationen zu den Neuerungen in Databricks Runtime 12.0 (einschließlich Apache Spark MLlib und SparkR) finden Sie in den Versionshinweisen zu Databricks Runtime 12.0 (nicht unterstützt).

Verbesserungen an Databricks AutoML

  • Vorhersagemodelle können jetzt optional nationale Feiertage einschließen.
  • Vorhersagen unterstützen jetzt die Häufigkeitsangaben „Monatlich“, „Vierteljährlich“ und „Jährlich“.
  • AutoML kann jetzt größere Datasets für das Training verwenden. AutoML weist automatisch weitere CPU-Kerne für große Datasets zu.

Weitere Informationen zu Databricks AutoML finden Sie unter Was ist AutoML?.

MLflow 2.0

Databricks Runtime 12.0 ML enthält MLflow 2.0. MLflow 2.0 baut auf der starken Plattformbasis von MLflow auf und berücksichtigt umfangreiches Benutzerfeedback, um Data Science-Workflows zu vereinfachen und innovative, erstklassige Tools für MLOps bereitzustellen. Zu den Features und Verbesserungen gehören Erweiterungen für MLflow-Rezepte (ehemals MLflow-Pipelines), z. B. AutoML, Hyperparameteroptimierung und Klassifizierungsunterstützung, sowie modernisierte Integrationen in das ML-Ökosystem, eine optimierte MLflow-Tracking-Benutzeroberfläche, eine Aktualisierung der Kern-APIs für die Plattformkomponenten von MLflow und vieles mehr. Weitere Informationen finden Sie in der Dokumentation zu MLflow 2.0 oder in diesem Blogbeitrag.

scikit-learn 1.0

Databricks Runtime ML 12.0 enthält Version 1.0 von scikit-learn. Lesen Sie die scikit-learn-Dokumentation, um mehr über Änderungen ab diesem Scikit-learn-Release zu erfahren.

Systemumgebung

Die Systemumgebung in Databricks Runtime 12.0 ML unterscheidet sich wie folgt von Databricks Runtime 12.0:

Databricks Runtime 12.0 ML enthält XGBoost 1.6.2, das keine GPU-Cluster mit Computefunktion 5.2 und niedriger unterstützt.

Bibliotheken

In den folgenden Abschnitten werden die in Databricks Runtime 12.0 ML enthaltenen Bibliotheken aufgeführt, die sich von denen unterscheiden, die in Databricks Runtime 12.0 enthalten sind.

Inhalt dieses Abschnitts:

Bibliotheken der obersten Ebene

Databricks Runtime 12.0 ML enthält die folgenden erstklassigen Bibliotheken:

Python-Bibliotheken

Databricks Runtime 12.0 ML verwendet Virtualenv für die Python-Paketverwaltung und enthält viele beliebte ML-Pakete.

Zusätzlich zu den Paketen, die in den folgenden Abschnitten aufgeführt sind, umfasst Databricks Runtime 12.0 ML auch die folgenden Pakete:

  • hyperopt 0.2.7.db1
  • sparkdl 2.3.0-db3
  • automl 1.14.1

Um die Databricks Runtime ML Python-Umgebung in Ihrer lokalen virtuellen Python-Umgebung zu reproduzieren, müssen Sie die Datei requirements-12.0.txt herunterladen und pip install -r requirements-12.0.txt ausführen. Dieser Befehl installiert alle Open-Source-Bibliotheken, die Databricks Runtime ML verwendet, installiert aber keine von Databricks entwickelten Bibliotheken wie databricks-automl, databricks-feature-store oder den Databricks-Fork von hyperopt.

Python-Bibliotheken in CPU-Clustern

Bibliothek Version Bibliothek Version Bibliothek Version
absl-py 1.0.0 argon2-cffi 21.3.0 argon2-cffi-bindings 21.2.0
astor 0.8.1 asttokens 2.0.5 astunparse 1.6.3
attrs 21.4.0 azure-core 1.26.1 azure-cosmos 4.2.0
backcall 0.2.0 backports.entry-points-selectable 1.2.0 bcrypt 3.2.0
beautifulsoup4 4.11.1 black 22.3.0 bleach 4.1.0
blis 0.7.9 boto3 1.21.32 botocore 1.24.32
cachetools 4.2.2 Katalog 2.0.8 category-encoders 2.5.1.post0
certifi 2021.10.8 cffi 1.15.0 chardet 4.0.0
charset-normalizer 2.0.4 Klicken 8.0.4 cloudpickle 2.0.0
cmdstanpy 1.0.8 confection 0.0.3 configparser 5.2.0
convertdate 2.4.0 cryptography 3.4.8 cycler 0.11.0
cymem 2.0.7 Cython 0.29.28 databricks-automl-runtime 0.2.13
databricks-cli 0.17.3 databricks-feature-store 0.8.0 dbl-tempo 0.1.12
dbus-python 1.2.16 debugpy 1.5.1 decorator 5.1.1
defusedxml 0.7.1 dill 0.3.4 diskcache 5.4.0
distlib 0.3.6 entrypoints 0,4 ephem 4.1.3
executing 0.8.3 facets-overview 1.0.0 fastjsonschema 2.16.2
fasttext 0.9.2 filelock 3.6.0 Flask 1.1.2
flatbuffers 22.10.26 fonttools 4.25.0 fsspec 2022.2.0
future 0.18.2 gast 0.4.0 gitdb 4.0.9
GitPython 3.1.27 google-auth 1.33.0 google-auth-oauthlib 0.4.6
google-pasta 0.2.0 grpcio 1.42.0 gunicorn 20.1.0
gviz-api 1.10.0 h5py 3.6.0 hijri-converter 2.2.4
holidays 0.16 horovod 0.25.0 htmlmin 0.1.12
huggingface-hub 0.11.0 idna 3.3 ImageHash 4.3.1
imbalanced-learn 0.8.1 importlib-metadata 4.11.3 ipykernel 6.15.3
ipython 8.5.0 ipython-genutils 0.2.0 ipywidgets 7.7.2
isodate 0.6.1 itsdangerous 2.0.1 jedi 0.18.1
Jinja2 2.11.3 jmespath 0.10.0 joblib 1.1.0
joblibspark 0.5.0 jsonschema 4.4.0 jupyter-client 6.1.12
jupyter_core 4.11.2 jupyterlab-pygments 0.1.2 jupyterlab-widgets 1.0.0
keras 2.10.0 Keras-Preprocessing 1.1.2 kiwisolver 1.3.2
korean-lunar-calendar 0.3.1 langcodes 3.3.0 libclang 14.0.6
lightgbm 3.3.3 llvmlite 0.38.0 LunarCalendar 0.0.9
Mako 1.2.0 Markdown 3.3.4 MarkupSafe 2.0.1
matplotlib 3.5.1 matplotlib-inline 0.1.2 missingno 0.5.1
mistune 0.8.4 mleap 0.20.0 mlflow-skinny 2.0.1
multimethod 1.8 murmurhash 1.0.9 mypy-extensions 0.4.3
nbclient 0.5.13 nbconvert 6.4.4 nbformat 5.3.0
nest-asyncio 1.5.5 networkx 2.7.1 nltk 3,7
Notebook 6.4.8 numba 0.55.1 numpy 1.21.5
oauthlib 3.2.0 opt-einsum 3.3.0 Packen 21,3
Pandas 1.4.2 pandas-profiling 3.3.0 pandocfilters 1.5.0
paramiko 2.9.2 parso 0.8.3 pathspec 0.9.0
pathy 0.6.1 patsy 0.5.2 petastorm 0.11.4
pexpect 4.8.0 phik 0.12.2 pickleshare 0.7.5
Pillow 9.0.1 pip 21.2.4 platformdirs 2.5.4
plotly 5.6.0 pmdarima 2.0.1 preshed 3.0.8
prometheus-client 0.13.1 prompt-toolkit 3.0.20 prophet 1.1.1
protobuf 3.19.4 psutil 5.8.0 psycopg2 2.9.3
ptyprocess 0.7.0 pure-eval 0.2.2 pyarrow 7.0.0
pyasn1 0.4.8 pyasn1-modules 0.2.8 pybind11 2.10.1
pycparser 2.21 pydantic 1.9.2 Pygments 2.11.2
PyGObject 3.36.0 PyJWT 2.6.0 PyMeeus 0.5.11
PyNaCl 1.5.0 pyodbc 4.0.32 pyparsing 3.0.4
pyrsistent 0.18.0 Python-dateutil 2.8.2 python-editor 1.0.4
pytz 2021.3 PyWavelets 1.3.0 PyYAML 6,0
pyzmq 22.3.0 regex 2022.3.15 requests 2.27.1
requests-oauthlib 1.3.1 requests-unixsocket 0.2.0 rsa 4.7.2
s3transfer 0.5.0 scikit-learn 1.0.2 scipy 1.7.3
seaborn 0.11.2 Send2Trash 1.8.0 setuptools 61.2.0
setuptools-git 1.2 shap 0.41.0 simplejson 3.17.6
sechs 1.16.0 slicer 0.0.7 smart-open 5.1.0
smmap 5.0.0 soupsieve 2.3.1 spacy 3.4.1
spacy-legacy 3.0.10 spacy-loggers 1.0.3 spark-tensorflow-distributor 1.0.0
sqlparse 0.4.2 srsly 2.4.5 ssh-import-id 5.10
stack-data 0.2.0 statsmodels 0.13.2 tabulate 0.8.9
tangled-up-in-unicode 0.2.0 tenacity 8.0.1 tensorboard 2.10.0
tensorboard-data-server 0.6.1 tensorboard-plugin-profile 2.8.0 tensorboard-plugin-wit 1.8.1
tensorflow-cpu 2.10.0 tensorflow-estimator 2.10.0 tensorflow-io-gcs-filesystem 0.28.0
termcolor 2.1.1 terminado 0.13.1 testpath 0.5.0
thinc 8.1.5 threadpoolctl 2.2.0 tokenize-rt 4.2.1
tokenizers 0.13.2 tomli 1.2.2 torch 1.12.1+cpu
torchvision 0.13.1+cpu tornado 6.1 tqdm 4.64.0
traitlets 5.1.1 transformers 4.23.1 typer 0.4.2
typing_extensions 4.1.1 unattended-upgrades 0,1 urllib3 1.26.9
virtualenv 20.8.0 Visionen 0.7.5 Wasabi 0.10.1
wcwidth 0.2.5 webencodings 0.5.1 websocket-client 0.58.0
Werkzeug 2.0.3 wheel 0.37.1 widgetsnbextension 3.6.1
wrapt 1.12.1 zipp 3.7.0

Python-Bibliotheken für GPU-Cluster

Bibliothek Version Bibliothek Version Bibliothek Version
absl-py 1.0.0 argon2-cffi 21.3.0 argon2-cffi-bindings 21.2.0
astor 0.8.1 asttokens 2.0.5 astunparse 1.6.3
attrs 21.4.0 azure-core 1.26.1 azure-cosmos 4.2.0
backcall 0.2.0 backports.entry-points-selectable 1.2.0 bcrypt 3.2.0
beautifulsoup4 4.11.1 black 22.3.0 bleach 4.1.0
blis 0.7.9 boto3 1.21.32 botocore 1.24.32
cachetools 4.2.2 Katalog 2.0.8 category-encoders 2.5.1.post0
certifi 2021.10.8 cffi 1.15.0 chardet 4.0.0
charset-normalizer 2.0.4 Klicken 8.0.4 cloudpickle 2.0.0
cmdstanpy 1.0.8 confection 0.0.3 configparser 5.2.0
convertdate 2.4.0 cryptography 3.4.8 cycler 0.11.0
cymem 2.0.7 Cython 0.29.28 databricks-automl-runtime 0.2.13
databricks-cli 0.17.3 databricks-feature-store 0.8.0 dbl-tempo 0.1.12
dbus-python 1.2.16 debugpy 1.5.1 decorator 5.1.1
defusedxml 0.7.1 dill 0.3.4 diskcache 5.4.0
distlib 0.3.6 entrypoints 0,4 ephem 4.1.3
executing 0.8.3 facets-overview 1.0.0 fastjsonschema 2.16.2
fasttext 0.9.2 filelock 3.6.0 Flask 1.1.2
flatbuffers 22.10.26 fonttools 4.25.0 fsspec 2022.2.0
future 0.18.2 gast 0.4.0 gitdb 4.0.9
GitPython 3.1.27 google-auth 1.33.0 google-auth-oauthlib 0.4.6
google-pasta 0.2.0 grpcio 1.42.0 gunicorn 20.1.0
gviz-api 1.10.0 h5py 3.6.0 hijri-converter 2.2.4
holidays 0.16 horovod 0.25.0 htmlmin 0.1.12
huggingface-hub 0.11.0 idna 3.3 ImageHash 4.3.1
imbalanced-learn 0.8.1 importlib-metadata 4.11.3 ipykernel 6.15.3
ipython 8.5.0 ipython-genutils 0.2.0 ipywidgets 7.7.2
isodate 0.6.1 itsdangerous 2.0.1 jedi 0.18.1
Jinja2 2.11.3 jmespath 0.10.0 joblib 1.1.0
joblibspark 0.5.0 jsonschema 4.4.0 jupyter-client 6.1.12
jupyter_core 4.11.2 jupyterlab-pygments 0.1.2 jupyterlab-widgets 1.0.0
keras 2.10.0 Keras-Preprocessing 1.1.2 kiwisolver 1.3.2
korean-lunar-calendar 0.3.1 langcodes 3.3.0 libclang 14.0.6
lightgbm 3.3.3 llvmlite 0.38.0 LunarCalendar 0.0.9
Mako 1.2.0 Markdown 3.3.4 MarkupSafe 2.0.1
matplotlib 3.5.1 matplotlib-inline 0.1.2 missingno 0.5.1
mistune 0.8.4 mleap 0.20.0 mlflow-skinny 2.0.1
multimethod 1.8 murmurhash 1.0.9 mypy-extensions 0.4.3
nbclient 0.5.13 nbconvert 6.4.4 nbformat 5.3.0
nest-asyncio 1.5.5 networkx 2.7.1 nltk 3,7
Notebook 6.4.8 numba 0.55.1 numpy 1.21.5
oauthlib 3.2.0 opt-einsum 3.3.0 Packen 21,3
Pandas 1.4.2 pandas-profiling 3.3.0 pandocfilters 1.5.0
paramiko 2.9.2 parso 0.8.3 pathspec 0.9.0
pathy 0.6.1 patsy 0.5.2 petastorm 0.11.4
pexpect 4.8.0 phik 0.12.2 pickleshare 0.7.5
Pillow 9.0.1 pip 21.2.4 platformdirs 2.5.4
plotly 5.6.0 pmdarima 2.0.1 preshed 3.0.8
prompt-toolkit 3.0.20 prophet 1.1.1 protobuf 3.19.4
psutil 5.8.0 psycopg2 2.9.3 ptyprocess 0.7.0
pure-eval 0.2.2 pyarrow 7.0.0 pyasn1 0.4.8
pyasn1-modules 0.2.8 pybind11 2.10.1 pycparser 2.21
pydantic 1.9.2 Pygments 2.11.2 PyGObject 3.36.0
PyJWT 2.6.0 PyMeeus 0.5.11 PyNaCl 1.5.0
pyodbc 4.0.32 pyparsing 3.0.4 pyrsistent 0.18.0
Python-dateutil 2.8.2 python-editor 1.0.4 pytz 2021.3
PyWavelets 1.3.0 PyYAML 6,0 pyzmq 22.3.0
regex 2022.3.15 requests 2.27.1 requests-oauthlib 1.3.1
requests-unixsocket 0.2.0 rsa 4.7.2 s3transfer 0.5.0
scikit-learn 1.0.2 scipy 1.7.3 seaborn 0.11.2
Send2Trash 1.8.0 setuptools 61.2.0 setuptools-git 1.2
shap 0.41.0 simplejson 3.17.6 sechs 1.16.0
slicer 0.0.7 smart-open 5.1.0 smmap 5.0.0
soupsieve 2.3.1 spacy 3.4.1 spacy-legacy 3.0.10
spacy-loggers 1.0.3 spark-tensorflow-distributor 1.0.0 sqlparse 0.4.2
srsly 2.4.5 ssh-import-id 5.10 stack-data 0.2.0
statsmodels 0.13.2 tabulate 0.8.9 tangled-up-in-unicode 0.2.0
tenacity 8.0.1 tensorboard 2.10.0 tensorboard-data-server 0.6.1
tensorboard-plugin-profile 2.8.0 tensorboard-plugin-wit 1.8.1 tensorflow 2.10.0
tensorflow-estimator 2.10.0 tensorflow-io-gcs-filesystem 0.28.0 termcolor 2.1.1
terminado 0.13.1 testpath 0.5.0 thinc 8.1.5
threadpoolctl 2.2.0 tokenize-rt 4.2.1 tokenizers 0.13.2
tomli 1.2.2 torch 1.12.1+cu113 torchvision 0.13.1+cu113
tornado 6.1 tqdm 4.64.0 traitlets 5.1.1
transformers 4.23.1 typer 0.4.2 typing_extensions 4.1.1
unattended-upgrades 0,1 urllib3 1.26.9 virtualenv 20.8.0
Visionen 0.7.5 Wasabi 0.10.1 wcwidth 0.2.5
webencodings 0.5.1 websocket-client 0.58.0 Werkzeug 2.0.3
wheel 0.37.1 widgetsnbextension 3.6.1 wrapt 1.12.1
zipp 3.7.0

R-Bibliotheken

Die R-Bibliotheken sind identisch mit den R-Bibliotheken in Databricks Runtime 12.0.

Java- und Scala-Bibliotheken (Scala 2.12-Cluster)

Zusätzlich zu den Java- und Scala-Bibliotheken in Databricks Runtime 12.0 enthält Databricks Runtime 12.0 ML die folgenden JARs:

CPU-Cluster

Gruppen-ID Artefakt-ID Version
com.typesafe.akka akka-actor_2.12 2.5.23
ml.combust.mleap mleap-databricks-runtime_2.12 v0.20.0-db1
ml.dmlc xgboost4j-spark_2.12 1.6.2
ml.dmlc xgboost4j_2.12 1.6.2
org.graphframes graphframes_2.12 0.8.2-db1-spark3.2
org.mlflow mlflow-client 2.0.1
org.scala-lang.modules scala-java8-compat_2.12 0.8.0
org.tensorflow spark-tensorflow-connector_2.12 1.15.0

GPU-Cluster

Gruppen-ID Artefakt-ID Version
com.typesafe.akka akka-actor_2.12 2.5.23
ml.combust.mleap mleap-databricks-runtime_2.12 v0.20.0-db1
ml.dmlc xgboost4j-gpu_2.12 1.6.2
ml.dmlc xgboost4j-spark-gpu_2.12 1.6.2
org.graphframes graphframes_2.12 0.8.2-db1-spark3.2
org.mlflow mlflow-client 2.0.1
org.scala-lang.modules scala-java8-compat_2.12 0.8.0
org.tensorflow spark-tensorflow-connector_2.12 1.15.0