Databricks Runtime 10.3 för ML (stöds inte)
Databricks Runtime 10.3 för Machine Learning ger en färdig miljö för maskininlärning och datavetenskap baserat på Databricks Runtime 10.3 (stöds inte). Databricks Runtime ML innehåller många populära maskininlärningsbibliotek, inklusive TensorFlow, PyTorch och XGBoost. Databricks Runtime ML innehåller AutoML, ett verktyg för att automatiskt träna maskininlärningspipelines. Databricks Runtime ML stöder även distribuerad djupinlärningsträning med Horovod.
Mer information, inklusive instruktioner för att skapa ett Databricks Runtime ML-kluster, finns i AI och Machine Learning på Databricks.
Nya funktioner och förbättringar
Databricks Runtime 10.3 ML bygger på Databricks Runtime 10.3. Information om nyheter i Databricks Runtime 10.3, inklusive Apache Spark MLlib och SparkR, finns i viktig information om Databricks Runtime 10.3 (stöds inte).
Förbättringar av Databricks AutoML
Följande förbättringar har gjorts i Databricks AutoML.
AutoML har nu stöd för ARIMA-modell för prognostisering
Förutom Prophet skapar och utvärderar AutoML nu ARIMA-modeller för prognostiseringsproblem.
Exkludera kolumner från datauppsättning
När du använder AutoML-API:et kan du ange kolumner som AutoML ska ignorera under dess beräkningar. Detta är endast tillgängligt för klassificerings- och regressionsproblem. Mer information finns i Parametrar för klassificering och regression .
Undanta algoritmramverk från en AutoML-körning
Du kan ange algoritmramverk, till exempel scikit-learn, som AutoML inte bör tänka på när det utvecklar modeller. Mer information finns i Avancerade konfigurationer och parametrar för klassificering och regression.
max_trials
Deprecated
Parametern max_trials
är inaktuell och tas bort i nästa större Databricks Runtime ML-version. Använd timeout_minutes
för att styra varaktigheten för en AutoML-körning. I Databricks Runtime 10.1 ML och senare innehåller AutoML även tidig stoppning. det stoppar tränings- och justeringsmodeller om valideringsmåttet inte längre förbättras.
Förbättringar av Databricks-funktionsarkivet
Nu kan du använda punkt-i-tid-sökningar för funktionstabeller för tidsserier. Mer information finns i Använda tidsseriefunktionstabeller med stöd för tidpunkter.
Automatisk loggning av Databricks (GA)
Databricks Autologging är nu allmänt tillgänglig i Databricks Runtime 10.3 ML. Databricks Autologging är en lösning utan kod som tillhandahåller automatisk experimentspårning för maskininlärningsträningssessioner i Azure Databricks. Med Databricks Autologging registreras modellparametrar, mått, filer och härkomstinformation automatiskt när du tränar modeller från en mängd populära maskininlärningsbibliotek. Utbildningssessioner registreras som MLflow Tracking Runs. Modellfiler spåras också så att du enkelt kan logga dem till MLflow Model Registry och distribuera dem för realtidsbedömning med MLflow Model Serving.
Mer information finns i Automatisk loggning av Databricks.
Systemmiljö
Systemmiljön i Databricks Runtime 10.3 ML skiljer sig från Databricks Runtime 10.3 på följande sätt:
- DBUtils: Databricks Runtime ML innehåller inte biblioteksverktyget (dbutils.library) (äldre).
Använd
%pip
kommandon i stället. Se Python-bibliotek med notebook-omfång. - För GPU-kluster innehåller Databricks Runtime ML följande NVIDIA GPU-bibliotek:
- CUDA 11.0
- cuDNN 8.0.5.39
- NCCL 2.10.3
- TensorRT 7.2.2
Bibliotek
I följande avsnitt visas de bibliotek som ingår i Databricks Runtime 10.3 ML som skiljer sig från de som ingår i Databricks Runtime 10.3.
I detta avsnitt:
- Bibliotek på den översta nivån
- Python-bibliotek
- R-bibliotek
- Java- och Scala-bibliotek (Scala 2.12-kluster)
Bibliotek på den översta nivån
Databricks Runtime 10.3 ML innehåller följande bibliotek på den översta nivån:
- GraphFrames
- Horovod och HorovodRunner
- MLflow
- PyTorch
- spark-tensorflow-connector
- TensorFlow
- TensorBoard
Python-bibliotek
Databricks Runtime 10.3 ML använder Virtualenv för Python-pakethantering och innehåller många populära ML-paket.
Förutom de paket som anges i följande avsnitt innehåller Databricks Runtime 10.3 ML även följande paket:
- hyperopt 0.2.7.db1
- sparkdl 2.2.0-db5
- feature_store 0.3.7
- automl 1.6.0
Python-bibliotek i CPU-kluster
Bibliotek | Version | Bibliotek | Version | Bibliotek | Version |
---|---|---|---|---|---|
absl-py | 0.11.0 | Antergos Linux | 2015.10 (ISO-Rullande) | appdirs | 1.4.4 |
argon2-cffi | 20.1.0 | Astor | 0.8.1 | astunparse | 1.6.3 |
async-generator | 1.10 | attrs | 20.3.0 | backcall | 0.2.0 |
Bcrypt | 3.2.0 | bidict | 0.21.4 | Blekmedel | 3.3.0 |
blis | 0.7.4 | boto3 | 1.16.7 | botocore | 1.19.7 |
cachetools | 4.2.4 | Katalog | 2.0.6 | Certifi | 2020.12.5 |
Cffi | 1.14.5 | Chardet | 4.0.0 | klicka | 7.1.2 |
cloudpickle | 1.6.0 | cmdstanpy | 0.9.68 | configparser | 5.0.1 |
konvertera | 2.3.2 | Kryptografi | 3.4.7 | Apparat | 0.10.0 |
cymem | 2.0.5 | Cython | 0.29.23 | databricks-automl-runtime | 0.2.5 |
databricks-cli | 0.16.2 | dbl-tempo | 0.1.2 | dbus-python | 1.2.16 |
Dekoratör | 5.0.6 | defusedxml | 0.7.1 | Dill | 0.3.2 |
diskcache | 5.2.1 | distlib | 0.3.4 | distro-info | 0.23ubuntu1 |
entrypoints | 0,3 | efem | 4.1.3 | facets-overview | 1.0.0 |
snabbtext | 0.9.2 | filelock | 3.0.12 | Flask | 1.1.2 |
flatbuffers | 2.0 | fsspec | 0.9.0 | Framtiden | 0.18.2 |
Gast | 0.4.0 | gitdb | 4.0.7 | GitPython | 3.1.12 |
google-auth | 1.22.1 | google-auth-oauthlib | 0.4.2 | google-pasta | 0.2.0 |
grpcio | 1.39.0 | gunicorn | 20.0.4 | gviz-api | 1.10.0 |
h5py | 3.1.0 | hijri-converter | 2.2.2 | Semester | 0,12 |
horovod | 0.23.0 | htmlmin | 0.1.12 | huggingface-hub | 0.1.2 |
idna | 2.10 | ImageHash | 4.2.1 | obalanserad inlärning | 0.8.1 |
importlib-metadata | 3.10.0 | ipykernel | 5.3.4 | Ipython | 7.22.0 |
ipython-genutils | 0.2.0 | ipywidgets | 7.6.3 | isodate | 0.6.0 |
itsdangerous | 1.1.0 | Jedi | 0.17.2 | Jinja2 | 2.11.3 |
jmespath | 0.10.0 | joblib | 1.0.1 | joblibspark | 0.3.0 |
jsonschema | 3.2.0 | jupyter-client | 6.1.12 | jupyter-core | 4.7.1 |
jupyterlab-pygments | 0.1.2 | jupyterlab-widgets | 1.0.0 | keras | 2.7.0 |
Keras-förbearbetning | 1.1.2 | kiwisolver | 1.3.1 | Koalas | 1.8.2 |
korean-lunar-calendar | 0.2.1 | langcodes | 3.3.0 | libclang | 12.0.0 |
lightgbm | 3.3.1 | llvmlite | 0.38.0 | LunarCalendar | 0.0.9 |
Mako | 1.1.3 | Markdown | 3.3.3 | Markering Valv | 2.0.1 |
matplotlib | 3.4.2 | missingno | 0.5.0 | mistune | 0.8.4 |
mleap | 0.18.1 | mlflow-skinny | 1.23.0 | multimethod | 1.6 |
murmurhash | 1.0.5 | nbclient | 0.5.3 | nbconvert | 6.0.7 |
nbformat | 5.1.3 | nest-asyncio | 1.5.1 | Networkx | 2.5 |
nltk | 3.6.1 | notebook-fil | 6.3.0 | numba | 0.55.0 |
Numpy | 1.20.1 | oauthlib | 3.1.0 | opt-einsum | 3.3.0 |
Förpackning | 21,3 | Pandas | 1.2.4 | pandas-profilering | 3.1.0 |
pandocfilter | 1.4.3 | Paramiko | 2.7.2 | parso | 0.7.0 |
Pati | 0.6.0 | Patsy | 0.5.1 | petastorm | 0.11.3 |
Pexpect | 4.8.0 | phik | 0.12.0 | pickleshare | 0.7.5 |
Kudde | 8.2.0 | Pip | 21.0.1 | plotly | 5.5.0 |
pmdarima | 1.8.4 | fördefinierad | 3.0.5 | prometheus-client | 0.10.1 |
prompt-toolkit | 3.0.17 | Profeten | 1.0.1 | protobuf | 3.17.2 |
Psutil | 5.8.0 | psycopg2 | 2.8.5 | ptyprocess | 0.7.0 |
pyarrow | 4.0.0 | pyasn1 | 0.4.8 | pyasn1-modules | 0.2.8 |
pybind11 | 2.9.0 | pycparser | 2,20 | pydantisk | 1.8.2 |
Pygments | 2.8.1 | Pygobject | 3.36.0 | PyMeeus | 0.5.11 |
PyNaCl | 1.4.0 | pyodbc | 4.0.30 | Pyparsing | 2.4.7 |
pyrsistent | 0.17.3 | pystan | 2.19.1.1 | python-apt | 2.0.0+ubuntu0.20.4.6 |
python-dateutil | 2.8.1 | python-editor | 1.0.4 | python-engineio | 4.3.0 |
python-socketio | 5.4.1 | pytz | 2020.5 | PyWavelets | 1.1.1 |
PyYAML | 5.4.1 | pyzmq | 20.0.0 | Regex | 2021.4.4 |
begäranden | 2.25.1 | requests-oauthlib | 1.3.0 | requests-unixsocket | 0.2.0 |
Rsa | 4.7.2 | s3transfer | 0.3.7 | sacremoses | 0.0.46 |
scikit-learn | 0.24.1 | Scipy | 1.6.2 | seaborn | 0.11.1 |
Send2Trash | 1.5.0 | Setuptools | 52.0.0 | setuptools-git | 1.2 |
Shap | 0.40.0 | Simplejson | 3.17.2 | Sex | 1.15.0 |
Slicer | 0.0.7 | smart-open | 5.2.0 | smmap | 3.0.5 |
spacy | 3.2.1 | spacy-legacy | 3.0.8 | spacy-loggers | 1.0.1 |
spark-tensorflow-distributor | 1.0.0 | sqlparse | 0.4.1 | srsly | 2.4.1 |
ssh-import-id | 5,10 | statsmodels | 0.12.2 | Tabellform | 0.8.7 |
trasslig upp-i-unicode | 0.1.0 | Uthållighet | 6.2.0 | tensorboard | 2.7.0 |
tensorboard-data-server | 0.6.1 | tensorboard-plugin-profile | 2.5.0 | tensorboard-plugin-wit | 1.8.1 |
tensorflow-cpu | 2.7.0 | tensorflow-estimator | 2.7.0 | tensorflow-io-gcs-filesystem | 0.23.1 |
termcolor | 1.1.0 | terminado | 0.9.4 | testpath | 0.4.4 |
thinc | 8.0.12 | threadpoolctl | 2.1.0 | tokenizers | 0.10.3 |
Fackla | 1.10.1+cpu | torchvision | 0.11.2+cpu | Tornado | 6.1 |
tqdm | 4.59.0 | traitlets | 5.0.5 | Transformatorer | 4.15.0 |
typer | 0.3.2 | skriva tillägg | 3.7.4.3 | ujson | 4.0.2 |
obevakade uppgraderingar | 0,1 | urllib3 | 1.25.11 | Virtualenv | 20.4.1 |
Visioner | 0.7.4 | Wasabi | 0.8.2 | wcwidth | 0.2.5 |
webencodings | 0.5.1 | websocket-client | 0.57.0 | Werkzeug | 1.0.1 |
Hjul | 0.36.2 | widgetsnbextension | 3.5.1 | wrapt | 1.12.1 |
xgboost | 1.5.1 | Zipp | 3.4.1 |
Python-bibliotek i GPU-kluster
Bibliotek | Version | Bibliotek | Version | Bibliotek | Version |
---|---|---|---|---|---|
absl-py | 0.11.0 | Antergos Linux | 2015.10 (ISO-Rullande) | appdirs | 1.4.4 |
argon2-cffi | 20.1.0 | Astor | 0.8.1 | astunparse | 1.6.3 |
async-generator | 1.10 | attrs | 20.3.0 | backcall | 0.2.0 |
Bcrypt | 3.2.0 | bidict | 0.21.4 | Blekmedel | 3.3.0 |
blis | 0.7.4 | boto3 | 1.16.7 | botocore | 1.19.7 |
cachetools | 4.2.4 | Katalog | 2.0.6 | Certifi | 2020.12.5 |
Cffi | 1.14.5 | Chardet | 4.0.0 | klicka | 7.1.2 |
cloudpickle | 1.6.0 | cmdstanpy | 0.9.68 | configparser | 5.0.1 |
konvertera | 2.3.2 | Kryptografi | 3.4.7 | Apparat | 0.10.0 |
cymem | 2.0.5 | Cython | 0.29.23 | databricks-automl-runtime | 0.2.5 |
databricks-cli | 0.16.2 | dbl-tempo | 0.1.2 | dbus-python | 1.2.16 |
Dekoratör | 5.0.6 | defusedxml | 0.7.1 | Dill | 0.3.2 |
diskcache | 5.2.1 | distlib | 0.3.4 | distro-info | 0.23ubuntu1 |
entrypoints | 0,3 | efem | 4.1.3 | facets-overview | 1.0.0 |
snabbtext | 0.9.2 | filelock | 3.0.12 | Flask | 1.1.2 |
flatbuffers | 2.0 | fsspec | 0.9.0 | Framtiden | 0.18.2 |
Gast | 0.4.0 | gitdb | 4.0.7 | GitPython | 3.1.12 |
google-auth | 1.22.1 | google-auth-oauthlib | 0.4.2 | google-pasta | 0.2.0 |
grpcio | 1.39.0 | gunicorn | 20.0.4 | gviz-api | 1.10.0 |
h5py | 3.1.0 | hijri-converter | 2.2.2 | Semester | 0,12 |
horovod | 0.23.0 | htmlmin | 0.1.12 | huggingface-hub | 0.1.2 |
idna | 2.10 | ImageHash | 4.2.1 | obalanserad inlärning | 0.8.1 |
importlib-metadata | 3.10.0 | ipykernel | 5.3.4 | Ipython | 7.22.0 |
ipython-genutils | 0.2.0 | ipywidgets | 7.6.3 | isodate | 0.6.0 |
itsdangerous | 1.1.0 | Jedi | 0.17.2 | Jinja2 | 2.11.3 |
jmespath | 0.10.0 | joblib | 1.0.1 | joblibspark | 0.3.0 |
jsonschema | 3.2.0 | jupyter-client | 6.1.12 | jupyter-core | 4.7.1 |
jupyterlab-pygments | 0.1.2 | jupyterlab-widgets | 1.0.0 | keras | 2.7.0 |
Keras-förbearbetning | 1.1.2 | kiwisolver | 1.3.1 | Koalas | 1.8.2 |
korean-lunar-calendar | 0.2.1 | langcodes | 3.3.0 | libclang | 12.0.0 |
lightgbm | 3.3.1 | llvmlite | 0.38.0 | LunarCalendar | 0.0.9 |
Mako | 1.1.3 | Markdown | 3.3.3 | Markering Valv | 2.0.1 |
matplotlib | 3.4.2 | missingno | 0.5.0 | mistune | 0.8.4 |
mleap | 0.18.1 | mlflow-skinny | 1.23.0 | multimethod | 1.6 |
murmurhash | 1.0.5 | nbclient | 0.5.3 | nbconvert | 6.0.7 |
nbformat | 5.1.3 | nest-asyncio | 1.5.1 | Networkx | 2.5 |
nltk | 3.6.1 | notebook-fil | 6.3.0 | numba | 0.55.0 |
Numpy | 1.20.1 | oauthlib | 3.1.0 | opt-einsum | 3.3.0 |
Förpackning | 21,3 | Pandas | 1.2.4 | pandas-profilering | 3.1.0 |
pandocfilter | 1.4.3 | Paramiko | 2.7.2 | parso | 0.7.0 |
Pati | 0.6.0 | Patsy | 0.5.1 | petastorm | 0.11.3 |
Pexpect | 4.8.0 | phik | 0.12.0 | pickleshare | 0.7.5 |
Kudde | 8.2.0 | Pip | 21.0.1 | plotly | 5.5.0 |
pmdarima | 1.8.4 | fördefinierad | 3.0.5 | prompt-toolkit | 3.0.17 |
Profeten | 1.0.1 | protobuf | 3.17.2 | Psutil | 5.8.0 |
psycopg2 | 2.8.5 | ptyprocess | 0.7.0 | pyarrow | 4.0.0 |
pyasn1 | 0.4.8 | pyasn1-modules | 0.2.8 | pybind11 | 2.9.0 |
pycparser | 2,20 | pydantisk | 1.8.2 | Pygments | 2.8.1 |
Pygobject | 3.36.0 | PyMeeus | 0.5.11 | PyNaCl | 1.4.0 |
pyodbc | 4.0.30 | Pyparsing | 2.4.7 | pyrsistent | 0.17.3 |
pystan | 2.19.1.1 | python-apt | 2.0.0+ubuntu0.20.4.6 | python-dateutil | 2.8.1 |
python-editor | 1.0.4 | python-engineio | 4.3.0 | python-socketio | 5.4.1 |
pytz | 2020.5 | PyWavelets | 1.1.1 | PyYAML | 5.4.1 |
pyzmq | 20.0.0 | Regex | 2021.4.4 | begäranden | 2.25.1 |
requests-oauthlib | 1.3.0 | requests-unixsocket | 0.2.0 | Rsa | 4.7.2 |
s3transfer | 0.3.7 | sacremoses | 0.0.46 | scikit-learn | 0.24.1 |
Scipy | 1.6.2 | seaborn | 0.11.1 | Send2Trash | 1.5.0 |
Setuptools | 52.0.0 | setuptools-git | 1.2 | Shap | 0.40.0 |
Simplejson | 3.17.2 | Sex | 1.15.0 | Slicer | 0.0.7 |
smart-open | 5.2.0 | smmap | 3.0.5 | spacy | 3.2.1 |
spacy-legacy | 3.0.8 | spacy-loggers | 1.0.1 | spark-tensorflow-distributor | 1.0.0 |
sqlparse | 0.4.1 | srsly | 2.4.1 | ssh-import-id | 5,10 |
statsmodels | 0.12.2 | Tabellform | 0.8.7 | trasslig upp-i-unicode | 0.1.0 |
Uthållighet | 6.2.0 | tensorboard | 2.7.0 | tensorboard-data-server | 0.6.1 |
tensorboard-plugin-profile | 2.5.0 | tensorboard-plugin-wit | 1.8.1 | tensorflow | 2.7.0 |
tensorflow-estimator | 2.7.0 | tensorflow-io-gcs-filesystem | 0.23.1 | termcolor | 1.1.0 |
terminado | 0.9.4 | testpath | 0.4.4 | thinc | 8.0.12 |
threadpoolctl | 2.1.0 | tokenizers | 0.10.3 | Fackla | 1.10.1+cu111 |
torchvision | 0.11.2+cu111 | Tornado | 6.1 | tqdm | 4.59.0 |
traitlets | 5.0.5 | Transformatorer | 4.15.0 | typer | 0.3.2 |
skriva tillägg | 3.7.4.3 | ujson | 4.0.2 | obevakade uppgraderingar | 0,1 |
urllib3 | 1.25.11 | Virtualenv | 20.4.1 | Visioner | 0.7.4 |
Wasabi | 0.8.2 | wcwidth | 0.2.5 | webencodings | 0.5.1 |
websocket-client | 0.57.0 | Werkzeug | 1.0.1 | Hjul | 0.36.2 |
widgetsnbextension | 3.5.1 | wrapt | 1.12.1 | xgboost | 1.5.1 |
Zipp | 3.4.1 |
Spark-paket som innehåller Python-moduler
Spark-paket | Python-modul | Version |
---|---|---|
graphframes | graphframes | 0.8.2-db1-spark3.2 |
R-bibliotek
R-biblioteken är identiska med R-biblioteken i Databricks Runtime 10.3.
Java- och Scala-bibliotek (Scala 2.12-kluster)
Förutom Java- och Scala-bibliotek i Databricks Runtime 10.3 innehåller Databricks Runtime 10.3 ML följande JAR:er:
CPU-kluster
Grupp-ID | Artefakt-ID | Version |
---|---|---|
com.typesafe.akka | akka-actor_2.12 | 2.5.23 |
ml.combust.mleap | mleap-databricks-runtime_2.12 | 0.18.1-23eb1ef |
ml.dmlc | xgboost4j-spark_2.12 | 1.5.1 |
ml.dmlc | xgboost4j_2.12 | 1.5.1 |
org.graphframes | graphframes_2.12 | 0.8.2-db1-spark3.2 |
org.mlflow | mlflow-client | 1.23.0 |
org.mlflow | mlflow-spark | 1.23.0 |
org.scala-lang.modules | scala-java8-compat_2.12 | 0.8.0 |
org.tensorflow | spark-tensorflow-connector_2.12 | 1.15.0 |
GPU-kluster
Grupp-ID | Artefakt-ID | Version |
---|---|---|
com.typesafe.akka | akka-actor_2.12 | 2.5.23 |
ml.combust.mleap | mleap-databricks-runtime_2.12 | 0.18.1-23eb1ef |
ml.dmlc | xgboost4j-spark_2.12 | 1.5.1 |
ml.dmlc | xgboost4j_2.12 | 1.5.1 |
org.graphframes | graphframes_2.12 | 0.8.2-db1-spark3.2 |
org.mlflow | mlflow-client | 1.23.0 |
org.mlflow | mlflow-spark | 1.23.0 |
org.scala-lang.modules | scala-java8-compat_2.12 | 0.8.0 |
org.tensorflow | spark-tensorflow-connector_2.12 | 1.15.0 |