Databricks Runtime 7,5 para (EoS) ML

nota

O suporte para essa versão do Databricks Runtime foi encerrado. Para saber a data do fim do suporte, consulte Histórico do fim do suporte. Para conhecer todas as versões compatíveis do site Databricks Runtime, consulte Databricks Runtime notas sobre as versões e a compatibilidade.

A Databricks lançou essa versão em dezembro de 2020.

O Databricks Runtime 7.5 for Machine Learning oferece um ambiente pronto para uso para aprendizado de máquina e ciência de dados com base no Databricks Runtime 7.5 (EoS). Databricks Runtime ML Contém muitas bibliotecas populares de aprendizado de máquina, incluindo TensorFlow, PyTorch, e XGBoost. Ele também oferece suporte ao treinamento de aprendizagem profunda distribuída usando o Horovod.

Para obter mais informações, incluindo instruções para criar um cluster Databricks Runtime ML , consulte AI e aprendizado de máquina em Databricks.

Novo recurso e grandes mudanças

O Databricks Runtime 7.5 ML foi desenvolvido com base no Databricks Runtime 7.5. Para obter informações sobre as novidades do Databricks Runtime 7.5, incluindo Apache Spark MLlib e SparkR, , consulte as notas sobre a versão Databricks Runtime 7.5 (EoS).

Depreciações

• O pacote gorilla está obsoleto e será removido em uma próxima versão principal do Databricks Runtime ML.

Principais mudanças no ambiente do Databricks Runtime ML Python

Consulte o Databricks Runtime 7.5 (EoS) para conhecer as principais alterações no ambiente Python do Databricks Runtime. Para obter uma lista completa do pacote Python instalado e suas versões, consulte Python biblioteca.

Python pacote atualizado

• databricks-CLI 0.13.0 -> 0.14.0
• Hyperopt 0.2.4.db2 -> 0.2.5.db1
• coalas 1.2.0 - > 1.4.0
• mlflow 1.11.0 - > 1.12.1
• petastorm 0.9.6 - > 0.9.7
• trama 4.10.0 - > 4.12.0
• PyTorch 1.6.0 - > 1,7,0
• torchvision 0.7.0 - > 0.8.1
• xgboost 1.2.0 - > 1.2.1

Ambiente do sistema

O ambiente do sistema no Databricks Runtime 7.5 ML difere do Databricks Runtime 7.5 da seguinte forma:

• DBUtils : Databricks Runtime ML não contém biblioteca utilidades (dbutils.biblioteca) (legado). Em vez disso, o senhor pode usar %pip e %conda comando. NotebookConsulte -scoped Pythonbiblioteca.
• Para o clustering de GPU, o site Databricks Runtime ML inclui a seguinte biblioteca de GPUs NVIDIA:
  • Atualização 2 do CUDA 10.1
  • cuDNN 7.6.5
  • NCCL 2.7.3
  • TensorRT 6.0.1

biblioteca

As seções a seguir listam as bibliotecas incluídas no Databricks Runtime 7.5 ML que diferem daquelas incluídas no Databricks Runtime 7.5.

Nesta secção:

• Biblioteca de primeira linha
• Bibliotecas Python
• R biblioteca
• Java e Scala biblioteca (Scala 2.12 clustering)

Biblioteca de primeira linha

Databricks Runtime 7,5 O site ML inclui as seguintes bibliotecas de primeira linha:

• GraphFrames
• Horovod e HorovodRunner
• MLflow
• PyTorch
• spark-tensorflow-connector
• TensorFlow
• TensorBoard

Python biblioteca

Databricks Runtime 7.5 O ML usa o Conda para o gerenciamento do pacote Python e inclui muitos pacotes populares do ML.

Além do pacote especificado nos ambientes Conda nas seções a seguir, o Databricks Runtime 7.5 ML também instala o seguinte pacote:

• Hyperopt 0.2.5.db1
• sparkdl 2.1.0-db2

Python biblioteca sobre clustering de CPU

YAML

name: databricks-ml
channels:
  - pytorch
  - defaults
dependencies:
  - _libgcc_mutex=0.1=main
  - absl-py=0.9.0=py37_0
  - asn1crypto=1.3.0=py37_1
  - astor=0.8.0=py37_0
  - backcall=0.1.0=py37_0
  - backports=1.0=pyhd3eb1b0_2
  - bcrypt=3.2.0=py37h7b6447c_0
  - blas=1.0=mkl
  - blinker=1.4=py37_0
  - boto3=1.12.0=py_0
  - botocore=1.15.0=py_0
  - c-ares=1.17.1=h27cfd23_0
  - ca-certificates=2020.10.14=h06a4308_1 # (updated from 0 in May 26, 2021 maintenance update)
  - cachetools=4.1.1=py_0
  - certifi=2020.11.8=py37h06a4308_0
  - cffi=1.14.0=py37he30daa8_1 # (updated from py37h2e261b9_0 in May 26, 2021 maintenance update)
  - chardet=3.0.4=py37h06a4308_1003
  - click=7.0=py37_0
  - cloudpickle=1.4.1=py_0
  - configparser=3.7.4=py37_0
  - cpuonly=1.0=0
  - cryptography=2.8=py37h1ba5d50_0
  - cycler=0.10.0=py37_0
  - cython=0.29.15=py37he6710b0_0
  - decorator=4.4.1=py_0
  - dill=0.3.1.1=py37_1
  - docutils=0.15.2=py37_0
  - entrypoints=0.3=py37_0
  - flask=1.1.1=py_1
  - freetype=2.9.1=h8a8886c_1
  - future=0.18.2=py37_1
  - gast=0.3.3=py_0
  - gitdb=4.0.5=py_0
  - gitpython=3.1.0=py_0
  - google-auth=1.11.2=py_0
  - google-auth-oauthlib=0.4.1=py_2
  - google-pasta=0.2.0=py_0
  - grpcio=1.27.2=py37hf8bcb03_0
  - gunicorn=20.0.4=py37_0
  - h5py=2.10.0=py37h7918eee_0
  - hdf5=1.10.4=hb1b8bf9_0
  - icu=58.2=he6710b0_3
  - idna=2.8=py37_0
  - intel-openmp=2020.0=166
  - ipykernel=5.1.4=py37h39e3cac_0
  - ipython=7.12.0=py37h5ca1d4c_0
  - ipython_genutils=0.2.0=pyhd3eb1b0_1
  - isodate=0.6.0=py_1
  - itsdangerous=1.1.0=py37_0
  - jedi=0.17.2=py37_0
  - jinja2=2.11.1=py_0
  - jmespath=0.10.0=py_0
  - joblib=0.14.1=py_0
  - jpeg=9b=h024ee3a_2
  - jupyter_client=5.3.4=py37_0
  - jupyter_core=4.6.1=py37_0
  - kiwisolver=1.1.0=py37he6710b0_0
  - krb5=1.17.1=h173b8e3_0 # (updated from 1.16.4 in May 26, 2021 maintenance update)
  - ld_impl_linux-64=2.33.1=h53a641e_7
  - libedit=3.1.20181209=hc058e9b_0
  - libffi=3.3=he6710b0_2 # (updated from 3.2.1 in May 26, 2021 maintenance update)
  - libgcc-ng=9.1.0=hdf63c60_0
  - libgfortran-ng=7.3.0=hdf63c60_0
  - libpng=1.6.37=hbc83047_0
  - libpq=12.2=h20c2e04_0 # (updated from 11.2 in May 26, 2021 maintenance update)
  - libprotobuf=3.11.4=hd408876_0
  - libsodium=1.0.16=h1bed415_0
  - libstdcxx-ng=9.1.0=hdf63c60_0
  - libtiff=4.1.0=h2733197_0
  - libuv=1.40.0=h7b6447c_0
  - lightgbm=2.3.0=py37he6710b0_0
  - lz4-c=1.8.1.2=h14c3975_0
  - mako=1.1.2=py_0
  - markdown=3.1.1=py37_0
  - markupsafe=1.1.1=py37h14c3975_1
  - matplotlib-base=3.1.3=py37hef1b27d_0
  - mkl=2020.0=166
  - mkl-service=2.3.0=py37he904b0f_0
  - mkl_fft=1.0.15=py37ha843d7b_0
  - mkl_random=1.1.0=py37hd6b4f25_0
  - ncurses=6.2=he6710b0_1
  - networkx=2.4=py_1
  - ninja=1.10.2=py37hff7bd54_0
  - nltk=3.4.5=py37_0
  - numpy=1.18.1=py37h4f9e942_0
  - numpy-base=1.18.1=py37hde5b4d6_1
  - oauthlib=3.1.0=py_0
  - olefile=0.46=py37_0
  - openssl=1.1.1k=h27cfd23_0 # (updated from 1.1.1h in May 26, 2021 maintenance update)
  - packaging=20.1=py_0
  - pandas=1.0.1=py37h0573a6f_0
  - paramiko=2.7.1=py_0
  - parso=0.7.0=py_0
  - patsy=0.5.1=py37_0
  - pexpect=4.8.0=pyhd3eb1b0_3
  - pickleshare=0.7.5=py37_1001
  - pillow=7.0.0=py37hb39fc2d_0
  - pip=20.0.2=py37_3
  - plotly=4.12.0=pyhd3eb1b0_0
  - prompt_toolkit=3.0.3=py_0
  - protobuf=3.11.4=py37he6710b0_0
  - psutil=5.6.7=py37h7b6447c_0
  - psycopg2=2.8.6=py37h3c74f83_1 # (updated from 2.8.4 in May 26, 2021 maintenance update)
  - ptyprocess=0.6.0=pyhd3eb1b0_2
  - pyasn1=0.4.8=py_0
  - pyasn1-modules=0.2.8=py_0
  - pycparser=2.19=py37_0
  - pygments=2.5.2=py_0
  - pyjwt=1.7.1=py37_0
  - pynacl=1.3.0=py37h7b6447c_0
  - pyodbc=4.0.30=py37he6710b0_0
  - pyopenssl=19.1.0=pyhd3eb1b0_1
  - pyparsing=2.4.6=py_0
  - pysocks=1.7.1=py37_1
  - python=3.7.10=hdb3f193_0 # (updated from 3.7.6 in May 26, 2021 maintenance update)
  - python-dateutil=2.8.1=py_0
  - python-editor=1.0.4=py_0
  - pytorch=1.7.0=py3.7_cpu_0
  - pytz=2019.3=py_0
  - pyzmq=18.1.1=py37he6710b0_0
  - readline=8.1=h27cfd23_0 # (updated from 7.0 in May 26, 2021 maintenance update)
  - requests=2.22.0=py37_1
  - requests-oauthlib=1.3.0=py_0
  - retrying=1.3.3=py37_2
  - rsa=4.0=py_0
  - s3transfer=0.3.3=py37_1
  - scikit-learn=0.22.1=py37hd81dba3_0
  - scipy=1.4.1=py37h0b6359f_0
  - setuptools=45.2.0=py37_0
  - simplejson=3.17.0=py37h7b6447c_0
  - six=1.14.0=py37h06a4308_0
  - smmap=3.0.4=py_0
  - sqlite=3.35.4=hdfb4753_0 # (updated from 3.31.1 in May 26, 2021 maintenance update)
  - sqlparse=0.4.1=py_0
  - statsmodels=0.11.0=py37h7b6447c_0
  - tabulate=0.8.3=py37_0
  - tk=8.6.10=hbc83047_0 # (updated from 8.6.8 in May 26, 2021 maintenance update)
  - torchvision=0.8.1=py37_cpu
  - tornado=6.0.3=py37h7b6447c_3
  - tqdm=4.42.1=py_0
  - traitlets=4.3.3=py37_0
  - typing_extensions=3.7.4.3=py_0
  - unixodbc=2.3.7=h14c3975_0
  - urllib3=1.25.8=py37_0
  - wcwidth=0.1.8=py_0
  - websocket-client=0.56.0=py37_0
  - werkzeug=1.0.0=py_0
  - wheel=0.34.2=py37_0
  - wrapt=1.11.2=py37h7b6447c_0
  - xz=5.2.5=h7b6447c_0 # (updated from 5.2.4 in May 26, 2021 maintenance update)
  - zeromq=4.3.1=he6710b0_3
  - zlib=1.2.11=h7b6447c_3
  - zstd=1.3.7=h0b5b093_0
  - pip:
      - astunparse==1.6.3
      - azure-core==1.9.0
      - azure-storage-blob==12.6.0
      - databricks-cli==0.14.0
      - diskcache==5.1.0
      - docker==4.4.0
      - gorilla==0.3.0
      - horovod==0.20.3
      - joblibspark==0.3.0
      - keras-preprocessing==1.1.2
      - koalas==1.4.0
      - mleap==0.16.1
      - mlflow==1.12.1
      - msrest==0.6.19
      - opt-einsum==3.3.0
      - petastorm==0.9.7
      - pyarrow==1.0.1
      - pyyaml==5.3.1
      - querystring-parser==1.2.4
      - seaborn==0.10.0
      - spark-tensorflow-distributor==0.1.0
      - tensorboard==2.3.0
      - tensorboard-plugin-wit==1.7.0
      - tensorflow-cpu==2.3.1
      - tensorflow-estimator==2.3.0
      - termcolor==1.1.0
      - xgboost==1.2.1
prefix: /databricks/conda/envs/databricks-ml

Python biblioteca sobre clustering de GPU

YAML

name: databricks-ml-gpu
channels:
  - pytorch
  - defaults
dependencies:
  - _libgcc_mutex=0.1=main
  - absl-py=0.9.0=py37_0
  - asn1crypto=1.3.0=py37_1
  - astor=0.8.0=py37_0
  - backcall=0.1.0=py37_0
  - backports=1.0=pyhd3eb1b0_2
  - bcrypt=3.2.0=py37h7b6447c_0
  - blas=1.0=mkl
  - blinker=1.4=py37_0
  - boto3=1.12.0=py_0
  - botocore=1.15.0=py_0
  - c-ares=1.17.1=h27cfd23_0
  - ca-certificates=2020.10.14=h06a4308_1 # (updated from 0 in May 26, 2021 maintenance update)
  - cachetools=4.1.1=py_0
  - certifi=2020.11.8=py37h06a4308_0
  - cffi=1.14.0=py37he30daa8_1 # (updated from py37h2e261b9_0 in May 26, 2021 maintenance update)
  - chardet=3.0.4=py37h06a4308_1003
  - click=7.0=py37_0
  - cloudpickle=1.4.1=py_0
  - configparser=3.7.4=py37_0
  - cryptography=2.8=py37h1ba5d50_0
  - cudatoolkit=10.1.243=h6bb024c_0
  - cycler=0.10.0=py37_0
  - cython=0.29.15=py37he6710b0_0
  - decorator=4.4.1=py_0
  - dill=0.3.1.1=py37_1
  - docutils=0.15.2=py37_0
  - entrypoints=0.3=py37_0
  - flask=1.1.1=py_1
  - freetype=2.9.1=h8a8886c_1
  - future=0.18.2=py37_1
  - gast=0.3.3=py_0
  - gitdb=4.0.5=py_0
  - gitpython=3.1.0=py_0
  - google-auth=1.11.2=py_0
  - google-auth-oauthlib=0.4.1=py_2
  - google-pasta=0.2.0=py_0
  - grpcio=1.27.2=py37hf8bcb03_0
  - gunicorn=20.0.4=py37_0
  - h5py=2.10.0=py37h7918eee_0
  - hdf5=1.10.4=hb1b8bf9_0
  - icu=58.2=he6710b0_3
  - idna=2.8=py37_0
  - intel-openmp=2020.0=166
  - ipykernel=5.1.4=py37h39e3cac_0
  - ipython=7.12.0=py37h5ca1d4c_0
  - ipython_genutils=0.2.0=pyhd3eb1b0_1
  - isodate=0.6.0=py_1
  - itsdangerous=1.1.0=py37_0
  - jedi=0.17.2=py37_0
  - jinja2=2.11.1=py_0
  - jmespath=0.10.0=py_0
  - joblib=0.14.1=py_0
  - jpeg=9b=h024ee3a_2
  - jupyter_client=5.3.4=py37_0
  - jupyter_core=4.6.1=py37_0
  - kiwisolver=1.1.0=py37he6710b0_0
  - krb5=1.17.1=h173b8e3_0 # (updated from 1.16.4 in May 26, 2021 maintenance update)
  - ld_impl_linux-64=2.33.1=h53a641e_7
  - libedit=3.1.20181209=hc058e9b_0
  - libffi=3.3=he6710b0_2 # (updated from 3.2.1 in May 26, 2021 maintenance update)
  - libgcc-ng=9.1.0=hdf63c60_0
  - libgfortran-ng=7.3.0=hdf63c60_0
  - libpng=1.6.37=hbc83047_0
  - libpq=12.2=h20c2e04_0 # (updated from 11.2 in May 26, 2021 maintenance update)
  - libprotobuf=3.11.4=hd408876_0
  - libsodium=1.0.16=h1bed415_0
  - libstdcxx-ng=9.1.0=hdf63c60_0
  - libtiff=4.1.0=h2733197_0
  - libuv=1.40.0=h7b6447c_0
  - lightgbm=2.3.0=py37he6710b0_0
  - lz4-c=1.8.1.2=h14c3975_0
  - mako=1.1.2=py_0
  - markdown=3.1.1=py37_0
  - markupsafe=1.1.1=py37h14c3975_1
  - matplotlib-base=3.1.3=py37hef1b27d_0
  - mkl=2020.0=166
  - mkl-service=2.3.0=py37he904b0f_0
  - mkl_fft=1.0.15=py37ha843d7b_0
  - mkl_random=1.1.0=py37hd6b4f25_0
  - ncurses=6.2=he6710b0_1
  - networkx=2.4=py_1
  - ninja=1.10.2=py37hff7bd54_0
  - nltk=3.4.5=py37_0
  - numpy=1.18.1=py37h4f9e942_0
  - numpy-base=1.18.1=py37hde5b4d6_1
  - oauthlib=3.1.0=py_0
  - olefile=0.46=py37_0
  - openssl=1.1.1k=h27cfd23_0 # (updated from 1.1.1h in May 26, 2021 maintenance update)
  - packaging=20.1=py_0
  - pandas=1.0.1=py37h0573a6f_0
  - paramiko=2.7.1=py_0
  - parso=0.7.0=py_0
  - patsy=0.5.1=py37_0
  - pexpect=4.8.0=pyhd3eb1b0_3
  - pickleshare=0.7.5=py37_1001
  - pillow=7.0.0=py37hb39fc2d_0
  - pip=20.0.2=py37_3
  - plotly=4.12.0=pyhd3eb1b0_0
  - prompt_toolkit=3.0.3=py_0
  - protobuf=3.11.4=py37he6710b0_0
  - psutil=5.6.7=py37h7b6447c_0
  - psycopg2=2.8.6=py37h3c74f83_1 # (updated from 2.8.4 in May 26, 2021 maintenance update)
  - ptyprocess=0.6.0=pyhd3eb1b0_2
  - pyasn1=0.4.8=py_0
  - pyasn1-modules=0.2.8=py_0
  - pycparser=2.19=py37_0
  - pygments=2.5.2=py_0
  - pyjwt=1.7.1=py37_0
  - pynacl=1.3.0=py37h7b6447c_0
  - pyodbc=4.0.30=py37he6710b0_0
  - pyopenssl=19.1.0=pyhd3eb1b0_1
  - pyparsing=2.4.6=py_0
  - pysocks=1.7.1=py37_1
  - python=3.7.10=hdb3f193_0 # (updated from 3.7.6 in May 26, 2021 maintenance update)
  - python-dateutil=2.8.1=py_0
  - python-editor=1.0.4=py_0
  - pytorch=1.7.0=py3.7_cuda10.1.243_cudnn7.6.3_0
  - pytz=2019.3=py_0
  - pyzmq=18.1.1=py37he6710b0_0
  - readline=8.1=h27cfd23_0 # (updated from 7.0 in May 26, 2021 maintenance update)
  - requests=2.22.0=py37_1
  - requests-oauthlib=1.3.0=py_0
  - retrying=1.3.3=py37_2
  - rsa=4.0=py_0
  - s3transfer=0.3.3=py37_1
  - scikit-learn=0.22.1=py37hd81dba3_0
  - scipy=1.4.1=py37h0b6359f_0
  - setuptools=45.2.0=py37_0
  - simplejson=3.17.0=py37h7b6447c_0
  - six=1.14.0=py37h06a4308_0
  - smmap=3.0.4=py_0
  - sqlite=3.35.4=hdfb4753_0 # (updated from 3.31.1 in May 26, 2021 maintenance update)
  - sqlparse=0.4.1=py_0
  - statsmodels=0.11.0=py37h7b6447c_0
  - tabulate=0.8.3=py37_0
  - tk=8.6.10=hbc83047_0 # (updated from 8.6.8 in May 26, 2021 maintenance update)
  - torchvision=0.8.1=py37_cu101
  - tornado=6.0.3=py37h7b6447c_3
  - tqdm=4.42.1=py_0
  - traitlets=4.3.3=py37_0
  - typing_extensions=3.7.4.3=py_0
  - unixodbc=2.3.7=h14c3975_0
  - urllib3=1.25.8=py37_0
  - wcwidth=0.1.8=py_0
  - websocket-client=0.56.0=py37_0
  - werkzeug=1.0.0=py_0
  - wheel=0.34.2=py37_0
  - wrapt=1.11.2=py37h7b6447c_0
  - xz=5.2.5=h7b6447c_0 # (updated from 5.2.4 in May 26, 2021 maintenance update)
  - zeromq=4.3.1=he6710b0_3
  - zlib=1.2.11=h7b6447c_3
  - zstd=1.3.7=h0b5b093_0
  - pip:
      - astunparse==1.6.3
      - azure-core==1.9.0
      - azure-storage-blob==12.6.0
      - databricks-cli==0.14.0
      - diskcache==5.1.0
      - docker==4.4.0
      - gorilla==0.3.0
      - horovod==0.20.3
      - joblibspark==0.3.0
      - keras-preprocessing==1.1.2
      - koalas==1.4.0
      - mleap==0.16.1
      - mlflow==1.12.1
      - msrest==0.6.19
      - opt-einsum==3.3.0
      - petastorm==0.9.7
      - pyarrow==1.0.1
      - pyyaml==5.3.1
      - querystring-parser==1.2.4
      - seaborn==0.10.0
      - spark-tensorflow-distributor==0.1.0
      - tensorboard==2.3.0
      - tensorboard-plugin-wit==1.7.0
      - tensorflow==2.3.1
      - tensorflow-estimator==2.3.0
      - termcolor==1.1.0
      - xgboost==1.2.1
prefix: /databricks/conda/envs/databricks-ml-gpu

Spark pacote contendo os módulos Python

Spark pacote	Módulo Python	Versão
graphframes	graphframes	0.8.1-db1-spark3.0

R biblioteca

A biblioteca R é idêntica à biblioteca R em Databricks Runtime 7.5.

Java e biblioteca ( 2.12 clustering) Scala Scala

Além de Java e Scala biblioteca em Databricks Runtime 7.5, Databricks Runtime 7.5 ML contém os seguintes JARs:

Agrupamento de CPU

ID do grupo	ID do artefato	Versão
com.typesafe.akka	também conhecido como actor_2.12	2.5.23
ml.combust.mleap	mleap-databricks-runtime_2.12	0,17.3-4882dc3
ml.dmlc	xgboost4j-spark_2.12	1.2.0
ml.dmlc	xgboost4j_2,12	1.2.0
org.mlflow	cliente mlflow	1.12.1
org.Scala-lang.modules	Scala-java8-compat_2.12	0,8.0
org.tensorflow	spark-tensorflow-connector_2.12	1,15.0

Agrupamento de GPU

ID do grupo	ID do artefato	Versão
com.typesafe.akka	também conhecido como actor_2.12	2.5.23
ml.combust.mleap	mleap-databricks-runtime_2.12	0,17.3-4882dc3
ml.dmlc	xgboost4j-spark-gpu_2.12	1.2.0
ml.dmlc	xgboost4j-gpu_2,12	1.2.0
org.mlflow	cliente mlflow	1.12.1
org.Scala-lang.modules	Scala-java8-compat_2.12	0,8.0
org.tensorflow	spark-tensorflow-connector_2.12	1,15.0