Databricks Runtime 7.3 LTS per Machine Learning (EoL)

Databricks ha rilasciato questa versione a settembre 2020. È stata dichiarata supporto a lungo termine (LTS) nell'ottobre 2020.

Databricks Runtime 7.3 LTS per Machine Learning offre un ambiente pronto all'uso per il machine learning e la data science basata su Databricks Runtime 7.3 LTS (EoL). Databricks Runtime ML contiene molte di queste popolari librerie per l’apprendimento automatico, tra cui TensorFlow, PyTorch e XGBoost. Supporta inoltre il training distribuito di apprendimento profondo usando Horovod.

Per altre informazioni, incluse le istruzioni per la creazione di un cluster di Databricks Runtime ML, vedere IA e Machine Learning in Databricks.

Per informazioni sulla migrazione da Databricks Runtime 6.x, vedere Guida alla migrazione di Databricks Runtime 7.x.

Nuove caratteristiche e modifiche importanti

Databricks Runtime 7.3 LTS per Machine Learning si basa su Databricks Runtime 7.3 LTS. Per informazioni sulle novità di Databricks Runtime 7.3 LTS, tra cui Apache Spark MLlib e SparkR, vedere le note sulla versione di Databricks Runtime 7.3 LTS (EoL).

Modifiche principali all'ambiente Python di Databricks Runtime ML

Attivazione di Conda nei ruoli di lavoro

In precedenza, quando è stato aggiornato l'ambiente Notebook usando %conda, il nuovo ambiente non è stato attivato nei processi di lavoro Python. Ciò causava problemi se una funzione UDF PySpark denominata funzione di terze parti usava le risorse installate all'interno dell'ambiente Conda. Questa limitazione non esiste più.

È anche consigliabile esaminare le principali modifiche apportate all'ambiente Python di Databricks Runtime in Databricks Runtime 7.3 LTS (EoL). Per un elenco completo dei pacchetti Python installati e delle relative versioni, vedere librerie Python.

Pacchetti Python aggiornati

• mlflow 1.9.1 -> 1.11.0
• tensorflow 2.2.0 -> 2.3.0
• tensorboard 2.2.2 -> 2.3.0
• pytorch 1.5.1 -> 1.6.0
• torchvision 0.6.1 -> 0.7.0
• petastorm 0.9.2 -> 0.9.5

Ambiente di sistema

L'ambiente di sistema in Databricks Runtime 7.3 LTS per Machine Learning differisce da Databricks Runtime 7.3 LTS come indicato di seguito:

• DBUtils: Databricks Runtime ML non contiene l'utilità libreria (dbutils.library) (legacy). È possibile usare i comandi %pip e %conda. Consulta le Librerie Python ad ambito notebook.
• Per i cluster GPU, Databricks Runtime ML include le librerie GPU NVIDIA seguenti:
  • CUDA 10.1 Update 2
  • cuDNN 7.6.5
  • NCCL 2.7.3
  • TensorRT 6.0.1

Librerie

Le sezioni seguenti elencano le librerie incluse in Databricks Runtime 7.3 LTS per Machine Learning che differiscono da quelle incluse in Databricks Runtime 7.3 LTS.

Contenuto della sezione:

• Librerie di livello superiore
• Librerie Python
• Librerie R
• Librerie Java e Scala (cluster Scala 2.12)

Librerie di livello superiore

Databricks Runtime 7.3 LTS per Machine Learning include le librerie di livello superiore seguenti:

• GraphFrames
• Horovod e HorovodRunner
• MLflow
• PyTorch
• spark-tensorflow-connector (connettore per Spark e TensorFlow)
• TensorFlow
• TensorBoard

Librerie Python

Databricks Runtime 7.3 LTS per Machine Learning usa Conda per la gestione dei pacchetti Python e include molti dei pacchetti ML più diffusi.

Oltre ai pacchetti specificati negli ambienti Conda nelle sezioni seguenti, Databricks Runtime 7.3 LTS per Machine Learning installa anche i pacchetti seguenti:

• hyperopt 0.2.4.db2
• sparkdl 2.1.0-db1

Librerie Python nei cluster CPU

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=py_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=1.15.0=h7b6447c_1001
  - ca-certificates=2020.7.22=0
  - cachetools=4.1.1=py_0
  - certifi=2020.6.20=pyhd3eb1b0_3 # (updated from py37_0 in June 15, 2021 maintenance update)
  - cffi=1.14.0=py37he30daa8_1 # (updated from py37h2e261b9_0 in June 15, 2021 maintenance update)
  - chardet=3.0.4=py37_1003
  - click=7.0=py37_0
  - cloudpickle=1.3.0=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=py37_0
  - isodate=0.6.0=py_1
  - itsdangerous=1.1.0=py37_0
  - jedi=0.14.1=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 June 15, 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 June 15, 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 June 15, 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
  - 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=1.10.0=py37hfd86e86_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 1.1.1g in June 15, 2021 maintenance update)
  - packaging=20.1=py_0
  - pandas=1.0.1=py37h0573a6f_0
  - paramiko=2.7.1=py_0
  - parso=0.5.2=py_0
  - patsy=0.5.1=py37_0
  - pexpect=4.8.0=py37_1
  - pickleshare=0.7.5=py37_1001
  - pillow=7.0.0=py37hb39fc2d_0
  - pip=20.0.2=py37_3
  - plotly=4.9.0=py_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 June 15, 2021 maintenance update)
  - ptyprocess=0.6.0=py37_0
  - pyasn1=0.4.8=py_0
  - pyasn1-modules=0.2.7=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=py_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 June 15, 2021 maintenance update)
  - python-dateutil=2.8.1=py_0
  - python-editor=1.0.4=py_0
  - pytorch=1.6.0=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 June 15, 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=py37_0
  - smmap=3.0.4=py_0
  - sqlite=3.35.4=hdfb4753_0 # (updated from 3.31.1 in June 15, 2021 maintenance update)
  - sqlparse=0.3.0=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 June 15, 2021 maintenance update)
  - torchvision=0.7.0=py37_cpu
  - tornado=6.0.3=py37h7b6447c_3
  - tqdm=4.42.1=py_0
  - traitlets=4.3.3=py37_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 June 15, 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==1.8.0
      - azure-storage-blob==12.4.0
      - databricks-cli==0.11.0
      - diskcache==5.0.2
      - docker==4.3.1
      - gorilla==0.3.0
      - horovod==0.19.5
      - joblibspark==0.2.0
      - keras-preprocessing==1.1.2
      - koalas==1.2.0
      - mleap==0.16.1
      - mlflow==1.11.0
      - msrest==0.6.18
      - opt-einsum==3.3.0
      - petastorm==0.9.5
      - 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==2.3.0
      - tensorflow-estimator==2.3.0
      - termcolor==1.1.0
      - xgboost==1.1.1
prefix: /databricks/conda/envs/databricks-ml

Librerie Python nei cluster GPU

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=py_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=1.15.0=h7b6447c_1001
  - ca-certificates=2020.7.22=0
  - cachetools=4.1.1=py_0
  - certifi=2020.6.20=pyhd3eb1b0_3 # (updated from py37_0 in June 15, 2021 maintenance update)
  - cffi=1.14.0=py37he30daa8_1 # (updated from py37h2e261b9_0 in June 15, 2021 maintenance update)
  - chardet=3.0.4=py37_1003
  - click=7.0=py37_0
  - cloudpickle=1.3.0=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=py37_0
  - isodate=0.6.0=py_1
  - itsdangerous=1.1.0=py37_0
  - jedi=0.14.1=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=1.16.4=h173b8e3_0 # (updated from 1.16.4 in June 15, 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 June 15, 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 June 15, 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
  - 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=1.10.0=py37hfd86e86_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 1.1.1g in June 15, 2021 maintenance update)
  - packaging=20.1=py_0
  - pandas=1.0.1=py37h0573a6f_0
  - paramiko=2.7.1=py_0
  - parso=0.5.2=py_0
  - patsy=0.5.1=py37_0
  - pexpect=4.8.0=py37_1
  - pickleshare=0.7.5=py37_1001
  - pillow=7.0.0=py37hb39fc2d_0
  - pip=20.0.2=py37_3
  - plotly=4.9.0=py_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 June 15, 2021 maintenance update)
  - ptyprocess=0.6.0=py37_0
  - pyasn1=0.4.8=py_0
  - pyasn1-modules=0.2.7=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=py_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 June 15, 2021 maintenance update)
  - python-dateutil=2.8.1=py_0
  - python-editor=1.0.4=py_0
  - pytorch=1.6.0=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 June 15, 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=py37_0
  - smmap=3.0.4=py_0
  - sqlite=3.35.4=hdfb4753_0 # (updated from 3.31.1 in June 15, 2021 maintenance update)
  - sqlparse=0.3.0=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 June 15, 2021 maintenance update)
  - torchvision=0.7.0=py37_cu101
  - tornado=6.0.3=py37h7b6447c_3
  - tqdm=4.42.1=py_0
  - traitlets=4.3.3=py37_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 June 15, 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==1.8.0
      - azure-storage-blob==12.4.0
      - databricks-cli==0.11.0
      - diskcache==5.0.2
      - docker==4.3.1
      - gorilla==0.3.0
      - horovod==0.19.5
      - joblibspark==0.2.0
      - keras-preprocessing==1.1.2
      - koalas==1.2.0
      - mleap==0.16.1
      - mlflow==1.11.0
      - msrest==0.6.18
      - opt-einsum==3.3.0
      - petastorm==0.9.5
      - 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==2.3.0
      - tensorflow-estimator==2.3.0
      - termcolor==1.1.0
      - xgboost==1.1.1
prefix: /databricks/conda/envs/databricks-ml-gpu

Pacchetti Spark contenenti moduli Python

Librerie R

Le librerie R sono identiche alle librerie R in Databricks Runtime 7.3 LTS.

Librerie Java e Scala (cluster Scala 2.12)

Oltre alle librerie Java e Scala in Databricks Runtime 7.3 LTS, Databricks Runtime 7.3 LTS per Machine Learning contiene i file JAR seguenti: