List of pre-installed software
Packages for data analysis and machine learning are pre-installed in DataSphere. If you are missing a package, you can install it right from the notebook cell.
List of pre-installed packages
- absl-py
- addict
- antlr4-python3-runtime
- appdirs
- argon2-cffi
- asn1crypto
- astor
- asttokens
- async-generator
- atpublic
- attrs
- audioread
- autograd
- aws-sam-translator
- aws-xray-sdk
- backcall
- bcrypt
- BeautifulSoup4
- bleach
- blis
- boto
- boto3
- botocore
- Bottleneck
- cached-property
- cachetools
- catalogue
- CatBoost
- certifi
- CFFI
- cfn-lint
- chardet
- Click
- cloud-ml
- cloudpickle
- colorama
- commonmark
- configobj
- configparser
- cryptography
- cycler
- cymem
- Cython
- Dask
- dataclasses
- DAWG-Python
- decorator
- defusedxml
- dill
- distro
- docker
- docopt
- docutils
- dpath
- DVC
- ecdsa
- enot
- enot-core
- enot-utils
- entrypoints
- executing
- fairseq
- fastai
- fastprogress
- filelock
- flatten-json
- flufl.lock
- FSSPEC
- funcy
- future
- gast
- Gensim
- gitdb
- GitPython
- google-api-core
- Google API Python Client
- googleapis-common-protos
- google-auth
- google-auth-httplib2
- googledrivedownloader
- google-pasta
- grandalf
- Graphviz
- gRPCio
- h5py
- httplib2
- hydra-core
- idna
- importlib-metadata
- importlib-resources
- iniconfig
- ipykernel
- ipystate
- ipython
- ipywidgets
- jax
- jaxlib
- jedi
- Jinja2
- jmespath
- joblib
- jsondiff
- jsonpatch
- jsonpath-ng
- jsonpickle
- jsonpointer
- jsonschema
- junit-xml
- jupyter-client
- jupyter-core
- jupyterlab-pygments
- kaggle
- Keras
- Keras-Applications
- Keras-Preprocessing
- keyring
- keyrings.alt
- kiwisolver
- librosa
- LightGBM
- littleutils
- llvmlite
- Markdown
- MarkupSafe
- Matplotlib
- midi2audio
- mistune
- ml-kernel
- MMCV-full
- mmdet
- mmpycocotools
- Mock
- more-itertools
- Moto
- murmurhash
- music21
- nanotime
- nbclient
- nbconvert
- nbformat
- nest-asyncio
- NetworkX
- NLTK
- notebook
- numba
- NumExpr
- NumPy
- nvidia-ml-py3
- nvidia-smi
- oauth2client
- OmegaConf
- OpenCV-Python
- opt-einsum
- packaging
- Pandas
- pandocfilters
- Paramiko
- parso
- pathspec
- Pexpect
- pickleshare
- Pillow
- pip
- Plac
- plotly
- pluggy
- ply
- Pooch
- portalocker
- preshed
- prometheus-client
- promise
- prompt-toolkit
- protobuf
- psutil
- Ptyprocess
- py
- PyArrow
- pyasn1
- pyasn1-modules
- pybase64
- pycparser
- PyCrypto
- PycURL
- pydot
- Pygments
- PyGObject
- pygtrie
- Pympler
- PyNaCl
- PyParsing
- Pyrsistent
- pytest
- python-apt
- python-dateutil
- python-jose
- python-slugify
- pytorch-ranger
- pytz
- pyxdg
- PyYAML
- PyZMQ
- regex
- Requests
- resampy
- responses
- retrying
- Rich
- RSA
- ruamel.yaml
- ruamel.yaml.clib
- S3Fs
- s3transfer
- sacrebleu
- sacremoses
- scikit-learn
- Scipy
- scp
- SecretStorage
- Send2Trash
- sentencepiece
- setuptools
- SharedArray
- shortuuid
- shtab
- Six
- slugify
- smart-open
- smmap
- sorcery
- sounddevice
- SoundFile
- soupsieve
- spaCy
- srsly
- sshpubkeys
- stt-metrics
- tabulate
- tblib
- tensorboard
- tensorboardX
- TensorFlow
- TensorFlow Datasets
- tensorflow-estimator
- tensorflow-gpu
- tensorflow-metadata
- termcolor
- terminado
- terminaltables
- testpath
- text-unidecode
- thinc
- Tokenizers
- toml
- Torch
- torch-optimizer
- torchvision
- Tornado
- tqdm
- Traitlets
- Transformers
- typing-extensions
- unattended-upgrades
- uritemplate
- urllib3
- Virtualenv
- Voluptuous
- wasabi
- wcwidth
- webcolors
- webencodings
- websocket-client
- Werkzeug
- wget
- wheel
- widgetsnbextension
- wrapt
- XGBoost
- xmltodict
- xxhash
- YaDisk
- yapf
- zc.lockfile
- zipp
Package versions
To view the version of the installed package, in the notebook cell, run the command:
%pip show <Package name>
To view the list of installed packages and their versions, in the notebook cell, run the command:
%pip list
Updating
You can both upgrade any pre-installed package to a later version or roll it back to an earlier version.
Upgraded versions are preserved when you change environments or restart your project.
To upgrade a library to the latest version, run the following command in the notebook cell:
%pip install <library name> - U
For example, to upgrade TensorFlow to its latest version, run the command:
%pip install tensorflow -U
To upgrade a library to a specific version, run the following command in the notebook cell:
%pip install <library name>==<version>
For example:
%pip install tensorflow==2.3.1
Note that TensorFlow version 1.15 is still installed by default.
Warning
Updating a pre-installed library may introduce new data types that are not supported in DataSphere and not versioned. You'll see a warning when you run cells with such types. In this case, please notify technical support about the libraries upgraded and non-supported data types. To be sure that versioning works correctly, roll back the library version.