- Added basic support for autoscaling deployments
- Normalized parameter names in
notebooks
group andNotebooksClient
- Fixed NotImplementedError raised when parsing experiment with dataset volume options set
- Fix output of
metrics stream
command - Fixed experiment status constants and added constants for deployment statuses
- Fixed reading logs with
--follow
. The bug caused all log lines in a loop instead of only the new one - Fixed
TypeError
occurring when a job was being created without--workspace
- Added whole directory upload to
models upload
command - Added a command group for managing
secrets
- Added
notebooks logs
command for reading logs of notebooks
- Fixed NotImplementedError raised when parsing experiment with dataset volume options set
- Fixed NotImplementedError raised when parsing experiment with dataset volume options set
- Fix output of
metrics stream
command - Fixed experiment status constants and added constants for deployment statuses
- Fixed reading logs with
--follow
. The bug caused all log lines in a loop instead of only the new one - Fixed
TypeError
occurring when a job was being created without--workspace
- Added whole directory upload to
models upload
command - Added a command group for managing
secrets
- Removed
entity
parameter from tag-related methods in clients
- Fix reading logs with
--follow true
- Fix creating job without workspace
- Added
deployments logs
command for reading depoyment logs - Added
--command
option todeployments update
- Added
--projectId
option todeployments create
anddeployments update
- Added
gradient-utils
to required
- Fix reading logs with
--follow true
- Fix creating job without workspace
- Added
deployments logs
command
- Fix reading logs with
--follow true
- Fix creating job without workspace
- Added
deployments logs
command
- Fix reading logs with
--follow true
- Fix creating job without workspace
- Removed
entity
parameter from tag-related methods in clients
- Added
--command
option todeployments update
- Added
--projectId
option todeployments create
anddeployments update
- Added
gradient-utils
to required
- Added
deployments logs
command for reading depoyment logs - Added
notebooks create
command for creating and starting a new notebook - Added
notebooks start
command for starting an existing notebook - Added
notebooks fork
command for forking a new notebook - Added
notebooks stop
command for stopping a running notebook - Added
notebooks artifacts list
command for listing the last workspace artifacts uploaded for a notebook.
- Removed experimental
projects wizard
command
- Added some datetime fields to deployment and experiment models
--modelId
is not required to create a deployment anymore- Added commands for reading and streaming experiment, deployment, notebook and job metrics
- Fix request field name for
deployment create
command - Fix cluster ID field again
- Fixed a bug in model upload introduced in 0.7.0a3
- Added some datetime fields to deployment and experiment models
--modelId
is not required to create a deployment anymore- Added commands for reading and streaming deployment, notebook and job metrics
- Fix cluster ID field again
- Removed experimental
projects wizard
command
- Added commands for reading experiment metrics
- Fix request field name for
deployment create
command
- Fix request field name for
deployment create
command
- Added support for uploading 100MB+ models
- Added
command
option/parameter todepoyments create
command and SDK method
--workspaceUrl
and--workspaceArchive
replaced by--workspace
- Added
clusters machineTypes list
command - Added
--projectId
option tomodels upload
command - Overwriting options provided in YAML file with values passed in terminal
- Fix raising KeyError when listing experiments
- Added
clusters list
command - Fixed URL to a newly created project
- Added
--projectId
option tomodels upload
command - Overwriting options provided in YAML file with values passed in terminal
--workspaceUrl
and--workspaceArchive
replaced by--workspace
- Added
clusters list
command - Fixed URL to a newly created project
- Removed the
--vpc
flag --name
is not required to create/run experiment. Parameters in methods for creating/running experiments have changed
- Added commands to
add
andremove
tags - Added filtering notebooks by tags
- Added showing tags in
details
commands - Added
projects details
command - Datasets for experiments in YAML options file can now be defined as list of objects
- Allow string instead of list of strings in YAML for options that take multiple values
- Removed empty lines in logs
- Fixed
ExperimentsClient.list
parameters (not a breaking change) - Fix showing cluster ID in
deployments details
- Add filtering entities by tags in
list
commands
- Removed the
--vpc
flag
- Added commands to
add
andremove
tags - Added filtering notebooks by tags
- Added showing tags in
details
commands - Added
projects details
command
--name
is not required to create/run experiment. Parameters in methods for creating/running experiments have changed
- Datasets for experiments in YAML options file can now be defined as list of objects
- Allow string instead of list of strings in YAML for options that take multiple values
- Removed empty lines in logs
- Fixed
ExperimentsClient.list
parameters (not a breaking change) - Fix showing cluster ID in
deployments details
- Add filtering entities by tags in
list
commands
- Update
notebooks list
command accepts--list
and--offset
arguments and show more notebooks - Add
deployments details
command - registry target options to
run
andjobs create
commands --workspace
option inexperiments
commands does not default to current working directory anymore and is required for VPC experiments
- Add
deployments details
command
- Update
notebooks list
command accepts--list
and--offset
arguments and show more notebooks
- Added
projects delete
command - Added
models upload
command for uploading model file from local machine - Added
--workspaceRef
option toexperiments create/run
for specifying branch, commit hash or tag - Added new options to
experiments create/run
for creating experiments with datasets - Added several new options to
deployments create
- Added
deployments update
command - Added
models details
command - Added pagination for
experiments list
- Fixed constant used to create deployment
- Fix listing experiments when filtering by project ID
- Fixed KeyError when setting dataset options without setting --datasetUri
- Fixed error occurring when not all dataset values were set
- Dropped dependencies: gradient-statsd and gradient-sdk
- Changed ID attribute in Job and Deployment models from
id_
toid
- Reverted changes to
--clusterId
and--vpc
from 0.5.1a2
- Update model upload to accept path to file or directory
- Change
--clusterId
to required parameter for creation and start of deployments and experiments - Remove
--vpc
flag for creation and start of deployments and experiments
- Added
model upload
command with directory upload command
- Fixed error occurring when not all dataset values were set
- Added
models details
command - Added pagination for
experiments list
- Added
deployments update
command
- Fixed KeyError when setting dataset options without setting --datasetUri
- Added
projects delete
command - Added
models upload
command for uploading model file from local machine - Added
--workspaceRef
option toexperiments create/run
for specifying branch, commit hash or tag - Added new options to
experiments create/run
for creating experiments with datasets - Added several new options to
deployments create
- Added filtering deleted projects in
projects list
- Changed ID attribute in Job and Deployment models from
id_
toid
- Fixed constant used to create deployment
- Fix listing experiments when filtering by project ID
- Dropped dependencies: gradient-statsd and gradient-sdk
- Fixed constant used to create deployment
- New
delete
commands for deployments and experiments - Printing url to instance web client's view to terminal after creating it
- Fix spinner not disappearing when some commands were executed
- base64-encoded
command
value so it's not blocked by Cloudflare's filter anymore
- Moved hyperparameters group to experiments group
- Dropped experiment_type_id parameter from create_mpi_multi_node() in ExperimentsClient
- Moved constants and config modules into api_sdk module
- Command for deleting a tensorboard
- Optional adding new experiment to tensorboard with new
--tensorboard
option. Available forexperiments create
andexperiments run
- Add MPI-specific options for creating experiments
- Fixed AttributeError raised when
--ignoreFiles
was used - Fixed showing help for some commands
- Fixed URL for fetching notebook details
- Fix StopIteration exception in Python 3.7
- Fix
job delete
andjob stop
commands - Fix how some error messages are printed
- Fix formatting of datetime in
projects list
table - Fix sending some bash commands using
--*command
options caused Cloudflare's filter to stop request
- Fix formatting of datetime in
projects list
table - Fix sending some bash commands using
--*command
options caused Cloudflare's filter to stop request
- Fix how some error messages are printed
- Fix formatting of datetime in
projects list
table - Fix sending some bash commands using
--*command
options caused Cloudflare's filter to stop request
- Fix how some error messages are printed
- Fix
job delete
andjob stop
commands - Fix how some error messages are printed
- Command for deleting a tensorboard
- Optional adding new experiment to tensorboard with new
--tensorboard
option. Available forexperiments create
andexperiments run
- Add MPI-specific options for creating experiments
- Fixed AttributeError raised when
--ignoreFiles
was used - Fixed showing help for some commands
- Fixed URL for fetching notebook details
- Fix StopIteration exception in Python 3.7
- Fix
job delete
andjob stop
commands
- Fix StopIteration exception in Python 3.7
- Fixed bug causing exception be raised when
--help
option was used with some commands (likegradient run
)
Fixed bug causing exception be raised when--help
option was used with some commands (likegradient run
)
- Fixed bug raising exception when
experiment create's
--ignoreFiles
option was used
- Support for tensorboard
- Create tensorboard
- List user tensorboards
- Details of user tensorboard
- Add experiments to existing tensorboard
- Remove experiments from existing tensorboard
- Add support for experiments workspace credentials
- Fixed bug preventing showing logs when 'experiments run' was executed
- Introduced Python SDK package
- Added
ClusterID
function to support new VPCs - Hyperparameter private registry and custom container functions added
--hyperparameterServerContainer
--hyperparameterServerRegistryPassword
--hyperparameterServerRegistryUsername
- Automated documentation generation based on sphinx
- Updated documentation and docstrings for SDK based functions
- Improved
--help
functions - Added support for reading options from file to all commands
- Renamed
--optionsFileTemplate
to--createOptionsFile
- Added
machines
support to the SDK - Added
notebooks
command group withcreate
,delete
andlist
commands - Added
--clusterId
optiondeployments create
- Added coloured options tags ([required], etc.)
- Showing deployment state in deployments list
- Added support for missing isPreemptible option to the SDK
- Fixed URL to logs of jobs
- Removed hardcoded choices for
--machineType
option - Fixed message shown while creating deployment
- Fixed exception when listing job artifacts
- Fixed exception when listing job artifacts
- Added
--clusterId
optiondeployments create
- Fixed message shown while creating deployment
- Added
machines
support to the SDK - Added
notebooks
command group withcreate
,delete
andlist
commands
- Removed hardcoded choices for
--machineType
option
- Added support for reading options from file to all commands
- Renamed
--optionsFileTemplate
to--createOptionsFile
- Fixed URL to logs of jobs
- Added --optionFile to read options from YAML file
- Added --optionFileTemplate to create YAML template ###Fixes
- Added support for missing isPreemptible option to the SDK
- Introduced Python SDK package
- Added
ClusterID
function to support new VPCs - Hyperparameter private registry and custom container functions added
--hyperparameterServerContainer
--hyperparameterServerRegistryPassword
--hyperparameterServerRegistryUsername
- Automated documentation generation based on sphinx
- Updated documentation and docstrings for SDK based functions
- Improved
--help
functions
- hardcoded
colorama
version to avoid errors during installation
- deprecated
--workspaceArchive
and--workspaceUrl
forexperiments create/run
,run
andjobs create
commands - single
--workspace
option for defining git, S3 and local workspaces - experimental wizard for creating projects
- add looking for new client version
- added
projects create
command - added spinner for some long running commands
- fix autocomplete executing prompt for password when double-tab on gradient apiKey
- deprecated
--workspaceArchive
and--workspaceUrl
forexperiments create/run
,run
andjobs create
commands - single
--workspace
option for defining git, S3 and local workspaces
###New features
- experimental wizard for creating projects
- fix autocomplete executing prompt for password when double-tab on gradient apiKey
- add looking for new client version
- added
projects create
command - added spinner for some long running commands
- renamed
createAndStart
commands torun
for experiments and hyperparameters
- removed command group aliases
- streaming logs for
experiments createAndStart
by default
- new params for
job create
andhyperparameters create
- fixed missing
--experimentType
when creating new single node experiment - deployments can be of
TFSERVING
type only - fixed autocomplete instructions
- autocomplete for bash (GNU/Linux and iOS)
- aliases in singular form for command groups
- showing logs for experiments and jobs
- restored and deprecated
run
command - workspace uploading for experiments
- workspace upload progress bar
- partial support for hyperparameter tuning
- prompt for apiKey if not provided
- deprecate
login
command --ignoreFiles
option forexperiments create
--ignoreFiles
and--workspace none
options forjobs create
- restore apiKey command
- colorful error and help messages
- removed
paperspace-python run
command
- partial support for managing jobs, models and projects
- filtering deployments list by project ID and model ID
- reading logs of experiments and jobs
- handling error response when listing experiments
- experiments and deployments management
- --apiKey for all commands for one-time api key overwrite
- help messages for all commands and parameters
- filtering experiments list with
--projectHandle
- paginating experiments list when does not fit the terminal width
- added
--modelType
and--modelPath
toexperiments create
- some minor bug fixes
- support for experiments API - creating, starting, listing, etc.
- Fix missing import
- Add statsd agent support with StatsdClient
- Fix crashing on converting to bytes in python < 3
-
Run gradient jobs from custom Dockerfile built containers (see: https://docs.paperspace.com/gradient/jobs/create-a-job#new-run-jobs-from-dockerfiles)
-
Push GPU-enabled container images to a registry of your choice
- Add custom headers to requests
- Allow user to select preemptible machine types
- Fix package.egg-info issue affecting Windows users downloading paperspace from PyPI
- Allow user to select custom port mappings between container and host
- Fix circular import bug causing problems for python2 users on v0.0.12
- New cluster and machineType functionality supported: send jobs to a gradient-node based on the cluster name or clusterId
- For more information on running gradient-node see the Gradient Private Cloud docs
- Send jobs to a specific node by specifying node attributes
- Changed default handling of machineType: cluster preferences come first
- Jobs create automatically records the git commit hash for local workspaces using git
- New jobs machineTypes method for discovering available job machine types.
- Minor doc fixes
paperspace-python run
command for running python scripts unmodified on paperspace- run a python script, module, or command string remotely
- run a executable or shell command remotely
- use pip, pipenv, or a shell script to manage python dependencies
- Improved docs
- Minor doc fixes
- The default machineType for paperspace-python is now P5000, since GPU+ machines are not yet available for jobs.
- The default container image for paperspace-python is 'docker.io/paperspace/tensorflow-python' which adds python3 support to Google's 'gcr.io/tensorflow/tensorflow:1.5.0-gpu' container.
- This is the first official release of the Paperspace Python package.
- It provides support for Gradient jobs and running python modules as jobs.
- There are python packages available on PyPI: https://pypi.python.org/pypi/paperspace
- And on Anaconda.org for Conda: https://anaconda.org/paperspace/paperspace