floyd restart
Restat/re-run a previous job
Usage¶
floyd restart JOB_NAME [OPTIONS] [COMMAND]
Options¶
Name, shorthand | Description |
---|---|
JOB_NAME | Name or ID of your job. |
All the [OPTIONS] and [COMMAND] of floyd run as documented here |
Description¶
Use this command to restart/re-run a previous job. Any [OPTIONS]
or
[COMMAND]
passed to floyd restart
will override the [OPTIONS]
or
[COMMAND]
of the original job. This can be useful if you want to
restart/re-run a previous job with different [OPTIONS]
(like a different
instance type), or override its [COMMAND]
.
When using floyd restart
, a copy of your code is not uploaded (since the code
of the previous job is used). This can save on startup time.
This command can take a shortened job name.
Examples¶
# Override the command $ floyd restart mckay/projects/mnist/1 "python train.py"
# Run the job on a GPU server $ floyd restart mckay/projects/mnist/1 --gpu
# Run the job with a new version of a dataset $ floyd restart mckay/projects/mnist/1 --data mckay/datasets/mnist/1:mnist
Help make this document better¶
This guide, as well as the rest of our docs, are open-source and available on GitHub. We welcome your contributions.
- Suggest an edit to this page (by clicking the edit icon at the top next to the title).
- Open an issue about this page to report a problem.