Skip to content

Conversation

@kiukchung
Copy link
Contributor

Summary:
Implements dumps and loads functions that dumps and loads INI file containing the runopts for all registered schedulers.

Here's a sample config file dumped by calling torchx.runner.config.dumps(std.out)

[default.local]
log_dir = None
prepend_cwd = False
image_type = dir
root_dir = /var/cache/fbpkg

[default.local_cwd]
log_dir = None
prepend_cwd = False

[default.local_par]
log_dir = None
prepend_cwd = False

[default.local_fbpkg]
log_dir = None
prepend_cwd = False

[default.mast]
hpcClusterUuid = TSCTestCluster
hpcEntitlementName = None
runningTimeoutSec = None
hpcIdentity = <FIXME_WITH_A_str_VALUE>
hpcJobOncall = pyper_training
useStrictName = False
mounts = None
localityConstraints = None
enableGracefulPreemption = False

[default.flow]
secure_group = <FIXME_WITH_A_str_VALUE>
entitlement = default
proxy_workflow_image = None

[default.quickflow]
entitlement = default
runningTimeoutSec = None
identity = <FIXME_WITH_A_str_VALUE>
oncall = None
useStrictName = False
mounts = None
localityConstraints = None
enableGracefulPreemption = False
perpetualRun = False

[default.slurm]
partition = None
time = None

[default.kubernetes]
namespace = default
queue = <FIXME_WITH_A_str_VALUE>

Added a //torchx/fb/example/.torchxconfig .

Next steps: document all this (will do this in a separate diff since its somewhat urgent we ship the functionality first)

Differential Revision: D31601700

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D31601700

@codecov
Copy link

codecov bot commented Oct 14, 2021

Codecov Report

Merging #249 (eda7182) into main (ffa8c48) will increase coverage by 0.29%.
The diff coverage is 100.00%.

Impacted file tree graph

@@            Coverage Diff             @@
##             main     #249      +/-   ##
==========================================
+ Coverage   91.26%   91.55%   +0.29%     
==========================================
  Files          58       60       +2     
  Lines        2667     2760      +93     
==========================================
+ Hits         2434     2527      +93     
  Misses        233      233              
Impacted Files Coverage Δ
torchx/cli/cmd_run.py 89.10% <ø> (ø)
torchx/cli/cmd_configure.py 100.00% <100.00%> (ø)
torchx/cli/main.py 93.33% <100.00%> (+0.15%) ⬆️
torchx/runner/api.py 96.35% <100.00%> (+0.08%) ⬆️
torchx/runner/config.py 100.00% <100.00%> (ø)
torchx/specs/api.py 99.27% <100.00%> (+<0.01%) ⬆️

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update ffa8c48...eda7182. Read the comment docs.

@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D31601700

@facebook-github-bot facebook-github-bot added the CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. label Oct 14, 2021
…m .INI file (#249)

Summary:
Pull Request resolved: #249

Implements `dumps` and `loads` functions that dumps and loads INI file containing the runopts for all registered schedulers.

Here's a sample config file dumped by calling `torchx.runner.config.dumps(std.out)`

```
[default.local]
log_dir = None
prepend_cwd = False
image_type = dir
root_dir = /var/cache/fbpkg

[default.local_cwd]
log_dir = None
prepend_cwd = False

[default.local_par]
log_dir = None
prepend_cwd = False

[default.local_fbpkg]
log_dir = None
prepend_cwd = False

[default.mast]
hpcClusterUuid = TSCTestCluster
hpcEntitlementName = None
runningTimeoutSec = None
hpcIdentity = <FIXME_WITH_A_str_VALUE>
hpcJobOncall = pyper_training
useStrictName = False
mounts = None
localityConstraints = None
enableGracefulPreemption = False

[default.flow]
secure_group = <FIXME_WITH_A_str_VALUE>
entitlement = default
proxy_workflow_image = None

[default.quickflow]
entitlement = default
runningTimeoutSec = None
identity = <FIXME_WITH_A_str_VALUE>
oncall = None
useStrictName = False
mounts = None
localityConstraints = None
enableGracefulPreemption = False
perpetualRun = False

[default.slurm]
partition = None
time = None

[default.kubernetes]
namespace = default
queue = <FIXME_WITH_A_str_VALUE>
```

Added a `//torchx/fb/example/.torchxconfig` .

Next steps: document all this (will do this in a separate diff since its somewhat urgent we ship the functionality first)

Reviewed By: aivanou

Differential Revision: D31601700

fbshipit-source-id: 03461b376452ae80939b29b72ab240c43419d29e
@facebook-github-bot
Copy link
Contributor

This pull request was exported from Phabricator. Differential Revision: D31601700

@facebook-github-bot
Copy link
Contributor

This pull request has been merged in 431c0e2.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

CLA Signed This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed. fb-exported Merged

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants