8c265c5239
Remove the deprecated `pl.utilities.cli` module ( #16116 )
2022-12-19 23:24:25 +01:00
7bc55c38a7
Distributed sampling parity between Lite and PyTorch ( #16101 )
2022-12-19 21:57:15 +00:00
07b775fbe5
Fix test failing on master due to bad auto-merge ( #16118 )
2022-12-19 18:42:47 +00:00
2a85d9b257
[App] Add status endpoint, enable `ready` ( #16075 )
...
Co-authored-by: thomas chaton <thomas@grid.ai>
2022-12-19 14:10:58 +00:00
f3157f306a
Set the default work start method to spawn on MacOS ( #16089 )
2022-12-19 14:06:52 +00:00
0fd3d54205
[App] Scale out/in interval for autoscaler ( #16093 )
...
* Adding arguments for scale out/in interval
* Tests
2022-12-19 13:49:00 +00:00
aba5f12352
[App] Min replica=0 would break autoscaler component ( #16092 )
...
* fixing the bug where num_replica=0 would fail
* changelog
2022-12-19 13:16:04 +00:00
8a727c6243
Re-enable Lite CLI on Windows + PyTorch 1.13 ( #15645 )
...
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Justus Schock <12886177+justusschock@users.noreply.github.com>
2022-12-19 10:50:08 +00:00
39d27f6370
[App] Improve the autoscaler UI ( #16063 )
...
[App] Improve the autoscaler UI (#16063 )
2022-12-19 10:25:24 +00:00
80b7fafcea
[App] Add work.delete ( #16103 )
...
Co-authored-by: thomas <thomas@thomass-MacBook-Pro.local>
2022-12-19 11:01:05 +01:00
e54f4f58be
Fix detection of whether app is running in cloud ( #16045 )
2022-12-19 00:39:21 +00:00
22b254f491
[App] Add display name property to the work ( #16095 )
...
Co-authored-by: thomas <thomas@thomass-MacBook-Pro.local>
2022-12-18 12:54:15 +00:00
3e8319d422
Remove redundant `find_unused_parameters=False` in Lite ( #16026 )
2022-12-17 19:00:37 +00:00
23013be28a
[App] Serve datatypes with better client code ( #16018 )
2022-12-16 16:03:08 +00:00
3d509f6902
Update Multinode Warning ( #16091 )
2022-12-16 15:13:34 +00:00
592b12658a
[App] PoC: Add support for Request ( #16047 )
2022-12-16 14:19:10 +00:00
b1ce2639f4
Better check for programmatic lightningignore ( #16080 )
...
Co-authored-by: Jirka Borovec <6035284+Borda@users.noreply.github.com>
2022-12-16 09:49:17 +00:00
5f7403e99c
[App] Fix bug where previously deleted apps cannot be re-run from the CLI ( #16082 )
2022-12-16 04:06:05 +01:00
cfe64875d6
Drop FairScale sharded parity tests ( #16069 )
2022-12-15 17:34:54 +01:00
3b323c842d
fix(cloud): detect and ignore venv ( #16056 )
...
Co-authored-by: Ethan Harris <ethanwharris@gmail.com>
2022-12-15 14:00:32 +00:00
06e1f19f42
Remove the deprecated profiler imports ( #16059 )
2022-12-15 11:02:33 +01:00
04f44486f6
Remove the deprecated `GPUAccelerator` ( #16050 )
2022-12-15 00:17:30 +01:00
64d0ebbd9b
Add guards to cluster deletion from cli ( #16053 )
...
Adds guards to cluster deletion.
- If cluster has running apps -> throw an error
- If cluster has stopped apps -> confirm w/ user that apps and logs will be deleted
2022-12-14 20:38:03 +00:00
e522a12d17
[App] Improve lightning connect experience ( #16035 )
2022-12-14 17:09:30 +00:00
d3a722608e
[App] Support running on multiple clusters ( #16016 )
2022-12-14 16:48:49 +00:00
9a24635c6d
Remove the deprecated `LightningDeepSpeedModule` ( #16041 )
2022-12-14 15:04:21 +01:00
a58639ce7e
Nightly PyTorch version is now 2.0 ( #16017 )
2022-12-14 12:37:00 +01:00
3e664c906b
Add info message for Ampere GPUs to enable tf32 matmuls ( #16037 )
2022-12-13 19:24:31 +01:00
53bf71475a
Remove deperecated code in pl.utilities.meta ( #16038 )
2022-12-13 12:07:30 -05:00
edd2b4259a
Introduce `{Work,Flow}.lightningignore` ( #15818 )
2022-12-13 13:50:08 +00:00
53ceb156c4
Integrate lightning_utilities==0.4.2 ( #15817 )
...
Co-authored-by: Jirka Borovec <6035284+Borda@users.noreply.github.com>
2022-12-13 13:13:51 +00:00
fcd3195e68
Fix DDPStrategy import in app framework after #14952 ( #16029 )
2022-12-13 10:01:33 +00:00
40596f05dd
Merge DDPStrategy and DDPSpawnStrategy in Lite ( #14952 )
...
Co-authored-by: Jirka Borovec <Borda@users.noreply.github.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2022-12-13 08:52:59 +00:00
ed52823c3f
LightningCLI support for optimizers and schedulers via dependency injection ( #15869 )
...
Co-authored-by: Jirka Borovec <6035284+Borda@users.noreply.github.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2022-12-12 15:36:19 +00:00
38acba08fc
Batch `MLFlowLogger` requests ( #15915 )
...
Co-authored-by: Jake Schmidt <jake.schmidt@utexas.edu>
Co-authored-by: awaelchli <aedu.waelchli@gmail.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2022-12-12 14:22:48 +00:00
2577285dd5
Use LRScheduler for torch >= 1.14 otherwise use _LRScheduler ( #15768 )
...
Co-authored-by: Jirka Borovec <6035284+Borda@users.noreply.github.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
2022-12-12 13:33:26 +00:00
c1d0156e1d
[App] Fix `AutoScaler` trying to replicate multiple works in a single machine ( #15991 )
...
* dont try to replicate new works in the existing machine
* update chglog
* Update comment
* Update src/lightning_app/components/auto_scaler.py
* add test
2022-12-11 00:56:46 +00:00
9ed43c64b6
Set the logger explicitly in tests ( #15815 )
2022-12-10 13:32:37 +00:00
1657ea8a39
[App] Install exact version whn upgrading and not when testing ( #15984 )
...
* [App] Install exact version whn upgrading and not when testing
* Update CHANGELOG.md
Co-authored-by: Jirka Borovec <6035284+Borda@users.noreply.github.com>
2022-12-09 14:47:36 +00:00
90a4c0289d
Fix cloudcomputes registration for structures ( #15964 )
...
* fix cloudcomputes
* updates cloudcompute registration
* changelog
2022-12-09 12:46:49 +00:00
d21b8992ee
Fix multinode cloud component ( #15965 )
...
* fix multinode cloud component
* add tests
2022-12-09 12:02:58 +00:00
dd83587102
[App] Resolve run installation ( #15974 )
2022-12-09 11:27:46 +01:00
edc9986084
Apply dynamo to training_step, validation_step, test_step, predict_step ( #15957 )
...
* Apply dynamo to training_step, validation_step, test_step, predict_step
* Add entry to CHANGELOG.md
2022-12-08 21:02:31 +01:00
4983083639
[App] Enable running an app from the Gallery ( #15941 )
...
Co-authored-by: thomas <thomas@thomass-MacBook-Pro.local>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Ethan Harris <ethanwharris@gmail.com>
Co-authored-by: Jirka <jirka.borovec@seznam.cz>
2022-12-08 18:36:16 +00:00
b5fa8968ec
Make LightningModule torch.jit.script-able again ( #15947 )
...
* Make LightningModule torch.jit.script-able again
* remove skip
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2022-12-08 18:02:27 +01:00
67a47d47e1
[App] Add automatic conversion to structures ( #15961 )
2022-12-08 16:13:11 +00:00
15184c694c
Fix restarting attribute for lr finder ( #15620 )
2022-12-08 14:37:29 +01:00
3004f13d36
Lite: Fix DataLoader shuffling when using DistributedSampler ( #15931 )
...
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
2022-12-08 12:50:52 +00:00
36aecde695
Multinode on MPS ( #15748 )
...
* Fix restarting attribute for lr finder
* update lite executor
* update trainer executor
* update spawn executor
* add multinode component tests
* add testing helpers
* add lite tests
* add trainer tests
* update changelog
* update trainer
* update workflow
* update tests
* debug
* add reason for skipif
* Apply suggestions from code review
* switch skipif
Co-authored-by: Jirka <jirka.borovec@seznam.cz>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: Carlos Mocholí <carlossmocholi@gmail.com>
Co-authored-by: Adrian Wälchli <aedu.waelchli@gmail.com>
Co-authored-by: Jirka Borovec <6035284+Borda@users.noreply.github.com>
2022-12-08 12:05:18 +00:00
df678335d6
[App] Multiprocessing-safe work pickling ( #15836 )
2022-12-08 07:23:29 +00:00
Carlos Mocholí