Add a 3-stage PP config #345
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This was referenced May 17, 2024
Merged
facebook-github-bot
added
the
CLA Signed
wconstab
added a commit
that referenced
this pull request
May 18, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. In any case, I only discovered this was not working when I tried to use it for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. I ran into issues with DCP loading my initial seed checkpoint. ghstack-source-id: 4d6072eb3e8adc1431afa27fb552c72ba4c26967 Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 18, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. ghstack-source-id: e7a6f933be1d765425f755332a2305c3f8787904 Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 18, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: e7a6f933be1d765425f755332a2305c3f8787904 Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 20, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: b6fd58261ae7d38107cf81718bfbb8bfd40acba5 Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 20, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: f447b55db41fa474b9882aedc7d618ca88f9d9ff Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 20, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: a120318e49d53a61f2b54080de1c9882c4c2016e Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 21, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: 72d366936cb4fd10de6c3a3ccea7f6d7e5089dfd Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 21, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: ec1892c5f68d7bb6cb9ec15cddcf8f606f0ecbfe Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 22, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: d41ffd0e6b11c63e3ca06141b704f659d5054737 Pull Request resolved: #345
wconstab
added a commit
that referenced
this pull request
May 22, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: 289eeb8473afa84e3b767986f9fb285f1d91fbf2 Pull Request resolved: #345
tianyu-l
pushed a commit
that referenced
this pull request
Aug 16, 2024
Pipelining is unique in that there is no need to stick to power-of-2 numbers of stages, and there maybe reasons an odd number is optimal depending on how you divide up your cluster. Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated than 2-stage but simpler than 4-stage setup. seems to run fine, if run with an even batch size (`--training.batch_size 12`) ghstack-source-id: 289eeb8473afa84e3b767986f9fb285f1d91fbf2 Pull Request resolved: #345
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
Pipelining is unique in that there is no need to stick to power-of-2
numbers of stages, and there maybe reasons an odd number is optimal
depending on how you divide up your cluster.
Anyway, I use this for validation of the 1f1b schedule in a slightly-more-complicated
than 2-stage but simpler than 4-stage setup.
seems to run fine, if run with an even batch size
(
--training.batch_size 12)