[WIP] zero bubble #546
Draft
[WIP] zero bubble #546
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
[ghstack-poisoned]
H-Huang
added a commit
that referenced
this pull request
Aug 20, 2024
ghstack-source-id: fd042c2482ffeac2a9b9bd53e29e803858875cca Pull Request resolved: #546
facebook-github-bot
added
the
CLA Signed
To run zb test:
`python test_runner.py ./test_out --test pp_zb`
TODO:
- zero bubble when AC is turned off is failing when using multiple hosts:
```
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/stage.py", line 668, in backward_weight_one_chunk
dweights = self.dw_runner.pop(bwd_chunk_id)(
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/_backward.py", line 251, in stage_backward_weight
dweight = all_dweights[grad_acc]
KeyError: <AccumulateGrad object at 0x7ff490125b10>
```
[ghstack-poisoned]
H-Huang
added a commit
that referenced
this pull request
Aug 28, 2024
ghstack-source-id: c6b2f05d19e207232306320decf506316c066347 Pull Request resolved: #546
To run zb test:
`python test_runner.py ./test_out --test pp_zb`
TODO:
- zero bubble when AC is turned off is failing when using multiple hosts:
```
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/stage.py", line 668, in backward_weight_one_chunk
dweights = self.dw_runner.pop(bwd_chunk_id)(
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/_backward.py", line 251, in stage_backward_weight
dweight = all_dweights[grad_acc]
KeyError: <AccumulateGrad object at 0x7ff490125b10>
```
[ghstack-poisoned]
H-Huang
added a commit
that referenced
this pull request
Sep 5, 2024
ghstack-source-id: bdcb4b3eba1b3ab35bedb745a1f334101c259ee7 Pull Request resolved: #546
To run zb test:
`python test_runner.py ./test_out --test pp_zb`
TODO:
- zero bubble when AC is turned off is failing when using multiple hosts:
```
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/stage.py", line 668, in backward_weight_one_chunk
dweights = self.dw_runner.pop(bwd_chunk_id)(
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/_backward.py", line 251, in stage_backward_weight
dweight = all_dweights[grad_acc]
KeyError: <AccumulateGrad object at 0x7ff490125b10>
```
[ghstack-poisoned]
H-Huang
added a commit
that referenced
this pull request
Sep 5, 2024
ghstack-source-id: d813ea1be1375567b1eda4f418bfae9d6fbdd84c Pull Request resolved: #546
To run zb test:
`python test_runner.py ./test_out --test pp_zb`
TODO:
- zero bubble when AC is turned off is failing when using multiple hosts:
```
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/stage.py", line 668, in backward_weight_one_chunk
dweights = self.dw_runner.pop(bwd_chunk_id)(
File "/packages/torchtitan_additional_packages/torchtitan/torchtitan/parallelisms/pipelining/_backward.py", line 251, in stage_backward_weight
dweight = all_dweights[grad_acc]
KeyError: <AccumulateGrad object at 0x7ff490125b10>
```
[ghstack-poisoned]
H-Huang
added a commit
that referenced
this pull request
Sep 26, 2024
ghstack-source-id: 954580ed2484b02b2d0a0e205de7239b2cb8e3df Pull Request resolved: #546
To run zb test: `python test_runner.py ./test_out --test pp_zb` internal mast run: ` torchx run mast.py:train --additional_folders /home/howardhuang/local/torchtitan --twtask_bootstrap_script run_torchtitan.sh --h "grandteton" --nodes 8 train_configs/debug_model_3d_mast.toml ` [ghstack-poisoned]
H-Huang
added a commit
that referenced
this pull request
Sep 26, 2024
ghstack-source-id: ec662ad4cdf9b4e12d9b334e09d2658926b25ace Pull Request resolved: #546
To run zb test: `python test_runner.py ./test_out --test pp_zb` internal mast run: ` torchx run mast.py:train --additional_folders /home/howardhuang/local/torchtitan --twtask_bootstrap_script run_torchtitan.sh --h "grandteton" --nodes 8 train_configs/debug_model_3d_mast.toml ` [ghstack-poisoned]
H-Huang
added a commit
that referenced
this pull request
Sep 30, 2024
ghstack-source-id: 9c557a4a76a019f3727985e777a1deb66c3ca941 Pull Request resolved: #546
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):
To run zb test:
python test_runner.py ./test_out --test pp_zbinternal mast run:
torchx run mast.py:train --additional_folders /home/howardhuang/local/torchtitan --twtask_bootstrap_script run_torchtitan.sh --h "grandteton" --nodes 8 train_configs/debug_model_3d_mast.toml