[FSDP] Re-support model dtype change after FSDP init #91192
Closed
Conversation
This file contains hidden or 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]
awgu
requested review from
H-Huang,
kwen2501,
mrshenli,
pritamdamania87,
rohan-varma,
wanchaol and
zhaojuanmao
as code owners
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/91192
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit f2ea4a2: This comment was automatically generated by Dr. CI and updates every 15 minutes. |
pytorch-bot
bot
added
the
release notes: distributed (fsdp)
To make mixed precision precise internally, #90660 changed the implementation to save `_orig_param_dtype`, `_low_prec_param_dtype`, and `_reduce_dtype` explicitly. However, these are computed at FSDP construction time, so it does not allow the user to change the model dtype after FSDP construction time but before lazy initialization. This PR recomputes those dtype attributes as needed if the model dtype changes in that window. Note that any mixed precision settings specified by the user take precedence over the model dtype. [ghstack-poisoned]
awgu
pushed a commit
to awgu/pytorch
that referenced
this pull request
Dec 20, 2022
ghstack-source-id: c4fb143 Pull Request resolved: pytorch#91192
To make mixed precision precise internally, #90660 changed the implementation to save `_orig_param_dtype`, `_low_prec_param_dtype`, and `_reduce_dtype` explicitly. However, these are computed at FSDP construction time, so it does not allow the user to change the model dtype after FSDP construction time but before lazy initialization. This PR recomputes those dtype attributes as needed if the model dtype changes in that window. Note that any mixed precision settings specified by the user take precedence over the model dtype. [ghstack-poisoned]
awgu
pushed a commit
to awgu/pytorch
that referenced
this pull request
Dec 21, 2022
ghstack-source-id: 177caa7 Pull Request resolved: pytorch#91192
This was referenced Jan 5, 2023
Closes #90838. To make mixed precision precise internally, #90660 changed the implementation to save `_orig_param_dtype`, `_low_prec_param_dtype`, and `_reduce_dtype` explicitly. However, these are computed at FSDP construction time, so it does not allow the user to change the model dtype after FSDP construction time but before lazy initialization. This PR recomputes those dtype attributes as needed if the model dtype changes in that window. Note that any mixed precision settings specified by the user take precedence over the model dtype. [ghstack-poisoned]
Closes #90838. To make mixed precision precise internally, #90660 changed the implementation to save `_orig_param_dtype`, `_low_prec_param_dtype`, and `_reduce_dtype` explicitly. However, these are computed at FSDP construction time, so it does not allow the user to change the model dtype after FSDP construction time but before lazy initialization. This PR recomputes those dtype attributes as needed if the model dtype changes in that window. Note that any mixed precision settings specified by the user take precedence over the model dtype. [ghstack-poisoned]
awgu
pushed a commit
to awgu/pytorch
that referenced
this pull request
Jan 11, 2023
ghstack-source-id: cd846ec Pull Request resolved: pytorch#91192
This was referenced Jan 11, 2023
This was referenced Jan 11, 2023
zhaojuanmao
approved these changes
Jan 11, 2023
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:
r's current device iscuda:r#92035 [FSDP][RFC] Enforce rankr's current device iscuda:rdevice_id+ CPU offload test #92031 [FSDP][BE] Improvedevice_id+ CPU offload testprefixed_param_names->fqnsfor consolidation #92028 [FSDP][BE] Renameprefixed_param_names->fqnsfor consolidationuse_orig_params=True#91767 [FSDP] Do not clean FQNs even foruse_orig_params=Trueuse_orig_params=True,no_sync(), mixed precision #91193 [FSDP] Testuse_orig_params=True,no_sync(), mixed precisionMixedPrecisiondocs #91974 [FSDP] ClarifyMixedPrecisiondocsCloses #90838.
To make mixed precision precise internally, #90660 changed the implementation to save
_orig_param_dtype,_low_prec_param_dtype, and_reduce_dtypeexplicitly. However, these are computed at FSDP construction time, so it does not allow the user to change the model dtype after FSDP construction time but before lazy initialization. This PR recomputes those dtype attributes as needed if the model dtype changes in that window.Note that any mixed precision settings specified by the user take precedence over the model dtype.