Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[Do not merge] Ensure that new model instance has correct name or path in hf checkpointer #1612

Draft
wants to merge 8 commits into
base: main
Choose a base branch
from
Draft

[Do not merge] Ensure that new model instance has correct name or path in hf checkpointer #1612

Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
68f91dd
yo
snarayan21 Sep 22, 2024
36cc16a
Merge branch 'main' of https://github.com/mosaicml/llm-foundry
snarayan21 Sep 30, 2024
3c55e39
Merge branch 'main' of https://github.com/mosaicml/llm-foundry
snarayan21 Oct 2, 2024
70b46f6
Merge branch 'main' of https://github.com/mosaicml/llm-foundry
snarayan21 Oct 21, 2024
f7cfa58
Merge branch 'main' of https://github.com/mosaicml/llm-foundry
snarayan21 Oct 23, 2024
aba03d1
y
snarayan21 Oct 23, 2024
65a3873
yo
snarayan21 Oct 24, 2024
b244d5e
yo
snarayan21 Oct 24, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
12 changes: 6 additions & 6 deletions llmfoundry/callbacks/hf_checkpointer.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -600,12 +600,6 @@ def tensor_hook(
new_model_instance.load_state_dict(state_dict, assign=True)
del state_dict

# Transform the model and tokenizer before saving
new_model_instance, original_tokenizer = self.transform_model_and_tokenizer(
new_model_instance,
original_tokenizer,
)

# Ensure that the pretrained model name is correctly set on the saved HF checkpoint.
if self.pretrained_model_name is not None:
new_model_instance.name_or_path = self.pretrained_model_name
Expand All @@ -616,6 +610,12 @@ def tensor_hook(
k
].base_model_name_or_path = self.pretrained_model_name

# Transform the model and tokenizer before saving
new_model_instance, original_tokenizer = self.transform_model_and_tokenizer(
new_model_instance,
original_tokenizer,
)

log.debug('Saving Hugging Face checkpoint to disk')

if upload_to_save_folder:
Expand Down
Loading