Skip to content

fix init weights issue for critic/reward model #983

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

Merged
merged 5 commits into from
Jul 8, 2025
Merged

fix init weights issue for critic/reward model #983

Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
1344ffd
Add file extension (#980)
hwchen2017 Jun 21, 2025
4c87e8d
fix init weights issue for critic/reward model
jouw Jun 26, 2025
9a7062b
Update submodule link to reflect https style (#981)
raviguptaamd Jul 4, 2025
c0c855b
Merge branch 'master' into fix-init-weights-issue
hwchen2017 Jul 6, 2025
8000b6f
fix formatting issue
jouw Jul 7, 2025
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
4 changes: 3 additions & 1 deletion applications/DeepSpeed-Chat/dschat/utils/model/model_utils.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,6 +11,7 @@
)
from huggingface_hub import snapshot_download
from transformers.integrations.deepspeed import HfDeepSpeedConfig
from transformers.modeling_utils import no_init_weights

from dschat.utils.model.reward_model import RewardModel
from dschat.utils.utils import load_state_dict_into_model, print_rank_0
Expand Down Expand Up @@ -99,7 +100,8 @@ def create_hf_model(model_class,
dschf = None
if rlhf_training:
# the weight loading is handled by create critic model
model = model_class.from_config(model_config)
with no_init_weights():
model = model_class.from_config(model_config)
else:
model = model_class.from_pretrained(
model_name_or_path,
Expand Down