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

fix backward error in DDP when running reward model finetune in RLHF #507

Merged
merged 1 commit into from
Jan 8, 2024
Merged

fix backward error in DDP when running reward model finetune in RLHF #507

Changes from all commits
Commits
Show all changes
1 commit
Select commit
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
fix backward error in DDP when running reward model finetune in RLHF 
Signed-off-by: Wang, Yi A <yi.a.wang@intel.com>
  • Loading branch information
@sywangyi
sywangyi committed Jan 8, 2024
commit 361b25272de9667649ed717b2ddda03d863122bc
5 changes: 4 additions & 1 deletion optimum/habana/transformers/models/llama/modeling_llama.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -599,6 +599,9 @@ def prepare_inputs_for_generation(

def apply_customized_rope(q, k, cos, sin, position_ids):
if q.device.type == "hpu" and FusedRoPE:
return FusedRoPE.apply(q, cos, sin, position_ids), FusedRoPE.apply(k, cos, sin, position_ids)
# TODO: remove `.clone()` when SynapseAI v1.15 is released
return FusedRoPE.apply(q, cos.clone(), sin.clone(), position_ids), FusedRoPE.apply(
k, cos.clone(), sin.clone(), position_ids
)
else:
return apply_rotary_pos_emb(q, k, cos, sin, position_ids)
Loading