Skip to content

Fix the llama3.2-11b/90b accuracy drop issue. #1175

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 1 commit into from
Apr 30, 2025
Merged

Fix the llama3.2-11b/90b accuracy drop issue. #1175

Show file tree
Hide file tree
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
2 changes: 1 addition & 1 deletion vllm/worker/hpu_enc_dec_model_runner.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -42,7 +42,7 @@
class HpuModelAdapterEncoderDecoder(HpuModelAdapter):

def __init__(self, model, vllm_config, layer_names, is_causal):
super().__init__(model, vllm_config, layer_names, False)
super().__init__(model, vllm_config, layer_names, is_causal)

# We only wrap the language model in HPU graph because some Ops in
# vision model will fallback to CPU and cause the graph building fail.
Expand Down
1 change: 0 additions & 1 deletion vllm/worker/hpu_worker.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -92,7 +92,6 @@ def __init__(
ModelRunnerClass = HPUPoolingModelRunner
elif is_encoder_decoder_model:
ModelRunnerClass = HPUEncoderDecoderModelRunner
is_causal = False
self.model_runner: HPUModelRunnerBase = ModelRunnerClass(
vllm_config=vllm_config,
kv_cache_dtype=self.cache_config.cache_dtype,
Expand Down