Skip to content

Fix early CUDA initialisation #41409

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
Oct 7, 2025
+2 −1
Merged

Fix early CUDA initialisation #41409

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
3 changes: 2 additions & 1 deletion src/transformers/generation/continuous_batching/continuous_api.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,7 +27,6 @@

from ...configuration_utils import PreTrainedConfig
from ...generation.configuration_utils import GenerationConfig
from ...integrations.hub_kernels import load_and_register_kernel
from ...utils.logging import logging
from ...utils.metrics import ContinuousBatchProcessorMetrics, attach_tracer, traced
from .cache import PagedAttentionCache
Expand Down Expand Up @@ -609,6 +608,8 @@ def __init__(
"""
self.model = model.eval()
if "paged|" not in model.config._attn_implementation:
from ...integrations.hub_kernels import load_and_register_kernel

attn_implementation = "paged|" + self.model.config._attn_implementation
load_and_register_kernel(attn_implementation)
model.set_attn_implementation(attn_implementation)
Expand Down