A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 15, 12:15 UTC
[Bugfix][ROCm] Fix import error on ROCm (#17495)
Signed-off-by: Gregory Shtrasberg <Gregory.Shtrasberg@amd.com>
Gregory Shtrasberg committed
Apr 30, 2025 at 22:51 UTC
08fb5587b4c4c1b031c08adf1b8608d710ab2585
1 file changed
+1
-1
vllm/model_executor/layers/rotary_embedding.py
+1
-1
@@ -32,7 +32,7 @@ from transformers import PretrainedConfig
32
from vllm.model_executor.custom_op import CustomOp
33
from vllm.platforms import current_platform
34
35
-if current_platform.is_cuda_alike():
35
+if current_platform.is_cuda():
36
from vllm.vllm_flash_attn.layers.rotary import apply_rotary_emb
37
38