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 17, 00:30 UTC
[Misc] Missing error message for custom ops import (#5282)
Jie Fu (傅杰) committed
Jun 7, 2024 at 11:17 UTC
15063741e30881d7a982c3436c3299a0551327dc
1 file changed
+4
-2
vllm/_custom_ops.py
+4
-2
@@ -5,8 +5,10 @@ import torch
5
try:
6
from vllm._C import cache_ops as vllm_cache_ops
7
from vllm._C import ops as vllm_ops
8
-except ImportError:
9
- pass
8
+except ImportError as e:
9
+ from vllm.logger import init_logger
10
+ logger = init_logger(__name__)
11
+ logger.warning("Failed to import from vllm._C with %r", e)
12
13
14
# activation ops