We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
There was an error while loading. Please reload this page.
1 parent f2ebb6f commit ad971afCopy full SHA for ad971af
vllm/model_executor/layers/fused_moe/layer.py
@@ -437,7 +437,7 @@ def __init__(
437
# Use expert parallelism instead of tensor parallelism?
438
vllm_config = get_current_vllm_config()
439
use_ep = (vllm_config.parallel_config.enable_expert_parallel
440
- and self.tp_size > 1)
+ and self.tp_size * self.dp_size > 1)
441
442
# For smuggling this layer into the fused moe custom op
443
self.use_direct_call = self.dp_size == 1
0 commit comments