Remove incorrect deep_gemm assertions from server_args.py (#14113)
Co-authored-by: Kangyan Zhou <zky314343421@gmail.com>
This commit is contained in:
@@ -1517,11 +1517,6 @@ class ServerArgs:
|
||||
self.ep_size == 1
|
||||
), "FP8 Cutlass MoE is only supported with ep_size == 1"
|
||||
|
||||
if self.moe_runner_backend == "deep_gemm":
|
||||
assert (
|
||||
self.ep_size > 1 and self.moe_a2a_backend == "deepep"
|
||||
), "DeepGemm MoE runner is only supported when ep is enabled and moe_a2a_backend is deepep"
|
||||
|
||||
def _handle_a2a_moe(self):
|
||||
if self.moe_a2a_backend == "deepep":
|
||||
if self.deepep_mode == "normal":
|
||||
|
||||
Reference in New Issue
Block a user