[Bugfix] Qwen3Next support FlashComm1 (#6830)
### What this PR does / why we need it?
Support FlashComm1 for Qwen3-Next. Fix some padding problems in Sequence
Parallel (SP)
and resolve precision problems in shared_out when both FlashComm1 is
enabled.
### Does this PR introduce _any_ user-facing change?
No
### How was this patch tested?
CI
- vLLM version: v0.15.0
- vLLM main:
83b47f67b1
---------
Signed-off-by: zhaojiangjiang <zhaojiangjiang1@h-partners.com>
Co-authored-by: zhaojiangjiang <zhaojiangjiang1@h-partners.com>
This commit is contained in:
@@ -705,7 +705,11 @@ def _get_row_parallel_op(
|
||||
|
||||
|
||||
def get_parallel_op(disable_tp, prefix, layer, direct):
|
||||
if disable_tp or ("shared_experts" in prefix and shared_expert_dp_enabled()):
|
||||
if (
|
||||
disable_tp
|
||||
or ("shared_experts" in prefix and shared_expert_dp_enabled())
|
||||
or ("shared_expert" in prefix and shared_expert_dp_enabled())
|
||||
):
|
||||
return None, 0, 1
|
||||
custom_op: (
|
||||
MLPColumnParallelOp
|
||||
|
||||
Reference in New Issue
Block a user