[Model] Add LongCat-Flash (#23991)

Signed-off-by: yangxurui <yangxurui@meituan.com>
Co-authored-by: yangxurui <yangxurui@meituan.com>
This commit is contained in:
XuruiYang
2025-09-24 21:53:40 -07:00
committed by GitHub
co-authored by yangxurui
parent 90b139cfff
commit 845adb3ec6
31 changed files with 1357 additions and 66 deletions
@@ -329,7 +329,7 @@ class QuarkW8A8Fp8MoEMethod(QuarkMoEMethod):
raise NotImplementedError(
"EPLB not supported for `QuarkW8A8Fp8MoEMethod` yet.")
topk_weights, topk_ids = FusedMoE.select_experts(
topk_weights, topk_ids, _ = FusedMoE.select_experts(
hidden_states=x,
router_logits=router_logits,
use_grouped_topk=use_grouped_topk,
@@ -531,7 +531,7 @@ class QuarkW4A4MXFp4MoEMethod(QuarkMoEMethod):
from vllm.model_executor.layers.fused_moe import fused_experts
topk_weights, topk_ids = FusedMoE.select_experts(
topk_weights, topk_ids, _ = FusedMoE.select_experts(
hidden_states=x,
router_logits=router_logits,
use_grouped_topk=use_grouped_topk,