xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-06-02 09:37:53 +08:00
Code Issues Packages Projects Releases Wiki Activity

remove modular kernel

Browse Source 
Signed-off-by: Sage Moore <sage@neuralmagic.com>
This commit is contained in:
Sage Moore 2025-06-02 18:03:21 +00:00
parent 6645882e95
commit d6eca0c130
1 changed files with 0 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
vllm/model_executor/layers/fused_moe/modular_kernel.py
View File

@ -3,7 +3,6 @@ from abc import ABC, abstractmethod
from typing import Optional from typing import Optional
import torch import torch
from vllm.v1.worker.ubatching import get_current_ubatch_context, dump_ubatching_state
# #
# This file defines a set of base classes used to make MoE kernels more modular. # This file defines a set of base classes used to make MoE kernels more modular.
@ -359,9 +358,6 @@ class FusedMoEModularKernel(torch.nn.Module):
expert_num_tokens=expert_num_tokens, expert_num_tokens=expert_num_tokens,
) )
# if (ubatch_ctx := get_current_ubatch_context()) is not None:
# print("in modular moe3, ubatch:", ubatch_ctx.id, self.fused_experts)
self.prepare_finalize.finalize(output, fused_out, topk_weights, self.prepare_finalize.finalize(output, fused_out, topk_weights,
topk_ids, apply_router_weight_on_input) topk_ids, apply_router_weight_on_input)