6415249693
MoE call chain from ds_vllm (vllm-project/vllm latest): ex_engine/moe/ — 20 files, 8736 lines - modular_kernel.py (1630 lines) — base classes for modular MoE - experts/fused_batched_moe.py (972 lines) — NaiveBatchedExperts - prepare_finalize/batched.py (171 lines) — token grouping by expert - topk_weight_and_reduce.py (176 lines) — scatter-add finalize - fused_moe.py (1740 lines) — main fused_moe dispatch - config.py (1407 lines) — FusedMoEQuantConfig - activation.py, utils.py, layer.py, etc. xllm layer code (jd-opensource/xllm): ex_engine/xllm_layers/ — 39 files, 5859 lines - ilu/fused_moe.cpp (797 lines) — production ixformer 7-step MoE pipeline - ilu/attention.cpp (189 lines) — paged_attention + flash_attn bridge - npu_torch/qwen3_gated_delta_net_base.cpp (576 lines) — GDN reference - common/rms_norm.cpp, rotary_embedding.cpp, activation.cpp, dense_mlp.cpp xllm ILU kernels — synced 10 files to upstream (diffs from prior edits) These are reference implementations, NOT hand-written. Source repos: vllm-project/vllm, jd-opensource/xllm
51 lines
1.9 KiB
C++
51 lines
1.9 KiB
C++
/* Copyright 2025 The xLLM Authors. All Rights Reserved.
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
https://github.com/jd-opensource/xllm/blob/main/LICENSE
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
==============================================================================*/
|
|
|
|
#include "ilu_ops_api.h"
|
|
#include "utils.h"
|
|
|
|
using namespace ixformer;
|
|
|
|
namespace xllm::kernel::ilu {
|
|
|
|
void residual_layer_norm(torch::Tensor& input,
|
|
torch::Tensor& output,
|
|
std::optional<torch::Tensor>& residual,
|
|
torch::Tensor& weight,
|
|
std::optional<torch::Tensor>& bias,
|
|
std::optional<torch::Tensor>& residual_out,
|
|
double eps) {
|
|
auto residual_ = residual.value_or(torch::zeros_like(input));
|
|
torch::Tensor residual_out_ = residual_out.value_or(torch::zeros_like(input));
|
|
infer::residual_rms_norm(input,
|
|
residual_,
|
|
weight,
|
|
output,
|
|
residual_out_,
|
|
bias,
|
|
/*alpha=*/1.0,
|
|
eps,
|
|
false);
|
|
}
|
|
|
|
void rms_norm(torch::Tensor& output,
|
|
torch::Tensor& input,
|
|
torch::Tensor& weight,
|
|
double eps) {
|
|
std::optional<torch::Tensor> fused_bias = std::nullopt;
|
|
infer::rms_norm(input, weight, output, fused_bias, eps);
|
|
}
|
|
|
|
} // namespace xllm::kernel::ilu
|