apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2026-04-23 00:17:25 +08:00
Code Issues Actions 19 Packages Projects Releases Wiki Activity

[OP]Unify MoE op with moe_permute path for bf16 GLM (#7164)

Browse Source
This commit is contained in:
fxyfxy777
2026-04-09 16:17:56 +08:00
committed by GitHub
parent 33682c6749
commit 39ff38aba1
5 changed files with 444 additions and 69 deletions
Download Patch File Download Diff File Expand all files Collapse all files
custom_ops/gpu_ops/cpp_extensions.cc
+3 -1
View File
@@ -537,7 +537,9 @@ std::vector<paddle::Tensor> TextImageGatherScatter(
const bool is_scatter);
std::vector<paddle::Tensor> count_tokens_per_expert_func(
const paddle::Tensor& topk_ids, int64_t num_experts);
const paddle::Tensor& topk_ids,
int64_t num_experts,
bool compute_padded_cumsum = false);
void GetPositionIdsAndMaskEncoderBatch(
const paddle::Tensor& seq_lens_encoder,
const paddle::Tensor& seq_lens_decoder,