apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-04 00:06:38 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

集中式支持fa3 (#3112)

Browse Source
This commit is contained in:
yangjianfengo1
2025-08-01 18:03:36 +08:00
committed by GitHub
parent bdb83e007d
commit 64d7a3194d
4 changed files with 257 additions and 25 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
custom_ops/gpu_ops/cpp_extensions.cc
View File

@@ -761,6 +761,17 @@ void SpeculateStepPaddle(
const int encoder_decoder_block_num,
const int max_draft_tokens);
void MergePrefillDecodeOutput(
const paddle::Tensor &encoder_res,
const paddle::Tensor &decoder_res,
const paddle::Tensor &seq_lens_encoder,
const paddle::Tensor &seq_lens_decoder,
const paddle::Tensor &seq_lens_this_time,
const paddle::Tensor &cu_seq_q,
const int head_num,
const int head_dim,
const int max_token);
PYBIND11_MODULE(fastdeploy_ops, m) {
m.def("get_expert_token_num", &GetExpertTokenNum, py::arg("topk_ids"),
@@ -1111,4 +1122,6 @@ PYBIND11_MODULE(fastdeploy_ops, m) {
m.def("mtp_step_paddle",&MTPStepPaddle, "mtp_step_paddle function");
m.def("speculate_step_paddle",&SpeculateStepPaddle, "speculate_step_paddle function");
m.def("merge_prefill_decode_output", &MergePrefillDecodeOutput, "merge_prefill_decode_output function");
}