apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-10-05 16:48:03 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Excutor] Experiment Feature-Support Prefill in cudagraph (#3459)

Browse Source 
* Support prefill in Cudagraph

* Refactor GetBlockShapeAndSplitKVBlock Kernel V2

* Refactor GetBlockShapeAndSplitKVBlock Kernel V2.1

* Refactor GetBlockShapeAndSplitKVBlock Kernel V2.2

* Refactor GetBlockShapeAndSplitKVBlock Kernel V2.3

* Refactor GetBlockShapeAndSplitKVBlock Kernel V2.4

* Refactor GetBlockShapeAndSplitKVBlock Kernel V2.5

* Solve problem about encoder_num_blocks_x_cpu

* Add early-exit mechanism for attention kernel

* fix test case about append-attention

* Update testcode, Add annotations to related tensors

* move get_input_length_list

* solve test_code

* Add annotations about early-exit for attention kernel

* Add annotations about early-exit for attention kernel2

* solve comment

* solve mtp

---------

Co-authored-by: RAM <gstian5555@outlook.com>
This commit is contained in:
Jundong Liu
2025-09-08 13:12:24 +08:00
committed by GitHub
parent 472402bf4e
commit 3d0aaa5923
21 changed files with 528 additions and 260 deletions
Download Patch File Download Diff File Expand all files Collapse all files

9
custom_ops/gpu_ops/cpp_extensions.cc
View File

@@ -299,7 +299,7 @@ paddle::Tensor OpenShmAndGetMetaSignalFunc(const int rank, const int device_id,
paddle::Tensor InitSignalLayerwiseFunc(const paddle::Tensor &kv_signal_metadata,
const int layer_id);
std::vector<paddle::Tensor> GetBlockShapeAndSplitKVBlock(
void GetBlockShapeAndSplitKVBlock(
const paddle::Tensor &seq_lens_encoder,
const paddle::Tensor &seq_lens_decoder,
const paddle::Tensor &seq_lens_this_time,
@@ -307,6 +307,13 @@ std::vector<paddle::Tensor> GetBlockShapeAndSplitKVBlock(
paddle::Tensor &decoder_tile_ids_per_batch, // Inplace
paddle::Tensor &decoder_num_blocks_x_cpu, // Inplace, Pinned Memory
paddle::Tensor &max_len_tensor_cpu, // Inplace, Pinned Memory
paddle::Tensor &encoder_batch_ids, // Inplace
paddle::Tensor &encoder_tile_ids_per_batch, // Inplace
paddle::Tensor &encoder_num_blocks_x_cpu, // Inplace, Pinned Memory
paddle::Tensor &kv_batch_ids, // Inplace
paddle::Tensor &kv_tile_ids_per_batch, // Inplace
paddle::Tensor &kv_num_blocks_x_cpu, // Inplace, Pinned Memory
paddle::Tensor &max_len_kv_cpu, // Inplace, Pinned Memory
const int encoder_block_shape_q,
const int decoder_block_shape_q,
const int group_size,