apps/FastDeploy
1
0
Fork 0
mirror of https://github.com/PaddlePaddle/FastDeploy.git synced 2025-12-24 13:28:13 +08:00
Code Issues Actions 2 Packages Projects Releases Wiki Activity

[Metax] adapt cutlass moe for ernie-vl (#4685)

Browse Source
This commit is contained in:
Neil Zhu
2025-11-03 17:44:27 +08:00
committed by GitHub
parent 69c2f3cda1
commit c95d0740ec
6 changed files with 174 additions and 101 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
custom_ops/setup_ops.py
View File

@@ -614,6 +614,8 @@ elif paddle.device.is_compiled_with_custom_device("metax_gpu"):
"gpu_ops/text_image_gather_scatter.cu",
"gpu_ops/text_image_index_out.cu",
"gpu_ops/get_position_ids_and_mask_encoder_batch.cu",
"gpu_ops/limit_thinking_content_length_v1.cu",
"gpu_ops/limit_thinking_content_length_v2.cu",
"gpu_ops/append_attn/mla_cache_kernel.cu",
"gpu_ops/append_attn/get_block_shape_and_split_kv_block.cu",
"gpu_ops/moe/tritonmoe_preprocess.cu",