mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2026-04-23 00:17:25 +08:00
[Feature] Support mtp overlap schedule (#7001)
This commit is contained in:
@@ -62,6 +62,7 @@ __global__ void RebuildAppendPaddingKernel(T *output_data,
|
||||
i += gridDim.x * blockDim.x * VecSize) {
|
||||
const int out_token_id = i / dim_embed;
|
||||
const int bi = batch_id_per_token_output[out_token_id];
|
||||
if (bi < 0) continue;
|
||||
if (seq_len_this_time[bi] == 0) continue;
|
||||
if (seq_len_decoder[bi] == 0 && seq_len_encoder[bi] == 0) continue;
|
||||
|
||||
|
||||
Reference in New Issue
Block a user