mirror of
https://github.com/PaddlePaddle/FastDeploy.git
synced 2026-04-23 00:17:25 +08:00
@@ -694,9 +694,9 @@ __global__ void quant2d_per_channel_cached(
|
||||
constexpr int PINGPONG_LM_LEN = 512;
|
||||
constexpr int LM_LEN = PINGPONG_LM_LEN * 2;
|
||||
constexpr int INPUT_BUF_LEN = 4 * 1024;
|
||||
constexpr int OUPUT_BUF_LEN = 2 * 1024;
|
||||
constexpr int OUTPUT_BUF_LEN = 2 * 1024;
|
||||
constexpr int INPUT_MAX_LEN = INPUT_BUF_LEN / sizeof(TX);
|
||||
constexpr int OUTPUT_MAX_LEN = OUPUT_BUF_LEN / sizeof(TY);
|
||||
constexpr int OUTPUT_MAX_LEN = OUTPUT_BUF_LEN / sizeof(TY);
|
||||
constexpr int INPUT_SIZE =
|
||||
INPUT_MAX_LEN < OUTPUT_MAX_LEN ? INPUT_MAX_LEN : OUTPUT_MAX_LEN;
|
||||
__simd__ __shared__ float
|
||||
|
||||
Reference in New Issue
Block a user