saymrwulf/onnxruntime
1
0
Fork 0
mirror of https://github.com/saymrwulf/onnxruntime.git synced 2026-05-28 22:56:32 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Fix 4byte to 8byte static warnings in attention_quant.cc (#8715)

Browse source
This commit is contained in:
Nick Kreeger 2021-08-13 09:31:27 -05:00 committed by GitHub
parent 436ac6dd5f
commit d7baef765a
No known key found for this signature in database
GPG key ID: 4AEE18F83AFDEB23
1 changed files with 3 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
onnxruntime/contrib_ops/cpu/quantization/attention_quant.cc
View file

@ -215,8 +215,8 @@ Status QAttention<T>::Compute(OpKernelContext* context) const {
BufferUniquePtr gemm_buffer(gemm_data, BufferDeleter(allocator));
auto Q = reinterpret_cast<T*>(gemm_data);
auto K = Q + batch_size * sequence_length * hidden_size;
auto V = K + batch_size * sequence_length * hidden_size;
auto K = Q + static_cast<int64_t>(batch_size) * sequence_length * hidden_size;
auto V = K + static_cast<int64_t>(batch_size) * sequence_length * hidden_size;
T* QKV[3] = {Q, K, V};
{
@ -272,7 +272,7 @@ Status QAttention<T>::Compute(OpKernelContext* context) const {
gemm_params.BIsPacked = true;
} else {
gemm_params.B = weights_data + weights_offset;
gemm_params.ldb = 3 * hidden_size;
gemm_params.ldb = static_cast<int64_t>(3) * hidden_size;
}
gemm_params.ZeroPointB = nullptr != weight_zp_data ? weight_zp_data + weights_zp_offset : &weight_zp_default;
gemm_params.PerColumnZeroPoints = is_weight_zp_per_column;