Skip to content

sycl: 2-3x prompt processing speedup by dequantizing PTQ1_0 and PQ2_0 to FP16 #278

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
dwymark wants to merge 1 commit into
base: prism
Choose a base branch
from
+2 −1
Open

sycl: 2-3x prompt processing speedup by dequantizing PTQ1_0 and PQ2_0 to FP16 #278

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 2 additions & 1 deletion ggml/src/ggml-sycl/ggml-sycl.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2682,7 +2682,8 @@ inline void ggml_sycl_op_mul_mat_sycl(
#ifdef GGML_SYCL_F16
bool use_fp16 = true; // TODO(Yu) SYCL capability check
#else
bool use_fp16 = false;
// ternary weights are exact in FP16 and the FP16 GEMM is much faster than FP32
bool use_fp16 = src0->type == GGML_TYPE_PQ2_0 || src0->type == GGML_TYPE_PTQ1_0;
#endif

#if GGML_SYCL_DNNL && defined(GGML_SYCL_HAS_BF16)
Expand Down