Skip to content
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

[Snippets][CPU] Disable MHA tokenization in LLM #28601

Open
wants to merge 2 commits into
base: master
Choose a base branch
from
+14 −5
Open

[Snippets][CPU] Disable MHA tokenization in LLM #28601

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
a764856
[Snippets][CPU] Disable MHA tokenization in LLM
a-sidorova Jan 22, 2025
8e13f0c
[Snippets][CPU] Added PagedAttentionExtension to check
a-sidorova Jan 23, 2025
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
19 changes: 14 additions & 5 deletions src/plugins/intel_cpu/src/transformations/transformation_pipeline.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1031,18 +1031,27 @@ void Transformations::MainSnippets(void) {
}
CPU_REGISTER_PASS_COMMON(snippetsManager, snippets::pass::SnippetsTokenization, tokenization_config);

// - CPU Plugin Subgraph supports f32, bf16, quantized and fp16(on avx_512_core_amx_fp16 target) BRGEMM
const bool isMHASupported =
#if defined(OPENVINO_ARCH_ARM64)
false;
#else
#if defined(OPENVINO_ARCH_X86_64)
// Currently, Snippets don't provide efficient execution for single token inference in LLM case.
// To avoid performance degradations, we disable MHA tokenization into Subgraphs in LLMs'.
// We consider the presence of `ScaledDotProductAttentionWithKVCache` and `PagedAttentionExtension` ops
// in the model as a sign that this model is LLM.
const auto is_LLM = ov::op::util::has_op_with_type<intel_cpu::ScaledDotProductAttentionWithKVCache>(model) ||
ov::op::util::has_op_with_type<ov::op::PagedAttentionExtension>(model);

// CPU Plugin Subgraph supports f32, bf16, quantized and fp16(on avx_512_core_amx_fp16 target) BRGEMM
const auto is_infer_prc_supported_by_MHA =
(dnnl::impl::cpu::x64::mayiuse(dnnl::impl::cpu::x64::avx2) &&
one_of(config.inferencePrecision, ov::element::f32, element::undefined)) ||
(dnnl::impl::cpu::x64::mayiuse(dnnl::impl::cpu::x64::avx512_core) &&
one_of(config.inferencePrecision, ov::element::bf16, ov::element::f32, element::undefined)) ||
(dnnl::impl::cpu::x64::mayiuse(dnnl::impl::cpu::x64::avx512_core_amx_fp16) &&
one_of(config.inferencePrecision, ov::element::f16));
const bool isMHASupported = !is_LLM && is_infer_prc_supported_by_MHA;
#else
const bool isMHASupported = false;
#endif

if (!isMHASupported) {
CPU_DISABLE_PASS_COMMON(snippetsManager, snippets::pass::TokenizeMHASnippets);
CPU_DISABLE_PASS_COMMON(snippetsManager, snippets::pass::ExtractReshapesFromMHA);
Expand Down
Loading