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.

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

StaticLLMPipeline: Decide when to enable NPUW_DQ_FULL property #1258

Merged
merged 4 commits into from
Nov 28, 2024
Merged
Changes from 3 commits
Commits
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
16 changes: 15 additions & 1 deletion src/cpp/src/llm_pipeline_static.cpp
Original file line number Diff line number Diff line change
Expand Up @@ -457,6 +457,7 @@ void merge_config_with(ov::AnyMap& lhs, const ov::AnyMap& rhs) {
struct NPUDesc {
std::string arch;
int64_t max_tiles;
bool compiler_dq;
Comment on lines 458 to +460
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think it is worth to have default values here, probably even incorrect (e.g., 0 as max_tiles)

};

std::optional<NPUDesc> extract_npu_descriptor(ov::Core& core) {
Expand All @@ -466,7 +467,14 @@ std::optional<NPUDesc> extract_npu_descriptor(ov::Core& core) {
}
const auto arch = core.get_property("NPU", ov::device::architecture);
const auto max_tiles = core.get_property("NPU", ov::intel_npu::max_tiles);
return std::make_optional(NPUDesc{arch, max_tiles});

bool compiler_dq = false;
const auto device_caps = core.get_property("NPU", ov::device::capabilities);
if (std::find(device_caps.begin(), device_caps.end(),
"COMPILER_DYNAMIC_QUANTIZATION") != device_caps.end()) {
compiler_dq = true;
}
Comment on lines +471 to +476
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Suggested change
bool compiler_dq = false;
const auto device_caps = core.get_property("NPU", ov::device::capabilities);
if (std::find(device_caps.begin(), device_caps.end(),
"COMPILER_DYNAMIC_QUANTIZATION") != device_caps.end()) {
compiler_dq = true;
}
const bool compiler_dq = caps.end() != std::find(caps.begin(), caps.end(), "COMPILER_DYNAMIC_QUANTIZATION")

return std::make_optional(NPUDesc{arch, max_tiles, compiler_dq});
}

ov::AnyMap get_baseline_common_config() {
Expand Down Expand Up @@ -508,6 +516,9 @@ ov::AnyMap get_default_prefill_config(const std::shared_ptr<ov::Model>& model,
npudesc->max_tiles != -1) {
config.emplace("NPU_DPU_GROUPS", npudesc->max_tiles);
}
if (npudesc.has_value() && npudesc->compiler_dq) {
config.emplace("NPUW_DQ_FULL", "NO");
}
return config;
}

Expand All @@ -523,6 +534,9 @@ ov::AnyMap get_default_generate_config(const std::shared_ptr<ov::Model>& model,
if (npudesc.has_value() && npudesc->arch == "4000") {
config.emplace("NPU_DPU_GROUPS", 4);
}
if (npudesc.has_value() && npudesc->compiler_dq) {
config.emplace("NPUW_DQ_FULL", "NO");
}
return config;
}

Expand Down
Loading