Skip to content

Make cache_dir and num_stream options session-local #809

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

Merged
merged 3 commits into from
Sep 12, 2025
Merged

Make cache_dir and num_stream options session-local #809

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
4d4d29b
Make cache_dir and num_stream options session-local
mklimenk Sep 5, 2025
2d0ec65
Merge branch 'ovep-develop' into private/mklimenk/per_model_cache_dir
preetha-intel Sep 11, 2025
e9791da
Merge branch 'ovep-develop' into private/mklimenk/per_model_cache_dir
mklimenk Sep 11, 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
12 changes: 6 additions & 6 deletions onnxruntime/core/providers/openvino/backends/basic_backend.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -242,13 +242,13 @@ void BasicBackend::PopulateConfigValue(ov::AnyMap& device_config) {
}
}

void BasicBackend::EnableCaching() {
void BasicBackend::EnableCaching(ov::AnyMap& device_config) {
// cache_dir argument has no effect when working with an embed-mode EPContext Graph
if (subgraph_context_.is_ep_ctx_graph) return;

if (!session_context_.cache_dir.empty() && !session_context_.so_context_enable) {
LOGS_DEFAULT(INFO) << log_tag << "Enables Caching";
OVCore::Get()->SetCache(session_context_.cache_dir.string());
device_config.emplace(ov::cache_dir(session_context_.cache_dir.string()));
}
}

Expand All @@ -262,7 +262,7 @@ void BasicBackend::EnableGPUThrottling(ov::AnyMap& device_config) {
}
}

void BasicBackend::EnableStreams() {
void BasicBackend::EnableStreams(ov::AnyMap& device_config) {
// Return silently for NPU as it's currently treated as a read-only flag by the NPU plugin
// and throws an exception for the same
if (session_context_.device_type.find("NPU") != std::string::npos)
Expand All @@ -279,7 +279,7 @@ void BasicBackend::EnableStreams() {
}
// Do nothing
} else {
OVCore::Get()->SetStreams(session_context_.device_type, session_context_.num_streams);
device_config.emplace(ov::num_streams(session_context_.num_streams));
}
}

Expand All @@ -293,13 +293,13 @@ void BasicBackend::SetOVDeviceConfiguration(ov::AnyMap& device_config) {
PopulateConfigValue(device_config);

// Enable caching
EnableCaching();
EnableCaching(device_config);

// Setting OpenCL queue throttling for GPU
EnableGPUThrottling(device_config);

// Enable streams; default=1 unless overridden by user configuration
EnableStreams();
EnableStreams(device_config);

// Set the inference_num_threads property of the CPU
SetNumThreads(device_config);
Expand Down
4 changes: 2 additions & 2 deletions onnxruntime/core/providers/openvino/backends/basic_backend.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -142,9 +142,9 @@ class BasicBackend : public IBackend {
private:
bool ValidateSubgraph(std::map<std::string, std::shared_ptr<ov::Node>>& const_outputs_map);
void PopulateConfigValue(ov::AnyMap& device_config);
void EnableCaching();
void EnableCaching(ov::AnyMap& device_config);
void EnableGPUThrottling(ov::AnyMap& device_config);
void EnableStreams();
void EnableStreams(ov::AnyMap& device_config);
void SetNumThreads(ov::AnyMap& device_config);
void SetOVDeviceConfiguration(ov::AnyMap& device_config);
void ValidateOrtDimsAgainstPartialShape(const std::vector<int64_t>& ort_dims,
Expand Down
8 changes: 0 additions & 8 deletions onnxruntime/core/providers/openvino/ov_interface.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -270,10 +270,6 @@ OVExeNetwork OVCore::ImportEPCtxOVIREncapsulation(std::istream& model_stream,
"Exception while Loading Network from OVIR model file: {}", model_file_path.string());
}

void OVCore::SetCache(const std::string& cache_dir_path) {
core.set_property(ov::cache_dir(cache_dir_path));
}

std::vector<std::string> OVCore::GetAvailableDevices() const {
std::vector<std::string> available_devices = core.get_available_devices();
return available_devices;
Expand Down Expand Up @@ -312,10 +308,6 @@ std::vector<std::string> OVCore::GetAvailableDevices(const std::string& device_t
return available_devices;
}

void OVCore::SetStreams(const std::string& device_type, int num_streams) {
core.set_property(device_type, {ov::num_streams(num_streams)});
}

std::shared_ptr<OVInferRequest> OVExeNetwork::CreateInferRequest() {
return OvExceptionBoundary([&]() {
auto infReq = compiled_model_obj.create_infer_request();
Expand Down
2 changes: 0 additions & 2 deletions onnxruntime/core/providers/openvino/ov_interface.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -95,8 +95,6 @@ struct OVCore : WeakSingleton<OVCore> {

std::vector<std::string> GetAvailableDevices() const;
std::vector<std::string> GetAvailableDevices(const std::string& device_type) const;
void SetCache(const std::string& cache_dir_path);
void SetStreams(const std::string& device_type, int num_streams);
};

class OVExeNetwork {
Expand Down
Loading