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

Revert "Support calling custom method names via MODULE_METHOD_NAME (f… #128

Merged
merged 1 commit into from
May 4, 2024
Merged

Revert "Support calling custom method names via MODULE_METHOD_NAME (f… #128

Show file tree
Hide file tree
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
14 changes: 0 additions & 14 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -217,20 +217,6 @@ key: "INTRA_OP_THREAD_COUNT"
}
```

* `MODULE_METHOD_NAME`:

String flag to specify which method on the PyTorch model is being called.
Default value is `forward`.

```
parameters: {
key: "MODULE_METHOD_NAME"
value: {
string_value:"custom_method"
}
}
```

* Additional Optimizations: Three additional boolean parameters are available to disable
certain Torch optimizations that can sometimes cause latency regressions in models with
complex execution modes and dynamic shapes. If not specified, all are enabled by default.
Expand Down
65 changes: 7 additions & 58 deletions src/libtorch.cc
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -61,8 +61,6 @@
// https://github.com/pytorch/pytorch/blob/v2.2.1-rc3/aten/src/ATen/Parallel.h#L133
#include <ATen/Parallel.h>

// Default forward method to call on PyTorch modules
const std::string DEFAULT_MODULE_METHOD_NAME = "forward";

//
// PyTorch C++ (LibTorch) Backend that implements the TRITONBACKEND API.
Expand Down Expand Up @@ -113,7 +111,6 @@ class ModelState : public BackendModel {
{
return model_outputs_;
}
const std::string& ModuleMethodName() { return module_method_name_; }

private:
ModelState(TRITONBACKEND_Model* triton_model);
Expand Down Expand Up @@ -156,10 +153,6 @@ class ModelState : public BackendModel {
// is specified both in the output section and state section, it indicates
// that the backend must return the output state to the client too.
std::map<std::string, std::pair<int64_t, int64_t>> model_outputs_;

// Method to call on PyTorch Module.
// Defaults to DEFAULT_MODULE_METHOD_NAME.
std::string module_method_name_;
};

TRITONSERVER_Error*
Expand Down Expand Up @@ -237,8 +230,7 @@ ModelState::ModelState(TRITONBACKEND_Model* triton_model)
enable_inference_mode_(true), enable_cache_cleaning_(false),
enable_weight_sharing_(false), enable_tensor_fuser_pair_({false, true}),
enable_jit_profiling_pair_({false, true}),
enable_jit_executor_pair_({false, true}),
module_method_name_(DEFAULT_MODULE_METHOD_NAME)
enable_jit_executor_pair_({false, true})
{
}

Expand Down Expand Up @@ -527,30 +519,6 @@ ModelState::ParseParameters()
.c_str());
}
}

// If 'MODULE_METHOD_NAME' is not present in 'parameters' then
// 'module_method_name_' is set to 'DEFAULT_MODULE_METHOD_NAME' ('forward').
std::string module_method_name = DEFAULT_MODULE_METHOD_NAME;
err = GetParameterValue(params, "MODULE_METHOD_NAME", &module_method_name);
if (err != nullptr) {
if (TRITONSERVER_ErrorCode(err) != TRITONSERVER_ERROR_NOT_FOUND) {
return err;
} else {
LOG_MESSAGE(
TRITONSERVER_LOG_INFO,
(std::string("module_method_name is not specified") +
" for model instance '" + Name() + "'")
.c_str());
TRITONSERVER_ErrorDelete(err);
}
} else {
module_method_name_ = module_method_name;
LOG_MESSAGE(
TRITONSERVER_LOG_INFO,
(std::string("module_method_name is ") + module_method_name_ +
" for model instance '" + Name() + "'")
.c_str());
}
}

return nullptr;
Expand Down Expand Up @@ -972,20 +940,7 @@ ModelInstanceState::ValidateInputs(const size_t expected_input_cnt)
// configuration specifies only those.
std::vector<std::string> allowed_inputs;

// First check if method exists in the model and throw an error if absent
const auto methodNameToExecute = model_state_->ModuleMethodName();
const auto optionalMethodHandle =
torch_model_->find_method(methodNameToExecute);
if (!optionalMethodHandle.has_value()) {
return TRITONSERVER_ErrorNew(
TRITONSERVER_ERROR_INVALID_ARG,
(std::string("unable to find method '") + methodNameToExecute +
"' in model '" + model_path_ + "'")
.c_str());
}

// Get the method schema and validate the inputs
const torch::jit::Method& method = optionalMethodHandle.value();
const torch::jit::Method& method = torch_model_->get_method("forward");
const auto& schema = method.function().getSchema();
const std::vector<c10::Argument>& arguments = schema.arguments();

Expand Down Expand Up @@ -1628,24 +1583,18 @@ ModelInstanceState::Execute(
torch::NoGradGuard no_grad;

// If input is a dictionary, prepare dictionary from 'input_tensors'.
std::string module_method_name = model_state_->ModuleMethodName();
std::vector<c10::IValue> inputs;
if (is_dict_input_) {
c10::Dict<std::string, at::Tensor> dict;
torch::Dict<std::string, torch::Tensor> input_dict;
for (auto& input_index : input_index_map_) {
torch::jit::IValue ival = (*input_tensors)[input_index.second];
dict.insert(input_index.first, ival.toTensor());
input_dict.insert(input_index.first, ival.toTensor());
}
inputs.push_back(dict);
std::vector<torch::jit::IValue> input_dict_ivalue = {input_dict};
model_outputs_ = torch_model_->forward(input_dict_ivalue);
} else {
for (auto& input_tensor : *input_tensors) {
inputs.push_back(input_tensor.toTensor());
}
model_outputs_ = torch_model_->forward(*input_tensors);
}

// Actually run the method on the model.
model_outputs_ = torch_model_->get_method(module_method_name)(inputs);

if (model_outputs_.isTuple()) {
auto model_outputs_tuple = model_outputs_.toTuple();
size_t op_index = 0;
Expand Down
Loading