feat(supportedops): Application to dump a list of supported operators

Signed-off-by: Naren Dasan <naren@narendasan.com>
Signed-off-by: Naren Dasan <narens@nvidia.com>

Author: narendasan

core/conversion/converters/NodeConverterRegistry.cpp

@@ -47,6 +47,7 @@ class NodeConverterRegistry {
  public:
   bool RegisterConverter(torch::jit::FunctionSchema* signature, OpConverter& converter) {
     LOG_DEBUG("Registering converter for " << canonical_schema_string(*signature));
+    registered_converter_schemas_.insert(c10::toString(*signature));
     auto name = signature->operator_name();
     auto iter = converter_lut_.find(name);
     if (iter != converter_lut_.end()) {
@@ -83,8 +84,15 @@ class NodeConverterRegistry {
     }
   }
 
+  std::vector<std::string> GetRegisteredConverterList() {
+    std::vector<std::string> converter_list;
+    std::copy(registered_converter_schemas_.begin(), registered_converter_schemas_.end(), std::back_inserter(converter_list));
+    return converter_list;
+  }
+
  private:
   ConverterLUT converter_lut_;
+  std::set<std::string> registered_converter_schemas_;
 };
 
 NodeConverterRegistry& get_converter_registry() {
@@ -115,6 +123,10 @@ bool node_is_convertable(const torch::jit::Node* n) {
   return get_converter_registry().Convertable(n);
 }
 
+std::vector<std::string> get_converter_list() {
+  return get_converter_registry().GetRegisteredConverterList();
+}
+
 RegisterNodeConversionPatterns&& RegisterNodeConversionPatterns::pattern(ConversionPattern p) && {
   register_node_converter(std::move(p));
   return std::move(*this);

core/conversion/converters/converters.h

@@ -39,6 +39,7 @@ class RegisterNodeConversionPatterns {
 
 bool node_is_convertable(const torch::jit::Node* n);
 OpConverter get_node_converter_for(const torch::jit::FunctionSchema* signature);
+std::vector<std::string> get_converter_list();
 
 } // namespace converters
 } // namespace conversion

core/conversion/evaluators/NodeEvaluatorRegistry.cpp

@@ -36,6 +36,9 @@ class NodeEvaluatorRegistry {
           "Attempting to override already registered evaluator " << node_kind.toQualString()
                                                                  << ", merge implementations instead");
     }
+    for (auto const& e : eval_reg.options.supported_variants) {
+      registered_evaluator_schemas_.insert(e);
+    }
     evaluator_lut_[node_kind] = std::move(eval_reg);
   }
 
@@ -76,6 +79,12 @@ class NodeEvaluatorRegistry {
     return evaluator;
   }
 
+  std::vector<std::string> GetRegisteredEvaluatorList() {
+    std::vector<std::string> evaluator_list;
+    std::copy(registered_evaluator_schemas_.begin(), registered_evaluator_schemas_.end(), std::back_inserter(evaluator_list));
+    return evaluator_list;
+  }
+
   bool EvalAtConversionTime(const torch::jit::Node* n) {
     auto evaluator = FindEvaluator(n);
     if (evaluator == nullptr) {
@@ -87,6 +96,7 @@ class NodeEvaluatorRegistry {
 
  private:
   EvaluatorLUT evaluator_lut_;
+  std::set<std::string> registered_evaluator_schemas_;
 };
 
 NodeEvaluatorRegistry& get_evaluator_registry() {
@@ -99,6 +109,10 @@ bool shouldEvalAtConversionTime(const torch::jit::Node* n) {
   return get_evaluator_registry().EvalAtConversionTime(n);
 }
 
+std::vector<std::string> getEvaluatorList() {
+  return get_evaluator_registry().GetRegisteredEvaluatorList();
+}
+
 c10::optional<torch::jit::IValue> EvalNode(const torch::jit::Node* n, kwargs& args) {
   auto evaluator = get_evaluator_registry().GetEvaluator(n);
   return evaluator(n, args);

core/conversion/evaluators/evaluators.h

@@ -38,13 +38,15 @@ typedef std::function<c10::optional<torch::jit::IValue>(const torch::jit::Node*,
 struct EvalOptions {
   std::set<c10::TypePtr> blacklisted_output_types;
   std::vector<c10::OperatorName> valid_schemas;
+  std::vector<std::string> supported_variants;
   EvalOptions() = default;
   EvalOptions& blacklistOutputTypes(std::set<c10::TypePtr> types) {
     use_options = true;
     blacklisted_output_types = types;
     return *this;
   }
   EvalOptions& validSchemas(std::set<std::string> schemas) {
+    std::copy(schemas.begin(), schemas.end(), std::back_inserter(supported_variants));
     use_options = true;
     for (auto s : schemas) {
       valid_schemas.push_back(torch::jit::parseSchema(s).operator_name());
@@ -72,6 +74,7 @@ struct EvalRegistration {
 
 c10::optional<torch::jit::IValue> EvalNode(const torch::jit::Node* n, kwargs& args);
 bool shouldEvalAtConversionTime(const torch::jit::Node* n);
+std::vector<std::string> getEvaluatorList();
 void register_node_evaluator(torch::jit::NodeKind node_kind, NodeEvaluator evaluator);
 void register_node_evaluator(EvalRegistration r);
 

cpp/supportedops/BUILD

@@ -0,0 +1,12 @@
+package(default_visibility = ["//visibility:public"])
+
+cc_binary(
+    name = "supportedops",
+    srcs = [
+        "main.cpp"
+    ],
+    deps = [
+        "//cpp/api:trtorch",
+        "//core/conversion/converters"
+    ],
+)

cpp/supportedops/main.cpp

@@ -0,0 +1,48 @@
+#include "core/conversion/converters/converters.h"
+#include "core/conversion/evaluators/evaluators.h"
+
+#include <string>
+#include <sstream>
+#include <vector>
+#include <iostream>
+
+int main(int argc, const char* argv[]) {
+  std::vector<std::string> converters = trtorch::core::conversion::converters::get_converter_list();
+  std::vector<std::string> evaluators = trtorch::core::conversion::evaluators::getEvaluatorList();
+
+  std::stringstream ss;
+
+  ss << R"TITLE(
+=================================
+Operators Supported
+=================================
+
+)TITLE";
+
+  ss << R"SEC(
+Operators Currently Supported Through Converters
+-------------------------------------------------
+
+)SEC";
+
+  for (auto c : converters) {
+    ss << "- " << c << std::endl;
+  }
+
+  ss << R"SEC(
+Operators Currently Supported Through Evaluators
+-------------------------------------------------
+
+)SEC";
+
+  for (auto e : evaluators) {
+    ss << "- " << e << std::endl;
+  }
+
+  std::ofstream ofs;
+  ofs.open(argv[1]);
+
+  ofs << ss.rdbuf();
+
+  return 0;
+}