Add GetOutput for C API

src/ort_genai.h

@@ -232,6 +232,12 @@ struct OgaGenerator : OgaAbstract {
     return OgaGenerator_GetSequenceData(this, index);
   }
 
+  std::unique_ptr<OgaTensor> GetOutput(const char* name) {
+    OgaTensor* out;
+    OgaCheckResult(OgaGenerator_GetOutput(this, name, &out));
+    return std::unique_ptr<OgaTensor>(out);
+  }
+
 #if __cplusplus >= 202002L
   std::span<const int32_t> GetSequence(size_t index) const {
     return {GetSequenceData(index), GetSequenceCount(index)};

src/ort_genai_c.cpp

@@ -208,6 +208,48 @@ OgaResult* OGA_API_CALL OgaGenerator_GenerateNextToken(OgaGenerator* generator)
   OGA_CATCH
 }
 
+OgaResult* OGA_API_CALL OgaGenerator_GetOutput(const OgaGenerator* oga_generator, const char* name, OgaTensor** out) {
+  OGA_TRY
+  auto& generator = *reinterpret_cast<const Generators::Generator*>(oga_generator);
+  auto* ortvalue_output = generator.state_->GetOutput(name);
+  auto type_info = ortvalue_output->GetTensorTypeAndShapeInfo();
+  std::unique_ptr<OrtValue> ortvalue_clone = OrtValue::CreateTensor(generator.model_->allocator_cpu_,
+                                                                    type_info->GetShape(),
+                                                                    type_info->GetElementType());
+  // Copy data to ortvalue_clone
+  auto element_size = Generators::SizeOf(type_info->GetElementType());
+  auto data_size = type_info->GetElementCount() * element_size;
+#if USE_CUDA
+  if (ortvalue_output->GetTensorMemoryInfo().GetDeviceType() == OrtMemoryInfoDeviceType_GPU && generator.model_->device_type_ == Generators::DeviceType::CUDA) {
+    cudaMemcpy(ortvalue_clone->GetTensorMutableRawData(), ortvalue_output->GetTensorMutableRawData(), data_size, cudaMemcpyDeviceToHost);
+  }
+#elif USE_DML
+  if (ortvalue_output->GetTensorMemoryInfo().GetDeviceType() == OrtMemoryInfoDeviceType_GPU && generator.model_->device_type_ == Generators::DeviceType::DML) {
+    ComPtr<ID3D12Resource> gpu_resource;
+    Ort::ThrowOnError(generator.model_->GetOrtDmlApi()->GetD3D12ResourceFromAllocation(
+        generator.model_->allocator_device_,
+        ortvalue_output->GetTensorMutableRawData(),
+        &gpu_resource));
+    auto cpu_tensor = ortvalue_clone->GetTensorMutableRawData();
+    generator.model_->GetDmlReadbackHeap()->ReadbackFromGpu(
+        std::span(reinterpret_cast<uint8_t*>(cpu_tensor), data_size),
+        gpu_resource.Get(),
+        0,
+        D3D12_RESOURCE_STATE_UNORDERED_ACCESS);
+  }
+#else
+  std::copy(static_cast<uint8_t*>(ortvalue_output->GetTensorMutableRawData()),
+            static_cast<uint8_t*>(ortvalue_output->GetTensorMutableRawData()) + data_size,
+            static_cast<uint8_t*>(ortvalue_clone->GetTensorMutableRawData()));
+#endif
+
+  auto tensor = std::make_shared<Generators::Tensor>(std::move(ortvalue_clone));
+  tensor->external_owner_ = tensor;
+  *out = reinterpret_cast<OgaTensor*>(tensor.get());
+  return nullptr;
+  OGA_CATCH
+}
+
 size_t OGA_API_CALL OgaGenerator_GetSequenceCount(const OgaGenerator* oga_generator, size_t index) {
   auto& generator = *reinterpret_cast<const Generators::Generator*>(oga_generator);
   return generator.GetSequence(static_cast<int>(index)).GetCPU().size();

src/ort_genai_c.h

@@ -224,6 +224,8 @@ OGA_EXPORT bool OGA_API_CALL OgaGenerator_IsDone(const OgaGenerator* generator);
 OGA_EXPORT OgaResult* OGA_API_CALL OgaGenerator_ComputeLogits(OgaGenerator* generator);
 OGA_EXPORT OgaResult* OGA_API_CALL OgaGenerator_GenerateNextToken(OgaGenerator* generator);
 
+OGA_EXPORT OgaResult* OGA_API_CALL OgaGenerator_GetOutput(const OgaGenerator* oga_generator, const char* name, OgaTensor** out);
+
 /*
  * \brief Returns the number of tokens in the sequence at the given index.
  * \param[in] generator The generator to get the count of the tokens for the sequence at the given index.

test/c_api_tests.cpp

@@ -173,6 +173,61 @@
 }
 #endif
 
+TEST(CAPITests, GetOutputCAPI) {
+  std::vector<int64_t> input_ids_shape{2, 4};
+  std::vector<int32_t> input_ids{0, 0, 0, 52, 0, 0, 195, 731};
+
+  auto input_sequence_length = input_ids_shape[1];
+  auto batch_size = input_ids_shape[0];
+  int max_length = 10;
+
+  // To generate this file:
+  // python convert_generation.py --model_type gpt2 -m hf-internal-testing/tiny-random-gpt2 --output tiny_gpt2_greedysearch_fp16.onnx --use_gpu --max_length 20
+  // And copy the resulting gpt2_init_past_fp32.onnx file into these two files (as it's the same for gpt2)
+
+  auto model = OgaModel::Create(MODEL_PATH "hf-internal-testing/tiny-random-gpt2-fp32");
+
+  auto params = OgaGeneratorParams::Create(*model);
+  params->SetSearchOption("max_length", max_length);
+  params->SetInputIDs(input_ids.data(), input_ids.size(), input_sequence_length, batch_size);
+
+  auto generator = OgaGenerator::Create(*model, *params);
+  generator->ComputeLogits();
+  auto prompt_logits = static_cast<float*>(generator->GetOutput("logits")->Data());
+
+  // check prompt
+  // full logits has shape [2, 4, 1000]. Sample 1 for every 200 tokens and the expected sampled logits has shape [2, 4, 5]
+  std::vector<float> expected_sampled_logits_prompt{0.29694548,  0.00955007,  0.0430819,  0.10063869,  0.0437237,
+                                                    0.27329233,  0.00841076, -0.1060291,  0.11328877,  0.13369876,
+                                                    0.30323744,  0.0545997,   0.03894716, 0.11702324,  0.0410665,
+                                                    -0.12675379, -0.04443946,  0.14492269, 0.03021223, -0.03212897,
+                                                    0.29694548,  0.00955007,  0.0430819,  0.10063869,  0.0437237,
+                                                    0.27329233,  0.00841076, -0.1060291,  0.11328877,  0.13369876,
+                                                    -0.04699047,  0.17915794,  0.20838135, 0.10888482, -0.00277808,
+                                                    0.2938929,  -0.10538938, -0.00226692, 0.12050669, -0.10622668};
+
+  int num_prompt_outputs_to_check = 40;
+  int sample_size = 200;
+  float tolerance = 0.001;
+  // Verify outputs match expected outputs
+  for (int i = 0; i < num_prompt_outputs_to_check; i++) {
+    EXPECT_NEAR(expected_sampled_logits_prompt[i], prompt_logits[i*sample_size], tolerance);
+  }
+
+  generator->GenerateNextToken();
+  auto token_gen_logits = static_cast<float*>(generator->GetOutput("logits")->Data());
+
+  // check for the 1st token generation
+  // full logits has shape [2, 1, 1000]. Sample 1 for every 200 tokens and the expected sampled logits has shape [2, 1, 5]
+  std::vector<float> expected_sampled_logits_token_gen{0.03742531, -0.05752287,  0.14159015, 0.04210977, -0.1484456,
+                                                        0.3041716,  -0.08701379, -0.03778192, 0.07471392, -0.02049096};
+  int num_token_gen_outputs_to_check = 10;
+
+  for (int i = 0; i < num_token_gen_outputs_to_check; i++) {
+    EXPECT_NEAR(expected_sampled_logits_token_gen[i], token_gen_logits[i*sample_size], tolerance);
+  }
+}
+
 #if TEST_PHI2
 
 struct Phi2Test {