integrate with runner

JacobSzwejbka · facebook-github-bot · commit 1ea084c2321a · 2025-07-23T17:07:11.000-07:00
Summary: Add to the runner code. The runner doesn't distinguish between decode and prefill today, but we can pipe that info down in a later diff

Differential Revision: D78499456
diff --git a/examples/models/llava/runner/llava_runner.cpp b/examples/models/llava/runner/llava_runner.cpp
@@ -49,7 +49,7 @@ Error LlavaRunner::load() {
   // Load the text decoder runner
   text_decoder_runner_ =
       // @lint-ignore CLANGTIDY facebook-hte-Deprecated
-      std::make_unique<LlavaTextDecoderRunner>(module_.get());
+      std::make_unique<LlavaTextDecoderRunner>(module_.get(), io_manager_.get());
   // @lint-ignore CLANGTIDY facebook-hte-Deprecated
   text_decoder_runner_->load();
 
diff --git a/examples/models/llava/runner/llava_text_decoder_runner.h b/examples/models/llava/runner/llava_text_decoder_runner.h
@@ -18,8 +18,8 @@ namespace example {
 class ET_EXPERIMENTAL LlavaTextDecoderRunner
     : public executorch::extension::llm::TextDecoderRunner {
  public:
-  explicit LlavaTextDecoderRunner(executorch::extension::Module* module)
-      : TextDecoderRunner(module) {}
+  explicit LlavaTextDecoderRunner(executorch::extension::Module* module, executorch::extension::llm::IOManager* io_manager)
+      : TextDecoderRunner(module, io_manager) {}
 
   inline executorch::runtime::Result<executorch::aten::Tensor> step(
       executorch::extension::TensorPtr& tokens,
diff --git a/extension/llm/runner/multimodal_runner.h b/extension/llm/runner/multimodal_runner.h
@@ -25,6 +25,7 @@
 #include <executorch/extension/llm/runner/text_decoder_runner.h>
 #include <executorch/extension/llm/runner/text_prefiller.h>
 #include <executorch/extension/llm/runner/text_token_generator.h>
+#include <executorch/extension/llm/runner/io_manager/io_manager.h>
 #include <executorch/extension/llm/sampler/sampler.h>
 #include <executorch/extension/module/module.h>
 #include <pytorch/tokenizers/tokenizer.h>
@@ -41,6 +42,7 @@ class ET_EXPERIMENTAL MultimodalRunner {
       const float temperature = 0.8f)
       : temperature_(temperature),
         module_(std::make_unique<Module>(model_path, Module::LoadMode::File)),
+        io_manager_(std::make_unique<IOManager>()),
         tokenizer_path_(tokenizer_path) {
     ET_LOG(
         Info,
@@ -127,6 +129,7 @@ class ET_EXPERIMENTAL MultimodalRunner {
   std::unique_ptr<TextDecoderRunner> text_decoder_runner_;
   std::unique_ptr<TextPrefiller> text_prefiller_;
   std::unique_ptr<ImagePrefiller> image_prefiller_;
+  std::unique_ptr<IOManager> io_manager_;
   std::unique_ptr<TextTokenGenerator> text_token_generator_;
   std::string tokenizer_path_;
   std::unique_ptr<::tokenizers::Tokenizer> tokenizer_;
diff --git a/extension/llm/runner/targets.bzl b/extension/llm/runner/targets.bzl
@@ -36,6 +36,7 @@ def define_common_targets():
                 ":stats",
                 "//executorch/kernels/portable/cpu/util:arange_util" + aten_suffix,
                 "//executorch/extension/llm/sampler:sampler" + aten_suffix,
+                "//executorch/extension/llm/runner/io_manager:io_manager" + aten_suffix,
                 "//executorch/extension/module:module" + aten_suffix,
                 "//executorch/extension/tensor:tensor" + aten_suffix,
             ],
@@ -102,6 +103,7 @@ def define_common_targets():
                 ":text_decoder_runner" + aten_suffix,
                 ":text_prefiller" + aten_suffix,
                 ":text_token_generator" + aten_suffix,
+                "//executorch/extension/llm/runner/io_manager:io_manager" + aten_suffix,
                 "//pytorch/tokenizers:hf_tokenizer",
                 "//pytorch/tokenizers:llama2c_tokenizer",
                 "//pytorch/tokenizers:sentencepiece",
diff --git a/extension/llm/runner/test/TARGETS b/extension/llm/runner/test/TARGETS
@@ -18,6 +18,7 @@ runtime.cxx_test(
     srcs = ["test_text_decoder_runner.cpp"],
     deps = [
         "//executorch/extension/llm/runner:runner_lib",
+        "//executorch/extension/llm/runner/io_manager:io_manager",
         "//executorch/kernels/portable:generated_lib",
         "//executorch/runtime/core/exec_aten/testing_util:tensor_util",
     ],
diff --git a/extension/llm/runner/test/test_text_decoder_runner.cpp b/extension/llm/runner/test/test_text_decoder_runner.cpp
@@ -8,6 +8,7 @@
  */
 
 #include <executorch/extension/llm/runner/text_decoder_runner.h>
+#include <executorch/extension/llm/runner/io_manager/io_manager.h>
 #include <executorch/extension/module/module.h>
 #include <executorch/extension/tensor/tensor.h>
 #include <executorch/runtime/core/exec_aten/testing_util/tensor_factory.h>
@@ -19,6 +20,7 @@ using namespace ::testing;
 using executorch::extension::Module;
 using executorch::extension::TensorPtr;
 using executorch::extension::llm::TextDecoderRunner;
+using executorch::extension::llm::IOManager;
 using executorch::runtime::Error;
 using executorch::runtime::EValue;
 using executorch::runtime::Result;
@@ -34,11 +36,13 @@ class TextDecoderRunnerTest : public Test {
  protected:
   void SetUp() override {
     mock_module_ = std::make_unique<MockModule>();
-    runner_ = std::make_unique<TextDecoderRunner>(mock_module_.get());
+    io_manager_ = std::make_unique<executorch::extension::llm::IOManager>();
+    runner_ = std::make_unique<TextDecoderRunner>(mock_module_.get(), io_manager_.get());
   }
 
   std::unique_ptr<MockModule> mock_module_;
   std::unique_ptr<TextDecoderRunner> runner_;
+  std::unique_ptr<IOManager> io_manager_;
 };
 
 // Test logits_to_token() method with Float tensor
@@ -150,15 +154,16 @@ TEST_F(TextDecoderRunnerTest, StepWithAllModels) {
 
     // Load the model
     auto module = std::make_unique<Module>(model_path);
+    
     auto load_result = module->load();
     if (load_result != Error::Ok) {
       ADD_FAILURE() << "Failed to load model " << model_name << " from "
                     << model_path << " with error: " << (int)load_result;
       continue;
     }
-
+    std::unique_ptr<executorch::extension::llm::IOManager> io_manager = std::make_unique<executorch::extension::llm::IOManager>();
     // Create TextDecoderRunner
-    TextDecoderRunner runner(module.get());
+    TextDecoderRunner runner(module.get(), io_manager.get());
     auto runner_load_result = runner.load();
     ASSERT_EQ(runner_load_result, Error::Ok)
         << "Failed to load runner for " << model_name;
diff --git a/extension/llm/runner/test/test_text_llm_runner.cpp b/extension/llm/runner/test/test_text_llm_runner.cpp
@@ -8,6 +8,7 @@
  */
 
 #include <executorch/extension/llm/runner/irunner.h>
+#include <executorch/extension/llm/runner/io_manager/io_manager.h>
 #include <executorch/extension/llm/runner/text_llm_runner.h>
 #include <executorch/extension/llm/runner/text_prefiller.h>
 #include <executorch/extension/llm/runner/text_token_generator.h>
@@ -63,7 +64,7 @@ class MockModule : public ::executorch::extension::Module {
 
 class MockTextDecoderRunner : public TextDecoderRunner {
  public:
-  MockTextDecoderRunner() : TextDecoderRunner(nullptr) {}
+  MockTextDecoderRunner() : TextDecoderRunner(nullptr, nullptr) {}
   MOCK_METHOD(
       Result<executorch::aten::Tensor>,
       step,
@@ -219,6 +220,7 @@ TEST_F(RunnerTest, GenerateCallsCallbackExactlyMaxNewTokensTimes) {
       std::move(text_decoder_runner),
       std::unique_ptr<::executorch::extension::llm::TextPrefiller>(
           text_prefiller.release()),
+      std::make_unique<executorch::extension::llm::IOManager>(),
       std::move(text_token_generator),
       std::move(stats));
 
@@ -278,6 +280,7 @@ TEST_F(RunnerTest, WarmupCallsGenerateWithWarmingFlag) {
       std::move(text_decoder_runner),
       std::unique_ptr<::executorch::extension::llm::TextPrefiller>(
           text_prefiller.release()),
+      std::make_unique<executorch::extension::llm::IOManager>(),
       std::move(text_token_generator),
       std::move(stats));
 
@@ -312,6 +315,7 @@ TEST_F(RunnerTest, IsLoadedReturnsTrueWhenComponentsInitialized) {
       std::move(text_decoder_runner),
       std::unique_ptr<::executorch::extension::llm::TextPrefiller>(
           text_prefiller.release()),
+      std::make_unique<executorch::extension::llm::IOManager>(),
       std::move(text_token_generator),
       std::move(stats));
 
@@ -356,6 +360,7 @@ TEST_F(RunnerTest, GenerateFromPosErrorsWithNegativeMaxNewTokens) {
       std::move(text_decoder_runner),
       std::unique_ptr<::executorch::extension::llm::TextPrefiller>(
           text_prefiller.release()),
+      std::make_unique<executorch::extension::llm::IOManager>(),
       std::move(text_token_generator),
       std::move(stats));
 
diff --git a/extension/llm/runner/test/test_text_prefiller.cpp b/extension/llm/runner/test/test_text_prefiller.cpp
@@ -24,7 +24,7 @@ using executorch::runtime::testing::TensorFactory;
 // Mock class for TextDecoderRunner
 class MockTextDecoderRunner : public TextDecoderRunner {
  public:
-  MockTextDecoderRunner() : TextDecoderRunner(nullptr) {}
+  MockTextDecoderRunner() : TextDecoderRunner(nullptr, nullptr) {}
   MOCK_METHOD(
       Result<executorch::aten::Tensor>,
       step,
diff --git a/extension/llm/runner/text_decoder_runner.cpp b/extension/llm/runner/text_decoder_runner.cpp
@@ -22,7 +22,7 @@ namespace llm {
 // NOTE: we observed ~2x loading performance increase on iPhone 15
 // and a ~5% improvement on Galaxy S22 by switching to
 // FileDataLoader instead of MmapDataLoader + UseMlockIgnoreErrors.
-TextDecoderRunner::TextDecoderRunner(Module* module) : module_(module) {}
+TextDecoderRunner::TextDecoderRunner(Module* module, IOManager* io_manager) : module_(module), io_manager_(io_manager) {}
 
 // This function is functional, meaning it shouldn't modify any state of the
 // input. It should be safe to call multiple times with the same inputs. The
@@ -71,8 +71,22 @@ ::executorch::runtime::Result<executorch::aten::Tensor> TextDecoderRunner::step(
         "Start pos tensor numel: %zu, tokens numel: %zu",
         start_pos_tensor->numel(),
         tokens->numel());
-    auto outputs_res = module_->forward({tokens, start_pos_tensor});
+
+
+    std::vector<runtime::EValue> inputs;
+    auto method_err = module_->method("forward");
+    ET_CHECK_OK_OR_RETURN_ERROR(method_err.error());
+    auto& method = *(method_err.get());
+
+    auto inputs_res = io_manager_->prepare_decode(tokens, start_pos_tensor, method);
+    ET_CHECK_OK_OR_RETURN_ERROR(inputs_res.error());
+    inputs = inputs_res.get();
+    auto outputs_res = module_->forward(inputs);
     ET_CHECK_OK_OR_RETURN_ERROR(outputs_res.error());
+
+    auto update_err = io_manager_->update_decode(method, outputs_res.get());
+    ET_CHECK_OK_OR_RETURN_ERROR(update_err);
+    
     ET_CHECK_MSG(
         outputs_res.get().size() == 1,
         "More then one output returned from executing LLM.");
diff --git a/extension/llm/runner/text_decoder_runner.h b/extension/llm/runner/text_decoder_runner.h
@@ -10,6 +10,7 @@
 
 #pragma once
 
+#include <executorch/extension/llm/runner/io_manager/io_manager.h>
 #include <executorch/extension/llm/sampler/sampler.h>
 #include <executorch/extension/module/module.h>
 #include <executorch/extension/tensor/tensor.h>
@@ -21,7 +22,7 @@ namespace llm {
 
 class ET_EXPERIMENTAL TextDecoderRunner {
  public:
-  explicit TextDecoderRunner(Module* module);
+  explicit TextDecoderRunner(Module* module, IOManager* io_manager);
 
   virtual ~TextDecoderRunner() = default;
 
@@ -94,13 +95,14 @@ class ET_EXPERIMENTAL TextDecoderRunner {
 
  protected:
   /**
-   * Note: TextDecoderRunner does not own the Module instance. It is expected
-   * that the outer class (likely Runner) manages the lifecycle of the Module.
+   * Note: TextDecoderRunner does not own the Module or IOManager instance. It is
+   * expected that the outer class (likely Runner) manages the lifecycle of them.
    * This means that the responsibility for creating, maintaining, and
    * destroying the Module lies outside of TextDecoderRunner. Ensure that the
    * Module remains valid for the duration of TextDecoderRunner's usage.
    */
   Module* module_;
+  IOManager* io_manager_;
   bool should_stop_{false};
 };
 
diff --git a/extension/llm/runner/text_llm_runner.cpp b/extension/llm/runner/text_llm_runner.cpp
@@ -12,6 +12,7 @@
 
 #include <executorch/extension/llm/runner/text_llm_runner.h>
 #include <executorch/extension/llm/runner/util.h>
+#include <executorch/extension/llm/runner/io_manager/io_manager.h>
 #include <executorch/runtime/platform/runtime.h>
 #include <pytorch/tokenizers/hf_tokenizer.h>
 #include <pytorch/tokenizers/llama2c_tokenizer.h>
@@ -39,6 +40,7 @@ TextLLMRunner::TextLLMRunner(
     std::unique_ptr<::executorch::extension::Module> module,
     std::unique_ptr<TextDecoderRunner> text_decoder_runner,
     std::unique_ptr<TextPrefiller> text_prefiller,
+    std::unique_ptr<IOManager> io_manager,
     std::unique_ptr<TextTokenGenerator> text_token_generator,
     std::unique_ptr<Stats> stats,
     float temperature)
@@ -47,9 +49,11 @@ TextLLMRunner::TextLLMRunner(
       module_(std::move(module)),
       text_decoder_runner_(std::move(text_decoder_runner)),
       text_prefiller_(std::move(text_prefiller)),
+      io_manager_(std::move(io_manager)),
       text_token_generator_(std::move(text_token_generator)),
       stats_(std::move(stats)),
       temperature_(temperature) {
+  
   // Note: This constructor assumes that text_prefiller and text_token_generator
   // already have references to the Module and TextDecoderRunner they need
 }
@@ -63,6 +67,14 @@ Error TextLLMRunner::load() {
     return Error::Ok;
   }
   ET_CHECK_OK_OR_RETURN_ERROR(text_prefiller_->load());
+  ET_CHECK_OK_OR_RETURN_ERROR(module_->load_method("forward"));
+  auto method_res = module_->method("forward");
+
+  Program& program = *module_->program();
+
+  ET_CHECK_OK_OR_RETURN_ERROR(method_res.error());
+  auto& forward = *(method_res.get());
+  ET_CHECK_OK_OR_RETURN_ERROR(io_manager_->load(program, forward, forward));
   ET_CHECK_OK_OR_RETURN_ERROR(text_token_generator_->load());
   return Error::Ok;
 }
@@ -393,9 +405,12 @@ std::unique_ptr<TextLLMRunner> create_text_llm_runner(
   auto eos_ids = std::make_unique<std::unordered_set<uint64_t>>(
       llm::get_eos_ids(tokenizer.get(), module.get()));
 
+  // Create IOManager
+  std::unique_ptr<IOManager> io_manager = std::make_unique<IOManager>();
+
   // Create text_decoder_runner. Use a shared_ptr so that it can be shared with
   // TextPrefiller and TextTokenGenerator
-  auto text_decoder_runner = std::make_unique<TextDecoderRunner>(module.get());
+  auto text_decoder_runner = std::make_unique<TextDecoderRunner>(module.get(), io_manager.get());
 
   // Create text_prefiller
   auto text_prefiller = std::make_unique<TextPrefiller>(
@@ -420,6 +435,7 @@ std::unique_ptr<TextLLMRunner> create_text_llm_runner(
       std::move(module),
       std::move(text_decoder_runner),
       std::move(text_prefiller),
+      std::move(io_manager),
       std::move(text_token_generator),
       std::move(stats),
       temperature);
diff --git a/extension/llm/runner/text_llm_runner.h b/extension/llm/runner/text_llm_runner.h
@@ -55,6 +55,7 @@ class ET_EXPERIMENTAL TextLLMRunner : public IRunner {
       std::unique_ptr<::executorch::extension::Module> module,
       std::unique_ptr<TextDecoderRunner> text_decoder_runner,
       std::unique_ptr<TextPrefiller> text_prefiller,
+      std::unique_ptr<IOManager> io_manager,
       std::unique_ptr<TextTokenGenerator> text_token_generator,
       std::unique_ptr<Stats> stats,
       float temperature = -1.0f);
@@ -155,6 +156,7 @@ class ET_EXPERIMENTAL TextLLMRunner : public IRunner {
                             // sure it outlives text_prefiller_ &
                             // text_token_generator_.
   std::unique_ptr<TextPrefiller> text_prefiller_;
+  std::unique_ptr<IOManager> io_manager_;
   std::unique_ptr<TextTokenGenerator> text_token_generator_;
 
   // Stats
diff --git a/extension/module/module.cpp b/extension/module/module.cpp
@@ -230,6 +230,7 @@ runtime::Error Module::load_method(
 
 ET_NODISCARD runtime::Result<Method*> Module::method(
     const std::string& method_name) {
+  ET_CHECK_OK_OR_RETURN_ERROR(load_method(method_name));
   ET_CHECK_OR_RETURN_ERROR(
       methods_.count(method_name) > 0,
       InvalidArgument,
