Added doxygen comments

include/MLModelRunner/MLModelRunner.h

@@ -4,6 +4,12 @@
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
+// Base MLModelRunner class which exposes APIs to set the features
+// to be send to the ML model and get the result back from the model.
+//
+// How to support a new ModelRunner:
+// 1. Create a new class which inherits from MLModelRunner.
+// 2. Implement the evaluateUntyped() method.
 //===----------------------------------------------------------------------===//
 //
 

include/MLModelRunner/ONNXModelRunner/ONNXModelRunner.h

@@ -4,6 +4,15 @@
 // See the LICENSE.txt file under ml-llvm-tools directory for license
 // information.
 //
+// ONNXModelRunner class supporting communication via ONNX Runtime
+//
+// How to use?
+// 1. Create agent objects with the path to the ONNX model
+// 2. Create an environment object inheriting from MLBridge::Environment
+// 3. Create an ONNXModelRunner object with the environment and the agents
+// 4. Populate the features to be sent to the model
+// 5. Call evaluate() to get the result back from the model
+//
 //===----------------------------------------------------------------------===//
 
 #ifndef ONNX_MODELRUNNER_H

include/MLModelRunner/ONNXModelRunner/agent.h

@@ -4,6 +4,12 @@
 // See the LICENSE.txt file under ml-llvm-tools directory for license
 // information.
 //
+// Agent class to support ML model inference
+//
+// How to use?
+// 1. Construct an agent object with the path to the ONNX model
+// 2. Call computeAction() to get the action from the model
+//
 //===----------------------------------------------------------------------===//
 
 #ifndef ONNX_MODELRUNNER_AGENT_H

include/MLModelRunner/ONNXModelRunner/environment.h

@@ -4,6 +4,12 @@
 // See the LICENSE.txt file under ml-llvm-tools directory for license
 // information.
 //
+// Base Environment class to support ML model inference
+//
+// How to use?
+// 1. Create an environment class inheriting from MLBridge::Environment
+// 2. Implement the step() and reset() methods
+// 3. Create an ONNXModelRunner object with the environment and the agents
 //===----------------------------------------------------------------------===//
 
 #ifndef ONNX_MODELRUNNER_ENVIRONMENT_H

include/MLModelRunner/PipeModelRunner.h

@@ -1,11 +1,18 @@
-//===- PipeModelRunner.h ---- "gym" ML model runner  -----*- C++ -*-===//
+//===- PipeModelRunner.h ---- PipeModelRunner  -----*- C++ -*-===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
-//===----------------------------------------------------------------------===//
+// PipeModelRunner class supporting communication via OS pipes
+// 
+// How to use?
+// 1. Create a PipeModelRunner object with the names of the pipes, and the
+// serialization technique
+// 2. Populate the features to be sent to the model
+// 3. Call evaluate() to get the result back from the model
 //
+//===----------------------------------------------------------------------===//
 
 #ifndef PipeModelRunner_H
 #define PipeModelRunner_H
@@ -23,9 +30,7 @@ namespace MLBridge {
 /// A MLModelRunner that asks for advice from an external agent, or host. It
 /// uses 2 files - ideally named pipes - one to send data to that agent, and
 /// one to receive advice.
-/// The data exchange uses the training logger (Utils/TrainingLogger.h) format.
-/// Specifically, the compiler will send the log header, set the context, and
-/// send observations; the host is expected to reply with a tensor value after
+/// The compiler will send observations; the host is expected to reply with a tensor value after
 /// each observation as a binary buffer that's conforming to the shape of the
 /// advice. Interleaved, the data closely resembles the training log for a
 /// log where we don't capture the reward signal.

include/MLModelRunner/gRPCModelRunner.h

@@ -4,7 +4,16 @@
 // See the LICENSE.txt file under ml-llvm-tools directory for license
 // information.
 //
-//===----------------------------------------------------------------------===//
+// gRPCModelRunner class supporting communication via gRPC
+//
+// How to use?
+// 1. Create a .proto file with the service and message definitions
+// 2. Generate the stubs using protoc
+// 3. Create a gRPCModelRunner object with the server address, stub, request and
+// response
+// 4. Populate the features to be sent to the model
+// 5. Call evaluate() to get the result back from the model
+// ===----------------------------------------------------------------------===//
 
 #ifndef GRPC_MODELRUNNER_H
 #define GRPC_MODELRUNNER_H

include/SerDes/baseSerDes.h

@@ -3,6 +3,13 @@
 //
 // Part of the MLCompilerBridge Project
 //
+// Base class for serialization and deserialization
+//
+// How to support a new serializtion technique:
+// 1. Create a new class which inherits from BaseSerDes.
+// 2. Implement the setFeature(), getSerializedData()
+// cleanDataStructures() and deserializeUntyped()
+// methods.
 //===------------------===//
 
 #ifndef BASE_SERDES_H

include/SerDes/bitstreamSerDes.h

@@ -3,6 +3,9 @@
 //
 // Part of the MLCompilerBridge Project
 //
+// Bitstream Serialization/Deserialization which sends header information
+// followed by the raw data.
+//
 //===------------------===//
 
 #ifndef BITSTREAM_SERIALIZER_H

include/SerDes/jsonSerDes.h

@@ -2,6 +2,7 @@
 //
 // Part of the MLCompilerBridge Project
 //
+// Json Serialization/Deserialization using LLVM's json library
 //===------------------===//
 
 #ifndef JSON_SERIALIZER_H

include/SerDes/protobufSerDes.h

@@ -2,6 +2,7 @@
 //
 // Part of the MLCompilerBridge Project
 //
+// Protobuf Serialization/Deserialization to support gRPC communication
 //===------------------===//
 
 #ifndef PROTOBUF_SERIALIZER_H