add config class

com.unity.ml-agents/Runtime/Policies/TrainingModelRunner.cs

             // barracudaModel = ModelLoader.Load(new NNModel());
             barracudaModel = ModelLoader.Load(model);
             m_Model = barracudaModel;
-            WorkerFactory.Type executionDevice = WorkerFactory.Type.CSharpBurst;
+            WorkerFactory.Type executionDevice = WorkerFactory.Type.CSharp;
             m_Engine = WorkerFactory.CreateWorker(executionDevice, barracudaModel, m_Verbose);
 
             m_InferenceInputs = barracudaModel.GetInputTensors();

com.unity.ml-agents/Runtime/Trainer.cs

 using System;
 using Unity.MLAgents.Actuators;
 using Unity.Barracuda;
-
+using UnityEngine;
+    internal class TrainerConfig
+    {
+        public int bufferSize = 1024;
+        public int batchSize = 64;
+        public float gamma = 0.99f;
+    }
+
-        int m_BufferSize = 1024;
-        int batchSize = 64;
-        float GAMMA;
+        TrainerConfig m_Config;
+        int m_TrainingStep;
-        public Trainer(string behaviorName, ActionSpec actionSpec, NNModel model, int seed=0)
+        public Trainer(string behaviorName, ActionSpec actionSpec, NNModel model, int seed=0, TrainerConfig config=null)
+            m_Config = config ?? new TrainerConfig();
-            m_Buffer = new ReplayBuffer(m_BufferSize);
+            m_Buffer = new ReplayBuffer(m_Config.bufferSize);
             m_ModelRunner = new TrainingModelRunner(actionSpec, model, seed);
             Academy.Instance.TrainerUpdate += Update;
         }
 
         public void Update()
         {
-            if (m_Buffer.Count < batchSize * 2)
+            if (m_Buffer.Count < m_Config.batchSize * 2)
-            var samples = m_Buffer.SampleBatch(batchSize);
+            var samples = m_Buffer.SampleBatch(m_Config.batchSize);
-            // q_values = policy_net(states).gather(1, action_batch)
+            // q_values = policy_net(states).gather(1, actions)
 
             // next_states = [s.next_state for s in samples]
             // rewards = [s.reward for s in samples]
             // loss = MSE(q_values, expected_q_values);
             // m_ModelRunner.model = Barracuda.ModelUpdate(m_ModelRunner.model, loss);
+
+            m_TrainingStep += 1;
         }
     }
 }