add wb

ml-agents/mlagents/trainers/learn.py

 )
 from mlagents_envs import logging_util
 
+import wandb
+
 logger = logging_util.get_logger(__name__)
 
 TRAINING_STATUS_FILE_NAME = "training_status.json"
     :param run_seed: Random seed used for training.
     :param run_options: Command line arguments for training.
     """
+    #wandb.init(project="mlagent-cloud-profiling", sync_tensorboard=True)
     with hierarchical_timer("run_training.setup"):
         checkpoint_settings = options.checkpoint_settings
         env_settings = options.env_settings

ml-agents/mlagents/trainers/ppo/optimizer_torch.py

 from mlagents.trainers.optimizer.torch_optimizer import TorchOptimizer
 from mlagents.trainers.settings import TrainerSettings, PPOSettings
 from mlagents.trainers.torch.utils import ModelUtils
+import wandb
 
 
 class TorchPPOOptimizer(TorchOptimizer):
             + 0.5 * value_loss
             - decay_bet * ModelUtils.masked_mean(entropy, loss_masks)
         )
+        wandb.log({"loss": loss})
 
         # Set optimizer learning rate
         ModelUtils.update_learning_rate(self.optimizer, decay_lr)

ml-agents/mlagents/trainers/stats.py

 from mlagents.tf_utils import tf, generate_session_config
 from mlagents.tf_utils.globals import get_rank
 
+import wandb
+
 
 logger = get_logger(__name__)
 
         self.summary_writers: Dict[str, tf.summary.FileWriter] = {}
         self.base_dir: str = base_dir
         self._clear_past_data = clear_past_data
+
+        with open('wandb_API', 'r') as f:
+            api_key = f.readline().strip()
+        os.environ['WANDB_API_KEY'] = api_key
+        wandb.init(project="mlagent-cloud-profiling")
+        wandb.tensorboard.patch(pytorch=True, tensorboardX=True)
 
     def write_stats(
         self, category: str, values: Dict[str, StatsSummary], step: int