feat: Wandb integration

#112
Obs: Unsolved Lint formatting issues.

experiments/solves/solve_collectables_sb3.py

@@ -7,6 +7,8 @@
 from gymnasium import spaces
 from pysc2.env import sc2_env
 from stable_baselines3 import PPO
+import wandb
+from wandb.integration.sb3 import WandbCallback
 
 from urnai.environments.stablebaselines3.custom_env import CustomEnv
 from urnai.sc2.actions.collectables import CollectablesActionSpace
@@ -36,6 +38,16 @@ def declare_trainer():
     models_dir = "saves/models/PPO"
     logdir = "saves/logs"
 
+    conf_dict = {
+             "policy":"MlpPolicy",
+             "model_save_name": "PPO"}
+
+    run = wandb.init(
+        project='solve_collectables',
+        config=conf_dict,
+        sync_tensorboard=True,  # auto-upload sb3's tensorboard metrics
+    )
+
     model=PPO("MlpPolicy", custom_env, verbose=1, tensorboard_log=logdir)
 
     trainer = SB3Trainer(custom_env, models_dir, logdir, model)
@@ -46,7 +58,8 @@ def main(unused_argv):
     try:
         trainer = declare_trainer()
         trainer.train_model(timesteps=10000, reset_num_timesteps=False, 
-                            tb_log_name="PPO", repeat_times=30)
+                            tb_log_name="PPO", repeat_times=30,
+                            callback=WandbCallback())
         # trainer.load_model(f"{trainer.models_dir}/290000")
         # trainer.test_model(total_steps=10000, deterministic=True)
     except KeyboardInterrupt: