thu-ml · Trinkle23897 · Aug 30, 2021 · Aug 26, 2021 · Aug 26, 2021 · Aug 27, 2021
diff --git a/README.md b/README.md
@@ -192,7 +192,7 @@ buffer_size = 20000
 eps_train, eps_test = 0.1, 0.05
 step_per_epoch, step_per_collect = 10000, 10
 writer = SummaryWriter('log/dqn')  # tensorboard is also supported!
-logger = ts.utils.BasicLogger(writer)
+logger = ts.utils.TensorboardLogger(writer)
 ```
 
 Make environments:

diff --git a/docs/tutorials/cheatsheet.rst b/docs/tutorials/cheatsheet.rst
@@ -40,8 +40,8 @@ This is related to `Issue 349 <https://github.com/thu-ml/tianshou/issues/349>`_.
 To resume training process from an existing checkpoint, you need to do the following things in the training process:
 
 1. Make sure you write ``save_checkpoint_fn`` which saves everything needed in the training process, i.e., policy, optim, buffer; pass it to trainer;
-2. Use ``BasicLogger`` which contains a tensorboard;
-3. To adjust the save frequency, specify ``save_interval`` when initializing BasicLogger.
+2. Use ``TensorboardLogger``;
+3. To adjust the save frequency, specify ``save_interval`` when initializing TensorboardLogger.
 
 And to successfully resume from a checkpoint:
 

diff --git a/docs/tutorials/dqn.rst b/docs/tutorials/dqn.rst
@@ -148,9 +148,9 @@ The trainer supports `TensorBoard <https://www.tensorflow.org/tensorboard>`_ for
 ::
 
     from torch.utils.tensorboard import SummaryWriter
-    from tianshou.utils import BasicLogger
+    from tianshou.utils import TensorboardLogger
     writer = SummaryWriter('log/dqn')
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
 Pass the logger into the trainer, and the training result will be recorded into the TensorBoard.
 

diff --git a/docs/tutorials/tictactoe.rst b/docs/tutorials/tictactoe.rst
@@ -176,7 +176,7 @@ So let's start to train our Tic-Tac-Toe agent! First, import some required modul
     import numpy as np
     from copy import deepcopy
     from torch.utils.tensorboard import SummaryWriter
-    from tianshou.utils import BasicLogger
+    from tianshou.utils import TensorboardLogger
 
     from tianshou.env import DummyVectorEnv
     from tianshou.utils.net.common import Net
@@ -323,7 +323,7 @@ With the above preparation, we are close to the first learned agent. The followi
     log_path = os.path.join(args.logdir, 'tic_tac_toe', 'dqn')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     # ======== callback functions used during training =========
 

diff --git a/examples/atari/atari_bcq.py b/examples/atari/atari_bcq.py
@@ -7,7 +7,7 @@
 import numpy as np
 from torch.utils.tensorboard import SummaryWriter
 
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offline_trainer
 from tianshou.utils.net.discrete import Actor
@@ -116,7 +116,7 @@ def test_discrete_bcq(args=get_args()):
         f'seed_{args.seed}_{datetime.datetime.now().strftime("%m%d-%H%M%S")}')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer, update_interval=args.log_interval)
+    logger = TensorboardLogger(writer, update_interval=args.log_interval)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_c51.py b/examples/atari/atari_c51.py
@@ -6,7 +6,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import C51Policy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offpolicy_trainer
 from tianshou.data import Collector, VectorReplayBuffer
@@ -101,7 +101,7 @@ def test_c51(args=get_args()):
     log_path = os.path.join(args.logdir, args.task, 'c51')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_cql.py b/examples/atari/atari_cql.py
@@ -7,7 +7,7 @@
 import numpy as np
 from torch.utils.tensorboard import SummaryWriter
 
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offline_trainer
 from tianshou.policy import DiscreteCQLPolicy
@@ -108,7 +108,7 @@ def test_discrete_cql(args=get_args()):
         f'seed_{args.seed}_{datetime.datetime.now().strftime("%m%d-%H%M%S")}')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer, update_interval=args.log_interval)
+    logger = TensorboardLogger(writer, update_interval=args.log_interval)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_crr.py b/examples/atari/atari_crr.py
@@ -7,7 +7,7 @@
 import numpy as np
 from torch.utils.tensorboard import SummaryWriter
 
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offline_trainer
 from tianshou.utils.net.discrete import Actor
@@ -117,7 +117,7 @@ def test_discrete_crr(args=get_args()):
         f'seed_{args.seed}_{datetime.datetime.now().strftime("%m%d-%H%M%S")}')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer, update_interval=args.log_interval)
+    logger = TensorboardLogger(writer, update_interval=args.log_interval)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_dqn.py b/examples/atari/atari_dqn.py
@@ -6,7 +6,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import DQNPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offpolicy_trainer
 from tianshou.data import Collector, VectorReplayBuffer
@@ -96,7 +96,7 @@ def test_dqn(args=get_args()):
     log_path = os.path.join(args.logdir, args.task, 'dqn')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_fqf.py b/examples/atari/atari_fqf.py
@@ -6,7 +6,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import FQFPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offpolicy_trainer
 from tianshou.data import Collector, VectorReplayBuffer
@@ -112,7 +112,7 @@ def test_fqf(args=get_args()):
     log_path = os.path.join(args.logdir, args.task, 'fqf')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_iqn.py b/examples/atari/atari_iqn.py
@@ -6,7 +6,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import IQNPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offpolicy_trainer
 from tianshou.data import Collector, VectorReplayBuffer
@@ -109,7 +109,7 @@ def test_iqn(args=get_args()):
     log_path = os.path.join(args.logdir, args.task, 'iqn')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_qrdqn.py b/examples/atari/atari_qrdqn.py
@@ -5,7 +5,7 @@
 import numpy as np
 from torch.utils.tensorboard import SummaryWriter
 
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.policy import QRDQNPolicy
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offpolicy_trainer
@@ -99,7 +99,7 @@ def test_qrdqn(args=get_args()):
     log_path = os.path.join(args.logdir, args.task, 'qrdqn')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/atari/atari_rainbow.py b/examples/atari/atari_rainbow.py
@@ -7,7 +7,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import RainbowPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offpolicy_trainer
 from tianshou.data import Collector, PrioritizedVectorReplayBuffer, VectorReplayBuffer
@@ -121,7 +121,7 @@ def test_rainbow(args=get_args()):
         f'seed_{args.seed}_{datetime.datetime.now().strftime("%m%d-%H%M%S")}')
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/box2d/acrobot_dualdqn.py b/examples/box2d/acrobot_dualdqn.py
@@ -7,7 +7,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import DQNPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import DummyVectorEnv
 from tianshou.utils.net.common import Net
 from tianshou.trainer import offpolicy_trainer
@@ -82,7 +82,7 @@ def test_dqn(args=get_args()):
     # log
     log_path = os.path.join(args.logdir, args.task, 'dqn')
     writer = SummaryWriter(log_path)
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/box2d/bipedal_hardcore_sac.py b/examples/box2d/bipedal_hardcore_sac.py
@@ -7,7 +7,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import SACPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.utils.net.common import Net
 from tianshou.env import SubprocVectorEnv
 from tianshou.trainer import offpolicy_trainer
@@ -134,7 +134,7 @@ def test_sac_bipedal(args=get_args()):
     # log
     log_path = os.path.join(args.logdir, args.task, 'sac')
     writer = SummaryWriter(log_path)
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/box2d/lunarlander_dqn.py b/examples/box2d/lunarlander_dqn.py
@@ -7,7 +7,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import DQNPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.utils.net.common import Net
 from tianshou.trainer import offpolicy_trainer
 from tianshou.data import Collector, VectorReplayBuffer
@@ -84,7 +84,7 @@ def test_dqn(args=get_args()):
     # log
     log_path = os.path.join(args.logdir, args.task, 'dqn')
     writer = SummaryWriter(log_path)
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/box2d/mcc_sac.py b/examples/box2d/mcc_sac.py
@@ -7,7 +7,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import SACPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import DummyVectorEnv
 from tianshou.exploration import OUNoise
 from tianshou.utils.net.common import Net
@@ -104,7 +104,7 @@ def test_sac(args=get_args()):
     # log
     log_path = os.path.join(args.logdir, args.task, 'sac')
     writer = SummaryWriter(log_path)
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/mujoco/mujoco_a2c.py b/examples/mujoco/mujoco_a2c.py
@@ -13,7 +13,7 @@
 from torch.distributions import Independent, Normal
 
 from tianshou.policy import A2CPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.utils.net.common import Net
 from tianshou.trainer import onpolicy_trainer
@@ -141,7 +141,7 @@ def dist(*logits):
     log_path = os.path.join(args.logdir, args.task, 'a2c', log_file)
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer, update_interval=100, train_interval=100)
+    logger = TensorboardLogger(writer, update_interval=100, train_interval=100)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/mujoco/mujoco_ddpg.py b/examples/mujoco/mujoco_ddpg.py
@@ -10,7 +10,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import DDPGPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.utils.net.common import Net
 from tianshou.exploration import GaussianNoise
@@ -110,7 +110,7 @@ def test_ddpg(args=get_args()):
     log_path = os.path.join(args.logdir, args.task, 'ddpg', log_file)
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/mujoco/mujoco_npg.py b/examples/mujoco/mujoco_npg.py
@@ -13,7 +13,7 @@
 from torch.distributions import Independent, Normal
 
 from tianshou.policy import NPGPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.utils.net.common import Net
 from tianshou.trainer import onpolicy_trainer
@@ -142,7 +142,7 @@ def dist(*logits):
     log_path = os.path.join(args.logdir, args.task, 'npg', log_file)
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer, update_interval=100, train_interval=100)
+    logger = TensorboardLogger(writer, update_interval=100, train_interval=100)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/mujoco/mujoco_ppo.py b/examples/mujoco/mujoco_ppo.py
@@ -13,7 +13,7 @@
 from torch.distributions import Independent, Normal
 
 from tianshou.policy import PPOPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.utils.net.common import Net
 from tianshou.trainer import onpolicy_trainer
@@ -149,7 +149,7 @@ def dist(*logits):
     log_path = os.path.join(args.logdir, args.task, 'ppo', log_file)
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer, update_interval=100, train_interval=100)
+    logger = TensorboardLogger(writer, update_interval=100, train_interval=100)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/mujoco/mujoco_reinforce.py b/examples/mujoco/mujoco_reinforce.py
@@ -13,7 +13,7 @@
 from torch.distributions import Independent, Normal
 
 from tianshou.policy import PGPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.utils.net.common import Net
 from tianshou.trainer import onpolicy_trainer
@@ -131,7 +131,7 @@ def dist(*logits):
     log_path = os.path.join(args.logdir, args.task, 'reinforce', log_file)
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer, update_interval=10, train_interval=100)
+    logger = TensorboardLogger(writer, update_interval=10, train_interval=100)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))

diff --git a/examples/mujoco/mujoco_sac.py b/examples/mujoco/mujoco_sac.py
@@ -10,7 +10,7 @@
 from torch.utils.tensorboard import SummaryWriter
 
 from tianshou.policy import SACPolicy
-from tianshou.utils import BasicLogger
+from tianshou.utils import TensorboardLogger
 from tianshou.env import SubprocVectorEnv
 from tianshou.utils.net.common import Net
 from tianshou.trainer import offpolicy_trainer
@@ -122,7 +122,7 @@ def test_sac(args=get_args()):
     log_path = os.path.join(args.logdir, args.task, 'sac', log_file)
     writer = SummaryWriter(log_path)
     writer.add_text("args", str(args))
-    logger = BasicLogger(writer)
+    logger = TensorboardLogger(writer)
 
     def save_fn(policy):
         torch.save(policy.state_dict(), os.path.join(log_path, 'policy.pth'))