rxnn 0.2.17__py3-none-any.whl → 0.2.18__py3-none-any.whl
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
rxnn/training/mrl.py
CHANGED
@@ -507,7 +507,7 @@ class MRLTrainer:
|
|
507
507
|
action=MrlActorAction.DECODE)
|
508
508
|
|
509
509
|
# 7. Calculate RL Algorithm (PPO etc.) loss
|
510
|
-
policy_loss = self.rl_algorithm.policy_loss(next_query, action,
|
510
|
+
policy_loss = self.rl_algorithm.policy_loss(next_query, action, logits, log_probs, advantages)
|
511
511
|
|
512
512
|
# 8. Reset gradients
|
513
513
|
self.optimizer.zero_grad()
|
@@ -16,7 +16,7 @@ rxnn/training/bml.py,sha256=FJszaQXOLx2ZHBa1CQpyMrG8i4Kj14E-gzDAEK_Ei5k,17272
|
|
16
16
|
rxnn/training/callbacks.py,sha256=-N0MQPpZQaUWCINdTOsjul4bDGbGr2JgQBqOIXBLS6o,35053
|
17
17
|
rxnn/training/dataset.py,sha256=7hTilFWPpqUEc6zNcMqBPjxFKxCfvTKKF3E8tVlwccQ,51250
|
18
18
|
rxnn/training/models.py,sha256=renPa5YH443XNTMFI-YTCwi5vNp3QzwF5UXedNd5hDk,5187
|
19
|
-
rxnn/training/mrl.py,sha256=
|
19
|
+
rxnn/training/mrl.py,sha256=Ar2g-vjqTq_4qLKc4L1Ai0j2LX-x98dmsx_VaWVV-Es,39448
|
20
20
|
rxnn/training/reward.py,sha256=C0ToTz-u-L-qyBd2yJ1HlvVPS110OChYj9ZhD6iSSMU,5654
|
21
21
|
rxnn/training/rl.py,sha256=DHFwnPUlnq2JVj6CS6DwifnC_eMeBAUVp36UCAWNMis,3934
|
22
22
|
rxnn/training/scheduler.py,sha256=LcjU35mEwz2U5x3U6tLfeeYlBqMxbFSxYzJYuXkWbSY,1408
|
@@ -32,7 +32,7 @@ rxnn/transformers/moe.py,sha256=j6jEx6Ip0zttlUZKKn82azxo95lkLZs-H2GLSMD88hY,5859
|
|
32
32
|
rxnn/transformers/positional.py,sha256=1PjcJybUzeQlIKJI4tahAGZcYgCRCL0otxs7mpsNuzM,4410
|
33
33
|
rxnn/transformers/sampler.py,sha256=t6iiQTdLQ0TakUWnnhKkb5DKF2F_9-thXHBydDF3fxg,17389
|
34
34
|
rxnn/utils.py,sha256=ihb6OTyDtPiocB_lOvnq7eOkjjpCkgs8wxvXUBNQ7mM,996
|
35
|
-
rxnn-0.2.
|
36
|
-
rxnn-0.2.
|
37
|
-
rxnn-0.2.
|
38
|
-
rxnn-0.2.
|
35
|
+
rxnn-0.2.18.dist-info/LICENSE,sha256=C8coDFIUYuOcke4JLPwTqahQUCyXyGq6WOaigOkx8tY,11275
|
36
|
+
rxnn-0.2.18.dist-info/METADATA,sha256=_hGNlaH_rclBfQdzA7tCFhkI-RZPiK5tNBM8tjUsbWQ,25960
|
37
|
+
rxnn-0.2.18.dist-info/WHEEL,sha256=b4K_helf-jlQoXBBETfwnf4B04YC67LOev0jo4fX5m8,88
|
38
|
+
rxnn-0.2.18.dist-info/RECORD,,
|
File without changes
|
File without changes
|