metacontroller-pytorch 0.0.29__py3-none-any.whl → 0.0.31__py3-none-any.whl

metacontroller/metacontroller.py

@@ -58,8 +58,10 @@ def straight_through(src, tgt):
 
 MetaControllerOutput = namedtuple('MetaControllerOutput', (
     'prev_hiddens',
+    'input_residual_stream',
     'action_dist',
     'actions',
+    'switch_beta',
     'kl_loss',
     'switch_loss'
 ))
@@ -150,7 +152,7 @@ class MetaController(Module):
         residual_stream,
         cache: MetaControllerOutput | None = None,
         discovery_phase = False,
-        hard_switch = False,
+        hard_switch = None,
         temperature = 1.,
         episode_lens: Tensor | None = None
     ):
@@ -166,6 +168,8 @@ class MetaController(Module):
 
         meta_embed = self.model_to_meta(residual_stream)
 
+        hard_switch = default(hard_switch, not discovery_phase) # think during internal RL phase, it needs to be a hard switch, then only the actions emitted during the switch is reinforced
+
         if discovery_phase:
             logger.warning('meta controller cache being passed back in for discovery phase, which does not make sense given bidirectional encoder')
 
@@ -268,7 +272,7 @@ class MetaController(Module):
             sampled_latent_action[:, -1:]
         )
 
-        return control_signal, MetaControllerOutput(next_hiddens, action_dist, sampled_latent_action, kl_loss, switch_loss)
+        return control_signal, MetaControllerOutput(next_hiddens, residual_stream, action_dist, sampled_latent_action, switch_beta, kl_loss, switch_loss)
 
 # main transformer, which is subsumed into the environment after behavioral cloning
 

metacontroller/metacontroller_with_binary_mapper.py

@@ -28,6 +28,8 @@ from torch_einops_utils.save_load import save_load
 
 from vector_quantize_pytorch import BinaryMapper
 
+from metacontroller.metacontroller import MetaControllerOutput
+
 # constants
 
 LinearNoBias = partial(Linear, bias = False)
@@ -50,14 +52,6 @@ def straight_through(src, tgt):
 
 # meta controller
 
-MetaControllerOutput = namedtuple('MetaControllerOutput', (
-    'prev_hiddens',
-    'action_dist',
-    'codes',
-    'kl_loss',
-    'switch_loss'
-))
-
 @save_load()
 class MetaControllerWithBinaryMapper(Module):
     def __init__(
@@ -148,7 +142,7 @@ class MetaControllerWithBinaryMapper(Module):
         residual_stream,
         cache: MetaControllerOutput | None = None,
         discovery_phase = False,
-        hard_switch = False,
+        hard_switch = None,
         temperature = 1.,
         episode_lens: Tensor | None = None
     ):
@@ -164,6 +158,8 @@ class MetaControllerWithBinaryMapper(Module):
 
         meta_embed = self.model_to_meta(residual_stream)
 
+        hard_switch = default(hard_switch, not discovery_phase) # think during internal RL phase, it needs to be a hard switch, then only the actions emitted during the switch is reinforced
+
         if discovery_phase:
             mask = maybe(lens_to_mask)(episode_lens, meta_embed.shape[1])
 
@@ -265,4 +261,4 @@ class MetaControllerWithBinaryMapper(Module):
             sampled_codes[:, -1:]
         )
 
-        return control_signal, MetaControllerOutput(next_hiddens, binary_logits, sampled_codes, kl_loss, switch_loss)
+        return control_signal, MetaControllerOutput(next_hiddens, residual_stream, binary_logits, sampled_codes, switch_beta, kl_loss, switch_loss)

{metacontroller_pytorch-0.0.29.dist-info → metacontroller_pytorch-0.0.31.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: metacontroller-pytorch
-Version: 0.0.29
+Version: 0.0.31
 Summary: Transformer Metacontroller
 Project-URL: Homepage, https://pypi.org/project/metacontroller/
 Project-URL: Repository, https://github.com/lucidrains/metacontroller

metacontroller_pytorch-0.0.31.dist-info/RECORD

@@ -0,0 +1,8 @@
+metacontroller/__init__.py,sha256=lj7IOGpN--qMxJWbB-4SGqoPXG7Hd4mgtToTRSyTZ58,57
+metacontroller/metacontroller.py,sha256=lxWgeWFcXxSDm-ygd14DjyEOYIJIALcuLkoRAfEzNtc,14719
+metacontroller/metacontroller_with_binary_mapper.py,sha256=BrsQdkhlOyR2O5xAXTLC4p-uKOAbW7wET-lVU0qktws,8242
+metacontroller/metacontroller_with_resnet.py,sha256=YKHcazRZrrRParHRH-H_EPvT1-55LHKAs5pM6gwuT20,7394
+metacontroller_pytorch-0.0.31.dist-info/METADATA,sha256=mtOtYymI01jBMO7pyaAIJ166B5Mk3khH8CUwUMNLTKw,4747
+metacontroller_pytorch-0.0.31.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
+metacontroller_pytorch-0.0.31.dist-info/licenses/LICENSE,sha256=1yCiA9b5nhslTavxPjsQAO-wpOnwJR9-l8LTVi7GJuk,1066
+metacontroller_pytorch-0.0.31.dist-info/RECORD,,

metacontroller_pytorch-0.0.29.dist-info/RECORD

@@ -1,8 +0,0 @@
-metacontroller/__init__.py,sha256=lj7IOGpN--qMxJWbB-4SGqoPXG7Hd4mgtToTRSyTZ58,57
-metacontroller/metacontroller.py,sha256=LWEq069EnBP3Sr6FTiDtz0cM5SFFT1zl35WkU6_kWGA,14451
-metacontroller/metacontroller_with_binary_mapper.py,sha256=uUFCSIRq20TdctRd7O20A_I2SiB9AgYS6z5iQMFqf2Q,8107
-metacontroller/metacontroller_with_resnet.py,sha256=YKHcazRZrrRParHRH-H_EPvT1-55LHKAs5pM6gwuT20,7394
-metacontroller_pytorch-0.0.29.dist-info/METADATA,sha256=8zeOj2sUZ-5V_qGXvzXoBH3lpCJqHgPfZq0-YllSrTs,4747
-metacontroller_pytorch-0.0.29.dist-info/WHEEL,sha256=WLgqFyCfm_KASv4WHyYy0P3pM_m7J5L9k2skdKLirC8,87
-metacontroller_pytorch-0.0.29.dist-info/licenses/LICENSE,sha256=1yCiA9b5nhslTavxPjsQAO-wpOnwJR9-l8LTVi7GJuk,1066
-metacontroller_pytorch-0.0.29.dist-info/RECORD,,