increase default frac gradient for state transition projection

lucidrains · Nov 10, 2023 · 78405cc · 78405cc
1 parent 4c02b43
commit 78405cc
Show file tree

Hide file tree

Showing 2 changed files with 3 additions and 3 deletions.
diff --git a/gateloop_transformer/gateloop_transformer.py b/gateloop_transformer/gateloop_transformer.py
@@ -181,7 +181,7 @@ def __init__(
         heads = None,
         dim_inner = None,
         checkpoint_gate_looped_attn = True,
-        frac_gradient_state_transition = 0.5
+        frac_gradient_state_transition = 0.9
     ):
         super().__init__()
         self.frac_gradient_state_transition = frac_gradient_state_transition
@@ -259,7 +259,7 @@ def __init__(
         dim_gate_looped_attn = None,
         attn_softmax_normalize = None,
         data_dependent_rel_pos = False,
-        frac_gradient_state_transition = 0.5,
+        frac_gradient_state_transition = 0.9,
         ablate_complex = False,
         ablate_state_transition = False,
         rotary_emb = False

diff --git a/setup.py b/setup.py
@@ -3,7 +3,7 @@
 setup(
   name = 'gateloop-transformer',
   packages = find_packages(exclude=[]),
-  version = '0.0.14',
+  version = '0.0.15',
   license='MIT',
   description = 'GateLoop Transformer',
   author = 'Phil Wang',