default learned value residual mix to true

setup.py

@@ -3,7 +3,7 @@
 setup(
   name = 'x-transformers',
   packages = find_packages(exclude=['examples']),
-  version = '1.42.26',
+  version = '1.42.28',
   license='MIT',
   description = 'X-Transformers - Pytorch',
   author = 'Phil Wang',

x_transformers/x_transformers.py

@@ -1584,7 +1584,7 @@ def __init__(
         unet_skips = False,
         reinject_input = False,              # seen first in DEQ paper https://arxiv.org/abs/1909.01377, but later used in a number of papers trying to achieve depthwise generalization https://arxiv.org/abs/2410.03020v1
         add_value_residual = False,          # resformer from Zhou et al - https://arxiv.org/abs/2410.17897v1
-        learned_value_residual_mix = False,  # seeing big improvements when the value residual mix value is learned per token - credit goes to @faresobeid for taking the first step with learned scalar mix, then @Blinkdl for taking it a step further with data dependent. here we will use per token learned
+        learned_value_residual_mix = True,   # seeing big improvements when the value residual mix value is learned per token - credit goes to @faresobeid for taking the first step with learned scalar mix, then @Blinkdl for taking it a step further with data dependent. here we will use per token learned
         rel_pos_kwargs: dict = dict(),
         **kwargs
     ):