x-transformers 1.42.25__py3-none-any.whl → 1.42.27__py3-none-any.whl

x_transformers/x_transformers.py

@@ -51,8 +51,8 @@ def default(val, d):
         return val
     return d() if callable(d) else d
 
-def first(it):
-    return it[0]
+def first(it, default = None):
+    return it[0] if len(it) > 0 else default
 
 def is_empty(x):
     return len(x) == 0
@@ -1077,7 +1077,7 @@ class Attention(Module):
         logit_softclamp_value = 50.,
         neutreno_value_residual = False, # Nguyen et al. https://arxiv.org/abs/2312.00751
         neutreno_alpha = 0.4,
-        learned_value_residual_mix = False,
+        learned_value_residual_mix = True,
         laser = False, # https://arxiv.org/abs/2411.03493v1
         laser_softclamp_value = 15.,
         onnxable = False,
@@ -1357,7 +1357,6 @@ class Attention(Module):
             k = k * self.qk_norm_k_scale
 
         if exists(rotary_pos_emb):
-
             freqs, xpos_scale = rotary_pos_emb
             q_xpos_scale, k_xpos_scale = (xpos_scale, xpos_scale ** -1.) if exists(xpos_scale) else (1., 1.)
 
@@ -1989,7 +1988,7 @@ class AttentionLayers(Module):
 
         if exists(self.rotary_pos_emb):
             if not exists(rotary_pos_emb):
-                maybe_mem = mems[0] # todo - handle edge case where different layers get different memory lengths. don't think this will ever come up but who knows
+                maybe_mem = first(mems, None) # todo - handle edge case where different layers get different memory lengths. don't think this will ever come up but who knows
                 mem_len = maybe_mem.shape[1] if exists(maybe_mem) else 0
 
                 if not exists(pos):

{x_transformers-1.42.25.dist-info → x_transformers-1.42.27.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: x-transformers
-Version: 1.42.25
+Version: 1.42.27
 Summary: X-Transformers - Pytorch
 Home-page: https://github.com/lucidrains/x-transformers
 Author: Phil Wang

{x_transformers-1.42.25.dist-info → x_transformers-1.42.27.dist-info}/RECORD

@@ -6,11 +6,11 @@ x_transformers/dpo.py,sha256=xt4OuOWhU8pN3OKN2LZAaC2NC8iiEnchqqcrPWVqf0o,3521
 x_transformers/multi_input.py,sha256=tCh-fTJDj2ib4SMGtsa-AM8MxKzJAQSwqAXOu3HU2mg,9252
 x_transformers/neo_mlp.py,sha256=XCNnnop9WLarcxap1kGuYc1x8GHvwkZiDRnXOxSl3Po,3452
 x_transformers/nonautoregressive_wrapper.py,sha256=2NU58hYMgn-4Jzg3mie-mXb0XH_dCN7fjlzd3K1rLUY,10510
-x_transformers/x_transformers.py,sha256=tj4s_p46Up89RcIFJF4aZ4iWtt4fpDVHKHqXv23Oekk,96643
+x_transformers/x_transformers.py,sha256=mLAqXQuZynqueJDkTEBs-kE9Uk8mSq_DF8UG9oY65Ns,96695
 x_transformers/xl_autoregressive_wrapper.py,sha256=CvZMJ6A6PA-Y_bQAhnORwjJBSl6Vjq2IdW5KTdk8NI8,4195
 x_transformers/xval.py,sha256=7S00kCuab4tWQa-vf-z-XfzADjVj48MoFIr7VSIvttg,8575
-x_transformers-1.42.25.dist-info/LICENSE,sha256=As9u198X-U-vph5noInuUfqsAG2zX_oXPHDmdjwlPPY,1066
-x_transformers-1.42.25.dist-info/METADATA,sha256=I2JJliI_WRW_0_tQoigduXIaYgDcU4YGdxJJKJ62BHE,739
-x_transformers-1.42.25.dist-info/WHEEL,sha256=PZUExdf71Ui_so67QXpySuHtCi3-J3wvF4ORK6k_S8U,91
-x_transformers-1.42.25.dist-info/top_level.txt,sha256=hO6KGpFuGucRNEtRfme4A_rGcM53AKwGP7RVlRIxS5Q,15
-x_transformers-1.42.25.dist-info/RECORD,,
+x_transformers-1.42.27.dist-info/LICENSE,sha256=As9u198X-U-vph5noInuUfqsAG2zX_oXPHDmdjwlPPY,1066
+x_transformers-1.42.27.dist-info/METADATA,sha256=g6KI8a3WyHUyq9w5Tq3aQatgr89gpc5IMZ5c1zAGlHU,739
+x_transformers-1.42.27.dist-info/WHEEL,sha256=PZUExdf71Ui_so67QXpySuHtCi3-J3wvF4ORK6k_S8U,91
+x_transformers-1.42.27.dist-info/top_level.txt,sha256=hO6KGpFuGucRNEtRfme4A_rGcM53AKwGP7RVlRIxS5Q,15
+x_transformers-1.42.27.dist-info/RECORD,,