SequentialLearning
/

SuperLinear

mixture-of-experts

Model card Files Files and versions

razmars commited on Apr 23

Commit

c15d2e4

·

verified ·

1 Parent(s): 5969221

Update configuration_super_linear.py

Files changed (1) hide show

configuration_super_linear.py +1 -1

configuration_super_linear.py CHANGED Viewed

@@ -54,7 +54,7 @@ class SuperLinearConfig(PretrainedConfig):
         noisy_gating_std_decay = 1,
         torch_dtype = "float32",
         transformers_version = "4.40.1",
-        use_fft = 1
         **kwargs,                          # any extra CLI args
     ):
         self.seq_len         = seq_len

         noisy_gating_std_decay = 1,
         torch_dtype = "float32",
         transformers_version = "4.40.1",
+        use_fft = 1,
         **kwargs,                          # any extra CLI args
     ):
         self.seq_len         = seq_len