update gru

Signed-off-by: Mayank Mishra <mayank31398@gmail.com>

Author: mayank31398

lm_engine/hf_models/modeling_utils/sequence_mixer_blocks/gru.py

@@ -132,7 +132,6 @@ def __init__(
 
         self.norm = get_normalization_function(normalization_function, self.state_size)
 
-        mark_parameter_as_mup_learning_rate(self.conv1d.weight)
         mark_parameter_as_mup_learning_rate(self.input_projection.weight)
         mark_parameter_as_mup_learning_rate(self.state_weight)
         mark_parameter_as_mup_learning_rate(self.output_projection.weight)

lm_engine/hf_models/modeling_utils/sequence_mixer_blocks/rnn.py

@@ -101,7 +101,6 @@ def __init__(
 
         self.norm = get_normalization_function(normalization_function, self.state_size)
 
-        mark_parameter_as_mup_learning_rate(self.conv1d.weight)
         mark_parameter_as_mup_learning_rate(self.input_projection.weight)
         mark_parameter_as_mup_learning_rate(self.state_weight)
         mark_parameter_as_mup_learning_rate(self.output_projection.weight)