Update modeling_neollm.py

modeling_neollm.py

@@ -12,6 +12,7 @@ import torch
 import torch.nn.functional as F
 from torch import nn
 from cut_cross_entropy import linear_cross_entropy
+from configuration_neollm import NeoLLMConfig
 
 from transformers.activations import ACT2FN
 from transformers.generation import GenerationMixin
@@ -28,7 +29,6 @@ from transformers.utils.import_utils import (
     is_causal_conv1d_available,
     is_flash_linear_attention_available,
 )
-from configuration_neollm import NeoLLMConfig
 
 from transformers import AutoConfig, AutoModel, AutoModelForCausalLM