diff --git a/internlm/core/naive_amp.py b/internlm/core/naive_amp.py
index a660359a..37c2dd8c 100644
--- a/internlm/core/naive_amp.py
+++ b/internlm/core/naive_amp.py
@@ -53,7 +53,7 @@ def __init__(
             self._world_size = 1
             self._sync_buf = False
         self._first_eval_run = False
-        
+
         # not-norm parameters
         self.not_norm = []
         # norm parameters
diff --git a/internlm/model/modeling_internlm.py b/internlm/model/modeling_internlm.py
index 02fbdc4d..ac36878f 100644
--- a/internlm/model/modeling_internlm.py
+++ b/internlm/model/modeling_internlm.py
@@ -195,7 +195,7 @@ def _forward(self, hidden_states=None, cu_seqlens=None, indexes=None, inference_
 
         if self.residual_in_fp32:
             residual1 = residual1.to(torch.float32)
-        
+
         if self.norm_fp32:
             hidden_states = hidden_states.to(self.dtype)
 
diff --git a/internlm/train/training_internlm.py b/internlm/train/training_internlm.py
index c5ce710b..827e75f4 100644
--- a/internlm/train/training_internlm.py
+++ b/internlm/train/training_internlm.py
@@ -101,12 +101,12 @@ def initialize_optimizer(model: Union[nn.Module, nn.ModuleList]):
 
     adam_cfg = gpc.config.adam
     if gpc.config.model.get("norm_fp32"):
-        params=[
+        params = [
             {"params": model.not_norm, "weight_decay": adam_cfg.weight_decay, "name": "default"},
             {"params": model.norm, "weight_decay": adam_cfg.weight_decay, "name": "norm"},
         ]
     else:
-        params=[{"params": model.parameters(), "weight_decay": adam_cfg.weight_decay}]
+        params = [{"params": model.parameters(), "weight_decay": adam_cfg.weight_decay}]
     naive_optimizer = torch.optim.AdamW(
         params=params,
         lr=adam_cfg.lr,