diff --git a/cleanrl/dqn.py b/cleanrl/dqn.py
index 081efe871..fcfafa6ea 100644
--- a/cleanrl/dqn.py
+++ b/cleanrl/dqn.py
@@ -214,7 +214,8 @@ def linear_schedule(start_e: float, end_e: float, duration: int, t: int):
             if global_step % args.target_network_frequency == 0:
                 for target_network_param, q_network_param in zip(target_network.parameters(), q_network.parameters()):
                     target_network_param.data.copy_(
-                        args.tau * q_network_param.data + (1. - args.tau) * target_network_param.data)
+                        args.tau * q_network_param.data + (1.0 - args.tau) * target_network_param.data
+                    )
 
     if args.save_model:
         model_path = f"runs/{run_name}/{args.exp_name}.cleanrl_model"
diff --git a/cleanrl/dqn_atari.py b/cleanrl/dqn_atari.py
index 7202c970a..e0e5a2b4d 100644
--- a/cleanrl/dqn_atari.py
+++ b/cleanrl/dqn_atari.py
@@ -236,7 +236,8 @@ def linear_schedule(start_e: float, end_e: float, duration: int, t: int):
             if global_step % args.target_network_frequency == 0:
                 for target_network_param, q_network_param in zip(target_network.parameters(), q_network.parameters()):
                     target_network_param.data.copy_(
-                        args.tau * q_network_param.data + (1. - args.tau) * target_network_param.data)
+                        args.tau * q_network_param.data + (1.0 - args.tau) * target_network_param.data
+                    )
 
     if args.save_model:
         model_path = f"runs/{run_name}/{args.exp_name}.cleanrl_model"
diff --git a/cleanrl/dqn_atari_jax.py b/cleanrl/dqn_atari_jax.py
index 1cea0bee2..12a4e16ae 100644
--- a/cleanrl/dqn_atari_jax.py
+++ b/cleanrl/dqn_atari_jax.py
@@ -264,7 +264,9 @@ def mse_loss(params):
 
             # update target network
             if global_step % args.target_network_frequency == 0:
-                q_state = q_state.replace(target_params=optax.incremental_update(q_state.params, q_state.target_params, args.tau)) 
+                q_state = q_state.replace(
+                    target_params=optax.incremental_update(q_state.params, q_state.target_params, args.tau)
+                )
 
     if args.save_model:
         model_path = f"runs/{run_name}/{args.exp_name}.cleanrl_model"
diff --git a/cleanrl/dqn_jax.py b/cleanrl/dqn_jax.py
index e5802a8ee..82c05499e 100644
--- a/cleanrl/dqn_jax.py
+++ b/cleanrl/dqn_jax.py
@@ -236,7 +236,9 @@ def mse_loss(params):
 
             # update target network
             if global_step % args.target_network_frequency == 0:
-                q_state = q_state.replace(target_params=optax.incremental_update(q_state.params, q_state.target_params, args.tau)) 
+                q_state = q_state.replace(
+                    target_params=optax.incremental_update(q_state.params, q_state.target_params, args.tau)
+                )
 
     if args.save_model:
         model_path = f"runs/{run_name}/{args.exp_name}.cleanrl_model"