diff --git a/conf/config.yaml b/conf/config.yaml
deleted file mode 100644
index 83cfa46..0000000
--- a/conf/config.yaml
+++ /dev/null
@@ -1,10 +0,0 @@
-params:
-  num_agents: 5
-  max_agents: 10
-  num_obs: 3
-  grid_size: 20
-  num_steps: 50
-  random_seed: 0
-  visualize: True
-  step_delay: 0.01
-  sim_type: "two_d"
\ No newline at end of file
diff --git a/simulate.py b/simulate.py
index 6c1cd3a..badfd52 100644
--- a/simulate.py
+++ b/simulate.py
@@ -1,56 +1,55 @@
 import time 
+import argparse
 
-import hydra
-from omegaconf import DictConfig, OmegaConf
 from jax import random
 
 from simulationsandbox.two_d_simulation import SimpleSimulation
 from simulationsandbox.three_d_simulation import ThreeDSimulation
 
 
-@hydra.main(version_base=None, config_path="conf", config_name="config")
-def main(cfg: DictConfig):
-    print(OmegaConf.to_yaml(cfg))
+def main():
+    parser = argparse.ArgumentParser()
+    parser.add_argument("--num_agents", type=int, default=5)
+    parser.add_argument("--max_agents", type=int, default=10)
+    parser.add_argument("--num_obs", type=int, default=3)
+    parser.add_argument("--grid_size", type=int, default=20)
+    parser.add_argument("--num_steps", type=int, default=60)
+    parser.add_argument("--step_delay", type=float, default=0.01)
+    parser.add_argument("--sim_type", type=str, default="two_d")
+    parser.add_argument("--random_seed", type=int, default=0)
+    parser.add_argument("--visualize", action="store_false")
+    args = parser.parse_args()  
 
-    num_agents = cfg.params.num_agents
-    max_agents = cfg.params.max_agents
-    num_obs = cfg.params.num_obs
-    grid_size = cfg.params.grid_size
-    num_steps = cfg.params.num_steps
-    visualize = cfg.params.visualize
-    step_delay = cfg.params.step_delay
-    sim_type = cfg.params.sim_type
-
-    key = random.PRNGKey(cfg.params.random_seed)
+    key = random.PRNGKey(args.random_seed)
 
     # Choose a simulation type
-    if sim_type == "two_d":
+    if args.sim_type == "two_d":
         Simulation = SimpleSimulation
-    elif sim_type == "three_d":
+    elif args.sim_type == "three_d":
         Simulation = ThreeDSimulation
     else:
-        raise(ValueError(f"Unknown sim type {sim_type}"))
+        raise(ValueError(f"Unknown sim type {args.sim_type}"))
 
-    sim = Simulation(max_agents, grid_size)
-    state = sim.init_state(num_agents, num_obs, key)
+    sim = Simulation(args.max_agents, args.grid_size)
+    state = sim.init_state(args.num_agents, args.num_obs, key)
 
     # Launch a simulation
     print("Simulation started")
-    for timestep in range(num_steps):
-        time.sleep(step_delay)
+    for timestep in range(args.num_steps):
+        time.sleep(args.step_delay)
         key, a_key, step_key = random.split(key, 3)
 
         if timestep % 10 == 0:
             print(f"\nstep {timestep}")
 
-        if timestep == (num_steps // 3):
+        if timestep == (args.num_steps // 3):
             # Add 3 agents and change the color of an agent
             state = sim.add_agent(state, 7)
             state = sim.add_agent(state, 9)
             state = sim.add_agent(state, 5)
             state = state.replace(colors=state.colors.at[0, 2].set(1.0))
 
-        if timestep ==  2* (num_steps // 3):
+        if timestep ==  2* (args.num_steps // 3):
             # Remove 3 other agents and change the color of another agent
             state = sim.remove_agent(state, 2)
             state = sim.remove_agent(state, 1)
@@ -61,7 +60,7 @@ def main(cfg: DictConfig):
         actions = sim.choose_action(state.obs, a_key)
         state = sim.step(state, actions, step_key)
 
-        if visualize:
+        if args.visualize:
             Simulation.visualize_sim(state)
     print("\nSimulation ended")