[EN] 10-3.md duplicate words "the the"

docs/en/week10/10-3.md

@@ -91,7 +91,7 @@ The emulator takes the current location ($\tcab^t$,$\xcab^t, \ycab^t$, $\ttraile
 |![]({{site.baseurl}}/images/week10/10-3/fig8.png)|
 |  <center><b>Fig. 5</b>: Training the Neural-net Emulator</center>|
 
-In this setup, the the simulator can tell us the location of next step given the current location and steering angle. Therefore, we don’t really need a neural-net that emulates the simulator. However, in a more complex system, we may not have access to the underlying equations of the system, *i.e.* we do not have the laws of the universe in a nice computable form. We may only observe data that records sequences of steering signals and their corresponding paths. In this case, we want to train a neural-net to emulate the dynamic of this complex system.
+In this setup, the simulator can tell us the location of next step given the current location and steering angle. Therefore, we don’t really need a neural-net that emulates the simulator. However, in a more complex system, we may not have access to the underlying equations of the system, *i.e.* we do not have the laws of the universe in a nice computable form. We may only observe data that records sequences of steering signals and their corresponding paths. In this case, we want to train a neural-net to emulate the dynamic of this complex system.
 
 In order to train enumlator, there are two important function in `Class truck` we need to look into when we train the emulator.