output.txt

Sending build context to Docker daemon    190kB
Step 1/6 : FROM python:3.8.5
 ---> 62aa40094bb1
Step 2/6 : RUN pip install numpy==1.19
 ---> Using cache
 ---> 2bdb87e7497a
Step 3/6 : RUN pip install gym==0.17
 ---> Using cache
 ---> cab2ee18195e
Step 4/6 : COPY ./*.py /usr/src/
 ---> 3e1d23334423
Step 5/6 : WORKDIR /usr/src
 ---> Running in 8216bc11f174
Removing intermediate container 8216bc11f174
 ---> 722ac26f28e8
Step 6/6 : CMD ["python", "main.py"]
 ---> Running in f010825a5388
Removing intermediate container f010825a5388
 ---> a0023a0c9065
Successfully built a0023a0c9065
Successfully tagged taxi:latest


Begin Taxi-v3 learning...
Parameters:
path_memory_decay=0.7, Q_weight=0.02, recency_weight=3, eps_start=-0.005, eps_decay=5e-05
learning_rate=0.7, discount_rate=0.5, min_stochasticity=0

Episode 10000/150000 || epsilon=0.6035056, Best average reward -87.38

Episode 20000/150000 || epsilon=0.3660446, Best average reward -18.69

Episode 30000/150000 || epsilon=0.2220173, Best average reward -3.92

Episode 40000/150000 || epsilon=0.1346603, Best average reward 1.88

Episode 50000/150000 || epsilon=0.0816756, Best average reward 5.42

Episode 60000/150000 || epsilon=0.0495388, Best average reward 6.34

Episode 70000/150000 || epsilon=0.0300468, Best average reward 7.82

Episode 80000/150000 || epsilon=0.0182243, Best average reward 8.15

Episode 90000/150000 || epsilon=0.0110536, Best average reward 8.29

Episode 100000/150000 || epsilon=0.0067043, Best average reward 8.34

Episode 110000/150000 || epsilon=0.0040664, Best average reward 8.72

Episode 120000/150000 || epsilon=0.0024664, Best average reward 8.72

Episode 130000/150000 || epsilon=0.0014959, Best average reward 8.72

Episode 140000/150000 || epsilon=0.0009073, Best average reward 9.26

Episode 150000/150000 || epsilon=0.0005503, Best average reward 9.26

Run 3/1, average so far=9.26

Local seed:  80
Average:  9.26
Median:  9.26
[9.26]