output.txt

Sending build context to Docker daemon  130.6kB
Step 1/6 : FROM python:3.6
 ---> b8e9b6d2fef3
Step 2/6 : RUN pip install numpy==1.15
 ---> Using cache
 ---> da63177ecd08
Step 3/6 : RUN pip install gym==0.11
 ---> Using cache
 ---> 089c688f04cb
Step 4/6 : COPY ./*.py /usr/src/
 ---> Using cache
 ---> 7e235a9e8e69
Step 5/6 : WORKDIR /usr/src
 ---> Using cache
 ---> 21fd2d3e5011
Step 6/6 : CMD ["python", "main.py"]
 ---> Using cache
 ---> 3628aba0a96c
Successfully built 3628aba0a96c
Successfully tagged taxi:latest


Begin Taxi-v2 learning...
Parameters:
path_memory_decay=0.7, Q_weight=0.02, recency_weight=8, eps_start=-0.005, eps_decay=0.00022
learning_rate=0.7, discount_rate=0.5, min_stochasticity=0

Episode 10000/80000 || epsilon=0.1102505, Best average reward 4.2

Episode 20000/80000 || epsilon=0.0122161, Best average reward 8.78

Episode 30000/80000 || epsilon=0.0013536, Best average reward 9.38

Episode 40000/80000 || epsilon=0.0001500, Best average reward 9.38

Episode 50000/80000 || epsilon=0.0000166, Best average reward 9.38

Episode 60000/80000 || epsilon=0.0000018, Best average reward 9.38

Episode 70000/80000 || epsilon=0.0000002, Best average reward 9.57

Episode 80000/80000 || epsilon=0.0000000, Best average reward 9.57

Run 8/1, average so far=9.57

Local seed:  0
Average:  9.57
Median:  9.57
[9.57]