Deep Reinforcement Learning Log and No Output files

AI & Machine learning
1

I ran the capstone program. But I only saw this log and no files were saved.  How can I remove these extra logs?

 

1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 16ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 19ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 21ms/step
1/1 [==============================] - 0s 21ms/step
1/1 [==============================] - 0s 21ms/step

D:\Github\Jupyter-Lab\Deep-Reinforcement-Learning-in-Trading-Resources\RL-Capstone-Solution\RL>


I was supposed to get this log. 
Trade 001 | pos 1 | len 2 | approx cum ret -0.23% | trade ret -0.23% | eps 1.0010 | 2008-01-22 08:45:00 | 5003
Trade 002 | pos 1 | len 73 | approx cum ret 1.17% | trade ret 1.41% | eps 0.0635 | 2008-01-22 15:15:00 | 5081
Trade 003 | pos -1 | len 356 | approx cum ret 1.06% | trade ret -0.11% | eps 0.0133 | 2008-01-24 09:30:00 | 5588
Trade 004 | pos 1 | len 124 | approx cum ret 1.86% | trade ret 0.80% | eps 0.0049 | 2008-01-25 02:45:00 | 5795
Trade 005 | pos 1 | len 140 | approx cum ret 1.63% | trade ret -0.22% | eps 0.0026 | 2008-01-30 11:15:00 | 7050
Trade 006 | pos -1 | len 543 | approx cum ret 0.29% | trade ret -1.35% | eps 0.0018 | 2008-02-03 04:00:00 | 7828
Trade 007 | pos -1 | len 449 | approx cum ret 0.06% | trade ret -0.22% | eps 0.0014 | 2008-02-10 14:00:00 | 9677
Trade 008 | pos 1 | len 468 | approx cum ret -0.27% | trade ret -0.33% | eps 0.0012 | 2008-02-14 02:40:00 | 10693
Trade 009 | pos -1 | len 4156 | approx cum ret -2.92% | trade ret -2.65% | eps 0.0012 | 2008-03-04 07:40:00 | 15364
Trade 010 | pos 1 | len 1973 | approx cum ret -6.75% | trade ret -3.84% | eps 0.0011 | 2008-03-21 06:50:00 | 19676
----saving weights, trade logs and replay buffer-----
Trade 011 | pos -1 | len 2685 | approx cum ret -7.08% | trade ret -0.33% | eps 0.0011 | 2008-04-04 07:10:00 | 23138
Trade 012 | pos 1 | len 635 | approx cum ret -7.73% | trade ret -0.64% | eps 0.0010 | 2008-04-13 22:10:00 | 25336
Trade 013 | pos -1 | len 2674 | approx cum ret -7.30% | trade ret 0.43% | eps 0.0010 | 2008-04-28 03:20:00 | 28856
Trade 014 | pos -1 | len 541 | approx cum ret -6.14% | trade ret 1.17% | eps 0.0010 | 2008-05-14 08:25:00 | 32951
Trade 015 | pos 1 | len 9109 | approx cum ret -4.10% | trade ret 2.03% | eps 0.0010 | 2008-06-22 02:15:00 | 42387



Test mode is on due to resource constraints and therefore stopped after 15 trades.

You can trade on full dataset on your local computer and set TEST_MODE flag to False in rl_config dictionary.

The full code file, quantra_reinforemcent_learning module and data file is available in last unit of the course.




----saving weights, trade logs and replay buffer-----

FINISHED







Am I missing something?

2

I was able to resolve this issue by reinstalling the packages. But can someone tell me which part of the code or environment setting disables printing the following, 

 

1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 16ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 19ms/step
1/1 [==============================] - 0s 18ms/step
1/1 [==============================] - 0s 17ms/step
1/1 [==============================] - 0s 21ms/step

 

3

Hi Marcus,



It can be due to the parameter "verbose" in Keras which is by default 1. While not exactly related, you can check a similar query on StackOverflow here.



Hope this helps.