1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
|
{ stdenv
, lib
, buildPythonPackage
, pythonOlder
, fetchPypi
, tensorflow
, annoy
, pillow
, matplotlib
, numpy
, pandas
, pygame
, pyopengl
, scipy
, scikitimage
, gym
, bokeh
, kubernetes
, redis
, minio
, pytest
, psutil
}:
buildPythonPackage rec {
version = "1.0.1";
pname = "rl-coach";
src = fetchPypi {
inherit pname version;
sha256 = "0i47hf0l76ydyrky6f8h760bfr0zg5g3vy675x6m6pgm9wrklkqc";
};
propagatedBuildInputs = [
tensorflow
annoy
pillow
matplotlib
numpy
pandas
pygame
pyopengl
scipy
scikitimage
gym
bokeh
kubernetes
redis
minio
psutil
];
checkInputs = [
pytest
];
# run only some tests that do not need any optional dependencies
# available tests: https://github.com/NervanaSystems/coach/tree/master/rl_coach/tests
testsToRun = [
# test only the tensorflow backend (not mxnet)
"architectures/tensorflow_components"
"agents"
"exploration_policies"
"filters"
"memories"
"utils"
];
checkPhase = let
fullTestPaths = map (testfile: "rl_coach/tests/${testfile}") testsToRun;
escapedPaths = map lib.escapeShellArg fullTestPaths;
pytestArgs = builtins.concatStringsSep " " escapedPaths;
in
''
pytest ${pytestArgs}
'';
postPatch = ''
# pinned to 8.0.1 for unknown reason, at least basic functionallity seems to work without it
# https://github.com/NervanaSystems/coach/pull/149#issuecomment-495915804
sed -i '/kubernetes/d' requirements.txt
# this is just an "intel-optimized" version of tensorflow, e.g. an implementation detail
sed -i 's/intel-tensorflow/tensorflow/g' setup.py
# backports of python3 features not needed
# https://github.com/NervanaSystems/coach/issues/324
sed -i '/futures/d' requirements.txt
'';
disabled = pythonOlder "3.5"; # minimum required version
meta = with stdenv.lib; {
description = "Enables easy experimentation with state of the art Reinforcement Learning algorithms";
homepage = "https://nervanasystems.github.io/coach/";
license = licenses.asl20;
maintainers = with maintainers; [ timokau ];
# pythonPackages.gym is too new
broken = true; # since 2020-04-20
};
}
|