Unit Tests for curriculums

[nmitra28]

DomainRandomization LearningProgressCurriculum

other methods commented out due to failure.

[nmitra28]

1. DomainRandomization with seed! SUCCESSFUL
2. DomainRandomization without seed! SUCCESSFUL
3. LearningProgressCurriculum with seed! SUCCESSFUL
4. LearningProgressCurriculum without seed! SUCCESSFUL
5. SequentialCurriculum with seed! SUCCESSFUL
6. SequentialCurriculum without seed! SUCCESSFUL
7. CentralizedPrioritizedLevelReplay with seed! SUCCESSFUL
8. CentralizedPrioritizedLevelReplay without seed! SUCCESSFUL
9. PrioritizedLevelReplay with seed! SUCCESSFUL
10. PrioritizedLevelReplay without seed! SUCCESSFUL

[nmitra28]

python -m syllabus.examples.training_scripts.cleanrl_procgen_plr --env-id="bigfish" --seed=1 --num-envs=4 --num-eval-episodes=3

For these metrics the expected return is :

global_step=164, episodic_return=1.0
global_step=300, episodic_return=0.0
global_step=380, episodic_return=3.0
global_step=420, episodic_return=0.0
global_step=592, episodic_return=1.0
global_step=728, episodic_return=4.0
global_step=740, episodic_return=0.0
global_step=760, episodic_return=3.0
global_step=908, episodic_return=1.0
global_step=936, episodic_return=4.0
SPS: 63
global_step=1040, episodic_return=0.0
global_step=1140, episodic_return=3.0
global_step=1200, episodic_return=0.0
global_step=1348, episodic_return=0.0
global_step=1484, episodic_return=4.0
global_step=1496, episodic_return=0.0
global_step=1564, episodic_return=2.0
global_step=1684, episodic_return=0.0
global_step=1836, episodic_return=4.0
global_step=1844, episodic_return=0.0
global_step=1864, episodic_return=3.0
global_step=1984, episodic_return=0.0
SPS: 60
global_step=2140, episodic_return=0.0
global_step=2200, episodic_return=2.0
global_step=2264, episodic_return=0.0
global_step=2364, episodic_return=4.0
global_step=2404, episodic_return=0.0
global_step=2540, episodic_return=0.0
global_step=2580, episodic_return=3.0
global_step=2680, episodic_return=0.0
global_step=2812, episodic_return=6.0
global_step=2820, episodic_return=0.0
global_step=2964, episodic_return=3.0
global_step=2968, episodic_return=0.0
SPS: 60
global_step=3116, episodic_return=0.0
global_step=3160, episodic_return=4.0
global_step=3244, episodic_return=0.0
global_step=3320, episodic_return=3.0
global_step=3388, episodic_return=0.0
global_step=3512, episodic_return=0.0
global_step=3656, episodic_return=0.0
global_step=3704, episodic_return=3.0
global_step=3728, episodic_return=4.0
global_step=3812, episodic_return=0.0
global_step=3972, episodic_return=0.0
global_step=4000, episodic_return=4.0
global_step=4060, episodic_return=2.0
SPS: 61
global_step=4120, episodic_return=0.0
global_step=4260, episodic_return=0.0
global_step=4400, episodic_return=0.0
global_step=4440, episodic_return=3.0
global_step=4528, episodic_return=0.0
global_step=4676, episodic_return=0.0
global_step=4724, episodic_return=6.0
global_step=4760, episodic_return=3.0
global_step=4792, episodic_return=0.0
global_step=4820, episodic_return=3.0
global_step=4972, episodic_return=1.0
global_step=5108, episodic_return=0.0
SPS: 60
global_step=5180, episodic_return=3.0
global_step=5288, episodic_return=1.0
global_step=5400, episodic_return=0.0
global_step=5540, episodic_return=0.0
global_step=5564, episodic_return=3.0
global_step=5572, episodic_return=3.0
global_step=5676, episodic_return=0.0
global_step=5712, episodic_return=6.0
global_step=5816, episodic_return=0.0
global_step=5948, episodic_return=3.0
global_step=5976, episodic_return=0.0
global_step=6120, episodic_return=0.0
SPS: 60
global_step=6252, episodic_return=0.0
global_step=6328, episodic_return=3.0
global_step=6404, episodic_return=1.0
global_step=6412, episodic_return=4.0
global_step=6556, episodic_return=0.0
global_step=6676, episodic_return=4.0
global_step=6688, episodic_return=4.0
global_step=6700, episodic_return=0.0
global_step=6836, episodic_return=0.0
global_step=6964, episodic_return=0.0
global_step=7068, episodic_return=3.0
global_step=7136, episodic_return=1.0

.....................

[nmitra28]

#11 DomainRandomization with seed with sample_weights
space = TaskSpace(gym.spaces.Discrete(4), ["a", "b", "c","d"])
c = DomainRandomization(task_space = space, seed = seed, sample_weights = [0.6,0.2,0.1,0.1])
if seed_test(c = c) :
    print("DomainRandomization with seed with sample weights! SUCCESSFUL")

#12: DomainRandomization without seed
c = DomainRandomization(task_space = space, sample_weights = [0.3,0.2,0.4,0.1])
if no_seed_test(c = c) :
    print("DomainRandomization without seed with sample weights! SUCCESSFUL")
    

[0]
[0]
[0]
[0]
[0]
DomainRandomization with seed with sample weights! SUCCESSFUL
[3]
[0]
[2]
[2]
[2]
DomainRandomization without seed with sample weights! SUCCESSFUL