Adding correct mpsf parameters for t1 experiment

Author: Federico-PizarroBejarano

experiments/crazyflie/config_overrides/ppo_crazyflie.yaml

@@ -20,10 +20,10 @@ algo_config:
   eval_batch_size: 10
 
   # misc
-  log_interval: 12500
+  log_interval: 10000
   save_interval: 0
   num_checkpoints: 0
-  eval_interval: 12500
+  eval_interval: 10000
   eval_save_best: True
   tensorboard: False
 

experiments/crazyflie/models/mpsc_parameters/nl_mpsc_crazyflie_track.pkl

@@ -9,13 +9,13 @@ algo_config:
   deque_size: 10
   entropy_coef: 0.01
   eval_batch_size: 10
-  eval_interval: 12500
+  eval_interval: 10000
   eval_save_best: true
   filter_train_actions: true
   gae_lambda: 0.95
   gamma: 0.99
   hidden_dim: 128
-  log_interval: 12500
+  log_interval: 10000
   max_env_steps: 250000
   max_grad_norm: 0.5
   mini_batch_size: 256

experiments/crazyflie/models/rl_models/ppo/mpsf_0.1_dm_t1/config.yaml

@@ -1,11 +1,26 @@
 step,loss/approx_kl
-25000,0.015373280962618687
-50000,0.005466316675301641
-75000,0.007334801884523281
-100000,0.008855676196981221
-125000,0.007131252387383333
-150000,0.016024184126096467
-175000,0.02016867712372914
-200000,0.014259486020697903
-225000,0.027465859936395037
-250000,0.018274522332164148
+10000,0.015022389074632278
+20000,0.015573783230502158
+30000,0.015440155026347685
+40000,0.014503382332623004
+50000,0.02573415245860815
+60000,0.01905389988096431
+70000,0.008920146268792452
+80000,0.010281776105208942
+90000,0.018977865506894884
+100000,0.015178235802644244
+110000,0.019403759688915063
+120000,0.013040242402348667
+130000,0.011364517908077689
+140000,0.017196612591700012
+150000,0.01807373953051865
+160000,0.011021253722719847
+170000,0.016570236758949856
+180000,0.024614000114767505
+190000,0.006834416506656757
+200000,0.01071121544421961
+210000,0.020620357982503876
+220000,0.021714769635582342
+230000,0.0165287251099168
+240000,0.02274632802194295
+250000,0.016948295312856014

experiments/crazyflie/models/rl_models/ppo/mpsf_0.1_dm_t1/logs/loss/approx_kl.log

@@ -1,11 +1,26 @@
 step,loss/entropy_loss
-25000,-1.9422575136025746
-50000,-1.9314473847548168
-75000,-1.718100424607595
-100000,-1.6027250786622367
-125000,-1.4119963268438975
-150000,-1.5207090298334758
-175000,-1.4677082558472954
-200000,-1.476611814896266
-225000,-1.4850205163160959
-250000,-1.5199606776237486
+10000,-1.8325788120428723
+20000,-1.8085498253504428
+30000,-1.8321571131547292
+40000,-1.8132434546947482
+50000,-1.7964839041233063
+60000,-1.8120882650216419
+70000,-1.7626775642236077
+80000,-1.6036625961462654
+90000,-1.6579803645610809
+100000,-1.607422707478205
+110000,-1.5929721852143606
+120000,-1.6025017718474068
+130000,-1.545069944858551
+140000,-1.4628295660018922
+150000,-1.455063772201538
+160000,-1.43400755127271
+170000,-1.3863306105136872
+180000,-1.31354029973348
+190000,-1.2574030061562855
+200000,-1.1542476773262025
+210000,-1.1533651630083723
+220000,-1.191436105966568
+230000,-1.2106483121713
+240000,-1.2057694375514985
+250000,-1.1224131127198538

experiments/crazyflie/models/rl_models/ppo/mpsf_0.1_dm_t1/logs/loss/entropy_loss.log

@@ -1,11 +1,26 @@
 step,loss/policy_loss
-25000,-0.008487398050761605
-50000,-0.00880940280480252
-75000,-0.009933911955918398
-100000,-0.005607683146801758
-125000,-0.004757624400151668
-150000,-0.003386248358056189
-175000,-0.002005064096720082
-200000,-0.005854441330401723
-225000,0.004003693091684401
-250000,-0.010062139415958049
+10000,0.002284406456391001
+20000,-0.001423316262468889
+30000,-0.016813184545244362
+40000,-0.004833155219460405
+50000,-0.012382710698938884
+60000,-0.006246745855067655
+70000,-0.010446838865443212
+80000,-0.007222142933292428
+90000,-0.01305004737691406
+100000,-0.005176228468921089
+110000,-0.0011207284944143468
+120000,-0.009628546985728589
+130000,-0.014381752366211017
+140000,-0.006747774420920974
+150000,-0.0067902182613011714
+160000,-0.012696297748397293
+170000,-0.013005741052013006
+180000,-0.007825781184775279
+190000,-0.014886107628898106
+200000,-0.014714001347183567
+210000,-0.012574326461046037
+220000,0.0033774217629884335
+230000,-0.004315913423119138
+240000,-0.013798009918731525
+250000,-0.012423067433133008

experiments/crazyflie/models/rl_models/ppo/mpsf_0.1_dm_t1/logs/loss/policy_loss.log

@@ -1,11 +1,26 @@
 step,loss/value_loss
-25000,7.793937429077428
-50000,13.893374047401428
-75000,23.89295015323139
-100000,19.281110565347028
-125000,27.118762816218055
-150000,12.193586144909709
-175000,18.288683568642096
-200000,11.609428511226813
-225000,5.738791902706543
-250000,4.967841796357956
+10000,16.000818497426486
+20000,5.13283882419356
+30000,10.46975174663164
+40000,7.622796251049335
+50000,16.70927202819223
+60000,11.259414334638953
+70000,11.998459610985089
+80000,30.419484205929546
+90000,19.23323946859423
+100000,15.947925519648427
+110000,24.232413715187626
+120000,12.315418160180055
+130000,24.538134771657603
+140000,9.818098521062183
+150000,19.958843225060594
+160000,21.223275025276525
+170000,24.341612772019694
+180000,8.468285359961438
+190000,23.882806073319767
+200000,22.90826762596818
+210000,13.029109533080453
+220000,7.9204959334016065
+230000,17.910784285702228
+240000,10.170660478601082
+250000,30.20223377740395

experiments/crazyflie/models/rl_models/ppo/mpsf_0.1_dm_t1/logs/loss/value_loss.log

@@ -1,11 +1,26 @@
 step,stat/ep_constraint_violation
-25000,309.5
-50000,126.5
-75000,100.5
-100000,105.5
-125000,219.0
-150000,31.5
-175000,34.5
-200000,31.5
-225000,18.5
-250000,12.0
+10000,291.0
+20000,431.5
+30000,224.5
+40000,314.0
+50000,300.5
+60000,359.0
+70000,227.5
+80000,219.5
+90000,263.5
+100000,206.5
+110000,159.5
+120000,47.0
+130000,218.0
+140000,65.0
+150000,162.0
+160000,61.5
+170000,214.5
+180000,60.5
+190000,57.0
+200000,99.5
+210000,61.5
+220000,45.0
+230000,88.0
+240000,29.0
+250000,149.0

experiments/crazyflie/models/rl_models/ppo/mpsf_0.1_dm_t1/logs/stat/ep_constraint_violation.log

@@ -1,11 +1,26 @@
 step,stat/ep_length
-25000,375.0
+10000,375.0
+20000,375.0
+30000,375.0
+40000,375.0
 50000,375.0
-75000,375.0
+60000,375.0
+70000,375.0
+80000,375.0
+90000,375.0
 100000,375.0
-125000,375.0
+110000,375.0
+120000,375.0
+130000,375.0
+140000,375.0
 150000,375.0
-175000,375.0
+160000,375.0
+170000,375.0
+180000,375.0
+190000,375.0
 200000,375.0
-225000,375.0
+210000,375.0
+220000,375.0
+230000,375.0
+240000,375.0
 250000,375.0

experiments/crazyflie/models/rl_models/ppo/mpsf_0.1_dm_t1/logs/stat/ep_length.log

@@ -1,11 +1,26 @@
 step,stat/ep_return
-25000,98.29296866306379
-50000,204.49077267324304
-75000,231.51700314647044
-100000,244.08289157053508
-125000,204.9318586508445
-150000,267.52341422784
-175000,279.615252885838
-200000,282.54841177443075
-225000,291.78329345184795
-250000,277.8975227464764
+10000,49.55504245197726
+20000,87.82771536449232
+30000,93.4982807082286
+40000,159.1211679466761
+50000,170.15868105017842
+60000,144.09704977951912
+70000,186.89882118270816
+80000,184.73422775877845
+90000,190.2953890708943
+100000,166.6781872505294
+110000,196.06540985033354
+120000,211.86574901277646
+130000,196.21003447908765
+140000,229.90488276419038
+150000,216.74168536598637
+160000,244.92908104364415
+170000,196.6948384054461
+180000,264.22386569827097
+190000,292.33070484504367
+200000,272.7727249144965
+210000,264.7823390332579
+220000,257.77452064337535
+230000,245.8423306097055
+240000,252.02534838376062
+250000,215.97332757259866