{"batches":[[{"__dict__":[["logp_new",{"__nd__":[-1.5859367723042639,-2.50163198529031,-0.8444215818966174,-0.5644756308600728,-3.7069483669259107],"dtype":"float64","shape":[5]}],["logp_old",{"__nd__":[-1.1952829202455686,-2.5399841062404955,-0.7495488041935428,-0.5594352836087861,-3.4510351886538366],"dtype":"float64","shape":[5]}],["advantages",{"__nd__":[0.8793979748628286,0.7777919354289483,0.06603069756121605,1.1272412069680329,0.4675093422520456],"dtype":"float64","shape":[5]}],["eps",0.2]]},{"__dict__":[["loss",-0.5893585173942038],["fraction",0.0],["grad",{"__nd__":[-0.11900265328658485,-0.16164026210524926,-0.012010833789911576,-0.22431476293199193,-0.07239000528046606],"dtype":"float64","shape":[5]}]]}],[{"__dict__":[["logp_new",{"__nd__":[-2.176570430701199,-0.924882910119507,-2.2434720993250576,0.44927410174250315,-1.7531334191785606,-1.9409837280810294,-2.5878159085278805,0.03053104996177808,-1.0900433357083563,-1.9853015509905447,-2.2722117889694937,-1.379931422292271],"dtype":"float64","shape":[12]}],["logp_old",{"__nd__":[-2.3592924628832384,-1.1312492159175012,-2.458882600828999,-0.6215496986927275,-1.5499259109862529,-1.6848623635452606,-2.1809295444039414,-0.27745866132596975,-1.6545294820688021,-1.9283278221631073,-1.8521335504882295,-0.9676908144466513],"dtype":"float64","shape":[12]}],["advantages",{"__nd__":[0.6505927878247011,0.7432541712034423,0.543154268305195,-0.6655097072886943,0.23216132306671977,0.11668580914072822,0.21868859672901295,0.8714287779481898,0.22359554877468227,0.6789135630718949,0.06757906948889146,0.28911939868998415],"dtype":"float64","shape":[12]}],["eps",0.1]]},{"__dict__":[["loss",-0.22466240473697519],["fraction",0.4166666666666667],["grad",{"__nd__":[0.0,0.0,0.0,0.1618176970678832,-0.015789076340679515,-0.007526701638109188,-0.012132111389599179,0.0,0.0,-0.05344288109622647,-0.0036999282363876645,-0.015953727315370387],"dtype":"float64","shape":[12]}]]}],[{"__dict__":[["logp_new",{"__nd__":[-1.0953692169621558,-3.1410462770559886,-1.7202390675465487,-1.9418875070786825,-2.0007807774297883,-1.8605927804939906,0.026649249449738732,-2.240713036899776,-0.5935909533525668,-3.0915147241043224,-1.967270218199105,-1.409857704208009,-0.9901252474403877,-1.027941349324953,-0.6092582686429584,-1.9426055402889824,-1.9598529689190298,-0.5458747869616644,-1.601998089675626,-2.6426093015433647],"dtype":"float64","shape":[20]}],["logp_old",{"__nd__":[-0.8687117741614596,-2.9571558198556662,-1.8196712163573014,-1.9703726542927955,-2.138877848243342,-1.7751422512266837,-0.0050586887656041135,-2.3658311156932434,-0.5317216454085192,-3.1828697716158048,-1.8348850299857749,-1.3372469348949945,-0.9137776686407219,-0.788773420207145,-0.7066527648000748,-1.8487250722484376,-1.9623517926645673,-0.6420241187428462,-1.6913043248816149,-2.775686323337922],"dtype":"float64","shape":[20]}],["advantages",{"__nd__":[-0.09848548450942361,-0.42329831204415375,-0.07971821090639905,-1.6873344339580298,-1.4471124724230873,-1.3226996123544024,-0.9972468276014818,0.3997742267234366,-0.9054790553600608,-0.3781625540393897,1.2992282977860654,-0.35626397106142593,0.7375155684670865,-0.933617680009877,-0.20543755786763002,-0.9500220549105812,-0.3390330759005625,0.8403081374573955,-1.7273204231923487,0.43442364354585733],"dtype":"float64","shape":[20]}],["eps",0.2]]},{"__dict__":[["loss",0.40640100974351273],["fraction",0.1],["grad",{"__nd__":[0.0,0.017609779993396982,0.004402611678504832,0.08680447523840153,0.08307055017645047,0.06071843053287176,0.051468706092914526,-0.022652851992900732,0.04255777502371235,0.02071684037644276,-0.056906427058746045,0.01656561074971518,-0.0341651919118575,0.0,0.011322641070841877,0.04324460381998839,0.016994065958248664,-0.04625574688005235,0.09443394166347038,-0.02481293136866238],"dtype":"float64","shape":[20]}]]}],[{"__dict__":[["logp_new",{"__nd__":[-1.4537488996461416,-2.9129479497942863,-2.802637346423402,-1.2518731451087184,-0.9421426488714533,-0.599234791913633,-1.1926510071569496,1.272421204296565],"dtype":"float64","shape":[8]}],["logp_old",{"__nd__":[-1.2622643976677221,-2.0941499556967944,-2.946057854388455,-1.4278704922861305,-2.0294927090638026,-1.267323788645296,-1.4781478544765572,0.10177889132091544],"dtype":"float64","shape":[8]}],["advantages",{"__nd__":[-1.188763054322851,-0.6397515327497477,-0.9265759414055249,-0.38980980315576796,-1.3766861475563088,0.6351509468144043,-0.22222269709877338,-1.4708062945026579],"dtype":"float64","shape":[8]}],["eps",0.3]]},{"__dict__":[["loss",1.4074310679007747],["fraction",0.25],["grad",{"__nd__":[0.12270002453738044,0.0,0.13368348779193415,0.058102851477996674,0.510475715511643,0.0,0.036956270111627346,0.5927464882119308],"dtype":"float64","shape":[8]}]]}]]}