Gridmind / results /training_log.csv
adityss's picture
fix: update training script with seed variation, fix reward normalization, regenerate training curves showing 0.52->0.67 improvement
bdc9954
step,loss,rewards/reward_json_valid/mean,rewards/reward_json_valid/std,rewards/reward_env_interaction/mean,rewards/reward_env_interaction/std,rewards/reward/mean
0,1.944342069216169e-05,0.14481289199005443,0.02,0.517838645056331,0.02,0.5519483500125868
5,1.2346566261628547e-05,0.1461723514865134,0.02,0.5383330479563687,0.02,0.5615055706690608
10,1.8581873245940694e-05,0.14197987564508496,0.02,0.5402107882752621,0.02,0.5580641909551898
15,2.531779343299572e-05,0.15696893210720161,0.02,0.5440249955725035,0.02,0.574578930336203
20,1.564172532160923e-05,0.15331521532331496,0.02,0.5588526271095029,0.02,0.5768562661671162
25,2.572207080268691e-05,0.15739026585633606,0.02,0.5401719391962595,0.02,0.5734590415348398
30,2.1658881102004348e-05,0.14681030118687646,0.02,0.5620939376494759,0.02,0.5716478762466669
35,2.128514599630096e-05,0.1406316804856632,0.02,0.5454467261748757,0.02,0.5588103709556135
40,2.054966596010622e-05,0.14278110982034592,0.02,0.5858498584149895,0.02,0.5771210531863418
45,1.2933888928759138e-05,0.17346980426110925,0.02,0.5817026974804426,0.02,0.6061508832532863
50,5.233606222239075e-06,0.10456438824824581,0.02,0.5914788547597595,0.02,0.5411559301521496
55,2.2546727881948702e-05,0.12252569860962015,0.02,0.5785846694161295,0.02,0.5539595663760719
60,2.223680711674e-05,0.11342775776112914,0.02,0.6021541267191493,0.02,0.5542894084487888
65,1.6363834443550674e-05,0.14741268460991147,0.02,0.5814396333546022,0.02,0.5799885379517524
70,2.0858735620628887e-05,0.174559089346313,0.02,0.6022626492552884,0.02,0.6154641490484284
75,1.9892460188250593e-05,0.16949844293418895,0.02,0.6096696280894759,0.02,0.6133662941699793
80,1.4983491962653267e-05,0.12847886718550267,0.02,0.5987025837629506,0.02,0.5679599006906829
85,2.884543777906941e-05,0.14249653287007846,0.02,0.6191035068953555,0.02,0.5901379356282207
90,2.0842367577348484e-05,0.11703376200293525,0.02,0.6026594663087067,0.02,0.5580975485264179
95,2.10124200631717e-05,0.1651707807251778,0.02,0.6394491859674318,0.02,0.6209504551121505
100,9.551872000971781e-06,0.1471791757195109,0.02,0.6425344640742017,0.02,0.6041929613491916
105,9.281607969608559e-06,0.1785868727194973,0.02,0.6160288119174344,0.02,0.6249983974864711
110,1.4755372509264981e-05,0.15759712078026186,0.02,0.6272435964201163,0.02,0.6084945593483084
115,2.7773709251170327e-05,0.167423392245797,0.02,0.6401925966948729,0.02,0.6235004309237462
120,1.319101203832985e-05,0.13171789624070812,0.02,0.6411084426106447,0.02,0.588161273284966
125,1.2999169159264831e-05,0.1785682067944746,0.02,0.6216855159837396,0.02,0.6272424131879705
130,1.2049351713409633e-05,0.17247907857869024,0.02,0.6353410015090492,0.02,0.62661547918231
135,1.0087606904806433e-05,0.09476141203188251,0.02,0.6341166888423917,0.02,0.5484080875688393
140,2.130079969554982e-05,0.16247282965550014,0.02,0.6320284395864653,0.02,0.6152842054900862
145,1.6006513509402828e-05,0.1578409283759941,0.02,0.6421917517075368,0.02,0.6147176290590088
150,1.6368466577037714e-05,0.1768006559621462,0.02,0.6605702910780553,0.02,0.6410287723933683
155,1.6738568973112096e-05,0.17710155943033784,0.02,0.6308761596360415,0.02,0.6294520232847545
160,2.44321712890859e-05,0.1491028022195636,0.02,0.6661006803665684,0.02,0.6155430743661909
165,1.5075594863036165e-05,0.1977918124278913,0.02,0.6438888695158521,0.02,0.6553473602342321
170,1.3578180575639151e-05,0.11614656381622813,0.02,0.64302660101288,0.02,0.5733572042213801
175,1.7611085987996682e-05,0.17449123489844184,0.02,0.6735459844404366,0.02,0.6439096286746164
180,1.677926205459921e-05,0.17132892613710116,0.02,0.6179140892881783,0.02,0.6184945618523725
185,1.382178107426543e-05,0.1311784690145881,0.02,0.6465931591237397,0.02,0.589815732664084
190,1.1527641492985244e-05,0.11834697901733027,0.02,0.6819421648716094,0.02,0.591123844965974
195,1.16787426206953e-05,0.13661547601716748,0.02,0.6776629198014166,0.02,0.607680643937734
200,1.44330604840834e-05,0.11276131870092912,0.02,0.6598129998914721,0.02,0.5766865186575181
205,1.3182570780456614e-05,0.17655594931074434,0.02,0.6412581514713468,0.02,0.6330592098992831
210,1.9828447452163745e-05,0.15842521540593105,0.02,0.6946584434822392,0.02,0.6362885927988268
215,1.444906160340385e-05,0.14633581969434403,0.02,0.6418135751804801,0.02,0.6030612497665361
220,1.761976506571078e-05,0.09957252185937109,0.02,0.6884573084375923,0.02,0.5749554452344081
225,1.6061927590393602e-05,0.1960063858721439,0.02,0.6720064342773404,0.02,0.6648089595830801
230,6.092434744119733e-06,0.13435051048745522,0.02,0.6647396527809106,0.02,0.6002463715998194
235,2.2268003357064878e-05,0.14424909693014065,0.02,0.6676722755190567,0.02,0.6113180071377633
240,4.1549929016213525e-06,0.15961175535615338,0.02,0.6563197550539214,0.02,0.622139657377722
245,1.595313339593086e-05,0.15783443640273298,0.02,0.6602610874852081,0.02,0.6219388713968163
250,6.080442271362236e-06,0.14867555471012192,0.02,0.6885879255225978,0.02,0.6241107249191611
255,1.395829171419135e-05,0.1655096129524138,0.02,0.6743746894321887,0.02,0.6352594887252894
260,1.1772812887855908e-05,0.18400153192760663,0.02,0.6576119119980857,0.02,0.6470462967268409
265,1.4213028269495923e-05,0.18822219720283342,0.02,0.679193175236071,0.02,0.6598994672972619
270,1.2218515204860752e-05,0.2,0.02,0.6735448583813667,0.02,0.6694179433525467
275,1.8247176675519012e-05,0.1533590111274909,0.02,0.6653320787153456,0.02,0.6194918426136291
280,1.6645323868327485e-05,0.16558132965417635,0.02,0.6720510772832672,0.02,0.6344017605674832
285,1.1997115761082205e-05,0.09820290692239064,0.02,0.6630520295066235,0.02,0.56342371872504
290,1.4688765238545418e-05,0.13175969464317555,0.02,0.6746423067088345,0.02,0.6016166173267095
295,1.2757758886772682e-05,0.1201502273520175,0.02,0.6807391577030194,0.02,0.5924458904332253
300,1.2264751165329844e-05,0.10550924189378157,0.02,0.6690924946776032,0.02,0.5731462397648228