Archive 4 checkpoints (1-4)

paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/lora64sat/history.json

@@ -0,0 +1,2608 @@
+[
+ {
+  "step": 0,
+  "inner_val_loss": 0.10827601102716995,
+  "inner_val_per_cohort": {
+   "immicro": 0.10636159498244524,
+   "imperial_eskd": 0.1086193590678952,
+   "liu2021": 0.10984707903116941
+  },
+  "train_loss": null
+ },
+ {
+  "step": 250,
+  "inner_val_loss": 0.10726830744269218,
+  "inner_val_per_cohort": {
+   "immicro": 0.10542473662644625,
+   "imperial_eskd": 0.10751504010774872,
+   "liu2021": 0.10886514559388161
+  },
+  "train_loss": 0.10270090734958649,
+  "sec": 11.6
+ },
+ {
+  "step": 500,
+  "inner_val_loss": 0.10546406937029325,
+  "inner_val_per_cohort": {
+   "immicro": 0.10382997244596481,
+   "imperial_eskd": 0.10550473224033009,
+   "liu2021": 0.10705750342458487
+  },
+  "train_loss": 0.10063257715106011,
+  "sec": 23.7
+ },
+ {
+  "step": 750,
+  "inner_val_loss": 0.10373948439203128,
+  "inner_val_per_cohort": {
+   "immicro": 0.1024517451102535,
+   "imperial_eskd": 0.10336958142844113,
+   "liu2021": 0.1053971266373992
+  },
+  "train_loss": 0.10037505868077277,
+  "sec": 35.5
+ },
+ {
+  "step": 1000,
+  "inner_val_loss": 0.1024205447786083,
+  "inner_val_per_cohort": {
+   "immicro": 0.10137436197449763,
+   "imperial_eskd": 0.10169565711509097,
+   "liu2021": 0.10419161524623632
+  },
+  "train_loss": 0.1000797398686409,
+  "sec": 47.0
+ },
+ {
+  "step": 1250,
+  "inner_val_loss": 0.1015513694628772,
+  "inner_val_per_cohort": {
+   "immicro": 0.1005761946241061,
+   "imperial_eskd": 0.10063335333358157,
+   "liu2021": 0.10344456043094397
+  },
+  "train_loss": 0.10017555391788482,
+  "sec": 57.4
+ },
+ {
+  "step": 1500,
+  "inner_val_loss": 0.10100688134329487,
+  "inner_val_per_cohort": {
+   "immicro": 0.09999394478897254,
+   "imperial_eskd": 0.10001225058328021,
+   "liu2021": 0.10301444865763187
+  },
+  "train_loss": 0.09994212421774865,
+  "sec": 69.0
+ },
+ {
+  "step": 1750,
+  "inner_val_loss": 0.10065855000243344,
+  "inner_val_per_cohort": {
+   "immicro": 0.09958950740595658,
+   "imperial_eskd": 0.09963068636980923,
+   "liu2021": 0.10275545623153448
+  },
+  "train_loss": 0.10005727735161782,
+  "sec": 79.8
+ },
+ {
+  "step": 2000,
+  "inner_val_loss": 0.10043430085660833,
+  "inner_val_per_cohort": {
+   "immicro": 0.09930831100791693,
+   "imperial_eskd": 0.09940264204686339,
+   "liu2021": 0.10259194951504469
+  },
+  "train_loss": 0.09982226571440697,
+  "sec": 90.6
+ },
+ {
+  "step": 2250,
+  "inner_val_loss": 0.1002833242951469,
+  "inner_val_per_cohort": {
+   "immicro": 0.09911872446537018,
+   "imperial_eskd": 0.09925826605070721,
+   "liu2021": 0.10247298236936331
+  },
+  "train_loss": 0.09994704848527908,
+  "sec": 101.0
+ },
+ {
+  "step": 2500,
+  "inner_val_loss": 0.10017923964657867,
+  "inner_val_per_cohort": {
+   "immicro": 0.09898490117241938,
+   "imperial_eskd": 0.09915754774754698,
+   "liu2021": 0.10239527001976967
+  },
+  "train_loss": 0.1001971900165081,
+  "sec": 111.3
+ },
+ {
+  "step": 2750,
+  "inner_val_loss": 0.10010535929427304,
+  "inner_val_per_cohort": {
+   "immicro": 0.09888610740502675,
+   "imperial_eskd": 0.0990916107865897,
+   "liu2021": 0.10233835969120264
+  },
+  "train_loss": 0.09980975806713105,
+  "sec": 121.7
+ },
+ {
+  "step": 3000,
+  "inner_val_loss": 0.10005084914390487,
+  "inner_val_per_cohort": {
+   "immicro": 0.09881323544929425,
+   "imperial_eskd": 0.0990413288501176,
+   "liu2021": 0.10229798313230276
+  },
+  "train_loss": 0.09977523875236512,
+  "sec": 132.1
+ },
+ {
+  "step": 3250,
+  "inner_val_loss": 0.10001074099405245,
+  "inner_val_per_cohort": {
+   "immicro": 0.09875891171395779,
+   "imperial_eskd": 0.09900419447909702,
+   "liu2021": 0.10226911678910255
+  },
+  "train_loss": 0.09998596471548081,
+  "sec": 142.5
+ },
+ {
+  "step": 3500,
+  "inner_val_loss": 0.09997580461956636,
+  "inner_val_per_cohort": {
+   "immicro": 0.09871138942738374,
+   "imperial_eskd": 0.09897003051909534,
+   "liu2021": 0.10224599391222
+  },
+  "train_loss": 0.09993550208210945,
+  "sec": 153.0
+ },
+ {
+  "step": 3750,
+  "inner_val_loss": 0.0999490846268306,
+  "inner_val_per_cohort": {
+   "immicro": 0.09867636455843846,
+   "imperial_eskd": 0.09894275258887898,
+   "liu2021": 0.10222813673317432
+  },
+  "train_loss": 0.09980573174357414,
+  "sec": 163.3
+ },
+ {
+  "step": 4000,
+  "inner_val_loss": 0.09992596615256384,
+  "inner_val_per_cohort": {
+   "immicro": 0.09864036335299413,
+   "imperial_eskd": 0.09891629760915582,
+   "liu2021": 0.10222123749554157
+  },
+  "train_loss": 0.09974821138381958,
+  "sec": 173.6
+ },
+ {
+  "step": 4250,
+  "inner_val_loss": 0.0999076326631687,
+  "inner_val_per_cohort": {
+   "immicro": 0.09861395973712206,
+   "imperial_eskd": 0.09889652038162405,
+   "liu2021": 0.10221241787075996
+  },
+  "train_loss": 0.09977313852310181,
+  "sec": 183.9
+ },
+ {
+  "step": 4500,
+  "inner_val_loss": 0.09989045167134868,
+  "inner_val_per_cohort": {
+   "immicro": 0.09859093030293782,
+   "imperial_eskd": 0.09887237474322319,
+   "liu2021": 0.10220804996788502
+  },
+  "train_loss": 0.09970952105522156,
+  "sec": 194.3
+ },
+ {
+  "step": 4750,
+  "inner_val_loss": 0.09987612898376856,
+  "inner_val_per_cohort": {
+   "immicro": 0.09857195522636175,
+   "imperial_eskd": 0.09885659881613472,
+   "liu2021": 0.10219983290880919
+  },
+  "train_loss": 0.09959850868582726,
+  "sec": 204.8
+ },
+ {
+  "step": 5000,
+  "inner_val_loss": 0.0998623613268137,
+  "inner_val_per_cohort": {
+   "immicro": 0.09855174459517002,
+   "imperial_eskd": 0.09884344041347504,
+   "liu2021": 0.10219189897179604
+  },
+  "train_loss": 0.09977317202091217,
+  "sec": 215.0
+ },
+ {
+  "step": 5250,
+  "inner_val_loss": 0.09984872853289349,
+  "inner_val_per_cohort": {
+   "immicro": 0.09853378714372714,
+   "imperial_eskd": 0.09882956404577602,
+   "liu2021": 0.1021828344091773
+  },
+  "train_loss": 0.09972456341981888,
+  "sec": 225.5
+ },
+ {
+  "step": 5500,
+  "inner_val_loss": 0.09983680739697783,
+  "inner_val_per_cohort": {
+   "immicro": 0.09851657195637624,
+   "imperial_eskd": 0.09881759841333736,
+   "liu2021": 0.10217625182121992
+  },
+  "train_loss": 0.09976890355348587,
+  "sec": 235.9
+ },
+ {
+  "step": 5750,
+  "inner_val_loss": 0.099825535214158,
+  "inner_val_per_cohort": {
+   "immicro": 0.0985001086567839,
+   "imperial_eskd": 0.0988028957085176,
+   "liu2021": 0.10217360127717257
+  },
+  "train_loss": 0.0993116979598999,
+  "sec": 246.5
+ },
+ {
+  "step": 6000,
+  "inner_val_loss": 0.09981175562874837,
+  "inner_val_per_cohort": {
+   "immicro": 0.09848416689783335,
+   "imperial_eskd": 0.09878998588431966,
+   "liu2021": 0.10216111410409212
+  },
+  "train_loss": 0.09960695725679397,
+  "sec": 256.9
+ },
+ {
+  "step": 6250,
+  "inner_val_loss": 0.09980083682169817,
+  "inner_val_per_cohort": {
+   "immicro": 0.09846683156987031,
+   "imperial_eskd": 0.09878119487654079,
+   "liu2021": 0.10215448401868343
+  },
+  "train_loss": 0.09966992327570916,
+  "sec": 267.3
+ },
+ {
+  "step": 6500,
+  "inner_val_loss": 0.09978956626634104,
+  "inner_val_per_cohort": {
+   "immicro": 0.0984572305654486,
+   "imperial_eskd": 0.0987721945751797,
+   "liu2021": 0.10213927365839481
+  },
+  "train_loss": 0.0997515520453453,
+  "sec": 277.8
+ },
+ {
+  "step": 6750,
+  "inner_val_loss": 0.09977524873396061,
+  "inner_val_per_cohort": {
+   "immicro": 0.09844368354727824,
+   "imperial_eskd": 0.09875918213616718,
+   "liu2021": 0.10212288051843643
+  },
+  "train_loss": 0.09959788158535958,
+  "sec": 288.3
+ },
+ {
+  "step": 7000,
+  "inner_val_loss": 0.09976577483155209,
+  "inner_val_per_cohort": {
+   "immicro": 0.09843429146955411,
+   "imperial_eskd": 0.09875817732377486,
+   "liu2021": 0.10210485570132732
+  },
+  "train_loss": 0.09952098730206489,
+  "sec": 298.7
+ },
+ {
+  "step": 7250,
+  "inner_val_loss": 0.09975827131608521,
+  "inner_val_per_cohort": {
+   "immicro": 0.0984239283328255,
+   "imperial_eskd": 0.09875076568939468,
+   "liu2021": 0.1021001199260354
+  },
+  "train_loss": 0.09974249657988549,
+  "sec": 309.0
+ },
+ {
+  "step": 7500,
+  "inner_val_loss": 0.09975153274570751,
+  "inner_val_per_cohort": {
+   "immicro": 0.09841166281451781,
+   "imperial_eskd": 0.09874993732029741,
+   "liu2021": 0.10209299810230732
+  },
+  "train_loss": 0.09955127349495888,
+  "sec": 319.4
+ },
+ {
+  "step": 7750,
+  "inner_val_loss": 0.09974195210808756,
+  "inner_val_per_cohort": {
+   "immicro": 0.09839833310494821,
+   "imperial_eskd": 0.09874412519010631,
+   "liu2021": 0.10208339802920818
+  },
+  "train_loss": 0.09954452291131019,
+  "sec": 329.9
+ },
+ {
+  "step": 8000,
+  "inner_val_loss": 0.09973331814575376,
+  "inner_val_per_cohort": {
+   "immicro": 0.09838834963738918,
+   "imperial_eskd": 0.09873570671135729,
+   "liu2021": 0.1020758980885148
+  },
+  "train_loss": 0.09924208855628967,
+  "sec": 340.4
+ },
+ {
+  "step": 8250,
+  "inner_val_loss": 0.09972917481391419,
+  "inner_val_per_cohort": {
+   "immicro": 0.09838081213335197,
+   "imperial_eskd": 0.09873436255888506,
+   "liu2021": 0.10207234974950552
+  },
+  "train_loss": 0.09968985944986343,
+  "sec": 350.9
+ },
+ {
+  "step": 8500,
+  "inner_val_loss": 0.0997211794660549,
+  "inner_val_per_cohort": {
+   "immicro": 0.09837040274093549,
+   "imperial_eskd": 0.09872928702018478,
+   "liu2021": 0.10206384863704443
+  },
+  "train_loss": 0.09927724292874336,
+  "sec": 361.4
+ },
+ {
+  "step": 8750,
+  "inner_val_loss": 0.09971236850539543,
+  "inner_val_per_cohort": {
+   "immicro": 0.09835935073594253,
+   "imperial_eskd": 0.09872640262950551,
+   "liu2021": 0.10205135215073824
+  },
+  "train_loss": 0.099745321393013,
+  "sec": 371.8
+ },
+ {
+  "step": 9000,
+  "inner_val_loss": 0.09970722291054147,
+  "inner_val_per_cohort": {
+   "immicro": 0.09835299849510193,
+   "imperial_eskd": 0.09872502664273436,
+   "liu2021": 0.10204364359378815
+  },
+  "train_loss": 0.09955113330483437,
+  "sec": 382.2
+ },
+ {
+  "step": 9250,
+  "inner_val_loss": 0.09969991944128215,
+  "inner_val_per_cohort": {
+   "immicro": 0.09834584624816974,
+   "imperial_eskd": 0.09871771254322746,
+   "liu2021": 0.10203619953244925
+  },
+  "train_loss": 0.0995605261027813,
+  "sec": 392.6
+ },
+ {
+  "step": 9500,
+  "inner_val_loss": 0.09969683978330306,
+  "inner_val_per_cohort": {
+   "immicro": 0.09833913141240676,
+   "imperial_eskd": 0.09871826083822684,
+   "liu2021": 0.10203312709927559
+  },
+  "train_loss": 0.09947743394970894,
+  "sec": 403.0
+ },
+ {
+  "step": 9750,
+  "inner_val_loss": 0.09969315957278013,
+  "inner_val_per_cohort": {
+   "immicro": 0.09833037573844194,
+   "imperial_eskd": 0.09871664270758629,
+   "liu2021": 0.10203246027231216
+  },
+  "train_loss": 0.09951701188087464,
+  "sec": 413.4
+ },
+ {
+  "step": 10000,
+  "inner_val_loss": 0.09969036514409864,
+  "inner_val_per_cohort": {
+   "immicro": 0.09832167749603589,
+   "imperial_eskd": 0.09871740097349341,
+   "liu2021": 0.10203201696276665
+  },
+  "train_loss": 0.09944971358776092,
+  "sec": 423.9
+ },
+ {
+  "step": 10250,
+  "inner_val_loss": 0.0996825485725445,
+  "inner_val_per_cohort": {
+   "immicro": 0.09831090333561103,
+   "imperial_eskd": 0.09870920601216229,
+   "liu2021": 0.10202753636986017
+  },
+  "train_loss": 0.09960869258642197,
+  "sec": 434.3
+ },
+ {
+  "step": 10500,
+  "inner_val_loss": 0.09967935987456579,
+  "inner_val_per_cohort": {
+   "immicro": 0.09830323824038108,
+   "imperial_eskd": 0.09870820560238579,
+   "liu2021": 0.10202663578093052
+  },
+  "train_loss": 0.09963642501831055,
+  "sec": 444.6
+ },
+ {
+  "step": 10750,
+  "inner_val_loss": 0.09967945546212824,
+  "inner_val_per_cohort": {
+   "immicro": 0.09829776299496491,
+   "imperial_eskd": 0.09871302883733403,
+   "liu2021": 0.10202757455408573
+  },
+  "train_loss": 0.09946285966038704,
+  "sec": 455.0
+ },
+ {
+  "step": 11000,
+  "inner_val_loss": 0.09967610754561845,
+  "inner_val_per_cohort": {
+   "immicro": 0.09828899490336578,
+   "imperial_eskd": 0.09871583702889355,
+   "liu2021": 0.10202349070459604
+  },
+  "train_loss": 0.09950033336877823,
+  "sec": 465.4
+ },
+ {
+  "step": 11250,
+  "inner_val_loss": 0.0996686208692163,
+  "inner_val_per_cohort": {
+   "immicro": 0.09828084024290244,
+   "imperial_eskd": 0.09870653193105351,
+   "liu2021": 0.10201849043369293
+  },
+  "train_loss": 0.0993753208220005,
+  "sec": 475.9
+ },
+ {
+  "step": 11500,
+  "inner_val_loss": 0.09966325551986394,
+  "inner_val_per_cohort": {
+   "immicro": 0.0982746339092652,
+   "imperial_eskd": 0.09870057146657597,
+   "liu2021": 0.10201456118375063
+  },
+  "train_loss": 0.09946013849973678,
+  "sec": 486.4
+ },
+ {
+  "step": 11750,
+  "inner_val_loss": 0.09965910831221729,
+  "inner_val_per_cohort": {
+   "immicro": 0.09826936076084773,
+   "imperial_eskd": 0.09869798137383028,
+   "liu2021": 0.10200998280197382
+  },
+  "train_loss": 0.09935733038187027,
+  "sec": 496.9
+ },
+ {
+  "step": 12000,
+  "inner_val_loss": 0.09965059354976573,
+  "inner_val_per_cohort": {
+   "immicro": 0.09825734173258145,
+   "imperial_eskd": 0.0986866642805663,
+   "liu2021": 0.1020077746361494
+  },
+  "train_loss": 0.09916825929284095,
+  "sec": 507.2
+ },
+ {
+  "step": 12250,
+  "inner_val_loss": 0.09963928670105006,
+  "inner_val_per_cohort": {
+   "immicro": 0.09824763083209594,
+   "imperial_eskd": 0.09867751598358154,
+   "liu2021": 0.10199271328747272
+  },
+  "train_loss": 0.09933510613441467,
+  "sec": 517.6
+ },
+ {
+  "step": 12500,
+  "inner_val_loss": 0.0996344072386773,
+  "inner_val_per_cohort": {
+   "immicro": 0.09824000174800555,
+   "imperial_eskd": 0.09867345338517969,
+   "liu2021": 0.10198976658284664
+  },
+  "train_loss": 0.09950089830160141,
+  "sec": 527.9
+ },
+ {
+  "step": 12750,
+  "inner_val_loss": 0.09962613017044285,
+  "inner_val_per_cohort": {
+   "immicro": 0.09823012538254261,
+   "imperial_eskd": 0.0986643900925463,
+   "liu2021": 0.10198387503623962
+  },
+  "train_loss": 0.09943969869613647,
+  "sec": 538.3
+ },
+ {
+  "step": 13000,
+  "inner_val_loss": 0.09961980148547828,
+  "inner_val_per_cohort": {
+   "immicro": 0.09822151561578114,
+   "imperial_eskd": 0.09866493263027885,
+   "liu2021": 0.10197295621037483
+  },
+  "train_loss": 0.09951297789812089,
+  "sec": 548.8
+ },
+ {
+  "step": 13250,
+  "inner_val_loss": 0.09961521937841117,
+  "inner_val_per_cohort": {
+   "immicro": 0.09821235047032435,
+   "imperial_eskd": 0.09866002269766548,
+   "liu2021": 0.10197328496724367
+  },
+  "train_loss": 0.09950392726063728,
+  "sec": 559.3
+ },
+ {
+  "step": 13500,
+  "inner_val_loss": 0.09961062080857126,
+  "inner_val_per_cohort": {
+   "immicro": 0.0982035684088866,
+   "imperial_eskd": 0.09865671430121768,
+   "liu2021": 0.10197157971560955
+  },
+  "train_loss": 0.09935772740840912,
+  "sec": 569.7
+ },
+ {
+  "step": 13750,
+  "inner_val_loss": 0.09960528922201406,
+  "inner_val_per_cohort": {
+   "immicro": 0.09819372774412234,
+   "imperial_eskd": 0.09865068407221274,
+   "liu2021": 0.10197145584970713
+  },
+  "train_loss": 0.09944565895199775,
+  "sec": 580.1
+ },
+ {
+  "step": 14000,
+  "inner_val_loss": 0.09960106923009711,
+  "inner_val_per_cohort": {
+   "immicro": 0.09818459705760081,
+   "imperial_eskd": 0.09865037622776898,
+   "liu2021": 0.10196823440492153
+  },
+  "train_loss": 0.0992985166311264,
+  "sec": 590.3
+ },
+ {
+  "step": 14250,
+  "inner_val_loss": 0.09959160736903096,
+  "inner_val_per_cohort": {
+   "immicro": 0.09817402344197035,
+   "imperial_eskd": 0.09864599393172697,
+   "liu2021": 0.10195480473339558
+  },
+  "train_loss": 0.09919915196299553,
+  "sec": 600.7
+ },
+ {
+  "step": 14500,
+  "inner_val_loss": 0.09958825715714031,
+  "inner_val_per_cohort": {
+   "immicro": 0.09816848828146855,
+   "imperial_eskd": 0.09864668548107147,
+   "liu2021": 0.1019495977088809
+  },
+  "train_loss": 0.09953417009115219,
+  "sec": 611.2
+ },
+ {
+  "step": 14750,
+  "inner_val_loss": 0.0995865830154431,
+  "inner_val_per_cohort": {
+   "immicro": 0.0981625368197759,
+   "imperial_eskd": 0.09864760427312418,
+   "liu2021": 0.10194960795342922
+  },
+  "train_loss": 0.09940882018208504,
+  "sec": 621.7
+ },
+ {
+  "step": 15000,
+  "inner_val_loss": 0.09958488608926835,
+  "inner_val_per_cohort": {
+   "immicro": 0.09816165392597516,
+   "imperial_eskd": 0.0986476990986954,
+   "liu2021": 0.1019453052431345
+  },
+  "train_loss": 0.09948887386918068,
+  "sec": 632.3
+ },
+ {
+  "step": 15250,
+  "inner_val_loss": 0.09958014113245288,
+  "inner_val_per_cohort": {
+   "immicro": 0.09815801742176215,
+   "imperial_eskd": 0.09864577041430907,
+   "liu2021": 0.1019366355612874
+  },
+  "train_loss": 0.09925769266486167,
+  "sec": 642.8
+ },
+ {
+  "step": 15500,
+  "inner_val_loss": 0.09957331853608291,
+  "inner_val_per_cohort": {
+   "immicro": 0.09814999252557755,
+   "imperial_eskd": 0.09863975644111633,
+   "liu2021": 0.10193020664155483
+  },
+  "train_loss": 0.09946393185853958,
+  "sec": 653.3
+ },
+ {
+  "step": 15750,
+  "inner_val_loss": 0.09956708879708642,
+  "inner_val_per_cohort": {
+   "immicro": 0.09814226751526196,
+   "imperial_eskd": 0.09863144836642525,
+   "liu2021": 0.10192755050957203
+  },
+  "train_loss": 0.09920621544122696,
+  "sec": 663.9
+ },
+ {
+  "step": 16000,
+  "inner_val_loss": 0.09956246233460579,
+  "inner_val_per_cohort": {
+   "immicro": 0.09813480544835329,
+   "imperial_eskd": 0.09863337197087028,
+   "liu2021": 0.10191920958459377
+  },
+  "train_loss": 0.0992486170232296,
+  "sec": 674.5
+ },
+ {
+  "step": 16250,
+  "inner_val_loss": 0.09955805192014784,
+  "inner_val_per_cohort": {
+   "immicro": 0.0981265048806866,
+   "imperial_eskd": 0.0986347808079286,
+   "liu2021": 0.10191287007182837
+  },
+  "train_loss": 0.09935080111026764,
+  "sec": 684.8
+ },
+ {
+  "step": 16500,
+  "inner_val_loss": 0.09955382875797121,
+  "inner_val_per_cohort": {
+   "immicro": 0.09811664714167516,
+   "imperial_eskd": 0.09863792766224254,
+   "liu2021": 0.10190691146999598
+  },
+  "train_loss": 0.09947746589779854,
+  "sec": 695.3
+ },
+ {
+  "step": 16750,
+  "inner_val_loss": 0.09954899039608661,
+  "inner_val_per_cohort": {
+   "immicro": 0.09810925989101331,
+   "imperial_eskd": 0.09863896938887509,
+   "liu2021": 0.10189874190837145
+  },
+  "train_loss": 0.09916770857572556,
+  "sec": 707.2
+ },
+ {
+  "step": 17000,
+  "inner_val_loss": 0.0995460928069672,
+  "inner_val_per_cohort": {
+   "immicro": 0.09810482834776242,
+   "imperial_eskd": 0.09863668815656142,
+   "liu2021": 0.10189676191657782
+  },
+  "train_loss": 0.0994631154537201,
+  "sec": 719.1
+ },
+ {
+  "step": 17250,
+  "inner_val_loss": 0.09954268551158785,
+  "inner_val_per_cohort": {
+   "immicro": 0.09809983676920335,
+   "imperial_eskd": 0.0986415720121427,
+   "liu2021": 0.10188664775341749
+  },
+  "train_loss": 0.09954756939411163,
+  "sec": 731.0
+ },
+ {
+  "step": 17500,
+  "inner_val_loss": 0.09953690065315575,
+  "inner_val_per_cohort": {
+   "immicro": 0.09809270439048608,
+   "imperial_eskd": 0.09863934428854422,
+   "liu2021": 0.10187865328043699
+  },
+  "train_loss": 0.09939650228619576,
+  "sec": 743.1
+ },
+ {
+  "step": 17750,
+  "inner_val_loss": 0.0995316659994047,
+  "inner_val_per_cohort": {
+   "immicro": 0.09808750015993913,
+   "imperial_eskd": 0.09863353656096892,
+   "liu2021": 0.10187396127730608
+  },
+  "train_loss": 0.09909506851434707,
+  "sec": 755.2
+ },
+ {
+  "step": 18000,
+  "inner_val_loss": 0.09953084043368245,
+  "inner_val_per_cohort": {
+   "immicro": 0.09808286372572184,
+   "imperial_eskd": 0.09863693030043082,
+   "liu2021": 0.10187272727489471
+  },
+  "train_loss": 0.0996696857213974,
+  "sec": 767.1
+ },
+ {
+  "step": 18250,
+  "inner_val_loss": 0.09952660424239708,
+  "inner_val_per_cohort": {
+   "immicro": 0.09808078967034817,
+   "imperial_eskd": 0.0986333347179673,
+   "liu2021": 0.10186568833887577
+  },
+  "train_loss": 0.09910059842467309,
+  "sec": 779.0
+ },
+ {
+  "step": 18500,
+  "inner_val_loss": 0.09952198269052638,
+  "inner_val_per_cohort": {
+   "immicro": 0.098074101532499,
+   "imperial_eskd": 0.09863057732582092,
+   "liu2021": 0.10186126921325922
+  },
+  "train_loss": 0.09939756116271019,
+  "sec": 791.1
+ },
+ {
+  "step": 18750,
+  "inner_val_loss": 0.09951426100098726,
+  "inner_val_per_cohort": {
+   "immicro": 0.09806749876588583,
+   "imperial_eskd": 0.09862439740787853,
+   "liu2021": 0.1018508868291974
+  },
+  "train_loss": 0.09918195790052414,
+  "sec": 803.0
+ },
+ {
+  "step": 19000,
+  "inner_val_loss": 0.09950775958861065,
+  "inner_val_per_cohort": {
+   "immicro": 0.09806019129852454,
+   "imperial_eskd": 0.09861610287969763,
+   "liu2021": 0.10184698458760977
+  },
+  "train_loss": 0.09929327410459518,
+  "sec": 813.4
+ },
+ {
+  "step": 19250,
+  "inner_val_loss": 0.09950298863002145,
+  "inner_val_per_cohort": {
+   "immicro": 0.09805366800477107,
+   "imperial_eskd": 0.09861622208898718,
+   "liu2021": 0.10183907579630613
+  },
+  "train_loss": 0.09919308805465699,
+  "sec": 825.0
+ },
+ {
+  "step": 19500,
+  "inner_val_loss": 0.09949899672746959,
+  "inner_val_per_cohort": {
+   "immicro": 0.0980461568882068,
+   "imperial_eskd": 0.09861674193631519,
+   "liu2021": 0.10183409135788679
+  },
+  "train_loss": 0.09918284976482392,
+  "sec": 837.0
+ },
+ {
+  "step": 19750,
+  "inner_val_loss": 0.09949418988003574,
+  "inner_val_per_cohort": {
+   "immicro": 0.09803942373643319,
+   "imperial_eskd": 0.09861331534656612,
+   "liu2021": 0.10182983055710793
+  },
+  "train_loss": 0.09910680323839188,
+  "sec": 849.1
+ },
+ {
+  "step": 20000,
+  "inner_val_loss": 0.09948734277765257,
+  "inner_val_per_cohort": {
+   "immicro": 0.09802971904476483,
+   "imperial_eskd": 0.0986113961447369,
+   "liu2021": 0.10182091314345598
+  },
+  "train_loss": 0.0993082580268383,
+  "sec": 859.8
+ },
+ {
+  "step": 20250,
+  "inner_val_loss": 0.09948586879505052,
+  "inner_val_per_cohort": {
+   "immicro": 0.09802489603559177,
+   "imperial_eskd": 0.0986083373427391,
+   "liu2021": 0.10182437300682068
+  },
+  "train_loss": 0.09930938351154327,
+  "sec": 872.2
+ },
+ {
+  "step": 20500,
+  "inner_val_loss": 0.09948031835709557,
+  "inner_val_per_cohort": {
+   "immicro": 0.09801554027944803,
+   "imperial_eskd": 0.0986083945767446,
+   "liu2021": 0.10181702021509409
+  },
+  "train_loss": 0.09896220970153809,
+  "sec": 882.7
+ },
+ {
+  "step": 20750,
+  "inner_val_loss": 0.09947731795560832,
+  "inner_val_per_cohort": {
+   "immicro": 0.0980114337677757,
+   "imperial_eskd": 0.09860677068883722,
+   "liu2021": 0.10181374941021204
+  },
+  "train_loss": 0.0991942221224308,
+  "sec": 893.0
+ },
+ {
+  "step": 21000,
+  "inner_val_loss": 0.09947251800372443,
+  "inner_val_per_cohort": {
+   "immicro": 0.09800740486631791,
+   "imperial_eskd": 0.09860613298687068,
+   "liu2021": 0.10180401615798473
+  },
+  "train_loss": 0.09934365773200989,
+  "sec": 903.4
+ },
+ {
+  "step": 21250,
+  "inner_val_loss": 0.09946806790928046,
+  "inner_val_per_cohort": {
+   "immicro": 0.09800160024315119,
+   "imperial_eskd": 0.09860232472419739,
+   "liu2021": 0.1018002787604928
+  },
+  "train_loss": 0.09877656608819961,
+  "sec": 913.9
+ },
+ {
+  "step": 21500,
+  "inner_val_loss": 0.09946146851048025,
+  "inner_val_per_cohort": {
+   "immicro": 0.09799354585508506,
+   "imperial_eskd": 0.09859578887170012,
+   "liu2021": 0.10179507080465555
+  },
+  "train_loss": 0.09902521017193794,
+  "sec": 924.4
+ },
+ {
+  "step": 21750,
+  "inner_val_loss": 0.09945632735587129,
+  "inner_val_per_cohort": {
+   "immicro": 0.0979912771532933,
+   "imperial_eskd": 0.09859123114835132,
+   "liu2021": 0.10178647376596928
+  },
+  "train_loss": 0.09934761926531792,
+  "sec": 934.9
+ },
+ {
+  "step": 22000,
+  "inner_val_loss": 0.09945220065613587,
+  "inner_val_per_cohort": {
+   "immicro": 0.09798451606184244,
+   "imperial_eskd": 0.09859200939536095,
+   "liu2021": 0.10178007651120424
+  },
+  "train_loss": 0.0992635957300663,
+  "sec": 945.2
+ },
+ {
+  "step": 22250,
+  "inner_val_loss": 0.09945068630681496,
+  "inner_val_per_cohort": {
+   "immicro": 0.09798089104394118,
+   "imperial_eskd": 0.09858499399640343,
+   "liu2021": 0.10178617388010025
+  },
+  "train_loss": 0.0991807531118393,
+  "sec": 955.7
+ },
+ {
+  "step": 22500,
+  "inner_val_loss": 0.09944832436928545,
+  "inner_val_per_cohort": {
+   "immicro": 0.0979754909252127,
+   "imperial_eskd": 0.09858833930709145,
+   "liu2021": 0.10178114287555218
+  },
+  "train_loss": 0.09914993885159493,
+  "sec": 966.3
+ },
+ {
+  "step": 22750,
+  "inner_val_loss": 0.09944552751352088,
+  "inner_val_per_cohort": {
+   "immicro": 0.09796964066723983,
+   "imperial_eskd": 0.09858348931778561,
+   "liu2021": 0.10178345255553722
+  },
+  "train_loss": 0.09922633677721024,
+  "sec": 976.8
+ },
+ {
+  "step": 23000,
+  "inner_val_loss": 0.09944343077714997,
+  "inner_val_per_cohort": {
+   "immicro": 0.09796551925440629,
+   "imperial_eskd": 0.09858719902959737,
+   "liu2021": 0.10177757404744625
+  },
+  "train_loss": 0.0995127004981041,
+  "sec": 987.2
+ },
+ {
+  "step": 23250,
+  "inner_val_loss": 0.09943665297833658,
+  "inner_val_per_cohort": {
+   "immicro": 0.0979584048812588,
+   "imperial_eskd": 0.09858134863051501,
+   "liu2021": 0.1017702054232359
+  },
+  "train_loss": 0.09934223687648773,
+  "sec": 997.7
+ },
+ {
+  "step": 23500,
+  "inner_val_loss": 0.09943218016526616,
+  "inner_val_per_cohort": {
+   "immicro": 0.09795270177225272,
+   "imperial_eskd": 0.09858120029622858,
+   "liu2021": 0.10176263842731714
+  },
+  "train_loss": 0.0993885167837143,
+  "sec": 1008.1
+ },
+ {
+  "step": 23750,
+  "inner_val_loss": 0.0994282720253022,
+  "inner_val_per_cohort": {
+   "immicro": 0.0979438719029228,
+   "imperial_eskd": 0.0985847616737539,
+   "liu2021": 0.10175618249922991
+  },
+  "train_loss": 0.09901119774580001,
+  "sec": 1018.3
+ },
+ {
+  "step": 24000,
+  "inner_val_loss": 0.09942546518839368,
+  "inner_val_per_cohort": {
+   "immicro": 0.09794353197018306,
+   "imperial_eskd": 0.09857995469461787,
+   "liu2021": 0.10175290890038013
+  },
+  "train_loss": 0.09947043040394783,
+  "sec": 1028.6
+ },
+ {
+  "step": 24250,
+  "inner_val_loss": 0.09941951851501608,
+  "inner_val_per_cohort": {
+   "immicro": 0.0979406600818038,
+   "imperial_eskd": 0.09856920824809508,
+   "liu2021": 0.1017486872151494
+  },
+  "train_loss": 0.09896109953522682,
+  "sec": 1038.9
+ },
+ {
+  "step": 24500,
+  "inner_val_loss": 0.09941576741108991,
+  "inner_val_per_cohort": {
+   "immicro": 0.09793770095954339,
+   "imperial_eskd": 0.09856476973403584,
+   "liu2021": 0.1017448315396905
+  },
+  "train_loss": 0.09891128227114677,
+  "sec": 1049.4
+ },
+ {
+  "step": 24750,
+  "inner_val_loss": 0.09941161414514287,
+  "inner_val_per_cohort": {
+   "immicro": 0.09793360934903224,
+   "imperial_eskd": 0.09856211055408824,
+   "liu2021": 0.1017391225323081
+  },
+  "train_loss": 0.09923951354622841,
+  "sec": 1059.7
+ },
+ {
+  "step": 25000,
+  "inner_val_loss": 0.09941053098876669,
+  "inner_val_per_cohort": {
+   "immicro": 0.09792589613546927,
+   "imperial_eskd": 0.09856514564969322,
+   "liu2021": 0.10174055118113756
+  },
+  "train_loss": 0.09904820841550827,
+  "sec": 1070.1
+ },
+ {
+  "step": 25250,
+  "inner_val_loss": 0.09940497362703989,
+  "inner_val_per_cohort": {
+   "immicro": 0.0979156931862235,
+   "imperial_eskd": 0.0985598201778802,
+   "liu2021": 0.10173940751701593
+  },
+  "train_loss": 0.09917536532878876,
+  "sec": 1080.6
+ },
+ {
+  "step": 25500,
+  "inner_val_loss": 0.0994048152363511,
+  "inner_val_per_cohort": {
+   "immicro": 0.09791059264292319,
+   "imperial_eskd": 0.09856709702448412,
+   "liu2021": 0.101736756041646
+  },
+  "train_loss": 0.09898550972342492,
+  "sec": 1090.9
+ },
+ {
+  "step": 25750,
+  "inner_val_loss": 0.09940233652602241,
+  "inner_val_per_cohort": {
+   "immicro": 0.09790635822961728,
+   "imperial_eskd": 0.09857056357643822,
+   "liu2021": 0.10173008777201176
+  },
+  "train_loss": 0.09921356397867202,
+  "sec": 1101.3
+ },
+ {
+  "step": 26000,
+  "inner_val_loss": 0.09939615393640718,
+  "inner_val_per_cohort": {
+   "immicro": 0.09790141818424065,
+   "imperial_eskd": 0.09856589003042741,
+   "liu2021": 0.10172115359455347
+  },
+  "train_loss": 0.09921703600883484,
+  "sec": 1111.6
+ },
+ {
+  "step": 26250,
+  "inner_val_loss": 0.09939450069772776,
+  "inner_val_per_cohort": {
+   "immicro": 0.0978979238619407,
+   "imperial_eskd": 0.09856893596324054,
+   "liu2021": 0.10171664226800203
+  },
+  "train_loss": 0.09909400057792664,
+  "sec": 1122.2
+ },
+ {
+  "step": 26500,
+  "inner_val_loss": 0.09939339353392522,
+  "inner_val_per_cohort": {
+   "immicro": 0.09789521247148514,
+   "imperial_eskd": 0.09856705367565155,
+   "liu2021": 0.10171791445463896
+  },
+  "train_loss": 0.09926339384913445,
+  "sec": 1132.6
+ },
+ {
+  "step": 26750,
+  "inner_val_loss": 0.09939015163767217,
+  "inner_val_per_cohort": {
+   "immicro": 0.09788955344508092,
+   "imperial_eskd": 0.09856496886773543,
+   "liu2021": 0.10171593260020018
+  },
+  "train_loss": 0.09880001661181449,
+  "sec": 1143.1
+ },
+ {
+  "step": 27000,
+  "inner_val_loss": 0.09938793762050795,
+  "inner_val_per_cohort": {
+   "immicro": 0.09788646269589663,
+   "imperial_eskd": 0.09856554425575516,
+   "liu2021": 0.10171180590987206
+  },
+  "train_loss": 0.09922973313927651,
+  "sec": 1153.5
+ },
+ {
+  "step": 27250,
+  "inner_val_loss": 0.09938393832381927,
+  "inner_val_per_cohort": {
+   "immicro": 0.09788307081907988,
+   "imperial_eskd": 0.09856181185353886,
+   "liu2021": 0.10170693229883909
+  },
+  "train_loss": 0.09929617580771447,
+  "sec": 1163.7
+ },
+ {
+  "step": 27500,
+  "inner_val_loss": 0.09938154324437633,
+  "inner_val_per_cohort": {
+   "immicro": 0.09787959698587656,
+   "imperial_eskd": 0.09855941175059839,
+   "liu2021": 0.10170562099665403
+  },
+  "train_loss": 0.09918298110365868,
+  "sec": 1174.1
+ },
+ {
+  "step": 27750,
+  "inner_val_loss": 0.09937542111547004,
+  "inner_val_per_cohort": {
+   "immicro": 0.09787333322068055,
+   "imperial_eskd": 0.09854933382435278,
+   "liu2021": 0.10170359630137682
+  },
+  "train_loss": 0.09928907465934754,
+  "sec": 1184.5
+ },
+ {
+  "step": 28000,
+  "inner_val_loss": 0.09937516129528634,
+  "inner_val_per_cohort": {
+   "immicro": 0.09787378429124753,
+   "imperial_eskd": 0.09855162555521185,
+   "liu2021": 0.10170007403939962
+  },
+  "train_loss": 0.09900567069649696,
+  "sec": 1194.9
+ },
+ {
+  "step": 28250,
+  "inner_val_loss": 0.09937044535058015,
+  "inner_val_per_cohort": {
+   "immicro": 0.09786721784621477,
+   "imperial_eskd": 0.09855217554352501,
+   "liu2021": 0.10169194266200066
+  },
+  "train_loss": 0.09908906435966491,
+  "sec": 1205.2
+ },
+ {
+  "step": 28500,
+  "inner_val_loss": 0.09936470890210734,
+  "inner_val_per_cohort": {
+   "immicro": 0.09786185466994841,
+   "imperial_eskd": 0.09854438900947571,
+   "liu2021": 0.10168788302689791
+  },
+  "train_loss": 0.09874538001418114,
+  "sec": 1215.8
+ },
+ {
+  "step": 28750,
+  "inner_val_loss": 0.0993581624569917,
+  "inner_val_per_cohort": {
+   "immicro": 0.09785419298956792,
+   "imperial_eskd": 0.09854104844006625,
+   "liu2021": 0.10167924594134092
+  },
+  "train_loss": 0.09896869030594825,
+  "sec": 1226.1
+ },
+ {
+  "step": 29000,
+  "inner_val_loss": 0.09935515945909,
+  "inner_val_per_cohort": {
+   "immicro": 0.09784995423008998,
+   "imperial_eskd": 0.09853472289713947,
+   "liu2021": 0.10168080125004053
+  },
+  "train_loss": 0.09896306955814362,
+  "sec": 1236.4
+ },
+ {
+  "step": 29250,
+  "inner_val_loss": 0.09935076578962382,
+  "inner_val_per_cohort": {
+   "immicro": 0.09784474068631728,
+   "imperial_eskd": 0.09852850911292163,
+   "liu2021": 0.10167904756963253
+  },
+  "train_loss": 0.09940092530846596,
+  "sec": 1246.9
+ },
+ {
+  "step": 29500,
+  "inner_val_loss": 0.09934828510376238,
+  "inner_val_per_cohort": {
+   "immicro": 0.09783962896714608,
+   "imperial_eskd": 0.09852382066574963,
+   "liu2021": 0.10168140567839146
+  },
+  "train_loss": 0.09870317733287812,
+  "sec": 1257.6
+ },
+ {
+  "step": 29750,
+  "inner_val_loss": 0.09934431924061342,
+  "inner_val_per_cohort": {
+   "immicro": 0.0978349894285202,
+   "imperial_eskd": 0.0985198315571655,
+   "liu2021": 0.10167813673615456
+  },
+  "train_loss": 0.09897947481274605,
+  "sec": 1269.5
+ },
+ {
+  "step": 30000,
+  "inner_val_loss": 0.09934141224419529,
+  "inner_val_per_cohort": {
+   "immicro": 0.09783248882740736,
+   "imperial_eskd": 0.09852027012543245,
+   "liu2021": 0.10167147777974606
+  },
+  "train_loss": 0.09870183727145194,
+  "sec": 1281.2
+ },
+ {
+  "step": 30250,
+  "inner_val_loss": 0.09933990139438949,
+  "inner_val_per_cohort": {
+   "immicro": 0.0978287026907007,
+   "imperial_eskd": 0.09852358800443736,
+   "liu2021": 0.10166741348803043
+  },
+  "train_loss": 0.09896448317170144,
+  "sec": 1293.2
+ },
+ {
+  "step": 30500,
+  "inner_val_loss": 0.09933794388603984,
+  "inner_val_per_cohort": {
+   "immicro": 0.09782554674893618,
+   "imperial_eskd": 0.09852322766726668,
+   "liu2021": 0.10166505724191666
+  },
+  "train_loss": 0.09899495866894722,
+  "sec": 1305.0
+ },
+ {
+  "step": 30750,
+  "inner_val_loss": 0.09933425873668507,
+  "inner_val_per_cohort": {
+   "immicro": 0.0978177400926749,
+   "imperial_eskd": 0.09852115234190767,
+   "liu2021": 0.10166388377547264
+  },
+  "train_loss": 0.09898275259137154,
+  "sec": 1316.8
+ },
+ {
+  "step": 31000,
+  "inner_val_loss": 0.09932816862081639,
+  "inner_val_per_cohort": {
+   "immicro": 0.0978074815745155,
+   "imperial_eskd": 0.09852251986210997,
+   "liu2021": 0.10165450442582369
+  },
+  "train_loss": 0.09862361484766007,
+  "sec": 1328.6
+ },
+ {
+  "step": 31250,
+  "inner_val_loss": 0.0993244137915999,
+  "inner_val_per_cohort": {
+   "immicro": 0.09779886932422717,
+   "imperial_eskd": 0.09852001849900592,
+   "liu2021": 0.1016543535515666
+  },
+  "train_loss": 0.09895282217860221,
+  "sec": 1340.3
+ },
+ {
+  "step": 31500,
+  "inner_val_loss": 0.09932184333188666,
+  "inner_val_per_cohort": {
+   "immicro": 0.0977964863801996,
+   "imperial_eskd": 0.09851717203855515,
+   "liu2021": 0.10165187157690525
+  },
+  "train_loss": 0.09869569772481919,
+  "sec": 1352.4
+ },
+ {
+  "step": 31750,
+  "inner_val_loss": 0.09931788083655063,
+  "inner_val_per_cohort": {
+   "immicro": 0.09779121602574985,
+   "imperial_eskd": 0.09851476075974377,
+   "liu2021": 0.10164766572415829
+  },
+  "train_loss": 0.09904776412248611,
+  "sec": 1364.3
+ },
+ {
+  "step": 32000,
+  "inner_val_loss": 0.09931599507766842,
+  "inner_val_per_cohort": {
+   "immicro": 0.0977865547562639,
+   "imperial_eskd": 0.0985180600122972,
+   "liu2021": 0.10164337046444416
+  },
+  "train_loss": 0.09901578316092491,
+  "sec": 1376.3
+ },
+ {
+  "step": 32250,
+  "inner_val_loss": 0.09931092645333271,
+  "inner_val_per_cohort": {
+   "immicro": 0.09778360432634751,
+   "imperial_eskd": 0.09851346469738266,
+   "liu2021": 0.10163571033626795
+  },
+  "train_loss": 0.09890797951817512,
+  "sec": 1388.3
+ },
+ {
+  "step": 32500,
+  "inner_val_loss": 0.09931160159634822,
+  "inner_val_per_cohort": {
+   "immicro": 0.09777931403368711,
+   "imperial_eskd": 0.09852011196992615,
+   "liu2021": 0.10163537878543139
+  },
+  "train_loss": 0.09879304331541061,
+  "sec": 1400.5
+ },
+ {
+  "step": 32750,
+  "inner_val_loss": 0.09930464670513615,
+  "inner_val_per_cohort": {
+   "immicro": 0.09777563810348511,
+   "imperial_eskd": 0.09850680591030554,
+   "liu2021": 0.10163149610161781
+  },
+  "train_loss": 0.09908109718561173,
+  "sec": 1412.4
+ },
+ {
+  "step": 33000,
+  "inner_val_loss": 0.09929824928103974,
+  "inner_val_per_cohort": {
+   "immicro": 0.09777248464524746,
+   "imperial_eskd": 0.09849728914824399,
+   "liu2021": 0.10162497404962778
+  },
+  "train_loss": 0.09883525639772416,
+  "sec": 1424.1
+ },
+ {
+  "step": 33250,
+  "inner_val_loss": 0.09929673671910559,
+  "inner_val_per_cohort": {
+   "immicro": 0.09776455846925576,
+   "imperial_eskd": 0.09849265522577545,
+   "liu2021": 0.10163299646228552
+  },
+  "train_loss": 0.09892024824023246,
+  "sec": 1435.9
+ },
+ {
+  "step": 33500,
+  "inner_val_loss": 0.0992907484184311,
+  "inner_val_per_cohort": {
+   "immicro": 0.09775580434749524,
+   "imperial_eskd": 0.0984879464588382,
+   "liu2021": 0.10162849444895983
+  },
+  "train_loss": 0.09849854728579521,
+  "sec": 1447.6
+ },
+ {
+  "step": 33750,
+  "inner_val_loss": 0.09928515508319392,
+  "inner_val_per_cohort": {
+   "immicro": 0.09775135293602943,
+   "imperial_eskd": 0.09847978739575906,
+   "liu2021": 0.10162432491779327
+  },
+  "train_loss": 0.09887439626455306,
+  "sec": 1459.4
+ },
+ {
+  "step": 34000,
+  "inner_val_loss": 0.09928189617854477,
+  "inner_val_per_cohort": {
+   "immicro": 0.09774434131880601,
+   "imperial_eskd": 0.0984826148910956,
+   "liu2021": 0.10161873232573271
+  },
+  "train_loss": 0.09906945377588272,
+  "sec": 1471.1
+ },
+ {
+  "step": 34250,
+  "inner_val_loss": 0.09927879166648244,
+  "inner_val_per_cohort": {
+   "immicro": 0.09773915447294712,
+   "imperial_eskd": 0.09848030148582025,
+   "liu2021": 0.10161691904067993
+  },
+  "train_loss": 0.09904856216907501,
+  "sec": 1483.0
+ },
+ {
+  "step": 34500,
+  "inner_val_loss": 0.0992772639964265,
+  "inner_val_per_cohort": {
+   "immicro": 0.0977344053486983,
+   "imperial_eskd": 0.09847925501790913,
+   "liu2021": 0.10161813162267208
+  },
+  "train_loss": 0.0990587807893753,
+  "sec": 1494.8
+ },
+ {
+  "step": 34750,
+  "inner_val_loss": 0.09927248769449164,
+  "inner_val_per_cohort": {
+   "immicro": 0.09773143629233043,
+   "imperial_eskd": 0.09847581792961467,
+   "liu2021": 0.10161020886152983
+  },
+  "train_loss": 0.09908203402161599,
+  "sec": 1506.5
+ },
+ {
+  "step": 35000,
+  "inner_val_loss": 0.09926872006194158,
+  "inner_val_per_cohort": {
+   "immicro": 0.0977262631058693,
+   "imperial_eskd": 0.09847140650857579,
+   "liu2021": 0.10160849057137966
+  },
+  "train_loss": 0.09904912787675857,
+  "sec": 1518.4
+ },
+ {
+  "step": 35250,
+  "inner_val_loss": 0.09926990183503037,
+  "inner_val_per_cohort": {
+   "immicro": 0.09772610353926818,
+   "imperial_eskd": 0.09847752445123413,
+   "liu2021": 0.10160607751458883
+  },
+  "train_loss": 0.09902628248929977,
+  "sec": 1530.2
+ },
+ {
+  "step": 35500,
+  "inner_val_loss": 0.09926755382030299,
+  "inner_val_per_cohort": {
+   "immicro": 0.09772276040166616,
+   "imperial_eskd": 0.09847286241975697,
+   "liu2021": 0.10160703863948584
+  },
+  "train_loss": 0.09917266005277633,
+  "sec": 1541.9
+ },
+ {
+  "step": 35750,
+  "inner_val_loss": 0.09926662124655765,
+  "inner_val_per_cohort": {
+   "immicro": 0.09771654692788918,
+   "imperial_eskd": 0.09847862510518594,
+   "liu2021": 0.1016046917065978
+  },
+  "train_loss": 0.09885487422347068,
+  "sec": 1553.7
+ },
+ {
+  "step": 36000,
+  "inner_val_loss": 0.09926429701348145,
+  "inner_val_per_cohort": {
+   "immicro": 0.09771231282502413,
+   "imperial_eskd": 0.09847958013415337,
+   "liu2021": 0.10160099808126688
+  },
+  "train_loss": 0.09899783197045327,
+  "sec": 1565.4
+ },
+ {
+  "step": 36250,
+  "inner_val_loss": 0.0992646421390501,
+  "inner_val_per_cohort": {
+   "immicro": 0.09771223925054073,
+   "imperial_eskd": 0.09847677194259384,
+   "liu2021": 0.10160491522401571
+  },
+  "train_loss": 0.09902246797084809,
+  "sec": 1577.1
+ },
+ {
+  "step": 36500,
+  "inner_val_loss": 0.09926129385565567,
+  "inner_val_per_cohort": {
+   "immicro": 0.09771480535467465,
+   "imperial_eskd": 0.09847315197641199,
+   "liu2021": 0.10159592423588037
+  },
+  "train_loss": 0.09859582942724228,
+  "sec": 1589.0
+ },
+ {
+  "step": 36750,
+  "inner_val_loss": 0.09925535452937839,
+  "inner_val_per_cohort": {
+   "immicro": 0.09771132965882619,
+   "imperial_eskd": 0.09846724468198689,
+   "liu2021": 0.10158748924732208
+  },
+  "train_loss": 0.09904875734448433,
+  "sec": 1600.8
+ },
+ {
+  "step": 37000,
+  "inner_val_loss": 0.09925095597752416,
+  "inner_val_per_cohort": {
+   "immicro": 0.0977093627055486,
+   "imperial_eskd": 0.09845911914652045,
+   "liu2021": 0.10158438608050346
+  },
+  "train_loss": 0.09862760493159294,
+  "sec": 1612.6
+ },
+ {
+  "step": 37250,
+  "inner_val_loss": 0.09924657757610383,
+  "inner_val_per_cohort": {
+   "immicro": 0.0977007004742821,
+   "imperial_eskd": 0.09845411811362613,
+   "liu2021": 0.10158491414040327
+  },
+  "train_loss": 0.09889650636911393,
+  "sec": 1624.3
+ },
+ {
+  "step": 37500,
+  "inner_val_loss": 0.09924213016273999,
+  "inner_val_per_cohort": {
+   "immicro": 0.09769877480963866,
+   "imperial_eskd": 0.09845046969977292,
+   "liu2021": 0.1015771459788084
+  },
+  "train_loss": 0.09900348865985871,
+  "sec": 1636.0
+ },
+ {
+  "step": 37750,
+  "inner_val_loss": 0.09923958511214064,
+  "inner_val_per_cohort": {
+   "immicro": 0.09769353891412418,
+   "imperial_eskd": 0.0984456247904084,
+   "liu2021": 0.10157959163188934
+  },
+  "train_loss": 0.09894430583715438,
+  "sec": 1647.8
+ },
+ {
+  "step": 38000,
+  "inner_val_loss": 0.09923493765256923,
+  "inner_val_per_cohort": {
+   "immicro": 0.09768956371893485,
+   "imperial_eskd": 0.09844448654489084,
+   "liu2021": 0.10157076269388199
+  },
+  "train_loss": 0.09905495619773864,
+  "sec": 1659.3
+ },
+ {
+  "step": 38250,
+  "inner_val_loss": 0.09923528945731996,
+  "inner_val_per_cohort": {
+   "immicro": 0.09768775633225839,
+   "imperial_eskd": 0.09845464439554648,
+   "liu2021": 0.10156346764415503
+  },
+  "train_loss": 0.09907284873723984,
+  "sec": 1671.1
+ },
+ {
+  "step": 38500,
+  "inner_val_loss": 0.09923375430846154,
+  "inner_val_per_cohort": {
+   "immicro": 0.0976846714814504,
+   "imperial_eskd": 0.09845528074286201,
+   "liu2021": 0.10156131070107222
+  },
+  "train_loss": 0.09897780102491378,
+  "sec": 1683.1
+ },
+ {
+  "step": 38750,
+  "inner_val_loss": 0.09923398165465003,
+  "inner_val_per_cohort": {
+   "immicro": 0.09768429740021627,
+   "imperial_eskd": 0.0984533845701001,
+   "liu2021": 0.10156426299363375
+  },
+  "train_loss": 0.09896312376856804,
+  "sec": 1695.1
+ },
+ {
+  "step": 39000,
+  "inner_val_loss": 0.09923235120985545,
+  "inner_val_per_cohort": {
+   "immicro": 0.09768208581954241,
+   "imperial_eskd": 0.09845148060809482,
+   "liu2021": 0.10156348720192909
+  },
+  "train_loss": 0.09879054906964302,
+  "sec": 1707.1
+ },
+ {
+  "step": 39250,
+  "inner_val_loss": 0.09922901097910874,
+  "inner_val_per_cohort": {
+   "immicro": 0.09767983108758926,
+   "imperial_eskd": 0.09844702549956062,
+   "liu2021": 0.10156017635017633
+  },
+  "train_loss": 0.09876596659421921,
+  "sec": 1719.3
+ },
+ {
+  "step": 39500,
+  "inner_val_loss": 0.09922444387668311,
+  "inner_val_per_cohort": {
+   "immicro": 0.09767690611382325,
+   "imperial_eskd": 0.09844353490255096,
+   "liu2021": 0.10155289061367512
+  },
+  "train_loss": 0.0991650185585022,
+  "sec": 1731.4
+ },
+ {
+  "step": 39750,
+  "inner_val_loss": 0.09922334098379419,
+  "inner_val_per_cohort": {
+   "immicro": 0.09767479045937459,
+   "imperial_eskd": 0.09844521500847557,
+   "liu2021": 0.10155001748353243
+  },
+  "train_loss": 0.09888976442813874,
+  "sec": 1743.3
+ },
+ {
+  "step": 40000,
+  "inner_val_loss": 0.09921769783691024,
+  "inner_val_per_cohort": {
+   "immicro": 0.09767392184585333,
+   "imperial_eskd": 0.09843612326817079,
+   "liu2021": 0.10154304839670658
+  },
+  "train_loss": 0.09892798399925232,
+  "sec": 1755.3
+ },
+ {
+  "step": 40250,
+  "inner_val_loss": 0.09921538155300147,
+  "inner_val_per_cohort": {
+   "immicro": 0.09767250344157219,
+   "imperial_eskd": 0.09843145243146202,
+   "liu2021": 0.10154218878597021
+  },
+  "train_loss": 0.0989858221411705,
+  "sec": 1768.5
+ },
+ {
+  "step": 40500,
+  "inner_val_loss": 0.09921165824764305,
+  "inner_val_per_cohort": {
+   "immicro": 0.09767217344293992,
+   "imperial_eskd": 0.09842510521411896,
+   "liu2021": 0.10153769608587027
+  },
+  "train_loss": 0.09911353090405464,
+  "sec": 1780.5
+ },
+ {
+  "step": 40750,
+  "inner_val_loss": 0.09920748057916308,
+  "inner_val_per_cohort": {
+   "immicro": 0.09766359813511372,
+   "imperial_eskd": 0.09842095659537749,
+   "liu2021": 0.10153788700699806
+  },
+  "train_loss": 0.09870276388525963,
+  "sec": 1792.3
+ },
+ {
+  "step": 41000,
+  "inner_val_loss": 0.09920869173124584,
+  "inner_val_per_cohort": {
+   "immicro": 0.09765898870925109,
+   "imperial_eskd": 0.09842761064117606,
+   "liu2021": 0.10153947584331036
+  },
+  "train_loss": 0.09889533534646035,
+  "sec": 1804.1
+ },
+ {
+  "step": 41250,
+  "inner_val_loss": 0.09920610574008237,
+  "inner_val_per_cohort": {
+   "immicro": 0.09765719001491864,
+   "imperial_eskd": 0.0984238436953588,
+   "liu2021": 0.10153728350996971
+  },
+  "train_loss": 0.09881549260020256,
+  "sec": 1815.9
+ },
+ {
+  "step": 41500,
+  "inner_val_loss": 0.09920833514758735,
+  "inner_val_per_cohort": {
+   "immicro": 0.09765264733384053,
+   "imperial_eskd": 0.09843285618857904,
+   "liu2021": 0.10153950192034245
+  },
+  "train_loss": 0.09884316202998161,
+  "sec": 1827.7
+ },
+ {
+  "step": 41750,
+  "inner_val_loss": 0.09920184699009464,
+  "inner_val_per_cohort": {
+   "immicro": 0.0976462649802367,
+   "imperial_eskd": 0.09842736714265564,
+   "liu2021": 0.1015319088473916
+  },
+  "train_loss": 0.09873170951008797,
+  "sec": 1839.5
+ },
+ {
+  "step": 42000,
+  "inner_val_loss": 0.09919669028522089,
+  "inner_val_per_cohort": {
+   "immicro": 0.09764220627645652,
+   "imperial_eskd": 0.09842182492667978,
+   "liu2021": 0.10152603965252638
+  },
+  "train_loss": 0.09872115179896354,
+  "sec": 1851.7
+ },
+ {
+  "step": 42250,
+  "inner_val_loss": 0.09919383323215174,
+  "inner_val_per_cohort": {
+   "immicro": 0.09764060098677874,
+   "imperial_eskd": 0.09842117435552857,
+   "liu2021": 0.10151972435414791
+  },
+  "train_loss": 0.09863546949625016,
+  "sec": 1863.6
+ },
+ {
+  "step": 42500,
+  "inner_val_loss": 0.09919021022741241,
+  "inner_val_per_cohort": {
+   "immicro": 0.0976384353513519,
+   "imperial_eskd": 0.0984137850728902,
+   "liu2021": 0.10151841025799513
+  },
+  "train_loss": 0.09911893022060395,
+  "sec": 1875.4
+ },
+ {
+  "step": 42750,
+  "inner_val_loss": 0.09918641708226818,
+  "inner_val_per_cohort": {
+   "immicro": 0.09763374272733927,
+   "imperial_eskd": 0.09841082820838148,
+   "liu2021": 0.1015146803110838
+  },
+  "train_loss": 0.09866088908910751,
+  "sec": 1887.4
+ },
+ {
+  "step": 43000,
+  "inner_val_loss": 0.09918631057283193,
+  "inner_val_per_cohort": {
+   "immicro": 0.09763490501791239,
+   "imperial_eskd": 0.0984035964039239,
+   "liu2021": 0.10152043029665947
+  },
+  "train_loss": 0.0985651224553585,
+  "sec": 1899.3
+ },
+ {
+  "step": 43250,
+  "inner_val_loss": 0.0991845711727034,
+  "inner_val_per_cohort": {
+   "immicro": 0.0976356090977788,
+   "imperial_eskd": 0.09840447184714404,
+   "liu2021": 0.10151363257318735
+  },
+  "train_loss": 0.09903169026970864,
+  "sec": 1910.9
+ },
+ {
+  "step": 43500,
+  "inner_val_loss": 0.09918464248002778,
+  "inner_val_per_cohort": {
+   "immicro": 0.09763293092449506,
+   "imperial_eskd": 0.09841158376498656,
+   "liu2021": 0.10150941275060177
+  },
+  "train_loss": 0.09896795332431793,
+  "sec": 1922.7
+ },
+ {
+  "step": 43750,
+  "inner_val_loss": 0.09918499867798704,
+  "inner_val_per_cohort": {
+   "immicro": 0.09763206262141466,
+   "imperial_eskd": 0.098413692956621,
+   "liu2021": 0.10150924045592546
+  },
+  "train_loss": 0.09932895737886428,
+  "sec": 1934.6
+ },
+ {
+  "step": 44000,
+  "inner_val_loss": 0.09918277307103078,
+  "inner_val_per_cohort": {
+   "immicro": 0.09763103909790516,
+   "imperial_eskd": 0.09840692207217216,
+   "liu2021": 0.101510358043015
+  },
+  "train_loss": 0.09858720603585243,
+  "sec": 1946.5
+ },
+ {
+  "step": 44250,
+  "inner_val_loss": 0.09918294490474944,
+  "inner_val_per_cohort": {
+   "immicro": 0.09762657371660073,
+   "imperial_eskd": 0.09841183403676207,
+   "liu2021": 0.10151042696088552
+  },
+  "train_loss": 0.09881421986222268,
+  "sec": 1958.4
+ },
+ {
+  "step": 44500,
+  "inner_val_loss": 0.09918107614485604,
+  "inner_val_per_cohort": {
+   "immicro": 0.09762147441506386,
+   "imperial_eskd": 0.09841357713395898,
+   "liu2021": 0.10150817688554525
+  },
+  "train_loss": 0.09897151693701745,
+  "sec": 1970.3
+ },
+ {
+  "step": 44750,
+  "inner_val_loss": 0.09917772072132187,
+  "inner_val_per_cohort": {
+   "immicro": 0.09761808191736539,
+   "imperial_eskd": 0.09841314974156293,
+   "liu2021": 0.10150193050503731
+  },
+  "train_loss": 0.09886547669768334,
+  "sec": 1982.2
+ },
+ {
+  "step": 45000,
+  "inner_val_loss": 0.09917506069471739,
+  "inner_val_per_cohort": {
+   "immicro": 0.09761708260824283,
+   "imperial_eskd": 0.09841622073541988,
+   "liu2021": 0.10149187874048948
+  },
+  "train_loss": 0.09877495992183685,
+  "sec": 1994.0
+ },
+ {
+  "step": 45250,
+  "inner_val_loss": 0.09917141792524342,
+  "inner_val_per_cohort": {
+   "immicro": 0.09761391611148913,
+   "imperial_eskd": 0.09841696748679335,
+   "liu2021": 0.1014833701774478
+  },
+  "train_loss": 0.09896727025508881,
+  "sec": 2005.8
+ },
+ {
+  "step": 45500,
+  "inner_val_loss": 0.09916833023343123,
+  "inner_val_per_cohort": {
+   "immicro": 0.09761261660605669,
+   "imperial_eskd": 0.0984171222556721,
+   "liu2021": 0.10147525183856487
+  },
+  "train_loss": 0.0992145713865757,
+  "sec": 2017.5
+ },
+ {
+  "step": 45750,
+  "inner_val_loss": 0.09916073129032597,
+  "inner_val_per_cohort": {
+   "immicro": 0.09760604333132505,
+   "imperial_eskd": 0.09840968014164404,
+   "liu2021": 0.10146647039800882
+  },
+  "train_loss": 0.09885516983270645,
+  "sec": 2029.3
+ },
+ {
+  "step": 46000,
+  "inner_val_loss": 0.09915942019510149,
+  "inner_val_per_cohort": {
+   "immicro": 0.09759808114419381,
+   "imperial_eskd": 0.0984134128825231,
+   "liu2021": 0.10146676655858755
+  },
+  "train_loss": 0.09903675276041031,
+  "sec": 2041.2
+ },
+ {
+  "step": 46250,
+  "inner_val_loss": 0.09915359865761164,
+  "inner_val_per_cohort": {
+   "immicro": 0.09759081620723009,
+   "imperial_eskd": 0.0984088924120773,
+   "liu2021": 0.10146108735352755
+  },
+  "train_loss": 0.09860858190059661,
+  "sec": 2053.1
+ },
+ {
+  "step": 46500,
+  "inner_val_loss": 0.0991513529190361,
+  "inner_val_per_cohort": {
+   "immicro": 0.09758647841711839,
+   "imperial_eskd": 0.0984077595851638,
+   "liu2021": 0.10145982075482607
+  },
+  "train_loss": 0.09888405001163483,
+  "sec": 2065.0
+ },
+ {
+  "step": 46750,
+  "inner_val_loss": 0.09914921861932134,
+  "inner_val_per_cohort": {
+   "immicro": 0.09758291579782963,
+   "imperial_eskd": 0.0983992950482802,
+   "liu2021": 0.10146544501185417
+  },
+  "train_loss": 0.09873066565394402,
+  "sec": 2076.6
+ },
+ {
+  "step": 47000,
+  "inner_val_loss": 0.09914722939896764,
+  "inner_val_per_cohort": {
+   "immicro": 0.0975803742185235,
+   "imperial_eskd": 0.0983934344893152,
+   "liu2021": 0.10146787948906422
+  },
+  "train_loss": 0.09863978362083435,
+  "sec": 2088.6
+ },
+ {
+  "step": 47250,
+  "inner_val_loss": 0.09914648413394737,
+  "inner_val_per_cohort": {
+   "immicro": 0.09757891607781251,
+   "imperial_eskd": 0.09839229827577417,
+   "liu2021": 0.10146823804825544
+  },
+  "train_loss": 0.09883225962519646,
+  "sec": 2100.3
+ },
+ {
+  "step": 47500,
+  "inner_val_loss": 0.09914414908247765,
+  "inner_val_per_cohort": {
+   "immicro": 0.09757769883920749,
+   "imperial_eskd": 0.09838525679978458,
+   "liu2021": 0.10146949160844088
+  },
+  "train_loss": 0.0990116263628006,
+  "sec": 2112.2
+ },
+ {
+  "step": 47750,
+  "inner_val_loss": 0.09914173734270865,
+  "inner_val_per_cohort": {
+   "immicro": 0.09757504394898812,
+   "imperial_eskd": 0.0983799397945404,
+   "liu2021": 0.1014702282845974
+  },
+  "train_loss": 0.09883168774843216,
+  "sec": 2124.1
+ },
+ {
+  "step": 48000,
+  "inner_val_loss": 0.09913776646546944,
+  "inner_val_per_cohort": {
+   "immicro": 0.09757277990380923,
+   "imperial_eskd": 0.09837393361059102,
+   "liu2021": 0.10146658588200808
+  },
+  "train_loss": 0.0988412564098835,
+  "sec": 2136.0
+ },
+ {
+  "step": 48250,
+  "inner_val_loss": 0.09913322208167026,
+  "inner_val_per_cohort": {
+   "immicro": 0.09756954945623875,
+   "imperial_eskd": 0.09837468815120784,
+   "liu2021": 0.10145542863756418
+  },
+  "train_loss": 0.09906010767817497,
+  "sec": 2148.0
+ },
+ {
+  "step": 48500,
+  "inner_val_loss": 0.09913473206600458,
+  "inner_val_per_cohort": {
+   "immicro": 0.09757044818252325,
+   "imperial_eskd": 0.09837075424465266,
+   "liu2021": 0.10146299377083778
+  },
+  "train_loss": 0.09871631351113319,
+  "sec": 2160.0
+ },
+ {
+  "step": 48750,
+  "inner_val_loss": 0.09913005727823034,
+  "inner_val_per_cohort": {
+   "immicro": 0.09756552334874868,
+   "imperial_eskd": 0.09836570038036867,
+   "liu2021": 0.10145894810557365
+  },
+  "train_loss": 0.0985575635433197,
+  "sec": 2171.0
+ },
+ {
+  "step": 49000,
+  "inner_val_loss": 0.09912713085574032,
+  "inner_val_per_cohort": {
+   "immicro": 0.09756414437045653,
+   "imperial_eskd": 0.09836000441150232,
+   "liu2021": 0.10145724378526211
+  },
+  "train_loss": 0.09898614141345025,
+  "sec": 2181.3
+ },
+ {
+  "step": 49250,
+  "inner_val_loss": 0.09912196589125828,
+  "inner_val_per_cohort": {
+   "immicro": 0.09755764342844486,
+   "imperial_eskd": 0.09835427288304675,
+   "liu2021": 0.10145398136228323
+  },
+  "train_loss": 0.09878795754909515,
+  "sec": 2192.9
+ },
+ {
+  "step": 49500,
+  "inner_val_loss": 0.09911951455616891,
+  "inner_val_per_cohort": {
+   "immicro": 0.09755662859727939,
+   "imperial_eskd": 0.09834918820045212,
+   "liu2021": 0.10145272687077522
+  },
+  "train_loss": 0.09872793969511985,
+  "sec": 2204.7
+ },
+ {
+  "step": 49750,
+  "inner_val_loss": 0.09912089547236459,
+  "inner_val_per_cohort": {
+   "immicro": 0.09755454088250796,
+   "imperial_eskd": 0.0983544100414623,
+   "liu2021": 0.10145373549312353
+  },
+  "train_loss": 0.09889029198884965,
+  "sec": 2216.4
+ },
+ {
+  "step": 50000,
+  "inner_val_loss": 0.0991194269365885,
+  "inner_val_per_cohort": {
+   "immicro": 0.097547915764153,
+   "imperial_eskd": 0.09835967963392084,
+   "liu2021": 0.10145068541169167
+  },
+  "train_loss": 0.09841051569581032,
+  "sec": 2228.2
+ },
+ {
+  "step": 50250,
+  "inner_val_loss": 0.09912143943059926,
+  "inner_val_per_cohort": {
+   "immicro": 0.09754828953494628,
+   "imperial_eskd": 0.09836896712129767,
+   "liu2021": 0.10144706163555384
+  },
+  "train_loss": 0.09863430169224739,
+  "sec": 2239.9
+ },
+ {
+  "step": 50500,
+  "inner_val_loss": 0.09911649046710343,
+  "inner_val_per_cohort": {
+   "immicro": 0.09754613911112149,
+   "imperial_eskd": 0.098364745690064,
+   "liu2021": 0.10143858660012484
+  },
+  "train_loss": 0.09901295855641365,
+  "sec": 2251.6
+ },
+ {
+  "step": 50750,
+  "inner_val_loss": 0.09911618968754103,
+  "inner_val_per_cohort": {
+   "immicro": 0.09754737000912428,
+   "imperial_eskd": 0.09836350652304562,
+   "liu2021": 0.1014376925304532
+  },
+  "train_loss": 0.09854860419034958,
+  "sec": 2263.4
+ },
+ {
+  "step": 51000,
+  "inner_val_loss": 0.09911762931468811,
+  "inner_val_per_cohort": {
+   "immicro": 0.09755073674023151,
+   "imperial_eskd": 0.09836432642557404,
+   "liu2021": 0.1014378247782588
+  },
+  "train_loss": 0.09890629333257675,
+  "sec": 2275.1
+ },
+ {
+  "step": 51250,
+  "inner_val_loss": 0.09911238422824277,
+  "inner_val_per_cohort": {
+   "immicro": 0.09754745631168286,
+   "imperial_eskd": 0.09835821017622948,
+   "liu2021": 0.10143148619681597
+  },
+  "train_loss": 0.09872068345546722,
+  "sec": 2286.9
+ },
+ {
+  "step": 51500,
+  "inner_val_loss": 0.09911214014646984,
+  "inner_val_per_cohort": {
+   "immicro": 0.09754582432409127,
+   "imperial_eskd": 0.09835764494809238,
+   "liu2021": 0.10143295116722584
+  },
+  "train_loss": 0.09874440547823905,
+  "sec": 2298.6
+ },
+ {
+  "step": 51750,
+  "inner_val_loss": 0.09911050132917937,
+  "inner_val_per_cohort": {
+   "immicro": 0.09754235049088796,
+   "imperial_eskd": 0.09835285795005885,
+   "liu2021": 0.10143629554659128
+  },
+  "train_loss": 0.09886522942781448,
+  "sec": 2310.4
+ },
+ {
+  "step": 52000,
+  "inner_val_loss": 0.09910877149627366,
+  "inner_val_per_cohort": {
+   "immicro": 0.09753622518231471,
+   "imperial_eskd": 0.09834894808855923,
+   "liu2021": 0.101441141217947
+  },
+  "train_loss": 0.0989722189605236,
+  "sec": 2322.0
+ },
+ {
+  "step": 52250,
+  "inner_val_loss": 0.09910904629287696,
+  "inner_val_per_cohort": {
+   "immicro": 0.09753741354991992,
+   "imperial_eskd": 0.0983479070392522,
+   "liu2021": 0.10144181828945875
+  },
+  "train_loss": 0.0988659348487854,
+  "sec": 2333.7
+ },
+ {
+  "step": 52500,
+  "inner_val_loss": 0.09911016592135032,
+  "inner_val_per_cohort": {
+   "immicro": 0.09753238223493099,
+   "imperial_eskd": 0.09835127741098404,
+   "liu2021": 0.10144683811813593
+  },
+  "train_loss": 0.09922552314400673,
+  "sec": 2345.5
+ },
+ {
+  "step": 52750,
+  "inner_val_loss": 0.09910709711939397,
+  "inner_val_per_cohort": {
+   "immicro": 0.09753176600982745,
+   "imperial_eskd": 0.09834623743187297,
+   "liu2021": 0.1014432879164815
+  },
+  "train_loss": 0.09894033402204513,
+  "sec": 2357.1
+ },
+ {
+  "step": 53000,
+  "inner_val_loss": 0.09910342249680648,
+  "inner_val_per_cohort": {
+   "immicro": 0.0975267244502902,
+   "imperial_eskd": 0.09834281998601827,
+   "liu2021": 0.101440723054111
+  },
+  "train_loss": 0.09856633412837983,
+  "sec": 2368.9
+ },
+ {
+  "step": 53250,
+  "inner_val_loss": 0.09910648380114574,
+  "inner_val_per_cohort": {
+   "immicro": 0.09752453056474526,
+   "imperial_eskd": 0.09834977679631927,
+   "liu2021": 0.1014451440423727
+  },
+  "train_loss": 0.09889911079406738,
+  "sec": 2380.7
+ },
+ {
+  "step": 53500,
+  "inner_val_loss": 0.09910155590059179,
+  "inner_val_per_cohort": {
+   "immicro": 0.09751605521887541,
+   "imperial_eskd": 0.0983499904925173,
+   "liu2021": 0.10143862199038267
+  },
+  "train_loss": 0.09865599223971366,
+  "sec": 2392.5
+ },
+ {
+  "step": 53750,
+  "inner_val_loss": 0.09910094293041362,
+  "inner_val_per_cohort": {
+   "immicro": 0.09751074357579152,
+   "imperial_eskd": 0.09835483506321907,
+   "liu2021": 0.10143725015223026
+  },
+  "train_loss": 0.09882554018497466,
+  "sec": 2404.2
+ },
+ {
+  "step": 54000,
+  "inner_val_loss": 0.0990963645392295,
+  "inner_val_per_cohort": {
+   "immicro": 0.0975094623863697,
+   "imperial_eskd": 0.0983570549975742,
+   "liu2021": 0.10142257623374462
+  },
+  "train_loss": 0.09864963078498841,
+  "sec": 2415.8
+ },
+ {
+  "step": 54250,
+  "inner_val_loss": 0.09909298338673332,
+  "inner_val_per_cohort": {
+   "immicro": 0.09750747587531805,
+   "imperial_eskd": 0.09835535118525679,
+   "liu2021": 0.10141612309962511
+  },
+  "train_loss": 0.09898491591215133,
+  "sec": 2427.5
+ },
+ {
+  "step": 54500,
+  "inner_val_loss": 0.09909435432178504,
+  "inner_val_per_cohort": {
+   "immicro": 0.0975024625658989,
+   "imperial_eskd": 0.09835845536806366,
+   "liu2021": 0.10142214503139257
+  },
+  "train_loss": 0.09884439074993133,
+  "sec": 2439.2
+ },
+ {
+  "step": 54750,
+  "inner_val_loss": 0.09909271068795762,
+  "inner_val_per_cohort": {
+   "immicro": 0.09750257867077987,
+   "imperial_eskd": 0.09836080399426547,
+   "liu2021": 0.10141474939882755
+  },
+  "train_loss": 0.09860664382576942,
+  "sec": 2451.1
+ },
+ {
+  "step": 55000,
+  "inner_val_loss": 0.0990911502240583,
+  "inner_val_per_cohort": {
+   "immicro": 0.09749670637150605,
+   "imperial_eskd": 0.09836337884718721,
+   "liu2021": 0.10141336545348167
+  },
+  "train_loss": 0.0987769887149334,
+  "sec": 2462.8
+ },
+ {
+  "step": 55250,
+  "inner_val_loss": 0.09908821708475701,
+  "inner_val_per_cohort": {
+   "immicro": 0.0974962670976917,
+   "imperial_eskd": 0.0983587039465254,
+   "liu2021": 0.10140968021005392
+  },
+  "train_loss": 0.09859942343831063,
+  "sec": 2474.5
+ },
+ {
+  "step": 55500,
+  "inner_val_loss": 0.09908765356874827,
+  "inner_val_per_cohort": {
+   "immicro": 0.09749284852296114,
+   "imperial_eskd": 0.09835493158210408,
+   "liu2021": 0.1014151806011796
+  },
+  "train_loss": 0.09878082996606827,
+  "sec": 2486.3
+ },
+ {
+  "step": 55750,
+  "inner_val_loss": 0.09908719807675088,
+  "inner_val_per_cohort": {
+   "immicro": 0.09749139628062646,
+   "imperial_eskd": 0.09835080125115135,
+   "liu2021": 0.10141939669847488
+  },
+  "train_loss": 0.0987787958085537,
+  "sec": 2498.1
+ },
+ {
+  "step": 56000,
+  "inner_val_loss": 0.09908281237526674,
+  "inner_val_per_cohort": {
+   "immicro": 0.0974938500051697,
+   "imperial_eskd": 0.0983438982882283,
+   "liu2021": 0.10141068883240223
+  },
+  "train_loss": 0.09859168410301208,
+  "sec": 2509.7
+ },
+ {
+  "step": 56250,
+  "inner_val_loss": 0.09908135770584899,
+  "inner_val_per_cohort": {
+   "immicro": 0.09749664397289355,
+   "imperial_eskd": 0.09834549474445256,
+   "liu2021": 0.10140193440020084
+  },
+  "train_loss": 0.0988267528116703,
+  "sec": 2521.3
+ },
+ {
+  "step": 56500,
+  "inner_val_loss": 0.09908284384268101,
+  "inner_val_per_cohort": {
+   "immicro": 0.0974943305676182,
+   "imperial_eskd": 0.09834775802764026,
+   "liu2021": 0.10140644293278456
+  },
+  "train_loss": 0.09843526512384415,
+  "sec": 2533.2
+ },
+ {
+  "step": 56750,
+  "inner_val_loss": 0.09907946491030732,
+  "inner_val_per_cohort": {
+   "immicro": 0.09749203423659007,
+   "imperial_eskd": 0.09834017367525534,
+   "liu2021": 0.10140618681907654
+  },
+  "train_loss": 0.09833121973276138,
+  "sec": 2544.9
+ },
+ {
+  "step": 57000,
+  "inner_val_loss": 0.09907439871305468,
+  "inner_val_per_cohort": {
+   "immicro": 0.09748964353154103,
+   "imperial_eskd": 0.09833593023094264,
+   "liu2021": 0.10139762237668037
+  },
+  "train_loss": 0.09849226334691048,
+  "sec": 2556.8
+ },
+ {
+  "step": 57250,
+  "inner_val_loss": 0.09907768409924977,
+  "inner_val_per_cohort": {
+   "immicro": 0.09748952277004719,
+   "imperial_eskd": 0.09834267266772011,
+   "liu2021": 0.10140085685998201
+  },
+  "train_loss": 0.09888469138741493,
+  "sec": 2568.5
+ },
+ {
+  "step": 57500,
+  "inner_val_loss": 0.09908233443745458,
+  "inner_val_per_cohort": {
+   "immicro": 0.09748918035378058,
+   "imperial_eskd": 0.09834627637808974,
+   "liu2021": 0.10141154658049345
+  },
+  "train_loss": 0.09866968008875847,
+  "sec": 2580.2
+ },
+ {
+  "step": 57750,
+  "inner_val_loss": 0.09908385999086831,
+  "inner_val_per_cohort": {
+   "immicro": 0.09749042956779401,
+   "imperial_eskd": 0.09834981709718704,
+   "liu2021": 0.10141133330762386
+  },
+  "train_loss": 0.09844868114590645,
+  "sec": 2591.9
+ },
+ {
+  "step": 58000,
+  "inner_val_loss": 0.09908586011428122,
+  "inner_val_per_cohort": {
+   "immicro": 0.09748956778397162,
+   "imperial_eskd": 0.09835320169275458,
+   "liu2021": 0.10141481086611748
+  },
+  "train_loss": 0.09874103501439095,
+  "sec": 2603.6
+ },
+ {
+  "step": 58250,
+  "inner_val_loss": 0.09908483060772973,
+  "inner_val_per_cohort": {
+   "immicro": 0.09749040038635333,
+   "imperial_eskd": 0.09835530715909871,
+   "liu2021": 0.10140878427773714
+  },
+  "train_loss": 0.0988630958199501,
+  "sec": 2615.4
+ },
+ {
+  "step": 58500,
+  "inner_val_loss": 0.09907948661294548,
+  "inner_val_per_cohort": {
+   "immicro": 0.09748632243523996,
+   "imperial_eskd": 0.09834658015858043,
+   "liu2021": 0.1014055572450161
+  },
+  "train_loss": 0.09892299008369446,
+  "sec": 2627.1
+ },
+ {
+  "step": 58750,
+  "inner_val_loss": 0.09907419696412605,
+  "inner_val_per_cohort": {
+   "immicro": 0.09748135165621837,
+   "imperial_eskd": 0.09833856536583467,
+   "liu2021": 0.10140267387032509
+  },
+  "train_loss": 0.09910489431023597,
+  "sec": 2638.9
+ },
+ {
+  "step": 59000,
+  "inner_val_loss": 0.09907614975001174,
+  "inner_val_per_cohort": {
+   "immicro": 0.0974842378248771,
+   "imperial_eskd": 0.09834039482203397,
+   "liu2021": 0.10140381660312414
+  },
+  "train_loss": 0.09887970399856567,
+  "sec": 2650.6
+ }
+]

paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/lora64sat/milestones.json

@@ -0,0 +1,32 @@
+[
+ {
+  "milestone_step": 20000,
+  "selected_step": 20000,
+  "selected_inner_val_loss": 0.09948734277765257,
+  "block_improvement": 0.008788668249517378,
+  "total_improvement": 0.008788668249517378,
+  "reason": "block_end",
+  "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/lora64sat/checkpoint_ft_s20000.pt",
+  "checkpoint_sha256": "86adfd23e1cac07cf2fe066b4d8dcf912d2c5b7439f8df9bdc7dbca12f8e176e"
+ },
+ {
+  "milestone_step": 40000,
+  "selected_step": 40000,
+  "selected_inner_val_loss": 0.09921769783691024,
+  "block_improvement": 0.0002696449407423307,
+  "total_improvement": 0.00905831319025971,
+  "reason": "block_end",
+  "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/lora64sat/checkpoint_ft_s40000.pt",
+  "checkpoint_sha256": "9c45ad32993702c16f072cb340e345fef38bff4228633463bdccbfd3ab200340"
+ },
+ {
+  "milestone_step": 59000,
+  "selected_step": 57000,
+  "selected_inner_val_loss": 0.09907439871305468,
+  "block_improvement": 0.0001432991238555642,
+  "total_improvement": 0.009201612314115273,
+  "reason": "early_stop",
+  "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/lora64sat/checkpoint_ft_s59000.pt",
+  "checkpoint_sha256": "1702ed8fcc77d0dfa0e7bc16bc4dd86f17def0e5c9aaf95d24d25c8ef9b3ad5b"
+ }
+]

paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/lora64sat/train.log

@@ -0,0 +1,242 @@
+2026-10-03T06:56:16 split=hoL_ye2026_sepsis K=1000 mode=lora trainable=5,184,896 balance=cohort immicro: train 95 rows / 347,061 cells, inner 12; imperial_eskd: train 146 rows / 462,172 cells, inner 11; giroux2022: train 24 rows / 254,250 cells, inner 0; liu2021: train 40 rows / 147,532 cells, inner 4
+2026-10-03T06:56:17 step 0 inner-val loss 0.10828 (pretrained) {'immicro': 0.10636159498244524, 'imperial_eskd': 0.1086193590678952, 'liu2021': 0.10984707903116941}
+2026-10-03T06:56:29 step 250 train 0.10270 inner-val 0.10727 best 0.10727@250 (12s) immicro=0.1054 imperial_eskd=0.1075 liu2021=0.1089
+2026-10-03T06:56:41 step 500 train 0.10063 inner-val 0.10546 best 0.10546@500 (24s) immicro=0.1038 imperial_eskd=0.1055 liu2021=0.1071
+2026-10-03T06:56:53 step 750 train 0.10038 inner-val 0.10374 best 0.10374@750 (36s) immicro=0.1025 imperial_eskd=0.1034 liu2021=0.1054
+2026-10-03T06:57:04 step 1000 train 0.10008 inner-val 0.10242 best 0.10242@1000 (47s) immicro=0.1014 imperial_eskd=0.1017 liu2021=0.1042
+2026-10-03T06:57:15 step 1250 train 0.10018 inner-val 0.10155 best 0.10155@1250 (57s) immicro=0.1006 imperial_eskd=0.1006 liu2021=0.1034
+2026-10-03T06:57:26 step 1500 train 0.09994 inner-val 0.10101 best 0.10101@1500 (69s) immicro=0.1000 imperial_eskd=0.1000 liu2021=0.1030
+2026-10-03T06:57:37 step 1750 train 0.10006 inner-val 0.10066 best 0.10066@1750 (80s) immicro=0.0996 imperial_eskd=0.0996 liu2021=0.1028
+2026-10-03T06:57:48 step 2000 train 0.09982 inner-val 0.10043 best 0.10043@2000 (91s) immicro=0.0993 imperial_eskd=0.0994 liu2021=0.1026
+2026-10-03T06:57:58 step 2250 train 0.09995 inner-val 0.10028 best 0.10028@2250 (101s) immicro=0.0991 imperial_eskd=0.0993 liu2021=0.1025
+2026-10-03T06:58:09 step 2500 train 0.10020 inner-val 0.10018 best 0.10018@2500 (111s) immicro=0.0990 imperial_eskd=0.0992 liu2021=0.1024
+2026-10-03T06:58:19 step 2750 train 0.09981 inner-val 0.10011 best 0.10011@2750 (122s) immicro=0.0989 imperial_eskd=0.0991 liu2021=0.1023
+2026-10-03T06:58:29 step 3000 train 0.09978 inner-val 0.10005 best 0.10005@3000 (132s) immicro=0.0988 imperial_eskd=0.0990 liu2021=0.1023
+2026-10-03T06:58:40 step 3250 train 0.09999 inner-val 0.10001 best 0.10001@3250 (142s) immicro=0.0988 imperial_eskd=0.0990 liu2021=0.1023
+2026-10-03T06:58:50 step 3500 train 0.09994 inner-val 0.09998 best 0.09998@3500 (153s) immicro=0.0987 imperial_eskd=0.0990 liu2021=0.1022
+2026-10-03T06:59:01 step 3750 train 0.09981 inner-val 0.09995 best 0.09995@3750 (163s) immicro=0.0987 imperial_eskd=0.0989 liu2021=0.1022
+2026-10-03T06:59:11 step 4000 train 0.09975 inner-val 0.09993 best 0.09993@4000 (174s) immicro=0.0986 imperial_eskd=0.0989 liu2021=0.1022
+2026-10-03T06:59:21 step 4250 train 0.09977 inner-val 0.09991 best 0.09991@4250 (184s) immicro=0.0986 imperial_eskd=0.0989 liu2021=0.1022
+2026-10-03T06:59:32 step 4500 train 0.09971 inner-val 0.09989 best 0.09989@4500 (194s) immicro=0.0986 imperial_eskd=0.0989 liu2021=0.1022
+2026-10-03T06:59:42 step 4750 train 0.09960 inner-val 0.09988 best 0.09988@4750 (205s) immicro=0.0986 imperial_eskd=0.0989 liu2021=0.1022
+2026-10-03T06:59:52 step 5000 train 0.09977 inner-val 0.09986 best 0.09986@5000 (215s) immicro=0.0986 imperial_eskd=0.0988 liu2021=0.1022
+2026-10-03T07:00:03 step 5250 train 0.09972 inner-val 0.09985 best 0.09985@5250 (226s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1022
+2026-10-03T07:00:13 step 5500 train 0.09977 inner-val 0.09984 best 0.09984@5500 (236s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1022
+2026-10-03T07:00:24 step 5750 train 0.09931 inner-val 0.09983 best 0.09983@5750 (246s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1022
+2026-10-03T07:00:34 step 6000 train 0.09961 inner-val 0.09981 best 0.09981@6000 (257s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1022
+2026-10-03T07:00:45 step 6250 train 0.09967 inner-val 0.09980 best 0.09980@6250 (267s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1022
+2026-10-03T07:00:55 step 6500 train 0.09975 inner-val 0.09979 best 0.09979@6500 (278s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1021
+2026-10-03T07:01:06 step 6750 train 0.09960 inner-val 0.09978 best 0.09978@6750 (288s) immicro=0.0984 imperial_eskd=0.0988 liu2021=0.1021
+2026-10-03T07:01:16 step 7000 train 0.09952 inner-val 0.09977 best 0.09977@7000 (299s) immicro=0.0984 imperial_eskd=0.0988 liu2021=0.1021
+2026-10-03T07:01:26 step 7250 train 0.09974 inner-val 0.09976 best 0.09976@7250 (309s) immicro=0.0984 imperial_eskd=0.0988 liu2021=0.1021
+2026-10-03T07:01:37 step 7500 train 0.09955 inner-val 0.09975 best 0.09975@7500 (319s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1021
+2026-10-03T07:01:47 step 7750 train 0.09954 inner-val 0.09974 best 0.09974@7750 (330s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1021
+2026-10-03T07:01:58 step 8000 train 0.09924 inner-val 0.09973 best 0.09973@8000 (340s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1021
+2026-10-03T07:02:08 step 8250 train 0.09969 inner-val 0.09973 best 0.09973@8250 (351s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1021
+2026-10-03T07:02:19 step 8500 train 0.09928 inner-val 0.09972 best 0.09972@8500 (361s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1021
+2026-10-03T07:02:29 step 8750 train 0.09975 inner-val 0.09971 best 0.09971@8750 (372s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1021
+2026-10-03T07:02:40 step 9000 train 0.09955 inner-val 0.09971 best 0.09971@9000 (382s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:02:50 step 9250 train 0.09956 inner-val 0.09970 best 0.09970@9250 (393s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:03:00 step 9500 train 0.09948 inner-val 0.09970 best 0.09970@9500 (403s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:03:11 step 9750 train 0.09952 inner-val 0.09969 best 0.09969@9750 (413s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:03:21 step 10000 train 0.09945 inner-val 0.09969 best 0.09969@10000 (424s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:03:32 step 10250 train 0.09961 inner-val 0.09968 best 0.09968@10250 (434s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:03:42 step 10500 train 0.09964 inner-val 0.09968 best 0.09968@10500 (445s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:03:52 step 10750 train 0.09946 inner-val 0.09968 best 0.09968@10500 (455s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:04:03 step 11000 train 0.09950 inner-val 0.09968 best 0.09968@11000 (465s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:04:13 step 11250 train 0.09938 inner-val 0.09967 best 0.09967@11250 (476s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:04:24 step 11500 train 0.09946 inner-val 0.09966 best 0.09966@11500 (486s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:04:34 step 11750 train 0.09936 inner-val 0.09966 best 0.09966@11750 (497s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:04:44 step 12000 train 0.09917 inner-val 0.09965 best 0.09965@12000 (507s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:04:55 step 12250 train 0.09934 inner-val 0.09964 best 0.09964@12250 (518s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:05:05 step 12500 train 0.09950 inner-val 0.09963 best 0.09963@12500 (528s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:05:16 step 12750 train 0.09944 inner-val 0.09963 best 0.09963@12750 (538s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:05:26 step 13000 train 0.09951 inner-val 0.09962 best 0.09962@13000 (549s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:05:37 step 13250 train 0.09950 inner-val 0.09962 best 0.09962@13250 (559s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:05:47 step 13500 train 0.09936 inner-val 0.09961 best 0.09961@13500 (570s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:05:57 step 13750 train 0.09945 inner-val 0.09961 best 0.09961@13750 (580s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:06:08 step 14000 train 0.09930 inner-val 0.09960 best 0.09960@14000 (590s) immicro=0.0982 imperial_eskd=0.0987 liu2021=0.1020
+2026-10-03T07:06:18 step 14250 train 0.09920 inner-val 0.09959 best 0.09959@14250 (601s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1020
+2026-10-03T07:06:28 step 14500 train 0.09953 inner-val 0.09959 best 0.09959@14500 (611s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:06:39 step 14750 train 0.09941 inner-val 0.09959 best 0.09959@14750 (622s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:06:50 step 15000 train 0.09949 inner-val 0.09958 best 0.09958@15000 (632s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:07:00 step 15250 train 0.09926 inner-val 0.09958 best 0.09958@15250 (643s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:07:11 step 15500 train 0.09946 inner-val 0.09957 best 0.09957@15500 (653s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:07:21 step 15750 train 0.09921 inner-val 0.09957 best 0.09957@15750 (664s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:07:32 step 16000 train 0.09925 inner-val 0.09956 best 0.09956@16000 (674s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:07:42 step 16250 train 0.09935 inner-val 0.09956 best 0.09956@16250 (685s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:07:53 step 16500 train 0.09948 inner-val 0.09955 best 0.09955@16500 (695s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:08:05 step 16750 train 0.09917 inner-val 0.09955 best 0.09955@16750 (707s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:08:16 step 17000 train 0.09946 inner-val 0.09955 best 0.09955@17000 (719s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:08:28 step 17250 train 0.09955 inner-val 0.09954 best 0.09954@17250 (731s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:08:40 step 17500 train 0.09940 inner-val 0.09954 best 0.09954@17500 (743s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:08:52 step 17750 train 0.09910 inner-val 0.09953 best 0.09953@17750 (755s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:09:04 step 18000 train 0.09967 inner-val 0.09953 best 0.09953@17750 (767s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:09:16 step 18250 train 0.09910 inner-val 0.09953 best 0.09953@18250 (779s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:09:28 step 18500 train 0.09940 inner-val 0.09952 best 0.09952@18500 (791s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:09:40 step 18750 train 0.09918 inner-val 0.09951 best 0.09951@18750 (803s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T07:09:51 step 19000 train 0.09929 inner-val 0.09951 best 0.09951@19000 (813s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:10:02 step 19250 train 0.09919 inner-val 0.09950 best 0.09950@19250 (825s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:10:14 step 19500 train 0.09918 inner-val 0.09950 best 0.09950@19500 (837s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:10:26 step 19750 train 0.09911 inner-val 0.09949 best 0.09949@19750 (849s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:10:37 step 20000 train 0.09931 inner-val 0.09949 best 0.09949@20000 (860s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:10:39 block end 20000: best 0.09949@20000; block improvement 0.008789 (100.00 % of total 0.008789) -> continue
+2026-10-03T07:10:49 step 20250 train 0.09931 inner-val 0.09949 best 0.09949@20250 (872s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:11:00 step 20500 train 0.09896 inner-val 0.09948 best 0.09948@20500 (883s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:11:10 step 20750 train 0.09919 inner-val 0.09948 best 0.09948@20750 (893s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:11:21 step 21000 train 0.09934 inner-val 0.09947 best 0.09947@21000 (903s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:11:31 step 21250 train 0.09878 inner-val 0.09947 best 0.09947@21250 (914s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:11:42 step 21500 train 0.09903 inner-val 0.09946 best 0.09946@21500 (924s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:11:52 step 21750 train 0.09935 inner-val 0.09946 best 0.09946@21750 (935s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:12:03 step 22000 train 0.09926 inner-val 0.09945 best 0.09945@22000 (945s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:12:13 step 22250 train 0.09918 inner-val 0.09945 best 0.09945@22250 (956s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:12:24 step 22500 train 0.09915 inner-val 0.09945 best 0.09945@22500 (966s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:12:34 step 22750 train 0.09923 inner-val 0.09945 best 0.09945@22750 (977s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:12:45 step 23000 train 0.09951 inner-val 0.09944 best 0.09944@23000 (987s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:12:55 step 23250 train 0.09934 inner-val 0.09944 best 0.09944@23250 (998s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:13:05 step 23500 train 0.09939 inner-val 0.09943 best 0.09943@23500 (1008s) immicro=0.0980 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:13:16 step 23750 train 0.09901 inner-val 0.09943 best 0.09943@23750 (1018s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:13:26 step 24000 train 0.09947 inner-val 0.09943 best 0.09943@24000 (1029s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T07:13:36 step 24250 train 0.09896 inner-val 0.09942 best 0.09942@24250 (1039s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:13:47 step 24500 train 0.09891 inner-val 0.09942 best 0.09942@24500 (1049s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:13:57 step 24750 train 0.09924 inner-val 0.09941 best 0.09941@24750 (1060s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:14:07 step 25000 train 0.09905 inner-val 0.09941 best 0.09941@25000 (1070s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:14:18 step 25250 train 0.09918 inner-val 0.09940 best 0.09940@25250 (1081s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:14:28 step 25500 train 0.09899 inner-val 0.09940 best 0.09940@25250 (1091s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:14:39 step 25750 train 0.09921 inner-val 0.09940 best 0.09940@25750 (1101s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:14:49 step 26000 train 0.09922 inner-val 0.09940 best 0.09940@26000 (1112s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:14:59 step 26250 train 0.09909 inner-val 0.09939 best 0.09939@26250 (1122s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:15:10 step 26500 train 0.09926 inner-val 0.09939 best 0.09939@26500 (1133s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:15:20 step 26750 train 0.09880 inner-val 0.09939 best 0.09939@26750 (1143s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:15:31 step 27000 train 0.09923 inner-val 0.09939 best 0.09939@27000 (1154s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:15:41 step 27250 train 0.09930 inner-val 0.09938 best 0.09938@27250 (1164s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:15:51 step 27500 train 0.09918 inner-val 0.09938 best 0.09938@27500 (1174s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:16:02 step 27750 train 0.09929 inner-val 0.09938 best 0.09938@27750 (1184s) immicro=0.0979 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:16:12 step 28000 train 0.09901 inner-val 0.09938 best 0.09938@27750 (1195s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:16:23 step 28250 train 0.09909 inner-val 0.09937 best 0.09937@28250 (1205s) immicro=0.0979 imperial_eskd=0.0986 liu2021=0.1017
+2026-10-03T07:16:33 step 28500 train 0.09875 inner-val 0.09936 best 0.09936@28500 (1216s) immicro=0.0979 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:16:43 step 28750 train 0.09897 inner-val 0.09936 best 0.09936@28750 (1226s) immicro=0.0979 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:16:54 step 29000 train 0.09896 inner-val 0.09936 best 0.09936@29000 (1236s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:17:04 step 29250 train 0.09940 inner-val 0.09935 best 0.09935@29250 (1247s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:17:15 step 29500 train 0.09870 inner-val 0.09935 best 0.09935@29500 (1258s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:17:27 step 29750 train 0.09898 inner-val 0.09934 best 0.09934@29750 (1270s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:17:39 step 30000 train 0.09870 inner-val 0.09934 best 0.09934@30000 (1281s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:17:50 step 30250 train 0.09896 inner-val 0.09934 best 0.09934@30250 (1293s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:18:02 step 30500 train 0.09899 inner-val 0.09934 best 0.09934@30500 (1305s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:18:14 step 30750 train 0.09898 inner-val 0.09933 best 0.09933@30750 (1317s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:18:26 step 31000 train 0.09862 inner-val 0.09933 best 0.09933@31000 (1329s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:18:38 step 31250 train 0.09895 inner-val 0.09932 best 0.09932@31250 (1340s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:18:50 step 31500 train 0.09870 inner-val 0.09932 best 0.09932@31500 (1352s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T07:19:02 step 31750 train 0.09905 inner-val 0.09932 best 0.09932@31750 (1364s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:19:14 step 32000 train 0.09902 inner-val 0.09932 best 0.09932@32000 (1376s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:19:26 step 32250 train 0.09891 inner-val 0.09931 best 0.09931@32250 (1388s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:19:38 step 32500 train 0.09879 inner-val 0.09931 best 0.09931@32250 (1400s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:19:50 step 32750 train 0.09908 inner-val 0.09930 best 0.09930@32750 (1412s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:20:01 step 33000 train 0.09884 inner-val 0.09930 best 0.09930@33000 (1424s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:20:13 step 33250 train 0.09892 inner-val 0.09930 best 0.09930@33250 (1436s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:20:25 step 33500 train 0.09850 inner-val 0.09929 best 0.09929@33500 (1448s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:20:37 step 33750 train 0.09887 inner-val 0.09929 best 0.09929@33750 (1459s) immicro=0.0978 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:20:48 step 34000 train 0.09907 inner-val 0.09928 best 0.09928@34000 (1471s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:21:00 step 34250 train 0.09905 inner-val 0.09928 best 0.09928@34250 (1483s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:21:12 step 34500 train 0.09906 inner-val 0.09928 best 0.09928@34500 (1495s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:21:24 step 34750 train 0.09908 inner-val 0.09927 best 0.09927@34750 (1506s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:21:36 step 35000 train 0.09905 inner-val 0.09927 best 0.09927@35000 (1518s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:21:47 step 35250 train 0.09903 inner-val 0.09927 best 0.09927@35000 (1530s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:21:59 step 35500 train 0.09917 inner-val 0.09927 best 0.09927@35500 (1542s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:22:11 step 35750 train 0.09885 inner-val 0.09927 best 0.09927@35500 (1554s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:22:23 step 36000 train 0.09900 inner-val 0.09926 best 0.09926@36000 (1565s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:22:34 step 36250 train 0.09902 inner-val 0.09926 best 0.09926@36000 (1577s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:22:46 step 36500 train 0.09860 inner-val 0.09926 best 0.09926@36500 (1589s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:22:58 step 36750 train 0.09905 inner-val 0.09926 best 0.09926@36750 (1601s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:23:10 step 37000 train 0.09863 inner-val 0.09925 best 0.09925@37000 (1613s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:23:22 step 37250 train 0.09890 inner-val 0.09925 best 0.09925@37250 (1624s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:23:33 step 37500 train 0.09900 inner-val 0.09924 best 0.09924@37500 (1636s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:23:45 step 37750 train 0.09894 inner-val 0.09924 best 0.09924@37750 (1648s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1016
+2026-10-03T07:23:57 step 38000 train 0.09905 inner-val 0.09923 best 0.09923@38000 (1659s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1016
+2026-10-03T07:24:08 step 38250 train 0.09907 inner-val 0.09924 best 0.09923@38000 (1671s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:24:20 step 38500 train 0.09898 inner-val 0.09923 best 0.09923@38500 (1683s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:24:32 step 38750 train 0.09896 inner-val 0.09923 best 0.09923@38500 (1695s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:24:44 step 39000 train 0.09879 inner-val 0.09923 best 0.09923@39000 (1707s) immicro=0.0977 imperial_eskd=0.0985 liu2021=0.1016
+2026-10-03T07:24:57 step 39250 train 0.09877 inner-val 0.09923 best 0.09923@39250 (1719s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1016
+2026-10-03T07:25:09 step 39500 train 0.09917 inner-val 0.09922 best 0.09922@39500 (1731s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1016
+2026-10-03T07:25:21 step 39750 train 0.09889 inner-val 0.09922 best 0.09922@39750 (1743s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1016
+2026-10-03T07:25:33 step 40000 train 0.09893 inner-val 0.09922 best 0.09922@40000 (1755s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:25:34 block end 40000: best 0.09922@40000; block improvement 0.000270 (2.98 % of total 0.009058) -> continue
+2026-10-03T07:25:46 step 40250 train 0.09899 inner-val 0.09922 best 0.09922@40250 (1768s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:25:58 step 40500 train 0.09911 inner-val 0.09921 best 0.09921@40500 (1780s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:26:10 step 40750 train 0.09870 inner-val 0.09921 best 0.09921@40750 (1792s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:26:21 step 41000 train 0.09890 inner-val 0.09921 best 0.09921@40750 (1804s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:26:33 step 41250 train 0.09882 inner-val 0.09921 best 0.09921@41250 (1816s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:26:45 step 41500 train 0.09884 inner-val 0.09921 best 0.09921@41250 (1828s) immicro=0.0977 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:26:57 step 41750 train 0.09873 inner-val 0.09920 best 0.09920@41750 (1840s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:27:09 step 42000 train 0.09872 inner-val 0.09920 best 0.09920@42000 (1852s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:27:21 step 42250 train 0.09864 inner-val 0.09919 best 0.09919@42250 (1864s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:27:33 step 42500 train 0.09912 inner-val 0.09919 best 0.09919@42500 (1875s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:27:45 step 42750 train 0.09866 inner-val 0.09919 best 0.09919@42750 (1887s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:27:57 step 43000 train 0.09857 inner-val 0.09919 best 0.09919@42750 (1899s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:28:08 step 43250 train 0.09903 inner-val 0.09918 best 0.09918@43250 (1911s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:28:20 step 43500 train 0.09897 inner-val 0.09918 best 0.09918@43250 (1923s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:28:32 step 43750 train 0.09933 inner-val 0.09918 best 0.09918@43250 (1935s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:28:44 step 44000 train 0.09859 inner-val 0.09918 best 0.09918@44000 (1946s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:28:56 step 44250 train 0.09881 inner-val 0.09918 best 0.09918@44000 (1958s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:29:08 step 44500 train 0.09897 inner-val 0.09918 best 0.09918@44500 (1970s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:29:19 step 44750 train 0.09887 inner-val 0.09918 best 0.09918@44750 (1982s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:29:31 step 45000 train 0.09877 inner-val 0.09918 best 0.09918@45000 (1994s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:29:43 step 45250 train 0.09897 inner-val 0.09917 best 0.09917@45250 (2006s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:29:55 step 45500 train 0.09921 inner-val 0.09917 best 0.09917@45500 (2018s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:30:07 step 45750 train 0.09886 inner-val 0.09916 best 0.09916@45750 (2029s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:30:18 step 46000 train 0.09904 inner-val 0.09916 best 0.09916@46000 (2041s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:30:30 step 46250 train 0.09861 inner-val 0.09915 best 0.09915@46250 (2053s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:30:42 step 46500 train 0.09888 inner-val 0.09915 best 0.09915@46500 (2065s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:30:54 step 46750 train 0.09873 inner-val 0.09915 best 0.09915@46750 (2077s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:31:06 step 47000 train 0.09864 inner-val 0.09915 best 0.09915@47000 (2089s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:31:18 step 47250 train 0.09883 inner-val 0.09915 best 0.09915@47000 (2100s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:31:29 step 47500 train 0.09901 inner-val 0.09914 best 0.09914@47500 (2112s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:31:41 step 47750 train 0.09883 inner-val 0.09914 best 0.09914@47750 (2124s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:31:53 step 48000 train 0.09884 inner-val 0.09914 best 0.09914@48000 (2136s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:32:05 step 48250 train 0.09906 inner-val 0.09913 best 0.09913@48250 (2148s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:32:17 step 48500 train 0.09872 inner-val 0.09913 best 0.09913@48250 (2160s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:32:28 step 48750 train 0.09856 inner-val 0.09913 best 0.09913@48750 (2171s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:32:39 step 49000 train 0.09899 inner-val 0.09913 best 0.09913@49000 (2181s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:32:50 step 49250 train 0.09879 inner-val 0.09912 best 0.09912@49250 (2193s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:33:02 step 49500 train 0.09873 inner-val 0.09912 best 0.09912@49500 (2205s) immicro=0.0976 imperial_eskd=0.0983 liu2021=0.1015
+2026-10-03T07:33:14 step 49750 train 0.09889 inner-val 0.09912 best 0.09912@49500 (2216s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:33:25 step 50000 train 0.09841 inner-val 0.09912 best 0.09912@49500 (2228s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1015
+2026-10-03T07:33:37 step 50250 train 0.09863 inner-val 0.09912 best 0.09912@49500 (2240s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:33:49 step 50500 train 0.09901 inner-val 0.09912 best 0.09912@50500 (2252s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:34:01 step 50750 train 0.09855 inner-val 0.09912 best 0.09912@50500 (2263s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:34:12 step 51000 train 0.09891 inner-val 0.09912 best 0.09912@50500 (2275s) immicro=0.0976 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:34:24 step 51250 train 0.09872 inner-val 0.09911 best 0.09911@51250 (2287s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:34:36 step 51500 train 0.09874 inner-val 0.09911 best 0.09911@51250 (2299s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:34:48 step 51750 train 0.09887 inner-val 0.09911 best 0.09911@51750 (2310s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:34:59 step 52000 train 0.09897 inner-val 0.09911 best 0.09911@52000 (2322s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:35:11 step 52250 train 0.09887 inner-val 0.09911 best 0.09911@52000 (2334s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:35:23 step 52500 train 0.09923 inner-val 0.09911 best 0.09911@52000 (2346s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:35:34 step 52750 train 0.09894 inner-val 0.09911 best 0.09911@52750 (2357s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:35:46 step 53000 train 0.09857 inner-val 0.09910 best 0.09910@53000 (2369s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:35:58 step 53250 train 0.09890 inner-val 0.09911 best 0.09910@53000 (2381s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:36:10 step 53500 train 0.09866 inner-val 0.09910 best 0.09910@53500 (2392s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:36:21 step 53750 train 0.09883 inner-val 0.09910 best 0.09910@53500 (2404s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:36:33 step 54000 train 0.09865 inner-val 0.09910 best 0.09910@54000 (2416s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:36:45 step 54250 train 0.09898 inner-val 0.09909 best 0.09909@54250 (2428s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:36:57 step 54500 train 0.09884 inner-val 0.09909 best 0.09909@54250 (2439s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:37:08 step 54750 train 0.09861 inner-val 0.09909 best 0.09909@54250 (2451s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:37:20 step 55000 train 0.09878 inner-val 0.09909 best 0.09909@55000 (2463s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:37:32 step 55250 train 0.09860 inner-val 0.09909 best 0.09909@55250 (2474s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:37:44 step 55500 train 0.09878 inner-val 0.09909 best 0.09909@55250 (2486s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:37:55 step 55750 train 0.09878 inner-val 0.09909 best 0.09909@55750 (2498s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:38:07 step 56000 train 0.09859 inner-val 0.09908 best 0.09908@56000 (2510s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:38:19 step 56250 train 0.09883 inner-val 0.09908 best 0.09908@56250 (2521s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:38:30 step 56500 train 0.09844 inner-val 0.09908 best 0.09908@56250 (2533s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:38:42 step 56750 train 0.09833 inner-val 0.09908 best 0.09908@56750 (2545s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:38:54 step 57000 train 0.09849 inner-val 0.09907 best 0.09907@57000 (2557s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:39:06 step 57250 train 0.09888 inner-val 0.09908 best 0.09907@57000 (2568s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:39:18 step 57500 train 0.09867 inner-val 0.09908 best 0.09907@57000 (2580s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:39:29 step 57750 train 0.09845 inner-val 0.09908 best 0.09907@57000 (2592s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:39:41 step 58000 train 0.09874 inner-val 0.09909 best 0.09907@57000 (2604s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:39:53 step 58250 train 0.09886 inner-val 0.09908 best 0.09907@57000 (2615s) immicro=0.0975 imperial_eskd=0.0984 liu2021=0.1014
+2026-10-03T07:40:04 step 58500 train 0.09892 inner-val 0.09908 best 0.09907@57000 (2627s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:40:16 step 58750 train 0.09910 inner-val 0.09907 best 0.09907@57000 (2639s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:40:28 step 59000 train 0.09888 inner-val 0.09908 best 0.09907@57000 (2651s) immicro=0.0975 imperial_eskd=0.0983 liu2021=0.1014
+2026-10-03T07:40:28 early stop: no improvement in 8 evaluations
+2026-10-03T07:40:33 saved /home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/lora64sat/checkpoint_ft.pt (selected step 57000, inner-val 0.09907 vs pretrained 0.10828); sha b402a1fd359b

paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/scores/AGGREGATE.md

@@ -0,0 +1,115 @@
+# Cross-cohort (LOCO) benchmark: scores
+
+Generated by `tools/aggregate_loco.py`; do not edit. Zero-pair arms first, then within-cohort references that used the held-out cohort's own pairs.
+
+## ImMicro (107 samples; reference c5e2547be3ec)
+
+| arm | comp_l1 | p_l1_vs_zs_K1000 | precision | coverage | coverage_dm | pb_r_median |
+|---|---|---|---|---|---|---|
+| locoL_fullsat_s20000_K1000 | 0.208 | 4.29e-19 | 0.74 | 0.386 | 0.719 | 0.886 |
+| locoL_fullsat_s40000_K1000 | 0.211 | 3.93e-18 | 0.729 | 0.368 | 0.738 | 0.883 |
+| locoL_fullsat_s60000_K1000 | 0.223 | 1.14e-16 | 0.709 | 0.349 | 0.72 | 0.882 |
+| locoL_fullsat_s65000_K1000 | 0.213 | 2.67e-17 | 0.704 | 0.346 | 0.713 | 0.882 |
+| locoL_lora16_K1000 | 0.163 | 2.81e-19 | 0.833 | 0.451 | 0.705 | 0.886 |
+| locoL_lora16sat_s20000_K1000 | 0.163 | 2.81e-19 | 0.833 | 0.451 | 0.705 | 0.886 |
+| locoL_lora16sat_s28500_K1000 | 0.175 | 3.24e-19 | 0.841 | 0.452 | 0.685 | 0.887 |
+| locoL_lora64sat_s20000_K1000 | 0.196 | 6.78e-18 | 0.855 | 0.437 | 0.713 | 0.878 |
+| locoL_lora64sat_s39000_K1000 | 0.199 | 1.48e-17 | 0.835 | 0.382 | 0.664 | 0.878 |
+| xcalL_K1000 | 0.172 | 5.67e-19 | 0.902 | 0.572 | 0.738 | 0.889 |
+| zs_K1000 | 0.446 |  | 0.671 | 0.186 | 0.211 | 0.892 |
+| [within-cohort calibration] zs__cal | 0.199 | 7.12e-17 | 0.905 | 0.527 | 0.769 | 0.88 |
+| [within-cohort LoRA r16] real_lora__raw | 0.162 | 1.05e-18 | 0.924 | 0.784 | 0.962 | 0.899 |
+| [within-cohort full FT] real_full__raw | 0.157 | 6e-19 | 0.89 | 0.695 | 0.95 | 0.914 |
+
+## Sepsis (82 samples; reference bac12da8a241)
+
+| arm | comp_l1 | p_l1_vs_zs_K1000 | precision | coverage | coverage_dm | pb_r_median | logfc_r | broadcast_r | partial_r | deg50 | deg50_broadcast | deg50_max |
+|---|---|---|---|---|---|---|---|---|---|---|---|---|
+| locoL_fullsat_s100000_K1000 | 0.577 | 0.059 | 0.819 | 0.353 | 0.553 | 0.877 | 0.0778 | 0.316 | 0.00597 | 7 | 21 | 250 |
+| locoL_fullsat_s108500_K1000 | 0.585 | 0.0381 | 0.821 | 0.346 | 0.551 | 0.877 | 0.073 | 0.316 | -7.02e-05 | 5 | 21 | 250 |
+| locoL_fullsat_s20000_K1000 | 0.498 | 0.972 | 0.838 | 0.329 | 0.587 | 0.876 | 0.121 | 0.316 | 0.0503 | 10 | 21 | 250 |
+| locoL_fullsat_s40000_K1000 | 0.53 | 0.399 | 0.835 | 0.335 | 0.59 | 0.876 | 0.112 | 0.316 | 0.045 | 7 | 21 | 250 |
+| locoL_fullsat_s60000_K1000 | 0.546 | 0.22 | 0.83 | 0.339 | 0.57 | 0.877 | 0.0831 | 0.316 | 0.0124 | 7 | 21 | 250 |
+| locoL_fullsat_s80000_K1000 | 0.576 | 0.0648 | 0.828 | 0.342 | 0.577 | 0.878 | 0.0873 | 0.316 | 0.0163 | 6 | 21 | 250 |
+| locoL_lora16_K1000 | 0.5 | 0.873 | 0.866 | 0.347 | 0.572 | 0.865 | 0.125 | 0.316 | 0.0724 | 9 | 21 | 250 |
+| locoL_lora16sat_s20000_K1000 | 0.5 | 0.873 | 0.866 | 0.347 | 0.572 | 0.865 | 0.125 | 0.316 | 0.0724 | 9 | 21 | 250 |
+| locoL_lora16sat_s26750_K1000 | 0.491 | 0.801 | 0.861 | 0.341 | 0.568 | 0.866 | 0.102 | 0.316 | 0.0504 | 8 | 21 | 250 |
+| locoL_lora64sat_s20000_K1000 | 0.55 | 0.0911 | 0.836 | 0.295 | 0.494 | 0.867 | 0.177 | 0.316 | 0.124 | 12 | 21 | 250 |
+| locoL_lora64sat_s40000_K1000 | 0.564 | 0.0682 | 0.827 | 0.291 | 0.489 | 0.865 | 0.215 | 0.316 | 0.162 | 4 | 21 | 250 |
+| locoL_lora64sat_s59000_K1000 | 0.57 | 0.0893 | 0.828 | 0.289 | 0.486 | 0.862 | 0.228 | 0.316 | 0.176 | 4 | 21 | 250 |
+| loco_lora16_K1000 | 0.538 | 0.184 | 0.851 | 0.288 | 0.513 | 0.866 | 0.0916 | 0.316 | 0.0336 | 7 | 21 | 250 |
+| xcalL_K1000 | 0.444 | 0.00276 | 0.896 | 0.464 | 0.652 | 0.866 | 0.0967 | 0.317 | 0.0165 | 13 | 27 | 250 |
+| zs_K1000 | 0.509 |  | 0.617 | 0.074 | 0.0712 | 0.879 | -0.0218 | 0.316 | -0.103 | 9 | 21 | 250 |
+| [within-cohort calibration] zs__cal | 0.401 | 0.000189 | 0.903 | 0.431 | 0.683 | 0.861 | 0.123 | 0.278 | 0.0394 | 11 | 22 | 250 |
+| [within-cohort LoRA r16] real_lora__raw | 0.362 | 0.000685 | 0.926 | 0.829 | 0.949 | 0.891 | 0.527 | 0.316 | 0.469 | 23 | 21 | 250 |
+| [within-cohort full FT] real_full__raw | 0.372 | 0.000402 | 0.866 | 0.795 | 0.927 | 0.901 | 0.534 | 0.316 | 0.468 | 24 | 21 | 250 |
+
+## Imperial (157 samples; reference 34b4caa3f8a6)
+
+| arm | comp_l1 | p_l1_vs_zs_K1000 | precision | coverage | coverage_dm | pb_r_median | logfc_r | broadcast_r | partial_r | deg50 | deg50_broadcast | deg50_max |
+|---|---|---|---|---|---|---|---|---|---|---|---|---|
+| locoL_fullsat_s20000_K1000 | 0.632 | 2.64e-13 | 0.763 | 0.6 | 0.692 | 0.877 | 0.532 | 0.661 | 0.325 | 98 | 84 | 250 |
+| locoL_fullsat_s40000_K1000 | 0.609 | 4.89e-13 | 0.787 | 0.603 | 0.704 | 0.878 | 0.503 | 0.661 | 0.293 | 99 | 84 | 250 |
+| locoL_fullsat_s60000_K1000 | 0.565 | 1.38e-15 | 0.798 | 0.611 | 0.703 | 0.881 | 0.498 | 0.661 | 0.291 | 99 | 84 | 250 |
+| locoL_fullsat_s80000_K1000 | 0.576 | 2.57e-15 | 0.809 | 0.605 | 0.701 | 0.88 | 0.466 | 0.661 | 0.251 | 93 | 84 | 250 |
+| locoL_fullsat_s93000_K1000 | 0.58 | 3.45e-15 | 0.802 | 0.608 | 0.701 | 0.879 | 0.459 | 0.661 | 0.244 | 97 | 84 | 250 |
+| locoL_lora16_K1000 | 0.657 | 8.24e-15 | 0.804 | 0.683 | 0.75 | 0.872 | 0.527 | 0.661 | 0.316 | 96 | 84 | 250 |
+| locoL_lora16sat_s20000_K1000 | 0.657 | 8.24e-15 | 0.804 | 0.683 | 0.75 | 0.872 | 0.527 | 0.661 | 0.316 | 96 | 84 | 250 |
+| locoL_lora16sat_s40000_K1000 | 0.635 | 1.01e-14 | 0.82 | 0.684 | 0.755 | 0.871 | 0.491 | 0.661 | 0.268 | 94 | 84 | 250 |
+| locoL_lora64sat_s20000_K1000 | 0.622 | 4.58e-16 | 0.808 | 0.683 | 0.761 | 0.87 | 0.511 | 0.661 | 0.297 | 101 | 84 | 250 |
+| locoL_lora64sat_s40000_K1000 | 0.603 | 6.15e-17 | 0.819 | 0.711 | 0.782 | 0.874 | 0.535 | 0.661 | 0.314 | 101 | 84 | 250 |
+| locoL_lora64sat_s60000_K1000 | 0.577 | 1.98e-18 | 0.818 | 0.723 | 0.783 | 0.876 | 0.568 | 0.661 | 0.361 | 109 | 84 | 250 |
+| locoL_lora64sat_s69250_K1000 | 0.594 | 6.3e-18 | 0.822 | 0.723 | 0.779 | 0.876 | 0.58 | 0.661 | 0.373 | 102 | 84 | 250 |
+| xcalL_K1000 | 0.651 | 1.06e-16 | 0.845 | 0.637 | 0.664 | 0.865 | 0.689 | 0.677 | 0.445 | 108 | 84 | 250 |
+| zs_K1000 | 0.803 |  | 0.494 | 0.113 | 0.157 | 0.884 | 0.642 | 0.661 | 0.428 | 108 | 84 | 250 |
+| [within-cohort calibration] zs__cal | 0.494 | 8.14e-16 | 0.842 | 0.639 | 0.688 | 0.867 | 0.679 | 0.69 | 0.412 | 111 | 97 | 250 |
+| [within-cohort LoRA r16] real_lora__raw | 0.376 | 9.15e-21 | 0.847 | 0.85 | 0.937 | 0.88 | 0.819 | 0.661 | 0.69 | 143 | 84 | 250 |
+
+## Giroux (24 samples; reference dd0152f7bde4)
+
+| arm | comp_l1 | p_l1_vs_zs_K1000 | precision | coverage | coverage_dm | pb_r_median | logfc_r | broadcast_r | partial_r | deg50 | deg50_broadcast | deg50_max |
+|---|---|---|---|---|---|---|---|---|---|---|---|---|
+| locoL_fullsat_s100000_K1000 | 0.49 | 0.0164 | 0.983 | 0.764 | 0.741 | 0.86 | 0.225 | 0.575 | 0.131 | 1 | 5 | 101 |
+| locoL_fullsat_s120000_K1000 | 0.465 | 0.0526 | 0.984 | 0.776 | 0.731 | 0.869 | 0.196 | 0.575 | 0.107 | 2 | 5 | 101 |
+| locoL_fullsat_s20000_K1000 | 0.543 | 0.000742 | 0.985 | 0.753 | 0.785 | 0.856 | 0.208 | 0.575 | 0.131 | 2 | 5 | 101 |
+| locoL_fullsat_s40000_K1000 | 0.532 | 0.000239 | 0.988 | 0.753 | 0.742 | 0.852 | 0.203 | 0.575 | 0.131 | 2 | 5 | 101 |
+| locoL_fullsat_s60000_K1000 | 0.499 | 0.00871 | 0.986 | 0.756 | 0.726 | 0.855 | 0.217 | 0.575 | 0.141 | 2 | 5 | 101 |
+| locoL_fullsat_s80000_K1000 | 0.466 | 0.0691 | 0.984 | 0.764 | 0.729 | 0.858 | 0.228 | 0.575 | 0.14 | 2 | 5 | 101 |
+| locoL_lora16_K1000 | 0.416 | 0.152 | 0.991 | 0.793 | 0.819 | 0.845 | 0.296 | 0.575 | 0.204 | 0 | 5 | 101 |
+| locoL_lora16sat_s20000_K1000 | 0.416 | 0.152 | 0.991 | 0.793 | 0.819 | 0.845 | 0.296 | 0.575 | 0.204 | 0 | 5 | 101 |
+| locoL_lora16sat_s35750_K1000 | 0.431 | 0.16 | 0.995 | 0.819 | 0.835 | 0.844 | 0.299 | 0.575 | 0.203 | 0 | 5 | 101 |
+| locoL_lora64_K1000 | 0.491 | 0.0211 | 0.994 | 0.802 | 0.819 | 0.835 | 0.264 | 0.575 | 0.171 | 2 | 5 | 101 |
+| locoL_lora64sat_s20000_K1000 | 0.491 | 0.0211 | 0.994 | 0.802 | 0.819 | 0.835 | 0.264 | 0.575 | 0.171 | 2 | 5 | 101 |
+| locoL_lora64sat_s40000_K1000 | 0.469 | 0.034 | 0.994 | 0.818 | 0.835 | 0.843 | 0.239 | 0.575 | 0.154 | 0 | 5 | 101 |
+| locoL_lora64sat_s60000_K1000 | 0.43 | 0.114 | 0.994 | 0.815 | 0.845 | 0.847 | 0.272 | 0.575 | 0.187 | 1 | 5 | 101 |
+| locoL_lora64sat_s69250_K1000 | 0.41 | 0.229 | 0.994 | 0.809 | 0.841 | 0.85 | 0.278 | 0.575 | 0.187 | 0 | 5 | 101 |
+| loco_lora16_K1000 | 0.347 | 0.833 | 0.991 | 0.808 | 0.826 | 0.847 | 0.296 | 0.575 | 0.188 | 0 | 5 | 101 |
+| xcalL_K1000 | 0.202 | 0.00015 | 0.988 | 0.891 | 0.889 | 0.885 | 0.238 | 0.569 | 0.115 | 0 | 2 | 101 |
+| zs_K1000 | 0.337 |  | 0.904 | 0.293 | 0.369 | 0.889 | 0.34 | 0.575 | 0.197 | 0 | 5 | 101 |
+| zs_K2000 | 0.322 | 0.0425 | 0.917 | 0.277 | 0.344 | 0.889 | 0.339 | 0.575 | 0.189 | 0 | 5 | 101 |
+| [within-cohort calibration] zs__cal | 0.266 | 0.00958 | 0.995 | 0.906 | 0.902 | 0.865 | 0.269 | 0.564 | 0.133 | 0 | 2 | 101 |
+| [within-cohort LoRA r16] real_lora__raw | 0.258 | 0.0315 | 0.998 | 0.928 | 0.974 | 0.865 | 0.461 | 0.575 | 0.323 | 5 | 5 | 101 |
+
+## Ji (whole blood) (33 samples; reference 19c30f56db69)
+
+| arm | comp_l1 | p_l1_vs_zs_K1000 | precision | coverage | coverage_dm | pb_r_median | logfc_r | broadcast_r | partial_r | deg50 | deg50_broadcast | deg50_max |
+|---|---|---|---|---|---|---|---|---|---|---|---|---|
+| locoL_fullsat_s20000_K1000 | 0.838 | 1e-06 | 0.714 | 0.168 | 0.392 | 0.826 | 0.187 | 0.24 | 0.14 | 4 | 4 | 250 |
+| locoL_fullsat_s40000_K1000 | 1.01 | 4.66e-10 | 0.611 | 0.112 | 0.282 | 0.818 | 0.149 | 0.24 | 0.105 | 2 | 4 | 250 |
+| locoL_fullsat_s60000_K1000 | 1.03 | 4.66e-10 | 0.541 | 0.0836 | 0.23 | 0.814 | 0.116 | 0.24 | 0.0787 | 0 | 4 | 250 |
+| locoL_fullsat_s80000_K1000 | 1.05 | 2.33e-10 | 0.512 | 0.0721 | 0.2 | 0.815 | 0.132 | 0.24 | 0.0953 | 0 | 4 | 250 |
+| locoL_fullsat_s86750_K1000 | 1.03 | 4.66e-10 | 0.511 | 0.0812 | 0.201 | 0.817 | 0.139 | 0.24 | 0.101 | 5 | 4 | 250 |
+| locoL_lora16_K1000 | 0.757 | 0.0898 | 0.837 | 0.271 | 0.4 | 0.831 | 0.144 | 0.24 | 0.102 | 2 | 4 | 250 |
+| locoL_lora16sat_s20000_K1000 | 0.757 | 0.0898 | 0.837 | 0.271 | 0.4 | 0.831 | 0.144 | 0.24 | 0.102 | 2 | 4 | 250 |
+| locoL_lora16sat_s40000_K1000 | 0.74 | 0.764 | 0.911 | 0.339 | 0.424 | 0.841 | 0.0453 | 0.24 | 0.00734 | 0 | 4 | 250 |
+| locoL_lora64_K1000 | 0.639 | 0.00548 | 0.931 | 0.381 | 0.538 | 0.843 | 0.199 | 0.24 | 0.151 | 2 | 4 | 250 |
+| locoL_lora64sat_s20000_K1000 | 0.639 | 0.00548 | 0.931 | 0.381 | 0.538 | 0.843 | 0.199 | 0.24 | 0.151 | 2 | 4 | 250 |
+| locoL_lora64sat_s40000_K1000 | 0.668 | 0.0307 | 0.948 | 0.387 | 0.553 | 0.847 | 0.177 | 0.24 | 0.129 | 0 | 4 | 250 |
+| locoL_lora64sat_s51750_K1000 | 0.62 | 0.00158 | 0.947 | 0.395 | 0.572 | 0.846 | 0.164 | 0.24 | 0.127 | 0 | 4 | 250 |
+| loco_lora16_K1000 | 0.494 | 0.000602 | 0.845 | 0.313 | 0.521 | 0.836 | 0.0892 | 0.24 | 0.0417 | 9 | 4 | 250 |
+| xcalL_K1000 | 0.565 | 0.000245 | 0.636 | 0.0711 | 0.274 | 0.828 | 0.127 | 0.256 | 0.0897 | 0 | 4 | 250 |
+| zs_K1000 | 0.725 |  | 0.721 | 0.059 | 0.195 | 0.834 | 0.0749 | 0.24 | 0.023 | 0 | 4 | 250 |
+| zs_K2000 | 0.729 | 0.538 | 0.719 | 0.0604 | 0.173 | 0.836 | 0.0756 | 0.24 | 0.0229 | 0 | 4 | 250 |
+| [within-cohort calibration] zs__cal | 0.501 | 0.0011 | 0.74 | 0.106 | 0.347 | 0.839 | 0.142 | 0.206 | 0.113 | 0 | 1 | 250 |
+| [within-cohort LoRA r16] real_lora__raw | 0.504 | 0.00879 | 0.971 | 0.857 | 0.965 | 0.873 | 0.479 | 0.24 | 0.452 | 15 | 4 | 250 |
+