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_immicro/K1000/fullsat/milestones.json

@@ -0,0 +1,42 @@
+[
+ {
+  "milestone_step": 20000,
+  "selected_step": 20000,
+  "selected_inner_val_loss": 0.10489554748077083,
+  "block_improvement": 0.009959028431712993,
+  "total_improvement": 0.009959028431712993,
+  "reason": "block_end",
+  "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_immicro/K1000/fullsat/checkpoint_ft_s20000.pt",
+  "checkpoint_sha256": "079c482ae6d6c25d1dab913b0101807b6f169e1c0c1975b197257918276747e8"
+ },
+ {
+  "milestone_step": 40000,
+  "selected_step": 40000,
+  "selected_inner_val_loss": 0.10458750774776594,
+  "block_improvement": 0.00030803973300488896,
+  "total_improvement": 0.010267068164717882,
+  "reason": "block_end",
+  "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_immicro/K1000/fullsat/checkpoint_ft_s40000.pt",
+  "checkpoint_sha256": "cb2a07994c9c8c6ebaf9a9b2f424d93a57fc6080b3dd081b288dc3a3c7956e03"
+ },
+ {
+  "milestone_step": 60000,
+  "selected_step": 60000,
+  "selected_inner_val_loss": 0.10440332228662791,
+  "block_improvement": 0.0001841854611380339,
+  "total_improvement": 0.010451253625855916,
+  "reason": "block_end",
+  "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_immicro/K1000/fullsat/checkpoint_ft_s60000.pt",
+  "checkpoint_sha256": "ed28603294a1a617e007b44d2f58dbc050869a6879bc2a73c7118f59c3fb26d3"
+ },
+ {
+  "milestone_step": 65000,
+  "selected_step": 63000,
+  "selected_inner_val_loss": 0.10436773711897007,
+  "block_improvement": 3.55851676578367e-05,
+  "total_improvement": 0.010486838793513753,
+  "reason": "early_stop",
+  "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_immicro/K1000/fullsat/checkpoint_ft_s65000.pt",
+  "checkpoint_sha256": "5c49c8ecff8e594daae931ae784fa3b289404ca3d1bcf4dfbb3cd9cc34f58be5"
+ }
+]

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

@@ -0,0 +1,267 @@
+2026-10-03T12:10:55 split=hoL_immicro K=1000 mode=full trainable=54,185,184 balance=cohort ye2026_sepsis: train 73 rows / 529,816 cells, inner 9; 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
+2026-10-03T12:10:57 step 0 inner-val loss 0.11485 (pretrained) {'ye2026_sepsis': 0.12609728963838684, 'imperial_eskd': 0.1086193590678952, 'liu2021': 0.10984707903116941}
+2026-10-03T12:11:10 step 250 train 0.10661 inner-val 0.11294 best 0.11294@250 (12s) ye2026_sepsis=0.1242 imperial_eskd=0.1065 liu2021=0.1081
+2026-10-03T12:11:23 step 500 train 0.10368 inner-val 0.11051 best 0.11051@500 (24s) ye2026_sepsis=0.1216 imperial_eskd=0.1039 liu2021=0.1060
+2026-10-03T12:11:37 step 750 train 0.10343 inner-val 0.10880 best 0.10880@750 (38s) ye2026_sepsis=0.1196 imperial_eskd=0.1022 liu2021=0.1047
+2026-10-03T12:11:51 step 1000 train 0.10294 inner-val 0.10774 best 0.10774@1000 (52s) ye2026_sepsis=0.1182 imperial_eskd=0.1012 liu2021=0.1038
+2026-10-03T12:12:05 step 1250 train 0.10313 inner-val 0.10709 best 0.10709@1250 (67s) ye2026_sepsis=0.1174 imperial_eskd=0.1006 liu2021=0.1033
+2026-10-03T12:12:18 step 1500 train 0.10295 inner-val 0.10668 best 0.10668@1500 (79s) ye2026_sepsis=0.1169 imperial_eskd=0.1002 liu2021=0.1030
+2026-10-03T12:12:31 step 1750 train 0.10274 inner-val 0.10641 best 0.10641@1750 (92s) ye2026_sepsis=0.1165 imperial_eskd=0.0999 liu2021=0.1028
+2026-10-03T12:12:44 step 2000 train 0.10254 inner-val 0.10622 best 0.10622@2000 (105s) ye2026_sepsis=0.1163 imperial_eskd=0.0997 liu2021=0.1026
+2026-10-03T12:12:57 step 2250 train 0.10253 inner-val 0.10608 best 0.10608@2250 (118s) ye2026_sepsis=0.1161 imperial_eskd=0.0996 liu2021=0.1025
+2026-10-03T12:13:11 step 2500 train 0.10302 inner-val 0.10597 best 0.10597@2500 (132s) ye2026_sepsis=0.1160 imperial_eskd=0.0995 liu2021=0.1025
+2026-10-03T12:13:24 step 2750 train 0.10246 inner-val 0.10588 best 0.10588@2750 (145s) ye2026_sepsis=0.1158 imperial_eskd=0.0994 liu2021=0.1024
+2026-10-03T12:13:37 step 3000 train 0.10245 inner-val 0.10581 best 0.10581@3000 (159s) ye2026_sepsis=0.1157 imperial_eskd=0.0993 liu2021=0.1023
+2026-10-03T12:13:50 step 3250 train 0.10277 inner-val 0.10575 best 0.10575@3250 (172s) ye2026_sepsis=0.1157 imperial_eskd=0.0993 liu2021=0.1023
+2026-10-03T12:14:04 step 3500 train 0.10267 inner-val 0.10570 best 0.10570@3500 (186s) ye2026_sepsis=0.1156 imperial_eskd=0.0992 liu2021=0.1023
+2026-10-03T12:14:19 step 3750 train 0.10231 inner-val 0.10565 best 0.10565@3750 (200s) ye2026_sepsis=0.1155 imperial_eskd=0.0992 liu2021=0.1022
+2026-10-03T12:14:33 step 4000 train 0.10239 inner-val 0.10561 best 0.10561@4000 (214s) ye2026_sepsis=0.1155 imperial_eskd=0.0992 liu2021=0.1022
+2026-10-03T12:14:47 step 4250 train 0.10236 inner-val 0.10558 best 0.10558@4250 (228s) ye2026_sepsis=0.1154 imperial_eskd=0.0991 liu2021=0.1022
+2026-10-03T12:15:01 step 4500 train 0.10238 inner-val 0.10555 best 0.10555@4500 (243s) ye2026_sepsis=0.1154 imperial_eskd=0.0991 liu2021=0.1022
+2026-10-03T12:15:15 step 4750 train 0.10236 inner-val 0.10552 best 0.10552@4750 (257s) ye2026_sepsis=0.1153 imperial_eskd=0.0991 liu2021=0.1022
+2026-10-03T12:15:30 step 5000 train 0.10234 inner-val 0.10549 best 0.10549@5000 (271s) ye2026_sepsis=0.1153 imperial_eskd=0.0990 liu2021=0.1021
+2026-10-03T12:15:44 step 5250 train 0.10226 inner-val 0.10547 best 0.10547@5250 (285s) ye2026_sepsis=0.1153 imperial_eskd=0.0990 liu2021=0.1021
+2026-10-03T12:15:58 step 5500 train 0.10229 inner-val 0.10544 best 0.10544@5500 (299s) ye2026_sepsis=0.1152 imperial_eskd=0.0990 liu2021=0.1021
+2026-10-03T12:16:11 step 5750 train 0.10189 inner-val 0.10542 best 0.10542@5750 (313s) ye2026_sepsis=0.1152 imperial_eskd=0.0990 liu2021=0.1021
+2026-10-03T12:16:24 step 6000 train 0.10211 inner-val 0.10540 best 0.10540@6000 (325s) ye2026_sepsis=0.1152 imperial_eskd=0.0990 liu2021=0.1021
+2026-10-03T12:16:37 step 6250 train 0.10222 inner-val 0.10538 best 0.10538@6250 (339s) ye2026_sepsis=0.1151 imperial_eskd=0.0989 liu2021=0.1021
+2026-10-03T12:16:51 step 6500 train 0.10223 inner-val 0.10536 best 0.10536@6500 (353s) ye2026_sepsis=0.1151 imperial_eskd=0.0989 liu2021=0.1021
+2026-10-03T12:17:05 step 6750 train 0.10216 inner-val 0.10534 best 0.10534@6750 (367s) ye2026_sepsis=0.1151 imperial_eskd=0.0989 liu2021=0.1020
+2026-10-03T12:17:19 step 7000 train 0.10196 inner-val 0.10533 best 0.10533@7000 (380s) ye2026_sepsis=0.1151 imperial_eskd=0.0989 liu2021=0.1020
+2026-10-03T12:17:32 step 7250 train 0.10222 inner-val 0.10531 best 0.10531@7250 (394s) ye2026_sepsis=0.1151 imperial_eskd=0.0989 liu2021=0.1020
+2026-10-03T12:17:46 step 7500 train 0.10208 inner-val 0.10530 best 0.10530@7500 (407s) ye2026_sepsis=0.1150 imperial_eskd=0.0988 liu2021=0.1020
+2026-10-03T12:17:59 step 7750 train 0.10203 inner-val 0.10528 best 0.10528@7750 (421s) ye2026_sepsis=0.1150 imperial_eskd=0.0988 liu2021=0.1020
+2026-10-03T12:18:12 step 8000 train 0.10165 inner-val 0.10527 best 0.10527@8000 (434s) ye2026_sepsis=0.1150 imperial_eskd=0.0988 liu2021=0.1020
+2026-10-03T12:18:24 step 8250 train 0.10204 inner-val 0.10526 best 0.10526@8250 (446s) ye2026_sepsis=0.1150 imperial_eskd=0.0988 liu2021=0.1020
+2026-10-03T12:18:39 step 8500 train 0.10171 inner-val 0.10525 best 0.10525@8500 (460s) ye2026_sepsis=0.1150 imperial_eskd=0.0988 liu2021=0.1020
+2026-10-03T12:18:53 step 8750 train 0.10231 inner-val 0.10523 best 0.10523@8750 (474s) ye2026_sepsis=0.1150 imperial_eskd=0.0988 liu2021=0.1020
+2026-10-03T12:19:07 step 9000 train 0.10210 inner-val 0.10522 best 0.10522@9000 (488s) ye2026_sepsis=0.1150 imperial_eskd=0.0988 liu2021=0.1020
+2026-10-03T12:19:21 step 9250 train 0.10194 inner-val 0.10521 best 0.10521@9250 (503s) ye2026_sepsis=0.1149 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:19:35 step 9500 train 0.10194 inner-val 0.10520 best 0.10520@9500 (517s) ye2026_sepsis=0.1149 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:19:49 step 9750 train 0.10197 inner-val 0.10519 best 0.10519@9750 (531s) ye2026_sepsis=0.1149 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:20:03 step 10000 train 0.10190 inner-val 0.10518 best 0.10518@10000 (545s) ye2026_sepsis=0.1149 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:20:17 step 10250 train 0.10197 inner-val 0.10517 best 0.10517@10250 (559s) ye2026_sepsis=0.1149 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:20:31 step 10500 train 0.10213 inner-val 0.10517 best 0.10517@10500 (572s) ye2026_sepsis=0.1149 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:20:45 step 10750 train 0.10199 inner-val 0.10515 best 0.10515@10750 (587s) ye2026_sepsis=0.1149 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:20:59 step 11000 train 0.10176 inner-val 0.10514 best 0.10514@11000 (600s) ye2026_sepsis=0.1148 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:21:11 step 11250 train 0.10177 inner-val 0.10513 best 0.10513@11250 (613s) ye2026_sepsis=0.1148 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:21:25 step 11500 train 0.10178 inner-val 0.10512 best 0.10512@11500 (627s) ye2026_sepsis=0.1148 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:21:39 step 11750 train 0.10174 inner-val 0.10512 best 0.10512@11750 (641s) ye2026_sepsis=0.1148 imperial_eskd=0.0987 liu2021=0.1019
+2026-10-03T12:21:53 step 12000 train 0.10153 inner-val 0.10511 best 0.10511@12000 (654s) ye2026_sepsis=0.1148 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T12:22:06 step 12250 train 0.10173 inner-val 0.10510 best 0.10510@12250 (668s) ye2026_sepsis=0.1148 imperial_eskd=0.0986 liu2021=0.1019
+2026-10-03T12:22:20 step 12500 train 0.10192 inner-val 0.10509 best 0.10509@12500 (682s) ye2026_sepsis=0.1148 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:22:33 step 12750 train 0.10169 inner-val 0.10508 best 0.10508@12750 (695s) ye2026_sepsis=0.1148 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:22:47 step 13000 train 0.10194 inner-val 0.10508 best 0.10508@13000 (709s) ye2026_sepsis=0.1148 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:23:01 step 13250 train 0.10184 inner-val 0.10507 best 0.10507@13250 (723s) ye2026_sepsis=0.1148 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:23:15 step 13500 train 0.10162 inner-val 0.10506 best 0.10506@13500 (737s) ye2026_sepsis=0.1148 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:23:29 step 13750 train 0.10185 inner-val 0.10505 best 0.10505@13750 (751s) ye2026_sepsis=0.1147 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:23:42 step 14000 train 0.10175 inner-val 0.10505 best 0.10505@14000 (764s) ye2026_sepsis=0.1147 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:23:54 step 14250 train 0.10135 inner-val 0.10504 best 0.10504@14250 (776s) ye2026_sepsis=0.1147 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:24:08 step 14500 train 0.10167 inner-val 0.10503 best 0.10503@14500 (790s) ye2026_sepsis=0.1147 imperial_eskd=0.0986 liu2021=0.1018
+2026-10-03T12:24:22 step 14750 train 0.10195 inner-val 0.10502 best 0.10502@14750 (804s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:24:36 step 15000 train 0.10191 inner-val 0.10501 best 0.10501@15000 (818s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:24:50 step 15250 train 0.10120 inner-val 0.10501 best 0.10501@15250 (832s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:25:04 step 15500 train 0.10177 inner-val 0.10500 best 0.10500@15500 (846s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:25:18 step 15750 train 0.10150 inner-val 0.10499 best 0.10499@15750 (860s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:25:32 step 16000 train 0.10147 inner-val 0.10499 best 0.10499@16000 (874s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:25:46 step 16250 train 0.10177 inner-val 0.10498 best 0.10498@16250 (888s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:26:00 step 16500 train 0.10169 inner-val 0.10498 best 0.10498@16500 (902s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:26:14 step 16750 train 0.10156 inner-val 0.10498 best 0.10498@16750 (916s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:26:28 step 17000 train 0.10184 inner-val 0.10497 best 0.10497@17000 (930s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:26:42 step 17250 train 0.10183 inner-val 0.10496 best 0.10496@17250 (944s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1018
+2026-10-03T12:26:56 step 17500 train 0.10190 inner-val 0.10496 best 0.10496@17500 (958s) ye2026_sepsis=0.1147 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T12:27:10 step 17750 train 0.10131 inner-val 0.10495 best 0.10495@17750 (972s) ye2026_sepsis=0.1146 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T12:27:23 step 18000 train 0.10195 inner-val 0.10494 best 0.10494@18000 (984s) ye2026_sepsis=0.1146 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T12:27:36 step 18250 train 0.10138 inner-val 0.10494 best 0.10494@18250 (997s) ye2026_sepsis=0.1146 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T12:27:49 step 18500 train 0.10167 inner-val 0.10493 best 0.10493@18500 (1011s) ye2026_sepsis=0.1146 imperial_eskd=0.0985 liu2021=0.1017
+2026-10-03T12:28:02 step 18750 train 0.10137 inner-val 0.10493 best 0.10493@18750 (1023s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:28:15 step 19000 train 0.10153 inner-val 0.10492 best 0.10492@19000 (1036s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:28:27 step 19250 train 0.10141 inner-val 0.10491 best 0.10491@19250 (1049s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:28:42 step 19500 train 0.10156 inner-val 0.10491 best 0.10491@19500 (1063s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:28:56 step 19750 train 0.10142 inner-val 0.10490 best 0.10490@19750 (1077s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:29:10 step 20000 train 0.10142 inner-val 0.10490 best 0.10490@20000 (1091s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:29:11 block end 20000: best 0.10490@20000; block improvement 0.009959 (100.00 % of total 0.009959) -> continue
+2026-10-03T12:29:26 step 20250 train 0.10151 inner-val 0.10489 best 0.10489@20250 (1107s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:29:40 step 20500 train 0.10129 inner-val 0.10488 best 0.10488@20500 (1121s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:29:54 step 20750 train 0.10141 inner-val 0.10488 best 0.10488@20750 (1136s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:30:08 step 21000 train 0.10164 inner-val 0.10488 best 0.10488@21000 (1150s) ye2026_sepsis=0.1146 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:30:22 step 21250 train 0.10117 inner-val 0.10487 best 0.10487@21250 (1163s) ye2026_sepsis=0.1145 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:30:35 step 21500 train 0.10145 inner-val 0.10486 best 0.10486@21500 (1177s) ye2026_sepsis=0.1145 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:30:49 step 21750 train 0.10146 inner-val 0.10486 best 0.10486@21750 (1191s) ye2026_sepsis=0.1145 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:31:03 step 22000 train 0.10138 inner-val 0.10485 best 0.10485@22000 (1205s) ye2026_sepsis=0.1145 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:31:17 step 22250 train 0.10149 inner-val 0.10485 best 0.10485@22250 (1219s) ye2026_sepsis=0.1145 imperial_eskd=0.0984 liu2021=0.1017
+2026-10-03T12:31:31 step 22500 train 0.10136 inner-val 0.10484 best 0.10484@22500 (1233s) ye2026_sepsis=0.1145 imperial_eskd=0.0984 liu2021=0.1016
+2026-10-03T12:31:45 step 22750 train 0.10131 inner-val 0.10484 best 0.10484@22750 (1247s) ye2026_sepsis=0.1145 imperial_eskd=0.0984 liu2021=0.1016
+2026-10-03T12:31:59 step 23000 train 0.10173 inner-val 0.10483 best 0.10483@23000 (1260s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:32:13 step 23250 train 0.10163 inner-val 0.10483 best 0.10483@23250 (1274s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:32:27 step 23500 train 0.10143 inner-val 0.10483 best 0.10483@23500 (1288s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:32:40 step 23750 train 0.10129 inner-val 0.10482 best 0.10482@23750 (1302s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:32:54 step 24000 train 0.10178 inner-val 0.10482 best 0.10482@24000 (1316s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:33:08 step 24250 train 0.10120 inner-val 0.10481 best 0.10481@24250 (1330s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:33:22 step 24500 train 0.10122 inner-val 0.10480 best 0.10480@24500 (1344s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:33:36 step 24750 train 0.10142 inner-val 0.10479 best 0.10479@24750 (1358s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:33:50 step 25000 train 0.10127 inner-val 0.10479 best 0.10479@25000 (1372s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:34:04 step 25250 train 0.10135 inner-val 0.10479 best 0.10479@25250 (1385s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:34:18 step 25500 train 0.10110 inner-val 0.10478 best 0.10478@25500 (1399s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:34:32 step 25750 train 0.10136 inner-val 0.10478 best 0.10478@25750 (1414s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:34:46 step 26000 train 0.10148 inner-val 0.10477 best 0.10477@26000 (1428s) ye2026_sepsis=0.1145 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:35:01 step 26250 train 0.10130 inner-val 0.10477 best 0.10477@26250 (1442s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:35:15 step 26500 train 0.10138 inner-val 0.10476 best 0.10476@26500 (1456s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:35:28 step 26750 train 0.10085 inner-val 0.10476 best 0.10476@26750 (1470s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:35:42 step 27000 train 0.10136 inner-val 0.10476 best 0.10476@27000 (1484s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:35:56 step 27250 train 0.10138 inner-val 0.10475 best 0.10475@27250 (1498s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:36:10 step 27500 train 0.10148 inner-val 0.10475 best 0.10475@27500 (1511s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:36:24 step 27750 train 0.10151 inner-val 0.10475 best 0.10475@27750 (1525s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1016
+2026-10-03T12:36:38 step 28000 train 0.10121 inner-val 0.10474 best 0.10474@28000 (1540s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1015
+2026-10-03T12:36:51 step 28250 train 0.10122 inner-val 0.10475 best 0.10474@28000 (1554s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1015
+2026-10-03T12:37:04 step 28500 train 0.10107 inner-val 0.10474 best 0.10474@28500 (1566s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1015
+2026-10-03T12:37:17 step 28750 train 0.10119 inner-val 0.10473 best 0.10473@28750 (1578s) ye2026_sepsis=0.1144 imperial_eskd=0.0983 liu2021=0.1015
+2026-10-03T12:37:29 step 29000 train 0.10115 inner-val 0.10473 best 0.10473@29000 (1591s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:37:41 step 29250 train 0.10162 inner-val 0.10473 best 0.10473@29000 (1604s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:37:53 step 29500 train 0.10093 inner-val 0.10473 best 0.10473@29000 (1616s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:38:07 step 29750 train 0.10114 inner-val 0.10472 best 0.10472@29750 (1628s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:38:19 step 30000 train 0.10078 inner-val 0.10472 best 0.10472@30000 (1641s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:38:32 step 30250 train 0.10102 inner-val 0.10472 best 0.10472@30250 (1654s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:38:45 step 30500 train 0.10109 inner-val 0.10471 best 0.10471@30500 (1666s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:38:57 step 30750 train 0.10103 inner-val 0.10471 best 0.10471@30750 (1679s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:39:10 step 31000 train 0.10099 inner-val 0.10471 best 0.10471@31000 (1692s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:39:22 step 31250 train 0.10101 inner-val 0.10471 best 0.10471@31000 (1705s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:39:35 step 31500 train 0.10102 inner-val 0.10470 best 0.10470@31500 (1716s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:39:48 step 31750 train 0.10099 inner-val 0.10469 best 0.10469@31750 (1729s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:40:01 step 32000 train 0.10104 inner-val 0.10469 best 0.10469@32000 (1743s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:40:13 step 32250 train 0.10102 inner-val 0.10469 best 0.10469@32250 (1755s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:40:26 step 32500 train 0.10114 inner-val 0.10469 best 0.10469@32500 (1768s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:40:39 step 32750 train 0.10107 inner-val 0.10468 best 0.10468@32750 (1781s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:40:53 step 33000 train 0.10080 inner-val 0.10468 best 0.10468@33000 (1795s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:41:07 step 33250 train 0.10111 inner-val 0.10467 best 0.10467@33250 (1808s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:41:21 step 33500 train 0.10058 inner-val 0.10467 best 0.10467@33500 (1822s) ye2026_sepsis=0.1144 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:41:35 step 33750 train 0.10092 inner-val 0.10467 best 0.10467@33750 (1836s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:41:48 step 34000 train 0.10119 inner-val 0.10466 best 0.10466@34000 (1850s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:42:02 step 34250 train 0.10132 inner-val 0.10466 best 0.10466@34250 (1864s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:42:16 step 34500 train 0.10131 inner-val 0.10466 best 0.10466@34500 (1878s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:42:30 step 34750 train 0.10112 inner-val 0.10465 best 0.10465@34750 (1892s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:42:44 step 35000 train 0.10119 inner-val 0.10465 best 0.10465@35000 (1906s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:42:58 step 35250 train 0.10117 inner-val 0.10465 best 0.10465@35250 (1920s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:43:12 step 35500 train 0.10139 inner-val 0.10465 best 0.10465@35500 (1934s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:43:26 step 35750 train 0.10098 inner-val 0.10464 best 0.10464@35750 (1948s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1015
+2026-10-03T12:43:41 step 36000 train 0.10123 inner-val 0.10464 best 0.10464@36000 (1962s) ye2026_sepsis=0.1143 imperial_eskd=0.0982 liu2021=0.1014
+2026-10-03T12:43:55 step 36250 train 0.10112 inner-val 0.10463 best 0.10463@36250 (1977s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:44:10 step 36500 train 0.10068 inner-val 0.10463 best 0.10463@36500 (1991s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:44:24 step 36750 train 0.10129 inner-val 0.10462 best 0.10462@36750 (2006s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:44:38 step 37000 train 0.10067 inner-val 0.10462 best 0.10462@37000 (2019s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:44:50 step 37250 train 0.10105 inner-val 0.10462 best 0.10462@37000 (2034s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:45:05 step 37500 train 0.10110 inner-val 0.10461 best 0.10461@37500 (2046s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:45:17 step 37750 train 0.10112 inner-val 0.10461 best 0.10461@37500 (2060s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:45:31 step 38000 train 0.10113 inner-val 0.10461 best 0.10461@38000 (2072s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:45:45 step 38250 train 0.10120 inner-val 0.10460 best 0.10460@38250 (2086s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:45:58 step 38500 train 0.10101 inner-val 0.10460 best 0.10460@38500 (2100s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:46:12 step 38750 train 0.10098 inner-val 0.10460 best 0.10460@38750 (2114s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:46:25 step 39000 train 0.10086 inner-val 0.10460 best 0.10460@38750 (2128s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:46:39 step 39250 train 0.10096 inner-val 0.10459 best 0.10459@39250 (2140s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:46:53 step 39500 train 0.10111 inner-val 0.10459 best 0.10459@39500 (2154s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:47:07 step 39750 train 0.10095 inner-val 0.10459 best 0.10459@39750 (2168s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:47:21 step 40000 train 0.10080 inner-val 0.10459 best 0.10459@40000 (2183s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:47:23 block end 40000: best 0.10459@40000; block improvement 0.000308 (3.00 % of total 0.010267) -> continue
+2026-10-03T12:47:35 step 40250 train 0.10122 inner-val 0.10459 best 0.10459@40000 (2198s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:47:49 step 40500 train 0.10121 inner-val 0.10458 best 0.10458@40500 (2211s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:48:03 step 40750 train 0.10079 inner-val 0.10458 best 0.10458@40750 (2224s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:48:17 step 41000 train 0.10096 inner-val 0.10458 best 0.10458@41000 (2238s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:48:31 step 41250 train 0.10099 inner-val 0.10457 best 0.10457@41250 (2252s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:48:43 step 41500 train 0.10099 inner-val 0.10457 best 0.10457@41250 (2266s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:48:56 step 41750 train 0.10081 inner-val 0.10457 best 0.10457@41250 (2279s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:49:09 step 42000 train 0.10084 inner-val 0.10457 best 0.10457@42000 (2291s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:49:23 step 42250 train 0.10055 inner-val 0.10457 best 0.10457@42250 (2305s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:49:37 step 42500 train 0.10112 inner-val 0.10456 best 0.10456@42500 (2319s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:49:51 step 42750 train 0.10071 inner-val 0.10456 best 0.10456@42750 (2333s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:50:04 step 43000 train 0.10072 inner-val 0.10456 best 0.10456@42750 (2347s) ye2026_sepsis=0.1143 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:50:18 step 43250 train 0.10119 inner-val 0.10456 best 0.10456@43250 (2360s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:50:30 step 43500 train 0.10085 inner-val 0.10456 best 0.10456@43250 (2373s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:50:44 step 43750 train 0.10134 inner-val 0.10456 best 0.10456@43750 (2386s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1014
+2026-10-03T12:50:58 step 44000 train 0.10072 inner-val 0.10455 best 0.10455@44000 (2400s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:51:12 step 44250 train 0.10073 inner-val 0.10454 best 0.10454@44250 (2413s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:51:24 step 44500 train 0.10090 inner-val 0.10454 best 0.10454@44250 (2427s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:51:36 step 44750 train 0.10088 inner-val 0.10454 best 0.10454@44250 (2440s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:51:51 step 45000 train 0.10077 inner-val 0.10454 best 0.10454@45000 (2452s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:52:04 step 45250 train 0.10096 inner-val 0.10453 best 0.10453@45250 (2466s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:52:18 step 45500 train 0.10130 inner-val 0.10453 best 0.10453@45500 (2480s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:52:32 step 45750 train 0.10098 inner-val 0.10453 best 0.10453@45750 (2494s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:52:44 step 46000 train 0.10103 inner-val 0.10453 best 0.10453@45750 (2507s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:52:57 step 46250 train 0.10065 inner-val 0.10453 best 0.10453@45750 (2520s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:53:11 step 46500 train 0.10093 inner-val 0.10452 best 0.10452@46500 (2532s) ye2026_sepsis=0.1142 imperial_eskd=0.0981 liu2021=0.1013
+2026-10-03T12:53:24 step 46750 train 0.10078 inner-val 0.10452 best 0.10452@46750 (2546s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:53:38 step 47000 train 0.10083 inner-val 0.10452 best 0.10452@47000 (2560s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:53:52 step 47250 train 0.10087 inner-val 0.10452 best 0.10452@47250 (2574s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:54:05 step 47500 train 0.10091 inner-val 0.10452 best 0.10452@47250 (2588s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:54:18 step 47750 train 0.10067 inner-val 0.10451 best 0.10451@47750 (2600s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:54:32 step 48000 train 0.10084 inner-val 0.10451 best 0.10451@48000 (2614s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:54:46 step 48250 train 0.10097 inner-val 0.10450 best 0.10450@48250 (2628s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:55:00 step 48500 train 0.10060 inner-val 0.10450 best 0.10450@48500 (2642s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:55:14 step 48750 train 0.10059 inner-val 0.10450 best 0.10450@48750 (2655s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:55:26 step 49000 train 0.10085 inner-val 0.10450 best 0.10450@48750 (2669s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:55:40 step 49250 train 0.10082 inner-val 0.10450 best 0.10450@49250 (2682s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:55:54 step 49500 train 0.10064 inner-val 0.10449 best 0.10449@49500 (2695s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:56:08 step 49750 train 0.10082 inner-val 0.10449 best 0.10449@49750 (2709s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:56:20 step 50000 train 0.10035 inner-val 0.10449 best 0.10449@49750 (2723s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:56:34 step 50250 train 0.10068 inner-val 0.10448 best 0.10448@50250 (2736s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:56:47 step 50500 train 0.10094 inner-val 0.10448 best 0.10448@50500 (2749s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:57:01 step 50750 train 0.10044 inner-val 0.10447 best 0.10447@50750 (2763s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:57:15 step 51000 train 0.10081 inner-val 0.10447 best 0.10447@51000 (2777s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:57:29 step 51250 train 0.10066 inner-val 0.10447 best 0.10447@51250 (2790s) ye2026_sepsis=0.1142 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:57:43 step 51500 train 0.10068 inner-val 0.10446 best 0.10446@51500 (2804s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:57:56 step 51750 train 0.10071 inner-val 0.10446 best 0.10446@51750 (2818s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:58:10 step 52000 train 0.10093 inner-val 0.10446 best 0.10446@52000 (2832s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:58:24 step 52250 train 0.10075 inner-val 0.10445 best 0.10445@52250 (2846s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:58:38 step 52500 train 0.10112 inner-val 0.10445 best 0.10445@52500 (2859s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:58:50 step 52750 train 0.10082 inner-val 0.10445 best 0.10445@52500 (2873s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:59:03 step 53000 train 0.10057 inner-val 0.10445 best 0.10445@52500 (2886s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:59:15 step 53250 train 0.10080 inner-val 0.10445 best 0.10445@52500 (2898s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:59:28 step 53500 train 0.10050 inner-val 0.10445 best 0.10445@52500 (2911s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:59:40 step 53750 train 0.10083 inner-val 0.10445 best 0.10445@52500 (2923s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1013
+2026-10-03T12:59:54 step 54000 train 0.10058 inner-val 0.10445 best 0.10445@54000 (2935s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:00:07 step 54250 train 0.10090 inner-val 0.10444 best 0.10444@54250 (2949s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:00:21 step 54500 train 0.10090 inner-val 0.10444 best 0.10444@54500 (2963s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:00:34 step 54750 train 0.10038 inner-val 0.10444 best 0.10444@54500 (2977s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:00:47 step 55000 train 0.10090 inner-val 0.10444 best 0.10444@55000 (2989s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:01:01 step 55250 train 0.10053 inner-val 0.10443 best 0.10443@55250 (3003s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:01:14 step 55500 train 0.10083 inner-val 0.10443 best 0.10443@55250 (3017s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:01:28 step 55750 train 0.10076 inner-val 0.10443 best 0.10443@55750 (3029s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:01:41 step 56000 train 0.10049 inner-val 0.10443 best 0.10443@56000 (3043s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:01:55 step 56250 train 0.10084 inner-val 0.10442 best 0.10442@56250 (3057s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:02:08 step 56500 train 0.10023 inner-val 0.10442 best 0.10442@56250 (3071s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:02:21 step 56750 train 0.10035 inner-val 0.10442 best 0.10442@56750 (3083s) ye2026_sepsis=0.1141 imperial_eskd=0.0980 liu2021=0.1012
+2026-10-03T13:02:35 step 57000 train 0.10044 inner-val 0.10442 best 0.10442@57000 (3097s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:02:48 step 57250 train 0.10078 inner-val 0.10441 best 0.10442@57000 (3111s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:03:01 step 57500 train 0.10071 inner-val 0.10441 best 0.10441@57500 (3123s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:03:16 step 57750 train 0.10034 inner-val 0.10441 best 0.10441@57750 (3137s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:03:28 step 58000 train 0.10072 inner-val 0.10441 best 0.10441@57750 (3151s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:03:41 step 58250 train 0.10098 inner-val 0.10441 best 0.10441@57750 (3164s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:03:54 step 58500 train 0.10086 inner-val 0.10441 best 0.10441@57750 (3177s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:04:06 step 58750 train 0.10081 inner-val 0.10441 best 0.10441@57750 (3189s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:04:18 step 59000 train 0.10090 inner-val 0.10441 best 0.10441@57750 (3202s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:04:32 step 59250 train 0.10034 inner-val 0.10441 best 0.10441@59250 (3214s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:04:45 step 59500 train 0.10067 inner-val 0.10440 best 0.10441@59250 (3228s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:04:57 step 59750 train 0.10070 inner-val 0.10441 best 0.10441@59250 (3240s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:05:11 step 60000 train 0.10047 inner-val 0.10440 best 0.10440@60000 (3253s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:05:13 block end 60000: best 0.10440@60000; block improvement 0.000184 (1.76 % of total 0.010451) -> continue
+2026-10-03T13:05:27 step 60250 train 0.10072 inner-val 0.10440 best 0.10440@60250 (3268s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:05:41 step 60500 train 0.10052 inner-val 0.10440 best 0.10440@60500 (3282s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:05:55 step 60750 train 0.10050 inner-val 0.10439 best 0.10439@60750 (3297s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:06:08 step 61000 train 0.10076 inner-val 0.10439 best 0.10439@60750 (3311s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:06:21 step 61250 train 0.10092 inner-val 0.10440 best 0.10439@60750 (3324s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:06:33 step 61500 train 0.10046 inner-val 0.10439 best 0.10439@60750 (3336s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:06:48 step 61750 train 0.10069 inner-val 0.10439 best 0.10439@61750 (3349s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:07:02 step 62000 train 0.10067 inner-val 0.10439 best 0.10439@62000 (3363s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:07:15 step 62250 train 0.10039 inner-val 0.10438 best 0.10438@62250 (3377s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:07:29 step 62500 train 0.10079 inner-val 0.10438 best 0.10438@62500 (3391s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:07:43 step 62750 train 0.10051 inner-val 0.10437 best 0.10437@62750 (3405s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:07:58 step 63000 train 0.10083 inner-val 0.10437 best 0.10437@63000 (3419s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:08:10 step 63250 train 0.10084 inner-val 0.10437 best 0.10437@63000 (3433s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:08:23 step 63500 train 0.10046 inner-val 0.10437 best 0.10437@63000 (3446s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:08:35 step 63750 train 0.10067 inner-val 0.10437 best 0.10437@63000 (3458s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:08:48 step 64000 train 0.10058 inner-val 0.10437 best 0.10437@63000 (3471s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:09:00 step 64250 train 0.10048 inner-val 0.10437 best 0.10437@63000 (3483s) ye2026_sepsis=0.1141 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:09:13 step 64500 train 0.10047 inner-val 0.10437 best 0.10437@63000 (3496s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:09:25 step 64750 train 0.10068 inner-val 0.10437 best 0.10437@63000 (3508s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:09:38 step 65000 train 0.10046 inner-val 0.10437 best 0.10437@63000 (3521s) ye2026_sepsis=0.1140 imperial_eskd=0.0979 liu2021=0.1012
+2026-10-03T13:09:38 early stop: no improvement in 8 evaluations
+2026-10-03T13:09:45 saved /home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_immicro/K1000/fullsat/checkpoint_ft.pt (selected step 63000, inner-val 0.10437 vs pretrained 0.11485); sha 72a9e1b9fdf5

paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_immicro/K1000/lora16/RUN.json

@@ -0,0 +1,516 @@
+{
+ "schema": "pbisc_sdd.ft_crosscohort.run.v1",
+ "source_checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/runs/f01_6k_1x_seed20260627/checkpoint_epoch_016.pt",
+ "source_checkpoint_sha256": "8c7cd0e83e2bc19fe03dfcbc084919a1e2f61be8cc8bc12f4caf543dfda91633",
+ "split": "hoL_immicro",
+ "split_file": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/splits/hoL_immicro.json",
+ "split_sha256": "d5646bee4a5f735e64f490824ca865a2cf8191234a8de40c0963cd590ebe6743",
+ "held_out": "immicro",
+ "recipe_k": 1000,
+ "balance": "cohort",
+ "mode": "lora",
+ "trainable": {
+  "mode": "lora",
+  "lora_layers": [
+   "time_encoder.0",
+   "time_encoder.2",
+   "condition_encoder.0",
+   "condition_encoder.3",
+   "context_mixer.2",
+   "down_blocks.0.input_projection",
+   "down_blocks.0.context_projection.1",
+   "down_blocks.1.input_projection",
+   "down_blocks.1.context_projection.1",
+   "down_blocks.2.input_projection",
+   "down_blocks.2.context_projection.1",
+   "down_blocks.3.input_projection",
+   "down_blocks.3.context_projection.1",
+   "up_blocks.0.input_projection",
+   "up_blocks.0.context_projection.1",
+   "up_blocks.1.input_projection",
+   "up_blocks.1.context_projection.1",
+   "up_blocks.2.input_projection",
+   "up_blocks.2.context_projection.1",
+   "output_trunk.0",
+   "value_head",
+   "active_head"
+  ],
+  "lora_rank": 16,
+  "lora_alpha": 16.0,
+  "n_trainable": 1296224,
+  "n_total_base": 54185184,
+  "trainable_tensors": 44
+ },
+ "lr": 0.0005,
+ "warmup": 200,
+ "batch_size": 256,
+ "ema_decay": 0.999,
+ "max_steps": 20000,
+ "steps_run": 20000,
+ "eval_every": 250,
+ "patience": 8,
+ "selected_step": 19750,
+ "selected_inner_val_loss": 0.10536502702130313,
+ "pretrained_inner_val_loss": 0.11485457591248383,
+ "cohorts": [
+  {
+   "cohort": "ye2026_sepsis",
+   "condition_npz": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/data/ye2026_sepsis/conditions_raw_K1000.npz",
+   "condition_npz_sha256": "ce8b61d7dcd3e62b9e63d52189cfb5bba881f87a661384277a7949a629231cb4",
+   "train_rows": [
+    "Abd-PS04",
+    "Abd-PS05",
+    "Abd-PS07",
+    "Abd-PS19",
+    "Abd-PS20",
+    "Abd-PS27",
+    "Abd-S103",
+    "Abd-S120",
+    "Abd-S123",
+    "Abd-S156",
+    "Abd-S39",
+    "Abd-S42",
+    "Abd-S47",
+    "Abd-S51",
+    "Abd-S57",
+    "Abd-S66",
+    "Abd-S79",
+    "Abd-S83",
+    "HC05",
+    "HC07",
+    "HC09",
+    "HC12",
+    "HC13",
+    "HC16",
+    "IC10",
+    "IC15",
+    "IC16",
+    "IC22",
+    "IC23",
+    "PHC01",
+    "PHC02",
+    "PHC03",
+    "PHC05",
+    "PIC01",
+    "PIC02",
+    "PIC04",
+    "PIC05",
+    "Res-PS02",
+    "Res-PS03",
+    "Res-PS12",
+    "Res-PS21",
+    "Res-PS22",
+    "Res-PS25",
+    "Res-PS26",
+    "Res-S10",
+    "Res-S125",
+    "Res-S127",
+    "Res-S136",
+    "Res-S146",
+    "Res-S149",
+    "Res-S150",
+    "Res-S164",
+    "Res-S166",
+    "Res-S177",
+    "Res-S184",
+    "Res-S19",
+    "Res-S26",
+    "Res-S40",
+    "Res-S77",
+    "Res-S87",
+    "Skin-S140",
+    "Skin-S142",
+    "Skin-S152",
+    "Skin-S38",
+    "Skin-S43",
+    "Skin-S46",
+    "Skin-S52",
+    "Skin-S74",
+    "Uri-S134",
+    "Uri-S147",
+    "Uri-S157",
+    "Uri-S20",
+    "Uri-S93"
+   ],
+   "inner_val_rows": [
+    "Abd-PS23",
+    "Abd-S102",
+    "Abd-S54",
+    "Abd-S98",
+    "HC01",
+    "IC19",
+    "Res-PS16",
+    "Res-S139",
+    "Res-S174"
+   ]
+  },
+  {
+   "cohort": "imperial_eskd",
+   "condition_npz": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/data/imperial_eskd/conditions_raw_K1000.npz",
+   "condition_npz_sha256": "41a62291e98e367d05108a0a95a859ef9ca1997f1fda4656cba58eeefbe56442",
+   "train_rows": [
+    "C20_1004",
+    "C20_1404",
+    "C21_1004",
+    "C21_1704",
+    "C21_2504",
+    "C23_2204",
+    "C31_1304",
+    "C31_2004",
+    "C31_2704",
+    "C33_1401",
+    "C33_1601",
+    "C33_1901",
+    "C34_1404",
+    "C34_2104",
+    "C34_2504",
+    "C36_1504",
+    "C36_1704",
+    "C36_2104",
+    "C40_1504",
+    "C40_2204",
+    "C40_0105",
+    "C42_1504",
+    "C42_2204",
+    "C42_2704",
+    "C58_1604",
+    "C58_2501",
+    "C58_2901",
+    "C58_0102",
+    "C58_0402",
+    "C58_1003",
+    "C60_2504",
+    "C65_1704",
+    "C65_2104",
+    "C65_2504",
+    "C69_1804",
+    "C82_2504",
+    "C85_2004",
+    "C93_2104",
+    "C93_2704",
+    "C93_0405",
+    "C101_0601",
+    "C101_1101",
+    "C101_1301",
+    "C101_1801",
+    "C101_2001",
+    "C101_2502",
+    "C103_2204",
+    "C104_2204",
+    "C106_2204",
+    "C108_2204",
+    "C108_1501",
+    "C108_1801",
+    "C108_2201",
+    "C113_0605",
+    "C123_2404",
+    "C123_2904",
+    "C123_0505",
+    "C124_2404",
+    "C124_2704",
+    "C124_0105",
+    "C126_2504",
+    "C126_3004",
+    "C126_0605",
+    "C127_2504",
+    "C127_0405",
+    "C127_1105",
+    "C132_2804",
+    "C132_0205",
+    "C132_1105",
+    "C137_3004",
+    "C137_2502",
+    "C138_3004",
+    "C138_2301",
+    "C138_3001",
+    "C138_0202",
+    "C138_0402",
+    "C140_3004",
+    "C140_2801",
+    "C140_3001",
+    "C140_0202",
+    "C140_0402",
+    "C140_0602",
+    "C140_0902",
+    "C140_1103",
+    "C141_2001",
+    "C141_2201",
+    "C141_3001",
+    "C141_0202",
+    "C145_3004",
+    "C145_0701",
+    "C145_1201",
+    "C145_1401",
+    "C145_1901",
+    "C145_2301",
+    "C146_3004",
+    "C146_1201",
+    "C146_1401",
+    "C146_1601",
+    "C146_1901",
+    "C146_2301",
+    "C146_2701",
+    "C146_2502",
+    "C152_0405",
+    "C161_1105",
+    "C164_1105",
+    "C167_1205",
+    "C168_1205",
+    "C168_2001",
+    "C168_2301",
+    "C168_3001",
+    "C168_0202",
+    "C168_0403",
+    "C169_1205",
+    "C169_2001",
+    "C169_2301",
+    "C169_0202",
+    "C169_0403",
+    "C170_1205",
+    "C170_2301",
+    "C170_0202",
+    "C170_0402",
+    "C170_0403",
+    "C187_2605",
+    "C187_1201",
+    "C187_1401",
+    "C187_1901",
+    "C187_2301",
+    "C187_2502",
+    "C190_2605",
+    "C190_2901",
+    "C190_0102",
+    "C190_0302",
+    "C190_0502",
+    "C190_0802",
+    "C190_1002",
+    "C190_0903",
+    "C193_2605",
+    "C195_2605",
+    "C197_2605",
+    "C212_2705",
+    "C214_2705",
+    "C237_2805",
+    "C238_2905",
+    "C240_2905",
+    "C246_2905",
+    "C251_2905"
+   ],
+   "inner_val_rows": [
+    "C63_1704",
+    "C63_2404",
+    "C63_2904",
+    "C128_2504",
+    "C128_0405",
+    "C128_1105",
+    "C142_3004",
+    "C147_3004",
+    "C147_1401",
+    "C147_1901",
+    "C234_2805"
+   ]
+  },
+  {
+   "cohort": "giroux2022",
+   "condition_npz": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/data/giroux2022/conditions_raw_K1000.npz",
+   "condition_npz_sha256": "b584655fcfa5fa4c0887d5dd83fcb1c5a0206da4afe5fefab96cba2272b1371c",
+   "train_rows": [
+    "450905_d0",
+    "450905_d14",
+    "0B943B_d0",
+    "0B943B_d14",
+    "7085CA_d0",
+    "7085CA_d7",
+    "7085CA_d14",
+    "2AD75E_d0",
+    "2AD75E_d7",
+    "2AD75E_d14",
+    "B85D75_d0",
+    "B85D75_d7",
+    "180E1A_d0",
+    "180E1A_d7",
+    "180E1A_d14",
+    "82CCF5_d0",
+    "82CCF5_d7",
+    "82CCF5_d14",
+    "BAAF62_d0",
+    "BAAF62_d7",
+    "BAAF62_d14",
+    "0BF51C_d0",
+    "0BF51C_d7",
+    "0BF51C_d14"
+   ],
+   "inner_val_rows": []
+  },
+  {
+   "cohort": "liu2021",
+   "condition_npz": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/data/liu2021/conditions_raw_K1000.npz",
+   "condition_npz_sha256": "ad3131eabcfb11b166ca37c7585c50abb11efb60e2a736fe62de604794535e0c",
+   "train_rows": [
+    "AA220014_HC",
+    "AA220534_HC",
+    "AA220907_HC",
+    "HGR0000071_T0",
+    "HGR0000074_T0",
+    "HGR0000083_T1",
+    "HGR0000089_T1",
+    "HGR0000091_T0",
+    "HGR0000093_T0",
+    "HGR0000136_T0",
+    "HGR0000137_T0",
+    "HGR0000141_T0",
+    "IM000040_HC",
+    "HGR0000042_T1",
+    "HGR0000069_T0",
+    "HGR0000069_T1",
+    "HGR0000092_T0",
+    "HGR0000094_T0",
+    "HGR0000124_T0",
+    "HGR0000144_T2",
+    "HGR0000391_T0",
+    "HGR0000395_T0",
+    "HGR0000418_T1",
+    "SHD4_HC",
+    "SHD7_HC",
+    "HGR0000051_T0",
+    "HGR0000051_T1",
+    "HGR0000051_T3",
+    "HGR0000101_T1",
+    "HGR0000101_T3",
+    "HGR0000102_T0",
+    "HGR0000102_T1",
+    "HGR0000134_T0",
+    "HGR0000135_T0",
+    "HGR0000135_T1",
+    "HGR0000392_T0",
+    "HGR0000429_T1",
+    "HGR0000429_T2",
+    "SHD3_HC",
+    "SHD5_HC"
+   ],
+   "inner_val_rows": [
+    "HGR0000098_T0",
+    "HGR0000143_T0",
+    "HGR0000115_T1",
+    "SHD1_HC"
+   ]
+  }
+ ],
+ "held_out_test_rows_never_loaded": [
+  "HS1",
+  "HS2",
+  "HS3",
+  "HS4",
+  "HS6",
+  "HS7",
+  "HS8",
+  "HS9",
+  "HS10",
+  "HS11",
+  "HS12",
+  "HS13",
+  "HS14",
+  "HS15",
+  "HS16",
+  "HS17",
+  "HS18",
+  "HS19",
+  "HS20",
+  "HS21",
+  "HS23",
+  "HS24",
+  "HS25",
+  "HS26",
+  "HS27",
+  "HS28",
+  "HS29",
+  "HS30",
+  "HS31",
+  "HS32",
+  "HS34",
+  "HS35",
+  "HS36",
+  "HS37",
+  "HS38",
+  "HS39",
+  "HS40",
+  "HS41",
+  "HS42",
+  "HS43",
+  "HS44",
+  "HS45",
+  "HS46",
+  "HS47",
+  "HS48",
+  "HS49",
+  "HS50",
+  "HS51",
+  "HS52",
+  "HS53",
+  "HS54",
+  "HS55",
+  "HS56",
+  "HS57",
+  "HS58",
+  "HS59",
+  "HS60",
+  "HS61",
+  "HS62",
+  "HS63",
+  "HS64",
+  "HS65",
+  "HS66",
+  "HS67",
+  "HS68",
+  "HS69",
+  "HS70",
+  "HS71",
+  "HS72",
+  "HS73",
+  "HS74",
+  "HS75",
+  "HS76",
+  "HS77",
+  "HS78",
+  "HS79",
+  "HS80",
+  "HS81",
+  "HS82",
+  "HS83",
+  "HS84",
+  "HS85",
+  "HS86",
+  "HS87",
+  "HS88",
+  "HS89",
+  "HS90",
+  "HS91",
+  "HS92",
+  "HS93",
+  "HS94",
+  "HS95",
+  "HS96",
+  "HS97",
+  "HS98",
+  "HS100",
+  "HS101",
+  "HS102",
+  "HS103",
+  "HS104",
+  "HS105",
+  "HS106",
+  "HS107",
+  "HS108",
+  "HS109",
+  "HS110",
+  "HS111"
+ ],
+ "max_abs_weight_change": 5.913187503814697,
+ "n_tensors_changed": 22,
+ "seed": 20261002,
+ "wall_seconds": 821.2,
+ "smoke": false,
+ "written_utc": "2026-10-02T14:33:59.618177+00:00",
+ "checkpoint_sha256": "8407719a28389a57ec8d8541ae43553189470ecc79d7d5fae04d83bd8ac72282",
+ "reload_forward_max_abs_diff": 0.0
+}

paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_immicro/K1000/lora16/history.json

@@ -0,0 +1,892 @@
+[
+ {
+  "step": 0,
+  "inner_val_loss": 0.11485457591248383,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.12609728963838684,
+   "imperial_eskd": 0.1086193590678952,
+   "liu2021": 0.10984707903116941
+  },
+  "train_loss": null
+ },
+ {
+  "step": 250,
+  "inner_val_loss": 0.11414438834169295,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.12543227771917978,
+   "imperial_eskd": 0.10781732811169191,
+   "liu2021": 0.10918355919420719
+  },
+  "train_loss": 0.10602236095070838,
+  "sec": 9.9
+ },
+ {
+  "step": 500,
+  "inner_val_loss": 0.11240342373548935,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.12371937392486466,
+   "imperial_eskd": 0.10594112561507658,
+   "liu2021": 0.1075497716665268
+  },
+  "train_loss": 0.10332092860341072,
+  "sec": 19.9
+ },
+ {
+  "step": 750,
+  "inner_val_loss": 0.11034579046058034,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.12138208167420493,
+   "imperial_eskd": 0.10381637344306166,
+   "liu2021": 0.10583891626447439
+  },
+  "train_loss": 0.10313452857732773,
+  "sec": 29.8
+ },
+ {
+  "step": 1000,
+  "inner_val_loss": 0.10861500124650873,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11921467921800083,
+   "imperial_eskd": 0.10211776434020563,
+   "liu2021": 0.10451256018131971
+  },
+  "train_loss": 0.10267900687456132,
+  "sec": 39.8
+ },
+ {
+  "step": 1250,
+  "inner_val_loss": 0.10749707844497179,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11774862640433842,
+   "imperial_eskd": 0.10105701197277416,
+   "liu2021": 0.10368559695780277
+  },
+  "train_loss": 0.10289887925982476,
+  "sec": 49.7
+ },
+ {
+  "step": 1500,
+  "inner_val_loss": 0.10686410504459132,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11692865068713824,
+   "imperial_eskd": 0.10043310611085458,
+   "liu2021": 0.1032305583357811
+  },
+  "train_loss": 0.10274808645248414,
+  "sec": 59.6
+ },
+ {
+  "step": 1750,
+  "inner_val_loss": 0.10646325236606578,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11642472114827898,
+   "imperial_eskd": 0.10000434788790616,
+   "liu2021": 0.1029606880620122
+  },
+  "train_loss": 0.10254462906718254,
+  "sec": 69.5
+ },
+ {
+  "step": 2000,
+  "inner_val_loss": 0.10620493388198542,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11609216406941414,
+   "imperial_eskd": 0.09972400150515816,
+   "liu2021": 0.10279863607138395
+  },
+  "train_loss": 0.102370800614357,
+  "sec": 79.5
+ },
+ {
+  "step": 2250,
+  "inner_val_loss": 0.10602658023424322,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11584894897209273,
+   "imperial_eskd": 0.0995442962104624,
+   "liu2021": 0.1026864955201745
+  },
+  "train_loss": 0.10236747318506241,
+  "sec": 89.4
+ },
+ {
+  "step": 2500,
+  "inner_val_loss": 0.10589890481797533,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11567362687653965,
+   "imperial_eskd": 0.09941506995396181,
+   "liu2021": 0.10260801762342453
+  },
+  "train_loss": 0.10288342818617821,
+  "sec": 99.1
+ },
+ {
+  "step": 2750,
+  "inner_val_loss": 0.10580066246253372,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11554202189048131,
+   "imperial_eskd": 0.0993189313872294,
+   "liu2021": 0.10254103410989046
+  },
+  "train_loss": 0.1023491935133934,
+  "sec": 109.0
+ },
+ {
+  "step": 3000,
+  "inner_val_loss": 0.10572898515221045,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11544580343696806,
+   "imperial_eskd": 0.09924783794717355,
+   "liu2021": 0.10249331407248974
+  },
+  "train_loss": 0.10235326462984085,
+  "sec": 118.9
+ },
+ {
+  "step": 3250,
+  "inner_val_loss": 0.10568079960717497,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11537433953748809,
+   "imperial_eskd": 0.09920252656394785,
+   "liu2021": 0.10246553272008896
+  },
+  "train_loss": 0.10269091421365738,
+  "sec": 128.8
+ },
+ {
+  "step": 3500,
+  "inner_val_loss": 0.10563749471923918,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11531577383478482,
+   "imperial_eskd": 0.09916334734721617,
+   "liu2021": 0.10243336297571659
+  },
+  "train_loss": 0.10259852829575539,
+  "sec": 138.6
+ },
+ {
+  "step": 3750,
+  "inner_val_loss": 0.1056082263927568,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.115277498960495,
+   "imperial_eskd": 0.0991301414641467,
+   "liu2021": 0.10241703875362873
+  },
+  "train_loss": 0.10225769150257111,
+  "sec": 148.4
+ },
+ {
+  "step": 4000,
+  "inner_val_loss": 0.10558213939833822,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11524279043078423,
+   "imperial_eskd": 0.09910677915269678,
+   "liu2021": 0.10239684861153364
+  },
+  "train_loss": 0.10236330050230026,
+  "sec": 158.2
+ },
+ {
+  "step": 4250,
+  "inner_val_loss": 0.10556400915049642,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11521883226103252,
+   "imperial_eskd": 0.09908798438581554,
+   "liu2021": 0.10238521080464125
+  },
+  "train_loss": 0.10233627352118492,
+  "sec": 168.0
+ },
+ {
+  "step": 4500,
+  "inner_val_loss": 0.10554716495251415,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11519643788536389,
+   "imperial_eskd": 0.09906585785475644,
+   "liu2021": 0.1023791991174221
+  },
+  "train_loss": 0.10235854291915894,
+  "sec": 177.9
+ },
+ {
+  "step": 4750,
+  "inner_val_loss": 0.10553413013980946,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11517886858847406,
+   "imperial_eskd": 0.09904907338998535,
+   "liu2021": 0.10237444844096899
+  },
+  "train_loss": 0.10237004610896111,
+  "sec": 187.8
+ },
+ {
+  "step": 5000,
+  "inner_val_loss": 0.10552329992756299,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151638784342342,
+   "imperial_eskd": 0.0990373200990937,
+   "liu2021": 0.10236870124936104
+  },
+  "train_loss": 0.10235879334807396,
+  "sec": 197.5
+ },
+ {
+  "step": 5250,
+  "inner_val_loss": 0.1055132161186199,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11514780546228091,
+   "imperial_eskd": 0.09902704202316025,
+   "liu2021": 0.10236480087041855
+  },
+  "train_loss": 0.10227865892648696,
+  "sec": 207.4
+ },
+ {
+  "step": 5500,
+  "inner_val_loss": 0.1055034135753659,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11513965162965986,
+   "imperial_eskd": 0.09901216964830052,
+   "liu2021": 0.10235841944813728
+  },
+  "train_loss": 0.10233222079277039,
+  "sec": 217.4
+ },
+ {
+  "step": 5750,
+  "inner_val_loss": 0.10549553676260641,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512907139129108,
+   "imperial_eskd": 0.09900357303294269,
+   "liu2021": 0.10235396586358547
+  },
+  "train_loss": 0.10193818452954292,
+  "sec": 227.2
+ },
+ {
+  "step": 6000,
+  "inner_val_loss": 0.10548896469200902,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512708332803515,
+   "imperial_eskd": 0.0989899259399284,
+   "liu2021": 0.10234988480806351
+  },
+  "train_loss": 0.10216974622011185,
+  "sec": 237.1
+ },
+ {
+  "step": 6250,
+  "inner_val_loss": 0.10548158889943703,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512675177719858,
+   "imperial_eskd": 0.09897803752259775,
+   "liu2021": 0.10233997739851475
+  },
+  "train_loss": 0.10230076742172241,
+  "sec": 247.0
+ },
+ {
+  "step": 6500,
+  "inner_val_loss": 0.10547515098005533,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512614041566849,
+   "imperial_eskd": 0.09896685928106308,
+   "liu2021": 0.10233245324343443
+  },
+  "train_loss": 0.10231144925951957,
+  "sec": 256.9
+ },
+ {
+  "step": 6750,
+  "inner_val_loss": 0.10546913769359541,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512955153981845,
+   "imperial_eskd": 0.09895387867634947,
+   "liu2021": 0.1023239828646183
+  },
+  "train_loss": 0.10225343719124794,
+  "sec": 266.8
+ },
+ {
+  "step": 7000,
+  "inner_val_loss": 0.10546116001264315,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512533451120059,
+   "imperial_eskd": 0.09894665262915871,
+   "liu2021": 0.10231149289757013
+  },
+  "train_loss": 0.1020633065700531,
+  "sec": 276.7
+ },
+ {
+  "step": 7250,
+  "inner_val_loss": 0.10545754334842317,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512686933080356,
+   "imperial_eskd": 0.09893873909657652,
+   "liu2021": 0.1023070216178894
+  },
+  "train_loss": 0.10232851794362068,
+  "sec": 287.4
+ },
+ {
+  "step": 7500,
+  "inner_val_loss": 0.1054532051632049,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512616897622745,
+   "imperial_eskd": 0.09892957183447751,
+   "liu2021": 0.10230387467890978
+  },
+  "train_loss": 0.10220364105701446,
+  "sec": 298.6
+ },
+ {
+  "step": 7750,
+  "inner_val_loss": 0.10544616896157412,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512511140770382,
+   "imperial_eskd": 0.0989149355075576,
+   "liu2021": 0.10229845996946096
+  },
+  "train_loss": 0.10216472792625428,
+  "sec": 309.8
+ },
+ {
+  "step": 8000,
+  "inner_val_loss": 0.10544278009509299,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11512821581628588,
+   "imperial_eskd": 0.09890479188073766,
+   "liu2021": 0.1022953325882554
+  },
+  "train_loss": 0.10179342824220658,
+  "sec": 321.2
+ },
+ {
+  "step": 8250,
+  "inner_val_loss": 0.10544220354997551,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151383225288656,
+   "imperial_eskd": 0.0989002524451776,
+   "liu2021": 0.1022880356758833
+  },
+  "train_loss": 0.102179168343544,
+  "sec": 332.1
+ },
+ {
+  "step": 8500,
+  "inner_val_loss": 0.10543868084585527,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11514172247714466,
+   "imperial_eskd": 0.09889226813207973,
+   "liu2021": 0.10228205192834139
+  },
+  "train_loss": 0.1018693950176239,
+  "sec": 341.9
+ },
+ {
+  "step": 8750,
+  "inner_val_loss": 0.10543396202565143,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11514601856470108,
+   "imperial_eskd": 0.09888599542054263,
+   "liu2021": 0.10226987209171057
+  },
+  "train_loss": 0.10248101574182511,
+  "sec": 351.6
+ },
+ {
+  "step": 9000,
+  "inner_val_loss": 0.10543470242396248,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11514870118763712,
+   "imperial_eskd": 0.09888359125364911,
+   "liu2021": 0.10227181483060122
+  },
+  "train_loss": 0.10226793873310089,
+  "sec": 362.4
+ },
+ {
+  "step": 9250,
+  "inner_val_loss": 0.10542872949167835,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151470500561926,
+   "imperial_eskd": 0.09886986423622478,
+   "liu2021": 0.10226927418261766
+  },
+  "train_loss": 0.10212115487456322,
+  "sec": 373.6
+ },
+ {
+  "step": 9500,
+  "inner_val_loss": 0.10542826642366972,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151539571583271,
+   "imperial_eskd": 0.09886692024090073,
+   "liu2021": 0.10226392187178135
+  },
+  "train_loss": 0.10213142389059067,
+  "sec": 383.3
+ },
+ {
+  "step": 9750,
+  "inner_val_loss": 0.10543130905452099,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515991017222404,
+   "imperial_eskd": 0.09886671907522461,
+   "liu2021": 0.10226729791611433
+  },
+  "train_loss": 0.1021658276617527,
+  "sec": 393.0
+ },
+ {
+  "step": 10000,
+  "inner_val_loss": 0.10543208207734406,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151674021449354,
+   "imperial_eskd": 0.09886749732223424,
+   "liu2021": 0.10226134676486254
+  },
+  "train_loss": 0.10209894260764121,
+  "sec": 402.8
+ },
+ {
+  "step": 10250,
+  "inner_val_loss": 0.10543248332685694,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11517156081067191,
+   "imperial_eskd": 0.09886711192401973,
+   "liu2021": 0.10225877724587917
+  },
+  "train_loss": 0.10218173372745513,
+  "sec": 413.2
+ },
+ {
+  "step": 10500,
+  "inner_val_loss": 0.10542481814696131,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151604296432601,
+   "imperial_eskd": 0.09886121174151247,
+   "liu2021": 0.10225281305611134
+  },
+  "train_loss": 0.10234721267223358,
+  "sec": 424.3
+ },
+ {
+  "step": 10750,
+  "inner_val_loss": 0.1054228725513876,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515307674805324,
+   "imperial_eskd": 0.09885951571843841,
+   "liu2021": 0.10225602518767118
+  },
+  "train_loss": 0.10221673774719238,
+  "sec": 435.3
+ },
+ {
+  "step": 11000,
+  "inner_val_loss": 0.10542102727211183,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515068014462788,
+   "imperial_eskd": 0.09886408969759941,
+   "liu2021": 0.10224831197410822
+  },
+  "train_loss": 0.10199678918719292,
+  "sec": 445.2
+ },
+ {
+  "step": 11250,
+  "inner_val_loss": 0.10542036100645058,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515592204199897,
+   "imperial_eskd": 0.0988620485771786,
+   "liu2021": 0.10224311240017414
+  },
+  "train_loss": 0.10201768520474434,
+  "sec": 455.3
+ },
+ {
+  "step": 11500,
+  "inner_val_loss": 0.10541562949580074,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151557279129823,
+   "imperial_eskd": 0.09885762496428056,
+   "liu2021": 0.10223353561013937
+  },
+  "train_loss": 0.10203186547756195,
+  "sec": 465.7
+ },
+ {
+  "step": 11750,
+  "inner_val_loss": 0.10541562218632962,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151649256547292,
+   "imperial_eskd": 0.09885498881340027,
+   "liu2021": 0.10222695209085941
+  },
+  "train_loss": 0.10199793341755867,
+  "sec": 475.6
+ },
+ {
+  "step": 12000,
+  "inner_val_loss": 0.10541427094388993,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516943035854234,
+   "imperial_eskd": 0.09884701059623198,
+   "liu2021": 0.10222637187689543
+  },
+  "train_loss": 0.10181100311875343,
+  "sec": 485.2
+ },
+ {
+  "step": 12250,
+  "inner_val_loss": 0.10541185696831958,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11517281126644877,
+   "imperial_eskd": 0.09884457222440025,
+   "liu2021": 0.1022181874141097
+  },
+  "train_loss": 0.10198975419998169,
+  "sec": 495.0
+ },
+ {
+  "step": 12500,
+  "inner_val_loss": 0.10540836088371898,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516620425714387,
+   "imperial_eskd": 0.09883916547352617,
+   "liu2021": 0.10221971292048693
+  },
+  "train_loss": 0.10219522455334663,
+  "sec": 504.7
+ },
+ {
+  "step": 12750,
+  "inner_val_loss": 0.10540560822344387,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516763228509161,
+   "imperial_eskd": 0.0988323488696055,
+   "liu2021": 0.10221684351563454
+  },
+  "train_loss": 0.10197097727656365,
+  "sec": 514.4
+ },
+ {
+  "step": 13000,
+  "inner_val_loss": 0.10540317388734333,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151683847937319,
+   "imperial_eskd": 0.09882856532931328,
+   "liu2021": 0.10221257153898478
+  },
+  "train_loss": 0.10222055840492249,
+  "sec": 524.1
+ },
+ {
+  "step": 13250,
+  "inner_val_loss": 0.10540056839798535,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516975777016746,
+   "imperial_eskd": 0.09882118823853406,
+   "liu2021": 0.10221075918525457
+  },
+  "train_loss": 0.10213792562484741,
+  "sec": 533.9
+ },
+ {
+  "step": 13500,
+  "inner_val_loss": 0.10539820979063962,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151690785255697,
+   "imperial_eskd": 0.0988198999654163,
+   "liu2021": 0.10220565088093281
+  },
+  "train_loss": 0.10193178761005402,
+  "sec": 545.0
+ },
+ {
+  "step": 13750,
+  "inner_val_loss": 0.10539732299594566,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516643439730008,
+   "imperial_eskd": 0.09881653467362578,
+   "liu2021": 0.10220899991691113
+  },
+  "train_loss": 0.10216229102015495,
+  "sec": 556.3
+ },
+ {
+  "step": 14000,
+  "inner_val_loss": 0.10539447725672112,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516422695583767,
+   "imperial_eskd": 0.09881145845759999,
+   "liu2021": 0.10220774635672569
+  },
+  "train_loss": 0.10207001346349716,
+  "sec": 566.8
+ },
+ {
+  "step": 14250,
+  "inner_val_loss": 0.10539028245755017,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515885260370043,
+   "imperial_eskd": 0.09880782562223347,
+   "liu2021": 0.1022041691467166
+  },
+  "train_loss": 0.10168200799822807,
+  "sec": 577.8
+ },
+ {
+  "step": 14500,
+  "inner_val_loss": 0.10538625630615939,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515445883075397,
+   "imperial_eskd": 0.09880769862370058,
+   "liu2021": 0.10219661146402359
+  },
+  "train_loss": 0.10201382622122765,
+  "sec": 589.0
+ },
+ {
+  "step": 14750,
+  "inner_val_loss": 0.10538591179833669,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515216819114155,
+   "imperial_eskd": 0.09880518202077258,
+   "liu2021": 0.10220038518309593
+  },
+  "train_loss": 0.10228333008289338,
+  "sec": 600.0
+ },
+ {
+  "step": 15000,
+  "inner_val_loss": 0.10538610344382648,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515264999535349,
+   "imperial_eskd": 0.09880211610685695,
+   "liu2021": 0.10220354422926903
+  },
+  "train_loss": 0.10224945750832558,
+  "sec": 610.3
+ },
+ {
+  "step": 15250,
+  "inner_val_loss": 0.10538751303974025,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515259825521046,
+   "imperial_eskd": 0.09880525110797449,
+   "liu2021": 0.1022046897560358
+  },
+  "train_loss": 0.10155924341082573,
+  "sec": 621.7
+ },
+ {
+  "step": 15500,
+  "inner_val_loss": 0.10538414901166353,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515089662538634,
+   "imperial_eskd": 0.09879969094287265,
+   "liu2021": 0.10220185946673155
+  },
+  "train_loss": 0.10213578280806541,
+  "sec": 632.9
+ },
+ {
+  "step": 15750,
+  "inner_val_loss": 0.1053860667679045,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516363794604938,
+   "imperial_eskd": 0.09879820048809052,
+   "liu2021": 0.1021963618695736
+  },
+  "train_loss": 0.10188705751299858,
+  "sec": 644.2
+ },
+ {
+  "step": 16000,
+  "inner_val_loss": 0.10538400699281031,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516176039973895,
+   "imperial_eskd": 0.09879548847675323,
+   "liu2021": 0.10219477210193872
+  },
+  "train_loss": 0.10184461978077888,
+  "sec": 654.7
+ },
+ {
+  "step": 16250,
+  "inner_val_loss": 0.10538500820534298,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515944533877903,
+   "imperial_eskd": 0.09880226376381787,
+   "liu2021": 0.10219331551343203
+  },
+  "train_loss": 0.1021460217833519,
+  "sec": 664.8
+ },
+ {
+  "step": 16500,
+  "inner_val_loss": 0.10538781408584279,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516572576430109,
+   "imperial_eskd": 0.09880549900911072,
+   "liu2021": 0.10219221748411655
+  },
+  "train_loss": 0.10209010773897172,
+  "sec": 674.5
+ },
+ {
+  "step": 16750,
+  "inner_val_loss": 0.10538657637381994,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516957191957368,
+   "imperial_eskd": 0.09880105406045914,
+   "liu2021": 0.10218910314142704
+  },
+  "train_loss": 0.10195132932066918,
+  "sec": 684.7
+ },
+ {
+  "step": 17000,
+  "inner_val_loss": 0.10538204521354701,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151627426346143,
+   "imperial_eskd": 0.09880103543400764,
+   "liu2021": 0.1021823575720191
+  },
+  "train_loss": 0.10225429475307464,
+  "sec": 694.5
+ },
+ {
+  "step": 17250,
+  "inner_val_loss": 0.10538269005251755,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516522781716453,
+   "imperial_eskd": 0.09880534999749878,
+   "liu2021": 0.10217749234288931
+  },
+  "train_loss": 0.10223693245649337,
+  "sec": 704.8
+ },
+ {
+  "step": 17500,
+  "inner_val_loss": 0.10538175629972409,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151622554494275,
+   "imperial_eskd": 0.09880792485042052,
+   "liu2021": 0.10217508859932423
+  },
+  "train_loss": 0.10230954352021217,
+  "sec": 714.7
+ },
+ {
+  "step": 17750,
+  "inner_val_loss": 0.10537756691915719,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516194376680586,
+   "imperial_eskd": 0.09880459139292891,
+   "liu2021": 0.10216616559773684
+  },
+  "train_loss": 0.10172887387871742,
+  "sec": 724.5
+ },
+ {
+  "step": 18000,
+  "inner_val_loss": 0.10537377367680423,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515356476108234,
+   "imperial_eskd": 0.09880166331475432,
+   "liu2021": 0.10216609295457602
+  },
+  "train_loss": 0.10238190338015557,
+  "sec": 734.5
+ },
+ {
+  "step": 18250,
+  "inner_val_loss": 0.10537126522686846,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515417612261242,
+   "imperial_eskd": 0.09879741580648856,
+   "liu2021": 0.10216220375150442
+  },
+  "train_loss": 0.10182084572315216,
+  "sec": 744.2
+ },
+ {
+  "step": 18500,
+  "inner_val_loss": 0.10537184036089113,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515499568647808,
+   "imperial_eskd": 0.09879844364794818,
+   "liu2021": 0.10216208174824715
+  },
+  "train_loss": 0.10211107203364372,
+  "sec": 754.3
+ },
+ {
+  "step": 18750,
+  "inner_val_loss": 0.10536940177269155,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515772135721312,
+   "imperial_eskd": 0.09879655594175513,
+   "liu2021": 0.10215392801910639
+  },
+  "train_loss": 0.10181015372276306,
+  "sec": 764.2
+ },
+ {
+  "step": 19000,
+  "inner_val_loss": 0.10536873857380967,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516206297609541,
+   "imperial_eskd": 0.0987899201837453,
+   "liu2021": 0.10215423256158829
+  },
+  "train_loss": 0.10198304224014282,
+  "sec": 774.6
+ },
+ {
+  "step": 19250,
+  "inner_val_loss": 0.1053676690925101,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11516604324181874,
+   "imperial_eskd": 0.0987897322259166,
+   "liu2021": 0.1021472318097949
+  },
+  "train_loss": 0.1018762274980545,
+  "sec": 785.2
+ },
+ {
+  "step": 19500,
+  "inner_val_loss": 0.1053670958274041,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11515933316614893,
+   "imperial_eskd": 0.09879281236366792,
+   "liu2021": 0.10214914195239544
+  },
+  "train_loss": 0.10202153268456458,
+  "sec": 796.7
+ },
+ {
+  "step": 19750,
+  "inner_val_loss": 0.10536502702130313,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.1151509251859453,
+   "imperial_eskd": 0.09879483418031172,
+   "liu2021": 0.10214932169765234
+  },
+  "train_loss": 0.10189098432660103,
+  "sec": 808.3
+ },
+ {
+  "step": 20000,
+  "inner_val_loss": 0.1053667031165827,
+  "inner_val_per_cohort": {
+   "ye2026_sepsis": 0.11514904432826573,
+   "imperial_eskd": 0.09879457307132808,
+   "liu2021": 0.1021564919501543
+  },
+  "train_loss": 0.10189062160253524,
+  "sec": 819.7
+ }
+]