Hugging Face's logo

Datasets:
junidude14
/
PBISC-data

Modalities:
Tabular
Text
Formats:
parquet
Languages:
English
Size:
< 1K
Tags:
single-cell-rna-seq
bulk-rna-seq
diffusion
Libraries:
Datasets
pandas
Polars
Dataset card Data Studio Files
xet
 Community
1
junidude14 commited on
Commit
9ade1e0
·
verified ·
1 Parent(s): 0e4833a

Archive 4 checkpoints (1-4)

Browse files
Files changed (4) hide show
  1. paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/RUN.json +556 -0
  2. paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/history.json +0 -0
  3. paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/milestones.json +62 -0
  4. paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/train.log +443 -0
paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/RUN.json ADDED
@@ -0,0 +1,556 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "schema": "pbisc_sdd.ft_crosscohort.run.v1",
3
+ "source_checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/runs/f01_6k_1x_seed20260627/checkpoint_epoch_016.pt",
4
+ "source_checkpoint_sha256": "8c7cd0e83e2bc19fe03dfcbc084919a1e2f61be8cc8bc12f4caf543dfda91633",
5
+ "split": "hoL_ye2026_sepsis",
6
+ "split_file": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/splits/hoL_ye2026_sepsis.json",
7
+ "split_sha256": "39cad7b25faf6c15cd65f0512977b9e526353cfb085dc2c45d18f8edc56db426",
8
+ "held_out": "ye2026_sepsis",
9
+ "recipe_k": 1000,
10
+ "balance": "cohort",
11
+ "mode": "full",
12
+ "trainable": {
13
+ "mode": "full",
14
+ "n_trainable": 54185184,
15
+ "n_total_base": 54185184,
16
+ "trainable_tensors": 64
17
+ },
18
+ "lr": 5e-05,
19
+ "warmup": 200,
20
+ "batch_size": 256,
21
+ "ema_decay": 0.999,
22
+ "max_steps": 200000,
23
+ "steps_run": 108500,
24
+ "eval_every": 250,
25
+ "patience": 8,
26
+ "selected_step": 106500,
27
+ "selected_inner_val_loss": 0.09843902954963422,
28
+ "pretrained_inner_val_loss": 0.10827601102716995,
29
+ "cohorts": [
30
+ {
31
+ "cohort": "immicro",
32
+ "condition_npz": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/data/immicro/conditions_raw_K1000.npz",
33
+ "condition_npz_sha256": "2be5717d19e86bdf38e4838d4bd2886a211ba78bfbfe749c458305a3e0ac2dcc",
34
+ "train_rows": [
35
+ "HS1",
36
+ "HS2",
37
+ "HS3",
38
+ "HS6",
39
+ "HS7",
40
+ "HS8",
41
+ "HS9",
42
+ "HS10",
43
+ "HS12",
44
+ "HS13",
45
+ "HS14",
46
+ "HS15",
47
+ "HS16",
48
+ "HS17",
49
+ "HS18",
50
+ "HS19",
51
+ "HS20",
52
+ "HS21",
53
+ "HS23",
54
+ "HS24",
55
+ "HS25",
56
+ "HS28",
57
+ "HS29",
58
+ "HS30",
59
+ "HS32",
60
+ "HS34",
61
+ "HS35",
62
+ "HS36",
63
+ "HS37",
64
+ "HS38",
65
+ "HS39",
66
+ "HS40",
67
+ "HS41",
68
+ "HS42",
69
+ "HS43",
70
+ "HS44",
71
+ "HS45",
72
+ "HS47",
73
+ "HS48",
74
+ "HS49",
75
+ "HS50",
76
+ "HS51",
77
+ "HS52",
78
+ "HS53",
79
+ "HS54",
80
+ "HS56",
81
+ "HS57",
82
+ "HS58",
83
+ "HS59",
84
+ "HS61",
85
+ "HS63",
86
+ "HS64",
87
+ "HS65",
88
+ "HS66",
89
+ "HS67",
90
+ "HS68",
91
+ "HS69",
92
+ "HS70",
93
+ "HS71",
94
+ "HS72",
95
+ "HS73",
96
+ "HS74",
97
+ "HS75",
98
+ "HS76",
99
+ "HS77",
100
+ "HS78",
101
+ "HS79",
102
+ "HS80",
103
+ "HS81",
104
+ "HS82",
105
+ "HS83",
106
+ "HS84",
107
+ "HS85",
108
+ "HS86",
109
+ "HS88",
110
+ "HS89",
111
+ "HS90",
112
+ "HS91",
113
+ "HS92",
114
+ "HS93",
115
+ "HS94",
116
+ "HS95",
117
+ "HS96",
118
+ "HS97",
119
+ "HS98",
120
+ "HS100",
121
+ "HS101",
122
+ "HS102",
123
+ "HS103",
124
+ "HS104",
125
+ "HS105",
126
+ "HS106",
127
+ "HS107",
128
+ "HS110",
129
+ "HS111"
130
+ ],
131
+ "inner_val_rows": [
132
+ "HS4",
133
+ "HS11",
134
+ "HS26",
135
+ "HS27",
136
+ "HS31",
137
+ "HS46",
138
+ "HS55",
139
+ "HS60",
140
+ "HS62",
141
+ "HS87",
142
+ "HS108",
143
+ "HS109"
144
+ ]
145
+ },
146
+ {
147
+ "cohort": "imperial_eskd",
148
+ "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",
149
+ "condition_npz_sha256": "41a62291e98e367d05108a0a95a859ef9ca1997f1fda4656cba58eeefbe56442",
150
+ "train_rows": [
151
+ "C20_1004",
152
+ "C20_1404",
153
+ "C21_1004",
154
+ "C21_1704",
155
+ "C21_2504",
156
+ "C23_2204",
157
+ "C31_1304",
158
+ "C31_2004",
159
+ "C31_2704",
160
+ "C33_1401",
161
+ "C33_1601",
162
+ "C33_1901",
163
+ "C34_1404",
164
+ "C34_2104",
165
+ "C34_2504",
166
+ "C36_1504",
167
+ "C36_1704",
168
+ "C36_2104",
169
+ "C40_1504",
170
+ "C40_2204",
171
+ "C40_0105",
172
+ "C42_1504",
173
+ "C42_2204",
174
+ "C42_2704",
175
+ "C58_1604",
176
+ "C58_2501",
177
+ "C58_2901",
178
+ "C58_0102",
179
+ "C58_0402",
180
+ "C58_1003",
181
+ "C60_2504",
182
+ "C65_1704",
183
+ "C65_2104",
184
+ "C65_2504",
185
+ "C69_1804",
186
+ "C82_2504",
187
+ "C85_2004",
188
+ "C93_2104",
189
+ "C93_2704",
190
+ "C93_0405",
191
+ "C101_0601",
192
+ "C101_1101",
193
+ "C101_1301",
194
+ "C101_1801",
195
+ "C101_2001",
196
+ "C101_2502",
197
+ "C103_2204",
198
+ "C104_2204",
199
+ "C106_2204",
200
+ "C108_2204",
201
+ "C108_1501",
202
+ "C108_1801",
203
+ "C108_2201",
204
+ "C113_0605",
205
+ "C123_2404",
206
+ "C123_2904",
207
+ "C123_0505",
208
+ "C124_2404",
209
+ "C124_2704",
210
+ "C124_0105",
211
+ "C126_2504",
212
+ "C126_3004",
213
+ "C126_0605",
214
+ "C127_2504",
215
+ "C127_0405",
216
+ "C127_1105",
217
+ "C132_2804",
218
+ "C132_0205",
219
+ "C132_1105",
220
+ "C137_3004",
221
+ "C137_2502",
222
+ "C138_3004",
223
+ "C138_2301",
224
+ "C138_3001",
225
+ "C138_0202",
226
+ "C138_0402",
227
+ "C140_3004",
228
+ "C140_2801",
229
+ "C140_3001",
230
+ "C140_0202",
231
+ "C140_0402",
232
+ "C140_0602",
233
+ "C140_0902",
234
+ "C140_1103",
235
+ "C141_2001",
236
+ "C141_2201",
237
+ "C141_3001",
238
+ "C141_0202",
239
+ "C145_3004",
240
+ "C145_0701",
241
+ "C145_1201",
242
+ "C145_1401",
243
+ "C145_1901",
244
+ "C145_2301",
245
+ "C146_3004",
246
+ "C146_1201",
247
+ "C146_1401",
248
+ "C146_1601",
249
+ "C146_1901",
250
+ "C146_2301",
251
+ "C146_2701",
252
+ "C146_2502",
253
+ "C152_0405",
254
+ "C161_1105",
255
+ "C164_1105",
256
+ "C167_1205",
257
+ "C168_1205",
258
+ "C168_2001",
259
+ "C168_2301",
260
+ "C168_3001",
261
+ "C168_0202",
262
+ "C168_0403",
263
+ "C169_1205",
264
+ "C169_2001",
265
+ "C169_2301",
266
+ "C169_0202",
267
+ "C169_0403",
268
+ "C170_1205",
269
+ "C170_2301",
270
+ "C170_0202",
271
+ "C170_0402",
272
+ "C170_0403",
273
+ "C187_2605",
274
+ "C187_1201",
275
+ "C187_1401",
276
+ "C187_1901",
277
+ "C187_2301",
278
+ "C187_2502",
279
+ "C190_2605",
280
+ "C190_2901",
281
+ "C190_0102",
282
+ "C190_0302",
283
+ "C190_0502",
284
+ "C190_0802",
285
+ "C190_1002",
286
+ "C190_0903",
287
+ "C193_2605",
288
+ "C195_2605",
289
+ "C197_2605",
290
+ "C212_2705",
291
+ "C214_2705",
292
+ "C237_2805",
293
+ "C238_2905",
294
+ "C240_2905",
295
+ "C246_2905",
296
+ "C251_2905"
297
+ ],
298
+ "inner_val_rows": [
299
+ "C63_1704",
300
+ "C63_2404",
301
+ "C63_2904",
302
+ "C128_2504",
303
+ "C128_0405",
304
+ "C128_1105",
305
+ "C142_3004",
306
+ "C147_3004",
307
+ "C147_1401",
308
+ "C147_1901",
309
+ "C234_2805"
310
+ ]
311
+ },
312
+ {
313
+ "cohort": "giroux2022",
314
+ "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",
315
+ "condition_npz_sha256": "b584655fcfa5fa4c0887d5dd83fcb1c5a0206da4afe5fefab96cba2272b1371c",
316
+ "train_rows": [
317
+ "450905_d0",
318
+ "450905_d14",
319
+ "0B943B_d0",
320
+ "0B943B_d14",
321
+ "7085CA_d0",
322
+ "7085CA_d7",
323
+ "7085CA_d14",
324
+ "2AD75E_d0",
325
+ "2AD75E_d7",
326
+ "2AD75E_d14",
327
+ "B85D75_d0",
328
+ "B85D75_d7",
329
+ "180E1A_d0",
330
+ "180E1A_d7",
331
+ "180E1A_d14",
332
+ "82CCF5_d0",
333
+ "82CCF5_d7",
334
+ "82CCF5_d14",
335
+ "BAAF62_d0",
336
+ "BAAF62_d7",
337
+ "BAAF62_d14",
338
+ "0BF51C_d0",
339
+ "0BF51C_d7",
340
+ "0BF51C_d14"
341
+ ],
342
+ "inner_val_rows": []
343
+ },
344
+ {
345
+ "cohort": "liu2021",
346
+ "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",
347
+ "condition_npz_sha256": "ad3131eabcfb11b166ca37c7585c50abb11efb60e2a736fe62de604794535e0c",
348
+ "train_rows": [
349
+ "AA220014_HC",
350
+ "AA220534_HC",
351
+ "AA220907_HC",
352
+ "HGR0000071_T0",
353
+ "HGR0000074_T0",
354
+ "HGR0000083_T1",
355
+ "HGR0000089_T1",
356
+ "HGR0000091_T0",
357
+ "HGR0000093_T0",
358
+ "HGR0000136_T0",
359
+ "HGR0000137_T0",
360
+ "HGR0000141_T0",
361
+ "IM000040_HC",
362
+ "HGR0000042_T1",
363
+ "HGR0000069_T0",
364
+ "HGR0000069_T1",
365
+ "HGR0000092_T0",
366
+ "HGR0000094_T0",
367
+ "HGR0000124_T0",
368
+ "HGR0000144_T2",
369
+ "HGR0000391_T0",
370
+ "HGR0000395_T0",
371
+ "HGR0000418_T1",
372
+ "SHD4_HC",
373
+ "SHD7_HC",
374
+ "HGR0000051_T0",
375
+ "HGR0000051_T1",
376
+ "HGR0000051_T3",
377
+ "HGR0000101_T1",
378
+ "HGR0000101_T3",
379
+ "HGR0000102_T0",
380
+ "HGR0000102_T1",
381
+ "HGR0000134_T0",
382
+ "HGR0000135_T0",
383
+ "HGR0000135_T1",
384
+ "HGR0000392_T0",
385
+ "HGR0000429_T1",
386
+ "HGR0000429_T2",
387
+ "SHD3_HC",
388
+ "SHD5_HC"
389
+ ],
390
+ "inner_val_rows": [
391
+ "HGR0000098_T0",
392
+ "HGR0000143_T0",
393
+ "HGR0000115_T1",
394
+ "SHD1_HC"
395
+ ]
396
+ }
397
+ ],
398
+ "held_out_test_rows_never_loaded": [
399
+ "Abd-PS04",
400
+ "Abd-PS05",
401
+ "Abd-PS07",
402
+ "Abd-PS19",
403
+ "Abd-PS20",
404
+ "Abd-PS23",
405
+ "Abd-PS27",
406
+ "Abd-S102",
407
+ "Abd-S103",
408
+ "Abd-S120",
409
+ "Abd-S123",
410
+ "Abd-S156",
411
+ "Abd-S39",
412
+ "Abd-S42",
413
+ "Abd-S47",
414
+ "Abd-S51",
415
+ "Abd-S54",
416
+ "Abd-S57",
417
+ "Abd-S66",
418
+ "Abd-S79",
419
+ "Abd-S83",
420
+ "Abd-S98",
421
+ "HC01",
422
+ "HC05",
423
+ "HC07",
424
+ "HC09",
425
+ "HC12",
426
+ "HC13",
427
+ "HC16",
428
+ "IC10",
429
+ "IC15",
430
+ "IC16",
431
+ "IC19",
432
+ "IC22",
433
+ "IC23",
434
+ "PHC01",
435
+ "PHC02",
436
+ "PHC03",
437
+ "PHC05",
438
+ "PIC01",
439
+ "PIC02",
440
+ "PIC04",
441
+ "PIC05",
442
+ "Res-PS02",
443
+ "Res-PS03",
444
+ "Res-PS12",
445
+ "Res-PS16",
446
+ "Res-PS21",
447
+ "Res-PS22",
448
+ "Res-PS25",
449
+ "Res-PS26",
450
+ "Res-S10",
451
+ "Res-S125",
452
+ "Res-S127",
453
+ "Res-S136",
454
+ "Res-S139",
455
+ "Res-S146",
456
+ "Res-S149",
457
+ "Res-S150",
458
+ "Res-S164",
459
+ "Res-S166",
460
+ "Res-S174",
461
+ "Res-S177",
462
+ "Res-S184",
463
+ "Res-S19",
464
+ "Res-S26",
465
+ "Res-S40",
466
+ "Res-S77",
467
+ "Res-S87",
468
+ "Skin-S140",
469
+ "Skin-S142",
470
+ "Skin-S152",
471
+ "Skin-S38",
472
+ "Skin-S43",
473
+ "Skin-S46",
474
+ "Skin-S52",
475
+ "Skin-S74",
476
+ "Uri-S134",
477
+ "Uri-S147",
478
+ "Uri-S157",
479
+ "Uri-S20",
480
+ "Uri-S93"
481
+ ],
482
+ "max_abs_weight_change": 1.4082739353179932,
483
+ "n_tensors_changed": 64,
484
+ "seed": 20261002,
485
+ "wall_seconds": 5600.4,
486
+ "smoke": false,
487
+ "block": 20000,
488
+ "saturation_frac": 0.01,
489
+ "stop_reason": "early_stop",
490
+ "milestones": [
491
+ {
492
+ "milestone_step": 20000,
493
+ "selected_step": 20000,
494
+ "selected_inner_val_loss": 0.09932147273135306,
495
+ "block_improvement": 0.008954538295816897,
496
+ "total_improvement": 0.008954538295816897,
497
+ "reason": "block_end",
498
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s20000.pt",
499
+ "checkpoint_sha256": "235563551146480bcf95b57d9456cd702fa24aaee2a58e987df763a37d9e09a0"
500
+ },
501
+ {
502
+ "milestone_step": 40000,
503
+ "selected_step": 40000,
504
+ "selected_inner_val_loss": 0.09898765926070587,
505
+ "block_improvement": 0.00033381347064719014,
506
+ "total_improvement": 0.009288351766464087,
507
+ "reason": "block_end",
508
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s40000.pt",
509
+ "checkpoint_sha256": "be3484436f57293db0bed630a4e032fc4a8a37f86e2f3cdbd92f0d207e8e2deb"
510
+ },
511
+ {
512
+ "milestone_step": 60000,
513
+ "selected_step": 59500,
514
+ "selected_inner_val_loss": 0.09878567852681935,
515
+ "block_improvement": 0.0002019807338865115,
516
+ "total_improvement": 0.009490332500350598,
517
+ "reason": "block_end",
518
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s60000.pt",
519
+ "checkpoint_sha256": "10def9a99a30b3a32b9027023d0bc8382fa779c3ae0e7df4d1d1dc96ed27e056"
520
+ },
521
+ {
522
+ "milestone_step": 80000,
523
+ "selected_step": 80000,
524
+ "selected_inner_val_loss": 0.0986245580092825,
525
+ "block_improvement": 0.00016112051753684742,
526
+ "total_improvement": 0.009651453017887446,
527
+ "reason": "block_end",
528
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s80000.pt",
529
+ "checkpoint_sha256": "9621e005c021486583ee7b97cc0a77bfadf57e5457f28b476d568777efb06b09"
530
+ },
531
+ {
532
+ "milestone_step": 100000,
533
+ "selected_step": 100000,
534
+ "selected_inner_val_loss": 0.09848370757706539,
535
+ "block_improvement": 0.00014085043221712112,
536
+ "total_improvement": 0.009792303450104567,
537
+ "reason": "block_end",
538
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s100000.pt",
539
+ "checkpoint_sha256": "3cde1860b41af88ef71cc29e7d7bb29371c30d6c7d12a5b7ba2765cbaa84e4e3"
540
+ },
541
+ {
542
+ "milestone_step": 108500,
543
+ "selected_step": 106500,
544
+ "selected_inner_val_loss": 0.09843902954963422,
545
+ "block_improvement": 4.4678027431160894e-05,
546
+ "total_improvement": 0.009836981477535728,
547
+ "reason": "early_stop",
548
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s108500.pt",
549
+ "checkpoint_sha256": "33441f5f75cb45b7c704f2e744ec842ad1371e46206c951443fbed98c7ea6095"
550
+ }
551
+ ],
552
+ "written_utc": "2026-10-03T09:10:38.958036+00:00",
553
+ "last_milestone_max_abs_diff": 0.0,
554
+ "checkpoint_sha256": "9240379953f47ef926556431f5e55c7cea43d0fa5b91eafef2d488bc25030a9a",
555
+ "reload_forward_max_abs_diff": 0.0
556
+ }
paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/history.json ADDED
The diff for this file is too large to render. See raw diff
 
paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/milestones.json ADDED
@@ -0,0 +1,62 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "milestone_step": 20000,
4
+ "selected_step": 20000,
5
+ "selected_inner_val_loss": 0.09932147273135306,
6
+ "block_improvement": 0.008954538295816897,
7
+ "total_improvement": 0.008954538295816897,
8
+ "reason": "block_end",
9
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s20000.pt",
10
+ "checkpoint_sha256": "235563551146480bcf95b57d9456cd702fa24aaee2a58e987df763a37d9e09a0"
11
+ },
12
+ {
13
+ "milestone_step": 40000,
14
+ "selected_step": 40000,
15
+ "selected_inner_val_loss": 0.09898765926070587,
16
+ "block_improvement": 0.00033381347064719014,
17
+ "total_improvement": 0.009288351766464087,
18
+ "reason": "block_end",
19
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s40000.pt",
20
+ "checkpoint_sha256": "be3484436f57293db0bed630a4e032fc4a8a37f86e2f3cdbd92f0d207e8e2deb"
21
+ },
22
+ {
23
+ "milestone_step": 60000,
24
+ "selected_step": 59500,
25
+ "selected_inner_val_loss": 0.09878567852681935,
26
+ "block_improvement": 0.0002019807338865115,
27
+ "total_improvement": 0.009490332500350598,
28
+ "reason": "block_end",
29
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s60000.pt",
30
+ "checkpoint_sha256": "10def9a99a30b3a32b9027023d0bc8382fa779c3ae0e7df4d1d1dc96ed27e056"
31
+ },
32
+ {
33
+ "milestone_step": 80000,
34
+ "selected_step": 80000,
35
+ "selected_inner_val_loss": 0.0986245580092825,
36
+ "block_improvement": 0.00016112051753684742,
37
+ "total_improvement": 0.009651453017887446,
38
+ "reason": "block_end",
39
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s80000.pt",
40
+ "checkpoint_sha256": "9621e005c021486583ee7b97cc0a77bfadf57e5457f28b476d568777efb06b09"
41
+ },
42
+ {
43
+ "milestone_step": 100000,
44
+ "selected_step": 100000,
45
+ "selected_inner_val_loss": 0.09848370757706539,
46
+ "block_improvement": 0.00014085043221712112,
47
+ "total_improvement": 0.009792303450104567,
48
+ "reason": "block_end",
49
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s100000.pt",
50
+ "checkpoint_sha256": "3cde1860b41af88ef71cc29e7d7bb29371c30d6c7d12a5b7ba2765cbaa84e4e3"
51
+ },
52
+ {
53
+ "milestone_step": 108500,
54
+ "selected_step": 106500,
55
+ "selected_inner_val_loss": 0.09843902954963422,
56
+ "block_improvement": 4.4678027431160894e-05,
57
+ "total_improvement": 0.009836981477535728,
58
+ "reason": "early_stop",
59
+ "checkpoint": "/home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft_s108500.pt",
60
+ "checkpoint_sha256": "33441f5f75cb45b7c704f2e744ec842ad1371e46206c951443fbed98c7ea6095"
61
+ }
62
+ ]
paper-releases/pbisc-diffusion/updates/20261003T103040Z-crosscohort-k1000-saturation/diffusion/dataset/server_1/model8/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/train.log ADDED
@@ -0,0 +1,443 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ 2026-10-03T16:37:15 split=hoL_ye2026_sepsis K=1000 mode=full trainable=54,185,184 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
2
+ 2026-10-03T16:37:18 step 0 inner-val loss 0.10828 (pretrained) {'immicro': 0.10636159498244524, 'imperial_eskd': 0.1086193590678952, 'liu2021': 0.10984707903116941}
3
+ 2026-10-03T16:37:31 step 250 train 0.10386 inner-val 0.10650 best 0.10650@250 (11s) immicro=0.1047 imperial_eskd=0.1066 liu2021=0.1081
4
+ 2026-10-03T16:37:44 step 500 train 0.10127 inner-val 0.10434 best 0.10434@500 (25s) immicro=0.1029 imperial_eskd=0.1041 liu2021=0.1060
5
+ 2026-10-03T16:37:58 step 750 train 0.10090 inner-val 0.10292 best 0.10292@750 (38s) immicro=0.1018 imperial_eskd=0.1024 liu2021=0.1046
6
+ 2026-10-03T16:38:13 step 1000 train 0.10053 inner-val 0.10204 best 0.10204@1000 (53s) immicro=0.1010 imperial_eskd=0.1013 liu2021=0.1038
7
+ 2026-10-03T16:38:27 step 1250 train 0.10057 inner-val 0.10147 best 0.10147@1250 (68s) immicro=0.1005 imperial_eskd=0.1006 liu2021=0.1033
8
+ 2026-10-03T16:38:40 step 1500 train 0.10029 inner-val 0.10110 best 0.10110@1500 (81s) immicro=0.1001 imperial_eskd=0.1002 liu2021=0.1030
9
+ 2026-10-03T16:38:54 step 1750 train 0.10038 inner-val 0.10084 best 0.10084@1750 (95s) immicro=0.0998 imperial_eskd=0.1000 liu2021=0.1027
10
+ 2026-10-03T16:39:08 step 2000 train 0.10010 inner-val 0.10066 best 0.10066@2000 (109s) immicro=0.0996 imperial_eskd=0.0998 liu2021=0.1026
11
+ 2026-10-03T16:39:20 step 2250 train 0.10021 inner-val 0.10052 best 0.10052@2250 (121s) immicro=0.0994 imperial_eskd=0.0996 liu2021=0.1025
12
+ 2026-10-03T16:39:33 step 2500 train 0.10044 inner-val 0.10041 best 0.10041@2500 (133s) immicro=0.0993 imperial_eskd=0.0995 liu2021=0.1024
13
+ 2026-10-03T16:39:46 step 2750 train 0.10003 inner-val 0.10033 best 0.10033@2750 (146s) immicro=0.0992 imperial_eskd=0.0994 liu2021=0.1024
14
+ 2026-10-03T16:39:59 step 3000 train 0.09998 inner-val 0.10026 best 0.10026@3000 (159s) immicro=0.0991 imperial_eskd=0.0993 liu2021=0.1023
15
+ 2026-10-03T16:40:12 step 3250 train 0.10017 inner-val 0.10020 best 0.10020@3250 (172s) immicro=0.0990 imperial_eskd=0.0993 liu2021=0.1023
16
+ 2026-10-03T16:40:24 step 3500 train 0.10012 inner-val 0.10015 best 0.10015@3500 (185s) immicro=0.0990 imperial_eskd=0.0992 liu2021=0.1022
17
+ 2026-10-03T16:40:37 step 3750 train 0.09997 inner-val 0.10011 best 0.10011@3750 (197s) immicro=0.0989 imperial_eskd=0.0992 liu2021=0.1022
18
+ 2026-10-03T16:40:50 step 4000 train 0.09991 inner-val 0.10007 best 0.10007@4000 (210s) immicro=0.0989 imperial_eskd=0.0991 liu2021=0.1022
19
+ 2026-10-03T16:41:02 step 4250 train 0.09992 inner-val 0.10003 best 0.10003@4250 (223s) immicro=0.0988 imperial_eskd=0.0991 liu2021=0.1022
20
+ 2026-10-03T16:41:15 step 4500 train 0.09985 inner-val 0.10000 best 0.10000@4500 (235s) immicro=0.0988 imperial_eskd=0.0991 liu2021=0.1022
21
+ 2026-10-03T16:41:27 step 4750 train 0.09973 inner-val 0.09997 best 0.09997@4750 (248s) immicro=0.0987 imperial_eskd=0.0990 liu2021=0.1021
22
+ 2026-10-03T16:41:40 step 5000 train 0.09989 inner-val 0.09995 best 0.09995@5000 (260s) immicro=0.0987 imperial_eskd=0.0990 liu2021=0.1021
23
+ 2026-10-03T16:41:52 step 5250 train 0.09984 inner-val 0.09993 best 0.09993@5250 (273s) immicro=0.0987 imperial_eskd=0.0990 liu2021=0.1021
24
+ 2026-10-03T16:42:05 step 5500 train 0.09986 inner-val 0.09990 best 0.09990@5500 (286s) immicro=0.0986 imperial_eskd=0.0990 liu2021=0.1021
25
+ 2026-10-03T16:42:18 step 5750 train 0.09942 inner-val 0.09988 best 0.09988@5750 (298s) immicro=0.0986 imperial_eskd=0.0990 liu2021=0.1021
26
+ 2026-10-03T16:42:31 step 6000 train 0.09970 inner-val 0.09986 best 0.09986@6000 (311s) immicro=0.0986 imperial_eskd=0.0989 liu2021=0.1021
27
+ 2026-10-03T16:42:44 step 6250 train 0.09974 inner-val 0.09984 best 0.09984@6250 (325s) immicro=0.0986 imperial_eskd=0.0989 liu2021=0.1021
28
+ 2026-10-03T16:42:57 step 6500 train 0.09981 inner-val 0.09982 best 0.09982@6500 (338s) immicro=0.0985 imperial_eskd=0.0989 liu2021=0.1020
29
+ 2026-10-03T16:43:10 step 6750 train 0.09966 inner-val 0.09980 best 0.09980@6750 (350s) immicro=0.0985 imperial_eskd=0.0989 liu2021=0.1020
30
+ 2026-10-03T16:43:22 step 7000 train 0.09958 inner-val 0.09978 best 0.09978@7000 (363s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1020
31
+ 2026-10-03T16:43:35 step 7250 train 0.09981 inner-val 0.09977 best 0.09977@7250 (376s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1020
32
+ 2026-10-03T16:43:48 step 7500 train 0.09960 inner-val 0.09975 best 0.09975@7500 (388s) immicro=0.0985 imperial_eskd=0.0988 liu2021=0.1020
33
+ 2026-10-03T16:44:01 step 7750 train 0.09957 inner-val 0.09974 best 0.09974@7750 (401s) immicro=0.0984 imperial_eskd=0.0988 liu2021=0.1020
34
+ 2026-10-03T16:44:13 step 8000 train 0.09927 inner-val 0.09972 best 0.09972@8000 (414s) immicro=0.0984 imperial_eskd=0.0988 liu2021=0.1020
35
+ 2026-10-03T16:44:26 step 8250 train 0.09972 inner-val 0.09971 best 0.09971@8250 (427s) immicro=0.0984 imperial_eskd=0.0988 liu2021=0.1020
36
+ 2026-10-03T16:44:39 step 8500 train 0.09930 inner-val 0.09970 best 0.09970@8500 (439s) immicro=0.0984 imperial_eskd=0.0988 liu2021=0.1019
37
+ 2026-10-03T16:44:51 step 8750 train 0.09975 inner-val 0.09968 best 0.09968@8750 (452s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1019
38
+ 2026-10-03T16:45:04 step 9000 train 0.09955 inner-val 0.09967 best 0.09967@9000 (464s) immicro=0.0984 imperial_eskd=0.0987 liu2021=0.1019
39
+ 2026-10-03T16:45:17 step 9250 train 0.09957 inner-val 0.09966 best 0.09966@9250 (477s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
40
+ 2026-10-03T16:45:30 step 9500 train 0.09946 inner-val 0.09965 best 0.09965@9500 (490s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
41
+ 2026-10-03T16:45:42 step 9750 train 0.09952 inner-val 0.09964 best 0.09964@9750 (503s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
42
+ 2026-10-03T16:45:55 step 10000 train 0.09944 inner-val 0.09963 best 0.09963@10000 (515s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
43
+ 2026-10-03T16:46:07 step 10250 train 0.09958 inner-val 0.09962 best 0.09962@10250 (528s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
44
+ 2026-10-03T16:46:20 step 10500 train 0.09963 inner-val 0.09961 best 0.09961@10500 (540s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
45
+ 2026-10-03T16:46:32 step 10750 train 0.09943 inner-val 0.09960 best 0.09960@10750 (553s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
46
+ 2026-10-03T16:46:45 step 11000 train 0.09944 inner-val 0.09959 best 0.09959@11000 (565s) immicro=0.0983 imperial_eskd=0.0987 liu2021=0.1019
47
+ 2026-10-03T16:46:58 step 11250 train 0.09934 inner-val 0.09958 best 0.09958@11250 (578s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1019
48
+ 2026-10-03T16:47:10 step 11500 train 0.09941 inner-val 0.09957 best 0.09957@11500 (591s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
49
+ 2026-10-03T16:47:23 step 11750 train 0.09929 inner-val 0.09956 best 0.09956@11750 (603s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
50
+ 2026-10-03T16:47:35 step 12000 train 0.09909 inner-val 0.09955 best 0.09955@12000 (616s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
51
+ 2026-10-03T16:47:48 step 12250 train 0.09928 inner-val 0.09954 best 0.09954@12250 (629s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
52
+ 2026-10-03T16:48:01 step 12500 train 0.09944 inner-val 0.09953 best 0.09953@12500 (641s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
53
+ 2026-10-03T16:48:13 step 12750 train 0.09938 inner-val 0.09952 best 0.09952@12750 (654s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
54
+ 2026-10-03T16:48:26 step 13000 train 0.09946 inner-val 0.09951 best 0.09951@13000 (667s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
55
+ 2026-10-03T16:48:39 step 13250 train 0.09942 inner-val 0.09950 best 0.09950@13250 (680s) immicro=0.0982 imperial_eskd=0.0986 liu2021=0.1018
56
+ 2026-10-03T16:48:52 step 13500 train 0.09928 inner-val 0.09950 best 0.09950@13500 (692s) immicro=0.0981 imperial_eskd=0.0986 liu2021=0.1018
57
+ 2026-10-03T16:49:05 step 13750 train 0.09937 inner-val 0.09949 best 0.09949@13750 (705s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1018
58
+ 2026-10-03T16:49:18 step 14000 train 0.09921 inner-val 0.09948 best 0.09948@14000 (718s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1018
59
+ 2026-10-03T16:49:31 step 14250 train 0.09910 inner-val 0.09947 best 0.09947@14250 (731s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1018
60
+ 2026-10-03T16:49:44 step 14500 train 0.09944 inner-val 0.09946 best 0.09946@14500 (744s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1018
61
+ 2026-10-03T16:49:57 step 14750 train 0.09931 inner-val 0.09945 best 0.09945@14750 (757s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1018
62
+ 2026-10-03T16:50:09 step 15000 train 0.09940 inner-val 0.09945 best 0.09945@15000 (770s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1018
63
+ 2026-10-03T16:50:22 step 15250 train 0.09915 inner-val 0.09944 best 0.09944@15250 (783s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1017
64
+ 2026-10-03T16:50:35 step 15500 train 0.09935 inner-val 0.09943 best 0.09943@15500 (796s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1017
65
+ 2026-10-03T16:50:48 step 15750 train 0.09908 inner-val 0.09943 best 0.09943@15750 (808s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1017
66
+ 2026-10-03T16:51:01 step 16000 train 0.09914 inner-val 0.09942 best 0.09942@16000 (821s) immicro=0.0981 imperial_eskd=0.0985 liu2021=0.1017
67
+ 2026-10-03T16:51:14 step 16250 train 0.09923 inner-val 0.09941 best 0.09941@16250 (834s) immicro=0.0980 imperial_eskd=0.0985 liu2021=0.1017
68
+ 2026-10-03T16:51:26 step 16500 train 0.09933 inner-val 0.09941 best 0.09941@16500 (847s) immicro=0.0980 imperial_eskd=0.0985 liu2021=0.1017
69
+ 2026-10-03T16:51:39 step 16750 train 0.09904 inner-val 0.09940 best 0.09940@16750 (860s) immicro=0.0980 imperial_eskd=0.0985 liu2021=0.1017
70
+ 2026-10-03T16:51:52 step 17000 train 0.09935 inner-val 0.09939 best 0.09939@17000 (872s) immicro=0.0980 imperial_eskd=0.0985 liu2021=0.1017
71
+ 2026-10-03T16:52:04 step 17250 train 0.09943 inner-val 0.09939 best 0.09939@17250 (885s) immicro=0.0980 imperial_eskd=0.0985 liu2021=0.1017
72
+ 2026-10-03T16:52:17 step 17500 train 0.09925 inner-val 0.09938 best 0.09938@17500 (898s) immicro=0.0980 imperial_eskd=0.0985 liu2021=0.1017
73
+ 2026-10-03T16:52:30 step 17750 train 0.09897 inner-val 0.09937 best 0.09937@17750 (911s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1017
74
+ 2026-10-03T16:52:43 step 18000 train 0.09953 inner-val 0.09937 best 0.09937@18000 (923s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1017
75
+ 2026-10-03T16:52:55 step 18250 train 0.09895 inner-val 0.09936 best 0.09936@18250 (936s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1017
76
+ 2026-10-03T16:53:08 step 18500 train 0.09925 inner-val 0.09936 best 0.09936@18500 (949s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1017
77
+ 2026-10-03T16:53:21 step 18750 train 0.09902 inner-val 0.09935 best 0.09935@18750 (961s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1016
78
+ 2026-10-03T16:53:33 step 19000 train 0.09913 inner-val 0.09934 best 0.09934@19000 (974s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1016
79
+ 2026-10-03T16:53:46 step 19250 train 0.09903 inner-val 0.09934 best 0.09934@19250 (987s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1016
80
+ 2026-10-03T16:53:59 step 19500 train 0.09901 inner-val 0.09933 best 0.09933@19500 (999s) immicro=0.0980 imperial_eskd=0.0984 liu2021=0.1016
81
+ 2026-10-03T16:54:11 step 19750 train 0.09895 inner-val 0.09933 best 0.09933@19750 (1012s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
82
+ 2026-10-03T16:54:24 step 20000 train 0.09916 inner-val 0.09932 best 0.09932@20000 (1024s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
83
+ 2026-10-03T16:54:25 block end 20000: best 0.09932@20000; block improvement 0.008955 (100.00 % of total 0.008955) -> continue
84
+ 2026-10-03T16:54:38 step 20250 train 0.09909 inner-val 0.09932 best 0.09932@20250 (1038s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
85
+ 2026-10-03T16:54:51 step 20500 train 0.09878 inner-val 0.09931 best 0.09931@20500 (1051s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
86
+ 2026-10-03T16:55:03 step 20750 train 0.09902 inner-val 0.09930 best 0.09930@20750 (1064s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
87
+ 2026-10-03T16:55:17 step 21000 train 0.09918 inner-val 0.09930 best 0.09930@21000 (1078s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
88
+ 2026-10-03T16:55:31 step 21250 train 0.09860 inner-val 0.09929 best 0.09929@21250 (1091s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
89
+ 2026-10-03T16:55:45 step 21500 train 0.09882 inner-val 0.09929 best 0.09929@21500 (1105s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
90
+ 2026-10-03T16:55:59 step 21750 train 0.09916 inner-val 0.09928 best 0.09928@21750 (1119s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
91
+ 2026-10-03T16:56:12 step 22000 train 0.09904 inner-val 0.09928 best 0.09928@22000 (1133s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
92
+ 2026-10-03T16:56:26 step 22250 train 0.09899 inner-val 0.09927 best 0.09927@22250 (1147s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
93
+ 2026-10-03T16:56:40 step 22500 train 0.09895 inner-val 0.09927 best 0.09927@22500 (1160s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
94
+ 2026-10-03T16:56:53 step 22750 train 0.09904 inner-val 0.09926 best 0.09926@22750 (1174s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
95
+ 2026-10-03T16:57:07 step 23000 train 0.09932 inner-val 0.09926 best 0.09926@23000 (1188s) immicro=0.0979 imperial_eskd=0.0984 liu2021=0.1016
96
+ 2026-10-03T16:57:21 step 23250 train 0.09913 inner-val 0.09925 best 0.09925@23250 (1202s) immicro=0.0979 imperial_eskd=0.0983 liu2021=0.1016
97
+ 2026-10-03T16:57:35 step 23500 train 0.09917 inner-val 0.09925 best 0.09925@23500 (1216s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1016
98
+ 2026-10-03T16:57:49 step 23750 train 0.09883 inner-val 0.09924 best 0.09924@23750 (1230s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
99
+ 2026-10-03T16:58:04 step 24000 train 0.09923 inner-val 0.09923 best 0.09923@24000 (1244s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
100
+ 2026-10-03T16:58:17 step 24250 train 0.09874 inner-val 0.09923 best 0.09923@24250 (1258s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
101
+ 2026-10-03T16:58:32 step 24500 train 0.09866 inner-val 0.09922 best 0.09922@24500 (1272s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
102
+ 2026-10-03T16:58:46 step 24750 train 0.09902 inner-val 0.09921 best 0.09921@24750 (1286s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
103
+ 2026-10-03T16:58:59 step 25000 train 0.09883 inner-val 0.09921 best 0.09921@25000 (1300s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
104
+ 2026-10-03T16:59:13 step 25250 train 0.09897 inner-val 0.09921 best 0.09921@25250 (1313s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
105
+ 2026-10-03T16:59:27 step 25500 train 0.09876 inner-val 0.09920 best 0.09920@25500 (1327s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
106
+ 2026-10-03T16:59:41 step 25750 train 0.09901 inner-val 0.09920 best 0.09920@25750 (1341s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
107
+ 2026-10-03T16:59:55 step 26000 train 0.09898 inner-val 0.09920 best 0.09920@26000 (1355s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
108
+ 2026-10-03T17:00:08 step 26250 train 0.09884 inner-val 0.09919 best 0.09919@26250 (1369s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
109
+ 2026-10-03T17:00:21 step 26500 train 0.09904 inner-val 0.09919 best 0.09919@26500 (1382s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
110
+ 2026-10-03T17:00:34 step 26750 train 0.09858 inner-val 0.09919 best 0.09919@26750 (1394s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
111
+ 2026-10-03T17:00:46 step 27000 train 0.09897 inner-val 0.09918 best 0.09918@27000 (1407s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
112
+ 2026-10-03T17:00:59 step 27250 train 0.09906 inner-val 0.09918 best 0.09918@27250 (1419s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
113
+ 2026-10-03T17:01:12 step 27500 train 0.09890 inner-val 0.09917 best 0.09917@27500 (1432s) immicro=0.0978 imperial_eskd=0.0983 liu2021=0.1015
114
+ 2026-10-03T17:01:24 step 27750 train 0.09903 inner-val 0.09917 best 0.09917@27750 (1445s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
115
+ 2026-10-03T17:01:37 step 28000 train 0.09876 inner-val 0.09916 best 0.09916@28000 (1458s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
116
+ 2026-10-03T17:01:50 step 28250 train 0.09886 inner-val 0.09916 best 0.09916@28250 (1471s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
117
+ 2026-10-03T17:02:04 step 28500 train 0.09847 inner-val 0.09916 best 0.09916@28500 (1484s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
118
+ 2026-10-03T17:02:17 step 28750 train 0.09874 inner-val 0.09915 best 0.09915@28750 (1498s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
119
+ 2026-10-03T17:02:31 step 29000 train 0.09871 inner-val 0.09914 best 0.09914@29000 (1512s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1014
120
+ 2026-10-03T17:02:45 step 29250 train 0.09917 inner-val 0.09914 best 0.09914@29250 (1526s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
121
+ 2026-10-03T17:03:00 step 29500 train 0.09846 inner-val 0.09914 best 0.09914@29500 (1540s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
122
+ 2026-10-03T17:03:14 step 29750 train 0.09871 inner-val 0.09914 best 0.09914@29750 (1554s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1015
123
+ 2026-10-03T17:03:28 step 30000 train 0.09841 inner-val 0.09914 best 0.09914@30000 (1568s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1014
124
+ 2026-10-03T17:03:42 step 30250 train 0.09868 inner-val 0.09913 best 0.09913@30250 (1582s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1014
125
+ 2026-10-03T17:03:56 step 30500 train 0.09871 inner-val 0.09913 best 0.09913@30500 (1596s) immicro=0.0977 imperial_eskd=0.0983 liu2021=0.1014
126
+ 2026-10-03T17:04:10 step 30750 train 0.09872 inner-val 0.09913 best 0.09913@30750 (1610s) immicro=0.0977 imperial_eskd=0.0982 liu2021=0.1014
127
+ 2026-10-03T17:04:24 step 31000 train 0.09834 inner-val 0.09912 best 0.09912@31000 (1624s) immicro=0.0977 imperial_eskd=0.0982 liu2021=0.1014
128
+ 2026-10-03T17:04:38 step 31250 train 0.09862 inner-val 0.09912 best 0.09912@31250 (1638s) immicro=0.0977 imperial_eskd=0.0982 liu2021=0.1014
129
+ 2026-10-03T17:04:52 step 31500 train 0.09843 inner-val 0.09911 best 0.09911@31500 (1652s) immicro=0.0977 imperial_eskd=0.0982 liu2021=0.1014
130
+ 2026-10-03T17:05:06 step 31750 train 0.09875 inner-val 0.09911 best 0.09911@31750 (1666s) immicro=0.0977 imperial_eskd=0.0982 liu2021=0.1014
131
+ 2026-10-03T17:05:20 step 32000 train 0.09873 inner-val 0.09910 best 0.09910@32000 (1681s) immicro=0.0977 imperial_eskd=0.0982 liu2021=0.1014
132
+ 2026-10-03T17:05:33 step 32250 train 0.09859 inner-val 0.09910 best 0.09910@32250 (1694s) immicro=0.0977 imperial_eskd=0.0982 liu2021=0.1014
133
+ 2026-10-03T17:05:46 step 32500 train 0.09852 inner-val 0.09909 best 0.09909@32500 (1706s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
134
+ 2026-10-03T17:05:58 step 32750 train 0.09876 inner-val 0.09909 best 0.09909@32750 (1719s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
135
+ 2026-10-03T17:06:11 step 33000 train 0.09851 inner-val 0.09909 best 0.09909@33000 (1731s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
136
+ 2026-10-03T17:06:24 step 33250 train 0.09864 inner-val 0.09908 best 0.09908@33250 (1744s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
137
+ 2026-10-03T17:06:36 step 33500 train 0.09820 inner-val 0.09908 best 0.09908@33500 (1757s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
138
+ 2026-10-03T17:06:47 step 33750 train 0.09858 inner-val 0.09908 best 0.09908@33500 (1769s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
139
+ 2026-10-03T17:07:01 step 34000 train 0.09876 inner-val 0.09907 best 0.09907@34000 (1782s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
140
+ 2026-10-03T17:07:15 step 34250 train 0.09877 inner-val 0.09907 best 0.09907@34250 (1796s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
141
+ 2026-10-03T17:07:29 step 34500 train 0.09876 inner-val 0.09907 best 0.09907@34500 (1809s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
142
+ 2026-10-03T17:07:42 step 34750 train 0.09877 inner-val 0.09906 best 0.09906@34750 (1823s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
143
+ 2026-10-03T17:07:56 step 35000 train 0.09868 inner-val 0.09906 best 0.09906@35000 (1837s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
144
+ 2026-10-03T17:08:10 step 35250 train 0.09869 inner-val 0.09906 best 0.09906@35250 (1850s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
145
+ 2026-10-03T17:08:24 step 35500 train 0.09885 inner-val 0.09905 best 0.09905@35500 (1864s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
146
+ 2026-10-03T17:08:36 step 35750 train 0.09855 inner-val 0.09905 best 0.09905@35500 (1878s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
147
+ 2026-10-03T17:08:50 step 36000 train 0.09867 inner-val 0.09904 best 0.09904@36000 (1890s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1014
148
+ 2026-10-03T17:09:04 step 36250 train 0.09865 inner-val 0.09904 best 0.09904@36250 (1904s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
149
+ 2026-10-03T17:09:18 step 36500 train 0.09829 inner-val 0.09904 best 0.09904@36500 (1918s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
150
+ 2026-10-03T17:09:32 step 36750 train 0.09875 inner-val 0.09903 best 0.09903@36750 (1932s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
151
+ 2026-10-03T17:09:45 step 37000 train 0.09828 inner-val 0.09903 best 0.09903@37000 (1946s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
152
+ 2026-10-03T17:09:58 step 37250 train 0.09852 inner-val 0.09903 best 0.09903@37000 (1960s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
153
+ 2026-10-03T17:10:11 step 37500 train 0.09867 inner-val 0.09903 best 0.09903@37500 (1972s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
154
+ 2026-10-03T17:10:25 step 37750 train 0.09860 inner-val 0.09902 best 0.09902@37750 (1986s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
155
+ 2026-10-03T17:10:39 step 38000 train 0.09870 inner-val 0.09902 best 0.09902@38000 (1999s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
156
+ 2026-10-03T17:10:53 step 38250 train 0.09873 inner-val 0.09901 best 0.09901@38250 (2013s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
157
+ 2026-10-03T17:11:07 step 38500 train 0.09865 inner-val 0.09901 best 0.09901@38500 (2028s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
158
+ 2026-10-03T17:11:22 step 38750 train 0.09859 inner-val 0.09901 best 0.09901@38750 (2042s) immicro=0.0976 imperial_eskd=0.0982 liu2021=0.1013
159
+ 2026-10-03T17:11:36 step 39000 train 0.09844 inner-val 0.09900 best 0.09900@39000 (2056s) immicro=0.0975 imperial_eskd=0.0982 liu2021=0.1013
160
+ 2026-10-03T17:11:51 step 39250 train 0.09845 inner-val 0.09900 best 0.09900@39250 (2071s) immicro=0.0975 imperial_eskd=0.0982 liu2021=0.1013
161
+ 2026-10-03T17:12:05 step 39500 train 0.09882 inner-val 0.09900 best 0.09900@39500 (2085s) immicro=0.0975 imperial_eskd=0.0982 liu2021=0.1013
162
+ 2026-10-03T17:12:18 step 39750 train 0.09848 inner-val 0.09899 best 0.09899@39750 (2099s) immicro=0.0975 imperial_eskd=0.0982 liu2021=0.1013
163
+ 2026-10-03T17:12:32 step 40000 train 0.09848 inner-val 0.09899 best 0.09899@40000 (2113s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
164
+ 2026-10-03T17:12:34 block end 40000: best 0.09899@40000; block improvement 0.000334 (3.59 % of total 0.009288) -> continue
165
+ 2026-10-03T17:12:48 step 40250 train 0.09863 inner-val 0.09898 best 0.09898@40250 (2128s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
166
+ 2026-10-03T17:13:02 step 40500 train 0.09877 inner-val 0.09898 best 0.09898@40500 (2142s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
167
+ 2026-10-03T17:13:16 step 40750 train 0.09833 inner-val 0.09898 best 0.09898@40750 (2156s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
168
+ 2026-10-03T17:13:29 step 41000 train 0.09855 inner-val 0.09897 best 0.09897@41000 (2170s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
169
+ 2026-10-03T17:13:43 step 41250 train 0.09842 inner-val 0.09897 best 0.09897@41250 (2184s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
170
+ 2026-10-03T17:13:57 step 41500 train 0.09845 inner-val 0.09897 best 0.09897@41500 (2198s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
171
+ 2026-10-03T17:14:11 step 41750 train 0.09839 inner-val 0.09897 best 0.09897@41750 (2212s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
172
+ 2026-10-03T17:14:25 step 42000 train 0.09836 inner-val 0.09896 best 0.09896@42000 (2225s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
173
+ 2026-10-03T17:14:39 step 42250 train 0.09825 inner-val 0.09896 best 0.09896@42250 (2239s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
174
+ 2026-10-03T17:14:51 step 42500 train 0.09874 inner-val 0.09896 best 0.09896@42250 (2253s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
175
+ 2026-10-03T17:15:05 step 42750 train 0.09826 inner-val 0.09896 best 0.09896@42750 (2265s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
176
+ 2026-10-03T17:15:17 step 43000 train 0.09821 inner-val 0.09896 best 0.09896@42750 (2279s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
177
+ 2026-10-03T17:15:31 step 43250 train 0.09866 inner-val 0.09896 best 0.09896@43250 (2292s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
178
+ 2026-10-03T17:15:45 step 43500 train 0.09859 inner-val 0.09895 best 0.09895@43500 (2305s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
179
+ 2026-10-03T17:15:59 step 43750 train 0.09892 inner-val 0.09895 best 0.09895@43750 (2319s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
180
+ 2026-10-03T17:16:12 step 44000 train 0.09817 inner-val 0.09895 best 0.09895@44000 (2333s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1013
181
+ 2026-10-03T17:16:26 step 44250 train 0.09842 inner-val 0.09894 best 0.09894@44250 (2347s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1012
182
+ 2026-10-03T17:16:40 step 44500 train 0.09857 inner-val 0.09894 best 0.09894@44500 (2361s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1012
183
+ 2026-10-03T17:16:54 step 44750 train 0.09846 inner-val 0.09894 best 0.09894@44750 (2374s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1012
184
+ 2026-10-03T17:17:08 step 45000 train 0.09836 inner-val 0.09893 best 0.09893@45000 (2388s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1012
185
+ 2026-10-03T17:17:20 step 45250 train 0.09856 inner-val 0.09893 best 0.09893@45000 (2402s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1012
186
+ 2026-10-03T17:17:34 step 45500 train 0.09881 inner-val 0.09893 best 0.09893@45500 (2414s) immicro=0.0975 imperial_eskd=0.0981 liu2021=0.1012
187
+ 2026-10-03T17:17:47 step 45750 train 0.09846 inner-val 0.09893 best 0.09893@45750 (2428s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
188
+ 2026-10-03T17:18:01 step 46000 train 0.09861 inner-val 0.09892 best 0.09892@46000 (2442s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
189
+ 2026-10-03T17:18:15 step 46250 train 0.09822 inner-val 0.09892 best 0.09892@46250 (2455s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
190
+ 2026-10-03T17:18:27 step 46500 train 0.09847 inner-val 0.09892 best 0.09892@46250 (2469s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
191
+ 2026-10-03T17:18:41 step 46750 train 0.09832 inner-val 0.09892 best 0.09892@46750 (2481s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
192
+ 2026-10-03T17:18:55 step 47000 train 0.09824 inner-val 0.09891 best 0.09891@47000 (2495s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
193
+ 2026-10-03T17:19:09 step 47250 train 0.09838 inner-val 0.09891 best 0.09891@47250 (2509s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
194
+ 2026-10-03T17:19:20 step 47500 train 0.09850 inner-val 0.09891 best 0.09891@47250 (2522s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
195
+ 2026-10-03T17:19:35 step 47750 train 0.09840 inner-val 0.09890 best 0.09890@47750 (2535s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
196
+ 2026-10-03T17:19:49 step 48000 train 0.09840 inner-val 0.09890 best 0.09890@48000 (2549s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
197
+ 2026-10-03T17:20:03 step 48250 train 0.09860 inner-val 0.09889 best 0.09889@48250 (2563s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
198
+ 2026-10-03T17:20:16 step 48500 train 0.09828 inner-val 0.09889 best 0.09889@48250 (2578s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
199
+ 2026-10-03T17:20:30 step 48750 train 0.09816 inner-val 0.09889 best 0.09889@48750 (2590s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
200
+ 2026-10-03T17:20:43 step 49000 train 0.09853 inner-val 0.09889 best 0.09889@48750 (2604s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
201
+ 2026-10-03T17:20:57 step 49250 train 0.09835 inner-val 0.09889 best 0.09889@49250 (2617s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
202
+ 2026-10-03T17:21:11 step 49500 train 0.09830 inner-val 0.09888 best 0.09888@49500 (2632s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
203
+ 2026-10-03T17:21:25 step 49750 train 0.09846 inner-val 0.09888 best 0.09888@49750 (2646s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
204
+ 2026-10-03T17:21:39 step 50000 train 0.09789 inner-val 0.09888 best 0.09888@50000 (2659s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
205
+ 2026-10-03T17:21:52 step 50250 train 0.09817 inner-val 0.09888 best 0.09888@50250 (2673s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
206
+ 2026-10-03T17:22:06 step 50500 train 0.09852 inner-val 0.09887 best 0.09887@50500 (2687s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
207
+ 2026-10-03T17:22:20 step 50750 train 0.09810 inner-val 0.09887 best 0.09887@50750 (2701s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
208
+ 2026-10-03T17:22:34 step 51000 train 0.09839 inner-val 0.09886 best 0.09886@51000 (2714s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
209
+ 2026-10-03T17:22:47 step 51250 train 0.09826 inner-val 0.09886 best 0.09886@51250 (2728s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
210
+ 2026-10-03T17:23:01 step 51500 train 0.09828 inner-val 0.09886 best 0.09886@51500 (2742s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1012
211
+ 2026-10-03T17:23:15 step 51750 train 0.09836 inner-val 0.09885 best 0.09885@51750 (2756s) immicro=0.0974 imperial_eskd=0.0981 liu2021=0.1011
212
+ 2026-10-03T17:23:29 step 52000 train 0.09851 inner-val 0.09885 best 0.09885@52000 (2769s) immicro=0.0974 imperial_eskd=0.0980 liu2021=0.1011
213
+ 2026-10-03T17:23:42 step 52250 train 0.09838 inner-val 0.09885 best 0.09885@52250 (2783s) immicro=0.0974 imperial_eskd=0.0980 liu2021=0.1011
214
+ 2026-10-03T17:23:56 step 52500 train 0.09871 inner-val 0.09885 best 0.09885@52500 (2796s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
215
+ 2026-10-03T17:24:08 step 52750 train 0.09847 inner-val 0.09885 best 0.09885@52500 (2810s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1012
216
+ 2026-10-03T17:24:22 step 53000 train 0.09809 inner-val 0.09884 best 0.09884@53000 (2822s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1012
217
+ 2026-10-03T17:24:34 step 53250 train 0.09845 inner-val 0.09884 best 0.09884@53000 (2836s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1012
218
+ 2026-10-03T17:24:47 step 53500 train 0.09817 inner-val 0.09885 best 0.09884@53000 (2849s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1012
219
+ 2026-10-03T17:25:00 step 53750 train 0.09835 inner-val 0.09884 best 0.09884@53750 (2861s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1012
220
+ 2026-10-03T17:25:14 step 54000 train 0.09812 inner-val 0.09884 best 0.09884@54000 (2875s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1012
221
+ 2026-10-03T17:25:28 step 54250 train 0.09847 inner-val 0.09883 best 0.09883@54250 (2889s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
222
+ 2026-10-03T17:25:42 step 54500 train 0.09837 inner-val 0.09883 best 0.09883@54500 (2902s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
223
+ 2026-10-03T17:25:56 step 54750 train 0.09811 inner-val 0.09883 best 0.09883@54750 (2916s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
224
+ 2026-10-03T17:26:09 step 55000 train 0.09832 inner-val 0.09883 best 0.09883@55000 (2929s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
225
+ 2026-10-03T17:26:22 step 55250 train 0.09813 inner-val 0.09882 best 0.09882@55250 (2942s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
226
+ 2026-10-03T17:26:35 step 55500 train 0.09829 inner-val 0.09882 best 0.09882@55500 (2955s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
227
+ 2026-10-03T17:26:49 step 55750 train 0.09824 inner-val 0.09882 best 0.09882@55750 (2969s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
228
+ 2026-10-03T17:27:03 step 56000 train 0.09803 inner-val 0.09881 best 0.09881@56000 (2984s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
229
+ 2026-10-03T17:27:17 step 56250 train 0.09831 inner-val 0.09881 best 0.09881@56250 (2998s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
230
+ 2026-10-03T17:27:31 step 56500 train 0.09793 inner-val 0.09880 best 0.09880@56500 (3012s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
231
+ 2026-10-03T17:27:45 step 56750 train 0.09785 inner-val 0.09880 best 0.09880@56750 (3026s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
232
+ 2026-10-03T17:27:59 step 57000 train 0.09801 inner-val 0.09880 best 0.09880@57000 (3039s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
233
+ 2026-10-03T17:28:11 step 57250 train 0.09834 inner-val 0.09880 best 0.09880@57000 (3053s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
234
+ 2026-10-03T17:28:25 step 57500 train 0.09816 inner-val 0.09879 best 0.09879@57500 (3065s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
235
+ 2026-10-03T17:28:37 step 57750 train 0.09789 inner-val 0.09879 best 0.09879@57500 (3079s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
236
+ 2026-10-03T17:28:50 step 58000 train 0.09817 inner-val 0.09880 best 0.09879@57500 (3092s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
237
+ 2026-10-03T17:29:02 step 58250 train 0.09832 inner-val 0.09880 best 0.09879@57500 (3104s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
238
+ 2026-10-03T17:29:14 step 58500 train 0.09840 inner-val 0.09880 best 0.09879@57500 (3116s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
239
+ 2026-10-03T17:29:27 step 58750 train 0.09860 inner-val 0.09880 best 0.09879@57500 (3129s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
240
+ 2026-10-03T17:29:39 step 59000 train 0.09833 inner-val 0.09879 best 0.09879@57500 (3141s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
241
+ 2026-10-03T17:29:53 step 59250 train 0.09784 inner-val 0.09879 best 0.09879@59250 (3153s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
242
+ 2026-10-03T17:30:07 step 59500 train 0.09813 inner-val 0.09879 best 0.09879@59500 (3167s) immicro=0.0973 imperial_eskd=0.0980 liu2021=0.1011
243
+ 2026-10-03T17:30:19 step 59750 train 0.09832 inner-val 0.09879 best 0.09879@59500 (3181s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
244
+ 2026-10-03T17:30:31 step 60000 train 0.09801 inner-val 0.09879 best 0.09879@59500 (3193s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
245
+ 2026-10-03T17:30:33 block end 60000: best 0.09879@59500; block improvement 0.000202 (2.13 % of total 0.009490) -> continue
246
+ 2026-10-03T17:30:45 step 60250 train 0.09817 inner-val 0.09879 best 0.09879@59500 (3207s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
247
+ 2026-10-03T17:30:59 step 60500 train 0.09820 inner-val 0.09878 best 0.09878@60500 (3220s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
248
+ 2026-10-03T17:31:13 step 60750 train 0.09796 inner-val 0.09878 best 0.09878@60750 (3233s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
249
+ 2026-10-03T17:31:25 step 61000 train 0.09833 inner-val 0.09878 best 0.09878@60750 (3247s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
250
+ 2026-10-03T17:31:38 step 61250 train 0.09850 inner-val 0.09878 best 0.09878@60750 (3260s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
251
+ 2026-10-03T17:31:52 step 61500 train 0.09794 inner-val 0.09877 best 0.09877@61500 (3272s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
252
+ 2026-10-03T17:32:06 step 61750 train 0.09821 inner-val 0.09877 best 0.09877@61750 (3286s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
253
+ 2026-10-03T17:32:20 step 62000 train 0.09810 inner-val 0.09877 best 0.09877@62000 (3301s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
254
+ 2026-10-03T17:32:35 step 62250 train 0.09808 inner-val 0.09876 best 0.09876@62250 (3315s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
255
+ 2026-10-03T17:32:49 step 62500 train 0.09826 inner-val 0.09876 best 0.09876@62500 (3329s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
256
+ 2026-10-03T17:33:03 step 62750 train 0.09801 inner-val 0.09875 best 0.09875@62750 (3343s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
257
+ 2026-10-03T17:33:16 step 63000 train 0.09826 inner-val 0.09875 best 0.09875@63000 (3357s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
258
+ 2026-10-03T17:33:30 step 63250 train 0.09842 inner-val 0.09875 best 0.09875@63250 (3371s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
259
+ 2026-10-03T17:33:43 step 63500 train 0.09794 inner-val 0.09875 best 0.09875@63250 (3384s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
260
+ 2026-10-03T17:33:56 step 63750 train 0.09811 inner-val 0.09875 best 0.09875@63750 (3397s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
261
+ 2026-10-03T17:34:10 step 64000 train 0.09819 inner-val 0.09874 best 0.09874@64000 (3411s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
262
+ 2026-10-03T17:34:22 step 64250 train 0.09813 inner-val 0.09874 best 0.09874@64000 (3424s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
263
+ 2026-10-03T17:34:36 step 64500 train 0.09801 inner-val 0.09874 best 0.09874@64500 (3437s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
264
+ 2026-10-03T17:34:49 step 64750 train 0.09814 inner-val 0.09874 best 0.09874@64500 (3451s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
265
+ 2026-10-03T17:35:02 step 65000 train 0.09812 inner-val 0.09874 best 0.09874@65000 (3463s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
266
+ 2026-10-03T17:35:16 step 65250 train 0.09820 inner-val 0.09873 best 0.09873@65250 (3477s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
267
+ 2026-10-03T17:35:30 step 65500 train 0.09788 inner-val 0.09873 best 0.09873@65500 (3490s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
268
+ 2026-10-03T17:35:42 step 65750 train 0.09834 inner-val 0.09873 best 0.09873@65500 (3504s) immicro=0.0972 imperial_eskd=0.0980 liu2021=0.1011
269
+ 2026-10-03T17:35:56 step 66000 train 0.09793 inner-val 0.09873 best 0.09873@66000 (3517s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
270
+ 2026-10-03T17:36:08 step 66250 train 0.09788 inner-val 0.09873 best 0.09873@66000 (3530s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
271
+ 2026-10-03T17:36:22 step 66500 train 0.09818 inner-val 0.09873 best 0.09873@66500 (3543s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
272
+ 2026-10-03T17:36:36 step 66750 train 0.09794 inner-val 0.09872 best 0.09872@66750 (3556s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
273
+ 2026-10-03T17:36:50 step 67000 train 0.09817 inner-val 0.09872 best 0.09872@67000 (3570s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
274
+ 2026-10-03T17:37:03 step 67250 train 0.09781 inner-val 0.09871 best 0.09871@67250 (3583s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1010
275
+ 2026-10-03T17:37:14 step 67500 train 0.09813 inner-val 0.09871 best 0.09871@67250 (3596s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
276
+ 2026-10-03T17:37:25 step 67750 train 0.09802 inner-val 0.09871 best 0.09871@67250 (3607s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1011
277
+ 2026-10-03T17:37:38 step 68000 train 0.09802 inner-val 0.09871 best 0.09871@68000 (3618s) immicro=0.0972 imperial_eskd=0.0979 liu2021=0.1010
278
+ 2026-10-03T17:37:50 step 68250 train 0.09799 inner-val 0.09871 best 0.09871@68250 (3631s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1011
279
+ 2026-10-03T17:38:02 step 68500 train 0.09807 inner-val 0.09871 best 0.09871@68250 (3644s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1011
280
+ 2026-10-03T17:38:14 step 68750 train 0.09827 inner-val 0.09870 best 0.09870@68750 (3655s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1011
281
+ 2026-10-03T17:38:26 step 69000 train 0.09793 inner-val 0.09871 best 0.09870@68750 (3668s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1011
282
+ 2026-10-03T17:38:38 step 69250 train 0.09806 inner-val 0.09870 best 0.09870@69250 (3679s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
283
+ 2026-10-03T17:38:51 step 69500 train 0.09758 inner-val 0.09870 best 0.09870@69500 (3692s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
284
+ 2026-10-03T17:39:04 step 69750 train 0.09813 inner-val 0.09870 best 0.09870@69750 (3704s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
285
+ 2026-10-03T17:39:17 step 70000 train 0.09793 inner-val 0.09870 best 0.09870@70000 (3717s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
286
+ 2026-10-03T17:39:29 step 70250 train 0.09799 inner-val 0.09869 best 0.09869@70250 (3730s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
287
+ 2026-10-03T17:39:42 step 70500 train 0.09802 inner-val 0.09869 best 0.09869@70500 (3743s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
288
+ 2026-10-03T17:39:55 step 70750 train 0.09783 inner-val 0.09868 best 0.09868@70750 (3755s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
289
+ 2026-10-03T17:40:06 step 71000 train 0.09786 inner-val 0.09869 best 0.09868@70750 (3768s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
290
+ 2026-10-03T17:40:17 step 71250 train 0.09839 inner-val 0.09868 best 0.09868@70750 (3779s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
291
+ 2026-10-03T17:40:30 step 71500 train 0.09788 inner-val 0.09868 best 0.09868@71500 (3791s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
292
+ 2026-10-03T17:40:43 step 71750 train 0.09790 inner-val 0.09867 best 0.09867@71750 (3804s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
293
+ 2026-10-03T17:40:56 step 72000 train 0.09807 inner-val 0.09867 best 0.09867@72000 (3816s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
294
+ 2026-10-03T17:41:08 step 72250 train 0.09759 inner-val 0.09867 best 0.09867@72250 (3829s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
295
+ 2026-10-03T17:41:19 step 72500 train 0.09839 inner-val 0.09867 best 0.09867@72250 (3841s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
296
+ 2026-10-03T17:41:32 step 72750 train 0.09802 inner-val 0.09867 best 0.09867@72750 (3852s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
297
+ 2026-10-03T17:41:43 step 73000 train 0.09757 inner-val 0.09867 best 0.09867@72750 (3865s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
298
+ 2026-10-03T17:41:56 step 73250 train 0.09809 inner-val 0.09866 best 0.09866@73250 (3876s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
299
+ 2026-10-03T17:42:07 step 73500 train 0.09805 inner-val 0.09866 best 0.09866@73250 (3889s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
300
+ 2026-10-03T17:42:19 step 73750 train 0.09819 inner-val 0.09866 best 0.09866@73250 (3901s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
301
+ 2026-10-03T17:42:32 step 74000 train 0.09764 inner-val 0.09866 best 0.09866@74000 (3912s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
302
+ 2026-10-03T17:42:43 step 74250 train 0.09789 inner-val 0.09866 best 0.09866@74000 (3925s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
303
+ 2026-10-03T17:42:56 step 74500 train 0.09798 inner-val 0.09866 best 0.09866@74500 (3936s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
304
+ 2026-10-03T17:43:07 step 74750 train 0.09794 inner-val 0.09866 best 0.09866@74500 (3949s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
305
+ 2026-10-03T17:43:18 step 75000 train 0.09822 inner-val 0.09866 best 0.09866@74500 (3960s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
306
+ 2026-10-03T17:43:30 step 75250 train 0.09771 inner-val 0.09866 best 0.09866@74500 (3972s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
307
+ 2026-10-03T17:43:42 step 75500 train 0.09800 inner-val 0.09865 best 0.09865@75500 (3983s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
308
+ 2026-10-03T17:43:55 step 75750 train 0.09787 inner-val 0.09865 best 0.09865@75750 (3995s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
309
+ 2026-10-03T17:44:08 step 76000 train 0.09793 inner-val 0.09865 best 0.09865@76000 (4008s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
310
+ 2026-10-03T17:44:19 step 76250 train 0.09806 inner-val 0.09865 best 0.09865@76000 (4021s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
311
+ 2026-10-03T17:44:32 step 76500 train 0.09783 inner-val 0.09864 best 0.09864@76500 (4032s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
312
+ 2026-10-03T17:44:45 step 76750 train 0.09792 inner-val 0.09864 best 0.09864@76750 (4046s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
313
+ 2026-10-03T17:44:56 step 77000 train 0.09818 inner-val 0.09864 best 0.09864@76750 (4058s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
314
+ 2026-10-03T17:45:08 step 77250 train 0.09786 inner-val 0.09864 best 0.09864@76750 (4070s) immicro=0.0971 imperial_eskd=0.0979 liu2021=0.1010
315
+ 2026-10-03T17:45:21 step 77500 train 0.09803 inner-val 0.09864 best 0.09864@77500 (4081s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
316
+ 2026-10-03T17:45:32 step 77750 train 0.09787 inner-val 0.09864 best 0.09864@77500 (4094s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
317
+ 2026-10-03T17:45:45 step 78000 train 0.09776 inner-val 0.09864 best 0.09864@78000 (4105s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
318
+ 2026-10-03T17:45:58 step 78250 train 0.09771 inner-val 0.09864 best 0.09864@78250 (4118s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
319
+ 2026-10-03T17:46:11 step 78500 train 0.09805 inner-val 0.09864 best 0.09864@78500 (4131s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
320
+ 2026-10-03T17:46:24 step 78750 train 0.09776 inner-val 0.09863 best 0.09863@78750 (4144s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
321
+ 2026-10-03T17:46:35 step 79000 train 0.09779 inner-val 0.09863 best 0.09863@78750 (4157s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
322
+ 2026-10-03T17:46:46 step 79250 train 0.09795 inner-val 0.09863 best 0.09863@78750 (4168s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
323
+ 2026-10-03T17:46:59 step 79500 train 0.09769 inner-val 0.09863 best 0.09863@79500 (4180s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
324
+ 2026-10-03T17:47:12 step 79750 train 0.09806 inner-val 0.09863 best 0.09863@79750 (4192s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
325
+ 2026-10-03T17:47:25 step 80000 train 0.09788 inner-val 0.09862 best 0.09862@80000 (4205s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
326
+ 2026-10-03T17:47:27 block end 80000: best 0.09862@80000; block improvement 0.000161 (1.67 % of total 0.009651) -> continue
327
+ 2026-10-03T17:47:40 step 80250 train 0.09796 inner-val 0.09862 best 0.09862@80250 (4220s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
328
+ 2026-10-03T17:47:52 step 80500 train 0.09787 inner-val 0.09862 best 0.09862@80500 (4233s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
329
+ 2026-10-03T17:48:04 step 80750 train 0.09786 inner-val 0.09861 best 0.09861@80750 (4245s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
330
+ 2026-10-03T17:48:17 step 81000 train 0.09791 inner-val 0.09861 best 0.09861@81000 (4257s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
331
+ 2026-10-03T17:48:29 step 81250 train 0.09753 inner-val 0.09860 best 0.09860@81250 (4270s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
332
+ 2026-10-03T17:48:40 step 81500 train 0.09804 inner-val 0.09861 best 0.09860@81250 (4282s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
333
+ 2026-10-03T17:48:52 step 81750 train 0.09796 inner-val 0.09860 best 0.09860@81250 (4294s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1010
334
+ 2026-10-03T17:49:03 step 82000 train 0.09780 inner-val 0.09860 best 0.09860@81250 (4305s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1009
335
+ 2026-10-03T17:49:16 step 82250 train 0.09783 inner-val 0.09860 best 0.09860@82250 (4316s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1009
336
+ 2026-10-03T17:49:27 step 82500 train 0.09791 inner-val 0.09860 best 0.09860@82250 (4329s) immicro=0.0970 imperial_eskd=0.0979 liu2021=0.1009
337
+ 2026-10-03T17:49:40 step 82750 train 0.09787 inner-val 0.09860 best 0.09860@82750 (4340s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
338
+ 2026-10-03T17:49:53 step 83000 train 0.09768 inner-val 0.09860 best 0.09860@83000 (4353s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
339
+ 2026-10-03T17:50:04 step 83250 train 0.09776 inner-val 0.09860 best 0.09860@83000 (4366s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
340
+ 2026-10-03T17:50:15 step 83500 train 0.09759 inner-val 0.09860 best 0.09860@83000 (4377s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1010
341
+ 2026-10-03T17:50:26 step 83750 train 0.09797 inner-val 0.09860 best 0.09860@83000 (4388s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1010
342
+ 2026-10-03T17:50:39 step 84000 train 0.09757 inner-val 0.09859 best 0.09859@84000 (4399s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
343
+ 2026-10-03T17:50:50 step 84250 train 0.09754 inner-val 0.09859 best 0.09859@84000 (4412s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
344
+ 2026-10-03T17:51:03 step 84500 train 0.09751 inner-val 0.09859 best 0.09859@84500 (4423s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
345
+ 2026-10-03T17:51:16 step 84750 train 0.09791 inner-val 0.09858 best 0.09858@84750 (4436s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
346
+ 2026-10-03T17:51:27 step 85000 train 0.09786 inner-val 0.09858 best 0.09858@84750 (4449s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
347
+ 2026-10-03T17:51:40 step 85250 train 0.09801 inner-val 0.09858 best 0.09858@85250 (4460s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
348
+ 2026-10-03T17:51:52 step 85500 train 0.09783 inner-val 0.09858 best 0.09858@85500 (4473s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
349
+ 2026-10-03T17:52:05 step 85750 train 0.09803 inner-val 0.09858 best 0.09858@85750 (4485s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
350
+ 2026-10-03T17:52:16 step 86000 train 0.09783 inner-val 0.09858 best 0.09858@85750 (4498s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
351
+ 2026-10-03T17:52:29 step 86250 train 0.09789 inner-val 0.09858 best 0.09858@86250 (4509s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
352
+ 2026-10-03T17:52:42 step 86500 train 0.09767 inner-val 0.09857 best 0.09857@86500 (4522s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
353
+ 2026-10-03T17:52:55 step 86750 train 0.09734 inner-val 0.09857 best 0.09857@86750 (4535s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
354
+ 2026-10-03T17:53:08 step 87000 train 0.09779 inner-val 0.09857 best 0.09857@87000 (4548s) immicro=0.0970 imperial_eskd=0.0978 liu2021=0.1009
355
+ 2026-10-03T17:53:20 step 87250 train 0.09768 inner-val 0.09857 best 0.09857@87250 (4561s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
356
+ 2026-10-03T17:53:33 step 87500 train 0.09779 inner-val 0.09856 best 0.09856@87500 (4574s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
357
+ 2026-10-03T17:53:46 step 87750 train 0.09778 inner-val 0.09856 best 0.09856@87750 (4586s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
358
+ 2026-10-03T17:53:57 step 88000 train 0.09785 inner-val 0.09856 best 0.09856@87750 (4599s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
359
+ 2026-10-03T17:54:08 step 88250 train 0.09764 inner-val 0.09856 best 0.09856@87750 (4610s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
360
+ 2026-10-03T17:54:21 step 88500 train 0.09770 inner-val 0.09855 best 0.09855@88500 (4621s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
361
+ 2026-10-03T17:54:34 step 88750 train 0.09769 inner-val 0.09855 best 0.09855@88750 (4634s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
362
+ 2026-10-03T17:54:47 step 89000 train 0.09809 inner-val 0.09855 best 0.09855@89000 (4647s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
363
+ 2026-10-03T17:54:59 step 89250 train 0.09813 inner-val 0.09855 best 0.09855@89250 (4660s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
364
+ 2026-10-03T17:55:12 step 89500 train 0.09787 inner-val 0.09854 best 0.09854@89500 (4673s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
365
+ 2026-10-03T17:55:25 step 89750 train 0.09783 inner-val 0.09854 best 0.09854@89750 (4686s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
366
+ 2026-10-03T17:55:38 step 90000 train 0.09768 inner-val 0.09854 best 0.09854@90000 (4698s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
367
+ 2026-10-03T17:55:49 step 90250 train 0.09758 inner-val 0.09854 best 0.09854@90000 (4711s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
368
+ 2026-10-03T17:56:02 step 90500 train 0.09778 inner-val 0.09854 best 0.09854@90500 (4722s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
369
+ 2026-10-03T17:56:13 step 90750 train 0.09761 inner-val 0.09853 best 0.09854@90500 (4735s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
370
+ 2026-10-03T17:56:26 step 91000 train 0.09787 inner-val 0.09853 best 0.09853@91000 (4746s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
371
+ 2026-10-03T17:56:38 step 91250 train 0.09767 inner-val 0.09853 best 0.09853@91250 (4759s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
372
+ 2026-10-03T17:56:49 step 91500 train 0.09761 inner-val 0.09853 best 0.09853@91250 (4771s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
373
+ 2026-10-03T17:57:02 step 91750 train 0.09754 inner-val 0.09853 best 0.09853@91750 (4783s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
374
+ 2026-10-03T17:57:13 step 92000 train 0.09772 inner-val 0.09853 best 0.09853@91750 (4795s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
375
+ 2026-10-03T17:57:25 step 92250 train 0.09771 inner-val 0.09853 best 0.09853@91750 (4807s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
376
+ 2026-10-03T17:57:36 step 92500 train 0.09754 inner-val 0.09853 best 0.09853@91750 (4818s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
377
+ 2026-10-03T17:57:47 step 92750 train 0.09758 inner-val 0.09853 best 0.09853@91750 (4829s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
378
+ 2026-10-03T17:57:58 step 93000 train 0.09775 inner-val 0.09853 best 0.09853@91750 (4840s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
379
+ 2026-10-03T17:58:09 step 93250 train 0.09761 inner-val 0.09853 best 0.09853@91750 (4851s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
380
+ 2026-10-03T17:58:21 step 93500 train 0.09778 inner-val 0.09853 best 0.09853@91750 (4862s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
381
+ 2026-10-03T17:58:33 step 93750 train 0.09750 inner-val 0.09853 best 0.09853@93750 (4874s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
382
+ 2026-10-03T17:58:45 step 94000 train 0.09807 inner-val 0.09853 best 0.09853@93750 (4886s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
383
+ 2026-10-03T17:58:56 step 94250 train 0.09753 inner-val 0.09853 best 0.09853@93750 (4898s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
384
+ 2026-10-03T17:59:09 step 94500 train 0.09783 inner-val 0.09852 best 0.09852@94500 (4909s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
385
+ 2026-10-03T17:59:21 step 94750 train 0.09753 inner-val 0.09852 best 0.09852@94750 (4922s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
386
+ 2026-10-03T17:59:32 step 95000 train 0.09778 inner-val 0.09852 best 0.09852@94750 (4934s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
387
+ 2026-10-03T17:59:45 step 95250 train 0.09755 inner-val 0.09851 best 0.09851@95250 (4946s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
388
+ 2026-10-03T17:59:58 step 95500 train 0.09764 inner-val 0.09851 best 0.09851@95500 (4958s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
389
+ 2026-10-03T18:00:09 step 95750 train 0.09779 inner-val 0.09851 best 0.09851@95500 (4971s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
390
+ 2026-10-03T18:00:20 step 96000 train 0.09748 inner-val 0.09851 best 0.09851@95500 (4982s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
391
+ 2026-10-03T18:00:32 step 96250 train 0.09791 inner-val 0.09851 best 0.09851@95500 (4994s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
392
+ 2026-10-03T18:00:43 step 96500 train 0.09772 inner-val 0.09851 best 0.09851@95500 (5005s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
393
+ 2026-10-03T18:00:54 step 96750 train 0.09736 inner-val 0.09851 best 0.09851@95500 (5016s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
394
+ 2026-10-03T18:01:07 step 97000 train 0.09750 inner-val 0.09851 best 0.09851@97000 (5027s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
395
+ 2026-10-03T18:01:20 step 97250 train 0.09761 inner-val 0.09850 best 0.09850@97250 (5040s) immicro=0.0969 imperial_eskd=0.0978 liu2021=0.1009
396
+ 2026-10-03T18:01:32 step 97500 train 0.09745 inner-val 0.09850 best 0.09850@97500 (5053s) immicro=0.0968 imperial_eskd=0.0978 liu2021=0.1009
397
+ 2026-10-03T18:01:45 step 97750 train 0.09790 inner-val 0.09850 best 0.09850@97750 (5065s) immicro=0.0968 imperial_eskd=0.0978 liu2021=0.1009
398
+ 2026-10-03T18:01:56 step 98000 train 0.09790 inner-val 0.09850 best 0.09850@97750 (5078s) immicro=0.0968 imperial_eskd=0.0978 liu2021=0.1009
399
+ 2026-10-03T18:02:08 step 98250 train 0.09759 inner-val 0.09850 best 0.09850@97750 (5090s) immicro=0.0968 imperial_eskd=0.0978 liu2021=0.1009
400
+ 2026-10-03T18:02:19 step 98500 train 0.09753 inner-val 0.09850 best 0.09850@97750 (5101s) immicro=0.0968 imperial_eskd=0.0978 liu2021=0.1009
401
+ 2026-10-03T18:02:31 step 98750 train 0.09747 inner-val 0.09850 best 0.09850@98750 (5112s) immicro=0.0968 imperial_eskd=0.0978 liu2021=0.1009
402
+ 2026-10-03T18:02:44 step 99000 train 0.09797 inner-val 0.09849 best 0.09849@99000 (5125s) immicro=0.0968 imperial_eskd=0.0978 liu2021=0.1009
403
+ 2026-10-03T18:02:57 step 99250 train 0.09767 inner-val 0.09849 best 0.09849@99250 (5138s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
404
+ 2026-10-03T18:03:08 step 99500 train 0.09737 inner-val 0.09849 best 0.09849@99250 (5150s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
405
+ 2026-10-03T18:03:21 step 99750 train 0.09755 inner-val 0.09849 best 0.09849@99750 (5162s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
406
+ 2026-10-03T18:03:34 step 100000 train 0.09790 inner-val 0.09848 best 0.09848@100000 (5174s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
407
+ 2026-10-03T18:03:36 block end 100000: best 0.09848@100000; block improvement 0.000141 (1.44 % of total 0.009792) -> continue
408
+ 2026-10-03T18:03:49 step 100250 train 0.09749 inner-val 0.09848 best 0.09848@100250 (5189s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
409
+ 2026-10-03T18:04:00 step 100500 train 0.09743 inner-val 0.09848 best 0.09848@100250 (5202s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
410
+ 2026-10-03T18:04:12 step 100750 train 0.09776 inner-val 0.09848 best 0.09848@100750 (5213s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
411
+ 2026-10-03T18:04:24 step 101000 train 0.09736 inner-val 0.09848 best 0.09848@100750 (5226s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
412
+ 2026-10-03T18:04:36 step 101250 train 0.09748 inner-val 0.09848 best 0.09848@101250 (5237s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
413
+ 2026-10-03T18:04:49 step 101500 train 0.09748 inner-val 0.09847 best 0.09847@101500 (5250s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
414
+ 2026-10-03T18:05:00 step 101750 train 0.09759 inner-val 0.09847 best 0.09847@101500 (5262s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
415
+ 2026-10-03T18:05:11 step 102000 train 0.09733 inner-val 0.09847 best 0.09847@101500 (5273s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
416
+ 2026-10-03T18:05:23 step 102250 train 0.09740 inner-val 0.09847 best 0.09847@101500 (5285s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
417
+ 2026-10-03T18:05:34 step 102500 train 0.09761 inner-val 0.09847 best 0.09847@101500 (5296s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
418
+ 2026-10-03T18:05:45 step 102750 train 0.09755 inner-val 0.09847 best 0.09847@101500 (5307s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
419
+ 2026-10-03T18:05:58 step 103000 train 0.09748 inner-val 0.09847 best 0.09847@103000 (5318s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
420
+ 2026-10-03T18:06:10 step 103250 train 0.09761 inner-val 0.09847 best 0.09847@103250 (5331s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
421
+ 2026-10-03T18:06:23 step 103500 train 0.09755 inner-val 0.09846 best 0.09846@103500 (5343s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
422
+ 2026-10-03T18:06:34 step 103750 train 0.09750 inner-val 0.09846 best 0.09846@103500 (5356s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1009
423
+ 2026-10-03T18:06:45 step 104000 train 0.09770 inner-val 0.09846 best 0.09846@103500 (5367s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
424
+ 2026-10-03T18:06:58 step 104250 train 0.09751 inner-val 0.09846 best 0.09846@104250 (5378s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
425
+ 2026-10-03T18:07:09 step 104500 train 0.09789 inner-val 0.09846 best 0.09846@104250 (5391s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
426
+ 2026-10-03T18:07:23 step 104750 train 0.09756 inner-val 0.09845 best 0.09845@104750 (5403s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
427
+ 2026-10-03T18:07:36 step 105000 train 0.09741 inner-val 0.09845 best 0.09845@104750 (5418s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
428
+ 2026-10-03T18:07:48 step 105250 train 0.09760 inner-val 0.09845 best 0.09845@104750 (5430s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
429
+ 2026-10-03T18:08:03 step 105500 train 0.09719 inner-val 0.09845 best 0.09845@105500 (5443s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
430
+ 2026-10-03T18:08:16 step 105750 train 0.09746 inner-val 0.09844 best 0.09844@105750 (5457s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
431
+ 2026-10-03T18:08:30 step 106000 train 0.09747 inner-val 0.09844 best 0.09844@106000 (5471s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
432
+ 2026-10-03T18:08:42 step 106250 train 0.09783 inner-val 0.09844 best 0.09844@106000 (5484s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
433
+ 2026-10-03T18:08:56 step 106500 train 0.09749 inner-val 0.09844 best 0.09844@106500 (5497s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
434
+ 2026-10-03T18:09:09 step 106750 train 0.09780 inner-val 0.09844 best 0.09844@106500 (5511s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
435
+ 2026-10-03T18:09:21 step 107000 train 0.09753 inner-val 0.09844 best 0.09844@106500 (5523s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
436
+ 2026-10-03T18:09:34 step 107250 train 0.09736 inner-val 0.09844 best 0.09844@106500 (5536s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
437
+ 2026-10-03T18:09:46 step 107500 train 0.09730 inner-val 0.09844 best 0.09844@106500 (5548s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
438
+ 2026-10-03T18:09:58 step 107750 train 0.09739 inner-val 0.09844 best 0.09844@106500 (5560s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
439
+ 2026-10-03T18:10:11 step 108000 train 0.09753 inner-val 0.09844 best 0.09844@106500 (5573s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
440
+ 2026-10-03T18:10:23 step 108250 train 0.09761 inner-val 0.09844 best 0.09844@106500 (5585s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
441
+ 2026-10-03T18:10:36 step 108500 train 0.09755 inner-val 0.09844 best 0.09844@106500 (5598s) immicro=0.0968 imperial_eskd=0.0977 liu2021=0.1008
442
+ 2026-10-03T18:10:36 early stop: no improvement in 8 evaluations
443
+ 2026-10-03T18:10:43 saved /home/sj_server_1/PBISC/immun-pbmc-fm/model_8_sparse_data_diffusion/plans/paired_crosscohort_20261002/runs/hoL_ye2026_sepsis/K1000/fullsat/checkpoint_ft.pt (selected step 106500, inner-val 0.09844 vs pretrained 0.10828); sha 9240379953f4