Upload folder using huggingface_hub

haar_v4_ungated_768/20260727_162937/vis_data/20260727_162937.json

@@ -0,0 +1,80 @@
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.02387575626373291, "grad_norm": 114.52095680236816, "loss": 3.641382517814636, "loss_cls": 0.8087327021360398, "loss_bbox": 1.3330383479595185, "loss_centerness": 0.6507055377960205, "loss_pos": 0.8346495192125439, "loss_offset": 0.014256374315591529, "time": 0.1746406936645508, "epoch": 1, "iter": 50, "memory": 6269, "step": 50}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.016263041496276855, "grad_norm": 16.76171932220459, "loss": 1.9775516772270203, "loss_cls": 0.32804525196552276, "loss_bbox": 0.7804536592960357, "loss_centerness": 0.627768611907959, "loss_pos": 0.23205391455441712, "loss_offset": 0.009230252052657307, "time": 0.13671158790588378, "epoch": 1, "iter": 100, "memory": 6269, "step": 100}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015863189697265623, "grad_norm": 12.959484844207763, "loss": 1.8052734732627869, "loss_cls": 0.25144625425338746, "loss_bbox": 0.7720166838169098, "loss_centerness": 0.6131773805618286, "loss_pos": 0.16167937211692332, "loss_offset": 0.006953791005071252, "time": 0.13558685302734375, "epoch": 1, "iter": 150, "memory": 6269, "step": 150}
+{"coco/bbox_mAP": 0.112, "coco/bbox_mAP_50": 0.47, "coco/bbox_mAP_75": 0.003, "coco/bbox_mAP_s": 0.113, "coco/bbox_mAP_m": 0.112, "coco/bbox_mAP_l": -1.0, "data_time": 0.0027687876191857743, "time": 0.014944686056816415, "step": 1}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.017955560684204102, "grad_norm": 10.754822750091552, "loss": 1.522691957950592, "loss_cls": 0.20511740639805795, "loss_bbox": 0.5785106360912323, "loss_centerness": 0.601420282125473, "loss_pos": 0.12888413108885288, "loss_offset": 0.008759502851171419, "time": 0.1390164613723755, "epoch": 2, "iter": 208, "memory": 6269, "step": 208}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015726680755615233, "grad_norm": 9.802234706878663, "loss": 1.4951458525657655, "loss_cls": 0.19281451269984246, "loss_bbox": 0.5602053409814834, "loss_centerness": 0.5924474763870239, "loss_pos": 0.14257961828261614, "loss_offset": 0.007098915867973119, "time": 0.13476008415222168, "epoch": 2, "iter": 258, "memory": 6269, "step": 258}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015753674507141113, "grad_norm": 10.286860389709473, "loss": 1.4608097052574158, "loss_cls": 0.17279406234622002, "loss_bbox": 0.5809448206424713, "loss_centerness": 0.5883310759067535, "loss_pos": 0.11235435826703906, "loss_offset": 0.006385399708524347, "time": 0.13440067291259766, "epoch": 2, "iter": 308, "memory": 6269, "step": 308}
+{"coco/bbox_mAP": 0.21, "coco/bbox_mAP_50": 0.674, "coco/bbox_mAP_75": 0.037, "coco/bbox_mAP_s": 0.213, "coco/bbox_mAP_m": 0.159, "coco/bbox_mAP_l": -1.0, "data_time": 0.0020721525208562868, "time": 0.013429028763730302, "step": 2}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.01754068851470947, "grad_norm": 9.125817556381225, "loss": 1.365412051677704, "loss_cls": 0.15463197439908982, "loss_bbox": 0.5052433621883392, "loss_centerness": 0.5871367955207825, "loss_pos": 0.11145446598529815, "loss_offset": 0.006945455521345139, "time": 0.13633635520935058, "epoch": 3, "iter": 366, "memory": 6269, "step": 366}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.016068544387817383, "grad_norm": 9.999114599227905, "loss": 1.393318359851837, "loss_cls": 0.1690808902680874, "loss_bbox": 0.5191915154457092, "loss_centerness": 0.5846550559997559, "loss_pos": 0.1127777398005128, "loss_offset": 0.007613137498265132, "time": 0.13512993335723877, "epoch": 3, "iter": 416, "memory": 6269, "step": 416}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015846266746520996, "grad_norm": 9.580743188858031, "loss": 1.3606600379943847, "loss_cls": 0.14454242944717408, "loss_bbox": 0.5217189937829971, "loss_centerness": 0.581095484495163, "loss_pos": 0.10637437952682376, "loss_offset": 0.006928750155493617, "time": 0.13455546855926515, "epoch": 3, "iter": 466, "memory": 6269, "step": 466}
+{"coco/bbox_mAP": 0.095, "coco/bbox_mAP_50": 0.403, "coco/bbox_mAP_75": 0.015, "coco/bbox_mAP_s": 0.091, "coco/bbox_mAP_m": 0.262, "coco/bbox_mAP_l": -1.0, "data_time": 0.001969819598727756, "time": 0.013254984423645541, "step": 3}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.017896056175231934, "grad_norm": 11.886852989196777, "loss": 1.5761951184272767, "loss_cls": 0.1883251576125622, "loss_bbox": 0.6712099689245224, "loss_centerness": 0.5823398959636689, "loss_pos": 0.12729708219412714, "loss_offset": 0.007023014086298645, "time": 0.13741453647613525, "epoch": 4, "iter": 524, "memory": 6269, "step": 524}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01574042797088623, "grad_norm": 11.804976673126221, "loss": 1.8784850573539733, "loss_cls": 0.3001200720667839, "loss_bbox": 0.8040754497051239, "loss_centerness": 0.6108310925960541, "loss_pos": 0.155565629824996, "loss_offset": 0.007892816206440329, "time": 0.13491064548492432, "epoch": 4, "iter": 574, "memory": 6269, "step": 574}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015125856399536133, "grad_norm": 8.498672256469726, "loss": 1.475808184146881, "loss_cls": 0.19704236343502998, "loss_bbox": 0.5570360046625137, "loss_centerness": 0.5899317514896393, "loss_pos": 0.12457641873508692, "loss_offset": 0.0072216421773191545, "time": 0.13200510025024415, "epoch": 4, "iter": 624, "memory": 6269, "step": 624}
+{"coco/bbox_mAP": 0.074, "coco/bbox_mAP_50": 0.327, "coco/bbox_mAP_75": 0.006, "coco/bbox_mAP_s": 0.075, "coco/bbox_mAP_m": 0.012, "coco/bbox_mAP_l": -1.0, "data_time": 0.0020367088480892346, "time": 0.013232316318740193, "step": 4}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.017707514762878417, "grad_norm": 10.190532112121582, "loss": 1.5970463275909423, "loss_cls": 0.2575735272467136, "loss_bbox": 0.6156666225194931, "loss_centerness": 0.5960467672348022, "loss_pos": 0.12227158192545176, "loss_offset": 0.005487838343251497, "time": 0.13671816349029542, "epoch": 5, "iter": 682, "memory": 6269, "step": 682}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016632404327392578, "grad_norm": 7.975882334709167, "loss": 1.4121679091453552, "loss_cls": 0.16413942188024522, "loss_bbox": 0.5541865700483322, "loss_centerness": 0.580131139755249, "loss_pos": 0.10705584756098688, "loss_offset": 0.006654911576770246, "time": 0.1359233331680298, "epoch": 5, "iter": 732, "memory": 6269, "step": 732}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015917253494262696, "grad_norm": 8.996517276763916, "loss": 1.4886412334442138, "loss_cls": 0.15306309029459952, "loss_bbox": 0.634658961892128, "loss_centerness": 0.5835314714908599, "loss_pos": 0.11071547254920006, "loss_offset": 0.006672249686671421, "time": 0.13437987804412843, "epoch": 5, "iter": 782, "memory": 6269, "step": 782}
+{"coco/bbox_mAP": 0.161, "coco/bbox_mAP_50": 0.613, "coco/bbox_mAP_75": 0.016, "coco/bbox_mAP_s": 0.161, "coco/bbox_mAP_m": 0.18, "coco/bbox_mAP_l": -1.0, "data_time": 0.001958650197738256, "time": 0.013422944810655382, "step": 5}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018230013847351074, "grad_norm": 7.757193851470947, "loss": 1.3803178119659423, "loss_cls": 0.1524515801668167, "loss_bbox": 0.5319341343641281, "loss_centerness": 0.581254894733429, "loss_pos": 0.10909361463040114, "loss_offset": 0.005583584911073558, "time": 0.13743532657623292, "epoch": 6, "iter": 840, "memory": 6269, "step": 840}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015194315910339356, "grad_norm": 7.847311778068542, "loss": 1.432754168510437, "loss_cls": 0.15826014041900635, "loss_bbox": 0.5617603421211242, "loss_centerness": 0.5805632305145264, "loss_pos": 0.1260861048847437, "loss_offset": 0.0060843450643005784, "time": 0.13362086296081543, "epoch": 6, "iter": 890, "memory": 6269, "step": 890}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015460410118103028, "grad_norm": 8.59046314239502, "loss": 1.3929545879364014, "loss_cls": 0.1539978276193142, "loss_bbox": 0.5722000098228455, "loss_centerness": 0.5774772417545319, "loss_pos": 0.08417099100537598, "loss_offset": 0.005108523470116779, "time": 0.1328950357437134, "epoch": 6, "iter": 940, "memory": 6269, "step": 940}
+{"coco/bbox_mAP": 0.181, "coco/bbox_mAP_50": 0.608, "coco/bbox_mAP_75": 0.032, "coco/bbox_mAP_s": 0.178, "coco/bbox_mAP_m": 0.253, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019063554258428067, "time": 0.012651062419271877, "step": 6}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.017206239700317382, "grad_norm": 6.785709195137024, "loss": 1.2810506629943847, "loss_cls": 0.12734990701079368, "loss_bbox": 0.4799131882190704, "loss_centerness": 0.575744092464447, "loss_pos": 0.09377483788877726, "loss_offset": 0.004268641428789124, "time": 0.13568567752838134, "epoch": 7, "iter": 998, "memory": 6269, "step": 998}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01695356845855713, "grad_norm": 7.346939630508423, "loss": 1.3278966546058655, "loss_cls": 0.13488397777080535, "loss_bbox": 0.5170796245336533, "loss_centerness": 0.5783039677143097, "loss_pos": 0.09240393787622452, "loss_offset": 0.005225151426857338, "time": 0.13568556785583497, "epoch": 7, "iter": 1048, "memory": 6269, "step": 1048}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015588855743408203, "grad_norm": 8.084754962921142, "loss": 1.3997604441642761, "loss_cls": 0.12984258115291594, "loss_bbox": 0.6016161435842514, "loss_centerness": 0.5777574527263641, "loss_pos": 0.08483651880174875, "loss_offset": 0.0057077554706484075, "time": 0.1334227752685547, "epoch": 7, "iter": 1098, "memory": 6269, "step": 1098}
+{"coco/bbox_mAP": 0.24, "coco/bbox_mAP_50": 0.698, "coco/bbox_mAP_75": 0.055, "coco/bbox_mAP_s": 0.241, "coco/bbox_mAP_m": 0.17, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019307633750459067, "time": 0.012958384375286918, "step": 7}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018718342781066894, "grad_norm": 7.671614689826965, "loss": 1.302541148662567, "loss_cls": 0.13622799679636954, "loss_bbox": 0.5055729216337204, "loss_centerness": 0.5755344927310944, "loss_pos": 0.07973715897649526, "loss_offset": 0.0054685806878842416, "time": 0.1391748857498169, "epoch": 8, "iter": 1156, "memory": 6269, "step": 1156}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015604238510131836, "grad_norm": 7.33595838546753, "loss": 1.3371935772895813, "loss_cls": 0.13143601655960083, "loss_bbox": 0.5303419882059097, "loss_centerness": 0.5752888548374177, "loss_pos": 0.0946077877189964, "loss_offset": 0.005518921547336504, "time": 0.13442244052886962, "epoch": 8, "iter": 1206, "memory": 6269, "step": 1206}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015556530952453613, "grad_norm": 6.322084312438965, "loss": 1.2781175518035888, "loss_cls": 0.12771683260798455, "loss_bbox": 0.4876609718799591, "loss_centerness": 0.5766118216514587, "loss_pos": 0.08130903821438551, "loss_offset": 0.004818873853655532, "time": 0.1342010259628296, "epoch": 8, "iter": 1256, "memory": 6269, "step": 1256}
+{"coco/bbox_mAP": 0.044, "coco/bbox_mAP_50": 0.199, "coco/bbox_mAP_75": 0.002, "coco/bbox_mAP_s": 0.045, "coco/bbox_mAP_m": 0.183, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018745055565467248, "time": 0.012709898826403495, "step": 8}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016962690353393553, "grad_norm": 7.528588318824768, "loss": 1.3156990897655487, "loss_cls": 0.13987344846129418, "loss_bbox": 0.5230728405714035, "loss_centerness": 0.5738614439964295, "loss_pos": 0.0742860192619264, "loss_offset": 0.004605343978619203, "time": 0.13479403018951416, "epoch": 9, "iter": 1314, "memory": 6269, "step": 1314}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015400733947753906, "grad_norm": 6.583463411331177, "loss": 1.2535515475273131, "loss_cls": 0.1289924655854702, "loss_bbox": 0.4492087346315384, "loss_centerness": 0.5746508133411408, "loss_pos": 0.09588879181072116, "loss_offset": 0.00481074296112638, "time": 0.133740873336792, "epoch": 9, "iter": 1364, "memory": 6269, "step": 1364}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016284708976745606, "grad_norm": 6.643251657485962, "loss": 1.2628607630729676, "loss_cls": 0.12052003189921379, "loss_bbox": 0.47820280969142914, "loss_centerness": 0.5739504361152649, "loss_pos": 0.08566969852894545, "loss_offset": 0.004517789883539081, "time": 0.13424815177917482, "epoch": 9, "iter": 1414, "memory": 6269, "step": 1414}
+{"coco/bbox_mAP": 0.264, "coco/bbox_mAP_50": 0.746, "coco/bbox_mAP_75": 0.091, "coco/bbox_mAP_s": 0.267, "coco/bbox_mAP_m": 0.145, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018993536631266275, "time": 0.01278637820838863, "step": 9}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018471441268920898, "grad_norm": 6.325474667549133, "loss": 1.2168771278858186, "loss_cls": 0.12575486034154892, "loss_bbox": 0.44242464542388915, "loss_centerness": 0.5722889256477356, "loss_pos": 0.07150071470066904, "loss_offset": 0.004907989402418025, "time": 0.1380993604660034, "epoch": 10, "iter": 1472, "memory": 6269, "step": 1472}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016656055450439453, "grad_norm": 7.229538035392761, "loss": 1.2897097182273864, "loss_cls": 0.12006522357463836, "loss_bbox": 0.5167982870340347, "loss_centerness": 0.5739198541641235, "loss_pos": 0.07485429923981428, "loss_offset": 0.004072059815516695, "time": 0.13608648777008056, "epoch": 10, "iter": 1522, "memory": 6269, "step": 1522}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01547323226928711, "grad_norm": 6.1348348808288575, "loss": 1.2192114770412446, "loss_cls": 0.11673817545175552, "loss_bbox": 0.44963080644607545, "loss_centerness": 0.5735272967815399, "loss_pos": 0.07516739720478655, "loss_offset": 0.004147799819184002, "time": 0.13438985347747803, "epoch": 10, "iter": 1572, "memory": 6269, "step": 1572}
+{"coco/bbox_mAP": 0.21, "coco/bbox_mAP_50": 0.651, "coco/bbox_mAP_75": 0.057, "coco/bbox_mAP_s": 0.214, "coco/bbox_mAP_m": 0.063, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019964625692775106, "time": 0.013408109061738364, "step": 10}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01780747890472412, "grad_norm": 6.169156465530396, "loss": 1.2160494685173036, "loss_cls": 0.11702308133244514, "loss_bbox": 0.45238163590431213, "loss_centerness": 0.5727790057659149, "loss_pos": 0.06946244772523641, "loss_offset": 0.004403298212491791, "time": 0.13704002380371094, "epoch": 11, "iter": 1630, "memory": 6269, "step": 1630}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015589737892150879, "grad_norm": 5.870634632110596, "loss": 1.159477950334549, "loss_cls": 0.10759207174181938, "loss_bbox": 0.41125225603580473, "loss_centerness": 0.5707440173625946, "loss_pos": 0.06594253187067807, "loss_offset": 0.003947081957012415, "time": 0.13382294178009033, "epoch": 11, "iter": 1680, "memory": 6269, "step": 1680}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015575742721557618, "grad_norm": 7.311523427963257, "loss": 1.2527956891059875, "loss_cls": 0.11181208297610283, "loss_bbox": 0.49854174852371214, "loss_centerness": 0.5704759740829468, "loss_pos": 0.06851734125986696, "loss_offset": 0.0034485349094029514, "time": 0.1344086742401123, "epoch": 11, "iter": 1730, "memory": 6269, "step": 1730}
+{"coco/bbox_mAP": 0.218, "coco/bbox_mAP_50": 0.655, "coco/bbox_mAP_75": 0.068, "coco/bbox_mAP_s": 0.218, "coco/bbox_mAP_m": 0.203, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019610025943853915, "time": 0.01311383777194553, "step": 11}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.0177852725982666, "grad_norm": 6.048166790008545, "loss": 1.199417164325714, "loss_cls": 0.1181537875533104, "loss_bbox": 0.44670230507850645, "loss_centerness": 0.5699885952472686, "loss_pos": 0.0609135265275836, "loss_offset": 0.0036589622881729156, "time": 0.13667559146881103, "epoch": 12, "iter": 1788, "memory": 6269, "step": 1788}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015840106010437012, "grad_norm": 7.208025574684143, "loss": 1.2247644114494323, "loss_cls": 0.11180497005581856, "loss_bbox": 0.47382911264896393, "loss_centerness": 0.5729782819747925, "loss_pos": 0.062134848015848544, "loss_offset": 0.004017203194380272, "time": 0.13444390773773193, "epoch": 12, "iter": 1838, "memory": 6269, "step": 1838}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018442683219909668, "grad_norm": 5.660691437721252, "loss": 1.1556305146217347, "loss_cls": 0.10315977282822132, "loss_bbox": 0.4162970733642578, "loss_centerness": 0.5688820195198059, "loss_pos": 0.0637389025464654, "loss_offset": 0.003552732290845597, "time": 0.13917843818664552, "epoch": 12, "iter": 1888, "memory": 6269, "step": 1888}
+{"coco/bbox_mAP": 0.157, "coco/bbox_mAP_50": 0.571, "coco/bbox_mAP_75": 0.02, "coco/bbox_mAP_s": 0.16, "coco/bbox_mAP_m": 0.002, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019042418553278997, "time": 0.012593824843056182, "step": 12}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01787266731262207, "grad_norm": 6.075709481239318, "loss": 1.1512815499305724, "loss_cls": 0.09467467859387398, "loss_bbox": 0.4316812252998352, "loss_centerness": 0.5683929216861725, "loss_pos": 0.05296651531010866, "loss_offset": 0.003566218525520526, "time": 0.1355909776687622, "epoch": 13, "iter": 1946, "memory": 6269, "step": 1946}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015505547523498536, "grad_norm": 5.894303050041199, "loss": 1.142732938528061, "loss_cls": 0.10125300385057927, "loss_bbox": 0.4011682283878326, "loss_centerness": 0.5691057133674622, "loss_pos": 0.06723949652630835, "loss_offset": 0.0039665168119245205, "time": 0.13348616123199464, "epoch": 13, "iter": 1996, "memory": 6269, "step": 1996}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015311622619628906, "grad_norm": 5.925397968292236, "loss": 1.1735150337219238, "loss_cls": 0.1086564926803112, "loss_bbox": 0.43431043744087217, "loss_centerness": 0.570909628868103, "loss_pos": 0.056253279503434894, "loss_offset": 0.003385203641373664, "time": 0.13318418979644775, "epoch": 13, "iter": 2046, "memory": 6269, "step": 2046}
+{"coco/bbox_mAP": 0.193, "coco/bbox_mAP_50": 0.632, "coco/bbox_mAP_75": 0.035, "coco/bbox_mAP_s": 0.197, "coco/bbox_mAP_m": 0.098, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018503535507071731, "time": 0.012584523258046208, "step": 13}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.017718935012817384, "grad_norm": 4.6375644540786745, "loss": 1.0503401231765748, "loss_cls": 0.09265139251947403, "loss_bbox": 0.3370001268386841, "loss_centerness": 0.5646761882305146, "loss_pos": 0.0524382827244699, "loss_offset": 0.003574130265333224, "time": 0.13637757778167725, "epoch": 14, "iter": 2104, "memory": 6269, "step": 2104}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015170278549194336, "grad_norm": 3.8037463188171388, "loss": 0.986722184419632, "loss_cls": 0.0801999495178461, "loss_bbox": 0.30282890737056734, "loss_centerness": 0.5650018358230591, "loss_pos": 0.03569386578630656, "loss_offset": 0.0029976315080421046, "time": 0.13306628704071044, "epoch": 14, "iter": 2154, "memory": 6269, "step": 2154}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01542081356048584, "grad_norm": 4.260610308647156, "loss": 0.9856020987033844, "loss_cls": 0.07983840823173523, "loss_bbox": 0.2959664106369019, "loss_centerness": 0.565356639623642, "loss_pos": 0.041319113122299314, "loss_offset": 0.003121536517282948, "time": 0.13338305473327636, "epoch": 14, "iter": 2204, "memory": 6269, "step": 2204}
+{"coco/bbox_mAP": 0.291, "coco/bbox_mAP_50": 0.774, "coco/bbox_mAP_75": 0.127, "coco/bbox_mAP_s": 0.291, "coco/bbox_mAP_m": 0.26, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018406700884175096, "time": 0.012511734269623064, "step": 14}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01903491497039795, "grad_norm": 4.047395715713501, "loss": 0.9670887410640716, "loss_cls": 0.07777763083577156, "loss_bbox": 0.2950610908865929, "loss_centerness": 0.5628402543067932, "loss_pos": 0.028107117956969886, "loss_offset": 0.003302650565165095, "time": 0.1387276029586792, "epoch": 15, "iter": 2262, "memory": 6269, "step": 2262}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.016506352424621583, "grad_norm": 4.267009086608887, "loss": 0.9722354638576508, "loss_cls": 0.0777778397500515, "loss_bbox": 0.29410224497318266, "loss_centerness": 0.5635494661331176, "loss_pos": 0.03424645137274638, "loss_offset": 0.002559460891716299, "time": 0.1359393835067749, "epoch": 15, "iter": 2312, "memory": 6269, "step": 2312}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015163192749023438, "grad_norm": 4.381349997520447, "loss": 0.9859471487998962, "loss_cls": 0.08225745782256126, "loss_bbox": 0.29726907461881635, "loss_centerness": 0.5630013751983642, "loss_pos": 0.03976258021779358, "loss_offset": 0.0036566603364190085, "time": 0.1327739477157593, "epoch": 15, "iter": 2362, "memory": 6269, "step": 2362}
+{"coco/bbox_mAP": 0.275, "coco/bbox_mAP_50": 0.745, "coco/bbox_mAP_75": 0.102, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.233, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018587308052258613, "time": 0.012624852677695772, "step": 15}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01761404037475586, "grad_norm": 4.326119842529297, "loss": 0.9812255465984344, "loss_cls": 0.08000751659274101, "loss_bbox": 0.29943086206912994, "loss_centerness": 0.5651274168491364, "loss_pos": 0.033120358306914566, "loss_offset": 0.003539396276464686, "time": 0.1372237253189087, "epoch": 16, "iter": 2420, "memory": 6269, "step": 2420}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01558643341064453, "grad_norm": 4.236466116905213, "loss": 0.9573487889766693, "loss_cls": 0.07560534067451954, "loss_bbox": 0.2844002604484558, "loss_centerness": 0.5619592118263245, "loss_pos": 0.03218893500044942, "loss_offset": 0.003195037094410509, "time": 0.13344262599945067, "epoch": 16, "iter": 2470, "memory": 6269, "step": 2470}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015976324081420898, "grad_norm": 4.18245527267456, "loss": 0.9447289884090424, "loss_cls": 0.07164965301752091, "loss_bbox": 0.2784549230337143, "loss_centerness": 0.5615593481063843, "loss_pos": 0.03012741846498102, "loss_offset": 0.0029376437258906664, "time": 0.13474231243133544, "epoch": 16, "iter": 2520, "memory": 6269, "step": 2520}
+{"coco/bbox_mAP": 0.276, "coco/bbox_mAP_50": 0.757, "coco/bbox_mAP_75": 0.109, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.302, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019612047407362196, "time": 0.013115301295223399, "step": 16}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01741969585418701, "grad_norm": 4.130473079681397, "loss": 0.9381736588478088, "loss_cls": 0.07050286196172237, "loss_bbox": 0.26982091248035434, "loss_centerness": 0.5628672671318055, "loss_pos": 0.031890174997970464, "loss_offset": 0.0030924458790104834, "time": 0.1356906032562256, "epoch": 17, "iter": 2578, "memory": 6269, "step": 2578}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015019917488098144, "grad_norm": 4.094593048095703, "loss": 0.944223929643631, "loss_cls": 0.07398053735494614, "loss_bbox": 0.27838924318552016, "loss_centerness": 0.5630177021026611, "loss_pos": 0.026048088371753694, "loss_offset": 0.002788354078656994, "time": 0.13370832443237304, "epoch": 17, "iter": 2628, "memory": 6269, "step": 2628}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01637794017791748, "grad_norm": 4.483822817802429, "loss": 0.9563675749301911, "loss_cls": 0.07527556575834751, "loss_bbox": 0.28489002108573913, "loss_centerness": 0.5595875239372253, "loss_pos": 0.03282955233938992, "loss_offset": 0.003784907319350168, "time": 0.1357308292388916, "epoch": 17, "iter": 2678, "memory": 6269, "step": 2678}
+{"coco/bbox_mAP": 0.276, "coco/bbox_mAP_50": 0.75, "coco/bbox_mAP_75": 0.115, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.333, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019330216269207816, "time": 0.013244418608836638, "step": 17}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.018779568672180176, "grad_norm": 4.255708332061768, "loss": 0.9483906233310699, "loss_cls": 0.07358938723802566, "loss_bbox": 0.2764634531736374, "loss_centerness": 0.5614309012889862, "loss_pos": 0.033469216162338854, "loss_offset": 0.0034376664576120674, "time": 0.13884978294372557, "epoch": 18, "iter": 2736, "memory": 6269, "step": 2736}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01540074348449707, "grad_norm": 5.14580801486969, "loss": 0.9381686961650848, "loss_cls": 0.06975050829350948, "loss_bbox": 0.2780903035402298, "loss_centerness": 0.5598645043373108, "loss_pos": 0.027314159641973673, "loss_offset": 0.003149225398665294, "time": 0.1336670732498169, "epoch": 18, "iter": 2786, "memory": 6269, "step": 2786}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015823354721069337, "grad_norm": 3.9810928440093996, "loss": 0.9237111282348632, "loss_cls": 0.06767715588212013, "loss_bbox": 0.266677650809288, "loss_centerness": 0.5617510569095612, "loss_pos": 0.02499782896018587, "loss_offset": 0.00260743027261924, "time": 0.1339794635772705, "epoch": 18, "iter": 2836, "memory": 6269, "step": 2836}
+{"coco/bbox_mAP": 0.276, "coco/bbox_mAP_50": 0.734, "coco/bbox_mAP_75": 0.107, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.256, "coco/bbox_mAP_l": -1.0, "data_time": 0.001917398485363039, "time": 0.013003039971376076, "step": 18}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.01727348804473877, "grad_norm": 4.144890036582947, "loss": 0.9193942558765411, "loss_cls": 0.06831951074302196, "loss_bbox": 0.2619258949160576, "loss_centerness": 0.560395987033844, "loss_pos": 0.02519456119509414, "loss_offset": 0.003558307944331318, "time": 0.13793036460876465, "epoch": 19, "iter": 2894, "memory": 6269, "step": 2894}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.015438718795776367, "grad_norm": 3.8387592744827272, "loss": 0.9199097931385041, "loss_cls": 0.06906423233449459, "loss_bbox": 0.2593508419394493, "loss_centerness": 0.5630311477184295, "loss_pos": 0.025483323871158065, "loss_offset": 0.002980243677739054, "time": 0.1348244094848633, "epoch": 19, "iter": 2944, "memory": 6269, "step": 2944}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.014845352172851562, "grad_norm": 3.838703947067261, "loss": 0.9138183081150055, "loss_cls": 0.06686785943806171, "loss_bbox": 0.2559739831089973, "loss_centerness": 0.5586423313617707, "loss_pos": 0.028648501732386648, "loss_offset": 0.0036856335669290274, "time": 0.13196611404418945, "epoch": 19, "iter": 2994, "memory": 6269, "step": 2994}
+{"coco/bbox_mAP": 0.284, "coco/bbox_mAP_50": 0.75, "coco/bbox_mAP_75": 0.121, "coco/bbox_mAP_s": 0.283, "coco/bbox_mAP_m": 0.282, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019409892905471672, "time": 0.013125775410578801, "step": 19}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.019097867012023925, "grad_norm": 3.9816743516921997, "loss": 0.9159980630874633, "loss_cls": 0.06774190671741963, "loss_bbox": 0.25821553051471713, "loss_centerness": 0.5594174206256867, "loss_pos": 0.027454992034472527, "loss_offset": 0.003168207489652559, "time": 0.13883081436157227, "epoch": 20, "iter": 3052, "memory": 6269, "step": 3052}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.016137003898620605, "grad_norm": 3.8565835833549498, "loss": 0.90747829079628, "loss_cls": 0.06646219410002231, "loss_bbox": 0.25374414771795273, "loss_centerness": 0.5608384728431701, "loss_pos": 0.023101599980145694, "loss_offset": 0.003331877125892788, "time": 0.13580811023712158, "epoch": 20, "iter": 3102, "memory": 6269, "step": 3102}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.016083950996398925, "grad_norm": 3.8600285196304323, "loss": 0.9185036134719848, "loss_cls": 0.06824196718633174, "loss_bbox": 0.25938189059495925, "loss_centerness": 0.5617206084728241, "loss_pos": 0.0259188624471426, "loss_offset": 0.0032402819058916068, "time": 0.13465497970581056, "epoch": 20, "iter": 3152, "memory": 6269, "step": 3152}
+{"coco/bbox_mAP": 0.283, "coco/bbox_mAP_50": 0.75, "coco/bbox_mAP_75": 0.118, "coco/bbox_mAP_s": 0.283, "coco/bbox_mAP_m": 0.294, "coco/bbox_mAP_l": -1.0, "data_time": 0.001965687417576456, "time": 0.013327921761406793, "step": 20}

haar_v4_ungated_768/20260727_162937/vis_data/config.py

@@ -0,0 +1,295 @@
+auto_scale_lr = dict(base_batch_size=16, enable=True)
+backend_args = None
+custom_imports = dict(imports=[
+    'projects.pahr',
+])
+data_root = 'data/coco/'
+dataset_type = 'CocoDataset'
+default_hooks = dict(
+    checkpoint=dict(
+        interval=1,
+        max_keep_ckpts=1,
+        rule='greater',
+        save_best='coco/bbox_mAP_75',
+        save_last=True,
+        type='CheckpointHook'),
+    logger=dict(interval=50, type='LoggerHook'),
+    param_scheduler=dict(type='ParamSchedulerHook'),
+    sampler_seed=dict(type='DistSamplerSeedHook'),
+    timer=dict(type='IterTimerHook'),
+    visualization=dict(type='DetVisualizationHook'))
+default_scope = 'mmdet'
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+launcher = 'none'
+load_from = None
+log_level = 'INFO'
+log_processor = dict(by_epoch=True, type='LogProcessor', window_size=50)
+model = dict(
+    backbone=dict(
+        depth=50,
+        frozen_stages=1,
+        init_cfg=dict(
+            checkpoint='open-mmlab://detectron/resnet50_caffe',
+            type='Pretrained'),
+        norm_cfg=dict(requires_grad=False, type='BN'),
+        norm_eval=True,
+        num_stages=4,
+        out_indices=(
+            0,
+            1,
+            2,
+            3,
+        ),
+        style='caffe',
+        type='ResNet'),
+    bbox_head=dict(
+        feat_channels=256,
+        in_channels=256,
+        loss_bbox=dict(loss_weight=1.0, type='IoULoss'),
+        loss_centerness=dict(
+            loss_weight=1.0, type='CrossEntropyLoss', use_sigmoid=True),
+        loss_cls=dict(
+            alpha=0.25,
+            gamma=2.0,
+            loss_weight=1.0,
+            type='FocalLoss',
+            use_sigmoid=True),
+        num_classes=1,
+        stacked_convs=4,
+        strides=[
+            8,
+            16,
+            32,
+            64,
+            128,
+        ],
+        type='FCOSHead'),
+    data_preprocessor=dict(
+        bgr_to_rgb=False,
+        mean=[
+            102.9801,
+            115.9465,
+            122.7717,
+        ],
+        pad_size_divisor=32,
+        std=[
+            1.0,
+            1.0,
+            1.0,
+        ],
+        type='DetDataPreprocessor'),
+    loss_offset_weight=0.1,
+    loss_pos_weight=0.1,
+    neck=dict(
+        add_extra_convs='on_output',
+        correction_gain=1.0,
+        correction_gate_floor=0.0,
+        detach_position_gate=False,
+        detail_channels=64,
+        gate_power=1.0,
+        guide_channels=0,
+        in_channels=[
+            256,
+            512,
+            1024,
+            2048,
+        ],
+        locator_channels=64,
+        num_outs=5,
+        out_channels=256,
+        relu_before_extra_convs=True,
+        start_level=1,
+        type='PAHRFPN',
+        use_output_gate=False),
+    position_stride=8,
+    test_cfg=dict(
+        max_per_img=100,
+        min_bbox_size=0,
+        nms=dict(iou_threshold=0.5, type='nms'),
+        nms_pre=1000,
+        score_thr=0.05),
+    tiny_max_sqrt_area=16.0,
+    type='PAHRFCOS',
+    use_phase_shift=True)
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=35, norm_type=2),
+    optimizer=dict(lr=0.01, momentum=0.9, type='SGD', weight_decay=0.0001),
+    paramwise_cfg=dict(bias_decay_mult=0.0, bias_lr_mult=2.0),
+    type='OptimWrapper')
+pahr_variant = dict(
+    correction_gain=1.0,
+    detail_lr_mult=1.0,
+    giou=False,
+    guide_channels=0,
+    image_size=768,
+    name='haar_v4_ungated_768',
+    phase_shift=True,
+    scaled_schedule=True,
+    use_output_gate=False,
+    v3_gate=False)
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=False,
+        end=500,
+        factor=0.3333333333333333,
+        type='ConstantLR'),
+    dict(
+        begin=0,
+        by_epoch=True,
+        end=20,
+        gamma=0.1,
+        milestones=[
+            13,
+            18,
+        ],
+        type='MultiStepLR'),
+]
+randomness = dict(seed=42)
+resume = False
+test_cfg = dict(type='TestLoop')
+test_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        ann_file=
+        '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/test.json',
+        backend_args=None,
+        data_prefix=dict(img='/marimo/LevirShipData/All Images/'),
+        data_root='',
+        metainfo=dict(classes=('ship', )),
+        pipeline=[
+            dict(backend_args=None, type='LoadImageFromFile'),
+            dict(keep_ratio=True, scale=(
+                768,
+                768,
+            ), type='Resize'),
+            dict(type='LoadAnnotations', with_bbox=True),
+            dict(
+                meta_keys=(
+                    'img_id',
+                    'img_path',
+                    'ori_shape',
+                    'img_shape',
+                    'scale_factor',
+                ),
+                type='PackDetInputs'),
+        ],
+        test_mode=True,
+        type='CocoDataset'),
+    drop_last=False,
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(shuffle=False, type='DefaultSampler'))
+test_evaluator = dict(
+    ann_file=
+    '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/test.json',
+    backend_args=None,
+    format_only=False,
+    metric='bbox',
+    type='CocoMetric')
+test_pipeline = [
+    dict(backend_args=None, type='LoadImageFromFile'),
+    dict(keep_ratio=True, scale=(
+        1333,
+        800,
+    ), type='Resize'),
+    dict(type='LoadAnnotations', with_bbox=True),
+    dict(
+        meta_keys=(
+            'img_id',
+            'img_path',
+            'ori_shape',
+            'img_shape',
+            'scale_factor',
+        ),
+        type='PackDetInputs'),
+]
+train_cfg = dict(max_epochs=20, type='EpochBasedTrainLoop', val_interval=1)
+train_dataloader = dict(
+    batch_sampler=dict(type='AspectRatioBatchSampler'),
+    batch_size=8,
+    dataset=dict(
+        ann_file=
+        '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/train.json',
+        backend_args=None,
+        data_prefix=dict(img='/marimo/LevirShipData/All Images/'),
+        data_root='',
+        filter_cfg=dict(filter_empty_gt=True, min_size=32),
+        metainfo=dict(classes=('ship', )),
+        pipeline=[
+            dict(backend_args=None, type='LoadImageFromFile'),
+            dict(type='LoadAnnotations', with_bbox=True),
+            dict(keep_ratio=True, scale=(
+                768,
+                768,
+            ), type='Resize'),
+            dict(prob=0.5, type='RandomFlip'),
+            dict(type='PackDetInputs'),
+        ],
+        type='CocoDataset'),
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(shuffle=True, type='DefaultSampler'))
+train_pipeline = [
+    dict(backend_args=None, type='LoadImageFromFile'),
+    dict(type='LoadAnnotations', with_bbox=True),
+    dict(keep_ratio=True, scale=(
+        1333,
+        800,
+    ), type='Resize'),
+    dict(prob=0.5, type='RandomFlip'),
+    dict(type='PackDetInputs'),
+]
+val_cfg = dict(type='ValLoop')
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        ann_file=
+        '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/val.json',
+        backend_args=None,
+        data_prefix=dict(img='/marimo/LevirShipData/All Images/'),
+        data_root='',
+        metainfo=dict(classes=('ship', )),
+        pipeline=[
+            dict(backend_args=None, type='LoadImageFromFile'),
+            dict(keep_ratio=True, scale=(
+                768,
+                768,
+            ), type='Resize'),
+            dict(type='LoadAnnotations', with_bbox=True),
+            dict(
+                meta_keys=(
+                    'img_id',
+                    'img_path',
+                    'ori_shape',
+                    'img_shape',
+                    'scale_factor',
+                ),
+                type='PackDetInputs'),
+        ],
+        test_mode=True,
+        type='CocoDataset'),
+    drop_last=False,
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(shuffle=False, type='DefaultSampler'))
+val_evaluator = dict(
+    ann_file=
+    '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/val.json',
+    backend_args=None,
+    format_only=False,
+    metric='bbox',
+    type='CocoMetric')
+vis_backends = [
+    dict(type='LocalVisBackend'),
+]
+visualizer = dict(
+    name='visualizer',
+    type='DetLocalVisualizer',
+    vis_backends=[
+        dict(type='LocalVisBackend'),
+    ])
+work_dir = '/marimo/mmdet_code/mmdetection/work_dirs/levir_pahr_v4_screen/haar_v4_ungated_768'

haar_v4_ungated_768/20260727_162937/vis_data/scalars.json

@@ -0,0 +1,80 @@
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.02387575626373291, "grad_norm": 114.52095680236816, "loss": 3.641382517814636, "loss_cls": 0.8087327021360398, "loss_bbox": 1.3330383479595185, "loss_centerness": 0.6507055377960205, "loss_pos": 0.8346495192125439, "loss_offset": 0.014256374315591529, "time": 0.1746406936645508, "epoch": 1, "iter": 50, "memory": 6269, "step": 50}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.016263041496276855, "grad_norm": 16.76171932220459, "loss": 1.9775516772270203, "loss_cls": 0.32804525196552276, "loss_bbox": 0.7804536592960357, "loss_centerness": 0.627768611907959, "loss_pos": 0.23205391455441712, "loss_offset": 0.009230252052657307, "time": 0.13671158790588378, "epoch": 1, "iter": 100, "memory": 6269, "step": 100}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015863189697265623, "grad_norm": 12.959484844207763, "loss": 1.8052734732627869, "loss_cls": 0.25144625425338746, "loss_bbox": 0.7720166838169098, "loss_centerness": 0.6131773805618286, "loss_pos": 0.16167937211692332, "loss_offset": 0.006953791005071252, "time": 0.13558685302734375, "epoch": 1, "iter": 150, "memory": 6269, "step": 150}
+{"coco/bbox_mAP": 0.112, "coco/bbox_mAP_50": 0.47, "coco/bbox_mAP_75": 0.003, "coco/bbox_mAP_s": 0.113, "coco/bbox_mAP_m": 0.112, "coco/bbox_mAP_l": -1.0, "data_time": 0.0027687876191857743, "time": 0.014944686056816415, "step": 1}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.017955560684204102, "grad_norm": 10.754822750091552, "loss": 1.522691957950592, "loss_cls": 0.20511740639805795, "loss_bbox": 0.5785106360912323, "loss_centerness": 0.601420282125473, "loss_pos": 0.12888413108885288, "loss_offset": 0.008759502851171419, "time": 0.1390164613723755, "epoch": 2, "iter": 208, "memory": 6269, "step": 208}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015726680755615233, "grad_norm": 9.802234706878663, "loss": 1.4951458525657655, "loss_cls": 0.19281451269984246, "loss_bbox": 0.5602053409814834, "loss_centerness": 0.5924474763870239, "loss_pos": 0.14257961828261614, "loss_offset": 0.007098915867973119, "time": 0.13476008415222168, "epoch": 2, "iter": 258, "memory": 6269, "step": 258}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015753674507141113, "grad_norm": 10.286860389709473, "loss": 1.4608097052574158, "loss_cls": 0.17279406234622002, "loss_bbox": 0.5809448206424713, "loss_centerness": 0.5883310759067535, "loss_pos": 0.11235435826703906, "loss_offset": 0.006385399708524347, "time": 0.13440067291259766, "epoch": 2, "iter": 308, "memory": 6269, "step": 308}
+{"coco/bbox_mAP": 0.21, "coco/bbox_mAP_50": 0.674, "coco/bbox_mAP_75": 0.037, "coco/bbox_mAP_s": 0.213, "coco/bbox_mAP_m": 0.159, "coco/bbox_mAP_l": -1.0, "data_time": 0.0020721525208562868, "time": 0.013429028763730302, "step": 2}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.01754068851470947, "grad_norm": 9.125817556381225, "loss": 1.365412051677704, "loss_cls": 0.15463197439908982, "loss_bbox": 0.5052433621883392, "loss_centerness": 0.5871367955207825, "loss_pos": 0.11145446598529815, "loss_offset": 0.006945455521345139, "time": 0.13633635520935058, "epoch": 3, "iter": 366, "memory": 6269, "step": 366}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.016068544387817383, "grad_norm": 9.999114599227905, "loss": 1.393318359851837, "loss_cls": 0.1690808902680874, "loss_bbox": 0.5191915154457092, "loss_centerness": 0.5846550559997559, "loss_pos": 0.1127777398005128, "loss_offset": 0.007613137498265132, "time": 0.13512993335723877, "epoch": 3, "iter": 416, "memory": 6269, "step": 416}
+{"base_lr": 0.003333333333333333, "lr": 0.0016666666666666666, "data_time": 0.015846266746520996, "grad_norm": 9.580743188858031, "loss": 1.3606600379943847, "loss_cls": 0.14454242944717408, "loss_bbox": 0.5217189937829971, "loss_centerness": 0.581095484495163, "loss_pos": 0.10637437952682376, "loss_offset": 0.006928750155493617, "time": 0.13455546855926515, "epoch": 3, "iter": 466, "memory": 6269, "step": 466}
+{"coco/bbox_mAP": 0.095, "coco/bbox_mAP_50": 0.403, "coco/bbox_mAP_75": 0.015, "coco/bbox_mAP_s": 0.091, "coco/bbox_mAP_m": 0.262, "coco/bbox_mAP_l": -1.0, "data_time": 0.001969819598727756, "time": 0.013254984423645541, "step": 3}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.017896056175231934, "grad_norm": 11.886852989196777, "loss": 1.5761951184272767, "loss_cls": 0.1883251576125622, "loss_bbox": 0.6712099689245224, "loss_centerness": 0.5823398959636689, "loss_pos": 0.12729708219412714, "loss_offset": 0.007023014086298645, "time": 0.13741453647613525, "epoch": 4, "iter": 524, "memory": 6269, "step": 524}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01574042797088623, "grad_norm": 11.804976673126221, "loss": 1.8784850573539733, "loss_cls": 0.3001200720667839, "loss_bbox": 0.8040754497051239, "loss_centerness": 0.6108310925960541, "loss_pos": 0.155565629824996, "loss_offset": 0.007892816206440329, "time": 0.13491064548492432, "epoch": 4, "iter": 574, "memory": 6269, "step": 574}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015125856399536133, "grad_norm": 8.498672256469726, "loss": 1.475808184146881, "loss_cls": 0.19704236343502998, "loss_bbox": 0.5570360046625137, "loss_centerness": 0.5899317514896393, "loss_pos": 0.12457641873508692, "loss_offset": 0.0072216421773191545, "time": 0.13200510025024415, "epoch": 4, "iter": 624, "memory": 6269, "step": 624}
+{"coco/bbox_mAP": 0.074, "coco/bbox_mAP_50": 0.327, "coco/bbox_mAP_75": 0.006, "coco/bbox_mAP_s": 0.075, "coco/bbox_mAP_m": 0.012, "coco/bbox_mAP_l": -1.0, "data_time": 0.0020367088480892346, "time": 0.013232316318740193, "step": 4}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.017707514762878417, "grad_norm": 10.190532112121582, "loss": 1.5970463275909423, "loss_cls": 0.2575735272467136, "loss_bbox": 0.6156666225194931, "loss_centerness": 0.5960467672348022, "loss_pos": 0.12227158192545176, "loss_offset": 0.005487838343251497, "time": 0.13671816349029542, "epoch": 5, "iter": 682, "memory": 6269, "step": 682}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016632404327392578, "grad_norm": 7.975882334709167, "loss": 1.4121679091453552, "loss_cls": 0.16413942188024522, "loss_bbox": 0.5541865700483322, "loss_centerness": 0.580131139755249, "loss_pos": 0.10705584756098688, "loss_offset": 0.006654911576770246, "time": 0.1359233331680298, "epoch": 5, "iter": 732, "memory": 6269, "step": 732}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015917253494262696, "grad_norm": 8.996517276763916, "loss": 1.4886412334442138, "loss_cls": 0.15306309029459952, "loss_bbox": 0.634658961892128, "loss_centerness": 0.5835314714908599, "loss_pos": 0.11071547254920006, "loss_offset": 0.006672249686671421, "time": 0.13437987804412843, "epoch": 5, "iter": 782, "memory": 6269, "step": 782}
+{"coco/bbox_mAP": 0.161, "coco/bbox_mAP_50": 0.613, "coco/bbox_mAP_75": 0.016, "coco/bbox_mAP_s": 0.161, "coco/bbox_mAP_m": 0.18, "coco/bbox_mAP_l": -1.0, "data_time": 0.001958650197738256, "time": 0.013422944810655382, "step": 5}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018230013847351074, "grad_norm": 7.757193851470947, "loss": 1.3803178119659423, "loss_cls": 0.1524515801668167, "loss_bbox": 0.5319341343641281, "loss_centerness": 0.581254894733429, "loss_pos": 0.10909361463040114, "loss_offset": 0.005583584911073558, "time": 0.13743532657623292, "epoch": 6, "iter": 840, "memory": 6269, "step": 840}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015194315910339356, "grad_norm": 7.847311778068542, "loss": 1.432754168510437, "loss_cls": 0.15826014041900635, "loss_bbox": 0.5617603421211242, "loss_centerness": 0.5805632305145264, "loss_pos": 0.1260861048847437, "loss_offset": 0.0060843450643005784, "time": 0.13362086296081543, "epoch": 6, "iter": 890, "memory": 6269, "step": 890}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015460410118103028, "grad_norm": 8.59046314239502, "loss": 1.3929545879364014, "loss_cls": 0.1539978276193142, "loss_bbox": 0.5722000098228455, "loss_centerness": 0.5774772417545319, "loss_pos": 0.08417099100537598, "loss_offset": 0.005108523470116779, "time": 0.1328950357437134, "epoch": 6, "iter": 940, "memory": 6269, "step": 940}
+{"coco/bbox_mAP": 0.181, "coco/bbox_mAP_50": 0.608, "coco/bbox_mAP_75": 0.032, "coco/bbox_mAP_s": 0.178, "coco/bbox_mAP_m": 0.253, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019063554258428067, "time": 0.012651062419271877, "step": 6}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.017206239700317382, "grad_norm": 6.785709195137024, "loss": 1.2810506629943847, "loss_cls": 0.12734990701079368, "loss_bbox": 0.4799131882190704, "loss_centerness": 0.575744092464447, "loss_pos": 0.09377483788877726, "loss_offset": 0.004268641428789124, "time": 0.13568567752838134, "epoch": 7, "iter": 998, "memory": 6269, "step": 998}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01695356845855713, "grad_norm": 7.346939630508423, "loss": 1.3278966546058655, "loss_cls": 0.13488397777080535, "loss_bbox": 0.5170796245336533, "loss_centerness": 0.5783039677143097, "loss_pos": 0.09240393787622452, "loss_offset": 0.005225151426857338, "time": 0.13568556785583497, "epoch": 7, "iter": 1048, "memory": 6269, "step": 1048}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015588855743408203, "grad_norm": 8.084754962921142, "loss": 1.3997604441642761, "loss_cls": 0.12984258115291594, "loss_bbox": 0.6016161435842514, "loss_centerness": 0.5777574527263641, "loss_pos": 0.08483651880174875, "loss_offset": 0.0057077554706484075, "time": 0.1334227752685547, "epoch": 7, "iter": 1098, "memory": 6269, "step": 1098}
+{"coco/bbox_mAP": 0.24, "coco/bbox_mAP_50": 0.698, "coco/bbox_mAP_75": 0.055, "coco/bbox_mAP_s": 0.241, "coco/bbox_mAP_m": 0.17, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019307633750459067, "time": 0.012958384375286918, "step": 7}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018718342781066894, "grad_norm": 7.671614689826965, "loss": 1.302541148662567, "loss_cls": 0.13622799679636954, "loss_bbox": 0.5055729216337204, "loss_centerness": 0.5755344927310944, "loss_pos": 0.07973715897649526, "loss_offset": 0.0054685806878842416, "time": 0.1391748857498169, "epoch": 8, "iter": 1156, "memory": 6269, "step": 1156}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015604238510131836, "grad_norm": 7.33595838546753, "loss": 1.3371935772895813, "loss_cls": 0.13143601655960083, "loss_bbox": 0.5303419882059097, "loss_centerness": 0.5752888548374177, "loss_pos": 0.0946077877189964, "loss_offset": 0.005518921547336504, "time": 0.13442244052886962, "epoch": 8, "iter": 1206, "memory": 6269, "step": 1206}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015556530952453613, "grad_norm": 6.322084312438965, "loss": 1.2781175518035888, "loss_cls": 0.12771683260798455, "loss_bbox": 0.4876609718799591, "loss_centerness": 0.5766118216514587, "loss_pos": 0.08130903821438551, "loss_offset": 0.004818873853655532, "time": 0.1342010259628296, "epoch": 8, "iter": 1256, "memory": 6269, "step": 1256}
+{"coco/bbox_mAP": 0.044, "coco/bbox_mAP_50": 0.199, "coco/bbox_mAP_75": 0.002, "coco/bbox_mAP_s": 0.045, "coco/bbox_mAP_m": 0.183, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018745055565467248, "time": 0.012709898826403495, "step": 8}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016962690353393553, "grad_norm": 7.528588318824768, "loss": 1.3156990897655487, "loss_cls": 0.13987344846129418, "loss_bbox": 0.5230728405714035, "loss_centerness": 0.5738614439964295, "loss_pos": 0.0742860192619264, "loss_offset": 0.004605343978619203, "time": 0.13479403018951416, "epoch": 9, "iter": 1314, "memory": 6269, "step": 1314}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015400733947753906, "grad_norm": 6.583463411331177, "loss": 1.2535515475273131, "loss_cls": 0.1289924655854702, "loss_bbox": 0.4492087346315384, "loss_centerness": 0.5746508133411408, "loss_pos": 0.09588879181072116, "loss_offset": 0.00481074296112638, "time": 0.133740873336792, "epoch": 9, "iter": 1364, "memory": 6269, "step": 1364}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016284708976745606, "grad_norm": 6.643251657485962, "loss": 1.2628607630729676, "loss_cls": 0.12052003189921379, "loss_bbox": 0.47820280969142914, "loss_centerness": 0.5739504361152649, "loss_pos": 0.08566969852894545, "loss_offset": 0.004517789883539081, "time": 0.13424815177917482, "epoch": 9, "iter": 1414, "memory": 6269, "step": 1414}
+{"coco/bbox_mAP": 0.264, "coco/bbox_mAP_50": 0.746, "coco/bbox_mAP_75": 0.091, "coco/bbox_mAP_s": 0.267, "coco/bbox_mAP_m": 0.145, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018993536631266275, "time": 0.01278637820838863, "step": 9}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018471441268920898, "grad_norm": 6.325474667549133, "loss": 1.2168771278858186, "loss_cls": 0.12575486034154892, "loss_bbox": 0.44242464542388915, "loss_centerness": 0.5722889256477356, "loss_pos": 0.07150071470066904, "loss_offset": 0.004907989402418025, "time": 0.1380993604660034, "epoch": 10, "iter": 1472, "memory": 6269, "step": 1472}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.016656055450439453, "grad_norm": 7.229538035392761, "loss": 1.2897097182273864, "loss_cls": 0.12006522357463836, "loss_bbox": 0.5167982870340347, "loss_centerness": 0.5739198541641235, "loss_pos": 0.07485429923981428, "loss_offset": 0.004072059815516695, "time": 0.13608648777008056, "epoch": 10, "iter": 1522, "memory": 6269, "step": 1522}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01547323226928711, "grad_norm": 6.1348348808288575, "loss": 1.2192114770412446, "loss_cls": 0.11673817545175552, "loss_bbox": 0.44963080644607545, "loss_centerness": 0.5735272967815399, "loss_pos": 0.07516739720478655, "loss_offset": 0.004147799819184002, "time": 0.13438985347747803, "epoch": 10, "iter": 1572, "memory": 6269, "step": 1572}
+{"coco/bbox_mAP": 0.21, "coco/bbox_mAP_50": 0.651, "coco/bbox_mAP_75": 0.057, "coco/bbox_mAP_s": 0.214, "coco/bbox_mAP_m": 0.063, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019964625692775106, "time": 0.013408109061738364, "step": 10}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01780747890472412, "grad_norm": 6.169156465530396, "loss": 1.2160494685173036, "loss_cls": 0.11702308133244514, "loss_bbox": 0.45238163590431213, "loss_centerness": 0.5727790057659149, "loss_pos": 0.06946244772523641, "loss_offset": 0.004403298212491791, "time": 0.13704002380371094, "epoch": 11, "iter": 1630, "memory": 6269, "step": 1630}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015589737892150879, "grad_norm": 5.870634632110596, "loss": 1.159477950334549, "loss_cls": 0.10759207174181938, "loss_bbox": 0.41125225603580473, "loss_centerness": 0.5707440173625946, "loss_pos": 0.06594253187067807, "loss_offset": 0.003947081957012415, "time": 0.13382294178009033, "epoch": 11, "iter": 1680, "memory": 6269, "step": 1680}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015575742721557618, "grad_norm": 7.311523427963257, "loss": 1.2527956891059875, "loss_cls": 0.11181208297610283, "loss_bbox": 0.49854174852371214, "loss_centerness": 0.5704759740829468, "loss_pos": 0.06851734125986696, "loss_offset": 0.0034485349094029514, "time": 0.1344086742401123, "epoch": 11, "iter": 1730, "memory": 6269, "step": 1730}
+{"coco/bbox_mAP": 0.218, "coco/bbox_mAP_50": 0.655, "coco/bbox_mAP_75": 0.068, "coco/bbox_mAP_s": 0.218, "coco/bbox_mAP_m": 0.203, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019610025943853915, "time": 0.01311383777194553, "step": 11}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.0177852725982666, "grad_norm": 6.048166790008545, "loss": 1.199417164325714, "loss_cls": 0.1181537875533104, "loss_bbox": 0.44670230507850645, "loss_centerness": 0.5699885952472686, "loss_pos": 0.0609135265275836, "loss_offset": 0.0036589622881729156, "time": 0.13667559146881103, "epoch": 12, "iter": 1788, "memory": 6269, "step": 1788}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015840106010437012, "grad_norm": 7.208025574684143, "loss": 1.2247644114494323, "loss_cls": 0.11180497005581856, "loss_bbox": 0.47382911264896393, "loss_centerness": 0.5729782819747925, "loss_pos": 0.062134848015848544, "loss_offset": 0.004017203194380272, "time": 0.13444390773773193, "epoch": 12, "iter": 1838, "memory": 6269, "step": 1838}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.018442683219909668, "grad_norm": 5.660691437721252, "loss": 1.1556305146217347, "loss_cls": 0.10315977282822132, "loss_bbox": 0.4162970733642578, "loss_centerness": 0.5688820195198059, "loss_pos": 0.0637389025464654, "loss_offset": 0.003552732290845597, "time": 0.13917843818664552, "epoch": 12, "iter": 1888, "memory": 6269, "step": 1888}
+{"coco/bbox_mAP": 0.157, "coco/bbox_mAP_50": 0.571, "coco/bbox_mAP_75": 0.02, "coco/bbox_mAP_s": 0.16, "coco/bbox_mAP_m": 0.002, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019042418553278997, "time": 0.012593824843056182, "step": 12}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.01787266731262207, "grad_norm": 6.075709481239318, "loss": 1.1512815499305724, "loss_cls": 0.09467467859387398, "loss_bbox": 0.4316812252998352, "loss_centerness": 0.5683929216861725, "loss_pos": 0.05296651531010866, "loss_offset": 0.003566218525520526, "time": 0.1355909776687622, "epoch": 13, "iter": 1946, "memory": 6269, "step": 1946}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015505547523498536, "grad_norm": 5.894303050041199, "loss": 1.142732938528061, "loss_cls": 0.10125300385057927, "loss_bbox": 0.4011682283878326, "loss_centerness": 0.5691057133674622, "loss_pos": 0.06723949652630835, "loss_offset": 0.0039665168119245205, "time": 0.13348616123199464, "epoch": 13, "iter": 1996, "memory": 6269, "step": 1996}
+{"base_lr": 0.009999999999999998, "lr": 0.004999999999999999, "data_time": 0.015311622619628906, "grad_norm": 5.925397968292236, "loss": 1.1735150337219238, "loss_cls": 0.1086564926803112, "loss_bbox": 0.43431043744087217, "loss_centerness": 0.570909628868103, "loss_pos": 0.056253279503434894, "loss_offset": 0.003385203641373664, "time": 0.13318418979644775, "epoch": 13, "iter": 2046, "memory": 6269, "step": 2046}
+{"coco/bbox_mAP": 0.193, "coco/bbox_mAP_50": 0.632, "coco/bbox_mAP_75": 0.035, "coco/bbox_mAP_s": 0.197, "coco/bbox_mAP_m": 0.098, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018503535507071731, "time": 0.012584523258046208, "step": 13}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.017718935012817384, "grad_norm": 4.6375644540786745, "loss": 1.0503401231765748, "loss_cls": 0.09265139251947403, "loss_bbox": 0.3370001268386841, "loss_centerness": 0.5646761882305146, "loss_pos": 0.0524382827244699, "loss_offset": 0.003574130265333224, "time": 0.13637757778167725, "epoch": 14, "iter": 2104, "memory": 6269, "step": 2104}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015170278549194336, "grad_norm": 3.8037463188171388, "loss": 0.986722184419632, "loss_cls": 0.0801999495178461, "loss_bbox": 0.30282890737056734, "loss_centerness": 0.5650018358230591, "loss_pos": 0.03569386578630656, "loss_offset": 0.0029976315080421046, "time": 0.13306628704071044, "epoch": 14, "iter": 2154, "memory": 6269, "step": 2154}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01542081356048584, "grad_norm": 4.260610308647156, "loss": 0.9856020987033844, "loss_cls": 0.07983840823173523, "loss_bbox": 0.2959664106369019, "loss_centerness": 0.565356639623642, "loss_pos": 0.041319113122299314, "loss_offset": 0.003121536517282948, "time": 0.13338305473327636, "epoch": 14, "iter": 2204, "memory": 6269, "step": 2204}
+{"coco/bbox_mAP": 0.291, "coco/bbox_mAP_50": 0.774, "coco/bbox_mAP_75": 0.127, "coco/bbox_mAP_s": 0.291, "coco/bbox_mAP_m": 0.26, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018406700884175096, "time": 0.012511734269623064, "step": 14}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01903491497039795, "grad_norm": 4.047395715713501, "loss": 0.9670887410640716, "loss_cls": 0.07777763083577156, "loss_bbox": 0.2950610908865929, "loss_centerness": 0.5628402543067932, "loss_pos": 0.028107117956969886, "loss_offset": 0.003302650565165095, "time": 0.1387276029586792, "epoch": 15, "iter": 2262, "memory": 6269, "step": 2262}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.016506352424621583, "grad_norm": 4.267009086608887, "loss": 0.9722354638576508, "loss_cls": 0.0777778397500515, "loss_bbox": 0.29410224497318266, "loss_centerness": 0.5635494661331176, "loss_pos": 0.03424645137274638, "loss_offset": 0.002559460891716299, "time": 0.1359393835067749, "epoch": 15, "iter": 2312, "memory": 6269, "step": 2312}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015163192749023438, "grad_norm": 4.381349997520447, "loss": 0.9859471487998962, "loss_cls": 0.08225745782256126, "loss_bbox": 0.29726907461881635, "loss_centerness": 0.5630013751983642, "loss_pos": 0.03976258021779358, "loss_offset": 0.0036566603364190085, "time": 0.1327739477157593, "epoch": 15, "iter": 2362, "memory": 6269, "step": 2362}
+{"coco/bbox_mAP": 0.275, "coco/bbox_mAP_50": 0.745, "coco/bbox_mAP_75": 0.102, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.233, "coco/bbox_mAP_l": -1.0, "data_time": 0.0018587308052258613, "time": 0.012624852677695772, "step": 15}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01761404037475586, "grad_norm": 4.326119842529297, "loss": 0.9812255465984344, "loss_cls": 0.08000751659274101, "loss_bbox": 0.29943086206912994, "loss_centerness": 0.5651274168491364, "loss_pos": 0.033120358306914566, "loss_offset": 0.003539396276464686, "time": 0.1372237253189087, "epoch": 16, "iter": 2420, "memory": 6269, "step": 2420}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01558643341064453, "grad_norm": 4.236466116905213, "loss": 0.9573487889766693, "loss_cls": 0.07560534067451954, "loss_bbox": 0.2844002604484558, "loss_centerness": 0.5619592118263245, "loss_pos": 0.03218893500044942, "loss_offset": 0.003195037094410509, "time": 0.13344262599945067, "epoch": 16, "iter": 2470, "memory": 6269, "step": 2470}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015976324081420898, "grad_norm": 4.18245527267456, "loss": 0.9447289884090424, "loss_cls": 0.07164965301752091, "loss_bbox": 0.2784549230337143, "loss_centerness": 0.5615593481063843, "loss_pos": 0.03012741846498102, "loss_offset": 0.0029376437258906664, "time": 0.13474231243133544, "epoch": 16, "iter": 2520, "memory": 6269, "step": 2520}
+{"coco/bbox_mAP": 0.276, "coco/bbox_mAP_50": 0.757, "coco/bbox_mAP_75": 0.109, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.302, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019612047407362196, "time": 0.013115301295223399, "step": 16}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01741969585418701, "grad_norm": 4.130473079681397, "loss": 0.9381736588478088, "loss_cls": 0.07050286196172237, "loss_bbox": 0.26982091248035434, "loss_centerness": 0.5628672671318055, "loss_pos": 0.031890174997970464, "loss_offset": 0.0030924458790104834, "time": 0.1356906032562256, "epoch": 17, "iter": 2578, "memory": 6269, "step": 2578}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015019917488098144, "grad_norm": 4.094593048095703, "loss": 0.944223929643631, "loss_cls": 0.07398053735494614, "loss_bbox": 0.27838924318552016, "loss_centerness": 0.5630177021026611, "loss_pos": 0.026048088371753694, "loss_offset": 0.002788354078656994, "time": 0.13370832443237304, "epoch": 17, "iter": 2628, "memory": 6269, "step": 2628}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01637794017791748, "grad_norm": 4.483822817802429, "loss": 0.9563675749301911, "loss_cls": 0.07527556575834751, "loss_bbox": 0.28489002108573913, "loss_centerness": 0.5595875239372253, "loss_pos": 0.03282955233938992, "loss_offset": 0.003784907319350168, "time": 0.1357308292388916, "epoch": 17, "iter": 2678, "memory": 6269, "step": 2678}
+{"coco/bbox_mAP": 0.276, "coco/bbox_mAP_50": 0.75, "coco/bbox_mAP_75": 0.115, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.333, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019330216269207816, "time": 0.013244418608836638, "step": 17}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.018779568672180176, "grad_norm": 4.255708332061768, "loss": 0.9483906233310699, "loss_cls": 0.07358938723802566, "loss_bbox": 0.2764634531736374, "loss_centerness": 0.5614309012889862, "loss_pos": 0.033469216162338854, "loss_offset": 0.0034376664576120674, "time": 0.13884978294372557, "epoch": 18, "iter": 2736, "memory": 6269, "step": 2736}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.01540074348449707, "grad_norm": 5.14580801486969, "loss": 0.9381686961650848, "loss_cls": 0.06975050829350948, "loss_bbox": 0.2780903035402298, "loss_centerness": 0.5598645043373108, "loss_pos": 0.027314159641973673, "loss_offset": 0.003149225398665294, "time": 0.1336670732498169, "epoch": 18, "iter": 2786, "memory": 6269, "step": 2786}
+{"base_lr": 0.0009999999999999998, "lr": 0.0004999999999999999, "data_time": 0.015823354721069337, "grad_norm": 3.9810928440093996, "loss": 0.9237111282348632, "loss_cls": 0.06767715588212013, "loss_bbox": 0.266677650809288, "loss_centerness": 0.5617510569095612, "loss_pos": 0.02499782896018587, "loss_offset": 0.00260743027261924, "time": 0.1339794635772705, "epoch": 18, "iter": 2836, "memory": 6269, "step": 2836}
+{"coco/bbox_mAP": 0.276, "coco/bbox_mAP_50": 0.734, "coco/bbox_mAP_75": 0.107, "coco/bbox_mAP_s": 0.275, "coco/bbox_mAP_m": 0.256, "coco/bbox_mAP_l": -1.0, "data_time": 0.001917398485363039, "time": 0.013003039971376076, "step": 18}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.01727348804473877, "grad_norm": 4.144890036582947, "loss": 0.9193942558765411, "loss_cls": 0.06831951074302196, "loss_bbox": 0.2619258949160576, "loss_centerness": 0.560395987033844, "loss_pos": 0.02519456119509414, "loss_offset": 0.003558307944331318, "time": 0.13793036460876465, "epoch": 19, "iter": 2894, "memory": 6269, "step": 2894}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.015438718795776367, "grad_norm": 3.8387592744827272, "loss": 0.9199097931385041, "loss_cls": 0.06906423233449459, "loss_bbox": 0.2593508419394493, "loss_centerness": 0.5630311477184295, "loss_pos": 0.025483323871158065, "loss_offset": 0.002980243677739054, "time": 0.1348244094848633, "epoch": 19, "iter": 2944, "memory": 6269, "step": 2944}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.014845352172851562, "grad_norm": 3.838703947067261, "loss": 0.9138183081150055, "loss_cls": 0.06686785943806171, "loss_bbox": 0.2559739831089973, "loss_centerness": 0.5586423313617707, "loss_pos": 0.028648501732386648, "loss_offset": 0.0036856335669290274, "time": 0.13196611404418945, "epoch": 19, "iter": 2994, "memory": 6269, "step": 2994}
+{"coco/bbox_mAP": 0.284, "coco/bbox_mAP_50": 0.75, "coco/bbox_mAP_75": 0.121, "coco/bbox_mAP_s": 0.283, "coco/bbox_mAP_m": 0.282, "coco/bbox_mAP_l": -1.0, "data_time": 0.0019409892905471672, "time": 0.013125775410578801, "step": 19}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.019097867012023925, "grad_norm": 3.9816743516921997, "loss": 0.9159980630874633, "loss_cls": 0.06774190671741963, "loss_bbox": 0.25821553051471713, "loss_centerness": 0.5594174206256867, "loss_pos": 0.027454992034472527, "loss_offset": 0.003168207489652559, "time": 0.13883081436157227, "epoch": 20, "iter": 3052, "memory": 6269, "step": 3052}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.016137003898620605, "grad_norm": 3.8565835833549498, "loss": 0.90747829079628, "loss_cls": 0.06646219410002231, "loss_bbox": 0.25374414771795273, "loss_centerness": 0.5608384728431701, "loss_pos": 0.023101599980145694, "loss_offset": 0.003331877125892788, "time": 0.13580811023712158, "epoch": 20, "iter": 3102, "memory": 6269, "step": 3102}
+{"base_lr": 9.999999999999999e-05, "lr": 4.9999999999999996e-05, "data_time": 0.016083950996398925, "grad_norm": 3.8600285196304323, "loss": 0.9185036134719848, "loss_cls": 0.06824196718633174, "loss_bbox": 0.25938189059495925, "loss_centerness": 0.5617206084728241, "loss_pos": 0.0259188624471426, "loss_offset": 0.0032402819058916068, "time": 0.13465497970581056, "epoch": 20, "iter": 3152, "memory": 6269, "step": 3152}
+{"coco/bbox_mAP": 0.283, "coco/bbox_mAP_50": 0.75, "coco/bbox_mAP_75": 0.118, "coco/bbox_mAP_s": 0.283, "coco/bbox_mAP_m": 0.294, "coco/bbox_mAP_l": -1.0, "data_time": 0.001965687417576456, "time": 0.013327921761406793, "step": 20}

haar_v4_ungated_768/best_coco_bbox_mAP_75_epoch_14.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:6295fdd9f9c5d5eddbb8addb1f660afca9148b43f630e798a59da3694b7c74fc
+size 130283909

haar_v4_ungated_768/epoch_20.pth

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:0151cc92ebe1d689e7bc0aa6c315da51fc5b830729fe6f240fd6240cde23b87b
+size 258712689

haar_v4_ungated_768/last_checkpoint

@@ -0,0 +1 @@
+/marimo/mmdet_code/mmdetection/work_dirs/levir_pahr_v4_screen/haar_v4_ungated_768/epoch_20.pth

haar_v4_ungated_768/patched_config.py

@@ -0,0 +1,295 @@
+auto_scale_lr = dict(base_batch_size=16, enable=True)
+backend_args = None
+custom_imports = dict(imports=[
+    'projects.pahr',
+])
+data_root = 'data/coco/'
+dataset_type = 'CocoDataset'
+default_hooks = dict(
+    checkpoint=dict(
+        interval=1,
+        max_keep_ckpts=1,
+        rule='greater',
+        save_best='coco/bbox_mAP_75',
+        save_last=True,
+        type='CheckpointHook'),
+    logger=dict(interval=50, type='LoggerHook'),
+    param_scheduler=dict(type='ParamSchedulerHook'),
+    sampler_seed=dict(type='DistSamplerSeedHook'),
+    timer=dict(type='IterTimerHook'),
+    visualization=dict(type='DetVisualizationHook'))
+default_scope = 'mmdet'
+env_cfg = dict(
+    cudnn_benchmark=False,
+    dist_cfg=dict(backend='nccl'),
+    mp_cfg=dict(mp_start_method='fork', opencv_num_threads=0))
+launcher = 'none'
+load_from = None
+log_level = 'INFO'
+log_processor = dict(by_epoch=True, type='LogProcessor', window_size=50)
+model = dict(
+    backbone=dict(
+        depth=50,
+        frozen_stages=1,
+        init_cfg=dict(
+            checkpoint='open-mmlab://detectron/resnet50_caffe',
+            type='Pretrained'),
+        norm_cfg=dict(requires_grad=False, type='BN'),
+        norm_eval=True,
+        num_stages=4,
+        out_indices=(
+            0,
+            1,
+            2,
+            3,
+        ),
+        style='caffe',
+        type='ResNet'),
+    bbox_head=dict(
+        feat_channels=256,
+        in_channels=256,
+        loss_bbox=dict(loss_weight=1.0, type='IoULoss'),
+        loss_centerness=dict(
+            loss_weight=1.0, type='CrossEntropyLoss', use_sigmoid=True),
+        loss_cls=dict(
+            alpha=0.25,
+            gamma=2.0,
+            loss_weight=1.0,
+            type='FocalLoss',
+            use_sigmoid=True),
+        num_classes=1,
+        stacked_convs=4,
+        strides=[
+            8,
+            16,
+            32,
+            64,
+            128,
+        ],
+        type='FCOSHead'),
+    data_preprocessor=dict(
+        bgr_to_rgb=False,
+        mean=[
+            102.9801,
+            115.9465,
+            122.7717,
+        ],
+        pad_size_divisor=32,
+        std=[
+            1.0,
+            1.0,
+            1.0,
+        ],
+        type='DetDataPreprocessor'),
+    loss_offset_weight=0.1,
+    loss_pos_weight=0.1,
+    neck=dict(
+        add_extra_convs='on_output',
+        correction_gain=1.0,
+        correction_gate_floor=0.0,
+        detach_position_gate=False,
+        detail_channels=64,
+        gate_power=1.0,
+        guide_channels=0,
+        in_channels=[
+            256,
+            512,
+            1024,
+            2048,
+        ],
+        locator_channels=64,
+        num_outs=5,
+        out_channels=256,
+        relu_before_extra_convs=True,
+        start_level=1,
+        type='PAHRFPN',
+        use_output_gate=False),
+    position_stride=8,
+    test_cfg=dict(
+        max_per_img=100,
+        min_bbox_size=0,
+        nms=dict(iou_threshold=0.5, type='nms'),
+        nms_pre=1000,
+        score_thr=0.05),
+    tiny_max_sqrt_area=16.0,
+    type='PAHRFCOS',
+    use_phase_shift=True)
+optim_wrapper = dict(
+    clip_grad=dict(max_norm=35, norm_type=2),
+    optimizer=dict(lr=0.01, momentum=0.9, type='SGD', weight_decay=0.0001),
+    paramwise_cfg=dict(bias_decay_mult=0.0, bias_lr_mult=2.0),
+    type='OptimWrapper')
+pahr_variant = dict(
+    correction_gain=1.0,
+    detail_lr_mult=1.0,
+    giou=False,
+    guide_channels=0,
+    image_size=768,
+    name='haar_v4_ungated_768',
+    phase_shift=True,
+    scaled_schedule=True,
+    use_output_gate=False,
+    v3_gate=False)
+param_scheduler = [
+    dict(
+        begin=0,
+        by_epoch=False,
+        end=500,
+        factor=0.3333333333333333,
+        type='ConstantLR'),
+    dict(
+        begin=0,
+        by_epoch=True,
+        end=20,
+        gamma=0.1,
+        milestones=[
+            13,
+            18,
+        ],
+        type='MultiStepLR'),
+]
+randomness = dict(seed=42)
+resume = False
+test_cfg = dict(type='TestLoop')
+test_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        ann_file=
+        '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/test.json',
+        backend_args=None,
+        data_prefix=dict(img='/marimo/LevirShipData/All Images/'),
+        data_root='',
+        metainfo=dict(classes=('ship', )),
+        pipeline=[
+            dict(backend_args=None, type='LoadImageFromFile'),
+            dict(keep_ratio=True, scale=(
+                768,
+                768,
+            ), type='Resize'),
+            dict(type='LoadAnnotations', with_bbox=True),
+            dict(
+                meta_keys=(
+                    'img_id',
+                    'img_path',
+                    'ori_shape',
+                    'img_shape',
+                    'scale_factor',
+                ),
+                type='PackDetInputs'),
+        ],
+        test_mode=True,
+        type='CocoDataset'),
+    drop_last=False,
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(shuffle=False, type='DefaultSampler'))
+test_evaluator = dict(
+    ann_file=
+    '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/test.json',
+    backend_args=None,
+    format_only=False,
+    metric='bbox',
+    type='CocoMetric')
+test_pipeline = [
+    dict(backend_args=None, type='LoadImageFromFile'),
+    dict(keep_ratio=True, scale=(
+        1333,
+        800,
+    ), type='Resize'),
+    dict(type='LoadAnnotations', with_bbox=True),
+    dict(
+        meta_keys=(
+            'img_id',
+            'img_path',
+            'ori_shape',
+            'img_shape',
+            'scale_factor',
+        ),
+        type='PackDetInputs'),
+]
+train_cfg = dict(max_epochs=20, type='EpochBasedTrainLoop', val_interval=1)
+train_dataloader = dict(
+    batch_sampler=dict(type='AspectRatioBatchSampler'),
+    batch_size=8,
+    dataset=dict(
+        ann_file=
+        '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/train.json',
+        backend_args=None,
+        data_prefix=dict(img='/marimo/LevirShipData/All Images/'),
+        data_root='',
+        filter_cfg=dict(filter_empty_gt=True, min_size=32),
+        metainfo=dict(classes=('ship', )),
+        pipeline=[
+            dict(backend_args=None, type='LoadImageFromFile'),
+            dict(type='LoadAnnotations', with_bbox=True),
+            dict(keep_ratio=True, scale=(
+                768,
+                768,
+            ), type='Resize'),
+            dict(prob=0.5, type='RandomFlip'),
+            dict(type='PackDetInputs'),
+        ],
+        type='CocoDataset'),
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(shuffle=True, type='DefaultSampler'))
+train_pipeline = [
+    dict(backend_args=None, type='LoadImageFromFile'),
+    dict(type='LoadAnnotations', with_bbox=True),
+    dict(keep_ratio=True, scale=(
+        1333,
+        800,
+    ), type='Resize'),
+    dict(prob=0.5, type='RandomFlip'),
+    dict(type='PackDetInputs'),
+]
+val_cfg = dict(type='ValLoop')
+val_dataloader = dict(
+    batch_size=1,
+    dataset=dict(
+        ann_file=
+        '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/val.json',
+        backend_args=None,
+        data_prefix=dict(img='/marimo/LevirShipData/All Images/'),
+        data_root='',
+        metainfo=dict(classes=('ship', )),
+        pipeline=[
+            dict(backend_args=None, type='LoadImageFromFile'),
+            dict(keep_ratio=True, scale=(
+                768,
+                768,
+            ), type='Resize'),
+            dict(type='LoadAnnotations', with_bbox=True),
+            dict(
+                meta_keys=(
+                    'img_id',
+                    'img_path',
+                    'ori_shape',
+                    'img_shape',
+                    'scale_factor',
+                ),
+                type='PackDetInputs'),
+        ],
+        test_mode=True,
+        type='CocoDataset'),
+    drop_last=False,
+    num_workers=4,
+    persistent_workers=True,
+    sampler=dict(shuffle=False, type='DefaultSampler'))
+val_evaluator = dict(
+    ann_file=
+    '/marimo/mmdet_code/mmdetection/data/levir_ship_coco/annotations/val.json',
+    backend_args=None,
+    format_only=False,
+    metric='bbox',
+    type='CocoMetric')
+vis_backends = [
+    dict(type='LocalVisBackend'),
+]
+visualizer = dict(
+    name='visualizer',
+    type='DetLocalVisualizer',
+    vis_backends=[
+        dict(type='LocalVisBackend'),
+    ])
+work_dir = '/marimo/mmdet_code/mmdetection/work_dirs/levir_pahr_v4_screen/haar_v4_ungated_768'