{"approach": "The current rule's own ONNX graph kept whole, plus a small evolved residual MLP: inputs are the 56 current-frame observations and the rule's 3 clipped outputs, 64 ReLU units, 3 outputs added to the rule's action (4,035 evolved float32 numbers, 35,530 bytes, repo validator PASS, no non-finite output on 20,000 random inputs). The last layer starts at zero, so training starts at exact mirror strength (0.5005 over 1,000) rather than at Starter's 0.025, and no distillation is needed. Antithetic ES with Adam, about 19.6 minutes on 3 threads: 655 s against the current model alone, then 520 s on a mix of 4 parts current, 1 Starter, 1 evader, so it is not a pure exploiter. How it wins, from every 7th decision of 600 matches against the current model, compared with the rule's mirror on the same seeds: it does not change the gun. It fired wherever the rule would and only 1 extra time in the sample. It changes the flying: throttle is 0.22 above what the rule asks in the same state, 0.65 against 0.35 in the first 3 s, so it arrives at 93 m/s against 54, then slows to 54 m/s in the 3-7 s fight at 86 m range (mirror 47 m/s, 99 m). Median range 111 m against 162, altitude about the same (median 269 against 280 m). It fires from 107 m on average, against 133 m in the mirror, when the target's nose is 21 degrees off it against 5, so it is mostly out of the head-on exchange the mirror fights. The rule's hit rate against it falls from 82% to 69%, its own stays near 88%. Matches end in about 7.6 s. Verdict: it beats the current model 0.828 but is not stronger overall. It scores 0.9745 against Starter where the current model scores 0.9913 on the same 3,000 seeds, and a fresh 10-minute exploiter beats it.", "blocked": null, "current_vs_starter": 0.9895, "records": {"starter": {"matches": 3000, "score": 0.9745}, "builtins": {"matches": 800, "score": 0.99875}, "current": {"matches": 1000, "wins": 810, "losses": 154, "draws": 36, "score": 0.828}, "exploiter": {"matches": 500, "score": 0.269, "how": "Same method from a fresh zero residual on the rule, trained only against the final model for 545 s on 4 threads (200 s at lr 0.01, 345 s from its best at lr 0.003). The rule alone scored 0.165 against the final model at the start and the exploiter reached 0.72 to 0.80 on validation. The checkpoint was chosen on separate seeds (0.772 against 0.7655 for the last one) and the record seeds were kept for this measurement. The other checkpoint gave the final model 0.25 on those seeds. The exploiter is itself weak elsewhere: 0.945 against Starter over 1,000."}}, "entry": "CAg6vpUCCiUKA29icwoEZ2lkeBIDc2VsIgZHYXRoZXIqCwoEYXhpcxgBoAECCt0BCgNzZWwSA2luMBIDaW4xEgNpbjISA2luMxIDaW40EgNpbjUSA2luOBIDaW45EgRpbjEwEgRpbjExEgRpbjEyEgRpbjE0EgRpbjE1EgRpbjE2EgRpbjE5EgRpbjIwEgRpbjIxEgRpbjIzEgRpbjI0EgRpbjI1EgRpbjUyEgRpbjU1EgRpbjU2EgRpbjU3EgRpbjU4EgRpbjY2EgRpbjY3EgRpbjY4EgRpbjc1EgRpbjc2EgRpbjc3EgRpbjc5EgRpbjgwEgRpbjgxIgVTcGxpdCoLCgRheGlzGAGgAQIKFAoDaW4wCgJjMRIEbjIyNSIDTXVsChQKA2luMwoCYzISBG4yMjciA011bAoUCgNpbjQKAmMzEgRuMjI5IgNNdWwKFAoDaW41CgJjNBIEbjIzMSIDTXVsChQKA2luOAoCYzUSBG4yMzMiA011bAoUCgNpbjkKAmM1EgRuMjM1IgNNdWwKFQoEaW4xMAoCYzESBG4yMzciA011bAoVCgRpbjE2CgJjNhIEbjIzOSIDTXVsChUKBG4yMzkKAmM3EgRuMjQxIgNNdWwKFQoEaW4xNAoCYzUSBG4yNDMiA011bAoVCgRuMjI1CgJjORIEbjI0NyIDU3ViChYKBG4yNDcKA2MxMBIEbjI0OSIDRGl2ChYKBG4yNDkKA2MxMRIEbjI1MSIDTWF4ChYKBG4yNTEKA2MxMhIEbjI1MyIDTWluChUKAmM4CgRuMjUzEgRuMjU0IgNNdWwKFQoCYzYKBG4yNTQSBG4yNTUiA1N1YgoVCgJjMgoEbjIyNxIEbjI1OSIDU3ViChYKA2MxMwoEbjI1ORIEbjI2MCIDTXVsChYKA2MxMgoEbjI2MBIEbjI2MSIDU3ViChcKBG4yNTUKBG4yNjESBG4yNjIiA011bAoWCgRuMjI5CgNjMTISBG4yNjQiA01heAoWCgRuMjY0CgNjMTQSBG4yNjYiA01pbgoWCgRuMjY2CgNjMTUSBG4yNjgiA0RpdgoXCgRuMjI1CgRuMjY4EgRuMjY5IgNNdWwKFgoEbjI2OQoDaW4xEgRuMjcwIgNNdWwKFwoEbjIyNQoEbjI2OBIEbjI3MSIDTXVsChYKBG4yNzEKA2luMhIEbjI3MiIDTXVsChUKBGluNjYKAmMxEgRuMjc0IgNNdWwKFwoEbjIzNwoEbjI3NBIEbjI3NSIDU3ViChYKBG4yNzUKA2MxNhIEbjI3NyIDTXVsChYKBG4yNzcKA2MxNxIEbjI3OSIDTXVsChYKBG4yNDMKA2MxOBIEbjI4MSIDRGl2ChYKBG4yODEKA2MxORIEbjI4MyIDTWluChcKBG4yNjgKBG4yODMSBG4yODQiA0FkZAoXCgRuMjQxCgRuMjg0EgRuMjg1IgNNdWwKFQoEbjI4NQoCYzASBG4yODciA011bAoXCgRuMjg3CgRuMjg3EgRuMjg4IgNNdWwKFwoEbjIzNwoEbjI4NBIEbjI4OSIDTXVsChYKBG4yODgKA2MyMBIEbjI5MiIDRGl2ChYKA2MxMgoEbjI5MhIEbjI5MyIDU3ViChcKBG4yODgKBG4yODgSBG4yOTQiA011bAoWCgRuMjk0CgNjMTUSBG4yOTYiA0RpdgoXCgRuMjkzCgRuMjk2EgRuMjk3IgNBZGQKFwoEbjI4OQoEbjI5NxIEbjI5OCIDTXVsChcKBG4yNzkKBG4yODQSBG4yOTkiA011bAoXCgRuMjk5CgRuMjg0EgRuMzAwIgNNdWwKFQoEbjMwMAoCYzASBG4zMDIiA011bAoXCgRuMjk4CgRuMzAyEgRuMzAzIgNBZGQKEQoEbjI4NxIEbjMwNCIDQ29zChEKBG4yODcSBG4zMDUiA1NpbgoXCgRpbjExCgRuMzA0EgRuMzA2IgNNdWwKFwoEaW4xMgoEbjMwNRIEbjMwNyIDTXVsChcKBG4zMDYKBG4zMDcSBG4zMDgiA1N1YgoXCgRuMzAzCgRuMzA4EgRuMzA5IgNNdWwKFwoEaW4xMgoEbjMwNBIEbjMxMCIDTXVsChcKBGluMTEKBG4zMDUSBG4zMTEiA011bAoXCgRuMzEwCgRuMzExEgRuMzEyIgNBZGQKFwoEbjMwMwoEbjMxMhIEbjMxMyIDTXVsChcKBG4yMzMKBG4zMDkSBG4zMTQiA0FkZAoXCgRuMzE0CgRuMjcwEgRuMzE1IgNTdWIKFwoEbjIzNQoEbjMxMxIEbjMxNiIDQWRkChcKBG4zMTYKBG4yNzISBG4zMTciA1N1YgoXCgRuMzE1CgRuMzE1EgRuMzE4IgNNdWwKFwoEbjMxNwoEbjMxNxIEbjMxOSIDTXVsChcKBG4zMTgKBG4zMTkSBG4zMjAiA0FkZAoSCgRuMzIwEgRuMzIxIgRTcXJ0ChYKBG4zMjEKA2MxNhIEbjMyMyIDU3ViChYKBG4zMjMKA2MxMRIEbjMyNSIDTWF4ChYKBG4zMjUKA2MxOBIEbjMyNyIDRGl2ChYKBG4zMjcKA2MxORIEbjMyOSIDTWluChcKBG4yNjgKBG4zMjkSBG4zMzAiA0FkZAoXCgRuMjQxCgRuMzMwEgRuMzMxIgNNdWwKFQoEbjMzMQoCYzASBG4zMzMiA011bAoXCgRuMzMzCgRuMzMzEgRuMzM0IgNNdWwKFwoEbjIzNwoEbjMzMBIEbjMzNSIDTXVsChYKBG4zMzQKA2MyMBIEbjMzOCIDRGl2ChYKA2MxMgoEbjMzOBIEbjMzOSIDU3ViChcKBG4zMzQKBG4zMzQSBG4zNDAiA011bAoWCgRuMzQwCgNjMTUSBG4zNDIiA0RpdgoXCgRuMzM5CgRuMzQyEgRuMzQzIgNBZGQKFwoEbjMzNQoEbjM0MxIEbjM0NCIDTXVsChcKBG4yNzkKBG4zMzASBG4zNDUiA011bAoXCgRuMzQ1CgRuMzMwEgRuMzQ2IgNNdWwKFQoEbjM0NgoCYzASBG4zNDgiA011bAoXCgRuMzQ0CgRuMzQ4EgRuMzQ5IgNBZGQKEQoEbjMzMxIEbjM1MCIDQ29zChEKBG4zMzMSBG4zNTEiA1NpbgoXCgRpbjExCgRuMzUwEgRuMzUyIgNNdWwKFwoEaW4xMgoEbjM1MRIEbjM1MyIDTXVsChcKBG4zNTIKBG4zNTMSBG4zNTQiA1N1YgoXCgRuMzQ5CgRuMzU0EgRuMzU1IgNNdWwKFwoEaW4xMgoEbjM1MBIEbjM1NiIDTXVsChcKBGluMTEKBG4zNTESBG4zNTciA011bAoXCgRuMzU2CgRuMzU3EgRuMzU4IgNBZGQKFwoEbjM0OQoEbjM1OBIEbjM1OSIDTXVsChcKBG4yMzMKBG4zNTUSBG4zNjAiA0FkZAoXCgRuMzYwCgRuMjcwEgRuMzYxIgNTdWIKFwoEbjIzNQoEbjM1ORIEbjM2MiIDQWRkChcKBG4zNjIKBG4yNzISBG4zNjMiA1N1YgoXCgRuMzYxCgRuMzYxEgRuMzY0IgNNdWwKFwoEbjM2MwoEbjM2MxIEbjM2NSIDTXVsChcKBG4zNjQKBG4zNjUSBG4zNjYiA0FkZAoSCgRuMzY2EgRuMzY3IgRTcXJ0ChYKBG4zNjcKA2MxNhIEbjM2OSIDU3ViChYKBG4zNjkKA2MxMRIEbjM3MSIDTWF4ChYKBG4zNzEKA2MxOBIEbjM3MyIDRGl2ChYKBG4zNzMKA2MxORIEbjM3NSIDTWluChcKBG4yNjgKBG4zNzUSBG4zNzYiA0FkZAoXCgRuMjQxCgRuMzc2EgRuMzc3IgNNdWwKFQoEbjM3NwoCYzASBG4zNzkiA011bAoXCgRuMzc5CgRuMzc5EgRuMzgwIgNNdWwKFwoEbjIzNwoEbjM3NhIEbjM4MSIDTXVsChYKBG4zODAKA2MyMBIEbjM4NCIDRGl2ChYKA2MxMgoEbjM4NBIEbjM4NSIDU3ViChcKBG4zODAKBG4zODASBG4zODYiA011bAoWCgRuMzg2CgNjMTUSBG4zODgiA0RpdgoXCgRuMzg1CgRuMzg4EgRuMzg5IgNBZGQKFwoEbjM4MQoEbjM4ORIEbjM5MCIDTXVsChcKBG4yNzkKBG4zNzYSBG4zOTEiA011bAoXCgRuMzkxCgRuMzc2EgRuMzkyIgNNdWwKFQoEbjM5MgoCYzASBG4zOTQiA011bAoXCgRuMzkwCgRuMzk0EgRuMzk1IgNBZGQKEQoEbjM3ORIEbjM5NiIDQ29zChEKBG4zNzkSBG4zOTciA1NpbgoXCgRpbjExCgRuMzk2EgRuMzk4IgNNdWwKFwoEaW4xMgoEbjM5NxIEbjM5OSIDTXVsChcKBG4zOTgKBG4zOTkSBG40MDAiA1N1YgoXCgRuMzk1CgRuNDAwEgRuNDAxIgNNdWwKFwoEaW4xMgoEbjM5NhIEbjQwMiIDTXVsChcKBGluMTEKBG4zOTcSBG40MDMiA011bAoXCgRuNDAyCgRuNDAzEgRuNDA0IgNBZGQKFwoEbjM5NQoEbjQwNBIEbjQwNSIDTXVsChcKBG4yMzMKBG40MDESBG40MDYiA0FkZAoXCgRuNDA2CgRuMjcwEgRuNDA3IgNTdWIKFwoEbjIzNQoEbjQwNRIEbjQwOCIDQWRkChcKBG40MDgKBG4yNzISBG40MDkiA1N1YgoXCgRuNDA3CgRuNDA3EgRuNDEwIgNNdWwKFwoEbjQwOQoEbjQwORIEbjQxMSIDTXVsChcKBG40MTAKBG40MTESBG40MTIiA0FkZAoSCgRuNDEyEgRuNDEzIgRTcXJ0ChYKBG40MTMKA2MxNhIEbjQxNSIDU3ViChYKBG40MTUKA2MxMRIEbjQxNyIDTWF4ChYKBG40MTcKA2MxOBIEbjQxOSIDRGl2ChcKBG40MDcKBG40MDcSBG40MjAiA011bAoXCgRuNDA5CgRuNDA5EgRuNDIxIgNNdWwKFwoEbjQyMAoEbjQyMRIEbjQyMiIDQWRkChIKBG40MjISBG40MjMiBFNxcnQKFgoEbjQyMwoDYzIxEgRuNDI1IgNBZGQKFgoDaW4xCgRuNDA5EgRuNDI2IgNNdWwKFgoDaW4yCgRuNDA3EgRuNDI3IgNNdWwKFwoEbjQyNgoEbjQyNxIEbjQyOCIDU3ViChYKA2luMQoEbjQwNxIEbjQyOSIDTXVsChYKA2luMgoEbjQwORIEbjQzMCIDTXVsChcKBG40MjkKBG40MzASBG40MzEiA0FkZAoRCgRuNDMxEgRuNDMzIgNBYnMKGwoEbjQzMwoDYzIyEgVuNDM0YiIHR3JlYXRlcgoeCgVuNDM0YhIEbjQzNCIEQ2FzdCoJCgJ0bxgBoAECChoKBG40MzQKAmMwEgVuNDM1YiIHR3JlYXRlcgofCgVuNDM1YgoEbjQzMQoDYzIyEgRuNDM1IgVXaGVyZQoXCgRuNDI4CgRuNDM1EgRuNDM2IgNEaXYKEgoEbjQzNhIEbjQzNyIEQXRhbgobCgNjMTEKBG40MzESBW40NDBiIgdHcmVhdGVyCh4KBW40NDBiEgRuNDQwIgRDYXN0KgkKAnRvGAGgAQIKGwoDYzExCgRuNDI4EgVuNDQyYiIHR3JlYXRlcgoeCgVuNDQyYhIEbjQ0MiIEQ2FzdCoJCgJ0bxgBoAECChYKBG40MzcKA2MyMxIEbjQ0MyIDU3ViChYKBG40MzcKA2MyMxIEbjQ0NCIDQWRkChoKBG40NDIKAmMwEgVuNDQ1YiIHR3JlYXRlcgogCgVuNDQ1YgoEbjQ0MwoEbjQ0NBIEbjQ0NSIFV2hlcmUKGgoEbjQ0MAoCYzASBW40NDZiIgdHcmVhdGVyCiAKBW40NDZiCgRuNDQ1CgRuNDM3EgRuNDQ2IgVXaGVyZQoXCgRuMjY4CgRuNDE5EgRuNDQ3IgNBZGQKFwoEbjI0MQoEbjQ0NxIEbjQ0OCIDTXVsChEKBG40NDgSBG40NDkiA0NvcwoRCgRuNDQ4EgRuNDUwIgNTaW4KFwoEaW4xMQoEbjQ0ORIEbjQ1MSIDTXVsChcKBGluMTIKBG40NTASBG40NTIiA011bAoXCgRuNDUxCgRuNDUyEgRuNDUzIgNTdWIKFwoEaW4xMgoEbjQ0ORIEbjQ1NCIDTXVsChcKBGluMTEKBG40NTASBG40NTUiA011bAoXCgRuNDU0CgRuNDU1EgRuNDU2IgNBZGQKFwoEbjQwNwoEbjQ1MxIEbjQ1NyIDTXVsChcKBG40MDkKBG40NTYSBG40NTgiA011bAoXCgRuNDU3CgRuNDU4EgRuNDU5IgNBZGQKFwoEbjQ1OQoEbjQyNRIEbjQ2MCIDRGl2ChcKBG40NTMKBG4yMzUSBG40NjEiA011bAoXCgRuNDU2CgRuMjMzEgRuNDYyIgNNdWwKFwoEbjQ2MQoEbjQ2MhIEbjQ2MyIDU3ViChsKBG40NjMKA2MxMRIFbjQ2NWIiB0dyZWF0ZXIKHgoFbjQ2NWISBG40NjUiBENhc3QqCQoCdG8YAaABAgoaCgRuNDY1CgJjMBIFbjQ2OGIiB0dyZWF0ZXIKHgoFbjQ2OGIKA2MyNAoDYzEyEgRuNDY4IgVXaGVyZQobCgRuNDYwCgNjMjUSBW40NzBiIgdHcmVhdGVyCh4KBW40NzBiEgRuNDcwIgRDYXN0KgkKAnRvGAGgAQIKEQoEbjQ1NhIEbjQ3MSIDTmVnChcKBG40NzEKBG40NjgSBG40NzIiA011bAoWCgRuNDcyCgNjMjYSBG40NzQiA011bAoaCgRuNDcwCgJjMBIFbjQ3NmIiB0dyZWF0ZXIKHwoFbjQ3NmIKBG40NzQKA2MxMRIEbjQ3NiIFV2hlcmUKFwoEbjQwNwoEbjQ3NhIEbjQ3NyIDQWRkChcKBG40NTMKBG40NjgSBG40NzgiA011bAoWCgRuNDc4CgNjMjYSBG40ODAiA011bAoaCgRuNDcwCgJjMBIFbjQ4MmIiB0dyZWF0ZXIKHwoFbjQ4MmIKBG40ODAKA2MxMRIEbjQ4MiIFV2hlcmUKFwoEbjQwOQoEbjQ4MhIEbjQ4MyIDQWRkChcKBG4yMzMKBGluMTESBG40ODQiA011bAoXCgRuMjM1CgRpbjEyEgRuNDg1IgNNdWwKFwoEbjQ4NAoEbjQ4NRIEbjQ4NiIDQWRkChYKBG4yNDMKA2MxMhIEbjQ4OCIDQWRkChcKBG40ODYKBG40ODgSBG40ODkiA0RpdgobCgRuNDg5CgNjMjcSBW40OTFiIgdHcmVhdGVyCh4KBW40OTFiEgRuNDkxIgRDYXN0KgkKAnRvGAGgAQIKGwoEbjI0MwoDYzI4EgVuNDkzYiIHR3JlYXRlcgoeCgVuNDkzYhIEbjQ5MyIEQ2FzdCoJCgJ0bxgBoAECChcKBG40OTEKBG40OTMSBG40OTQiA011bAobCgRuNDc3CgNjMTESBW40OTZiIgdHcmVhdGVyCh4KBW40OTZiEgRuNDk2IgRDYXN0KgkKAnRvGAGgAQIKGgoEbjQ5NgoCYzASBW40OTliIgdHcmVhdGVyCh0KBW40OTliCgJjNQoDYzI5EgRuNDk5IgVXaGVyZQoXCgRuNDc3CgRuNDk5EgRuNTAwIgNTdWIKGgoEbjQ5NAoCYzASBW41MDFiIgdHcmVhdGVyCiAKBW41MDFiCgRuNTAwCgRuNDc3EgRuNTAxIgVXaGVyZQoWCgNpbjEKBG40ODMSBG41MDIiA011bAoWCgNpbjIKBG41MDESBG41MDMiA011bAoXCgRuNTAyCgRuNTAzEgRuNTA0IgNTdWIKFgoDaW4xCgRuNTAxEgRuNTA1IgNNdWwKFgoDaW4yCgRuNDgzEgRuNTA2IgNNdWwKFwoEbjUwNQoEbjUwNhIEbjUwNyIDQWRkChEKBG41MDcSBG41MDkiA0FicwobCgRuNTA5CgNjMjISBW41MTBiIgdHcmVhdGVyCh4KBW41MTBiEgRuNTEwIgRDYXN0KgkKAnRvGAGgAQIKGgoEbjUxMAoCYzASBW41MTFiIgdHcmVhdGVyCh8KBW41MTFiCgRuNTA3CgNjMjISBG41MTEiBVdoZXJlChcKBG41MDQKBG41MTESBG41MTIiA0RpdgoSCgRuNTEyEgRuNTEzIgRBdGFuChsKA2MxMQoEbjUwNxIFbjUxNmIiB0dyZWF0ZXIKHgoFbjUxNmISBG41MTYiBENhc3QqCQoCdG8YAaABAgobCgNjMTEKBG41MDQSBW41MThiIgdHcmVhdGVyCh4KBW41MThiEgRuNTE4IgRDYXN0KgkKAnRvGAGgAQIKFgoEbjUxMwoDYzIzEgRuNTE5IgNTdWIKFgoEbjUxMwoDYzIzEgRuNTIwIgNBZGQKGgoEbjUxOAoCYzASBW41MjFiIgdHcmVhdGVyCiAKBW41MjFiCgRuNTE5CgRuNTIwEgRuNTIxIgVXaGVyZQoaCgRuNTE2CgJjMBIFbjUyMmIiB0dyZWF0ZXIKIAoFbjUyMmIKBG41MjEKBG41MTMSBG41MjIiBVdoZXJlChYKBG4yNjIKA2MxNhIEbjUyNCIDRGl2ChcKBG41MjIKBG41MjQSBG41MjUiA0RpdgoWCgRuNTI1CgNjMzASBG41MjciA011bAoWCgRuNTI3CgNjMjQSBG41MjkiA01heAoWCgRuNTI5CgNjMTISBG41MzEiA01pbgoaCgJjMAoEaW4yMRIFbjUzM2IiB0dyZWF0ZXIKHgoFbjUzM2ISBG41MzMiBENhc3QqCQoCdG8YAaABAgoRCgRpbjE5EgRuNTM0IgNBYnMKEQoEaW4yMBIEbjUzNSIDQWJzChcKBG41MzQKBG41MzUSBG41MzYiA0FkZAobCgRuNTM2CgNjMzESBW41MzhiIgdHcmVhdGVyCh4KBW41MzhiEgRuNTM4IgRDYXN0KgkKAnRvGAGgAQIKFwoEbjUzMwoEbjUzOBIEbjUzOSIDTXVsChoKAmMwCgRpbjI1EgVuNTQxYiIHR3JlYXRlcgoeCgVuNTQxYhIEbjU0MSIEQ2FzdCoJCgJ0bxgBoAECChEKBGluMjMSBG41NDIiA0FicwoRCgRpbjI0EgRuNTQzIgNBYnMKFwoEbjU0MgoEbjU0MxIEbjU0NCIDQWRkChsKBG41NDQKA2MzMRIFbjU0NmIiB0dyZWF0ZXIKHgoFbjU0NmISBG41NDYiBENhc3QqCQoCdG8YAaABAgoXCgRuNTQxCgRuNTQ2EgRuNTQ3IgNNdWwKGgoEbjUzOQoCYzASBW41NDhiIgdHcmVhdGVyCiAKBW41NDhiCgRpbjE5CgRpbjIzEgRuNTQ4IgVXaGVyZQoVCgRuNTQ4CgJjNRIEbjU1MCIDTXVsChoKBG41MzkKAmMwEgVuNTUxYiIHR3JlYXRlcgogCgVuNTUxYgoEaW4yMAoEaW4yNBIEbjU1MSIFV2hlcmUKFQoEbjU1MQoCYzUSBG41NTMiA011bAoXCgRuNTM5CgRuNTQ3EgRuNTU0IgNNYXgKGgoCYzAKBGluNzcSBW41NTZiIgdHcmVhdGVyCh4KBW41NTZiEgRuNTU2IgRDYXN0KgkKAnRvGAGgAQIKEQoEaW43NRIEbjU1NyIDQWJzChEKBGluNzYSBG41NTgiA0FicwoXCgRuNTU3CgRuNTU4EgRuNTU5IgNBZGQKGwoEbjU1OQoDYzMxEgVuNTYxYiIHR3JlYXRlcgoeCgVuNTYxYhIEbjU2MSIEQ2FzdCoJCgJ0bxgBoAECChcKBG41NTYKBG41NjESBG41NjIiA011bAoaCgJjMAoEaW44MRIFbjU2NGIiB0dyZWF0ZXIKHgoFbjU2NGISBG41NjQiBENhc3QqCQoCdG8YAaABAgoRCgRpbjc5EgRuNTY1IgNBYnMKEQoEaW44MBIEbjU2NiIDQWJzChcKBG41NjUKBG41NjYSBG41NjciA0FkZAobCgRuNTY3CgNjMzESBW41NjliIgdHcmVhdGVyCh4KBW41NjliEgRuNTY5IgRDYXN0KgkKAnRvGAGgAQIKFwoEbjU2NAoEbjU2ORIEbjU3MCIDTXVsChoKBG41NjIKAmMwEgVuNTcxYiIHR3JlYXRlcgogCgVuNTcxYgoEaW43NQoEaW43ORIEbjU3MSIFV2hlcmUKFQoEbjU3MQoCYzUSBG41NzMiA011bAoaCgRuNTYyCgJjMBIFbjU3NGIiB0dyZWF0ZXIKIAoFbjU3NGIKBGluNzYKBGluODASBG41NzQiBVdoZXJlChUKBG41NzQKAmM1EgRuNTc2IgNNdWwKFwoEbjU2MgoEbjU3MBIEbjU3NyIDTWF4ChUKBGluNTYKAmMxEgRuNTc5IgNNdWwKFgoEbjIyNQoDaW4xEgRuNTgwIgNNdWwKFwoEbjU3OQoEaW41NxIEbjU4MSIDTXVsChcKBG41ODAKBG41ODESBG41ODIiA0FkZAoWCgRuNTgyCgNjMzISBG41ODQiA011bAoWCgRuMjI1CgNpbjISBG41ODUiA011bAoXCgRuNTc5CgRpbjU4EgRuNTg2IgNNdWwKFwoEbjU4NQoEbjU4NhIEbjU4NyIDQWRkChYKBG41ODcKA2MzMhIEbjU4OSIDTXVsChcKBG41NTAKBG41ODQSBG41OTAiA0FkZAoXCgRuNTkwCgRuNTczEgRuNTkxIgNTdWIKFwoEbjU1MwoEbjU4ORIEbjU5MiIDQWRkChcKBG41OTIKBG41NzYSBG41OTMiA1N1YgoXCgRpbjExCgRpbjY3EgRuNTk0IgNBZGQKFwoEaW4xMgoEaW42OBIEbjU5NSIDQWRkChoKBG41NzcKAmMwEgVuNTk2YiIHR3JlYXRlcgogCgVuNTk2YgoEbjU5MQoEbjU5NBIEbjU5NiIFV2hlcmUKGgoEbjU3NwoCYzASBW41OTdiIgdHcmVhdGVyCiAKBW41OTdiCgRuNTkzCgRuNTk1EgRuNTk3IgVXaGVyZQoXCgRuNTk2CgRuNTk2EgRuNTk4IgNNdWwKFwoEbjU5NwoEbjU5NxIEbjU5OSIDTXVsChcKBG41OTgKBG41OTkSBG42MDAiA0FkZAoSCgRuNjAwEgRuNjAxIgRTcXJ0ChYKBG42MDEKA2MzMRIEbjYwMyIDQWRkChcKBG41OTYKBG42MDMSBG42MDQiA0RpdgoXCgRuNTk3CgRuNjAzEgRuNjA1IgNEaXYKFgoEbjYwNAoDYzE4EgRuNjA3IgNNdWwKFgoEbjYwNQoDYzE4EgRuNjA5IgNNdWwKFgoEbjIyNQoDaW4xEgRuNjEwIgNNdWwKFwoEbjYwNwoEbjYxMBIEbjYxMSIDU3ViChYKBG4yMjUKA2luMhIEbjYxMiIDTXVsChcKBG42MDkKBG42MTISBG42MTMiA1N1YgoXCgRuNTUwCgRuNjExEgRuNjE0IgNNdWwKFwoEbjU1MwoEbjYxMxIEbjYxNSIDTXVsChcKBG42MTQKBG42MTUSBG42MTYiA0FkZAoRCgRuNjE2EgRuNjE3IgNOZWcKFwoEbjYxMQoEbjYxMRIEbjYxOCIDTXVsChcKBG42MTMKBG42MTMSBG42MTkiA011bAoXCgRuNjE4CgRuNjE5EgRuNjIwIgNBZGQKFgoEbjYyMAoDYzEyEgRuNjIyIgNBZGQKFwoEbjYxNwoEbjYyMhIEbjYyMyIDRGl2ChYKBG42MjMKA2MxMRIEbjYyNSIDTWF4ChUKBG42MjUKAmMwEgRuNjI3IgNNaW4KFwoEbjYwNwoEbjYyNxIEbjYyOCIDTXVsChcKBG41NTAKBG42MjgSBG42MjkiA0FkZAoXCgRuNjA5CgRuNjI3EgRuNjMwIgNNdWwKFwoEbjU1MwoEbjYzMBIEbjYzMSIDQWRkChYKA2MxMgoEbjI2MhIEbjYzMyIDTXVsChcKBG42MzMKBG42MjcSBG42MzQiA011bAoVCgRuNjM0CgJjMBIEbjYzNiIDTXVsChcKBG42MzYKBG42MzYSBG42MzciA011bAoXCgRuMjI1CgRuNjI3EgRuNjM4IgNNdWwKFgoEbjYzNwoDYzIwEgRuNjQxIgNEaXYKFgoDYzEyCgRuNjQxEgRuNjQyIgNTdWIKFwoEbjYzNwoEbjYzNxIEbjY0MyIDTXVsChYKBG42NDMKA2MxNRIEbjY0NSIDRGl2ChcKBG42NDIKBG42NDUSBG42NDYiA0FkZAoXCgRuNjM4CgRuNjQ2EgRuNjQ3IgNNdWwKEQoEbjYzNhIEbjY0OCIDQ29zChEKBG42MzYSBG42NDkiA1NpbgoWCgNpbjEKBG42NDgSBG42NTAiA011bAoWCgNpbjIKBG42NDkSBG42NTEiA011bAoXCgRuNjUwCgRuNjUxEgRuNjUyIgNTdWIKFwoEbjY0NwoEbjY1MhIEbjY1MyIDTXVsChYKA2luMgoEbjY0OBIEbjY1NCIDTXVsChYKA2luMQoEbjY0ORIEbjY1NSIDTXVsChcKBG42NTQKBG42NTUSBG42NTYiA0FkZAoXCgRuNjQ3CgRuNjU2EgRuNjU3IgNNdWwKFwoEbjYyOQoEbjY1MxIEbjY1OCIDU3ViChcKBG42MzEKBG42NTcSBG42NTkiA1N1YgoXCgRuNjU4CgRuNjU4EgRuNjYwIgNNdWwKFwoEbjY1OQoEbjY1ORIEbjY2MSIDTXVsChcKBG42NjAKBG42NjESBG42NjIiA0FkZAoSCgRuNjYyEgRuNjYzIgRTcXJ0ChYKBG42MzYKA2MzMxIEbjY2NSIDTXVsChEKBG42NjUSBG42NjYiA0NvcwoWCgRuNjM2CgNjMzMSBG42NjgiA011bAoRCgRuNjY4EgRuNjY5IgNTaW4KFgoDaW4xCgRuNjY2EgRuNjcwIgNNdWwKFgoDaW4yCgRuNjY5EgRuNjcxIgNNdWwKFwoEbjY3MAoEbjY3MRIEbjY3MiIDU3ViChcKBG42MDQKBG42NzISBG42NzMiA011bAoWCgNpbjIKBG42NjYSBG42NzQiA011bAoWCgNpbjEKBG42NjkSBG42NzUiA011bAoXCgRuNjc0CgRuNjc1EgRuNjc2IgNBZGQKFwoEbjYwNQoEbjY3NhIEbjY3NyIDTXVsChcKBG42NzMKBG42NzcSBG42NzgiA0FkZAobCgRuNjc4CgNjMzQSBW42ODBiIgdHcmVhdGVyCh4KBW42ODBiEgRuNjgwIgRDYXN0KgkKAnRvGAGgAQIKGgoEbjY4MAoCYzASBW42ODJiIgdHcmVhdGVyCh8KBW42ODJiCgNjMzUKBG42NjMSBG42ODIiBVdoZXJlChYKA2MyNAoEbjI2MhIEbjY4NCIDTXVsChcKBG42ODQKBG42MjcSBG42ODUiA011bAoVCgRuNjg1CgJjMBIEbjY4NyIDTXVsChcKBG42ODcKBG42ODcSBG42ODgiA011bAoXCgRuMjI1CgRuNjI3EgRuNjg5IgNNdWwKFgoEbjY4OAoDYzIwEgRuNjkyIgNEaXYKFgoDYzEyCgRuNjkyEgRuNjkzIgNTdWIKFwoEbjY4OAoEbjY4OBIEbjY5NCIDTXVsChYKBG42OTQKA2MxNRIEbjY5NiIDRGl2ChcKBG42OTMKBG42OTYSBG42OTciA0FkZAoXCgRuNjg5CgRuNjk3EgRuNjk4IgNNdWwKEQoEbjY4NxIEbjY5OSIDQ29zChEKBG42ODcSBG43MDAiA1NpbgoWCgNpbjEKBG42OTkSBG43MDEiA011bAoWCgNpbjIKBG43MDASBG43MDIiA011bAoXCgRuNzAxCgRuNzAyEgRuNzAzIgNTdWIKFwoEbjY5OAoEbjcwMxIEbjcwNCIDTXVsChYKA2luMgoEbjY5ORIEbjcwNSIDTXVsChYKA2luMQoEbjcwMBIEbjcwNiIDTXVsChcKBG43MDUKBG43MDYSBG43MDciA0FkZAoXCgRuNjk4CgRuNzA3EgRuNzA4IgNNdWwKFwoEbjYyOQoEbjcwNBIEbjcwOSIDU3ViChcKBG42MzEKBG43MDgSBG43MTAiA1N1YgoXCgRuNzA5CgRuNzA5EgRuNzExIgNNdWwKFwoEbjcxMAoEbjcxMBIEbjcxMiIDTXVsChcKBG43MTEKBG43MTISBG43MTMiA0FkZAoSCgRuNzEzEgRuNzE0IgRTcXJ0ChYKBG42ODcKA2MzMxIEbjcxNiIDTXVsChEKBG43MTYSBG43MTciA0NvcwoWCgRuNjg3CgNjMzMSBG43MTkiA011bAoRCgRuNzE5EgRuNzIwIgNTaW4KFgoDaW4xCgRuNzE3EgRuNzIxIgNNdWwKFgoDaW4yCgRuNzIwEgRuNzIyIgNNdWwKFwoEbjcyMQoEbjcyMhIEbjcyMyIDU3ViChcKBG42MDQKBG43MjMSBG43MjQiA011bAoWCgNpbjIKBG43MTcSBG43MjUiA011bAoWCgNpbjEKBG43MjASBG43MjYiA011bAoXCgRuNzI1CgRuNzI2EgRuNzI3IgNBZGQKFwoEbjYwNQoEbjcyNxIEbjcyOCIDTXVsChcKBG43MjQKBG43MjgSBG43MjkiA0FkZAobCgRuNzI5CgNjMzQSBW43MzFiIgdHcmVhdGVyCh4KBW43MzFiEgRuNzMxIgRDYXN0KgkKAnRvGAGgAQIKGgoEbjczMQoCYzASBW43MzNiIgdHcmVhdGVyCh8KBW43MzNiCgNjMzUKBG43MTQSBG43MzMiBVdoZXJlChcKBG41MzEKBG4yNjISBG43MzQiA011bAoXCgRuNzM0CgRuNjI3EgRuNzM1IgNNdWwKFQoEbjczNQoCYzASBG43MzciA011bAoXCgRuNzM3CgRuNzM3EgRuNzM4IgNNdWwKFwoEbjIyNQoEbjYyNxIEbjczOSIDTXVsChYKBG43MzgKA2MyMBIEbjc0MiIDRGl2ChYKA2MxMgoEbjc0MhIEbjc0MyIDU3ViChcKBG43MzgKBG43MzgSBG43NDQiA011bAoWCgRuNzQ0CgNjMTUSBG43NDYiA0RpdgoXCgRuNzQzCgRuNzQ2EgRuNzQ3IgNBZGQKFwoEbjczOQoEbjc0NxIEbjc0OCIDTXVsChEKBG43MzcSBG43NDkiA0NvcwoRCgRuNzM3EgRuNzUwIgNTaW4KFgoDaW4xCgRuNzQ5EgRuNzUxIgNNdWwKFgoDaW4yCgRuNzUwEgRuNzUyIgNNdWwKFwoEbjc1MQoEbjc1MhIEbjc1MyIDU3ViChcKBG43NDgKBG43NTMSBG43NTQiA011bAoWCgNpbjIKBG43NDkSBG43NTUiA011bAoWCgNpbjEKBG43NTASBG43NTYiA011bAoXCgRuNzU1CgRuNzU2EgRuNzU3IgNBZGQKFwoEbjc0OAoEbjc1NxIEbjc1OCIDTXVsChcKBG42MjkKBG43NTQSBG43NTkiA1N1YgoXCgRuNjMxCgRuNzU4EgRuNzYwIgNTdWIKFwoEbjc1OQoEbjc1ORIEbjc2MSIDTXVsChcKBG43NjAKBG43NjASBG43NjIiA011bAoXCgRuNzYxCgRuNzYyEgRuNzYzIgNBZGQKEgoEbjc2MxIEbjc2NCIEU3FydAoWCgRuNzM3CgNjMzMSBG43NjYiA011bAoRCgRuNzY2EgRuNzY3IgNDb3MKFgoEbjczNwoDYzMzEgRuNzY5IgNNdWwKEQoEbjc2ORIEbjc3MCIDU2luChYKA2luMQoEbjc2NxIEbjc3MSIDTXVsChYKA2luMgoEbjc3MBIEbjc3MiIDTXVsChcKBG43NzEKBG43NzISBG43NzMiA1N1YgoXCgRuNjA0CgRuNzczEgRuNzc0IgNNdWwKFgoDaW4yCgRuNzY3EgRuNzc1IgNNdWwKFgoDaW4xCgRuNzcwEgRuNzc2IgNNdWwKFwoEbjc3NQoEbjc3NhIEbjc3NyIDQWRkChcKBG42MDUKBG43NzcSBG43NzgiA011bAoXCgRuNzc0CgRuNzc4EgRuNzc5IgNBZGQKGwoEbjc3OQoDYzM0EgVuNzgxYiIHR3JlYXRlcgoeCgVuNzgxYhIEbjc4MSIEQ2FzdCoJCgJ0bxgBoAECChoKBG43ODEKAmMwEgVuNzgzYiIHR3JlYXRlcgofCgVuNzgzYgoDYzM1CgRuNzY0EgRuNzgzIgVXaGVyZQobCgRuNjI3CgNjMTESBW43ODViIgdHcmVhdGVyCh4KBW43ODViEgRuNzg1IgRDYXN0KgkKAnRvGAGgAQIKFwoEbjU1NAoEbjc4NRIEbjc4NiIDTXVsChsKA2MzNgoEbjYyNxIFbjc4OGIiB0dyZWF0ZXIKHgoFbjc4OGISBG43ODgiBENhc3QqCQoCdG8YAaABAgoXCgRuNzg2CgRuNzg4EgRuNzg5IgNNdWwKGwoDYzM3CgRuNzgzEgVuNzkxYiIHR3JlYXRlcgoeCgVuNzkxYhIEbjc5MSIEQ2FzdCoJCgJ0bxgBoAECChcKBG43ODkKBG43OTESBG43OTIiA011bAobCgRuMjI5CgNjMzgSBW43OTRiIgdHcmVhdGVyCh4KBW43OTRiEgRuNzk0IgRDYXN0KgkKAnRvGAGgAQIKFwoEbjc5MgoEbjc5NBIEbjc5NSIDTXVsChwKBG42ODIKBG43MzMSBW43OTZiIgdHcmVhdGVyCh4KBW43OTZiEgRuNzk2IgRDYXN0KgkKAnRvGAGgAQIKGgoEbjc5NgoCYzASBW43OTliIgdHcmVhdGVyCh4KBW43OTliCgNjMTIKA2MyNBIEbjc5OSIFV2hlcmUKEQoEaW41MhIEbjgwMCIDQWJzChUKA2MzOQoDYzIzEgRuODAzIgNEaXYKHAoEbjgwMwoEbjgwMBIFbjgwNGIiB0dyZWF0ZXIKHgoFbjgwNGISBG44MDQiBENhc3QqCQoCdG8YAaABAgobCgNjNDAKBG4yNDMSBW44MDZiIgdHcmVhdGVyCh4KBW44MDZiEgRuODA2IgRDYXN0KgkKAnRvGAGgAQIKFwoEbjgwNAoEbjgwNhIEbjgwNyIDTXVsChsKBG4yMjkKA2M0MRIFbjgwOWIiB0dyZWF0ZXIKHgoFbjgwOWISBG44MDkiBENhc3QqCQoCdG8YAaABAgoXCgRuODA5CgRuODA3EgRuODEwIgNNdWwKEQoEbjIzNRIEbjgxMSIDTmVnChYKBG44MTEKA2luMRIEbjgxMiIDTXVsChYKBG4yMzMKA2luMhIEbjgxMyIDTXVsChcKBG44MTIKBG44MTMSBG44MTQiA0FkZAobCgRuODE0CgNjMTESBW44MTZiIgdHcmVhdGVyCh4KBW44MTZiEgRuODE2IgRDYXN0KgkKAnRvGAGgAQIKGgoEbjgxNgoCYzASBW44MTliIgdHcmVhdGVyCh4KBW44MTliCgNjMTIKA2MyNBIEbjgxOSIFV2hlcmUKFwoEbjgxMQoEbjgxORIEbjgyMCIDTXVsChcKBG4yMzMKBG44MTkSBG44MjEiA011bAoWCgNpbjEKBG44MjESBG44MjIiA011bAoWCgNpbjIKBG44MjASBG44MjMiA011bAoXCgRuODIyCgRuODIzEgRuODI0IgNTdWIKFgoDaW4xCgRuODIwEgRuODI1IgNNdWwKFgoDaW4yCgRuODIxEgRuODI2IgNNdWwKFwoEbjgyNQoEbjgyNhIEbjgyNyIDQWRkChEKBG44MjcSBG44MjkiA0FicwobCgRuODI5CgNjMjISBW44MzBiIgdHcmVhdGVyCh4KBW44MzBiEgRuODMwIgRDYXN0KgkKAnRvGAGgAQIKGgoEbjgzMAoCYzASBW44MzFiIgdHcmVhdGVyCh8KBW44MzFiCgRuODI3CgNjMjISBG44MzEiBVdoZXJlChcKBG44MjQKBG44MzESBG44MzIiA0RpdgoSCgRuODMyEgRuODMzIgRBdGFuChsKA2MxMQoEbjgyNxIFbjgzNmIiB0dyZWF0ZXIKHgoFbjgzNmISBG44MzYiBENhc3QqCQoCdG8YAaABAgobCgNjMTEKBG44MjQSBW44MzhiIgdHcmVhdGVyCh4KBW44MzhiEgRuODM4IgRDYXN0KgkKAnRvGAGgAQIKFgoEbjgzMwoDYzIzEgRuODM5IgNTdWIKFgoEbjgzMwoDYzIzEgRuODQwIgNBZGQKGgoEbjgzOAoCYzASBW44NDFiIgdHcmVhdGVyCiAKBW44NDFiCgRuODM5CgRuODQwEgRuODQxIgVXaGVyZQoaCgRuODM2CgJjMBIFbjg0MmIiB0dyZWF0ZXIKIAoFbjg0MmIKBG44NDEKBG44MzMSBG44NDIiBVdoZXJlChcKBG44NDIKBG41MjQSBG44NDMiA0RpdgoWCgRuODQzCgNjMjQSBG44NDUiA01heAoWCgRuODQ1CgNjMTISBG44NDciA01pbgoaCgRuODEwCgJjMBIFbjg0OGIiB0dyZWF0ZXIKIAoFbjg0OGIKBG44NDcKBG41MzESBG44NDgiBVdoZXJlChoKBG43OTUKAmMwEgVuODQ5YiIHR3JlYXRlcgogCgVuODQ5YgoEbjc5OQoEbjg0OBIEbjg0OSIFV2hlcmUKFwoEbjg0OQoEbjI2MhIEbjg1MCIDTXVsChcKBG44NTAKBG4yNjgSBG44NTEiA011bAoXCgRuNDQ2CgRuODUxEgRuODUyIgNTdWIKEQoEbjg1MhIEbjg1MyIDU2luChEKBG44NTMSBG44NTQiA0FicwoXCgRuNDI1CgRuODU0EgRuODU1IgNNdWwKEQoEbjg1MRIEbjg1NiIDQ29zChYKA2luMQoEbjg1NhIEbjg1NyIDTXVsChEKBG44NTESBG44NTgiA1NpbgoWCgNpbjIKBG44NTgSBG44NTkiA011bAoXCgRuODU3CgRuODU5EgRuODYwIgNTdWIKEQoEbjg1MRIEbjg2MSIDQ29zChYKA2luMgoEbjg2MRIEbjg2MiIDTXVsChEKBG44NTESBG44NjMiA1NpbgoWCgNpbjEKBG44NjMSBG44NjQiA011bAoXCgRuODYyCgRuODY0EgRuODY1IgNBZGQKFwoEbjg2MAoEbjQ1MxIEbjg2NiIDTXVsChcKBG44NjUKBG40NTYSBG44NjciA011bAoXCgRuODY2CgRuODY3EgRuODY4IgNBZGQKFQoDYzQyCgNjNDMSBG44NzEiA1N1YgocCgRuODcxCgRuODY4EgVuODcyYiIHR3JlYXRlcgoeCgVuODcyYhIEbjg3MiIEQ2FzdCoJCgJ0bxgBoAECChEKBG44NTISBG44NzMiA0NvcwobCgRuODczCgNjMTESBW44NzViIgdHcmVhdGVyCh4KBW44NzViEgRuODc1IgRDYXN0KgkKAnRvGAGgAQIKGwoDYzQ0CgRuODU1EgVuODc3YiIHR3JlYXRlcgoeCgVuODc3YhIEbjg3NyIEQ2FzdCoJCgJ0bxgBoAECChsKA2M0NQoEbjQxORIFbjg3OWIiB0dyZWF0ZXIKHgoFbjg3OWISBG44NzkiBENhc3QqCQoCdG8YAaABAgoXCgRuODc3CgRuODc5EgRuODgwIgNNdWwKFwoEbjg4MAoEbjg3MhIEbjg4MSIDTXVsChcKBG44ODEKBG44NzUSBG44ODIiA011bAobCgNjMTEKBGluMTUSBW44ODRiIgdHcmVhdGVyCh4KBW44ODRiEgRuODg0IgRDYXN0KgkKAnRvGAGgAQIKGwoEbjI0MwoDYzQ2EgVuODg2YiIHR3JlYXRlcgoeCgVuODg2YhIEbjg4NiIEQ2FzdCoJCgJ0bxgBoAECChcKBG44ODQKBG44ODYSBG44ODciA011bAoWCgRuMjQzCgNjNDcSBG44ODkiA1N1YgoVCgNjNDgKA2M0NxIEbjg5MiIDU3ViChYKBG44OTIKA2MxMhIEbjg5NCIDTWF4ChcKBG44ODkKBG44OTQSBG44OTUiA0RpdgoWCgRuODk1CgNjMTESBG44OTciA01heAoWCgRuODk3CgNjMTISBG44OTkiA01pbgoXCgRuODg3CgRuNDk0EgRuOTAwIgNNYXgKFQoDYzQ5CgNjNTASBG45MDUiA1N1YgoXCgRuODk5CgRuOTA1EgRuOTA2IgNNdWwKFgoDYzUwCgRuOTA2EgRuOTA3IgNBZGQKGgoEbjkwMAoCYzASBW45MDhiIgdHcmVhdGVyCh8KBW45MDhiCgNjNDkKBG45MDcSBG45MDgiBVdoZXJlChYKBG4yMzcKA2M1MRIEbjkxMCIDTXVsChcKBG45MDgKBG45MTASBG45MTEiA01heAoWCgNjNTIKBGluNTUSBG45MTMiA1N1YgoWCgRuOTEzCgNjMTESBG45MTUiA01heAoWCgRuOTE1CgNjNTMSBG45MTciA011bAoXCgRuOTExCgRuOTE3EgRuOTE4IgNBZGQKGwoDYzU0CgRuMjMxEgVuOTIwYiIHR3JlYXRlcgoeCgVuOTIwYhIEbjkyMCIEQ2FzdCoJCgJ0bxgBoAECChYKBG45MTgKA2M1NRIEbjkyMiIDTWF4ChoKBG45MjAKAmMwEgVuOTIzYiIHR3JlYXRlcgogCgVuOTIzYgoEbjkyMgoEbjkxOBIEbjkyMyIFV2hlcmUKFwoEbjkyMwoEbjIyNRIEbjkyNCIDU3ViChYKBG45MjQKA2MxNBIEbjkyNiIDRGl2ChYKBG45MjYKA2MxMRIEbjkyOCIDTWF4ChYKBG45MjgKA2MxMhIEbjkzMCIDTWluChoKA2luMQoDYzExEgVuOTMyYiIHR3JlYXRlcgoeCgVuOTMyYhIEbjkzMiIEQ2FzdCoJCgJ0bxgBoAECChoKBG45MzIKAmMwEgVuOTM1YiIHR3JlYXRlcgoeCgVuOTM1YgoDYzEyCgNjMjQSBG45MzUiBVdoZXJlChcKBG4yMjUKBG4yNjISBG45MzYiA0RpdgoVCgNpbjIKA2MxMRIEbjkzOSIDTWluChYKA2MxMgoEbjkzORIEbjk0MCIDQWRkChcKBG45MzYKBG45NDASBG45NDEiA011bAoWCgRuOTQxCgNjMTESBG45NDMiA011bAoVCgNpbjIKA2luMhIEbjk0NiIDTXVsChYKA2MxMgoEbjk0NhIEbjk0NyIDU3ViChYKBG45NDcKA2MxMRIEbjk0OSIDTWF4ChIKBG45NDkSBG45NTAiBFNxcnQKFgoDYzEyCgRuOTUwEgRuOTUxIgNTdWIKFwoEbjkzNgoEbjk1MRIEbjk1MiIDTXVsChYKBG45NTIKA2M1NhIEbjk1NCIDTXVsChcKBG45NDMKBG45NTQSBG45NTUiA0FkZAoaCgNjMTEKA2luMhIFbjk1N2IiB0dyZWF0ZXIKHgoFbjk1N2ISBG45NTciBENhc3QqCQoCdG8YAaABAgoWCgRuMjI1CgNpbjISBG45NTgiA011bAoWCgRuOTU4CgNjNTcSBG45NjAiA011bAoXCgRuMjMxCgRuOTYwEgRuOTYxIgNBZGQKFwoEbjk2MQoEbjk1NRIEbjk2MiIDU3ViChsKA2M1OAoEbjk2MhIFbjk2NGIiB0dyZWF0ZXIKHgoFbjk2NGISBG45NjQiBENhc3QqCQoCdG8YAaABAgoXCgRuOTU3CgRuOTY0EgRuOTY1IgNNdWwKGwoDYzU4CgRuMjMxEgVuOTY3YiIHR3JlYXRlcgoeCgVuOTY3YhIEbjk2NyIEQ2FzdCoJCgJ0bxgBoAECChoKA2M1OQoDaW4yEgVuOTY5YiIHR3JlYXRlcgoeCgVuOTY5YhIEbjk2OSIEQ2FzdCoJCgJ0bxgBoAECChcKBG45NjcKBG45NjkSBG45NzAiA011bAoXCgRuOTY1CgRuOTcwEgRuOTcxIgNNYXgKGwoEbjIzMQoDYzYwEgVuOTczYiIHR3JlYXRlcgoeCgVuOTczYhIEbjk3MyIEQ2FzdCoJCgJ0bxgBoAECChoKA2luMgoDYzYxEgVuOTc1YiIHR3JlYXRlcgoeCgVuOTc1YhIEbjk3NSIEQ2FzdCoJCgJ0bxgBoAECChcKBG45NzMKBG45NzUSBG45NzYiA011bAobCgNjNjIKBG4yMjUSBW45NzhiIgdHcmVhdGVyCh4KBW45NzhiEgRuOTc4IgRDYXN0KgkKAnRvGAGgAQIKGgoDaW4yCgNjMTESBW45ODBiIgdHcmVhdGVyCh4KBW45ODBiEgRuOTgwIgRDYXN0KgkKAnRvGAGgAQIKFwoEbjk3OAoEbjk4MBIEbjk4MSIDTXVsChEKBG45MzUSBG45ODIiA05lZwoaCgRuOTc2CgJjMBIFbjk4M2IiB0dyZWF0ZXIKIAoFbjk4M2IKBG45ODIKBG44NDkSBG45ODMiBVdoZXJlChEKBG45MzUSBG45ODQiA05lZwoaCgRuOTgxCgJjMBIFbjk4NWIiB0dyZWF0ZXIKIAoFbjk4NWIKBG45ODQKBG45ODMSBG45ODUiBVdoZXJlChoKBG45NzEKAmMwEgVuOTg2YiIHR3JlYXRlcgogCgVuOTg2YgoEbjkzNQoEbjk4NRIEbjk4NiIFV2hlcmUKGgoEbjk4MQoCYzASBW45ODhiIgdHcmVhdGVyCh8KBW45ODhiCgNjMTIKBG45MzASBG45ODgiBVdoZXJlChoKA2luMgoDYzYzEgVuOTkwYiIHR3JlYXRlcgoeCgVuOTkwYhIEbjk5MCIEQ2FzdCoJCgJ0bxgBoAECChoKBG45OTAKAmMwEgVuOTkzYiIHR3JlYXRlcgoeCgVuOTkzYgoDYzEyCgNjNjQSBG45OTMiBVdoZXJlChoKBG45NzEKAmMwEgVuOTk0YiIHR3JlYXRlcgogCgVuOTk0YgoEbjk5MwoEbjk4OBIEbjk5NCIFV2hlcmUKGgoEbjg4MgoCYzASBW45OTdiIgdHcmVhdGVyCh4KBW45OTdiCgNjMTIKA2MyNBIEbjk5NyIFV2hlcmUKGQoDaW4wCgNjMTESCHplcm9fY29sIgNNdWwKGQoEbjk4NgoIemVyb19jb2wSAm8wIgNBZGQKGQoEbjk5NAoIemVyb19jb2wSAm8xIgNBZGQKGQoEbjk5NwoIemVyb19jb2wSAm8yIgNBZGQKLgoCbzAKAm8xCgJvMhILcnVsZV9hY3Rpb24iBkNvbmNhdCoLCgRheGlzGAGgAQIKJAoDb2JzCgRyX3N0CgRyX2VuCgRyX2F4EgRyX2YwIgVTbGljZQoeCgtydWxlX2FjdGlvbgoEcl9oaRIEcl9jMSIDTWluChcKBHJfYzEKBHJfbG8SBHJfcmMiA01heAomCgRyX2YwCgRyX3JjEgNyX3oiBkNvbmNhdCoLCgRheGlzGAGgAQIKLAoDcl96CgRyX1cxCgRyX2IxEgRyX2cxIgRHZW1tKg0KBnRyYW5zQhgBoAECChEKBHJfZzESA3JfaCIEUmVsdQorCgNyX2gKBHJfVzIKBHJfYjISA3JfZCIER2VtbSoNCgZ0cmFuc0IYAaABAgoYCgRyX3JjCgNyX2QSBmFjdGlvbiIDQWRkEgRydWxlKp0CCCIQB0IEZ2lkeEqQAgAAAAAAAAAAAQAAAAAAAAACAAAAAAAAAAMAAAAAAAAABAAAAAAAAAAFAAAAAAAAAAgAAAAAAAAACQAAAAAAAAAKAAAAAAAAAAsAAAAAAAAADAAAAAAAAAAOAAAAAAAAAA8AAAAAAAAAEAAAAAAAAAATAAAAAAAAABQAAAAAAAAAFQAAAAAAAAAXAAAAAAAAABgAAAAAAAAAGQAAAAAAAAA0AAAAAAAAADcAAAAAAAAAOAAAAAAAAAA5AAAAAAAAADoAAAAAAAAAQgAAAAAAAABDAAAAAAAAAEQAAAAAAAAASwAAAAAAAABMAAAAAAAAAE0AAAAAAAAATwAAAAAAAABQAAAAAAAAAFEAAAAAAAAAKg4IARABQgJjMEoEAAAAPyoOCAEQAUICYzFKBAAAekMqDggBEAFCAmMySgQAAKBAKg4IARABQgJjM0oEAAC0QioOCAEQAUICYzRKBAAAFkQqDggBEAFCAmM1SgQAAHpEKg4IARABQgJjNkoEAACAQCoOCAEQAUICYzdKBHStSz4qDggBEAFCAmM4SgTNzExAKg4IARABQgJjOUoEAACgQSoPCAEQAUIDYzEwSgQAAGZDKg8IARABQgNjMTFKBAAAAAAqDwgBEAFCA2MxMkoEAACAPyoPCAEQAUIDYzEzSgQK16M8Kg8IARABQgNjMTRKBAAAIEEqDwgBEAFCA2MxNUoEAADwQioPCAEQAUIDYzE2SgQAAEBBKg8IARABQgNjMTdKBBWPHD8qDwgBEAFCA2MxOEoEAADIQyoPCAEQAUIDYzE5SgTafEU+Kg8IARABQgNjMjBKBAAAwEAqDwgBEAFCA2MyMUoEbxKDOioPCAEQAUIDYzIySgRfcIkwKg8IARABQgNjMjNKBNsPSUAqDwgBEAFCA2MyNEoEAACAvyoPCAEQAUIDYzI1SgQqd2Q/Kg8IARABQgNjMjZKBDKmfEIqDwgBEAFCA2MyN0oECphgPyoPCAEQAUIDYzI4SgQYiuZCKg8IARABQgNjMjlKBAAAesQqDwgBEAFCA2MzMEoEuWl5PyoPCAEQAUIDYzMxSgS9N4Y1Kg8IARABQgNjMzJKBKuqKj0qDwgBEAFCA2MzM0oEAAAAQCoPCAEQAUIDYzM0SgQAAEA/Kg8IARABQgNjMzVKBAAAyEIqDwgBEAFCA2MzNkoEgQsFPyoPCAEQAUIDYzM3SgQdKWlBKg8IARABQgNjMzhKBLyVNkIqDwgBEAFCA2MzOUoEInchPyoPCAEQAUIDYzQwSgQhmIRDKg8IARABQgNjNDFKBE42wkIqDwgBEAFCA2M0MkoE9P00PyoPCAEQAUIDYzQzSgSSvzg7Kg8IARABQgNjNDRKBLRI5UAqDwgBEAFCA2M0NUoE6zAJPyoPCAEQAUIDYzQ2SgT7p65DKg8IARABQgNjNDdKBIaXqUMqDwgBEAFCA2M0OEoE3qv/QyoPCAEQAUIDYzQ5SgScFjtDKg8IARABQgNjNTBKBAblJEIqDwgBEAFCA2M1MUoE0PVkPioPCAEQAUIDYzUySgTi2Lc+Kg8IARABQgNjNTNKBCLSEEMqDwgBEAFCA2M1NEoEC3ugQioPCAEQAUIDYzU1SgRrLuFBKg8IARABQgNjNTZKBCevkj8qDwgBEAFCA2M1N0oEoGSDPioPCAEQAUIDYzU4SgQruuBBKg8IARABQgNjNTlKBM3MTD4qDwgBEAFCA2M2MEoEo0QGRCoPCAEQAUIDYzYxSgTNzMw9Kg8IARABQgNjNjJKBAJjlUEqDwgBEAFCA2M2M0oEzczMvSoPCAEQAUIDYzY0SgRuyBY+KhQIARAHQgRyX3N0SggAAAAAAAAAACoUCAEQB0IEcl9lbkoIOAAAAAAAAAAqFAgBEAdCBHJfYXhKCAEAAAAAAAAAKhgIAxABQgRyX2hpSgwAAIA/AACAPwAAgD8qGAgDEAFCBHJfbG9KDAAAgL8AAAAAAACAvyqPdghACDsQAUIEcl9XMUqAdhsehD5XP4I+7WQpPIObL751vO29rSdHPfbcZ73e3JA+TxWhv6R7yj9Sykw9i1VwPh+66r370+287GMwv0KwOz2SPC8+bXTQvR0lc74gh6G/+lECwBK1PT5IJx49kEGrPuimZb/okKG+PdecPXoHEMDBIlE8FGJ2vKZ2sT4ISJe/ypu6vVcmCD8P6Vc9/EANPhgWpD3JTLW+jW+VPcorwT4Qj+m/gk3nOPyJaj0gC5a/Y8wxP04PKj4/iSG+vdUovytupj8s+58++IhlPoie4L7DsQA+YpmJvj596j0wNLe+4U4wPXeDBLyqaQC+p0EIvq8MgT2Paim98nfovo/wMb6Z6b+9qBpyvvVxWD7b2zw/TA6gv+zhXj6+tV+9eP4kPmt5Pb4f/60+7PJVPgSALT4h3RA9F+Xgvg7aDr54/8O+dHu0O1vTmr5A0Py/yGhEvmOIXj4Dwb4+RkQ5vgaTBb943M69i2TIPuS5Uj+WoGu//7aHPJdN8DqUoANAJrd6wFZvUT2Y2V+9uH3rv2baPj9F5em9DbWXPrp48j09bow/C9lbPctBmj52QXO/uCejv7nbjz71shO/DxQAviFrnD0rEmI+vSOiPn+Hbz5TPoI9dBUSvkQQAj7OAq69/i6yvdXLKj7OZOW+K7kIv5eBoD5tpUY/PXqbvlFZ0r+WixG+7ok6PkPS1zwtSJW9SWkTvbWWl75933U+F8Q6va6dVr5OWJe+zmGyv8o7Dz/a1GK+KhywvnI43D+jLWe+JnyWvjzP8r1hGcK+oeHLvy3w3z7CeEk9klaPv+r5ID92NK6+T127vcSBQb9GF4c/WNfkPjKEdb0+LRc/RGxPwGzcKz5XnMg+DCOTvpY58L5hHyM+hDxjPg8IST9eDyNA+5+CvuLpgb5184G+UQSTPlIbhj7+6AE+JvzCviGKITyg6ig+L6lkvqmeAL4lyhM++nWvPu8z370WiYE+EP3Uvf3UAz+fgJE+zE4jvhTzub/i8Aq+UsHAvgXlmj2e/zy+gerUPyPeDz41uIO+At69PeLaOb+ivaQ/7ESUPqrXhz307z++si44QJR+F0BWrbG+x5vcvdndqL4BJBa/mJaQvYk+lD6VnIS/smXpvlgQ6D4In9O7ivm0v/J+dL4BxdU8fHknv3Lcs7z6o4G/hDaAPKIh7j2Y+HC/fpc6P7txJr2lcDi+iPngv4kQlb6JMIy+wchyPaedML4tK2K+aAnavTZbMTxTEHe+a1vFviSGqr3WR8c+yI7evpgTsz7eRYk+LLeyvfW3Kzwxlvu88oYnvnJWQb7yPqw/fvxNv8G8mr7Gr4U+6r5SPRP+uTt8P5y/+qRsvkJAmD2+TlC+vIdVvpxZVr9BBGY+lpSSPv0uSz3W7po/QAiGv/rcaz61JM2+qlLnvxjWHr+SWQC/uJcUvtVzlz6mwIY/4vkBvQ0tDz1CNw4/2Y8IwEKNHD+umGy+yDyOvm6CPT9WMo6+wHmGvlja/r1jvDdAjDhVvdO+Fj0JVP2+HRYfv/T+3D0mtky+ALmYvr9ZBL86gcq9Kc2GvkGz0T1Vnta90MmDPqnhsTwh0Fk+ays9vspiCT42lc0+glmxPj+7Bb43uc49B8kwvhhXAb9ic4++QCNJPST9YD3tj6G94sqOvgQvQj/vrDw+bSLaPoxPT711MPc9ik0BwJBcvj5Eb8a+iCXgvaihEcCWe68/1atEvjO9vb58CQbAi0eFwHQukjsm8pW+heqEvyjLwL582xk9D3bFva2KKj8EXEK97jxBPuQVZjzOLfI9YVQ4P8c8+T5SHrc+TpXMv5Pg0j+s50Q+g5yKPuYmL8Crr0i/1CFhPi834b2qlrA9mZ2JPBR4Bb+H+78+DhJpvebJEb+ZWhc+o+R1PuoHKz645Da+9oyRPukYPb2xE/29sMqxPcFCeD6ITcU+GL3+vtS7wT/8jaA+vS6NPq3+ij69CyY+mm/Gv1klGb0T+c8+qw/rvRbYZb32HKM/9j9Qv+FpHb5R74C+K0WtP4qbvb8wxIw9wZCuPjNQPkDfHqy/bps8PkIu273riSy/HCnpP2A0jz2ps0o+MmoPv0YS7z41zYW9AyISPnbWYr7+ZGC/gbGfvv1A4D3jEba/AXO4v/Ilhz6YTQy+ZO5wv7z2iL8+Mr0+uqG/vrYQZbu0NbU9ScGRvij2bD5xbFC8XudKvznJqz5ntN6+5qJePc/6Ub5B6b8+hPgHP11Pnr5z7oA+nsSjPiVAkL0ruxw+hgJ6PxtXST4iMgE+P+iQvvUXKz569YO/ud1bPp1U8T6DA3k+qLi7PeBrIEDjiiS+/ebXPWKXKj1cYkm+qsBHvzosNr7V+Ou+PiZKvXTVIb866h4+0oOgPrAayj9y3EQ/iiCbPXgyqD1OQ70+TgOhP7ABnz0dOgi+cO8EvpZegL9HTtG9JgGYvsxNL7/fw/w/0l6gPo4Abj0wddk/Y3YLP/2MET53BxQ+PnRRvpjK9j3tafe9z4OBvufOub30z8e+jjIgPrpm3b34Tv2948G8vrx2nz7ijFW8SwWAPlhOxr423CI+xjCBvquMl74Y+yQ/GmzOPRamCr3oTQY/iI9yPTC4rLyzjNM9SMVtPcaN/r3Rglq+iCVMv98FwD9sBtw83ZA6vqEkX74+Vus/oj85vu/7Tj1M/PI+rGnvvq9NnT4CmC++Ab/Pv9rCOr/e4Uk+7DOjvrLsCUDsWAw/6hsOvjWGqb0ioBrAZl/KP6ojrr7fkmC+cnXKvqrTEz/rAuM74YonPtZ4Kb+5kxtADH6qvUCwdL58gTQ+10HLPonAH745Vte9fuK/PoGeZL6mYLK+kGcyvpg9fz32A5+9rH6oPjuzVr2s1hC/pMe9PWPLkD6kZ/w8jNdHvaqqor0eoHY+E2y8voWaYD6pDK68UgxqP04Iwj4pgza+btHQPTQzaz5lxAW/vxxPvxamPb7C0/m+9teJP4sbaT+rK8M9Lp1jPpDxjL73XIM+ZbZcvgV75T6uYY++SHrrPgWglr75DIQ9afwxv2kSez+lEE0++lfyvnh6Dz+qiuY/LI30PRVMqD6A9fI+Qjjsv4NsC76vMpG+upfivWOHsD0xdI4+syXGvZyPhj25QkO+T960vikLkz4exbG+xNUfvyrhCL8u5Ky8eB+dvk7BMLurSva93f2cvg0Gfb7vXCs9Z0Ervs7g3T61iJa+09PbvwL+zr5CFt4+LvfCPsgx+r1qSS+/OIeNPpJDPzruSzw+6JUFvX3arL8gV4o/JHrmvvcGkrxMrxe/hSSGPsr5vT7zes69kJY2PmpS9r3y5hE+fCOCPrRp6j9OqrK/TwjKvaGnzT0NHA2/jiJnvzpS+T5/icQ9CkcuPT73nb24XI4+uvgHv6flBL7vxBi+4MbAvbibYL4RQOU+6IxXv2+M/byVGrA8+BhgvoZ3RT5BijU+gvl4PS/TUr7RbYy++9mKvsjQRL5dfSk+oM4MPrneLb4gR9m+oJuFvXBalb0kKHc9oa8YvqK5+L+ww7Q/PBWwvhGjxrwpeJG+ZxMTvcBTCT0Zcpq+KZ65PU26tT3VHQo9SjBoPxgGdL/HqQ++pUv7PXdHpb7B+R6/2RufPp/nXL2VKnS+j8COPxdLzb6/gLq9odp6v32Jsj76x5+90bGEPl/hmz9LrJo/HVXwvVebrb0Bg3Q+61IzwB35Wb6eioc++NXcPzYOoj4T3pm8AjK0PgMSPb6oMQDAXEbbvo1zjzz5V92916QbvjCK076wKfO+xZIPv04wlz2Io2O+/tAZvdBUJ78fSiI/8KZrPdD0Cj6LocI+c0s5Pm4mtL7Z3zU+PgYsPriNWD8rhfe8UadjPZDVVz4Dxoo+GWs2v4YvmD5Xq9K+2Y+DvvuTx74iH0O/XALJP4J7iD6P+Ho+X6p5vgDh9D5KW0K+clKqvudPvT5E69M+vXUGv6HYBz7uAWa/U2sWv+52Er17qpK+nM8nvynyHEAUR3U98rUyPqw3Db+PC6q/EfRMPS7NS74OBvA+F6EEQOSGbD61arC9VvnMvoYHUj9Nu9i+EtAUvcyTzD22AO++whIHvTKL1r1MrIE8RQuqvk60ez74OGI93wXMPi+U273mm7S+CU6kPh2fsT4p6k4+eiyAvp4YnL0ZI4C+Sv9Xvjky6L5ZSeG976/vvlXDoT3bVJc+YcYMPoRfgT37B56+jXsNvppWyb9PERtAcOa1vY7LVz7KltW+tKe5PuYwAT7u7u0+9sOGPx/iTL/gtT6+8mJPvie20T+K+z0/+clnvhBq0722uifArMtWvz4npD4QmUc+65rov5vRjr5Jg0A+sInPPgopDb/m9ro+GvsFPXV/Lb3o5Fc+yCJePwMtHL4BPiQ8oXWJvnEMVr17hv067TNuvUbi871edr49jb1hPS3ISD460FI+AyNAu2ibqz2N4Gi9UaaqvU8HIb47PCQ90FwfPUDLhr8l6+C/knkyPSfW27wHjB0+Ho6FPtMrEb+kBTs+X6PuPdOqDj6MnKE+7lWTP+1ajz8pEdO+DUWCPXjTVMAjuJe7MtRNvpiY2T02nkFAntR9v9b7iz4JEN49DKejP2NukD8HzN6+NGlCPxEaaL3JDU++hGQIPrkryD7OpVk/AFbCv2VhET9ok9k9xIQ+QAip/D7UkoO+MQykPZnFub4PYMy8JnUgPYq5sT3kfki9vSGVvdpn/D2LP0u+xU39vRYQ6LyRSFc+/Xg1PGQpnz78gcg++6gkPd3kx73rQvc+iMXMPYtAG74IKpu+alC/vhRDmj6DJb+8HINTvf8CTr8ahPy8ENAfP03KEL545c48VL+IvilgLDtxkV0/Wmy7v664zr0D1VG+Vu2HPhyfeLwzpIe+zftgPny9mr+wkKu/Q+zYva4LQj64WeO/1/sgvzxmhb50+ZY+mucgP57smr/SfAM/z5POvkXynb6m+RxAz8pjvSe6RT4c926/Nn/Lv2zdhD1tzlO+0bimvg1lv78YJri+ofJ7vq6ePT5SeSo+EiAZPEpjmb6SnQU9gxNiPsgqBb4yuGQ99pF9PmemRD3P1wK81y+jPahh1Lp1HSG+oQsVv1z2vL0bFO0+RlSIv/IXiz74eJS+nZzxvfCooz4aVdm+VoA/PiK5a75bb1y+dLkLPj5Ve78eTVM/EaofvkW9Jz5Nssk+/E4Uv7yBKz78fkU+y3NyPhhOw70OIDW+lU4MP2iHnb8RZfE/+l5xvpOpCz8dLg6/SR6TvyppD75SMr0+dx22P3R1Vj/U0oA+RgoLvmu4vb5Q/52/VBELP+B9tj2PIwpAg6WBPdFLuj3xKag99mlvvpVlmb4bfs09COZ2vo9E4b0WiG6+GuGoPq2qDr7YcEe9MlUCv4W9tb0i3Gk+C02ivoaOE76C8HE+en/uvfpNAj52Yd6/sZYOvxQY1L6Zkxe/4HyiPkLhDb+RU8E9776XvaZOyj5Zvya9exKVv1Ixjz+crcE99b2wviZFi75+Lqs/YeAePC5ikD2078O/dm5Bv5GCKD+RCJ++YCTuvIp/Cz/aBqs+Dz2tvm8a4L9dFrW+PPBovbcpxr5M/5W+NO9GwG2bVb1q5Fi+1inzPYA6Q7+loZi9BJxSvaccmL+6cR2/M8ZKvq6DF7665J0+MYqAvkgkKr6YwDM+W4KUPDjwGD5A20G+nHaaPom37b7xZ948J+LpPaD3Ez/5SIk9TLVcPcL7Er7g5wi+cymFP+Rjxb6RVlQ9zVkMv70akj5/KqM+gMYnv1Zrk72YyyY/vLbKPS2zdr7lJZy/vj1Iv98a1D2ExZE+2gS4v8Xr1D8CSxm9C68tvTlv7D9z684+fLC6vVnz0z1048o9AGvnv2KDwb3ulBq9WhNKv3Jp+7zevTC+8NZAvtzKhD4qbJw+O2WjvH3iBT8UHGu+ENj1P/oFUr4/MtS9AeN7P6qV+D+mhns+lWCvPFo9JD+G+pS9pn29va19kD44+sQ9bOOfPo6lnL3yAB0/6Is5PrjqWj0nYDk+7WUyPgx3q7065s++ZPwsPYZAPz7UOhq/XGMDwDT5Lz5Wpg4+96eRPk59er0CEg0+L08FP/u/Hj2vh5W+21dKPd5tBb5i/ka9pQWIPgSsnD5XXVY+/gdQv2yLnD7EmaG8Ih4BQCJpjb9NpeI918TGPG2KPcCuv9K/7dPpPuSHc76KdrA//wutvw9oEr+6ad++40exv7j1X78A9Ke+6Ruqu/zlVr+Rqa8/Cipbvr4bYz6Hf4s/tI7hv6ihsr4oikc+JEmavljEgr1Yg00+HcIiPZMozDwB94y7Gj++vtNyyD4vJRs/RV/IvvrKsj4BgQO/BRRaPDrZnb7CZ5w+DT3KPa4YBb6tOSS/ylcAvnpOtryt2xO+bB9fPdVTCkDBvWc+i714vi1i+r6Haly+dw9eP+UP5z7R4qy+Vk0hPhTjDsAmNJe/4tniPQeqS70GkP6+rIQRvVWiOr1ITby8q0OSvrl1iT/mZC87M7BdvoC2Wr4/PCM/cUnjPVj3mL7bTCZAYb6RP6P3972VRji+SyiJPqiqJT/8UT29sE54vrAzoD6QiAFAwYMqPjGRnz7Nic29fGl1vBV/or5sKki/8PcUPsUoZj0cABO+p+gWP135+T2jmba+f76GPrQc0jvvm/49myrWvaHKk71wlZ49WFenv1C4n79ItgC+SHHwPbP3Ob4ct3Q+ly4ZP0RJIjyfjbi9KsayPqXKCT7PVcY/1tMRQBksVr5yhrm7eKHwvwIdJL/bs8k+QRZQPhswJr+R8JW9uX8uPkX0sj49Hmu/Wb+XvpjgdT7LDZ+9seJVQCLCob9B1cM8GLaYPgm98z/gOmW+I2SaPldUpb7Rg/G/JvwIv6oTvr60xd++zC/DP3IG1b9ut989RZeevSlRrT5aZQO/U1lJPoQApD75mDW+ZhDXvqBbIruHlYQ+59ZMPZuNTz5RNOW6b8aBvpAMZ76YakU8g0ewPqMv2T1GT60/g11EwEVzTD58e1a+4ketPWl5rz7wiBw/LKkdPbyYAL6gTsc+EtpWPodBKr7S7MQ8DW62vT+dRj7LMG8/gFyIvgG2jb7YrV++ZTOPP+QNpr5hkIo+0mtqPg+GfD5iQS+/dcOhPupt7r7EMxe+Aa6GPwPB173iymC+CdEQQMOr7j9q36K9oZy+PWxSnj9knNK/0PoqPhlolj1SQu4/4KIFwODBCD6tIIw8frKLPexm3r6Uj/s9skMJPlbsoT53YJq+vv3uPic1yz6EUzY8D5dIPvELnT5TqtA+J2UdP+RTpT0flgi+MgOLvsH80z5Y9Iw/2soWvjsFIT09n10+5UMCP+Y6LsDAKrQ+y4xKvZlXGj7k/0w+i92Ev0wErb+QLgW+hKEBv0IGXD/pB8M9jGSQvqhYg75WbrG/vB5JPlpsiD60bVK+BzkRwCyknD/sxEm+GrkDvxIRnLzc71y/BQlVvv9MEr40+S7AajQavzvrLD+Rjw6+tc4EP3GAdL5q77U+jtDfvUuTnz6b1ui/YgpwvsVJLr4Of0Q/daI4PcfFvb4BqNo9RnWVvVoa/73OJw4+xIemPgN+IL50xVM+NGoCvctlPz4GdOu9ErvfvbTtdb6bPf0+EMXSPj/ebL6tono9I2BoPk8lcj0EuM8+PqIuwMFUqr4bI3K+6flkvuphCz3UyC6+MvlUv5A1xL67e/68AkRAv+5oCz1Lrs2+/e8uvt3zWD8WgyS/Xs02vDwaLz0qd56/LBIiPfqEoD4WjdK+Vxrev6/FqL+lLDa9P2ZGv6z6Wz/XRUhApU8KPT5RgD6AR2K/wvjHPry7b76a3Re+mum4P/o+T78Hkkg+kwJqvsY/ij1RCJg+kdVkPXyE5D054Ng+ye6qvkYYpL3LPaS+hzlsPa6x2b5Lu2c9vqHfvc5COr62x42+DhABPol7gz6hKSw/Pt6bP60YADwcs989N3qaPVVARr79Ynq/3umBvlyv3rwDzCU+EpWkvjoalr+AJtU9ndrkveJngz6apSq+hqRUv4n8FT4rPcw+4fKXQEi3OkCsVYM+WFS3PIAdz79qGYg/sZ8JP4z+BD6Aeug/3gmsvjrZ2z2nAcW7Yqn3vnw3Wz6qp8c+ch53PZ9M3T+pX9w/gNNsvTD3Sr70/Rw+KFAdPwhEzD6UBEO+5HR8vtVkWr5bGtu+OM0wvkt8vD1a1nk+2t+dvewbBj9NQao+E4YyvZHLxz5BCdc9AltYPuuazL0ZYLk9v0CaPk0ts79/Jos/VeqvvoMdKz7AP26+0KU4vvj7DsB4TPm97OYIPY7NLz699kI9uJQSQAD70j+KJ528Z5Z7Ptx1sD2O/9S/ofckPfcxiz5GvgS/0HxDvheScr0RycM+XxakPwBBor8Y6x8+L8UdvgC7fb+Ez1k/L3MJv41kOz6iNKw/bVTPPQbZMz2KJFs+BX/Av5M607+pfxG+2DfNvCvNSj4ra/A+CIR+Ples0D7VIr89l7g7vg8Btzvfm7A+iB2mvj76O75AqIk+KoirvXoFxr3Sz3u+/X0JP69VqT3/YsS+9C1XPk6KDjp/UD6+r7C8P9F7Kb8uqe09hbksvqtZKj5dTAO/cWC1v4SQCj2SaJM9Q5vXvbl5nT7EHFU/GY8/wOMflr1E1BS+0m8jvyX2lD+irgY/Mi4CvgGyhb+2PBc/MkZOPp5umb40Lqs+i+i3PxV2Zj57k1q+LhajP2IERj+uboW+jmQBP4xhjL8H3CS/F8Hvvb9vY712MOM/I0Xcv1ppV75KU5i+CAZoPuQL77+hoWq+44MRveevgz2fzkS+OWbMPeqyor4bpgI/lzkCPzw3UD5kNck+QO/1vlbAq76BIYm+VqS6vghfqb6P6au9+AQ3vS304j4dhjw/TLSav4Emxrtnd1S9sOkePbbMibxtc3U/bv4cvkOX4L6j88i+/QeFvpVLmz8B1QS/zYidu9Tmez5dg7m+FYCdPqc6Mr4LW729AsgXQIQCJkCXmaY+MqurPmCxgT+Ruck/JqrAvuZFRb23y4A/5stOP3kdyr7+KRW+GpMgP594Hr//55w+dk9NPm10kL/wiqS/56ehvSlyUb7spi+/uglcv9Ckv77XIdW8OWwKvpXozD5hdOM+NNI/Pe7QEz7XBJI+v9evPYUpWr3fKQW/9wGpvkngTb7JJoS9GwquvRbo3r53mp29T+JpPSOVdT9VNfI/qi8YPupbDb5SgCO9mLmwPrU4hz6tW1k9NUSJPZueF74hlWy+CJj5vIWqlT9c3+S9gM/+PhjsXEB+2gS/x1XuvpxDEj7Qbbi/Cb1tPssr071JHGO+MF6OvyGRMr9/xXw+EEsGvrsAHcDW+SXAWdizvT6xn70m7Cu/rjMCwFp/Tr0k2C69dSkxPxLbsT8OaIK+N8bFvPUshL+IBDU+2Ui/voGNFL5ypHg+X7DoPYTuUr31XJ2+8RkqvQQ4vT5AAku+tYflPrUQsb0HRrY+5I+Dvfs5sT7wynO9eqHKvLrHp75ZCl4+VPk1PyPIB75WvRA+eWTpPnzzDL2C4yw+sTsOwI5JWD5K/0s+ILWOPvtVur03bwu7XR9iPlcV+7vgr5G+/TEWvgEri783YQy/70Q7vaaNNj031iBAxxU7vbV+hz6iLhs/gfIzP+WrELwh19m9DNKIv5a0CL/pibu99mkPPl70vz+GWRK/bFehPWzm3D7rS5Y/ogizv6BQFz4hqC4+K1LTvhFpuL84wTg+ythDvYdCkz6inn683mn7PrjAJ75PW9y+jyPJPQChtz66p3W+P++MPi++Dz/5lTS+DPQ0vGCuAL71e9495hc2PnWRSL7llZO+FBdqPRXgOj5AOK67T4osvrjZZj4/7++/C6FHvvf3p77VMSa+AEuGu8kz6T02B/++69vpvn48xj3mHlW+dFQ9QLvH0D47vhQ9HSQpP/qzzz7JBpM+/+SKPqhcsj21TNC/UepGPppS/D4SYt8/TCJBv6lCVr5fSaW+zXh6vx08Fr90/5y+wUb6vX7FsL7QFdE/Y6PJvubD2L3yGmE+sNngP7PVST5ijKc+qhr7PgB5Qj7IsUQ+Tc0yviAAmL0dlri8XYp4vpwEFj1ZdgM+zWDYPdkWkL0Z8Yu9oIyGvc1Qfz3cwRu/Vk2XOtP0uD/ZKDO/8RZjPmWeAD3Ts2q+jjq/PW46nT909jQ+jqA5PoVinL5gkoa+2OuQPrXBq76Zfve7SpztPExsTj86e+2/o7obPla9uL7+E90/57ZQv/3TOr6eSLC8iluZv63UZz9GSSi+ROU3PubFfT5vfh4/UhvrvXn1Pb6xFro/9z3AvyAy872ARzW+BbHQv9lQ2j+hEvI+CZw4v7Q9y70jEtk/6Ew0PqQyvb5/64K+ypJRPhTa+73Y3Jc+lzvBPR/ioD4nWhe9IbO3PYHLFD97ZOE8yArBPne4ID4ouny9wCd2vv1zDDzNhw69SIgRQHghEz/m3TI+ShGMPsnbGj5mIZM+msPkP59Cxz1JTZy+9ZRGPiztpD1EFQ6/mq4vPe7MoD5srZy+8wJlvht3AD+Q958+RjAsvSy3gT8B1J4/SK5bPXv+kr7KPBVAIMV5P9zBFL1A5ww++cDNvqu3r79A4Jg+PXrWPTBIl7+cEbS/NbWSPgpvRj7QGRc/o6JCP6ZPMb0/aRc9B8KLvyOxeb5TY4I+u17ZvbxBET4UZxu+RZ8OvXUChD4gVOa+aQaAvvwHPT5np+Y8PTSVvqgrpb3mcGO+AvK8PmFcO7zO2RC9K2KjPtT/iT5MBgc/hhihP3opmD6XLDo/f7K1PRp51j1AgD09wzIUPh6OobxU0cE+cLi8vkkhw7+8UMc/TyohPiS8Rr1dqg4/VmKKv0M/rT7IWFI+motevzSzUD8xQFI+Fd5pvTLi/L8Gol8/AGbiviLlMb1HNj2/gseqv2IjUT3f7zC+DSjFP5j9H77TjYc95+49PkiQO75im8A/HQtCPk1vKz51moE9ERguv7bnjL0r6Nk+17GFvub/Ab7vKVM+xIQWPrb1q77gejc9tsTDPQG07725RYw+Bz4OPhvQGT6dbgs9XvkAv76AlL3FFjq+O2DHu4mIFD5LdoA/SpYrvqvh7b1yG9s9rGHhPkOK9L7EamM9kjIKPvK/6D7zrww8OWfdvp89mj+IFiS+UubCPj2+qr9xZna/i0jWPOaYjrz5822+Z3Sxv/JRkz6dhT2+eY1QvpATfr7xT2I9c/0aPpGvUj+PTIU/18aFPSrLpL1KCXi/3KoDwPZ+OL6m3em+SOpFQP6SpL48eoo9RNoivhcKtz8cbNq/Z88VvmZabL2WI5G+ra/dvTemnrx2kbc+5xg5PuZhkD6l5Mm9HEA7vqK09b3gY0W+DOaEvkV9Dr1EFMq9eJtNvmLmF75BfYg9elCxv5JAQj+3AHY9F92ePZ5Ner5rVJW9cNHVvmv8Ub69ka++hgqvPbfHhL37fb29hqoRvof0e74BE8E+MBjKP+xVlj/2JiC+9G2+vci9gD7TXCu/kYueveDBhz52aga/PNUEPUDVRL7ti3Q+E5QuP3uETr+0KXG9wY+wvKKRVr95soI/5NI4Pdkpob5yADq/z1Asv6wnUr7B9q++XvzrvnPY9T/GNYs+TxMIPos25r6Rg9W+n1CdPZzGVj4Cp9c90H6WvZdf8r06sfU+NpcGv1qChD1yYb2+ftKMPjq5nL6wIUC+NXrNvpTH3r4OHEM9G1C9vqXFuryXLZe+tcDyvRy7Hz6R7JS9sQ2lPgCpRb5VeqK+wtmYPeDD4z+c58G+cgbavslRCD4a/Ic+upnLPdRMFL8/aig+e2Vtv6dBmbsIe1o+pLDWvlBelr+VB4C/Hp1BPsvlkD4QRJU+Xa/OP3QzVL22EV+8MVWTPli+AsCgGww+nKprPm5G2T9dWOs8dx8Pv6dknb1ini4/6XJLPhrTBr8pb4K+TyLyvcAwOr6KOc89DqJ6PbYiFT8UqCo+2Gm/vhIXZ73m7tO+oGXoPdumj75fRXg8NRxKvoJanj5B6Ww+yNPbPUTjEr7j1Bg/Rlhsvl3xyr5x0c88T9uKPhIRWL8YKbO+LAiuPbrJAj9Wmqu9QKtwvmbWm79ZjV++XdMpvv8bmb5ixDNA1F/aPg+/bT4h2as/MsPJPrXyMr5+rfW9EPyzP6SMB8CYRBK9jBf4vsusEL+ymMs/RNrMvRiq7L6uRrM/lpPxP332dz7Cp8U9bveDv9sniT9AQEY/2pszPvxlOj+Q/R9Am4CRPt7atL4gAPQ9SK6GPhjjDr5v8w8+EncEPv8L4D7c9Vu+pHo/vUCQxr4CGB69HZo6vL6r4b5+Sgk7jdnJPSwfsD6tv9i9qphfPqQ4rr4Be04+rEgqvrhJg76s6na+FOsPPob2Ur4iTbk9LAMDuwFzEz9N4H++HUB2P3CNDr9igIQ+l3MEwON/gT+QBw0+rHcIPqyqGcALZQU/c4gYviKbBT5X9Ag/4DKdP9kFNT6Aves9XBe0P+5QH8BqhBo9at0/vSCN3D5RXxO/4uoPPhOK+L4X4JW+gAHlv1l/gj2gcIk+840AwOq8dr7mKxu+Rn6WPlg4Ar1Nvl4+Q1bjvoBXk77vw6K92eKcPiKhwz37EgY+tEX+vCRu1bwbmbq+iD4qvp2wBT4TLEe+Nt6PPm+rSry03wVArG7SPQOTir6bFtg96fJFvuQ9Ob7DHAhAMLWBvuHPSLxbpei+4n3dPGrzsT7wJXU/Iy6XvrkDkz6pC2C/SYPAP3CUtD118pA+N5Nkv8yGUD6zSUI+E90ZPdmgxb5z1ZG/aeLvPTLpJr7bdbM/0wUGP+6ci779ySu/ZtFVvrhhR77Uqqc9e3KGve5MWr9Y1CC+rJ25PHw3Ob4oHgPAYwBTPfnLYT7bbs+9WAinO//6LL3S4pS+z4gRv7bzSz6L3jG+8gDLPfShzTyG+MO9LY4LPpYz1T5JbdM902iPvgLgmL3gLP48JXrFvh6F8L/ghH07UueQPlmryb4Wijc+7cWHPpXPKj/ydY69xtBEvRCmiLy5Z749+0F0v39Mzz8HAIK9uO4EvlFCp7+4oYc+Oad/vtGxNb5kDO496COCPlpGMj2Xcrg+LIuWv1h3iz+JG/K+MTeMPrUU4b8rIBe/T+kzPlnnzL76SgxAYoRywJy5E7/jHcy+oMAaP1ZnpL51qZQ9Xea6PN9AlT+4v+e+4Zm2PmhWgj7yIpa+h3VtvtqTEb7DeRm8UJyiPmX55z4Hspu+6SZIPUARnj5xATy+LZGGPmeQqzt87Q+8m/6gvuwU9L1UCDe9g+IJwDgneEBAdkk+d9LZvQTFGT6+QqG4qv0fvdAHBj7iVVS+m2WMPGur5jz/RaY/P8X0PwrVJT43DYk+aEGuP1TRbj+zBGk99sGtPpCUCD/TDYC/UHrGvY0TJ78UxSpAImdLvxSmlr2Nq7s+ssLJP2ZarD9fwWM+22rEPIQWEUBCYvA/63+LPkdaR74F3sI+QaorQCMjar7EYsG9vju4PW6KBj/wEw6+VhsXPc1wDj9SjRo+ahbrPr0/GL8nFYM9V7WOvIV5sb1ywJ8+dbVaPayY0D0oXfQ+Uky0vdO0SL678eU902iOPcw5tr6E9iE/6GHLP1FUDj5s+os+7LhePhJmsz6qybE/pAwRPW++pr4mU/E8YBThPatki78iJRs/ScZrPjVg7z2rCpk9RSqjP+Lw5j2CiJQ+yX7PPyuagz4l2oM+cWThvLj13L/qbhG+CF16vdM9+T6u/bc+Mk/SPyRmg75yhR66zifxPxEI7L8BAGO8cmMBPiQ6+z8ARVm/kAE/PnmsO721yRC/0uRpvzH76b20fnM+ZpGePLRbE77l7gg8PqZrPXtMUj6MIYS+LsAkPjJf0r4uBfa7xkiqvo8NxjzcFQs+TBoBPzaGuz3XTCU+xOvgPtg0ZT08wGK/8oYEv5r9/r2rMHk+/J3Ivg1I0T8S+pk+8Z4gvnRGlr3O/ea+7TuZPYDLAr+eDWw+kxucPkeKjr5CNDI/+yobviuBqz7eCoQ+JMQzu74SDb0EO46+0uDJvk6dYj+9DrA8ODmqPsUeLsBuCFo/w78mPtq2YTwmPiDAmft2vjv4Z73gQzi/6FXDvWgQjT/2aO2+h4pBvkW5+r4jVKA/WKf+vRHPKb3yld8+QECHPZ0eaz4rSPK+nZ0XO7i1Hr+lpeW93ZUkvtZRnr4YOeQ+1KmFvrdoNj4Yj3w+utm4PRZ2AT3vt++9HjSXPyDYFzuEnwm+rHdpPjGF2z2Wcqc9E6Mcv8hJhj1ZzNy+9fkdPsovBb7dNr0/36qivgnGBT7fhhm95NIovyHk4T4mKUC+Qg9tvtxLzD8WU3o/hvxbvjZEJj8OTbC+jWWpPt2ABb4ihPY9wJ7ZPsthvb/n7QU+ZLMWvvoUVT/KSoq/3gFFPhuhn71y+fA/ScA5Pv8D4j0s2sq9KHJ3vrNo8T4EFCw+F+ybvggvlT0p6gU/SgJevmPpRT4Pwuq9W8cCPymer743cAA+s26WPo8h7L7RAUI+a5kMPkRET77Mnw2+G+TCPknEI75SAIA9EV/UP5cc3j4k/BW+fKuKPt8LLT5DbfO/vd9XPO3BIj5G0Zw+9+e6vmJPIUD4e1O/owVWup9HDr4OKAw/SOwCv7yKpz1a2c8+VBnovnKSPL/xv669rXZjPisvlb2+TJ0+1ylBvjdzpD12iVE/g5+Hv8BYXT6xW16+C8KVPlv8oT+NQ2A8xk9tvsieET9Tuhs/bN/qvDQFn70sma6/c1ioP3KtqD6lNgO+oHIVvrAxYL5wJ849OrlAPqrKmD1/SOK+3u6mPu8Uob2DuhU+DZekvgpH8T0kexa+U6TTvGzvlz5ecZI9VQocviUR8j991bW/IaWKvvyRqD3cB9C9xcy1PdDVDMCz/fU9cn28Ptt/Gb6xGSy+IdXlvzTLhT+HJ1u96D3BOpY7mr7HiFY/tnL4vTOZMz56TlI/ODkbv6OGdT4EExq+9ui7PzNzmL0Rx7G9NJqXPKF8qj8x42E+nbeGvsHXAr1+OM+9JSskva96B79T+CK+C6N2P+daV73zvkc+bgSEvSgB1r8woRS+3WSDPpg2nTxGxAI+cxm7vvoPkr7uJ7a9HvwuPXIQML4YF0Y+/sRyPqchCz8dpYG9crkBvzTEwz2PcQe8V9pFPUXW7j6u+S6+TLjDv8X4i79yEmc+wtp+vvvoD76Y8ge7bsjNvicxkD1rRo6+TXACPye0nD60jjXAzPiNvUrPsT6d2wq/RJu6v5jSPD8f3E0+HukzPobSsL8EBh2+RXG+vkspGzso4qm8EPvSPi64PLus27s+9wkAwBg6Lz9Fifm8nqaGvtLesT90rEW/c04pvtiQwL3YN8y+yrB4P/QitL7O4Aw+FiBmPwB00D7S7Ca9rXixvtvh1T4WIyY+ndOBvg6wGj4AtrK+tbA2vkMBXj4yi0I+0dq8PNivMr7+k5U+RhFuPbLA3D0xypw9MEFQPiWi8D2IKUO8abAEv/g3Rbw9yNW9Ots1ulK53z4UpsI/aPgBPjvrLj4DNc89VAYcPg30b77v8Fy/EWIvvlWmbj4yi3Y/BJ2fv+YdtT4KDAg+tdWlvhGv5L9Xs0s+CnxGPgL7db8lCnk+NEWPPeN1Kj0w5Ee/zJGBv4E2ir3Hl9S930XlP/BnV77ORe49Vq6yvv4CGD5O/T4/dqkovohTYT6HDzNAltOmPjAJAz/uSsO+tuVaPioPkr4eWKi9xtk5PUsezj1ibyG+pgSRvp0gcj2BHdu+iekEPg4QrD4D1MO98PuNPiyHGzzf2/Y9DI2JPtC8Gb9FBoG/c1TVPVp/rL6UE2O7vRbrvSaV2b4Zuk09W5txvjIPFT5Db+K+tAUJQA6ysz5QEdE9DCecO6h2Q77OF/s7bzYZP8gWVD7QhCi/ls0fvxolJT5mSgc+Icghvwg/hj73bia/Veu8PoPeGL+dSg1ApTXEvBtgOr6cnhI/USzDPBp5Gz2v9Ko9zqDVv2oCSb2dSSM9qK2EvQgl078RnGg/XLFevsjc+zycZgE9YV+GPsaNwDxA3I48CpqXvv90zD2riv09lSeePhV1Fr4clxw+w4Cuu/TpCz6NH9q+GH6vvv3GUD7dMxW+RDFLv7IQN8C4OY29qomPvgFtRz1LcaI9WUybP3F4nr6Js9C+q0TIve81gb4ODJs+1oiFvun9+r0+2Mm+g3yluk3ytb12k6s9ZgGcvhR21T4C/O+7gfAVPouQgjmkPXk+SOKovxFix70k49S+sW6PvMx6UT17608+lGTEvtgeY79Ai0o/BduNvk/JDL6p50Q+LJINQM6aw74ATBK+N48AwDqfa7971DU+rjGFvmJvn7157c8+Jp+svoE5PD4Tblm+RWUmPYlI1r6ws+u+CYJxPoIQ6T1PYKq+tjdlPntsGL9+FIa9wYnyPbWsVz6knZA/fuciQPiZ3z6e0bA9w2uMPqQgT75hc1u/PbqVPmpAHr6gsG09x3Cfvshqxb9lb5A+8wDCOzWhJz6U1rS+DftmP11Ouz6wk3i8mmHUvpwajr+cwoA8W5POPGIHZb+Q9OU/FgoYvq+GED6zvNG/xJKHvqLCaD5vp1o91ZYawB5Avb5T94Y70hcxvoLlnj7HV4I/wpnHvh41ar41m/K+bqcLPxh0gz5DJbQ+W6IqvtL5S73UCYU+j/CcviNfHr0Jy7i8mooSPgWcxLybtbS++W6HPqQokT51RRe+LPxavvZ54z46eQC+BwVIPvxa071S/i++lIE+Ph7Amj7cgpC+P7vuvQot/z7pvFm+A33avmd2oz5YGMc+rr0Fv/WV5j0YUom9DgxWPsNSIkBWl1C+t6IOvkNiXT5Oiq0/znafvraSWj4qWSg+JaYyPmcPyb/2zp871naBPnhG472LWylA8aWBvQ1xY76qEXQ/NNYxvzhZgD7V2rs+l9rnPFihx79Hkem8oFVBvkj3A8DzXNG/h/mIvrkSbT1Wfks+8bULP6Hgvb5D2Xc+Nx6svkTsor1OGQg+dXoqvN3ENr6Tp0Y+Ty9avnRNWz5N/Tq+AK0lPzlu6L1iBdC7qgS9vhvmJ7+19EE+fXGdvZEm8T4gqCu+ZhNuv2DvlD68RiE+Gf5fvisFr77MxUO/uhXFv6uLDz08TZU+VOaOP3onfD57W40+aMJHvTd28T+n7/4/c1xVPoZg870TGxG/IgJfPqInMT9ejrM97r7Ev86Ror/qCcM+2KjQvWix7T5LgDa/MngVvrzMrLunreW+oS2+v9H7ODwsQDI+YRjXPmmOnz/t+Hs7d9efvKRDIb4XzXQ+RDv3PWTho7zTzyI9XENePrkhOD4bpIk9g6lgPftSk76yDUQ+EpPevj+u8D3p8xc+VykovqPIkLlOFiE/7sOMPgler7wUYYI+I4jAPa4+Lr5WBta/Sj4tPkiobDyfgus8oKl7PtqME0BD14+/KlNkvicyAD/WBrc+nK1bv/jnjb7opxS+FosJwH3MAD/SBeS8u8HWPUocuD/IRlXAli8VviJzwb50Y0w/jVS0vsudG70PZgY9Mn9FQD1ZHj9k4ZC+TmdnPAy7ir52KSw//8G5PnoH4j5Zf46/OFGCvyfZiD4ZOsU+KJXkvlR86D2ZZiO+BXw/Pq6Fvz5dCm6+oqIuvgyt7D4sYQO+eYUvPXhNCj6UEX2+gVQNPQIghT09UyE+a5U4vpypvr896Wu++Nl7PpcgwL0pMUq96Io9PJKCE8BpBbU+OYzVPHZhFT8g37S9T+3fP6BEvj2wKD2+Nfmfvpuy3zzX6ko/dML8vQ2ktb7E9Xm/HuduPviQhD2Noeq9JmATvZAfDcD+9ci8rj7HvkE+jj/yLLQ+y8uoPRvffz5+o8U+hnqdvzIwmj46lz8+k3q2vaSPVEAQDbk8CXo/PiV8qr23BBW/1gkCPpFRiz2y5jS/rqChPtRd2z6DVb8+r94PPkVKkz5MSA0+mnLCPb1CFj9dpis+8X4APXA+J77UoFi+D3hnPjkQ0T0QlXm9jbAHv37LD8BMo8I7IJWJPp6kej7WfKU9Zhwqv0Lnnr1AsjG+MxYmvY3sI71UrtU/mdaOv68M1r5vDpC+YXBjv21/V7+fDfa9WxlTPdWBMr2TZ+W/TZhTPMRRtb3YfPg+638Gv8+4sb7mY2E+hX8LQHwY/T8Pp8C++4bWvm3Zgj96E7k9AX9xPksvhD3NyWQ/7JFFwIhRCL9yOCO/rM1evZGlmj782bM+N8+cvrKbRL60e6o+PwRnvqIYjD4/VDI8o3QMPUZRLT7QoxU9/ha7vhCc9T32Goa8UuOZPkKb/b0bd4w+1ypcP+rQib5HpsG+sDWBPyXP5T3FV149rLDvvd84GT9c1r4+DFezvhSaDr/Qtj896QzkuzBXtj5suie/wFotPjX1Yj46LoHAYI5TQDPP1b1cczG+xx/hvk4ej78NWRw/qkIRPswkk75tnCs+jPYZvTBvLT4Sn1c+J0iBPjojAD3HWRm+coEhwDivCb+Os8A+Q22vvszaoz3ToO8+fe2DPuuPEj6GZ9E+anhSvxdUUT4ZHPs+18rEvlvSZL642iK/ILNnv1+7rb5ZmHs9Wc/UPsfMAr3NhWk+wSnAvYhBQD0H/Ac+u0MmvtVWLD6+0Ug+RZi9vr1czD6+Tk8+//SoPJnUGr7B7rk+8rLsPf6gv7+II48+7tKVvpiNUj2tDlI+YJdSv+0Agj8E7+a98L6vvWYFrb4QAts/YHuMvh+mtD5D9NE+zQoDwMe+8j7gTp6+Qm8+viOYir9wkuy+qVf+vATDkb95Zhc++i8qPuShhr3+U7m/uIH/vtSutT0Faao+2+dYP/T4rL9bTyi/P+7VPVp5bT95tdA/NdnlPDwbWr1DIsM+hAiVviceaD6ZhD4+TMsePV4fpz0ZGKK9SiPiPViw471cNxE9aya3vRVhHb70hAQ+eldeviEWNT7QVg6/Hp5PPyI+mT1iA1M++lepvsHXDD6s5TC+jAH7P19jVT41eWA+glEhvrHqjz5qszS+lBmWP87Fnr2kNh++lpvJPq4bpb+C0/O+h2iqvqQV/T6ESLA93pkEPYFxTr5gTZK/sP7bvlXGt77XRfe9LiVNP7Nthb9oKlU9nKm/vQVwDb+dat4+iHZHvhgPlL5xlB+/Nz/tv0FUeL7pMza+cNynPHSYnz8yVw+9MEBgPXAq1z5lkla+hvFcPmAmnL1Cvj4+cmV1vVKHWz5zBYa+G0CJvszySr6XJZM+M+hCPkpC1L7IIXS93sGXPcL6HT9Za9u+GJC6PsuF1z5WkFg7OifHPfZ6jL2dGwdArWLqPEIyor6s2eC+l3PAPcJvqb5Acvi/65fCPXSMT75a75A/S4MzP4MmZz6aX26+2rJNPykl3ryVug6+ly/1vjRrgj8SPSa/1m5+vjkOTD28qyXAJ7VRv9aSPr62HkA+QKAXv556SD/dnY49+DH2PiTOgb/OK6E9BrlsuxOxZb7/ByLABOWFvk3XSj3Nty8/1Iwqvs8RozwJfXm+TWdOPtRN9TzJjp49/54WPswukj7CpIc+PSwSPUkMBD6XP/C+9KJFPUtLDj4k0C6+o5OEPqTFmD4tSP0/VchbPViO0z4K5wA+SifbvmS9R79GaDA+xUkwPmn+7D6HzUS9/7S+v3NDrr8wj/69I2iUPogirb9bSZm/v+WFPnxbsL47hpQ+2WIGvwD/cT1Og7C+rFWMPwruJcArYO49s35oPkVaKj+EdMw9jELWPUVYir4kpgM/9L36Ppt9Br5vCzG+mg+9v3+CLr8Pxk67hMwMPVo6sT++bI++31QNPjo27r2Pl6a+0IJ6vk91eD7kxzu+MnBSPiauzr20wWq+cpugvWwGbj3cLMM+8Iuru5AFGT2umRu+0DbbvSxwtL67JNC+0gGxPeaSFkDufWy9Si8Nv5QrgD6uvpA9CI8Hvka9yb6KLq+80ys1vnIS57tcbzBAS4X8Py1Eur3Fm+A8HIiRv2YUfb595Co+zTFGvG2tkT87Qx6/cEs9PmTtvr5BP92+qLBNPxheUz7Kb+g9Ic9iP/Z88jyAAlO+AcOuPo5xY7/PUqy/Vh9xPZiZ1j4xYT6/W+DWv9CORL5nYfi+1/ZMP67hxb46rZs+lEfHvorS/Lyw7Sw+LcoBvsu1oz04wJY+gFrZvOCNMz23ObI+Ko0CCEAQAUIEcl9iMUqAAoChJT12nRu+D9WPvnSsMr5SZx68FNgNPsz+Lb4hoxY9qFG+vFwjj70r/gi+M6n2O5eoPr3rhU09eMz/PZeQsL0HBbC71gU+PWP02b05tU68Rs68vo+9Zb5/kGG+hdSCvlx01rw2xzs95vkBvrgUyj3XLQI+flf3va3JoT3e9X8+rY8FvoIEfTqJrrK+TkHFvshrUz5myww+x9UMPgqyV7yZJMK7YgTXPjBwpz02rU2+q7UJvX6zZ76fAok+WLw3vYjcOL4z2iQ+bBfLvWe1Pjw199k9A3rcPkv+ozwwGFK9Cilyvr+YDz9ljXI9c26pvfnaND7895g9Vs9EvmWFtz4qjwYIAwhAEAFCBHJfVzJKgAaJTDu+fXw+PZyCZTxZhZO9Dj94vJDmo7s379M8eHYZvt3Ddb4n1Dq+RFB7vqnDIj7yJbS8ZH+4vTfFYr7udMu9DLa8vBhDhD7pU7e9+gAHPgQepT6eD/O98IGQvmUIjbxUVGi+faUDvfohmj2Zbwg+aQMvPd+ZLD4QVgI+LYhxPl6xkT3R1Qe9wmXFPWhIFD1SqUo9h+CfPSXOPD790ZQ+RROvvhsbLT4Yn4E9kqLAPR2gMr1b5zk9pYtTvqMtpz5OIoA94N6uPWg+Yz6eEAK+T3AMPCakuzpKVnc+/suPvhx63r1fG7a9PxW5Pcen3bzEJSc+99HAPSvWGL1wTdi+Sm6ePV5Ylj7Jx688Px9BPUGm1bwjdS+9T+IMvtUsPT78/eE9LhZUPdiYhj214pa9gfyQvaNkq70N+tK8ZkanvmyY0z3Z5Ru+eprrPR9P9z28b9A+4cSbPfgSPD5ZTga+Dmt7vUulhLwgcDo+lS5Yvr+G2b1CrMA+1qgdvSglnT1tyOq9zhkcPtO8X71Irga+oCqbvlzgWL0O49i9DOu/Pc1tAb46iIs+jWiTPfDC9D7QZCm+uambvRoGPb5oHhW+pcBUvatlMz5AIIq9pCoJvvYzY70bf6u8cwLXvYHZSL6y17g9K3y7vvHMmz0luVm+6xiFPXruLj7ww0Q991g2vS0xojzFhQs+RnKgvTqs/z0JX12+mTjuPYna6b0fnJs9zhuAPQ6w4j0aWBe+LwZMPR72Cr734nI+F/5dvbCmBT5F7sS9Q2ZgPHmykT329Wg+p5H0vTRzj722JJm9/IQwvkzFBj6SbUc98bqavKrOnz5ExtM91HQXvupAXD4A9he+tNh1Ps7QSL3Pw/y9dJi6PU1tHD7iqBI87toGvsyk0b0BRdY9o6uPvVt1jzzAzUw9+ChRvhoiGj7WEa+9IykUvn+kcD1yAd67WHe2PfGInj7JDFq9rl+pPcPk3z0zHho+EwQQvohMrD1fnHK++LgiPi4dab1w9lS+R2z6vUPMSz0qGAgDEAFCBHJfYjJKDPw7vj08HTk+hV7AvVobCgNvYnMSFAoSCAESDgoHEgViYXRjaAoDCOABYh0KBmFjdGlvbhITChEIARINCgcSBWJhdGNoCgIIA0IECgAQEQ==", "entry_sha256": "2071a2c1574c68ee0479f2220b9ff5c34fecbaf1d52a0933fcd0a08dabbb5eda", "entry_bytes": 35530, "ops": ["Abs", "Add", "Atan", "Cast", "Concat", "Cos", "Div", "Gather", "Gemm", "Greater", "Max", "Min", "Mul", "Neg", "Relu", "Sin", "Slice", "Split", "Sqrt", "Sub", "Where"], "tried": ["Harness: batched lockstep loop over the repo's SimState, 880 matches a second against Starter; final tick equal to the repo's run_match with OnnxPolicy on 40 of 40 and 30 of 30 matches. Each seed is played from both sides. The current model against Starter gave 0.9895, 0.993 and 0.9913 on three separate 3,000-match seed sets, against the seat's 0.9933, so differences near 0.003 here are noise.", "The premise did not carry over: the 27 percent in the current entry's notes was taken from rule n7, an earlier version. Its final rule b1 was already hardened and scored 0.9875 against a fresh neural exploiter.", "Starter head by ES against the current model, as briefed: three 45 s runs, about 110,000 matches, ended at 0 to 0.005, below Starter's own 0.025, because almost every match is a loss and there is little signal.", "Starter head, about 18 minutes on one thread (135,808 matches, about a quarter of the machine, so not the full compute): 0.47775 over 1,000 fresh matches against the current model, with 1 HP win, 954 draws and 45 losses. It never eliminates. It is a pure evader: median 432 m from the target, beyond the 200 m bullet range, altitude about 370 m, throttle 0.81, and 1 hit from 63,092 shots. It cannot beat anything: 0.059 against Starter over 400 for the earlier checkpoint. It does expose that the rule cannot force a kill on a runner, which costs a board entry 0.5 per match.", "Rule plus evolved residual, zero start, 90 s: 0.50 to 0.71 over 1,000 validation matches.", "Same, 565 s at lr 0.01 on 3 threads: peak 0.8335 at 226 s then drift to about 0.78. Best checkpoint 0.81 against current over 3,000 unused matches, but Starter fell to 0.981 from about 0.99.", "Same, 520 s at lr 0.003 on the mix with Starter and the evader, to win Starter back: validation flat near 0.82 on the mix, and the evader's 80 s matches cut it to 54 generations. Starter did not recover (0.980). The val-best checkpoint gave 0.829 against current over 3,000 unused matches.", "BEST, the last checkpoint of that run, chosen on those unused matches at 0.8475 against current. On fresh record seeds: 0.828 against current, 0.9745 against Starter, 0.99875 against built-ins, 0.63375 over 400 against the evader, of which 142 were HP wins and 202 draws. It scores 0.269 against a 10-minute exploiter. Not better than the current model against Starter: 72 losses and 9 draws in 3,000 against 26 losses on the same seeds."], "next": "Do not upload this in place of the current entry on this evidence: against Starter it loses about three times as often as the current model on the same seeds, and third place allows no loss in 60. Use it and its exploiter as sparring partners. The useful finding is the method: a zero-started residual on the rule took a hardened model from 0.5 to about 0.8 in minutes and costs 16 KB, while a Starter head gets nothing but draws. A next seat should (1) run it as a league, training a residual against the current model, Starter, this entry and its exploiter from the start with fixed shares, because adding Starter late at low lr did not recover it; (2) keep a held-out Starter check on every checkpoint and reject any below 0.99; (3) use a learning-rate decay, since lr 0.01 drifted after 226 s; (4) fit fitness to the board score alone once past 0.7, since the hit bonus may have caused that drift (not tested); (5) give the rule a way to force a kill on a runner beyond 200 m. The gun was never changed by ES, so a residual with a larger shoot scale is untested."}