{"credit":"Taylor (gtmcknight): original replica, Gaussian filter, solved policy, fee rule and qm tables; IdentityMD TC/COMBO/qt1 references retain their credited authorship.","item":"history","status":"complete replica experiment; no contract/submission","hashes":{"pinned_commit":"4812a6382311624e1506aca131ca35c378f930d9","commit_api_sha_verified":true,"all_four_artifact_url_suffixes_match":true,"downloaded":{"gates.py":{"sha256":"603280c1f595ca163ffaf2ba1ebdc7945fdb08dabae67fc53c922d39e1cce604","bytes":792},"replica2.py":{"sha256":"6e37b4e61e1ffa5ffb1e43912ddb8659105989aa986eb48fcd1d7608f8dca685","bytes":7503},"evalpol.py":{"sha256":"6e67667bfed7798bdd3384652a6efc6ba0b1fce653740d0f9af2bcbcb8e6a27c","bytes":4610},"real.py":{"sha256":"1305b5333731c046c3e6736cc2578eff4635c88132a39c22f5780d78aef2c81e","bytes":9586},"latt.py":{"sha256":"925c0fe996b19fd51fb79a85d5e4dba4bc2f5453d0a156cc1e480f65a6a00cc8","bytes":8893},"pol_mid.json":{"sha256":"efbb5ee102db4d312f64565d9bf7232b6becbe02b624164f744a402bbbf2983b","bytes":780061},"qm_k1e-5.json":{"sha256":"5ff53e42ca87390e5f65d1caa3f7a480e67a8212fb0670274dfd562b05ba1afd","bytes":7892018},"dp.py":{"sha256":"ca4e7c378d23c46c7def4716611135c471a831d15234b984fe30d58d9eec5734","bytes":15595},"qmdp.py":{"sha256":"63b8c35048718c37eb0a5dae9e70849fa3af257c335e4a2e3a625c833f1dc1eb","bytes":8609},"calib_out_2048.json":{"sha256":"fdf24c7aa2b604d770336a5783276da8ed4d50bca9dbcc02e98b4da58f608d26","bytes":2948144},"qt1.artifact":{"sha256":"8a6e303b662e1a5ea9bb47f31acedd3988ef65b3ddb66d074517aa59ad41ca40","bytes":62223},"TC.artifact":{"sha256":"d3c849db2ac31e567b57e52b80b86e8c24c842e5b4125b9d75c78e3cb07115b9","bytes":23156},"COMBO.artifact":{"sha256":"d360bea6c8c982717b7b43a8069a87013169c89ee3e4da25622db4d880b936d4","bytes":25626},"QMS1.artifact":{"sha256":"17d249ab0c33fc44931d9b105cf3f9e3846a560bd518b67310c2a56defacdecc","bytes":29486},"commit.json":{"sha256":"e1c47ec0c784d923ea92481aec80ea8e6c2f71b9fce9fcfcd01ab0e05ff9a75f","bytes":5480}},"extracted":{"qmref.py":{"sha256":"1dfc9b7a3c2237ff7a93dd9f0af54b9bdab16f2ce0a7eb00784cb4432598db92","bytes":7739},"twocomp.py":{"sha256":"9d3c1833321493fb51bbde4e1b4dabc2f8978bd6d774ee2af8dc8aed13f0f0af","bytes":11202},"combo.py":{"sha256":"c5b711a0b4fb4783e8879a484bf87a427171f4a11aa5eb5ec3513c24ead1a796","bytes":5929},"blob.bin":{"sha256":"3ac7cb8811ecfb6a17f499aa4e9b70dd55f7fb6432b3aec3b904e11a3dece38e","bytes":10167},"Strategy.sol":{"sha256":"c68887cff61180e12ba2b1ceb333edf76a830f91f63b5d775a8806fb9ec08aca","bytes":42990}},"extracted_all_published_hashes_match":true,"source_url_prefix":"https://raw.githubusercontent.com/gtmcknight/amm-sim/4812a6382311624e1506aca131ca35c378f930d9/research/solved/"},"gates_output":"seed   perfect-info      gaussian+plain    gaussian+qm       lattice+qm\n10000  448.261189  443.455598  443.522608  443.663520\n10001  576.779199  571.463409  571.623551  571.977293\n10002  531.886497  525.168028  525.865859  526.411631\n","task_findings":{"nominee":"compact_half","nominee_report_gain_over_COMBO":{"n":2048,"mean_diff":0.03748018749726159,"se":0.007526575137643317},"nominee_report_gain_over_actual_qt1":{"n":2048,"mean_diff":0.6348711508829178,"se":0.013727885580878667},"interpretation":"History features yield only a small policy correction. The large part of the gain over qt1 is inherited from COMBO, not this classifier. Replica returns do not establish a leaderboard score.","teacher_report_gain_over_COMBO":{"n":2048,"mean_diff":0.04528076658034333,"se":0.008814927377648872},"compact_distillation_has_large_rare_probability_errors":true,"probability_loss_vs_return":"Full-strength compact has lower choose Brier loss than half-strength, but half-strength has the higher paired choose return; nomination used return.","promotion_to_contract_or_board":"none; this assignment is replica only"},"seed_coverage":{"training":[73000,73511,512],"calibration_and_student_distillation":[64000,64127,128],"choose":[64000,65023,1024],"report":[10000,12047,2048],"steps_per_run":10000,"configs":"replica2 defaults, cfg=None, qcap=None; fresh policy per seed; immutable decoded-table cache only","mandatory_pre_edit_gates":[10000,10001,10002],"invariance_only_seeds":[76000,76001,76002,76003],"no_hidden_configuration_used":true,"training_first_own_callbacks":3578409,"training_sampled_callbacks":298136,"training_COMBO_raw_error":0.160925986940006,"training_COMBO_raw_brier":0.11198611428655915},"frozen_spec":{"manifest":{"credit":"Taylor (gtmcknight), original replica, filter, policy and qm tables","frozen_utc":"2026-11-10T04:01:48Z","nominee":"compact_half","nomination_rule":"largest paired choose mean over COMBO among compact and compact_half; no report rows accessed","choose_scores":{"compact":0.017316605071657665,"compact_half":0.03182334700248585},"model_binary_bytes":1252,"binary_format":"HIST1 + LE uint16 node count + LE float64 strength,temperature,bias + nodes <biHHh: signed feature id (-1 leaf), signed Q20 ceil threshold, uint16 left,right, signed Q12 ties-even leaf residual; all nodes 11 bytes; header 31 bytes","selected_hashes":{"history.py":"f56aeadc82a3b99fc18d85875fdc2c80fb1c9f1951c72137bf8791cbf68ba51e","patched_combo.py":"ad8b984a4c63d684987b287ff0b7be44b980d00aeae28401f8f7a9f2b5076502","driver.py":"4326c0eb1808f02db678f6ab133ba433bfc6d871f28de34561c75c8e48aca9c4","scratch/teacher.json":"3c7ffb676853930bc4c942731f0d672f10ec87c96598bbe00c5ddd40732c4e04","scratch/compact.json":"73d749f82533722834937cb9a54e5a93eb36e8715d553914a4cc5c4afd71e6c8","scratch/nominee.json":"25a7fa43a33abfbf7c831cd22494d71ecaffd8f95795ea437253aa4df0979f74"},"complete_hash_manifest":"scratch/frozen.json in source bundle; report driver checks every entry"},"feature_formulas":{"credit":"Taylor (gtmcknight), original replica, filter, fee rule and qm tables","scope":"First own callback with dt>=1; all later own callbacks keep COMBO and update history only","definitions":{"side":"-1 if is_buy else +1","s":"reserveY/reserveX after callback trade","sp":"spot after previous own callback","dt":"t - previous own timestamp","raw_pi":"unmodified COMBO la/(la+lr), before posterior, phase and branch quoting","pm_pv":"mixture mean and variance of COMBO propagated prior","lp":"ln(s*(1-fb)) if is_buy else ln(s/(1-fa))","mt_vt":"COMBO retail-cut posterior moments","fb_fa":"stored fees in force during callback trade","q":"min(1-1e-7,max(1e-7,raw_pi))","history":"updated after live callback; ignores arb,p,s2,j and init configuration"},"features":["ln(q/(1-q))","ln(max(abs(ln(s/sp))*1e4,1e-5))","ln(dt)","ln(1+min(100,t-last inferred pin))","sqrt(pv)*1e4","side*(lp-pm)*1e4","side*(mt-lp)*1e4","sqrt(vt)*1e4","(fb if is_buy else fa)*1e4","(fa if is_buy else fb)*1e4","side*ln(inferred_rival/sp)*1e4","min(2,previous inferred retail count)","float(previous synced)","previous own callback pi; zero for a later same-step callback","ln(1+previous callback amountY)","side*previous ln(s/sp)*1e4","float(side==previous side)","ln(1+second previous callback amountY)","side*second previous ln(s/sp)*1e4","EWMA of ln(1+amountY), decay .9, initialized zero","side*EWMA of own side, decay .9, initialized zero","pi at previous first-own callback","ln(1+number of own callbacks in previous step)","weight of first carried COMBO component"],"pin":"timestamp of first-own callback whose predicted pi>.9; initialized 0","replacement":"one hook before is_arb, .005/.995 posterior collapse, carried m/v, phase and branch-weighted offsets; remaining COMBO logic identical","sample":"one in 12 first-own callbacks, random.Random(seed+78901234).randrange(12)==0; row is 24 float32 features followed by float32 true-arb label; live features/inference are binary64"},"nominee_calibration":{"strength":0.6769214287241393,"temperature":0.9789059573928485,"bias":0.00892241043931171},"replacement":"pi hook before posterior collapse, component weights, is_arb/phase and COMBO branch quoting; other policy logic retained","teacher_training":{"algorithm":"histogram Newton gradient boosting with raw filter logit as offset","trees":64,"max_depth":3,"quantile_bins":32,"learning_rate":0.15,"l2":20.0,"minimum_count":180,"minimum_split_side_hessian":5,"hessian_floor":0.002,"unscaled_leaf_clip":[-3,3],"rows":"deterministic 1/12 COMBO training callbacks only"},"distillation":{"algorithm":"adaptive binary partition table of teacher residual logits","fit_seeds":[64000,64127],"weighted_loss":"squared residual logit error, weight=max(.01,raw_pi*(1-raw_pi))","max_depth":7,"min_count":300,"l2":2.0,"leaf_clip":[-3,3],"thresholds":"Q20 ceiling","leaves":"Q12 ties-to-even","calibration":"three-variable Newton logistic regression on choose callbacks only; candidates full and half fitted residual strength"},"quantization_and_distillation":{"seeds":[64000,64127],"n":74359,"nodes":111,"leaves":56,"distill_logit_rmse":0.4978802326063741,"quant_residual_max_abs":0.00012165660485918828,"quant_residual_rmse":7.738155458680028e-05,"quant_probability_max_abs":3.0414104634557404e-05,"quant_probability_rmse":1.108604457203921e-05,"quant_route_or_leaf_differences":0,"distill_probability_rmse":0.019084579802878095,"distill_probability_max_abs":0.7847612278173357},"costs":{"json_model_bytes":4664,"packed_model_and_calibration_bytes":1252,"node_count":111,"cell_count":56,"max_threshold_comparisons":7,"teacher_tree_count":64,"teacher_max_threshold_comparisons":192,"extra_history_scalar_count":11,"compact_table_referenced_history_only":["h_disp","h_size","h_first","h_pi"],"ephemeral_feature_scalars":24,"model_inference_arithmetic_excluding_features_and_COMBO":{"multiplications":2,"additions":3,"division":1,"exp":1,"clamps":2},"feature_and_history_extra_transcendentals_upper_bound_first_callback":{"log_or_log1p":9,"sqrt":2},"gas":{"measured":false,"estimate":"No reliable total without EVM math and storage layout; static operation/state counts only","meets_245000_peak":null}}},"probability_calibration_and_error":{"choose":{"teacher":{"n":7203331,"brier":0.10912372205411557,"error":0.15735956045890437,"logloss":0.3411949256369909,"raw_filter_brier_same_trajectory":0.11071360870125109,"raw_filter_error_same_trajectory":0.1590789039126482,"ece10":0.004461512151068042},"compact":{"n":7202533,"brier":0.10988784469628321,"error":0.15794137978958236,"logloss":0.344662910118196,"raw_filter_brier_same_trajectory":0.11095276005909577,"raw_filter_error_same_trajectory":0.15939711765291462,"ece10":0.0052061233171682205},"compact_half":{"n":7202010,"brier":0.11013982324988658,"error":0.1583169143058674,"logloss":0.3463611030245533,"raw_filter_brier_same_trajectory":0.11112039062197036,"raw_filter_error_same_trajectory":0.15961863424238512,"ece10":0.0029133978814069}},"report":{"teacher":{"n":14386213,"brier":0.10958253032049846,"error":0.15812215487147313,"logloss":0.34237903707849526,"raw_filter_brier_same_trajectory":0.11116552997275599,"raw_filter_error_same_trajectory":0.1598385203944916,"ece10":0.004184811814579195},"compact":{"n":14384498,"brier":0.1103701330707829,"error":0.158780236891131,"logloss":0.34587660367106354,"raw_filter_brier_same_trajectory":0.1114263060456804,"raw_filter_error_same_trajectory":0.16012501791859543,"ece10":0.004998002936409234},"compact_half":{"n":14383077,"brier":0.11062073522104728,"error":0.15916003230741238,"logloss":0.3475867525587428,"raw_filter_brier_same_trajectory":0.11159877718713777,"raw_filter_error_same_trajectory":0.16046934880484892,"ece10":0.0029229735448886335}},"calibration_fit":{"seeds":[64000,64127],"sample_n":74359,"raw":{"n":74359,"brier":0.1129667208036103,"logloss":0.3568200560756736,"error":0.1625492542933606,"ece10":0.007330033790757929},"teacher":{"affine_logit_coefficients":[1.0081996854081747,1.1238238461475387,-0.00673001879969805],"calibrated":{"n":74359,"brier":0.11113814052645826,"logloss":0.34714596639311895,"error":0.16084132384781935,"ece10":0.003030555729281214}},"compact":{"affine_logit_coefficients":[0.9789059573928485,1.3252848385298768,0.00892241043931171],"calibrated":{"n":74359,"brier":0.11161795139263683,"logloss":0.35010495857866425,"error":0.16139270296803346,"ece10":0.004236834097412573}}}},"closed_loop_distribution_shift":{"seeds":[64000,64127],"method":"same deterministic 1/12 callback sampling, live-model and COMBO closed loops; callback populations may differ","base_callback_sample":{"n":74359,"brier":0.1119112824162901,"logloss":0.35189768579107694,"error":0.16124477198456139,"ece10":0.004503560093946247},"nominee_callback_sample":{"n":74369,"brier":0.11107711645342601,"logloss":0.34951324802884026,"error":0.1595019430138902,"ece10":0.005037240378706193},"max_abs_standardized_mean_shift":0.043640954759162034,"largest5_standardized_mean_shifts":{"ret_sd_bps":0.043640954759162034,"prior_arb_weight":-0.03743939725501468,"prior_sd_bps":0.03146517322993003,"pin_age":0.026599588360091787,"same_side":-0.009926938485675177}},"policy":{"immediate_base":"COMBO","reference":"actual qt1 QMRef, not Taylor Gaussian","nomination_before_report":"compact_half","choose":{"seed_range":[64000,65023],"n":1024,"policies":{"compact":{"vs_COMBO":{"n":1024,"mean_diff":0.017316605071657665,"se":0.012595535065420339,"worst5":[{"seed":64522,"diff":-1.52485680436655},{"seed":64338,"diff":-1.3074356271954457},{"seed":65018,"diff":-1.252868943695944},{"seed":64844,"diff":-1.2337628187001428},{"seed":64589,"diff":-1.2046690510048848}],"sum_d":17.73220359337745,"sum_d2":166.49857124134752},"vs_qt1":{"n":1024,"mean_diff":0.6269212516011892,"se":0.019779548323703437,"worst5":[{"seed":64101,"diff":-2.050174847154267},{"seed":64522,"diff":-1.9796749351207836},{"seed":64670,"diff":-1.1829985213810232},{"seed":64845,"diff":-1.1141547012176716},{"seed":64262,"diff":-1.1015939985968544}],"sum_d":641.9673616396177,"sum_d2":812.2973079883617},"mean_edge":536.3474167232292},"compact_half":{"vs_COMBO":{"n":1024,"mean_diff":0.03182334700248585,"se":0.010272398266490089,"worst5":[{"seed":64173,"diff":-1.1663718633955682},{"seed":64338,"diff":-1.1403415102038252},{"seed":64715,"diff":-0.909648381090733},{"seed":64611,"diff":-0.8820244988987724},{"seed":64011,"diff":-0.8451540207144035}],"sum_d":32.58710733054551,"sum_d2":111.5769870143213},"vs_qt1":{"n":1024,"mean_diff":0.6414279935320173,"se":0.01950272653892826,"worst5":[{"seed":64101,"diff":-2.223622407412222},{"seed":64374,"diff":-1.1882642012438964},{"seed":64845,"diff":-1.0162863930569301},{"seed":64011,"diff":-1.0154347943157518},{"seed":64651,"diff":-0.9875919225893313}],"sum_d":656.8222653767857,"sum_d2":819.7472350362904},"mean_edge":536.36192346516},"teacher":{"vs_COMBO":{"n":1024,"mean_diff":0.04092274425463932,"se":0.012314111040783652,"worst5":[{"seed":64522,"diff":-1.569772002101331},{"seed":64715,"diff":-1.5420209471435555},{"seed":65018,"diff":-1.344316146227186},{"seed":64338,"diff":-1.3115883237182402},{"seed":64817,"diff":-1.165114319436384}],"sum_d":41.904890116750664,"sum_d2":160.56285217663967},"vs_qt1":{"n":1024,"mean_diff":0.6505273907841709,"se":0.01998675326523482,"worst5":[{"seed":64101,"diff":-2.0859247762688256},{"seed":64522,"diff":-2.0245901328555647},{"seed":64651,"diff":-1.3035601902498684},{"seed":64205,"diff":-1.0771882734086375},{"seed":64670,"diff":-1.0471686248620813}],"sum_d":666.140048162991,"sum_d2":851.8082655086589},"mean_edge":536.3710228624121}},"COMBO_vs_qt1":{"n":1024,"mean_diff":0.6096046465295315,"se":0.01940052280944413,"worst5":[{"seed":64101,"diff":-2.7769830832371554},{"seed":64845,"diff":-1.1638684799705743},{"seed":64374,"diff":-1.023433823667915},{"seed":64795,"diff":-0.9831714544297938},{"seed":64555,"diff":-0.9461893679592208}],"sum_d":624.2351580462403,"sum_d2":774.8145734774697},"means":{"COMBO":536.3301001181576,"compact":536.3474167232292,"compact_half":536.36192346516,"qt1":535.720495471628,"teacher":536.3710228624121}},"report":{"seed_range":[10000,12047],"n":2048,"policies":{"compact":{"vs_COMBO":{"n":2048,"mean_diff":0.01565358910994738,"se":0.00914684430835077,"worst5":[{"seed":11478,"diff":-1.5525986101494027},{"seed":11696,"diff":-1.5154533127258674},{"seed":11370,"diff":-1.470958950100794},{"seed":11076,"diff":-1.2529142185363753},{"seed":10184,"diff":-1.228918715343866}],"sum_d":32.058550497172234,"sum_d2":351.24592683437595},"vs_qt1":{"n":2048,"mean_diff":0.6130445524956035,"se":0.01424573435325453,"worst5":[{"seed":10146,"diff":-2.782326918699823},{"seed":10519,"diff":-2.3389081751689673},{"seed":11746,"diff":-1.483256392576095},{"seed":10437,"diff":-1.476412915890137},{"seed":11062,"diff":-1.3861797337478947}],"sum_d":1255.515243510996,"sum_d2":1620.467184420683},"mean_edge":533.8758752189987},"compact_half":{"vs_COMBO":{"n":2048,"mean_diff":0.03748018749726159,"se":0.007526575137643317,"worst5":[{"seed":11696,"diff":-1.5498139185848459},{"seed":11478,"diff":-1.1609677507263996},{"seed":11918,"diff":-1.1071584596961657},{"seed":10772,"diff":-1.068682665353208},{"seed":10711,"diff":-0.9894338044483675}],"sum_d":76.75942399439174,"sum_d2":240.3654650372765},"vs_qt1":{"n":2048,"mean_diff":0.6348711508829178,"se":0.013727885580878667,"worst5":[{"seed":11062,"diff":-2.1171435619281738},{"seed":11555,"diff":-1.7652117186887608},{"seed":10519,"diff":-1.6447365610819134},{"seed":11376,"diff":-1.5763696499136586},{"seed":11929,"diff":-1.4548483537895436}],"sum_d":1300.2161170082156,"sum_d2":1615.5206468921613},"mean_edge":533.8977018173861},"teacher":{"vs_COMBO":{"n":2048,"mean_diff":0.04528076658034333,"se":0.008814927377648872,"worst5":[{"seed":11696,"diff":-1.5373526885083493},{"seed":11478,"diff":-1.290418517701596},{"seed":10856,"diff":-1.2528973170594213},{"seed":10546,"diff":-1.1816222884603462},{"seed":11911,"diff":-1.1498719641137995}],"sum_d":92.73500995654314,"sum_d2":329.9497483636579},"vs_qt1":{"n":2048,"mean_diff":0.6426717299659995,"se":0.013854124575007502,"worst5":[{"seed":10519,"diff":-2.7452242133089158},{"seed":10146,"diff":-2.599006940872755},{"seed":11062,"diff":-2.0471980103233136},{"seed":11808,"diff":-1.2890364370200587},{"seed":10903,"diff":-1.2388568368851338}],"sum_d":1316.191702970367,"sum_d2":1650.5272648927562},"mean_edge":533.905502396469}},"COMBO_vs_qt1":{"n":2048,"mean_diff":0.5973909633856562,"se":0.013387993974623986,"worst5":[{"seed":10146,"diff":-3.552496929485528},{"seed":11062,"diff":-2.3348491437070606},{"seed":11808,"diff":-1.8227381435300458},{"seed":10519,"diff":-1.782691465320454},{"seed":11376,"diff":-1.7502956875490554}],"sum_d":1223.456693013824,"sum_d2":1482.2951576559892},"means":{"COMBO":533.8602216298888,"compact":533.8758752189987,"compact_half":533.8977018173861,"qt1":533.2628306665031,"teacher":533.905502396469}},"se_definition":"sample_sd(diff)/sqrt(n), equivalently sqrt((sum_d2-sum_d^2/n)/(n-1)/n); worst5 sorted by diff ascending"},"checks":{"Gate_P":{"status":"pass","before_policy_edits":true,"all_12_readme_values_match_6dp":true},"actual_qt1_three_seed_edges":[[10000,443.1017411146794],[10001,571.454483381923],[10002,525.908441161234]],"qt1_TC_prefix_AST_equivalence":{"all_13_module_definitions_equal":true},"identity_hook":{"model":null,"hidden_invariance_pass":true,"unmodified_combo_equivalence":true,"seeds":[76000,76001,76002,76003],"callbacks":46326,"first_callbacks":29011,"trace_sha256_concatenated":"90657f67eab460ad29743b5803649a20557fe1d1ba104eab9486398439fe6fa9"},"nominee_hidden_input_invariance":{"model":"scratch/nominee.json","hidden_invariance_pass":true,"unmodified_combo_equivalence":false,"seeds":[76000,76001,76002,76003],"callbacks":46343,"first_callbacks":29054,"trace_sha256_concatenated":"47fd81be67e325741995500fa3ecc3bbe564fbf2ed170e697538443b8572ec20"},"packed_table_roundtrip":"exact equality, including threshold/leaf scales and binary64 calibration","offline_replay":{"status":"pass","network_used_by_replay":false,"decoded_sources":18,"compact_replay_seed":10000,"edge_equality":"exact binary64 equality for COMBO, actual qt1 and frozen compact_half","regenerated_training_seed":73000,"training_rows_bytes_exact":true,"wall_seconds":6.126407042844221,"bundle_payload_sha256":"f78c65965a6a64c14b28d8d0896b05ff58612ea83dc8855d80acf6f00955f20b"},"teacher_regeneration":{"status":"pass","teacher_sha256":"3c7ffb676853930bc4c942731f0d672f10ec87c96598bbe00c5ddd40732c4e04","exact_checkpoint_sequence":[8,8,8,16,24],"all_64_trees_retrained":true},"Gate_Q":{"status":"complete","all_tested_policies_both_windows":true,"paired_with_immediate_base_COMBO_and_actual_qt1":true,"report_never_used_for_training_selection_or_retuning":true},"Gate_F":{"status":"not applicable; replica-only item, no contract"},"engine":"not run; replica-only item","pinned_git_tree":{"url":"https://api.github.com/repos/gtmcknight/amm-sim/git/trees/4812a6382311624e1506aca131ca35c378f930d9?recursive=1","status":"all 11 git blob hashes match pinned commit tree","verified_files":11}},"reproducibility":{"versions":{"python":"CPython 3.12.3","system":"Linux x86_64, little endian","dependencies":"Python standard library only; no package installation"},"bundle":{"codec":"base64(zlib(JSON object mapping relative paths to UTF-8 text; blob.bin value is base64))","sha256_uncompressed_json":"f78c65965a6a64c14b28d8d0896b05ff58612ea83dc8855d80acf6f00955f20b","uncompressed_bytes":84790,"zlib_base64":"eNrsfQlzo0iy8F/hzcSEJAuLQ0hC9mjiFTp8yrZ8dLfb7XAAAoStqwEJ2RP937/MqgKBLHf3zM6+ePs+9+5YQF1ZWXlXFvz5S+DMx75tqpX58y97v3yh/7tYBM7uxXM0mk0FXi7MXMFyps7QtyMrMIfPkjmZ7Nojczx2pp4j/HfdqjdVxdWFItz7U6cShKJgu5MJXnyZmoHlR4HpOYuAlgSzReTwSycy/TFclkQhDvwocqaCG8wmQjRyhHC2CGxHcGeBYDv+2J96grOaO4E/caZRWPkyvTw7EMIocMxJKAx913WCdePLRRgJDBwRehJMK5yNYVwhtGeBEwqT2dIRrGc2kOMM4U+0/2VqzyZzM3CE+Qxm7kO9uekHUArYoDXNCauOw3+ZElbvWfCcKBRMwQaUWKb9JJguzFBwlk4AIwDKHOzAhzqzeAptZmOhCFgRYGYMAzD9sf/ksHZXsTkHtM3HC+hzCpVMe+wIS9+JhWIULAAjAJMwD3xAzhxqDCloVVmw5iHtvBAK4XwWCaHKAcHyuTMdAgphOezIB2hgUdMZAaDTCGDAvgCgRTANhaLlDx9cB7Bnhk94UaowCvky9SfzWRAJEzMawRJCo9kEsXEitARFliuysIO/ePVl2iefevBcrihfpmfsSpargvCrMJ0FE3Psv+CqOQ528GU6dFzBHpuTuVt0S3tfpgL8YwBhO8F3BVf4nV4649ARirR3+vQPgV7Tx25p3VuwmBZxwUS+VKIwfQgjZx62KIhIp17rbIZU8tU25/QyGdkOAF42v8ol/Skq8gO0wv+EMiWEEqsawjweJg4sVwuaVRZTH8gWqjcrMISqVGReD+aWryFX6qKwLg99b2Ju1oDRdF0VKeoAap3XhXkD8BzWHBAijiOuO3M9Vmn+ekKZOQTbS2GiCq8xDHw3gkq7cqUGi8z657+sxmTxMEYcIG1UxjOvmMJUEnYF1kypqOwva+JMGFJoE2Dw4i7c8wHnCUmx25UiCs9K8kxcUxktBAw9q28UupYIbAOFjAoq/tSPihTuFFcJnBsNEnK0gD+Sa5NXchAWR2VUzds5TqiwB0jkcAsEDOwFbOtPhVvBjCjbZViYM+JkhmKNdRKb0+gROgGxYkZRUExIt4AF4cNjQRR6JlB6KeUQHBIkSmZkEJ5cuAgSFgnO0HN4/SWtvlSygIeLyYNrwRN6YWaLPIpV6HZXOOvx2l+DKFkzvOaTB4EW4TyBijynyDmtlCHRubCTWWlGT+WUkOZBxTMXYVhkCwhcUVo3/VXY3d0VUmUi5gSIH4C4L6LwA5kZIK6HDko8JtcAv1NESJjhFZwTkIsEdLN+CiwFBb8L8wzI+G+6QsTAPIsn0KLoIQHPS6V9wcSClQqYma7yTaArKPxDkDe6wn+01cSfFs2ViK13ANfNZul1xWmQ9m+u9uH2GW4RBCiAWyd6wAf4eBcmsy9E9JY+lmDZXnc4D4DcWggATAAaRc+v6wDktNp22HOsBmCIOPw+csFui7ZbN3HGDJ9/vMLnarnG5w7SlwTo5NgsQuEuDFHaMoXvITXFVZlNz1NfIex5xhZ9FwveRA7U2uUY2oocKAcIqLL8Z/BEBQbINaBFJU+LyhZa9FJmdM1X2PHeQs4GAefpV9lCvz9C92s6Vt6m48z6KD9Fy8omLW/v82fI+edIOqdeMsulJMu1z+Tr+g7EZ7kFI+5zmU9vAI7NVcn+21RDs+kDNQ+LVJyDqQXjRiDoI5EDIwrXYPGB8SImwgrMgBLOh+kIau/8/c5KPwHq9xTgmsuVLVyeoVTrpyl1QzRskQxKaTtB/IhaUwJMBMQb8kF5Sz78rIz4y3Li7xLf8yxDfBQa0OBvgvRdAmSUgSTzPPtX6e+n+/onyI+ZBczSWT9FG1TeF76iRQuuJrNmM63ikQ8+1VdYE2p+bqp7xKeCzXfeaI+WzmOwNnVgclNqKG/0hGYlgwGMYTRYlmbgm5FTpIYy2jibSADUoiMi+CEYOJGA/gglIdrTH7RsL+kWb/KtrcVzfsrUXaq9GgLq7SHu6IIFeBsKn4TYj0bC7etF8cw3dQ5VA0rCslQHeGjKPSvItWpaoDI9DwUqFKCnYaLCM7fYKEO0fE3AZkDbvK6wwPGKWFp8RnVP+0WUlKjqQPEwxEkOhf8CKmA0SjEmCW/1xr3LBSrcxLmkTdjDP1h7+nyhbCHbBZ0p1tmFCltl0wINIcXZ1d4UT8/MOilj1cR+WaXyCQTCapZam8HqTbmzmv1A1IANAvS9wH6gLsiL/deWyvYpKD81BZSwC0YJb0xB+ZengESyUDIzcKgQdIb7zB1Krpmg3LYk3xG8qUyls/hXFPoKBOBCyQlBXpKVqI/BT6r0n+nunxCqTK+HTlZGhM4YnKtPr3uPVgntS9tUj2e9aQZQiZUKj2dKMqjAVig8rFR4PKfCY5UIDwuFh/WW8LC+IzyWa+GxSoVHtCqllsUr0QHz2yo4lhnBscwKDmjAHv2Bbemz5TahscQZQo1dKN7Kb8sfiozU5Vmqf8vlyZkqPxQZdBC0s9Hw+DmRsVR+agooMpZs/f+OVfZTU0DSoGMkM/iuyHie/T2RQUf4F82wpfLKdPpbUuPne/snhAaPIAEOXFhEHkbCOzMXyMVthCIGo1qOwoI4LUdNtgJauBIiWrsttHjh8bIV4ATMZcvEX8RziyJbfA2ya7U4EFIS5sVIo6PhDFocos2i193QcFRrHR1spRHCVjZMiP8DPISh0PNXzpCTHgadHx4wvvjwUASZ6cLYpT2Unm4FPQd3XY0FIWmVHRPqcASxqiL/XVdPl/XHTdagndPdg8FiFjkcQBbFp08wtC/Mi0p5UqI2puUP8XYXb4PZgu8sXF/edLMBy+zmBu5nUKDSzYE3cDBJcEBDvT/EAd0lYBHv6psY8MMHMFyZp2FyRxcompIPJe5QzdrjIYoRpPaMGb82bEOMi8xRLSg0tI+gZsOP7lqN8Vq7SS1omYks8hkwtnm9FO3ZZO5sLMbJdBb/aCunIhBYrK90IQVYGiewFyyWHPhLEyviYlrPwoJTtAXFU1wlwXLG2D+u83iGmhWWmXc0W0R4l+0G72k3vLVp4VYZJQto/cN1HqOJAr8jv4UB3Cz+KbawnF4t+C/URJJYt/vfQhwBEkeI8pEu9lkPbQ+LPdtJn63rfzVpRHy1SUWANZFGyQIzRzgLeLxRdeSXsjT31eLxtQ2KYz3CSIGVa5/pcXd7jxyDjI6/opwMxQwDfAWC/UX85eskcNzsrvBXUI/wyAmcqQ1Uc20+j2cB0ErRiyb209T3RlFJOMDQvW8C6ftjlBBFE2k4iHarSMURULc5Bm+20p6ZQYjGYhT4k121hMaaY45hPKj3dcJ2A4MF+OezJd+87HTb551uB0p3w/nYB5c4Mq2xAwAM+leKYI1nsKbhyFRrdaFq2g3b0vVKpVKqCFd03GAojH0rMFFiTcfPlS/TRWh6MBO6X0ynK/BdzUH/0nH32U+xgD1XLH9aKKFDbiZb4mqy8VvkOzLAF9MjoFaqn9ADTyeENLzH9nwW47EwH5s+bgGPHybIdzgL3BMqyiL8r8T3R1VdrdbluiJrVb1e05uCY9oj3HHetgEbRsHCjlDMXB0doD2G+4VNrSbs7KDVetp56IsC/L3KbsspSkVRwSor4eorjTpAf3Z0/cbw+V3cZGd1PvKLqzVTrrfv6CYfBqNWKBnVSk2u16HLhlav4p6cyppf5JvTRpSSWUeBW0SzW6loiqYq1VpdrTaqcrOmlDKbuzDzqV0E3bwEPQ3yA4h9L90bXQq/A8TpAO54ZkbF8QyfTvAPVAYmQjHCt7SG2Q2t4hLjF+gqjDHUx4Q8WIfIkUWQWJlHL/AIZ2OhD4EX5nqD9gXcArB+FTWF4wVBxSIAAq+o4TbyYQ6AgIrKQzVzsDzmONac9ifSX4s7P8W5iVrIQgBecrYVdD5BYTCEvgImImTWM0VtkW6yYWNqn7IO4A69oSCL2DlyaBGJP0Eo2uB38v1a3AbDojuJsiKUBm4pMVYW07lpPz0gdxULvxdgaKgrcj6dQT8wE/xB85A3sc2xjdbVun5pP5nWkg0CStQBssde7va0e6EFfmABBUCB99YStHQjElZ2CuwFf2NRiAELwCpTyprDYsEwDg8No4BLzPrEBtgf1yfIK2M/jIpY+behURB+g8a4YTJVky4KJRhyX/ionuYrj2hlFStTm2FdmfUNQCEYd/frPdOHzJ4pFGeROoVlnNrw31MKOnS2Ln8K8qMf4ujFaQCrqpUQhid7awV7XSHcWuGJV1gPxVGFkACiqbWCEcqnjO2Dc6uYc9x5Ld6tQRcBTPjPvi9t7BNj/cxko6TtJjqLES4v+Ol3CvtR70spbBwuRgEwFWfKKPe113H2sYXU8FFuIUEctphMGAF7XbXu2E1YotCh3Bau7kVc3qRoyYqWWASPobBntNDMw1m08E+Ggywz9MPiEyAIFOyqlDP3TNAD9sLybcFI1BmtLoABtgD1AfpgJRTbs9Xu0BGMGc0SshdB6M+mAGqA2lYALAEc9ngxdIZZkwx35emo+zQGjsiIUC5xxjCQ8hQWs4juJvcoCFcgiOAaw9f3TBxhUAOnOlnTZdqRUrpPpdtK+KMFTXeV+8wigq0bcUNo8p1e1hCkgwMxUvhgeFzlbG0oucOO75kRviajYS4Gr2VZx5hu7wz+G26E8reCSOttBBRQIzBwh/e43wgTYDqBP2RI3E2n9Hr7mkYpADacehn3tHg3KItRPBtw/Tpcv9lqYziqaS3WemNcxKgxzfGCcbc3vV+TargIXNN2wuJFnkx7d9H93Rh8gPH9XTCL7+/s2fieryywD6PcUpLR4gU+yoOhA/QOyAQHJkuW4GdsyLwIVF7xtYgQhXYJ2zvTxcQJcKfk4q6AvFW4z4lFgAU6REMcyntG4R6nTWnKRyHAQ3G4mFAO3Fq4zyz5SZChDA1a3i0Bf3Vtzd9PAXZ4B23PPhYQwwqvnOnE/mEnNu2Ewrq1h3CzBw7vFUC7tYVBAWey5QTw5TOR5K/Jl0NcArI07ExlQOwjq/yYUTYAWSlLK2GmBazjmLUY57mDA5ht2MssbrLAb7TaYKgOtrwLF5OiEd4BoYU44/ZdEaMzoMXANSjhlQ1XgM1UNHP4n0pI+DoXV3Z2ZiVWO7NHNw1KG5zVTQfv3AVA3zi2Yd890svXPb6Bvx+P00t14h1lHsxuolMO7nzgLRy2iwA8buvru8NuW/kMoQPbJUP38voQSrKxiWvqTuXYn0iG4KxMOxo/g4oFJyl8GDo2MHglmj2Am4b5uqCsgPcnQpFGixxwaGZCHcxD25+YY1CH348VZO1L/HeB9u7a8txnLiw+vVhXOvuI92yu+0KMK1hEcxzZXEa+oRLhsPAKP2cfs8tCuybQ+s9vfByD3uRpODI90PLYx4s/L94VZLsgCgUlpH/ptUqvVbtwL2YF6SaZE0ppdxRNxR4Suk+XuyzEcCUK9VeLjDzDK5ZKWyezjct6pU3aYxO9g5mg7CZbCo2kEOGL7/b2QJ8zOpzF2DtBKOjTrUMSOuSalKjnvLdOZATvGPrGUDR79mCDM+tkkP0GZTzMwRfL4hH1YPKcWp4ABh2swrrcy93dpXVxaoy+i7M5Nw5pASxcYBVKFXDWh8WspUvRgt7wGx3uU1JltSoXG+2uGH1eIXnSBx9l9gTJkz/KEzF9NGJPDtMHPYM9QeWW9ER9jUShbUbWcv69LCf8s6SOO7+JuHUm83t7mr0L1RHPn+UPWEIovTJZZKG6MWo4z7cIn6c23WmhWwyZBZ7PQpWv7jK7qh+TEWBOa5v1Yym37tSX/wgGfjYGuFHjD6wxRbV5v+H004eZDlmoAOeNWRtQtpkXQv37MW4oKRt8jEEbFhMoYwBBkF7tDAIsH++gGjWtl3vYDcbwhts2V1nkM1vG4ab9FzHb6ePdeEZtw+LHu5F/nz5ZY3Y8mz0t5hy16OuqGC0A7R2C9boe8dckJiVcwu8F2ylibdfDXyWLcQV+ISJoEzNPGLSA9blC+0RliSlDRP3V3RM1PPegCibPZJy6iAcxFVR7qf4rQrtdbMZml7RPnmXZEcQTdIHClybyjcKMhVcMMT96FGIUsVgIaY0sIdJKBbuQ7Y+I1Crm/Mtko5jcMmmYyViOOfgJ41K7LDORGMNAjIPLCcNnRxtjxCBW01o5sk0ms0FmtE06aM9gw+2md2MaXphjrbGaDNszeHiCpsNwEVLOLgWluvxItA9amXHoWE2CHLjMCWNmuZFD5k+j4jwuicIaL9j0MVusQvFU5SW4+QXI3BX8fZoZOMeCx3XH1GseF6+tTd3JeQJKuNLkAW4X1xwfUsJhBW4pKfUoRfA2jLTeaJcU0raZvEKUeAAQoSTK20YRNmZPmc+K4Yho3chijYytjYw3GmW2odZh1CLG6eIS3aXc1UqJ8xinLvrWvalMe8yCyLdHN3J3Wwcbe1Zb918CL3y1m7OhHzCGnAbt8Eakf//WHk0r2aPGgzoV3K7hF48teevGzb4wREkTJVsfVNXt4xgm3bXJKCTOfFAfJNemiM/rRSa+09MsXOVlt1hcC3mRifLNelK2non1MFid4GzCmWeZB2AjIojcNMRZKaKgqK8sSxZ5BZGL0f994UFMhtgSHM+3TFttyJ55biav5oqYYwvHCClT9fV0XwXVN8YyeYi7OJ6vQ+pJYP0zzAVIYhltn82+MOTRCNMKizkwOP5LuL3h7G5mpI/Ri/5Mw+11jIPj+GnzIWo23DRBKHDbhKonDKjTmw3pias+xpyC4hgzFscsJYNf0/RfzmwZQoRGfwhbcjLBuqphvr8PfyrNZm3LQk9U1n4HV6nMsD33kccn0f4GUSFu0BTwafSLolct0U2hfMviEo/gFCdRpsarVDC25gD/3utRxvP1JkrevnkNEF3ODRw62cNffEs9Z5DSPVUH16vZbLCt1DnuH7LbpGKp9B0TlKOYeg1sHnkgbLRkgC03qZ2PsknzIQdmk+JfrWmRdvz7GsjNnliFP9YV9rIzx8LNWa09qVcSi2Utz6kRnxpCVJThAzqeumENsNryhlxP3ISEqKiVH2YVDs8JoqXcAs2wYCmTHMETcIobwhFx8arSfCRmpAUjG15YeqV6xIzmQZrATcqMXYayStxaYG7ZEU9U3y/iL1E8w2PAbP/7V+E6nu3i/WzqACbHZhRhEky4sMLIjxYRZroPzXkEa8M2kyOFmsVsjzy3Q57u4L5vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp79vp/9nbafngiLxDF9+UKReYD42aiTxjtxeO1hQkRP4s2B/2246fbcu27fOvZSgc0NOH9o312h8q3/NJQ4Xcycolipp3XWtDXQhkDRWQo3XPGMy/x2r/ncYmZFvT5xoNBuuQWGv+g2L2El2+Al7+24xphu+qLdiqsGp9qKVN3YHed3iAiXOikl8Km8yjRdbGid7zJlVXJYSoOmGZTCbmx6Gixm2htHmjjDlcoaEFCH5YA++Y+Wn1S5yF9Zv8e7Wi4j8jLYzncIWF5IiLmsFZJH7ump+4bDhZoQnYhXuX1sRMTcGOEK3APKyVcXC6qBd8IbdwMdMAot0RdN+ShuVo1lkjvnSY3RJzoWWoldTxgg+bQLeM6juqrztDUk+VwrrIBRFHrD7k/OML7+xhqbg79FJo/0v32+ZwkN26rTeD5HPZni/RdV9pyn6h3RKfPU2VwaXD9+KxF47BRO+f0X30AmvzxamtG1177P88J9maG8YaP+o0b1hbqB8TnhvLTYyAuPfZg5vsX1BlIGdm1zw15//kJnpHLbQ3A+sbT4qWAlxYva+Nrt/VkZ8Rzh8ToZ4eV3GP6DxU7IDMPs55Yq9rX2lHPb5De5i421h7XRKOSEM1Us/xd6vRRBdlNcyiD7+KSHEQH0lh/6DnJyN5fsHHJ687TL3Re5twJy5C/O2WF77LD/ql0p33vHPSfct7beD8coFypFb2k3pJ3yiLR3/Nf9oWwf/2b7SGwvzt52njVX4K47Ulqb/fq/q7Zcg0n8/43P9oIv/FQnOwCHWLPt6r/Z53zhHTObcL553XBZiBx0uZ7hrLh389IYwc90QlAW+Q86feiCn2oEzxLzcbW4afTs3TfMVN3OaRfZ9FMxNRlL8OuH+3Jdp0fOj0cJCZpbWndEPT4X+RPhvTVdUs17V8b2RoWMG9kgKZ+OlMwQzJpddLXB3ExaG+59CMZmm4EcOf/MhptzE5tLbZVMLk7kJxf7Rp6Re4AhH+EURP3rud+jDULAWMBP+eagJYGJ7grRwht+MEtaZ5QAEvlmxDZ1GvmvaUSEU2rOhg36p/QRe23QyG/quDxMSLCf0ocSPNvKs6TImedZtvNlnP38tz/rLtMcWGvFgm0Hg49ewwHF16Iw51gq4A4YPJ5Q0qH+X3BVTb62lYo/nMGX2SkFKes5QCqMZfkWLZnOzL3Wxj00hchB5JvX+AAfsQy7pWyXZG8m3KjbUhYDhqT2irImaS2BKRyjtCckymhGwnInaFDfw8C99zaFgj30wmWiCxRjHtQAa7JWbjLzjorKb0YhbelVor8w6jEpbOp9EvHP6DZofCAc0o2ZAw+y7ZLBwDAcmW0EKGPDNWCgmXxLjctgdmx5leBr+9XdMaiiUGfgluIXGomDRT4zFfuiwt9UC/UyBnyP6dTN8radn+tOQZZ8lZDDxV7CCYIU75hQW1pjBajDcOPQMAqb6Jl9cmwfOLvXhGRKcMPJB0jlA+9AKWQkntABq2J55/2XKemIskpA1pz6RO9Dr7HtRYEJvjrne1IzOhpoYG/DGpb1/VxRo6sT0+QN++C1kKmpzJ/mBE83WHWWRL2F2eP5+VpoqwboQ4P9AS0WWpIFJqoDOEjsEAfKTrSOms+4nFDHHJN4hfi4PP2MX4RGNTITs/9DW9Puu9Puu9H/IrjSfIVhB1v+q0NI+Snm4wS+N/ESYiX5vZTcRzKkhBwJ4hL77kBa/R6beI1Pvkan3yNT/RGQqI5TQV0scN386X0ThHj3rmdjz1GINf2iybhFh7BBhNoo0xs+S0Q+qpo/oGbvZePgQpvo2VEdb5qpyfNFIk5mPNJlppIn1VCpt7YCxBO0gyHcQ/EQHGMvhXfyzcScKHTXKePfFdKjfOdjb41C0yh9Jla2fEln3u7fGQNJ2C99TrZaSEiMekSNf5D2ImV5L7xHP94jnf3TEk/qgmQBnbjmQH8AFRzNvL5Ga/CMeIAuD9OPle+uQD/tKxIaFsjWqCiP/O1NU+LeoNoU+6F7fW8wWSdyIGtKgif3dJGSZSH7uhINMnGYd5k3997Oigvo/G58vM9Gy30BM6vwzBIHO2IIcOkopXyyjKMl/5TkdB6xtK/juOKCIto5DpW+uWMlif7l+/qZUNJOMOdPM572ZATu5SQutjUIryPcS06z/VzSzQWn/YhrvP5LK+1asJ8/s/+aA/cjHAOpz8kaS70TbxXWs/fsh9jQId+ov8YQGHUCg0TMMOwcVgW0KgMSAWdsjYCUTxDlYT+EM+oJVxprelAaArWcaBx8CgNDv2Qz6Gw6daSpSuDEm4KGrqR/5GCOdTV3g3cCMgA0F6AS/4u2YGG2k9Wh0ubI1TnhG+t2r1l0BiNqPHuZ+QcTLh6EfzsemTUPSD9Y8TB5HcDH3pw+m5+AVWukP4ZDXAM57gAVZhA+0AJ4A9vkTKOP3afWpYwYPgFB+6+buqBHJ2+JHe+hoztKfYe8YLs4+YAyWq+Lnyv2XfH3ANTBkdpJQHJoT5wF3B+AarD0nSNolN1sb4TesnHhi0uther0eC0ziNV7Tx5TwUxQi5piUxaNTLN4J61Fk4dZ5i8YBd4HBGiLSOb3AT05n32+CYmkuYZga6Z8enva9ycwf4qtMkk/VSPjdopR3sdtqrUL73MWLFXtzQfIGE1gNeuy/iJcie1OA30KGZmFNfMyctD9azNV0RXsRiWMxEJctXroPTcb0RP6de/97Cz+0RCUDSjEOVtKPdp+MHeA2zcIcFyezoTPmQyff51pMilnYoHT91gZa/66A5SE9V51EsQ8ZV/b9lT/FzvgoQJCcX6m+ZEYMCzvCqk1b8n5606rUkhuqITP3uPIUL0L64AUeVJJyoKCNJ0hG+QobT5yJmbnDAfJPKAFlxkRr5iFh+hYaJvvr54EZ4wQ2ns5R9NkoM1o8XoVI4WKL4UOEliIPo5kr0XwWIzEUh5FI6VYEBQ/6fclftUKxsKtsGopUtM8n4nzZypnbzPXHwlXrjtE7jFYSU2p+7b9Lqfe+g5oV+KBWyjQYRusbZU7TJBVZFqPd9YIm1amOni9ZPwj4Djihu/NJ9sEk2gUrgj7I6XX6pJgcMtk0sLlBna1kbq9kZQZ7Zd6lU2Xjw1RUkUffEhEyhenyN4OszeqSmE5WzBAjGyVDfElLXLNWhsrS9pxmcw05mYprGs0VcyoVs1ySXZAs6fJxaBqkjCLk/g1SXr2mY/jByokrQHk+NcfnSfHaEXhprWCAMsMSiIapF412UhGz7gMkCa0/byWik3kKzmSO73sAcHZeWC+WbzKPYCs3zfNyYs43Q1F4zTmX5SPVGxy2EuE6DVHzCHUSoE7j078KZDxOrAMz8BaUraD3sW/R11OMnwVYLdsMMBvcckBGOjRSM0YrBayAMVVi1GigQaiw9Syt9odRizMMi3BTubSFBXe4jfmG9MGi2SJqJRuA6YRfz7WUk7mtbShlIfQ/Wko2aJY+o/tpSfs/Ks29jACPsjX33hCe/Em5pWQeJUI7w0Wb4jrDT5vCP0P25nMpJ/bxT05zvCEzM5qg0txZ35Urys4WoBIdsa7Ln+TqD50MNdI3LYBNzK3Sh0wyyw8s41nge/4UbE+WBCExA1nasIffc2Hec2Hec2Hec2H+z+TCXDBB+Z4S854S854S854S854S878mJSbrFqfnKKfRe/LMe/LMe/LM/y/JM7QdY/Ikiol29hahF7Jz2CyMmdmoLL1n47xn47xn47xn47xn47xn47xn47xn47xn4/xbs3GSVJe/sOPA4vFJxBnFHo9CV9K4H85m7FviyAzx98t03uKPKhfwW3x4AL/SeXjA9+vS7oqlytwMHPQXw1ZxLhWSnRD+Ct6HyFlFRZgJGB6twt80Vb9MC1+m/IsQIai2xTQqQn+lFm47ha2wgvMwbadY2HrwriAWvheEpJCy9tCpCP+VC/+0bQwzABz8KqC4j82QJVDQt+zP5iCIaLAd/Q94HCzZZiPAwWPa4dyxfddHIqNv7TanwszCirg1U9nAwJteO4AgFn7a+/+O+8/mErYK/9pOF3RTxn0loJnNXTRYEfoickY8qLOA4IdgkDvBX8k+Y7FRSvS7nOjTXag1yYfPoZiQ/WM4m4pg8Dt08z3hAtEMAlhh0BrD2UQETWEvAiT5irug27biDOZxeX5+/bO8sg9jVrBuxZ8iURdlEWOo2AUyPfVFkmQ4DmU2C0ekuWe8Xg53Se0sofN6r/emtsf3aXybF6WbVLwsWIB2NFoJrFJmV2u/02KPQjvAwQsb6UPF3ASy8JWSpKL8FgDNK6Ab4mKSedBSKmImpQBvMZ2gBQY2S6PZ3B8w+PYx64v+3c/lMrSSi/3NfIVW5no/zVxo4R92m0+BKq3P/I7Hjh0l031jcqHjDDMzfAN8WoFPITQx7SBsUVKs0L/FgltIJuhEgW+HrTswEPL/v9/n7sACkxuTrKxg6rUYOVcu6U8RISo39KasqFVqZf1DuRZ/JZMhsYlzORFon6LXQG1UKpV+FU5NyxmD9IxMmhJKetfdS5acgV8pQCOOfcgITVmB73T6U1YjyWKlXZn8a0FobO3Pg60ZFPuBGbc2M2lyWOXZD0J2KTBtpqXkVudOgUfgi++apZ0dNV+kQhEG7aH4j0qt9F8tClGuShWq7Jo7uSSveSDSD5GVSrtFBfvNlSq76/J8XxpCAjPZBkotAQXK17DQCSYLBORTQfLhn+hRQRHLe1kyBZsqwpjh6+Us35kstpb/ZAVjnr4D2PwB6+zsPMVv8Ast+kmGAKEFpWnx/eYLehW5xGuGiwnt535H1dZP1OTR/2pWGSOjZGl8+fMk/vMkvdyl43yHqpffJWraOk+6yyxd8943aHv5A9J+A6bvkLfVSknjDi3ypogVlztAC0AMFp+/RSe9hF+cHh2FtkYCehJf8t85er1Ipb2Uhu6eoIOXDE2xBzsvr/kDSSyeBU9AHJEZPlGCQfWQ0STUE2lh6X7UQhsG1B5Qw2zq26ieqLs/w5zFAlAWqGxKIfNWorCwo9J+0AIdT6/FeWl/2JonHL0/rEQztGfY5zY6UpF18/DbkBoAv9H2YiG2+GZv8pZD2ldwV3CGnlO4F+fJQogYhx+WJIkG5KlFUyorJXET8t0Ig19oArfQOgMCMIdhMWdoraef8zlKyZRZGfMm+GbaGhnhyHejbcgQccy7Au2gcA/qIWM10JLktnAPZVnTgRVnntAa1H5gRXgJz/LITlH9CtEUxK2I3v8pLKc9A7bV2hsYxpZh689v+0Pf9KazMELxSb/mQk1n30mAT24L91QKwTOGWVoXmWBqgv1su57gr9tWaG5WkQk6jANCHUD+10gp7M1b7LusRokln66LaRYcVmAeW1IhxFxVpiuYDQWDpQu1thbhacVzouJ6oUSkr8xCpTWya4WV6FqlpXS5RJ5WQPF0hwDeZ5cvQT2fHxghU/w6Lnhkc5GCWtrLIJa3/7PAlwnmmK5YAeUPPsBfsYCCAe/wl92pya36bZ3bzyiAArdlgcXM2ACij0oVQXh4ACw/PEyQkR8Ka6GCeyIitW5EqpXAXTEDb3mn7NVAK2IplaH0qrRPK9IH9Kq0z1YlbVWjlEIzGvij0h+1N9gRXA7wNBg/gg/y4kyzbA/sUGBPK/g0H2NgCVcsSoDfpJV/b7GEKOoG4lWZwvd7S1FlTc/Ufi1NMj5bqwWDMrAekNrZyBQ+IHZXdFbwMGIWJoPtroD8CwySUj0fh/uTFfZR4yJzolw+Ces5cqBuqTJyVkMf1jHCsZPeRbD2A8d5cYSJH06ohyW6OOPXsn6fKoKORFUHy9ejooR+HakQF0oYQ3C3+LGVi2BmO2F4MZuNuyvHXoAZhsr2gWmesKXTps5qL5m9LwZ5feesKhNzXmQNxLsiIyemoEqbX3pjZJZZmVL6KTmXBQCKdO2Hi8k8LAag68B9NgGosFUsiPC/PRCB5QINSOy7FXe8CEfF1FL1f6uqrVZV2ZsHSJoZwvbLilhwxuY8xLNP7PNdW3hGxUMC0GUL4/vYLW/Sel0XClFSA8bLBWRkk5NnLo6RmcufVFoV9ihQBQpVYY8BV6CYKOwx7ivE5nj8EDqwWEMQAxyCb5TW2bwKnfOzbkF8xbe8an4Kv4i/jB0zmP6dM3xboikbScsdBxMFnKn9vIu0ylwtLzAnwpkTA66EA+NapGmX7GPl6I7NMd0WbdBd9tnsIbTxx+Mkt/4abQzMR2UJ0cl+BN+vbVSByyuVRrWmKDytt23SNF+a+r4IMadyNJuBoOFfCN9Ic0ziPhgDorEfFumhwR8aBrJ8wH20Dof+m6M81AwS6ZEekZ+kENcntr5MN4Ms+2e91tp82j9rt856ZSUxGFFmMpUbzGLwYe65gkZVseZCikQRcagy5jNfBxnMCkLLJNRbtt+GGKM2IAybuIJ35p2/55fP2q8+eydTI8UsiWdt9p1DrtKwdTKVrwtzGiGeQSiF+M5x9tkavFvPK/OhvbMemwv4OzNMqC0G+e8wYgf72DrJbuHVQtD3S2r8I0zL0q4isoudJ0mqqvz7fE/Z79PCU5YqycHGXhOwXfBLcf1AMY5DPOEXip44EofOHCR0VRyrLRU8QBhu2lJ05gZidTYldtDOLvpDNOP499UOWniKz7vzM2iECqX9Q1owel3wocVOKdJxivBzIBUPy5jDSNdo6lMKwmFL+/g3/awlzJ1+wF78ANavZYZO62AnacvNHADrjxadDZ6WxX5wyN/VHZxScnZySl1JCxRaer6FrpZo53UH4oiL/6nH3Ouquj8drS+ncIlXqb+VzJJvlFotG89NTr07Cxwpj16P6PWIXk/pNQvWeOPWaIzHAqdjdp5lc2XpG+hhxYBw0g8FeONyC3p/ut8f4dUIr6Z4NYWr9AvK0/HvOP8sSnYzz0bj32v4e7iLVyDogbKnyQ4ypny3vPGON5aKMAagulw82PWoIwu432XPdnE50vGwzR8ylc9FRHOypYuD0DJ8iDlNe3QRivhMfBSfklXMtskv2gNWa2H5vt3C9QFk4PNxCwnzzs8tAnaF+P+99XSPBAveHLo4b1f8I62HfdKzrFP/vnX3KHK832EFMQAC3M+ChV2yj4mnC+UBNYOFvp85F5vwIBq3xdRhztiy91v85H0qKZng3M8ZplwxpZbpBJURVTVOIr2xGRpI+PvA9E4iBVuZZ1x6FsS6hpK3rimqjhiYUkakommt3bEdWilTMBO4FbDNvs8bKlQsbhGZ+7iErTsmoplqq7Cfh7HjRgm5iygsSxvScouERVL4Fb/p4M3wm7EzmHgMdgKoBDH9yCw1VTH3igZfmY4Hi4ufUsrg1wLUcLsfXQ6wuSNM03ICZkpxnyLjC+BncDe9fXzmrGClQANxr7RgUxsHbMWf2aRav7Agu4eUHoPC0/Y8jlPYY8qan5jeu7sXC0/+dAgDgZmTnsNm59HxEDvaXTDrB6SYwl5FqdEvpWIQrnUX0FOWG8ezAzHg57ODLQe088tzn1rkkcOtn3SlqKeWUL2aCrN5MLOAGJIDm3PW45zlrDrD+30P4dpV7nfnbCxxnnwzF0cUafsSSMIW/f5xRZZVcb7DjtFnu4JaTDRGQWu7Piyl4hc/NU4/nxHs4eWddr/TAkzR8vz8Ey0VBfzkKcXjvLwVgXPmpiDoWE+kBM2AQorZbqOzg61bfY5kSMqhOGKBWgkbK5Th2O/5GAUr8J2gIOLyF9ersUtxj2L/7SmUpOkrF+UNicWZadOZzvFY3pveB2fCc6LW3Q/oknfxXcpMaURJzNoiknwJySX3oLSt9a/Che/Y9KAU5ulgXCVKGWyXMhjjz/2c+8BPSYKvMLMXoTBbOgH4RNDQS87tAoUvd2L2bXcxTvDL5g1kec/UUuu7VlyDWnHciANTmtLGYsqk7wsSJaXkvfsNAqed4xLg5ivIyodMm2Q53nIevy/SuOTK77cX1oIqM9D9t+8Q+K/C9QiwPMLjvHvCQJUF2/HHIML2wawxlw4+VFQB30aCBkTkO+FuNNt1lg7D7gZDM8MOcaHcs0O+2F2RP9lRZE2vNeoliV/scwHQYryTiANNbkId/Iv9MaXwPyfjYdlSKf/ahd1Kk9AdRzclysLen4UoRSt0w96OgugVE/yCKo3xwGrZn7q4tfXM1MfSydZXVHEr6tfgP5hAPqOJE/k2tAOtbAbPdY2+h4b1Q9kufX/Ew9wvldND/hR2TF2eI9SAO9MFWBymzlLTByf07dsGJX+fen8kVp0gaN1tipvdJVdBy5wKYpzK+PQreE2Ydv+qaeZBhvKhZJuomfNeskLplU4p7b5VvDHAthFANN/9ld75MYnvTR1xzwXfQybQ+yMpgl47EvXaDG0AdQOFgq2J7AZlzDMsrckPdQFlRfn+d3mDw6FaAgV7KVIwCTHGtc7khMbLHbaUdDaw1qDCMu3QWjAtYILo+ceth87bjTFqCeuI7dlpDI7GpCFnyoeEY/MNvm6p8ENwvnJwWLs3gZk7W+v8sPt5rns0oZ2HWQB2u+k+sAPcmNbEl2j5hyLpSlPNQfdN9DE6F7V4+j+3X9hSi+lSb1lD8SfWMG+H5GKTr+2e0tt7EMxdY18ISRysv5bmmEj3RNpvlfKVn48B/k8F/xJSSyN/PM3rdWZVGtwDU9D1V+J4JvJvIn4v0Leu9VMhvt/CNL6XjMPifBuBvv0tUT619naUT629GeXjW2DFF/GZRYKSLI3vJWlYrfG4hVqDnmnCGivxOZGVaU+ZV9OsSvsWJuLsPtPsgPG4TCNkK1Eu7a52nsvZFyGl+cEoQ/AlX/swUpKqw9IHwHEdF59LmFuxkTkw55kDS545sKSZA/P9Jc0ceE5dfgAxCVz8yeUv8wn2LDDw8fVn41mI0niMtwDADIrgh97ZjiIzpqdiDobYpV7emkMRKvQUCgF4B4m8YZuMd3Q8hE8s4BsFHuZ4p9xL+Agdavqb7tSxN9W5gWkzUwYHeqPqt/z499/4m9UokxCRHTPut+5YHGcWl8p3y/QrkmtdhzUTavbzsTmLO7JPrfXZPl/kJWLmYN/THmKmf/d0f+eDh7Hfhx8Rb1v4R8RbWDv84ef5MH+cqozfMcOktheY6Hl8QBXdReQXwXiaeouxGRSoIMWmrbuVxKTiCsHER6lP/rQdcBzn6b9aPo8t2i0O4j6F7W61a+88b9JzCjHFSrL3i9/gBGcx+xHOPn8dXipDi6sw4QV75ritO6UiKpSrOHqzfMV3ApIoLPr47GqzZrWUzjID5Xqo3CuhkEZ3LFrZFK20sojwIGsNW8+78/24xaMI+ShvMl4SiQUclFsr+LvDj8FuxKeg5oidWOfVVnC5E6dxktc1scMKe39Q5ADYlFRHoodhM0SYWaaB9EqNBtLhx+IGXmYyWFPE5tn1wYff3XRfBx23hQb3n0MeidkwJ/cxUPYj5+d72jB1eLabgjRrBCgduh5iCgNMG20DW1Ak8Pvo8ZpdfOlJuiuHr580Y9DhXI7TuNYG0EgZ3/bTnJI/kySPP7+JLB3kz2/ZRBJsd5eEKTCRn/kYhfu9tf+Xj2kUsVm5kO7A5mIaK0AltavFjRclguWjVDbt9X1YWdFubfIQjmy7XutPnnOyx7pY55mA3DbFXFLJnimyNJI9ez15ngUCXfGoZ5oWwpwtboIh+bjYYBrhGlGYYHUW0xSuYQbl9IVtqztOLFQYrRjWAXnbGpg7tIm1g43K9qtm3/IpO8kCcO5OppKGMUbm2C3cfx83O8A8b6KH7aUXWKw8k3Hxd8MhqV2YxEX4cYSU+NOEpr3k6jtREQra2vv9PmiwpBmbe58ZxW+WU5sX3+EU/jV7N8vRab5V7k0+zK7dONuDBu9ftGm3W6LIYEWa8UKFMxVKCQu7NCsGy9KckTGGEXyKSTfHnZUQ5hFhcZIoJmSSmkLHFnegXmBm+R1rZ8LHmHnGvBJsk0s6S5N9qFRhb+yh0d/QiWAW5mLM0/7+/MbPXSFD4pN7zpbLtUTPZGXOTT/A/WV0yun4uDVHhwSfPky2n4s8QY/2tBvcYSVuo7GSROYkMJf2mW2InZT2QxYwc6n3xXQwmycvVrPlO1tqTFqhNF3Dn5iZ1OZD7xH4FLVA1gOlJ9HEYqjuhjvQuASCdVdBV4/utEz/ULJ2YQycEtWoQYnzAYMUHUnsd8kswZBBw9GBQGVNtNUeCp/S3V7t/p4muT2wLuiVCpeY6vYjBUrpMJOqdo+ZwaDX0hQAugWWHCOjJa3MNiJ74Q88LdHX+rLLsoIrAWOmvUyRVnGXhfZXoYEbXC8qHXjnoA+ThIKg1GrR9nnVsm6XcMxflGJvZuNwbf5AJwULsnVaLM7D3jFEizJC8LX+pfBnMzeZUkZtLXL1fZ/bxkbOyeSJpmptGT4kKZ2bXMP7AUigEksLfVUFH5c40dJsy7011XPWegvLUrq1+o0h/I4N+MBHu29lh+PA8BH3aXUcNUTFhuP8nYFfo/RbGij5jk4QWArwYjIxg2f0yKnG+W29rKWf0z5UuyydwHefH3Av9q+lgGXOjG4JoWSP0PFTcmu1k6gbjKX88zEUdipw45ScaPzo0NvWk238BNfIsZ+Q2vkbGVv8eP8bR9S4ExqDzkvzzdMEDL4FnTWF1onKGZuIl6wzyjfMo2RnMZtSnlhMvIzllH9LX39qtpKTd5ljMdbWh5jak2RY84NN7DWv7Ia9i5ffgP61ndZGImtyXgyRw0/pgFGbhJoYOBVaSJ/vJwct2LNdRWw2m5i9YbdSgLLVE6naallJnMT8m8du+Kmb9aGbFLzvnLjBTsQ5NBUfqWmQgv+9Nmhh0HY71QY03anIalV8LDfxH/MgMpP90XGfDAbW7+w1N14enECVe5ppzRuhoZ7WpSc/Mz1O0pJJ9vEyfbzMnj4yN94gvQbTTukweT9whpbKrW3wZ08vZWitspgP0fUCjg2KRf6K7I2mIAKdqQ2gFHNnT8wfZrqP3zxTkkkTen2oJJ/KruRS2bdHYO8adZplCX8V+lelf6vMjbVbVOgkBz4z50JsHl7l2Q6pdYVlhbXfvWczPIsFiuCHXAl7mzg7n7C3PiNSoOh9YCyM9Ri6M9nn3/4l44SpDFCiXPplMLoWlBkTkr2PGwTs0gx8PDnxgC/LLOyh+EVvN3n9Ljvp/OB8XeALmnDTg8vAhAzFAqIMpgp/31KxYNUV874MKNnM0LRDpmo3gM3W8vm7h3+Q7U3Ho/qXpfD/nfdE/2DT+jvbGdk9DNSfC5u9Wv6f18XUd13vZphDkRlVW/KWqUvYuku8fCoEEVrQi+X0KcV6mNg848Jv/mboTsMw29r4Z05oauJnzP6cpc+jXDVZ1RD2rNHuRNRi/zNnAYpb4k9iPuDx7W3z/lcgSrDWWEoKfbO2j++mWWVeLT9bsBw5dspkjyVkMtwl+etofgrLkOXB4zsf7BnK/D+neTs5taPvMr4dTz1hVvsbE7j/tj+dTWgonXVN3TN2iaeg8I0PG2fw3ojS5OJtGCbbPDZ2BwPBsmHzKWLGcb4fRYE+3ozHFLedzUk6Sjez3gRznaweBblTZGk2050MoFotq3B4dHWtFMqMgYATQFwXfj8cDofrrVPaQcaGo/d5m40+YqYacg09QJR++IUnyljlVn4Uyz88xOM+dFOJpcNAmzRjpkRbr8uWLEkmQU+SlUFf05DDjEWps4uZmgIOBa5kmqCJ0eHAn4vr9xojg7ASekNfTb4+a5HkZaEkmoJoECM8PJFMZDHF/h9QSqRYs8RayqrTFk9CZ+9+LybtgRWLfwurvpjkJazzy6F31LbZL+1sBY+j2xKrSllRdvwUyjvaNElQYl3QVKR78PFn+H0AFim4K6w/lYXxxc23m4Dlzl/eA5fpa02SHFF2md37RjE0CRyXXa5fWc94OOkzcWn4CMmbOcRE6EqvmGBdlEsDzLTIJvSsH+e4dv34FR++7ggBAn6amGCYYSr6XzUsNrczXmdu0QN4D4vILuzRtANYYRcvioXfbnd/m+z+Nrz+7XDvt/7eb1efeWaCN6EVgI8SeVTYgwt+R8n7AeU2ADnGrJvtshnTC/kpJXMyAw7hc2ackhG1+2DlcklPdYVg2njojn6GK1kiKnbBZGKSmB8n3ANhv/UE4fQ7JwjXop+FHLg59sDSwVgTFgmySiJPEnuANmAhwHyp0BPKwmlXWICAUeos3YeeMGOP6bsF6pqQMGj2fC09RgvVaIaJwPhqT+AJbNx2F3xg910FMwpdfJ11mg1Hk+GENFVOTADARHUxQPpY11ZUlgKH+W+0pzSXYp9KLQaAogh0uvvCCLAFq1XlDwrfmKTMnyx9I32Nk+534/pbKlHzD5P1YEUpi4T/HiuQHXC7ZAMldLdLv+pAXxGF2SYTGJB+5WZfcBeAHjZv/jE8oEcXT/txkYDpss6U55RX3jg+R+1Mblr+vWyZLW8PemVqZLbBttoaP9pTxE7Q0PwH9pQ2V4OhP/tKtdy+0laA3t5k+qm9T0Rla3O3k52oQC8ai9OjENng7Cu3l9bMIRJ3GNkJeBbbxJSBv7vdiX/f2PHcsqf4j28n0uGpe/KXTpJTuzUFJjnVnWiH+3/Kek0mSS3rf3AfE72CFOM56z/n1ryeV8ImCWAOfkvyx+ySncebk4XlTmk5+2qIveT1Hv+Ojdy/upn83RkI/5i3IUzCFr57h45X2p+EWf+o9efT3pK/Tma57SUaNHEo2cLMrym4cRkImKr/8awn4Ztz/o5ftP9/2C2a+CHmeNFAHgUHhKQ4+t4rHujnRtkLHaBkLYjnaynMO02CeVgPc4tS0bz9LRHz770eYiSy0Xi+7Qfc38HPBJlL0x9TX44rdwouflGCgiAUU7XupJqevS+XzhatAkbo+PZM3ognKGzxDcCO+fPLL4x7v/yy9+WXf8CW+fKL+OWXNIr9y97dl1+ST+3Som0f210XRPSSf3CXXWc+uUsfbHx0lz7LfXY3fZJplP30LoNw437z87t87OwHePOP2Ft1N6r5G3X8l81Wr7+pSyukn+Kld+uP8eZu32iafJCX3w0zd5sf5U0Rnf8sbwbR6w/zfvnlHp5SsYHLiNm8u/VKXVebjapeV+pNrVqti4qo1MVduSKrNUWra7pWbdTuxTtFrFYaaq2pNqo1tSnX4FcTVbEhKpWGomhqraErKlZUq6JcaTYVFVvKmqpW69WqWBXr0IEM5U2xqldqTVlpNDUYsNaE/kVNrIlqRdVrmixXm3Kd9oTH3ivs4Ltc0fUqvhTxVYlaqTW0htyQq8pm0XfG00UFwVS1Zk1RGopOp6jSlk1RUXH+ilKvN5sqL9MAK5VGXW3o0KShg1zTVFGRRQWHU5Sa1tBrSm0L3FVZrlUbitxkaMyUIY4b1WZDq28vrNb0ajNBP+BZheo4MgABbWo6jA9Q1aBJI0V+pgelIit1Xattwwx0DpNW6+nUYdErdU3TZeioVqsp2KOoNMQ6A1Nj651QAjxSq6qiqhr8qVMiAUh0UW3iCM16U63qKocIl0EH1NebekOpqXKzCSTWFFWdTlFXFLmaIg7wWa9Ad3JTU+oKNKjKQBawxlVaWWVo5yDjDFVV1jVFA7w0ajCIqCqiShevBtPQ1GZd34bXZhPbvForKAJqaMrrdgCPikDCM1h6ta7L9ToQPfy/RtECd1u7AagamqwmCEMiBU0oq3IdHgDa5Iao1kW1gW2Aoqrq1iWSG029uqaA/BAKQKXU5S0kt6tUmnXg53RpkUIUpAW1Xld5ETCFLGq0pNpcM3lNrGqVqgbYrGm4wMhwYhVYX0dmUZoNtZ4sFWIf1rOpwDxVqN2E3qqqWKV95pCoIWc3dLlWVzStoWhVIDARRYKGnaq1ml57Tf9QAhwPzLOFshHohLxUzn9VBVZDByAQkyKIsWqD1kvRkG8P+JDXpJQvqwGpqhky2wXGA1EIcwfcNZpspk1RQ3GhyVnGh7rVitYAQQUrrcqIhqqoAaKx/7qqZWRNbsim/Kb4gKd6vZmS61a04yCIS5Deiq5so0kAVKnKb+C5WstwIEpFYHa1Iet6U9VVFVdc1OpijZY0NK2upmJ4OzQNscb61fR6dS0GgGjr2LKuw4RkRUcq0HRRa1LIQTbo29kAINcVvboV9qrcXMs3hKamoIxo1qA2SjdVrCliTaUM0KjXt/ILlGnN5ppb8zyYJ+Vt862B7qKcLGtKba28tssOmDOgExFZ304K8hskotSaWkaqAlODWNX1hq7WZCpR5KpY08VakwomEBy6vo30qV5vbic2qvN0TV6vwxuio66IusaqaxntBbIXVrgByALUyHqT8nkdTAQ6RJ6Ot3JKHf6vUfoDGgGsbVurBgiQ1CQBFaCDBKkikQJ/agpIgzqMqdNOQDU1a3wmQGdgptQ1GKpRrWuAsybUq4t1unIggjOw5blDbTZ1WatvW6r6miqpkAM20mrAMw2QQnUFJFy9KTZoTaANPV27jeUAoShvVVU4hyx5g85RGpUqkHEVTAAuTkSg8QZb9Bqgs5qCClKn0qzWQYnAcgPtgPYVG1WxQakvjxw07YBAgSZgjfRaFXQSVNTERo0t3FtSVMZZvjErECCg6RPSkMVaRQeJAlQMWkCpNUCjNwB2agSotaqmNLbr6px22lDjwNLAiAkhb11hXRZ1aq6AK6o1t4r7lNiZ9QdWlV7ltP2GVKAM9oY4xemo1SzzMYkqNwE2MGgAI0woinpNbLLpozNcb6xlzBYu0utik81Dq2asjq1cpDdEXWdz3m6QZml9qxzRm2KTYV/+jhkF4g7cgW1cii25WZhyh1JBwxW0Khg1Dbkuw6SaqtisMQzUspSE5AgcU9dAugJETeDshtisik2N21Za8w2LGQS+ojTfsO02RgEfR9PQyQHsA5aBmhAkwHODEhcI8lptm/FBrVbtjWmDMVMH/ZAQzfYxwMGQ2VT0DDy6qFdAWaCuA9rSmlXwwjA/DlYR62ra2vbaGBO0jr52jWBFFdCDMkjbBozLAIVe0FuhhK2BMb6NFejEasCMza3oq1WrWb3wmmpgcvAfI+l6TroAQcOagGpWErEOFXF2FNXghyiqvn1Qpf42m1VBcObG2GQa8LkAFazym7YkhbZerW01BhElWlNP7aN79J/xpRM0tPKTr52grnj2xRPQ+E9ww9NXT9DOfublE8zXp+f/c23eegFFNnCTeQUFbZu8hIJHZ5L+/pEXUXz55du3XHxqY7uIBal4sAsjEeCqgqQBfYH+v66APw/yQkFzB1gNaKdRA8Obyhy5ji9JhJWCyqCXa7ggPP6FHYEpjZyAarMK1is4XdARWGp4icSNEpgaB0Aj4L0CH1aB0cDius8CnNmNoKBO8YvTaUxN+FFQ7ad2yBHvtN/1PjnrW5XV+q6i7Crytaztycqepn9OK/P9AVYzuzGTr7HeKWc1/62b5enQuS1zHPhPdnonu0L4+SbQzwqY4iBdlXoNrPKauFGPz4hWroJ4qIJHKssqrKZOX0D1LUGdP94YaJ3twSbu1uqmYw5tXTWrVrPp2oo+hF4aNXdoq7Yuu5ZiN12wFBQbVEe1YblAWYptuXXdMmuKw+eGXW/mjrABzKFugYduana9OqyDZtIbFpharitbDcvRNCiUh7IMQDhAg7Li6m7DbLqqBbPHtLfMAEkmCutYHlYbULVuq04TwFdrlq7bVUeVkRkcRYGpgEw1HWeoDxvDOgzvNNWaM2yargMd65mO07wW1jM49nVbdiwFDDJXVodWvaG7ddMCg9UyNfjj2vUhWIuuqg+dqgYKzW7UbN3RdNM2m7aW6TnJkmEdO8CSjt207JoDyACroVlX5KFZG+p2DRBRt926o9bdZt3S4H8WcLZVtayqYrmAG8utZTrO5txwYlfNOgZDhgByXZVrmgsus2zrIANgXNUGMeGCNwjmUQ19V3jqynbDUlSoBhivZzpPMnhYx8rQBZgbZtUGww0WDipXh8OmK5suQm0NATN1V7Ud2Ww4FuoQzbY0QGINgLGa2RVc5wOxrpvDKhBcFRQn2JJV16opljV0NEexNOjVVl0dpJU1rA8bDc1xVNPVgVJNZ6hUXRnHzyIkR3V2zWooiilbmmvB/KoOqOKmCeYW0q+pqQ2QaK5mQhWz5lg1xwYFBvPTgBUUs5FDRiZXifVdd4ANNKeuOIrrmjUXWMQBKamow6Gl12tgccPEm6YJdqxjabprD5UhmLU6Am+CAM9SNM98Yh1XTbthAxUrimO7Vt1UGi5Y2KapOYB+eTis1dwGPAfMWlUTQLbAmgVCN6tDx3aqepYX38qjYgNZmgnrqSpD0ARNG2NDtl5VzKYGlkcNRH5Nd4eyW1VNoB7Q4cCWgBqwicxqtQF06W4ZKJuVxWdjA60A66C/1KzKlq3Z4GSBZHPlYb2huors2HrDbgLGdMtyZNmuDYdDTW5UgVYdWds2m8w+DhukUR02tKYLRg2Apqo6GCNVwGHTrGlODejUsap1B8OqgDxYIhRDds3WTHfYUJy6rW8ZJLtTzhVOzWy4wPgwfdNyLbeB2LKHKuIE+7FBzQ91t4lhFQfqNQAaWDVgxGaj6Ta2zWRzZ/bfti6ZrUMujzUbfLaGrTTBCddNMBVUE1jbBoYegtXQGA5d8Od1kPHq0K0CD4KQUqu2BYY7GJlmFVNq1grmdYIWjgI2pMor5LK0GAT/fyRqIZ7QZEpYHOykmxvluj/o9waDA4+QmyNCBkcdg9wc+Eer24tmWaq9SOGF1JX0hqReSL5UdqXGhbSUpFOpeSEpkmRJTUnSJGkqlZdSQ5JeJMmVyhL8I4QYhECXMelC38TwyCEhHmkPyAkhIekQck6MLukR8pkYGj58Jp0Yaz6RoyMyI+TggMy68mX38NDv+lH7/OSY2OObq4+fjy/JRPJv2ycumX9+/ro8WT59BUuvf3qmR+Xr7vLCC8lB3NZI14M/pzEZeedG/+ZKbXfko6b+NHz+VDv+ZE8az5+9Yz2enrV978Tozv2rp8OT2dPXo6uwf9r2lucv1ZiM7W7b+/BiDEmvT8wzY3l78HVgD9qXR32j3b/9fHXUPx72x7Z/e3xcD6cfrh6rJ0fa/PQqkE/mYVS9HkpGn3QvyceGcXNzsKTT7R4eEK0brY77R8c3j1H7cnrs2uPhsxUdj7Xps2+rJydHs+7z5Phk8PR1cjXvn6hhGLWbxxdW2D4cXHvGQb/3Mrh5MebywRFxHw355rhDvG440rTjC2180Xa7J+To63M7Ou/f6l0n/tgzpt3DixEsdefp9nI1Ojg46N96xujqABbrduQfw1qQr6NLrdcPbz5fDuzeRf+GkPZTrwdrRYjbmXrnvUHjuDO2LzqeRsixfX7p2bBSs8O+51wYz/rhJDZXRkzIxcCpw9rDOl58Ig250wEiM55uYYjOmXHbbQN5LMnXm7YVH0WGE7YbXh/h6gAJEtIbtJVBd0IUr/2RXBwC7ZBJ/KlPBoREg8MXcjlozwdnH0gdBiCdz8Q6AlrrfCWfb4wxOe0T12t/xf56cmccA/kdesaKHChk4HVgvBXC1Y3PfOxv6HVf4Lcdewc+mfahn7ZOXKA+aE187K8TkdsjYxwDXOeyYWJ/hlt2vYNAKh+RWtzpYuUyMQwJ+jsedCOk5xep+UKOsJ92l5zPDARUalpSgBwC/HKEjQydhEsZLjTS9jqytBqQxlLvE/rPOMO/saQfkQtgOVfvkIsG8FWnKemn+NuQABZpqbfpcw9rAygDsv53QfmQ/pMknfzP/OtIUvx26RnSX//2kz3yv/bkwefAeLR6UP/jyDeQ/vzLj3FXDj9+NHqL7vPtB8DDSa97A/TXcTtt+aIfx6ddvztoD4D+yODikgD9nfcPFfLZbh/Petfe7QnSX0RMSn/O4CIk6lPn2RtExuMMhmjXjVsb6K89IVOv3SH9sdEmne4AEHeidTq4zmehcQ4rSaZx54mcn+PiXcQXJuJWig+PyMe43SQnL8TvgxgjdaIj3cCoDtCZ122Setiexn1cjk7kIb2EbXvQ7ZPJU/swBrrFtYqPJ9jfZ6/jkKvQmJPjgJg30J/xQqRbXP85qVel48HxnJzM2o+4rieecYRtz9z6ITn5KPmD9vmAVLG/Hjl0Ee/NWD6X+l57vtRM8sGC9Tdk0vFgvp3TxrnX7S4bHXJgQb9ItjPyvAQ8Hh4dmJLcJx8DyR6QM+xHvmiauF4rV9HJiSVdxp1zSaELe0POl9ivS84C6TDuzKTVEdA70NfFNaPD86nUHnRUKZ6RsqtfUGrmdNGOO3UptvF5j1xQOi4jHcPvsxR7BGynPrlwpRNKxzajW+OW6JJ+Rc4DqecZF9hPbanjshCwYo/JBSipQSeUVKQJcs1JDfqvSjAujHdDVhcIXyDdUPrTbw9W3vDAfvp0eflYQ/nnXvoRo79zo+fqV96l7feCW5R/H3ofUf514o7mncWjzqQteX0F5d9pfPYJ5d/R04Hk2bHhe8en8aQNyCYN75HS3/XgekaaT111cH1KdCo8Vsa5jPQ3JY/dTntw0SEzr/uI8upc61wh/R3H7U+D4yUZaZ364ChCnB7G/Sv8nXm9I7LwjBk5rZL5UYeiVg9RblWh0Pg4OBqT8Lbd96Cfo9v2M8oEUK6n3vEJAbofekB30E9jcDxD+ruOz29BM7fNQc8hE+SLTggyGOXpiXR9224uGw5ZzsgHlMtnt8Yz4vH8pjtcNibk+lqakh6VfxLqW1wxyZyRaUMfxQdL6bmLfGGTs2kD4JN7fVdVBx8tqQvE7UL17gxEdRdBr30gZ7Z02zdOXOkU+3leNp9wnW8GZw3pkRhHUtMnfaRn4iA9QL+OJIP8vNA1cv5Jel4CXP1PkoS/HZtoru6Qc0sC3cGEuUcovdRd/ZxcdKSOB3K0HiD9XXF75Qnp6yDujCUwZS6Q/toxk6dnn6QDr2NKzzdIh8fYzxhEMsJ3OOgspNUTqSHdnrvYDxe0HQ3prnzB5HePCr8LCQ0hcq5/6q68/kFvcHtjeM9If1eXoz7W8YxBvde/ubkYXX7unRwB/RnPPUND+vvY+UDOq5763Ank8w9U/8rn1zHSX//wPHavjKp8+Dl2nlD+ucS+wgkDqX412l4X6NAyDm8R2VdGn8qDZzDD2kfAzORZ7swHQH9HMEmkm85R+7N3AuU37a7XbyIdG4P+AunG9k7L5HbQfvROfcAs9nNFfErPJ+Taa9fJgQo2U9tA+XcCep3SX9yeer0T4vXbATmm9OwMTj8weUpeyGiA9PyZfMR+jAPyhPQHz5c3JPAOH8lF1/iIdHLSNRa4HsfX0tHgMJZeBsZjbNA1ag4I1Zt1r3MpNY5gfcpgX6KWNR7ZyhsfpLFM+lLzAzm9BtGJz2TQv7BcJ0dgUDY75PxUAvk0x3XVcN1QHiz1BchbIFmQU80qs2svkG5gVJB3uJy4vocDJhqpvAK5QzyQd0BHIH9G2F+Z61/pAvX1C5NLK4/JKU4wVF8PUB4OiIpwXTxy/d3FfukswA74hL/ttV43WDuE6xrpucr1b/KcpL/QjzHy0K6+PRp54UE7vu2sRv5BV7sdXvrXB2fap/alFxwY4c3Z5cDqHXRvTo3L517/6SYyznpduX81Hp2Ne5b+ITQurc710Zk1kPXO5+7ZJakOO4DBHpGmndXNBfGeG23v6EwnnyRjqPeOCfDi9e2BF5tVY6D3pt5V2TgdHPje5NZQj9qXgxODTEAoxieH5OmprXont+Ql7j57VwujPesO44FCQD48xQcNYt8YNa/7mTzPOuC/3JA47lygnq8ddfSB0SGXsjEfHNwiPXa8o4BMn9qrGH6Bzi8GZ21SGxjaRfmADPrGiBx8JZdH7YF3/JFMb9vjAdD9o97pEZC3V4SopFMjHdkIvIOPZHhkNEn3mUy09lN8opIA3AACek4DeQP6DkxHcF4OwCSxDW3QjYl31D4YnN+Q6AjorHEN/gioYjDLr2zjI+k9EcszloMezNduh4NuTYoeQY6VR6QzM7yYHBA3JoNBb0TMvnHoHSnk8enkSgq/gpHYPCYHMgHxoHmdAyAGo0bAJLFn4Orc1g31pt27UENy5oEdhCbFDfEGnQk5PiJfByBtjuSDS++6Kc0/SS6wEOnqpD8AoXooEysG6XGkGTVXmhsjuadI2inpgggbgInbDckKpJ6kHBGTdB7xXo9vDmEdSFWSbqXnGPSy9FVa6QCidCCt0B4x2lKsHS69i2dJqkoHpPNJimPgA+Ys9gagZ176RAcaNc6Ax2OgX/hnDJB/ppIRox1xRJpLfUB6LqiCTzLgFe2HI7QTDI/ZGSB/P5P+o/TJ65xJUQxy61Cn9sQS7YNHqQP6G/QRUSS9DKaWNJQUwJul0/EAHh3rnTekbtzpSy/gf7rNmJyGHVjHc+I2EJ4a9ldG+X9hSQeDzlR6BrkPRn25+ahTfkQ7PO7oCC+z1y2EP5JWXbDz2ibqh/LK/dxr3n5aGsOPBwc3n2ojp3xw3TUvVhO1V7/5dHr5aPSetZvZ5fmwd27fPBv9q97MBsNQv+6eHl01Li+tnhVeB5c3nY4tn330VuBRyWcXJAJKsD8cjfrL7k2/XyfLF/CMTlbep5nhPPWaA3NuXA0OG7H5mTi37Quv3zGetEPZs+T2cb99AMKV+GH7KzkKyRLti8tT42zQvfUuy8Zx2J0MrnRJGRhyfFAlwxvj0+BURj549k6fwI7unMTnTVJ96tTiUwf8IsOPDyRidY2rGPg41Nr1+FQj3i3a04cgtKQb1Ms3tuHEh2fEuTF6YDqB/9VeIl+Nup2hd9KTYhn1w5JcycbF4PAU7WvDO6qineKi/lncdJagnyRJJkEMpH4zANV28EKsJ6M8OGySjzCfwalLDh8X/fj4VHqWDcsDuu14xuXgYEE+3RiX8UlInCPw/z7O2/JLo+wdPEvVJ1L1QK8BX117B3PyCZhkcAIuhdeL3aFDxo9onz6BnwqmWOeGgOi59A5qKJ8uvSuPyJ8kh1y+SOFUuok7itQIiTUAfr6aGWceGD0f9d6jNzOJs6x3BqePkqQBHekROdKB3ztA10+gcnuP5IN21F7qV8bRp+ZHcjAAOjc0jLc0BmRIOmC13RhAaytQ4uRZUsrkZioNvXZPqiJ87Sna5TUP7Xa7exZPQF/eGifL5pj0O1LXM0JyNJUcUNnSM5qIBxoZdM+6XnsovcQIzxXqOeCXBjmZSp8HbRWDP87g6LwH+qt+oZ+R8xep7bU1ST0iz27jnPQb5UuvV5VGt51qDA4G6MX2APXbjID91JWqHmgB/ZB8vpYWXvuxPI8N4LsnciqBPwJwqzJoC61OZp8al4PrYcO5+WQv62M0zTRJj8nZi/TBMxZS9Qjsx+e50b+oOkZvXLU942Up6cjH9SX4G2BnnYChJYHIOHVrc3IT1D55FujhW5DOo+CAdG/HI696cPF0G49GXm9lf5qM3N7B6RPqQf2AaDegw2q94/718PKq3J3H1y8ju96T5A8SDNo7sj98WF1fgN7rSwNl1An1Mz8uGx315mgQBx/bi8G5G9frnbOjc23wYWTMvd44Bj147fXk+PK8DUb6J+K9GF734JD4luHp7Rvv1EH/EIyKLqnfdsGVcIhy1LmJLwcG2NuT+Ooc/ENyRE5vgR/a595pl1i37UUMXa0GnTm5MEjjpgNCFfzI2JjEhwZxQ6M66Prkqd/249Mm6NXOx/isQUAOfh0YY3LbNRzv4BqFPsi5Dhl77afB6YI83Xauvb4mrWbGMcq/Kw30JOgh0LvNQddFfboAOiDNbvvsonECfgaYjoeHYBIZ4EcMiCXDuL0yeYrb5/GZDXKhW5X0G7AbiTzoLMk5uBhxzyJO39DjwyUZz9rXg/OvTd1rX7rSmBzOCMjxCTmbgV7qtcnlTXviHVYNo3vgeB+XUr8P5cYViiozBlV9roEf03PJdb89hfkaZY2o7tdb8sEFOU0u0I4HUMA6icllfLhAFdKPQR+dhke6u/wsSaflAcZVAC+PAzBZO0eGSshX49PN4afBkV9u33ZflqpNjqdNldpv4Cej3u3Kxue46xiTatkeXH6SYKmuls0VxrCaAKUEKvgc5f7lkzGN/Uty8qn5Oe59lZQZqK5yTao/EQ/0vVQDOSc1e8QGu+eifExOGuiHgL46gnmRkdToElUqx1LNI7p3rIC9A/yuT9F/7gw6tlS+kEAfW1I1KLe9U7kcymCGVWsYTyovJdA/MxIs9RG5XjaXg+PnshofeUtQs0MdXIOOL61uMP5UluSYNC8aFok6zT6xRnqjb9TJYCEBH4P9q6A+Jqgf6zPwGrQaMS1Quh9d4gWNyaALeJkyexgDHx7IC21Apm7zhly5zWncdsjJqXS+cqPe49Mn1XDPek3yKV4NjYO2fntqeDdgd95+uhxdHXSfrg9GZ5+68c2NYVyd9U6ebj5fXrhd6eb6enU57PVmN97lldE57PZvvbjcGZCzwFsdd040WOFq2Pl4e2Z51WbHImeP8SfZ+OCBk/oZnIc+2D2Ob9zcwq85MSJyEAysmTEn7dvBMbC9jsG6Cfo188HZI9iJneHg/Bzsxm4vvgQ5EkJXR5dkGRsf4tNjjB/9v6a9pElVJlj7v9wti3JsYZkFxaCglFOLO0UpZ7ultdSI779/T3reuHFXHafbg0JlPlOmtUYfQrHFJo+pZ4Olxl2B93omnsDcyoGFXTxZ8NXgQXXp38xoQGgFxwQLfl1t44Jg2RyrfuF70H8Dj36zYKdHb5IZeDGqaFHIysQh91/LqiEX84UGGXWzYGnkLxAW/BguaI7PRdGBVko2tRrTNfFvdrQn6JaUrZnswFzFLqlM5jZsEqRBS6s5PTpBkzJ2d/TWwZ5GJQnCT2lkTuGc8eFq1Qk4FHznHjjZlTMdbOE7IdED+Gngqv2nZyMaX2U0FWeuSxz93vgpv+WR8LoRTJqOHlQlQWXGLvyW12JwTwxNNf5rVLNOutOoAXzA57cdVeTemuax0Ba+rqs4H6oovFJhIFXCBM8tPVEDn6P6arD/T40sWG/tH47DvHZGq5ACuJlccz+eK3fKfZMyhQ4vYq59F7oAj0g6ZFQAvbYVzwJ9AxoYrQV84J3zrwb7t/gCOTLa0uPGfXdnPvGYZNAfUNkR52K/FEzFteTn//F58LrQgQu8XvZEr6RT3hG0qnoDm9TisHbEw1XsF8UnJ4ihG9DaOPk5NCd/CEXAbfrkvPB/Gf8qowH8q6E98EGmIHc6587uk0ewT/zvZ848eIFP9m+iramqoC/Ipf7OtCLZKQaS/d+x0M/dhv3f7+6wC5sWv8+u6nmd9p9ZVzX0tCMH36quJxc5OCmRTDyq18Gokwp6vHwnAWm9y6Bvs8w2v/xulpV2ImVYq62ZxHKQqJGdPulp1JHyl5QaRY46+T3KjgEOl4k0ZrAkWyhhZm+6WDUhpehvRmcd1vBvyjH6RlUpnza4kgP9RuGYRoV0KZhTcRN9i/64XP2tRd+ure9wKO/E3orUE30BvYnrlQVwLI5oc5TfNsZhFugfalK5Fj8Wvm56lKVVfcK5dYElkELyoydBXb4OzxQBJ6364iKFD/rjXBW8uKVlDj2Cc1/iXAzO27dSc347CMTaQCenb5Fa6J7elV7GhxoHDgsxIpqKYy4SAd/UewBCRzlLoz7rTvj3J9cVfH3Euab/+Tf+XwVIzj75kvqX64Ov7KdO+4y7Q/ZRT0Of0OFTB5yvLqDPggfr6JPwStZP/xsdaPYpJedTGfsUiF2X6/7/zA3kfznGx8/0OMdlH7U75OHbLdTTxJFfLsPdbh15WTHcrV6cp3oyaSpvNonHYO5Mjb6N9xf8qPFKlmXQa0zR2Eu/W2cdcy/8V5KPn/5f8CxHrv0+q7FNZ3oKMCuDHU1genWw0mmXTnXg2eWEiquSth/SuPbfFiJhNfNzq9cyvKrYpKG818Eerk2GwKuPmTUgmOBFPwo6ZLyk1OWQ0qc7JJRWJW1df6fhS+a1j2M8Ukv50qoZVRw94CgXGc11yHmqjHmORaXfM5Ci2wJmGfgIET0yCjq+hi8Z/PLfNw8nRwdDf8sRMTVo/4X6hfToA+Rq2c3dB21Sr8341MmoQbQHBMrYygTXIeB9KFal6j1AnqrDuXVJ3xndTJQKh1sdPD5JnZWBr+xBx6OpxRO3ypH8MYCf9v9Yj0vOw98zelRug7vgItybONfwybASsELwxQllsZhxPvUViyH4RYiFCK1fM6QkfH3grVP1+jzH2lbdK/tbL3cP7JPvD/gCXPVAai52GdXG/5Qe6uWLfT7PdHgOdhbekhT8wsN9cT3OdvtOxEZQ7rZRnC36z/0iDK7fZncPwvux2MpsrZrHqStHA5jOhR2v/qKiLObmWUdxZ9Kjv15QZYMj/VVBdOw/IZo4sbDj/A6RNXyYSUndRnjUwKGMwtBUX/Kooq5OSukZ9bThgLrAF0qf9KdAiuEX3ZT/1rogmDWHooJk4W8sfj+/4vCSOedSngmn1GpASUYu7KYcwOdQaclw3+8yv6QMPnHmP0wkRH2VXwYlHCUSJvEOX+Vr5neRyc3Dvct+zNEI8KWmlkXpTa2sYMZYD59NcILMElsLCwjIa2guTZI3npsf4b9zbysaCY+sgPgHEUK/cZ1B96FvYeEqdy2ch/euvJ149YSGjxVtSxXRnI6GtJEV+/h2jvMLStQXdCr4YQmfCNGGPocOBd9BUs4+/FaJigbCy3RWi+8SPg66rsW4KAc850m1n4syQbFUnizaTk+4T8oeolW5C851YpIKR9ok3c29X5KfXMdpiylvAoAK3lWvLYJGvM+XWv71HODaWzwtNXMnhF+lR+61aTp1ldaOmIH3n6UJu3bR2m1taFEX4/05GpTFZLdZR6NisRn3Q3W1E2cMvmpfJzcZbdS0M36jEdWhmOb0OPo/s/Rk7g//mGQ33VLBRg0FHcNgXaRLPUllrgJF3yAbrSI73ss0UXOzWEpqhAudftGuhE5MI5oXfqDzX7rDt9n+mnpXv2XiMVWZbFuVUq05p4TvMlxHG7q4/k1HG1qV8leHD9Kl36b4TLrhN20kaVf4qYlcmivoOPDNvIC/gC+HdHVNCmCc+QMdtznXUzqUtE1oRmpNixL4AR13rmWPc4Y0k1cTwCo26IEjp+0Mpd2PYR6h2+SGlrGAz4kpmtEJuoQyBUkJP7j5RKYTxhmycsW44VaeQB1Deaia+bDx8H4oXHNONvyXy7E0gnvhuu18Ro1folGyLvqlYSzGwFHRKaiRf51pUImvyktoEAu8ro06pCXrmXEgjIlLccB9Mf60jtSunJTUQjh4dGJZw8y3drR84624vmLgnp+KtqKZ8OAi1k5Lx3NxODKvNZjvwIsLqH/wrwTfd6jMgXujqZBPNm5uEYzNLpJuMXge2uHLFPnOOFFULk/jjY38Yjbc6VUYlrNM6nboH2fheHYIB7Pp85lv1fg67+2Gd/YZW92sgqIeXnRrGhQ2u5qmDP7gjWUwC8ZqNDBFLXdlNKaiJfcu7Oj6Ik8metn5Qc4ygMeqKYuaN1DO9FP6fzoFmCZ+agc57RVOIJvLsRtqO2/xHsId+MM59J+J4DOvQW3zJbmdYGGHPlnW1XHJemWnoYuWHSBwv0Un+H+TjoFTgdEjn16FnFhAiG7IsYl/aFXLxPaBgzPg4OCP/jT7izf7AdcGJ8pdWut4SvCL8P8HOhX+UkMC/mTAQ7n48JINOqQS+a05/3altMkvHa6+0v0Nndd8bj/Uv9Lhvzx9bkOHVgX8R9wn7QaBSUE6sdfX8Qy6Hf0RXEgm8B8R6yvgY3qhtCODvBvQde3sLFxfrNl/fCAmYn03LuGn9N6fZrKuWnB/Lt0IlO8f6UdzTmcgCSIPzyeK9OAgfkqGcJdzvCEOjOKE639PIGPgVFNeiuCHoPOdm6gMbZkamihhrruNRjfNQBW0HsgS9S++hpSuOd/W4lnTF69BvXkDKtwLMYtSmz65LxqcY0Ov+Z/cOWOnnIg3+xqvI6elhL824pmxz4nYp0C//f6Xa5X8qDOzdqgSnHu7nznPR8fN+O8FZRexoGAjLkd/9pkHXf7l6P9e57COBD6XPOdsPTr40JUjKnfwydHp33U++ydDIRI0v/ibBQNSAd/Fvznmm6/X4lyh/XBvOBIBc1c6l5vT/7zPJ/eWnxxc89y94Pn8iPv+xnoIetYbV/3QFsWfLK/s99fPchB6deGhw6Oi+M7kpQ5fNOvIvKuAaX8ybyvXzqfP4qB27iTZqUP4spOx1Idg6Q7hO9KAktHDNiJfmP4GXlxNisnZwIl5djiw61+eO11MAY3WiS566/tJpr7N9IjDTt40asmV8jd6ULKfv0NH0q0jhybc0VcZnM33idpZNDELj/wa9cr5t5Y/FC3p6CpPDyqeQ9Umr9BnwdGg/jOSVx39kC78hUmhJxt+yyZ/5Cp/qMchfWk547wHvmOroZTGmb9i/rgU/q8d9OjH9f9s0hG9Dq843Qh1MWefs8E96H5Nx6t/5f2Ah/Y7JpxwNGqBv0zKU0I/QRI1rDqTKfk633SY9oSRS+q7vGnksE+ZU5Tz9b6MujE+HE0K8J1B58GfDHrC0dxXHbmz4YSmiW91hOs25MmmU6cmv5d3BY/2XxRsOf82FDZpepVDO1v6g7j7a27COWiePvU5qtgamvC8bW5Vj76T8GTLiRzG7R4lPuNLL3ebvC64sNA56yNtzOAlyiw6Pxp3GV0cSdFGeIoalXvmvMDYYEpZAnIbf0n/KCeP7o6yqRgQeKkFtVs5O+YRiPI2FL+amf2f8Dp0BBJR1kYfQRIF3MT+lnOvH/gGqjsKUmvH863XA++TPqCvfMU+CT6uK54XMTCLU/D4zH8e3E88fOC51+HhKBqlzp3dk1OD/5OYRkIE/82XyPjdT+7HrTVvO33y32L9Bm7A/6HfgBuBeLvc2mseUuOtv5xdkr2qzh/pGbk5+Ax9m5J/F92CllX3RPe292dWYS+eySfvXfK+AeMG3ldp6D7gTc7+IZ16oYV5gy8cyB3096yoxrsoUrp47Ha/0Zj57/AMW+7ya7cFL+rZS463nH+v5QSCqjN1pJqEy2J2kfNQ/aj5VOoVij5b6M4teFyHc2rtgk5jnOoWTuaYNXZBFyJq9NSLlbw0woFev+RSRysNG9UtIk/PLj58ntJ7KQ8NHw+7w/PUb5v9UnMGU7Xok63RHPEUZOa3aHzBeclAp0fgPHwY+OwIm6NHB97RvIA/IIKCbx3vOf+6cP5dKjk1IImHlQ0z6KOvg8Kgb2wOPA8znv8seU4FnrxYVUE/4X0Gls4AS+o3oLN4leJAVU3fJjS0ubKem3L+fdI43y7E8APns3LJs+Cr/AidpV7QZTKkfpvOym+aQRt8Sd+crw8btLXBhnKikY4CKrUsTXLiPZ+JAT8cU+iV+AH8h7iGGB/i7nnfbG5lk/cWvdqv89aNuj2v0vFKdEFBBk8Z/TOkeM15InTkUA5JlY9VQK8FnJj45NpG+5DchudjHbqs3Vh/n0S/jjqiPRV/badRiSfPM4/QX/B7UpFcSu8Ybm1fObPCn1ZQltRzeg9RMF/tNaghT2SdN2pwltPkYVCN4uf+3MbcH99c75WR0KlH3MfpQXHlXKwP/wFfLXgEm1CBLmJ99tZ+Jl4P0Wa9mfH8KqhEo6A+Q+UX0TsXc9FO+Hl4PlzzKwf/cL7FfgR3mYGvRRumgefolfF9XIr3csB3Q+ZBmI4L9IXzC70sug/gW9Z2WqQ+81mrSTzES7xx/zzH2h/cyk5vYtb21nZhnQfvkfE+2QW6F++H87nyvlz70PvR6Vu0et7ZqM++hcg5z8tFZOVTuFcqeVVv23NuJHFfON9nqaA3F/fn9hR6dnEf83C2XHRk1Q7dxiLh/YzQTLNdliuRTN1x/lTvenbe6TL01cyVw1ZI15neTUyQNrKteY6Drcoq3r+IjpzD+cGqA+XzLlDsw5td/KDIYRZWQ3msowYt99Ito45enuShg2ZZ7yT0JurrQm7tRzRMyRZB79MnNpBmsqfzTF1pFMAkykInVzKuH/P8Z18HhU57hEOd2WxLzUZwo9CnNUmPYuD9jPeEWrx3trSDJ/vopc6/eK+n1NBha9Zr0LPbqwxMsqSDC74cLD6Zqc1KKhqcv3gEa7jlPlhn3H8DOil/afD+ovR37G9gMiaw+jTP5IL3OZZWtnlfCH7LMekvPdDElVfQxlJuIbnBp2MDX74o5JNUg0wWoEhQh+gPTT/4/Oi++EDSyjGBr3Tmw1+dZXL0lVEb8Sjlu8I5F+AjJhHwH1rO0nQGnoRva6fOVYeVcGvoU/jcBPqY88qsA36DX5sXYaAHoB7Oq/2SfTD6RFFkoUeDlHDr0Pl7CrKgyp0389oMepd9OvD9xFZsacGsfoL7HE1pz7myGHIeoJhXXDQVLDjvMRy15L0Pf6TTO0fDEetGFHuiwYrw//cKfYXXPSjo0dWFh+ARYpt5acD+LSJ/JZoNanK/dnpiQvmGqjbrzy7z4md+O+Q5ty95nw5UuOT86Kq5H/7bc8LrOsINeO+qtBJCkbUo540LiGSepwnOjRzer4g4t3IOopGzr2t7cxN//9/94c++0ygX0Jtr0XQ5P/ilVe7cq//W7j46VNFH147A09YXollQWnm/nD8NpEkin4roufN5D2q0M2mUHcf1Ltqq0h0XY1WpqzseyWQDUhqM7HXn/9YDh4vu5vaPZjDn/YMaUpNuDdRF/g3+geh+t6maAReTX79I3b3+NuK3iK6iBsEpOc496JQZ/VkVi5sBCHY7JI/8sF9UXMTI0pVkBhxy3sLp0Il9hnWh6/FTvHFu8sG+AkfYYlzzed6QsW6Pfll3+zq48fcxHBw+wfKOLO/PrP/lvXi+3sM1n3n9czeL+vX8umu54f466jzFLdwc9d8z7UBvw9CJYbJP4tVYO373GOd0GQQ0G6rHqx30sgTm/y4HxocP7NB2lkw57zga/yi8IfdxYadcDNHZDlOxNvHZhD/Chb4z4FSYtS8jL+LWdhKKTpyvHkzi0ObtaBqteG7zEr07VbdeZcMrJZV35ki/zX0rTzSF2Mo7bZk9PPiOp1i6NK++HObdhbiv6fnwjBmmTstV/dy7iQ7vjXb6dFq7hR3XvKR3l9skOh+/5/Adap+tw91uFQ2Bt/SsQ++Ye5O8A1IYvuTwAB+Z9ame+ONEXsh7+8fGYEz5lb5r+aXzQhapGNvvKf0R+jXZ0oqHccFU3g9fsZ01OIOf5c43f19BG3/uPwOnYfpL79rA53IrGgAvqvNR+hdx5O/PdHLeh3IoVsHl0YV+jL2c82FIEjcvI/k39b5y50pZBh/Zs4Ku3wdnP5ewsjWpXxFfw6dVZe9P64P4K2g09Tom7ztsXk1oRdih5XM7Dveq8J+HZfRtFn/jahd6Roe79Dd41eOXrVfqBH81DpW/NsndnI5+qxitab/279QPaWToUfgX3XfoogHGgya9tDyZ0Z024EWdpLQ6AuTUhG49iOI0Zt220WktWgr4yLgFXDTBkkC/HRvhfh9iynu43uXzTSJerVwBgTn/a+ZexnO146M7pP4FeAS13CxkUzQrGj4c+K0t75EVWvqiq3l/OEffO7kOnvw9JXzaI/fPPvdq3gv5n//3/wHQeYlw","contains":"Offline compact weights, replica, qt1/TC/COMBO references, blob, stdlib trainer and drivers. Regenerate teacher with the exact checkpoint sequence; Gate P large tables remain external pinned inputs."},"extract_python_source":"import pathlib,json,zlib,base64,hashlib,sys; r=json.loads(pathlib.Path(sys.argv[1]).read_text()); b=zlib.decompress(base64.b64decode(r['reproducibility']['bundle']['zlib_base64'])); assert hashlib.sha256(b).hexdigest()==r['reproducibility']['bundle']['sha256_uncompressed_json']; root=pathlib.Path(sys.argv[2]); root.mkdir(parents=True,exist_ok=True); files=json.loads(b);\nfor n,v in files.items():\n p=root/n; assert not pathlib.Path(n).is_absolute() and '..' not in pathlib.Path(n).parts; p.parent.mkdir(parents=True,exist_ok=True); p.write_bytes(base64.b64decode(v) if n=='blob.bin' else v.encode())","recipe":{"extract":"python3 -I -c <extract_python_source> <result.json> test/scratch/history-replay; subsequent commands from that fresh extraction root","compact_only_offline":["python3 -I build.py","python3 -I restore_specs.py","python3 -I driver.py eval 10000 512 replay_0.jsonl scratch/compact_eval_spec.json","repeat driver chunks at 10512,11024,11536; compare COMBO and qt1 edges and paired stats"],"data_regeneration":["python3 -I driver.py train 73000 64 train_0.jsonl","python3 -I driver.py train 73064 448 train_1.jsonl","python3 -I driver.py train 64000 128 calib_collect.jsonl"],"teacher_training_exact_checkpoint_sequence":["python3 -I learn.py boost 8","python3 -I learn.py boost 8","python3 -I learn.py boost 8","python3 -I learn.py boost 16","python3 -I learn.py boost 24"],"teacher_training_note":"Start from fresh extraction with no teacher.json. Checkpoint grouping matters for byte-exact binary64 sums. Then hash teacher.json against frozen manifest.","student_and_calibration":["python3 -I learn.py distill","python3 -I calibrate.py","python3 -I restore_specs.py"],"choose_and_freeze":["python3 -I driver.py eval 64000 1024 choose_base.jsonl scratch/base_spec.json","driver.py eval chunks of 256 at 64000,64256,64512,64768 with scratch/choose_model_spec.json; names choose_models_0..3.jsonl","python3 -I freeze.py (do this before report; recorded source/model hashes must match)"],"full_report":["python3 -I driver.py report 10000 512 report_0.jsonl scratch/report_spec.json","python3 -I driver.py report 10512 512 report_1.jsonl scratch/report_spec.json","python3 -I driver.py report 11024 512 report_2.jsonl scratch/report_spec.json","python3 -I driver.py report 11536 512 report_3.jsonl scratch/report_spec.json","python3 -I stats.py report_0.jsonl report_1.jsonl report_2.jsonl report_3.jsonl"],"checks":["python3 -I verify_live.py scratch/nominee.json","Original Gate P inputs: fetch listed pinned files to scratch/gate_inputs, verify SHA-256, run python3 -I gates.py there"],"command_limit":"Use timeout 470s for each scoring/training command; all measured chunks were below 8 minutes. No timer waiting."},"data_and_score_checksums":{"training":{"seed_range":[73000,73511],"seeds":512,"rows":298136,"bytes":29813600,"sha256_concatenated_seed_order":"950ed288ab3a85de446046e491caa31d070b905a4c05453b800d3225f5dfcf97"},"choose_calibration_and_distillation":{"seed_range":[64000,64127],"seeds":128,"rows":74359,"bytes":7435900,"sha256_concatenated_seed_order":"7c45337a9532fbb0125119cd23efead8a9a08514bda0873c10e89f04ee4502b6"},"choose_scores":{"seed_range":[64000,65023],"n":1024,"policies_in_sorted_order":["COMBO","compact","compact_half","qt1","teacher"],"format":"rows ascending seed: little-endian uint32 seed, then float64 edge in sorted policy-name order","sha256":"ccd031bd994307247c7eebed65069b8f42f6a19c5031e233ab890f9544ebcafe"},"report_scores":{"seed_range":[10000,12047],"n":2048,"policies_in_sorted_order":["COMBO","compact","compact_half","qt1","teacher"],"format":"rows ascending seed: little-endian uint32 seed, then float64 edge in sorted policy-name order","sha256":"f64812d196d83f7eb9f2612426d14d27ee1b4aa171d9dfb76ec4d9ca661936e7"},"closed_loop_shift":{"seed_range":[64000,64127],"seeds":128,"rows":74369,"bytes":7436900,"sha256_concatenated_seed_order":"9a165336c2f1a9de176cbe3d004ab32d413bd18a74ed5bc509b8b9a40db94d25"},"callback_row_format":"Ascending seeds, concatenated little-endian 25 float32 rows (24 features, label); sample RNG in driver"}},"runtime":{"wall_seconds_since_source_fetch":2748.4232547283173,"teacher_original_training_chunks_seconds":[54.465198538964614,59.89048095792532,78.52010818989947,100.00732086179778,156.13883417891338],"hardware_parallelism_used":8,"max_command_limit_seconds":470,"chunk_wall_seconds":[119.06385720381513,79.21533363405615,79.9191495559644,81.79821365512908,85.1877262420021,215.0384338318836,228.50972522608936,262.18882543896325,233.61552505707368,14.172662195051089]},"limitations":["Replica only: no engine, contract, measured gas, Gate F, bytecode-size check or board submission.","Compact distillation has rare large probability errors; the reported quantization error is sampled, not a global bound.","Closed-loop callback populations differ. Raw filter errors on a model trajectory are not counterfactual COMBO performance.","Calibration and distillation share 128 choose seeds. Report never affects fitting or nomination.","Teacher weights regenerate offline from included stdlib source; compact weights are embedded. Original Gate P large tables require the pinned external inputs."]}