Training Report: mein_modell2

Basic Information

Model Type: wakeword
Version: 1.0.0
Author: Trixy ML Trainer
Created: 2025-10-01T13:37:43.215229+00:00
Last Modified: 2025-10-01T13:40:25.348344+00:00
Description: RepCNN wakeword detection model (standard)

Model Architecture

Architecture: RepCNN
Total Parameters: 2,586,371
Trainable Parameters: 2,586,371
Model Size: 9.88 MB
Layers: 45
Activations: ReLU

Training Results

Total Epochs: 94
Best Epoch: 20
Final Train Loss: 0.001067
Final Train Accuracy: 0.0000
Best Validation Loss: 0.119300
Best Validation Accuracy: 0.9011
Training Duration: 148.79 seconds

Dataset Information

Total Samples: 1,415
Training Samples: 989
Validation Samples: 283
Test Samples: 143
Number of Classes: 0

Hardware Information

Hostname: BPGamer
Platform: Linux 6.6.87.2-microsoft-standard-WSL2
CPU Cores: 32
CUDA Available: True
CUDA Version: 12.8
GPU Count: 1
GPUs: NVIDIA GeForce RTX 5070 Ti
Total Memory: 31.2 GB

Wakeword Information

Wakewords: custom, system_command, negative
Classes:
- 0: custom
- 1: system_command
- 2: negative
Detection Threshold: 0.5

Test Results

validation_loss: 0.718315
validation_time: 0.436567
num_samples: 283
metrics: {'accuracy': 0.13427561837455831, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(72), 'precision_Class_1': np.float64(0.13427561837455831), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2367601246105919), 'support_Class_1': np.int64(38), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(173), 'precision_macro': np.float64(0.04475853945818611), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07892004153686397), 'precision_weighted': np.float64(0.018029941689870023), 'recall_weighted': np.float64(0.13427561837455831), 'f1_weighted': np.float64(0.03179111213852471), 'confusion_matrix': [[0, 72, 0], [0, 38, 0], [0, 173, 0]], 'auc_roc_ovr': 0.3374660565425811, 'auc_roc_ovo': np.float64(0.35195643049499153), 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 72.0}, 'Class_1': {'precision': 0.13427561837455831, 'recall': 1.0, 'f1-score': 0.2367601246105919, 'support': 38.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 173.0}, 'accuracy': 0.13427561837455831, 'macro avg': {'precision': 0.04475853945818611, 'recall': 0.3333333333333333, 'f1-score': 0.07892004153686397, 'support': 283.0}, 'weighted avg': {'precision': 0.018029941689870023, 'recall': 0.13427561837455831, 'f1-score': 0.03179111213852471, 'support': 283.0}}, 'intra_class_similarity_mean': np.float32(0.9960452), 'intra_class_similarity_std': np.float32(0.0040368335), 'inter_class_similarity_mean': np.float32(0.99553347), 'inter_class_similarity_std': np.float32(0.0044998657), 'class_separation': np.float32(0.0005117059), 'eer': 1.0, 'eer_threshold': 0.0}
predictions: [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]
targets: [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2]
probabilities: [[0.26378169655799866, 0.4948451519012451, 0.2413731813430786], [0.2597719728946686, 0.4915142059326172, 0.24871376156806946], [0.2572769224643707, 0.48541060090065, 0.2573123872280121], [0.2689773738384247, 0.48552438616752625, 0.24549822509288788], [0.2618688642978668, 0.48936912417411804, 0.24876199662685394], [0.26263755559921265, 0.4827612340450287, 0.25460121035575867], [0.27078720927238464, 0.47540581226348877, 0.2538069188594818], [0.2573886513710022, 0.4904750883579254, 0.2521362900733948], [0.26967811584472656, 0.4686349034309387, 0.2616870105266571], [0.265425443649292, 0.4835914671421051, 0.2509830892086029], [0.2643943130970001, 0.4882180392742157, 0.24738766252994537], [0.26142215728759766, 0.49202829599380493, 0.24654951691627502], [0.26545584201812744, 0.4894888401031494, 0.24505528807640076], [0.27168896794319153, 0.4703938066959381, 0.2579171657562256], [0.2680344879627228, 0.4737197756767273, 0.2582457661628723], [0.25869446992874146, 0.4888080954551697, 0.25249743461608887], [0.26071494817733765, 0.4895656108856201, 0.24971939623355865], [0.26501041650772095, 0.49302011728286743, 0.24196943640708923], [0.26591190695762634, 0.4772942066192627, 0.25679388642311096], [0.26204779744148254, 0.4870642423629761, 0.250887930393219], [0.2553892433643341, 0.49251696467399597, 0.2520937919616699], [0.26082295179367065, 0.48556482791900635, 0.2536122798919678], [0.2689482569694519, 0.4741143584251404, 0.2569373548030853], [0.2731628119945526, 0.4671652615070343, 0.2596719563007355], [0.2782571613788605, 0.45616570115089417, 0.265577107667923], [0.25703200697898865, 0.49324139952659607, 0.24972651898860931], [0.2792070806026459, 0.4592638313770294, 0.2615290582180023], [0.26623815298080444, 0.475988507270813, 0.25777333974838257], [0.26293042302131653, 0.486819326877594, 0.25025030970573425], [0.2565838694572449, 0.49522140622138977, 0.24819470942020416], [0.28250789642333984, 0.4563349783420563, 0.26115715503692627], [0.25830382108688354, 0.49065983295440674, 0.2510363459587097], [0.27217599749565125, 0.4590441584587097, 0.26877984404563904], [0.26598212122917175, 0.5054824948310852, 0.22853538393974304], [0.26271799206733704, 0.4940319359302521, 0.2432500571012497], [0.25981512665748596, 0.482341468334198, 0.25784337520599365], [0.2614932656288147, 0.48502442240715027, 0.2534823715686798], [0.26463842391967773, 0.48165401816368103, 0.25370752811431885], [0.25570693612098694, 0.4926699697971344, 0.25162312388420105], [0.2620331346988678, 0.4845897853374481, 0.2533770799636841], [0.2712179124355316, 0.46262678503990173, 0.2661552429199219], [0.2573585510253906, 0.4907333254814148, 0.2519081234931946], [0.27149698138237, 0.4719278812408447, 0.2565751373767853], [0.27089056372642517, 0.46680334210395813, 0.2623061537742615], [0.27238357067108154, 0.4771067500114441, 0.250509649515152], [0.25821441411972046, 0.4898906946182251, 0.25189483165740967], [0.2575710117816925, 0.4918721616268158, 0.2505568563938141], [0.26595088839530945, 0.47699034214019775, 0.2570587992668152], [0.2563515901565552, 0.4912757873535156, 0.2523725628852844], [0.26200535893440247, 0.48469793796539307, 0.25329670310020447], [0.259110689163208, 0.48871129751205444, 0.25217804312705994], [0.2633714973926544, 0.4814513623714447, 0.25517719984054565], [0.2763628661632538, 0.45975324511528015, 0.2638838589191437], [0.2676157057285309, 0.4755809009075165, 0.25680336356163025], [0.2582278251647949, 0.488240122795105, 0.2535320222377777], [0.26997673511505127, 0.47180113196372986, 0.25822213292121887], [0.26875999569892883, 0.47478771209716797, 0.2564522624015808], [0.26703450083732605, 0.4847593903541565, 0.24820609390735626], [0.2675831913948059, 0.4790405333042145, 0.2533762454986572], [0.26484423875808716, 0.4707363247871399, 0.2644193768501282], [0.26398715376853943, 0.4999687075614929, 0.23604412376880646], [0.2670154571533203, 0.48436763882637024, 0.24861694872379303], [0.25826898217201233, 0.49177491664886475, 0.24995605647563934], [0.2852296233177185, 0.4585736393928528, 0.25619667768478394], [0.258920282125473, 0.4883047044277191, 0.25277507305145264], [0.25736844539642334, 0.49051231145858765, 0.252119243144989], [0.26583996415138245, 0.5026752352714539, 0.23148484528064728], [0.27614784240722656, 0.47211185097694397, 0.25174033641815186], [0.2654113173484802, 0.5012249946594238, 0.23336370289325714], [0.25669804215431213, 0.4914531111717224, 0.25184884667396545], [0.26123109459877014, 0.4920278787612915, 0.24674105644226074], [0.2635180354118347, 0.5118021368980408, 0.22467981278896332], [0.27106812596321106, 0.46984562277793884, 0.2590862810611725], [0.2734253704547882, 0.46262478828430176, 0.26394984126091003], [0.26966872811317444, 0.46796929836273193, 0.262362003326416], [0.2596326768398285, 0.49029040336608887, 0.25007694959640503], [0.2763848602771759, 0.4644748568534851, 0.2591402530670166], [0.26387447118759155, 0.48803040385246277, 0.2480950802564621], [0.27854815125465393, 0.45711275935173035, 0.2643390893936157], [0.28685393929481506, 0.4410839080810547, 0.27206215262413025], [0.2573487460613251, 0.4927464425563812, 0.2499048113822937], [0.2624478042125702, 0.48635274171829224, 0.25119954347610474], [0.2681894898414612, 0.47303009033203125, 0.25878044962882996], [0.26465103030204773, 0.49228909611701965, 0.243059903383255], [0.26140353083610535, 0.48698845505714417, 0.2516080439090729], [0.27144375443458557, 0.49135297536849976, 0.23720324039459229], [0.26296553015708923, 0.49757009744644165, 0.2394644021987915], [0.25712278485298157, 0.4935651123523712, 0.2493121325969696], [0.26652979850769043, 0.4922057092189789, 0.2412644475698471], [0.2700376510620117, 0.4643547832965851, 0.2656075060367584], [0.2688497304916382, 0.4666457176208496, 0.2645045220851898], [0.2611611783504486, 0.4903285801410675, 0.2485102117061615], [0.25829118490219116, 0.4878886938095093, 0.25382018089294434], [0.26154249906539917, 0.48754775524139404, 0.25090980529785156], [0.27045589685440063, 0.4695446193218231, 0.25999945402145386], [0.2669876217842102, 0.4742473065853119, 0.25876498222351074], [0.2781559228897095, 0.4574403464794159, 0.264403760433197], [0.26473280787467957, 0.4946274161338806, 0.24063970148563385], [0.27634620666503906, 0.45997124910354614, 0.2636825144290924], [0.2579871714115143, 0.49160510301589966, 0.25040769577026367], [0.2587699294090271, 0.48767903447151184, 0.25355109572410583], [0.2582254409790039, 0.4955754578113556, 0.2461990863084793], [0.26986435055732727, 0.4724223017692566, 0.25771334767341614], [0.2789592742919922, 0.4560183584690094, 0.2650223672389984], [0.2606169581413269, 0.4866025149822235, 0.2527805268764496], [0.2640111446380615, 0.4953792095184326, 0.24060966074466705], [0.2640824615955353, 0.4954056143760681, 0.24051201343536377], [0.2639096677303314, 0.4941117465496063, 0.24197864532470703], [0.25958961248397827, 0.49311667680740356, 0.24729366600513458], [0.26698753237724304, 0.47227296233177185, 0.2607395052909851], [0.2659629285335541, 0.49893131852149963, 0.2351057529449463], [0.265018492937088, 0.4993705451488495, 0.23561102151870728], [0.2653740644454956, 0.50153648853302, 0.233089417219162], [0.2657981216907501, 0.4974399209022522, 0.23676201701164246], [0.26596561074256897, 0.5017416477203369, 0.23229274153709412], [0.26425355672836304, 0.4764607548713684, 0.25928571820259094], [0.2709636688232422, 0.4971366226673126, 0.23189963400363922], [0.2674328684806824, 0.48093903064727783, 0.25162816047668457], [0.2596820592880249, 0.49577775597572327, 0.24454019963741302], [0.2652026116847992, 0.4776196777820587, 0.2571777403354645], [0.262648344039917, 0.5006135106086731, 0.23673813045024872], [0.274915874004364, 0.4610024392604828, 0.2640816867351532], [0.26994597911834717, 0.49711254239082336, 0.23294150829315186], [0.26222267746925354, 0.4921218156814575, 0.24565547704696655], [0.2759972810745239, 0.46866774559020996, 0.2553349733352661], [0.2635006606578827, 0.49024009704589844, 0.24625930190086365], [0.26800107955932617, 0.49892714619636536, 0.23307180404663086], [0.2613588571548462, 0.5080662369728088, 0.23057492077350616], [0.2690945267677307, 0.47086185216903687, 0.2600436508655548], [0.2681240141391754, 0.49966150522232056, 0.23221451044082642], [0.26458293199539185, 0.4865855276584625, 0.24883151054382324], [0.266436368227005, 0.4980922043323517, 0.23547136783599854], [0.26303964853286743, 0.4806388020515442, 0.25632160902023315], [0.26489466428756714, 0.4839283227920532, 0.251177042722702], [0.2618263363838196, 0.48090264201164246, 0.2572709918022156], [0.2638579308986664, 0.4882756471633911, 0.24786639213562012], [0.2647974193096161, 0.48888781666755676, 0.24631477892398834], [0.26310405135154724, 0.5055026412010193, 0.23139327764511108], [0.2659326195716858, 0.4913879334926605, 0.2426794320344925], [0.26549723744392395, 0.49725663661956787, 0.23724618554115295], [0.27166807651519775, 0.48350396752357483, 0.24482795596122742], [0.2702019214630127, 0.4697669744491577, 0.2600310742855072], [0.2577150762081146, 0.49135488271713257, 0.2509300410747528], [0.26214149594306946, 0.5009613633155823, 0.23689712584018707], [0.2691740393638611, 0.4956934154033661, 0.23513254523277283], [0.26333460211753845, 0.4982220232486725, 0.23844343423843384], [0.2608487606048584, 0.5116605758666992, 0.2274906486272812], [0.2658911347389221, 0.48045188188552856, 0.2536569833755493], [0.2663593292236328, 0.50249183177948, 0.231148824095726], [0.2672879695892334, 0.48089757561683655, 0.25181445479393005], [0.2662883996963501, 0.4986937940120697, 0.2350178360939026], [0.2657853960990906, 0.4980829954147339, 0.23613162338733673], [0.2664678692817688, 0.49995315074920654, 0.23357893526554108], [0.26515042781829834, 0.49569982290267944, 0.23914974927902222], [0.26305243372917175, 0.49697351455688477, 0.2399740368127823], [0.2607244849205017, 0.4923610985279083, 0.24691440165042877], [0.26099202036857605, 0.4845621883869171, 0.25444573163986206], [0.2642607092857361, 0.49842071533203125, 0.23731866478919983], [0.27516430616378784, 0.47298017144203186, 0.2518555819988251], [0.266123503446579, 0.49410584568977356, 0.23977063596248627], [0.26386404037475586, 0.49422550201416016, 0.24191045761108398], [0.26186710596084595, 0.4995032250881195, 0.23862965404987335], [0.263101726770401, 0.5088292956352234, 0.2280689924955368], [0.2585914433002472, 0.5090664625167847, 0.23234204947948456], [0.27383315563201904, 0.4747675359249115, 0.25139930844306946], [0.2638143002986908, 0.49912890791893005, 0.23705677688121796], [0.2630397081375122, 0.4895302951335907, 0.2474299818277359], [0.2716045677661896, 0.4683646261692047, 0.2600308060646057], [0.31141552329063416, 0.38850170373916626, 0.300082802772522], [0.26769474148750305, 0.49712684750556946, 0.2351783663034439], [0.2652944028377533, 0.5050656199455261, 0.22963997721672058], [0.26737451553344727, 0.49786582589149475, 0.23475968837738037], [0.26517537236213684, 0.4938091039657593, 0.2410154491662979], [0.2693014144897461, 0.503767192363739, 0.22693143784999847], [0.2634860575199127, 0.49680235981941223, 0.23971158266067505], [0.26536428928375244, 0.49899256229400635, 0.2356431931257248], [0.2904083728790283, 0.4383469521999359, 0.27124470472335815], [0.26468920707702637, 0.4979380667209625, 0.2373727411031723], [0.2649855613708496, 0.49172186851501465, 0.24329257011413574], [0.2631325125694275, 0.5055278539657593, 0.23133961856365204], [0.2598416805267334, 0.509870707988739, 0.2302876114845276], [0.27295076847076416, 0.4633517861366272, 0.2636975049972534], [0.25707635283470154, 0.4917657971382141, 0.2511577904224396], [0.26236632466316223, 0.4889219105243683, 0.24871180951595306], [0.26128318905830383, 0.49236252903938293, 0.24635423719882965], [0.2713295817375183, 0.4666386544704437, 0.26203182339668274], [0.26157131791114807, 0.49827030301094055, 0.24015839397907257], [0.2680545747280121, 0.4785217046737671, 0.25342366099357605], [0.263088583946228, 0.49660223722457886, 0.24030916392803192], [0.26607730984687805, 0.4989796280860901, 0.23494304716587067], [0.264016717672348, 0.48819059133529663, 0.24779273569583893], [0.2675474286079407, 0.49402105808258057, 0.23843148350715637], [0.26570188999176025, 0.48959892988204956, 0.24469922482967377], [0.2660980224609375, 0.4979240298271179, 0.23597800731658936], [0.26706352829933167, 0.4864121675491333, 0.24652430415153503], [0.2665886878967285, 0.4887083172798157, 0.24470292031764984], [0.2649840712547302, 0.49176061153411865, 0.2432553619146347], [0.2880231738090515, 0.4406386911869049, 0.2713381052017212], [0.2627614140510559, 0.5018942356109619, 0.2353443205356598], [0.26218265295028687, 0.5112866759300232, 0.22653065621852875], [0.26407191157341003, 0.5051990151405334, 0.23072905838489532], [0.2737196981906891, 0.46314918994903564, 0.26313111186027527], [0.2648954689502716, 0.505029559135437, 0.23007498681545258], [0.2637994885444641, 0.4779782295227051, 0.25822222232818604], [0.26327717304229736, 0.5063818693161011, 0.23034097254276276], [0.2677401006221771, 0.49958449602127075, 0.2326754331588745], [0.27003782987594604, 0.4916228652000427, 0.2383393496274948], [0.26571592688560486, 0.4767899513244629, 0.25749415159225464], [0.26169922947883606, 0.5029582381248474, 0.23534250259399414], [0.26901137828826904, 0.47154754400253296, 0.2594410479068756], [0.2640586793422699, 0.4915614426136017, 0.2443798929452896], [0.261364221572876, 0.49596118927001953, 0.2426745444536209], [0.2657190263271332, 0.49835771322250366, 0.23592326045036316], [0.2727929651737213, 0.47807642817497253, 0.24913068115711212], [0.26392003893852234, 0.5024784803390503, 0.23360148072242737], [0.2882295548915863, 0.4508384168148041, 0.26093199849128723], [0.26959967613220215, 0.4882473647594452, 0.24215301871299744], [0.26265403628349304, 0.49857741594314575, 0.2387685924768448], [0.26012223958969116, 0.5113822221755981, 0.2284955531358719], [0.2618471086025238, 0.4937022924423218, 0.24445058405399323], [0.2654992938041687, 0.49910515546798706, 0.23539555072784424], [0.2705202102661133, 0.4695005714893341, 0.2599792182445526], [0.26706987619400024, 0.4888845682144165, 0.24404557049274445], [0.25459176301956177, 0.49515074491500854, 0.2502574622631073], [0.2580134868621826, 0.48860710859298706, 0.25337934494018555], [0.264854371547699, 0.4996492266654968, 0.2354964166879654], [0.27659347653388977, 0.45903250575065613, 0.2643740475177765], [0.26730427145957947, 0.4722718596458435, 0.26042383909225464], [0.26838740706443787, 0.4746171236038208, 0.25699546933174133], [0.26373758912086487, 0.4984777569770813, 0.23778462409973145], [0.26465412974357605, 0.4975045323371887, 0.23784132301807404], [0.26640331745147705, 0.4764774739742279, 0.25711923837661743], [0.2582195997238159, 0.48847150802612305, 0.2533089220523834], [0.26972150802612305, 0.4736069142818451, 0.25667157769203186], [0.26270028948783875, 0.48681652545928955, 0.2504831850528717], [0.26568374037742615, 0.4922421872615814, 0.2420741617679596], [0.26679688692092896, 0.49473056197166443, 0.23847262561321259], [0.2647838592529297, 0.5037209391593933, 0.2314952164888382], [0.26148009300231934, 0.4949418008327484, 0.24357807636260986], [0.27138444781303406, 0.47037598490715027, 0.2582395374774933], [0.27077361941337585, 0.492900013923645, 0.23632632195949554], [0.26787838339805603, 0.4914984703063965, 0.24062316119670868], [0.26383960247039795, 0.5059162378311157, 0.23024415969848633], [0.2651820480823517, 0.5000778436660767, 0.23474015295505524], [0.264621376991272, 0.47667333483695984, 0.2587052583694458], [0.2771778404712677, 0.45717525482177734, 0.26564690470695496], [0.2636473774909973, 0.4943254590034485, 0.24202710390090942], [0.2680599093437195, 0.4933520555496216, 0.23858806490898132], [0.25713178515434265, 0.495186448097229, 0.24768179655075073], [0.26530563831329346, 0.49383771419525146, 0.24085664749145508], [0.2679969072341919, 0.49818456172943115, 0.23381853103637695], [0.2642267942428589, 0.5044577121734619, 0.2313154935836792], [0.2612784504890442, 0.5012822151184082, 0.2374393194913864], [0.2720484435558319, 0.47649505734443665, 0.25145649909973145], [0.2746206223964691, 0.46282318234443665, 0.2625562250614166], [0.2609817683696747, 0.48249077796936035, 0.25652751326560974], [0.26583966612815857, 0.49911120533943176, 0.2350490838289261], [0.27743831276893616, 0.4588219225406647, 0.26373985409736633], [0.26956209540367126, 0.46637120842933655, 0.2640667259693146], [0.26460763812065125, 0.5071075558662415, 0.22828476130962372], [0.2638445794582367, 0.5025988817214966, 0.23355653882026672], [0.2659156918525696, 0.49367037415504456, 0.24041391909122467], [0.27349424362182617, 0.46856406331062317, 0.25794169306755066], [0.26320213079452515, 0.5105735063552856, 0.2262244075536728], [0.26467496156692505, 0.4967818260192871, 0.23854325711727142], [0.2676398456096649, 0.47956886887550354, 0.25279131531715393], [0.262666255235672, 0.509223222732544, 0.22811050713062286], [0.2637774646282196, 0.49790677428245544, 0.23831576108932495], [0.25997570157051086, 0.49001243710517883, 0.2500118911266327], [0.26995253562927246, 0.48180198669433594, 0.24824552237987518], [0.2642759680747986, 0.5049718618392944, 0.23075219988822937], [0.2652055621147156, 0.49859917163848877, 0.23619523644447327], [0.2679162323474884, 0.4812210202217102, 0.2508626878261566], [0.2684822976589203, 0.4952399432659149, 0.23627768456935883], [0.2649422287940979, 0.49794816970825195, 0.23710963129997253], [0.2651507258415222, 0.49102452397346497, 0.24382470548152924], [0.2596917450428009, 0.5077255964279175, 0.23258259892463684], [0.26286980509757996, 0.5069637298583984, 0.2301664650440216], [0.26287633180618286, 0.5081875324249268, 0.22893615067005157], [0.2646990418434143, 0.4776306450366974, 0.2576703131198883], [0.2641565203666687, 0.49107638001441956, 0.24476709961891174], [0.2662995457649231, 0.5011040568351746, 0.23259639739990234], [0.2665804922580719, 0.5111191868782043, 0.22230029106140137]]
standard_test: {'validation_loss': 0.7267700036366781, 'validation_time': 0.47025108337402344, 'num_samples': 143, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'auc_roc_ovr': 0.3549749065790557, 'auc_roc_ovo': np.float64(0.37556612874217227), 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}, 'intra_class_similarity_mean': np.float32(0.9960874), 'intra_class_similarity_std': np.float32(0.004027016), 'inter_class_similarity_mean': np.float32(0.9958434), 'inter_class_similarity_std': np.float32(0.0043481365), 'class_separation': np.float32(0.00024396181), 'eer': 1.0, 'eer_threshold': 0.0}, 'predictions': [1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 'targets': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2], 'probabilities': [[0.261184960603714, 0.4855104088783264, 0.253304660320282], [0.26502761244773865, 0.48785409331321716, 0.24711836874485016], [0.25821352005004883, 0.4888419508934021, 0.25294455885887146], [0.26847562193870544, 0.4840279519557953, 0.24749639630317688], [0.26119062304496765, 0.4895646572113037, 0.2492447793483734], [0.25792446732521057, 0.48690104484558105, 0.2551744878292084], [0.2545399069786072, 0.49605169892311096, 0.24940842390060425], [0.2672329246997833, 0.4720095098018646, 0.26075759530067444], [0.27588826417922974, 0.4695679545402527, 0.2545437812805176], [0.2702755928039551, 0.4702548384666443, 0.259469598531723], [0.26947078108787537, 0.47766730189323425, 0.2528618276119232], [0.27236494421958923, 0.46338117122650146, 0.2642538547515869], [0.26384732127189636, 0.492117702960968, 0.24403499066829681], [0.26074376702308655, 0.48340320587158203, 0.2558530867099762], [0.2635514736175537, 0.4993364214897156, 0.2371121644973755], [0.2793697714805603, 0.4600013792514801, 0.260628879070282], [0.2687304615974426, 0.4728254973888397, 0.25844407081604004], [0.2650108337402344, 0.506981611251831, 0.22800754010677338], [0.27055731415748596, 0.47412461042404175, 0.2553181052207947], [0.26973575353622437, 0.4751695394515991, 0.2550947070121765], [0.26710304617881775, 0.48145002126693726, 0.25144699215888977], [0.2723742723464966, 0.47737863659858704, 0.250247061252594], [0.26162436604499817, 0.4906712472438812, 0.24770431220531464], [0.26857271790504456, 0.4828528165817261, 0.24857451021671295], [0.27049124240875244, 0.4739808440208435, 0.25552791357040405], [0.2610608637332916, 0.49210599064826965, 0.2468331903219223], [0.26558998227119446, 0.484563946723938, 0.24984605610370636], [0.2584162950515747, 0.49431928992271423, 0.24726447463035583], [0.25656038522720337, 0.4907509982585907, 0.25268861651420593], [0.25652649998664856, 0.4931589961051941, 0.25031447410583496], [0.2569662928581238, 0.49172067642211914, 0.25131306052207947], [0.26626524329185486, 0.4882645010948181, 0.24547025561332703], [0.25839996337890625, 0.48575443029403687, 0.25584566593170166], [0.2654722034931183, 0.5001062750816345, 0.2344214767217636], [0.26393747329711914, 0.49040552973747253, 0.24565696716308594], [0.278157114982605, 0.4497531056404114, 0.27208977937698364], [0.26113638281822205, 0.486333429813385, 0.25253015756607056], [0.25509271025657654, 0.494528204202652, 0.2503790557384491], [0.25768232345581055, 0.4907529950141907, 0.2515646517276764], [0.2644014060497284, 0.48996326327323914, 0.24563533067703247], [0.25658878684043884, 0.4917580187320709, 0.2516532242298126], [0.2661421000957489, 0.4784094989299774, 0.25544843077659607], [0.272273987531662, 0.4712011218070984, 0.25652486085891724], [0.26034530997276306, 0.486104816198349, 0.25354981422424316], [0.2601625323295593, 0.49204105138778687, 0.2477964162826538], [0.2805248200893402, 0.45387014746665955, 0.26560503244400024], [0.2686176896095276, 0.4865526258945465, 0.2448296844959259], [0.26190024614334106, 0.49207592010498047, 0.24602381885051727], [0.27862921357154846, 0.4576135277748108, 0.26375722885131836], [0.2648402750492096, 0.483436644077301, 0.251723051071167], [0.25684425234794617, 0.49441447854042053, 0.2487412393093109], [0.26523271203041077, 0.4796966016292572, 0.2550707161426544], [0.26762428879737854, 0.4724019765853882, 0.25997382402420044], [0.2627633810043335, 0.4845157265663147, 0.2527208924293518], [0.2647069990634918, 0.48218438029289246, 0.2531086504459381], [0.2571265995502472, 0.4892635643482208, 0.25360989570617676], [0.2663002908229828, 0.48734143376350403, 0.24635827541351318], [0.26340848207473755, 0.49968579411506653, 0.2369057536125183], [0.26341256499290466, 0.47922176122665405, 0.25736570358276367], [0.26679688692092896, 0.5011491775512695, 0.2320539355278015], [0.2640385925769806, 0.4975512623786926, 0.23841015994548798], [0.25706249475479126, 0.49138399958610535, 0.251553475856781], [0.26380684971809387, 0.5037788152694702, 0.2324143946170807], [0.27741947770118713, 0.4646858870983124, 0.25789469480514526], [0.272723525762558, 0.46751105785369873, 0.2597653865814209], [0.26680609583854675, 0.49715644121170044, 0.23603740334510803], [0.2605472505092621, 0.5107880234718323, 0.22866474092006683], [0.26449471712112427, 0.49040287733078003, 0.24510237574577332], [0.2634216248989105, 0.5062246918678284, 0.2303536981344223], [0.2573058605194092, 0.4900277256965637, 0.2526664435863495], [0.26361083984375, 0.48846107721328735, 0.24792800843715668], [0.2669690251350403, 0.493459552526474, 0.23957140743732452], [0.26458585262298584, 0.4916524291038513, 0.24376168847084045], [0.2636070251464844, 0.5047328472137451, 0.2316601574420929], [0.26479360461235046, 0.49664634466171265, 0.2385600507259369], [0.26409250497817993, 0.49917179346084595, 0.2367357462644577], [0.2646801173686981, 0.4877946674823761, 0.24752521514892578], [0.2659223973751068, 0.4933943450450897, 0.24068324267864227], [0.2631888687610626, 0.5029144287109375, 0.23389671742916107], [0.2650880813598633, 0.4884413480758667, 0.2464705854654312], [0.27053481340408325, 0.48566949367523193, 0.2437957376241684], [0.2690788507461548, 0.49307289719581604, 0.2378482073545456], [0.2560591995716095, 0.4919504225254059, 0.25199034810066223], [0.2705286741256714, 0.46598443388938904, 0.2634868919849396], [0.26320552825927734, 0.5119261741638184, 0.2248682826757431], [0.267072856426239, 0.48259010910987854, 0.25033700466156006], [0.2632271647453308, 0.5095873475074768, 0.22718550264835358], [0.26186418533325195, 0.49134260416030884, 0.24679327011108398], [0.26875242590904236, 0.49779438972473145, 0.23345324397087097], [0.26529884338378906, 0.5020872950553894, 0.23261387646198273], [0.2641504406929016, 0.5029255151748657, 0.23292405903339386], [0.2656996250152588, 0.4966418743133545, 0.23765847086906433], [0.2702951729297638, 0.4644054174423218, 0.26529937982559204], [0.2607957124710083, 0.4978049099445343, 0.24139931797981262], [0.2626131772994995, 0.4935840964317322, 0.24380265176296234], [0.2630622088909149, 0.4934960603713989, 0.24344179034233093], [0.26722970604896545, 0.47454768419265747, 0.25822266936302185], [0.2621440589427948, 0.4918625056743622, 0.24599343538284302], [0.26442575454711914, 0.4920050799846649, 0.24356913566589355], [0.26070359349250793, 0.49127236008644104, 0.2480241060256958], [0.2659478783607483, 0.4728337526321411, 0.2612183094024658], [0.26263391971588135, 0.4970043897628784, 0.24036170542240143], [0.27313950657844543, 0.465713769197464, 0.2611466348171234], [0.2642384469509125, 0.5111084580421448, 0.22465308010578156], [0.26563170552253723, 0.4980522394180298, 0.23631611466407776], [0.2641066014766693, 0.5080733895301819, 0.22781997919082642], [0.2674694359302521, 0.4995659291744232, 0.2329646348953247], [0.26015999913215637, 0.5042216777801514, 0.23561836779117584], [0.27067387104034424, 0.46767735481262207, 0.2616487741470337], [0.2673896849155426, 0.5060433149337769, 0.22656702995300293], [0.26705485582351685, 0.49741557240486145, 0.2355296015739441], [0.2634315490722656, 0.49230554699897766, 0.24426288902759552], [0.26798489689826965, 0.46927541494369507, 0.2627396583557129], [0.26304322481155396, 0.5099937915802002, 0.22696296870708466], [0.2669427692890167, 0.49238675832748413, 0.24067053198814392], [0.2624886929988861, 0.5015621781349182, 0.23594912886619568], [0.2643607556819916, 0.4931698441505432, 0.24246934056282043], [0.264608770608902, 0.4988705515861511, 0.2365206927061081], [0.2675653398036957, 0.4921415448188782, 0.24029314517974854], [0.2560276389122009, 0.5037729144096375, 0.24019943177700043], [0.2578684389591217, 0.4896048903465271, 0.2525266110897064], [0.2636682391166687, 0.48901861906051636, 0.24731312692165375], [0.2645813524723053, 0.49427860975265503, 0.2411399632692337], [0.2641514539718628, 0.4983564019203186, 0.23749211430549622], [0.263198584318161, 0.4949099123477936, 0.24189145863056183], [0.26477959752082825, 0.49476197361946106, 0.24045851826667786], [0.2636794149875641, 0.5007743835449219, 0.23554623126983643], [0.26832929253578186, 0.4985484480857849, 0.23312224447727203], [0.2628946006298065, 0.48811763525009155, 0.2489878237247467], [0.26218992471694946, 0.48601770401000977, 0.25179240107536316], [0.27199631929397583, 0.4659452438354492, 0.26205840706825256], [0.26468029618263245, 0.5033774971961975, 0.23194217681884766], [0.2655929625034332, 0.48889675736427307, 0.2455102950334549], [0.2641844153404236, 0.4904041290283203, 0.2454114705324173], [0.2649112939834595, 0.5009515285491943, 0.2341371327638626], [0.2633408010005951, 0.4997940957546234, 0.23686519265174866], [0.26026803255081177, 0.5087888836860657, 0.23094309866428375], [0.26711276173591614, 0.49061185121536255, 0.2422754168510437], [0.26479628682136536, 0.5051676630973816, 0.23003609478473663], [0.2681112289428711, 0.4686660170555115, 0.2632228136062622], [0.30452442169189453, 0.39548295736312866, 0.2999926209449768], [0.272600919008255, 0.4655373990535736, 0.2618615925312042], [0.26831212639808655, 0.48833921551704407, 0.24334867298603058]], 'test_type': 'standard'}
augmentation_robustness: {'noise_light': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}, 'noise_moderate': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}, 'noise_heavy': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}, 'speed_light': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}, 'speed_moderate': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}, 'volume_light': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}, 'volume_moderate': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}, 'combined_light': {'loss': 0.7267700036366781, 'accuracy': 0.13286713286713286, 'metrics': {'accuracy': 0.13286713286713286, 'precision_Class_0': np.float64(0.0), 'recall_Class_0': np.float64(0.0), 'f1_Class_0': np.float64(0.0), 'support_Class_0': np.int64(37), 'precision_Class_1': np.float64(0.13286713286713286), 'recall_Class_1': np.float64(1.0), 'f1_Class_1': np.float64(0.2345679012345679), 'support_Class_1': np.int64(19), 'precision_Class_2': np.float64(0.0), 'recall_Class_2': np.float64(0.0), 'f1_Class_2': np.float64(0.0), 'support_Class_2': np.int64(87), 'precision_macro': np.float64(0.04428904428904429), 'recall_macro': np.float64(0.3333333333333333), 'f1_macro': np.float64(0.07818930041152262), 'precision_weighted': np.float64(0.01765367499633234), 'recall_weighted': np.float64(0.13286713286713286), 'f1_weighted': np.float64(0.03116636449969783), 'confusion_matrix': [[0, 37, 0], [0, 19, 0], [0, 87, 0]], 'classification_report': {'Class_0': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 37.0}, 'Class_1': {'precision': 0.13286713286713286, 'recall': 1.0, 'f1-score': 0.2345679012345679, 'support': 19.0}, 'Class_2': {'precision': 0.0, 'recall': 0.0, 'f1-score': 0.0, 'support': 87.0}, 'accuracy': 0.13286713286713286, 'macro avg': {'precision': 0.04428904428904429, 'recall': 0.3333333333333333, 'f1-score': 0.07818930041152262, 'support': 143.0}, 'weighted avg': {'precision': 0.01765367499633234, 'recall': 0.13286713286713286, 'f1-score': 0.03116636449969783, 'support': 143.0}}}}}
performance_profile: {'inference_performance': {'mean_inference_time': np.float64(0.0008207382052205503), 'std_inference_time': np.float64(0.000325613196581573), 'min_inference_time': np.float64(0.000614962016697973), 'max_inference_time': np.float64(0.00167938502272591), 'median_inference_time': np.float64(0.0006612160068470985), 'throughput_samples_per_second': np.float64(1218.4153164056474)}, 'memory_usage': {'initial_memory_mb': 67.14697265625, 'peak_memory_mb': 75.1025390625, 'final_memory_mb': 67.1474609375, 'memory_increase_mb': 7.95556640625}, 'model_parameters': {'total_parameters': 2586371, 'trainable_parameters': 2586371, 'non_trainable_parameters': 0}, 'model_size': {'total_size_bytes': 10360412, 'total_size_mb': 9.88045883178711, 'parameter_size_mb': 9.866222381591797, 'buffer_size_mb': 0.0142364501953125}, 'batch_performance': {'batch_1': {'inference_time': np.float64(0.0006596447550691664), 'throughput': np.float64(1515.9674844911729)}, 'batch_8': {'inference_time': np.float64(0.0015107915416592733), 'throughput': np.float64(5295.237482739514)}, 'batch_16': {'inference_time': np.float64(0.007534609900903888), 'throughput': np.float64(2123.5339600104)}, 'batch_32': {'inference_time': np.float64(0.006903560695354827), 'throughput': np.float64(4635.289151804187)}, 'batch_64': {'inference_time': np.float64(0.0142083060083678), 'throughput': np.float64(4504.407489697084)}}}
robustness_tests: {'confidence_thresholds': {'threshold_0.5': {'accuracy': np.float64(0.0), 'coverage': np.float64(0.16783216783216784), 'num_samples': np.int64(24)}}, 'input_corruptions': {'zero_out_10': {'loss': 0.7175495425860087, 'accuracy': np.float64(0.13286713286713286)}, 'zero_out_25': {'loss': 0.7003007133801779, 'accuracy': np.float64(0.13286713286713286)}, 'gaussian_noise_01': {'loss': 0.7303666273752848, 'accuracy': np.float64(0.13286713286713286)}, 'gaussian_noise_02': {'loss': 0.7397465109825134, 'accuracy': np.float64(0.13286713286713286)}, 'dropout_10': {'loss': 0.7523035208384196, 'accuracy': np.float64(0.13286713286713286)}, 'dropout_25': {'loss': 0.7997564077377319, 'accuracy': np.float64(0.13286713286713286)}}, 'adversarial_robustness': {'epsilon_0.01': {'loss': 0.7407722671826681, 'accuracy': np.float64(0.13286713286713286)}, 'epsilon_0.05': {'loss': 0.7932684421539307, 'accuracy': np.float64(0.13286713286713286)}, 'epsilon_0.1': {'loss': 0.8522491653760275, 'accuracy': np.float64(0.13286713286713286)}, 'epsilon_0.2': {'loss': 0.9539903402328491, 'accuracy': np.float64(0.13286713286713286)}}}

Wakeword Detection Specific Results

Classification Performance

Confusion Matrix

Actual\Predicted    custom system_c negative 
      custom         0       37        0 
system_comma         0       19        0 
    negative         0       87        0

Robustness Tests

Input Corruption Robustness

zero_out_10: 0.1329 accuracy
zero_out_25: 0.1329 accuracy
gaussian_noise_01: 0.1329 accuracy
gaussian_noise_02: 0.1329 accuracy
dropout_10: 0.1329 accuracy
dropout_25: 0.1329 accuracy

Performance Profile

Inference time: 0.82 ms
Throughput: 1218.4 samples/sec
Model size: 9.88 MB

Deployment Recommendations

Detection Threshold

Recommended threshold: 0.5
Monitor false positive/negative rates in production
Consider threshold adjustment based on use case requirements

Hardware Requirements

Suitable for real-time processing on most devices

Model Optimization

Consider using lightweight variant for edge deployment
Model has been reparameterized for efficient inference

training_report.md 49 KB Bezpośredni odnośnik Historia Czysty

Training Report: mein_modell2

Basic Information

Model Architecture

Training Results

Dataset Information

Hardware Information

Wakeword Information

Test Results

Wakeword Detection Specific Results

Classification Performance

Confusion Matrix

Robustness Tests

Input Corruption Robustness

Performance Profile

Deployment Recommendations

Detection Threshold

Hardware Requirements

Model Optimization

training_report.md 49 KB

Bezpośredni odnośnik Historia Czysty