nommis_segformer5

• Loss: 0.4968
• Mean Iou: 0.3573
• Mean Accuracy: 0.4923
• Overall Accuracy: 0.8411
• Per Category Iou: [0.74975317392031, 0.7794410147887274, 0.9477603899927828, 0.7656932606023603, 0.723857664111405, 0.8716512692421345, 0.863022621721695, 0.9240294561369442, 0.7242633568910231, 0.6134294305818301, 0.48159993522529454, 0.6128290042426419, 0.7376616340096556, 0.14249931825756734, 0.35176633571923344, 0.5620357131338759, 0.8638289260658392, 0.415410472072499, 0.5517729659566936, 0.6172647443104129, 0.8124566178707076, 0.0, 0.6428092209430266, 0.0, 0.2104813929668829, 0.4251821264997899, 0.9380425651078129, 0.8474509980895648, 0.5821050652926282, 0.4506812149650353, 0.6943803190572994, 0.0, 0.17265479960685814, 0.23995127892813642, 0.0, 0.0, 0.2724044481824129, 0.0, 0.392571160462934, 0.21584808909542505, 0.2496028068317225, 0.0, 0.09854609688158163, 0.022266401590457258, nan, 0.0, 0.9471716274248482, 0.6674884632404071, nan, 0.8374251999328897, 0.0, 0.49230769230769234, 0.0, 0.4042453273222504, 0.0, nan, nan, 0.5250632076508739, 0.9178690960601052, 0.2001023541453429, nan, nan, 0.5349142857142857, 0.0, 0.0, 0.8767808047988003, 0.6694305040366572, 0.7116823374785158, 0.6645531400966184, 0.0, 0.7045357686453577, nan, 0.33535395047775607, nan, 0.7918331226295828, nan, 0.0, nan, 0.0, 0.10583069982872169, nan, 0.18884178859565157, 0.11131725417439703, nan, nan, 0.617096297708105, 0.5133936387510942, 0.023268853378188955, 0.0, 0.8122001370801919, nan, 0.0, 0.6120563928050559, 0.015886896953796824, 0.0, 0.0, nan, nan, nan, nan, nan, nan, 0.13397003128602009, nan, 0.6251570200019325, nan, nan, nan, nan, nan, nan, nan, 0.0, nan, nan, 0.0, 0.0, nan, nan, nan, nan, 0.2817006400534663, 0.0, 0.0, nan, 0.4339781328847771, nan, 0.0, nan, nan, nan, nan, nan, nan, 0.37268373932958565, 0.0, 0.0, nan, 0.0, nan, nan, nan, 0.36669593210418494, 0.0, nan, nan, nan, 0.0009633911368015414, 0.0, 0.029017047515415305]
• Per Category Accuracy: [0.8500475032918948, 0.9090727442266007, 0.9790694696013319, 0.9066653475744558, 0.8036125736948979, 0.9619617217520251, 0.9428051445352432, 0.9501946899936256, 0.7538857522417801, 0.66818192208844, 0.6985935343688522, 0.6854295277054667, 0.8803769797226216, 0.33735832815264694, 0.3681730304494925, 0.7064864864864865, 0.9939117530177338, 0.49297531916389165, 0.6516846102923243, 0.7121253549269113, 0.9651162072522219, nan, 0.9039209026755727, 0.0, 0.23541766109785203, 0.9826890059795157, 0.966677789100616, 0.9145306082693463, 0.662943420936646, 0.6884258672110558, 0.9477743551155736, nan, 0.27695728338498854, 0.6679707876890976, 0.0, nan, 0.6349379220213461, nan, 0.41777903531839816, 0.21584808909542505, 0.3558586192251427, 0.0, 0.09857623432668457, 0.022798208712172616, nan, nan, 0.9908985282726569, 0.7127235909551131, nan, 0.884569943289225, nan, 0.494096872882574, 0.0, 0.4246669955599408, nan, nan, nan, 0.9346443596516975, 0.9835706951203179, 0.22040586245772267, nan, nan, 0.665907878356749, nan, 0.0, 0.9194746192154236, 0.713322483143455, 0.8259364641694289, 0.7627227722772277, 0.0, 0.9854381333560419, nan, 0.35627353599301426, nan, 0.9498058959116826, nan, 0.0, nan, nan, 0.13212661506490378, nan, 0.47752420470262796, 0.12829650748396293, nan, nan, 0.7659234727068425, 0.5280946091967823, 0.027880416526704737, nan, 0.8304134548002803, nan, nan, 0.9665496819477956, 0.016129032258064516, nan, 0.0, nan, nan, nan, nan, nan, nan, 0.13411578530924437, nan, 0.6578626264680462, nan, nan, nan, nan, nan, nan, nan, 0.0, nan, nan, 0.0, 0.0, nan, nan, nan, nan, 0.2817223650385604, nan, 0.0, nan, 0.48074534161490684, nan, 0.0, nan, nan, nan, nan, nan, nan, 0.463010863942059, 0.0, 0.0, nan, 0.0, nan, nan, nan, 0.3707466219548279, 0.0, nan, nan, nan, 0.0009633911368015414, 0.0, 0.029017047515415305]

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

• learning_rate: 6e-05
• train_batch_size: 2
• eval_batch_size: 2
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Framework versions

• Transformers 4.28.0
• Pytorch 2.0.1+cu118
• Datasets 2.14.5
• Tokenizers 0.13.3