--- license: other base_model: nvidia/mit-b0 tags: - vision - image-segmentation - generated_from_trainer model-index: - name: segformer-b0-finetuned results: [] --- # segformer-b0-finetuned This model is a fine-tuned version of [nvidia/mit-b0](https://huggingface.co/nvidia/mit-b0) on the segments/sidewalk-semantic dataset. It achieves the following results on the evaluation set: - Loss: 0.5754 - Mean Iou: 0.2781 - Mean Accuracy: 0.3329 - Overall Accuracy: 0.8463 - Per Category Iou: [0.0, 0.7432161129630078, 0.854265404236928, 0.4606401052721709, 0.6557337899613191, 0.4079867997829282, nan, 0.37471812939221005, 0.2905341043386837, 0.0, 0.7537587486511262, 0.0, 0.0, nan, 0.0, 0.019848656872972055, 0.0, 0.0, 0.7115931639469374, nan, 0.3661808713379434, 0.13378413732653244, 0.0, nan, 0.0, 0.23570903658727577, 0.0, 0.0, 0.8461792428096935, 0.7553019453875489, 0.9045825383881589, 0.0, 0.0, 0.10651182264386322, 0.0] - Per Category Accuracy: [0.0, 0.8511274737458464, 0.9523527728262475, 0.7305783824446481, 0.7179823443918317, 0.5112934364530293, nan, 0.4671955914617317, 0.39620749876026823, 0.0, 0.9325380267720194, 0.0, 0.0, nan, 0.0, 0.019920987025907694, 0.0, 0.0, 0.9114075726560573, nan, 0.4767221960460328, 0.14080931640440494, 0.0, nan, 0.0, 0.2902864462270403, 0.0, 0.0, 0.9417630123717813, 0.8946072183599384, 0.9626510283976625, 0.0, 0.0, 0.12104456389804058, 0.0] ## Model description More information needed ## Intended uses & limitations More information needed ## Training and evaluation data More information needed ## Training procedure ### Training hyperparameters The following hyperparameters were used during training: - learning_rate: 5e-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: 5 ### Training results | Training Loss | Epoch | Step | Validation Loss | Mean Iou | Mean Accuracy | Overall Accuracy | Per Category Iou | Per Category Accuracy | |:-------------:|:-----:|:----:|:---------------:|:--------:|:-------------:|:----------------:|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------:| | 0.6702 | 1.0 | 400 | 0.7027 | 0.2195 | 0.2629 | 0.8084 | [0.0, 0.6792133271879363, 0.7849474176188894, 0.058328120930117175, 0.6300690246185523, 0.25673461142351706, nan, 0.3004378389548008, 0.0, 0.0, 0.6990982425959871, 0.0, 0.0, nan, 0.0, 0.0, 0.0, 0.0, 0.6732565754528083, nan, 0.2872208279956378, 0.00018886917594771138, 0.0, nan, 0.0, 0.04832964515878562, 0.0, 0.0, 0.8103176781510323, 0.6868793807107686, 0.8837386972387465, 0.0, 0.0, 0.005328297121957592, 0.0] | [0.0, 0.766565968404771, 0.9690642801889813, 0.05881921997258986, 0.6774284161746986, 0.2997799346472589, nan, 0.3604784648706302, 0.0, 0.0, 0.9238506174053699, 0.0, 0.0, nan, 0.0, 0.0, 0.0, 0.0, 0.879793151515724, nan, 0.3814629549547224, 0.0001889329509116015, 0.0, nan, 0.0, 0.049252727470549255, 0.0, 0.0, 0.92204060605898, 0.9165746690826654, 0.9399161153753854, 0.0, 0.0, 0.005432287151031732, 0.0] | | 0.3787 | 2.0 | 800 | 0.6242 | 0.2529 | 0.3048 | 0.8336 | [0.0, 0.7200155470846057, 0.8500725277201905, 0.4283923004744409, 0.6393695507210657, 0.35502977816991826, nan, 0.35184539253673836, 0.007092160389449536, 0.0, 0.7043921336269122, 0.0, 0.0, nan, 0.0, 0.0, 0.0, 0.0, 0.6907551328541987, nan, 0.30912243068319983, 0.08139332433133045, 0.0, nan, 0.0, 0.188100947571913, 0.0, 0.0, 0.8385996121617959, 0.7447284921504436, 0.8944303097872178, 0.0, 0.0, 0.03573598106370054, 0.0] | [0.0, 0.8883665924421738, 0.9304929976181545, 0.6379432034836245, 0.684408431327974, 0.47546406343261166, nan, 0.46904807570859863, 0.007098812013540027, 0.0, 0.9393400531924911, 0.0, 0.0, nan, 0.0, 0.0, 0.0, 0.0, 0.9030339520321862, nan, 0.38990531390988364, 0.08456558485802895, 0.0, nan, 0.0, 0.22015360876747014, 0.0, 0.0, 0.9441960488646456, 0.8731757915405423, 0.9605664489132197, 0.0, 0.0, 0.0402316629096584, 0.0] | | 0.5272 | 3.0 | 1200 | 0.5910 | 0.2640 | 0.3116 | 0.8409 | [0.0, 0.7417372358583919, 0.8491040334276788, 0.5026923409983705, 0.6531274799797995, 0.39671746797276214, nan, 0.3489393985838212, 0.10917691003765771, 0.0, 0.7340253134142348, 0.0, 0.0, nan, 0.0, 0.0015906711475390872, 0.0, 0.0, 0.6888929372303201, nan, 0.3000933215998536, 0.09430463167198322, 0.0, nan, 0.0, 0.22079460109263602, 0.0, 0.0, 0.8375430585634192, 0.7432674869295846, 0.8979452744557967, 0.0, 0.0, 0.06465850622764936, 0.0] | [0.0, 0.8742236634719254, 0.9469015516965127, 0.6845976124235456, 0.7153136135405865, 0.4838119970613863, nan, 0.426013369696913, 0.11595265302602359, 0.0, 0.9307602006332041, 0.0, 0.0, nan, 0.0, 0.0015906711475390872, 0.0, 0.0, 0.9211760279107508, nan, 0.3640489760089845, 0.09822101537391639, 0.0, nan, 0.0, 0.27522724799952525, 0.0, 0.0, 0.9529837430876597, 0.8328291532204077, 0.9626684254773068, 0.0, 0.0, 0.07271163516559737, 0.0] | | 1.0028 | 4.0 | 1600 | 0.5819 | 0.2749 | 0.3265 | 0.8451 | [0.0, 0.7442319582171808, 0.8549546101758252, 0.4558465282708946, 0.6592549345415454, 0.40147520263994, nan, 0.3560786579426865, 0.2724675418610539, 0.0, 0.7615078761694535, 0.0, 0.0, nan, 0.0, 0.01480792989181263, 0.0, 0.0, 0.6971675525618446, nan, 0.3289306001269004, 0.1400376526683254, 0.0, nan, 0.0, 0.2330975509072671, 0.0, 0.0, 0.8412274001878343, 0.7610379911113287, 0.9042555512089849, 0.0, 0.0, 0.09514392306437187, 0.0] | [0.0, 0.8831147123698461, 0.9454704608007805, 0.7031575260834061, 0.7194367374187804, 0.4940337653992012, nan, 0.4313144685568867, 0.3654837110023501, 0.0, 0.911745245214873, 0.0, 0.0, nan, 0.0, 0.01483662361250094, 0.0, 0.0, 0.9249937012782616, nan, 0.39416005628239303, 0.1487585698177601, 0.0, nan, 0.0, 0.28903522220353906, 0.0, 0.0, 0.9446123320713813, 0.8885032163816529, 0.9544933330809051, 0.0, 0.0, 0.10758314548292006, 0.0] | | 1.3105 | 5.0 | 2000 | 0.5754 | 0.2781 | 0.3329 | 0.8463 | [0.0, 0.7432161129630078, 0.854265404236928, 0.4606401052721709, 0.6557337899613191, 0.4079867997829282, nan, 0.37471812939221005, 0.2905341043386837, 0.0, 0.7537587486511262, 0.0, 0.0, nan, 0.0, 0.019848656872972055, 0.0, 0.0, 0.7115931639469374, nan, 0.3661808713379434, 0.13378413732653244, 0.0, nan, 0.0, 0.23570903658727577, 0.0, 0.0, 0.8461792428096935, 0.7553019453875489, 0.9045825383881589, 0.0, 0.0, 0.10651182264386322, 0.0] | [0.0, 0.8511274737458464, 0.9523527728262475, 0.7305783824446481, 0.7179823443918317, 0.5112934364530293, nan, 0.4671955914617317, 0.39620749876026823, 0.0, 0.9325380267720194, 0.0, 0.0, nan, 0.0, 0.019920987025907694, 0.0, 0.0, 0.9114075726560573, nan, 0.4767221960460328, 0.14080931640440494, 0.0, nan, 0.0, 0.2902864462270403, 0.0, 0.0, 0.9417630123717813, 0.8946072183599384, 0.9626510283976625, 0.0, 0.0, 0.12104456389804058, 0.0] | ### Framework versions - Transformers 4.38.2 - Pytorch 2.2.1+cu121 - Datasets 2.18.0 - Tokenizers 0.15.2