nateraw / mit-b0-finetuned-sidewalks

nateraw/mit-b0-finetuned-sidewalks

This model is a fine-tuned version of nvidia/mit-b0 on an unknown dataset. It achieves the following results on the evaluation set:

• Train Loss: 0.5197
• Validation Loss: 0.6268
• Validation Mean Iou: 0.2719
• Validation Mean Accuracy: 0.3442
• Validation Overall Accuracy: 0.8180
• Validation Per Category Iou: [0. 0.62230678 0.81645513 0.18616589 0.66669478 0.30574734 nan 0.36681201 0.31128062 0. 0.76635363 0.

1.            nan 0.         0.37874505 0.         0.
   

0.68193241 0. 0.48867838 0.25809644 0. nan 0. 0.25765818 0. 0. 0.81965205 0.71604385 0.9214592 0. 0.00636635 0.12957446 0. ]

• Validation Per Category Accuracy: [0. 0.89469845 0.88320521 0.45231002 0.72104833 0.3386303 nan 0.53522723 0.72026843 0. 0.93197124 0.

1.            nan 0.         0.45525816 0.         0.
   

0.87276184 0. 0.60762821 0.29654901 0. nan 0. 0.32162193 0. 0. 0.90797988 0.89199119 0.96388697 0. 0.00646084 0.21171965 0. ]

• Epoch: 5

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:

• optimizer: {'name': 'Adam', 'learning_rate': 6e-05, 'decay': 0.0, 'beta_1': 0.9, 'beta_2': 0.999, 'epsilon': 1e-07, 'amsgrad': False}
• training_precision: float32

Training results

    nan 0.279315   0.28998136 0.         0.71924852 0.

1.            nan 0.         0.10241989 0.         0.
   

0.60537245 0. 0.37966409 0.0624908 0. 0. 0. 0.11869763 0. 0. 0.79675107 0.70541969 0.89177953 0. 0. 0.01097213 0. ] | [0. 0.70687024 0.92710849 0.47653578 0.6809956 0.28562204 nan 0.35954555 0.53804171 0. 0.87451178 0. 0. nan 0. 0.10473185 0. 0. 0.88548482 0. 0.52011987 0.06421075 0. nan 0. 0.13802701 0. 0. 0.9278545 0.83106582 0.94693817 0. 0. 0.01170072 0. ] | 1 | | 0.7051 | 0.6513 | 0.2568 | 0.3210 | 0.8151 | [0.00000000e+00 6.31500555e-01 8.33347761e-01 2.40727740e-01 6.71879162e-01 2.32727132e-01 nan 3.15720178e-01 3.22578864e-01 0.00000000e+00 7.51066980e-01 0.00000000e+00 0.00000000e+00 nan 0.00000000e+00 3.01090014e-01 0.00000000e+00 0.00000000e+00 6.56592309e-01 0.00000000e+00 3.82317489e-01 2.25385079e-01 0.00000000e+00 nan 0.00000000e+00 2.34975219e-01 0.00000000e+00 0.00000000e+00 7.92710603e-01 6.82508692e-01 9.02369099e-01 0.00000000e+00 5.10019193e-04 4.02361131e-02 0.00000000e+00] | [0.00000000e+00 7.76355941e-01 9.39707165e-01 3.90888278e-01 7.70256989e-01 2.84066636e-01 nan 4.57106724e-01 6.33498392e-01 0.00000000e+00 9.05789013e-01 0.00000000e+00 0.00000000e+00 nan 0.00000000e+00 3.57230962e-01 0.00000000e+00 0.00000000e+00 8.45761217e-01 0.00000000e+00 5.16681541e-01 2.82796479e-01 0.00000000e+00 nan 0.00000000e+00 3.07634724e-01 0.00000000e+00 0.00000000e+00 9.04391068e-01 8.86212453e-01 9.64570665e-01 0.00000000e+00 5.17411580e-04 4.71742075e-02 0.00000000e+00] | 2 | | 0.6294 | 0.6365 | 0.2695 | 0.3320 | 0.8244 | [0. 0.63840754 0.83879521 0.31781353 0.69394774 0.22324776 nan 0.35012894 0.31369877 0. 0.7683448 0. 0. nan 0. 0.36532292 0. 0. 0.65554136 0. 0.37438724 0.25682621 0. nan 0. 0.23051151 0. 0. 0.81818163 0.7633018 0.91092518 0. 0.00145576 0.10215516 0. ] | [0. 0.76103704 0.95305272 0.43848725 0.78760908 0.25645014 nan 0.48971828 0.61853472 0. 0.90793733 0. 0. nan 0. 0.48772201 0. 0. 0.84205031 0. 0.53308407 0.36285878 0. nan 0. 0.27953916 0. 0. 0.93079576 0.87079757 0.96477884 0. 0.00147054 0.13899972 0. ] | 3 | | 0.5686 | 0.6122 | 0.2715 | 0.3360 | 0.8256 | [0.00000000e+00 6.38345814e-01 8.56252996e-01 3.07043269e-01 6.87537894e-01 3.06534041e-01 nan 3.84145525e-01 3.19438916e-01 0.00000000e+00 7.57233152e-01 0.00000000e+00 0.00000000e+00 nan 0.00000000e+00 4.06585843e-01 0.00000000e+00 0.00000000e+00 6.47648546e-01 2.91885581e-04 4.00547422e-01 1.97261484e-01 0.00000000e+00 nan 0.00000000e+00 2.20793008e-01 0.00000000e+00 0.00000000e+00 8.19526784e-01 7.19306080e-01 9.20192720e-01 0.00000000e+00 2.23374930e-03 9.77508243e-02 0.00000000e+00] | [0.00000000e+00 7.89438910e-01 9.16367241e-01 4.32251205e-01 7.89740409e-01 4.88566404e-01 nan 5.36825005e-01 6.47787376e-01 0.00000000e+00 9.32641501e-01 0.00000000e+00 0.00000000e+00 nan 0.00000000e+00 4.73813253e-01 0.00000000e+00 0.00000000e+00 9.09004353e-01 2.91885581e-04 4.37175308e-01 2.25663128e-01 0.00000000e+00 nan 0.00000000e+00 2.60992057e-01 0.00000000e+00 0.00000000e+00 9.19328058e-01 9.02898346e-01 9.65529369e-01 0.00000000e+00 2.23984750e-03 1.20880721e-01 0.00000000e+00] | 4 | | 0.5197 | 0.6268 | 0.2719 | 0.3442 | 0.8180 | [0. 0.62230678 0.81645513 0.18616589 0.66669478 0.30574734 nan 0.36681201 0.31128062 0. 0.76635363 0. 0. nan 0. 0.37874505 0. 0. 0.68193241 0. 0.48867838 0.25809644 0. nan 0. 0.25765818 0. 0. 0.81965205 0.71604385 0.9214592 0. 0.00636635 0.12957446 0. ] | [0. 0.89469845 0.88320521 0.45231002 0.72104833 0.3386303 nan 0.53522723 0.72026843 0. 0.93197124 0. 0. nan 0. 0.45525816 0. 0. 0.87276184 0. 0.60762821 0.29654901 0. nan 0. 0.32162193 0. 0. 0.90797988 0.89199119 0.96388697 0. 0.00646084 0.21171965 0. ] | 5 |

Framework versions

• Transformers 4.24.0
• TensorFlow 2.9.2
• Datasets 2.6.1
• Tokenizers 0.13.2