CoolFace
Modelpublic

li1212/conditional-detr-resnet-50_finetuned_cppe5

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face
0likes105downloads
Model Card

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

conditional-detr-resnet-50finetunedcppe5

This model is a fine-tuned version of microsoft/conditional-detr-resnet-50 on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 1.0625
  • Map: 0.2889
  • Map 50: 0.5213
  • Map 75: 0.2755
  • Map Small: 0.1022
  • Map Medium: 0.229
  • Map Large: 0.4272
  • Mar 1: 0.3222
  • Mar 10: 0.4831
  • Mar 100: 0.5016
  • Mar Small: 0.2614
  • Mar Medium: 0.4179
  • Mar Large: 0.6587
  • Map Coverall: 0.5718
  • Mar 100 Coverall: 0.6954
  • Map Face Shield: 0.2333
  • Mar 100 Face Shield: 0.4986
  • Map Gloves: 0.2051
  • Mar 100 Gloves: 0.4271
  • Map Goggles: 0.1597
  • Mar 100 Goggles: 0.4629
  • Map Mask: 0.2745
  • Mar 100 Mask: 0.4239

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
  • trainbatchsize: 8
  • evalbatchsize: 8
  • seed: 42
  • optimizer: Use OptimizerNames.ADAMWTORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizerargs=No additional optimizer arguments
  • lrschedulertype: cosine
  • num_epochs: 30

Training results

Training LossEpochStepValidation LossMapMap 50Map 75Map SmallMap MediumMap LargeMar 1Mar 10Mar 100Mar SmallMar MediumMar LargeMap CoverallMar 100 CoverallMap Face ShieldMar 100 Face ShieldMap GlovesMar 100 GlovesMap GogglesMar 100 GogglesMap MaskMar 100 Mask
1.91471.01061.81940.0180.03880.01430.00210.01240.02390.03270.09140.13690.05110.11580.17840.0660.27230.00160.01410.0020.16330.00.02420.02040.2105
1.77792.02121.67950.02580.05660.02130.00280.02690.0450.05810.13980.20030.07360.14510.2760.09480.46720.00180.01830.00340.21560.00.01610.02880.2842
1.53843.03181.51980.05640.11020.04970.0110.04610.07280.10320.21120.26110.09540.190.35630.19850.56720.01950.08030.0060.27040.00260.04520.05540.3426
1.43964.04241.46930.0680.12680.06050.00830.05590.08440.10450.23280.29140.09590.21470.38650.22210.63080.02430.11690.00620.25030.00220.08870.08540.3703
1.3465.05301.38650.10920.2110.08860.02990.09470.13060.1470.30560.34980.16880.27210.46760.37440.63850.02430.26760.01410.2990.0170.17740.11610.3665
1.49456.06361.37850.11920.23910.10040.02840.10220.140.14680.31660.35470.16250.29610.4680.40670.61490.03580.2690.03760.33470.01480.20320.10110.3517
1.33747.07421.34940.12270.24460.11040.02240.09710.15040.17710.34920.38490.19140.30820.51580.43170.6190.03430.35350.03290.33970.0090.24030.10560.3718
1.21468.08481.29110.14380.27890.12790.070.10740.20040.18870.37930.40990.20980.32610.56310.44330.61280.04260.3690.0730.39350.03020.29030.12980.3837
1.29489.09541.29160.1420.27610.11660.05210.11470.18010.19820.37790.41070.19760.31970.56160.42290.61030.05090.37180.06720.39350.03680.28710.13210.3909
1.207810.010601.27530.15070.31440.12350.03280.11910.22690.20930.38510.41320.17230.3320.57460.43890.62560.07350.3690.06620.37140.02710.31770.14770.3823
1.258711.011661.25080.17340.35240.14420.05150.14260.24460.21130.39370.42570.19540.34090.58440.4630.64360.06270.38590.10960.37990.05530.33870.17620.3804
1.169812.012721.17860.20050.38770.16830.10160.15680.28340.2380.4240.45210.23520.37160.59950.5060.66310.11290.46760.12150.40950.05750.30810.20480.412
1.206713.013781.20970.19840.3870.16820.06140.14820.29110.2290.41810.44540.2170.35190.61790.50250.66970.1030.43660.11960.38990.06180.35650.2050.3742
1.091214.014841.15910.22160.4190.20060.08190.18570.30550.26250.44020.47140.24370.39020.6150.50250.67590.15050.4690.130.40550.0830.38870.24220.4177
1.015415.015901.14740.23060.43270.20990.10350.18470.3330.27060.44340.46830.22410.39220.61410.52120.66720.1590.47180.14420.39350.08760.39350.24110.4153
1.077216.016961.13450.23880.44980.21750.10530.19060.35050.28110.44830.47910.22890.39740.63130.52510.66920.16430.48450.14850.41310.10730.41940.24890.4096
0.918517.018021.12530.24090.44710.21730.11590.19910.34760.2790.45380.47830.2530.40550.62260.52420.65590.14690.46060.16760.42110.11370.4290.25230.4249
1.073618.019081.11870.25070.46560.23780.10030.21440.35540.28780.45920.48580.22650.4260.61520.53530.67850.16350.46060.16990.43870.14120.43390.24370.4172
1.002119.020141.09870.26080.47360.25220.12620.21720.37340.29810.45260.48090.2370.40490.63210.53960.67130.15290.48030.18880.42010.16160.4210.26110.412
1.061320.021201.11440.25630.47850.22760.09170.21020.36910.28950.44720.47320.24930.39230.61890.54210.66210.16230.45210.18720.43170.13830.41450.25150.4057
1.057621.022261.08880.28270.50590.2710.10410.23590.40610.31060.46580.49220.26390.40760.64320.55160.68310.21450.50560.18970.42760.18960.42740.26780.4172
0.894922.023321.08930.27620.50780.24890.12070.22410.40140.30380.46620.4890.24130.40260.65190.55830.68560.19080.47890.20280.42260.16520.44030.26350.4177
1.085323.024381.07470.2750.5010.26020.11440.22670.39660.30490.47220.49580.26320.4120.6460.55690.69080.17430.4930.2020.42810.17710.45810.26490.4091
1.009824.025441.06230.280.50810.2630.11360.21990.41010.3110.4790.50110.26290.41750.6530.57050.69490.20570.50140.20230.42560.15510.46130.26650.4225
0.846625.026501.06820.28620.5140.27480.1050.23060.42170.31730.48130.50090.26690.41950.65650.56910.69440.2210.49720.20880.42410.16130.46450.2710.4244
0.906726.027561.06380.28810.52250.2760.10850.23090.42190.31830.48020.49910.24750.42050.65390.57230.69380.2320.49440.20510.42460.1580.45970.27340.423
0.846227.028621.06290.2880.51820.26790.10130.22520.42720.31910.48070.50040.2610.41760.65460.570.69330.23370.49860.20570.42860.1580.45650.27280.4249
0.877128.029681.06420.29010.5240.27810.10280.22990.42670.32150.48210.50090.26030.41920.65550.57250.69490.23460.49720.20550.42510.16340.46290.27440.4244
0.802229.030741.06260.28850.52120.27580.10170.2290.4270.3220.48290.50130.25890.41830.65880.57180.69590.23090.49720.2060.42760.15950.46130.27440.4244
0.824530.031801.06250.28890.52130.27550.10220.2290.42720.32220.48310.50160.26140.41790.65870.57180.69540.23330.49860.20510.42710.15970.46290.27450.4239

Framework versions

  • Transformers 4.52.4
  • Pytorch 2.6.0+cu124
  • Datasets 3.6.0
  • Tokenizers 0.21.2