EfficientNetV2-S-FacesMTL-EXP1

thethinkmachine/EfficientNetV2-S-FacesMTL-EXP1

EfficientNetV2-S-FacesMTL-EXP1 is a machine learning model from thethinkmachine. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for transformers.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

36% of all-time downloads

All-time downloads

Public

Parameters

21.6M

86.7 MB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors86.7 MB · 100%

At a glance

Library: transformers
Model type: facesmultitasknet
Access: Public
Created: Nov 12, 2025
Updated: Nov 12, 2025
SHA: 438b5386

Model card

From the Hugging Face model README

EfficientNetV2-S-FacesMTL-EXP1

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

Gender Accuracy: 0.9219
Gender F1: 0.8960
Age Mae: 5.4856
Age Rmse: 7.7077
Loss: 59.6107

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: 0.0001
train_batch_size: 32
eval_batch_size: 32
seed: 42
optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: cosine
num_epochs: 5
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Gender Accuracy	Gender F1	Age Mae	Age Rmse	Validation Loss
141.0082	0.1728	150	0.7669	0.4982	9.8306	13.2003	174.7533
115.015	0.3456	300	0.8179	0.6641	7.2638	9.8329	97.0897
114.0637	0.5184	450	0.8727	0.8262	7.6025	10.2955	106.4104
108.9411	0.6912	600	0.8355	0.7141	6.4249	8.5998	74.3246
102.9268	0.8641	750	0.8862	0.8405	6.3497	8.6367	74.9009
74.0264	1.0369	900	0.8911	0.8478	6.0467	8.1970	67.4731
69.2823	1.2097	1050	0.8741	0.8048	7.7731	9.7514	95.3764
75.1102	1.3825	1200	0.8971	0.8558	6.1987	8.4372	71.4451
84.3635	1.5553	1350	0.8969	0.8597	6.1119	8.4568	71.7744
69.7893	1.7281	1500	0.9072	0.8736	6.0104	8.2477	68.2638
61.8971	1.9009	1650	0.9092	0.8763	6.1341	8.2888	68.9466
42.5042	2.0737	1800	0.8997	0.8730	5.7658	7.9370	63.2745
39.7624	2.2465	1950	0.9107	0.8726	5.9121	8.1481	66.6196
41.98	2.4194	2100	0.9136	0.8830	5.6534	7.8299	61.5340
48.5888	2.5922	2250	0.9069	0.8794	5.7673	7.9777	63.8842
49.4607	2.7650	2400	0.9136	0.8835	5.6717	7.7749	60.6700
52.7909	2.9378	2550	0.9182	0.8907	5.7793	7.9226	62.9845
29.6541	3.1106	2700	0.9182	0.8910	5.5971	7.6659	58.9833
29.6989	3.2834	2850	0.9242	0.8964	5.6225	7.7525	60.3134
34.2387	3.4562	3000	0.9225	0.8959	5.6239	7.7780	60.7093
29.4395	3.6290	3150	0.9205	0.8945	5.6094	7.7657	60.5210
30.939	3.8018	3300	0.9231	0.8977	5.5582	7.6809	59.2081
43.7756	3.9747	3450	0.9202	0.8945	5.5838	7.7157	59.7454
21.8149	4.1475	3600	0.9225	0.8951	5.5411	7.6636	58.9366
28.7175	4.3203	3750	0.9213	0.8965	5.5546	7.6726	59.0829
28.0368	4.4931	3900	0.9245	0.8990	5.5760	7.7432	60.1640
24.2052	4.6659	4050	0.9165	0.8918	5.5538	7.6908	59.3667
24.5022	4.8387	4200	0.9245	0.8992	5.5598	7.7017	59.5255

Framework versions

Transformers 4.57.1
Pytorch 2.9.0+cu130
Datasets 4.4.1
Tokenizers 0.22.1

Library: transformers
Type: facesmultitasknet
Created: Nov 12, 2025
Updated: Nov 12, 2025