detr-resnet-50_find_tuned_beyond_words

This model is a fine-tuned version of facebook/detr-resnet-50 on the beyond_words_23 dataset. It achieves the following results on the evaluation set:

• Loss: 0.9310

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: 16
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 200
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
2.7439	0.56	100	2.2690
1.7644	1.12	200	1.5053
1.557	1.69	300	1.3136
1.3207	2.25	400	1.2063
1.3705	2.81	500	1.2007
1.1924	3.37	600	1.2704
1.2604	3.93	700	1.1784
1.1982	4.49	800	1.1167
1.1912	5.06	900	1.1562
1.1206	5.62	1000	1.2124
1.1344	6.18	1100	1.0622
1.1388	6.74	1200	1.0425
1.0124	7.3	1300	0.9908
1.0776	7.87	1400	1.1182
0.9614	8.43	1500	0.9967
1.0136	8.99	1600	0.8933
1.0206	9.55	1700	0.9354
0.9529	10.11	1800	0.9751
1.0126	10.67	1900	0.9310

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.0
• Tokenizers 0.13.2