detr_finetuned_trashify_box_detector

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: 1.1302

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: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_ratio: 0.05
• num_epochs: 25
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
101.8783	1.0	50	7.5132
4.1455	2.0	100	3.0556
2.5964	3.0	150	2.2737
2.2773	4.0	200	2.0691
2.0818	5.0	250	1.8494
1.9253	6.0	300	1.6872
1.7802	7.0	350	1.6033
1.675	8.0	400	1.4511
1.5263	9.0	450	1.4097
1.4322	10.0	500	1.3397
1.386	11.0	550	1.2897
1.3098	12.0	600	1.2813
1.248	13.0	650	1.2096
1.209	14.0	700	1.2200
1.1757	15.0	750	1.1987
1.144	16.0	800	1.1757
1.0732	17.0	850	1.1935
1.0501	18.0	900	1.1531
0.9864	19.0	950	1.1576
0.9941	20.0	1000	1.1513
0.9589	21.0	1050	1.1450
0.9279	22.0	1100	1.1355
0.9071	23.0	1150	1.1233
0.8851	24.0	1200	1.1338
0.8709	25.0	1250	1.1302

Framework versions

• Transformers 4.45.0.dev0
• Pytorch 2.4.0+cu124
• Datasets 2.21.0
• Tokenizers 0.19.1