rtdetr-v2-r34-cppe5-finetune-2

This model is a fine-tuned version of PekingU/rtdetr_v2_r34vd on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 8.3132
Map: 0.3055
Map 50: 0.5372
Map 75: 0.3008
Map Small: 0.1094
Map Medium: 0.241
Map Large: 0.3888
Mar 1: 0.2863
Mar 10: 0.4999
Mar 100: 0.578
Mar Small: 0.3884
Mar Medium: 0.4832
Mar Large: 0.7099
Map Coverall: 0.5596
Mar 100 Coverall: 0.7275
Map Face Shield: 0.2013
Mar 100 Face Shield: 0.6342
Map Gloves: 0.2742
Mar 100 Gloves: 0.5192
Map Goggles: 0.146
Mar 100 Goggles: 0.4769
Map Mask: 0.3462
Mar 100 Mask: 0.5324

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
train_batch_size: 8
eval_batch_size: 8
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: linear
lr_scheduler_warmup_steps: 300
num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Small	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Small	Mar Medium	Mar Large	Map Coverall	Mar 100 Coverall	Map Face Shield	Mar 100 Face Shield	Map Gloves	Mar 100 Gloves	Map Goggles	Mar 100 Goggles	Map Mask	Mar 100 Mask
No log	1.0	107	21.6622	0.0574	0.1029	0.0519	0.0001	0.0074	0.0702	0.0808	0.1764	0.2386	0.0255	0.1314	0.3738	0.263	0.6131	0.0098	0.1962	0.0056	0.1446	0.0002	0.0692	0.0085	0.1698
No log	2.0	214	12.3093	0.1473	0.2769	0.1347	0.04	0.108	0.2006	0.1941	0.3837	0.4464	0.1571	0.3411	0.6232	0.3946	0.6973	0.0871	0.438	0.067	0.392	0.0198	0.3308	0.1682	0.3742
No log	3.0	321	9.8181	0.2151	0.3903	0.2003	0.0862	0.1856	0.2692	0.2513	0.4746	0.5437	0.2899	0.4531	0.6932	0.4504	0.7221	0.1277	0.5734	0.1439	0.471	0.0493	0.4523	0.304	0.4996
No log	4.0	428	9.0262	0.2471	0.44	0.2372	0.0808	0.2084	0.3213	0.2685	0.4722	0.5494	0.2294	0.4611	0.7006	0.5037	0.7387	0.147	0.5823	0.2081	0.4938	0.0661	0.4338	0.3107	0.4982
27.2158	5.0	535	8.6126	0.276	0.4857	0.2769	0.104	0.2124	0.3508	0.2741	0.4908	0.5667	0.3254	0.4704	0.6999	0.5335	0.7261	0.156	0.6139	0.2473	0.5156	0.0958	0.4523	0.3474	0.5253
27.2158	6.0	642	8.4669	0.2826	0.5022	0.2795	0.095	0.2191	0.3575	0.2766	0.4948	0.5704	0.3249	0.461	0.7107	0.5455	0.7356	0.1524	0.6215	0.2571	0.5125	0.1208	0.4646	0.3372	0.5178
27.2158	7.0	749	8.3188	0.3003	0.5202	0.2958	0.1107	0.2348	0.3887	0.2879	0.5044	0.5766	0.3707	0.4823	0.7142	0.5545	0.7288	0.2019	0.6329	0.2655	0.5263	0.1341	0.4631	0.3453	0.532
27.2158	8.0	856	8.3084	0.2972	0.5265	0.2912	0.107	0.2409	0.3732	0.2811	0.5029	0.5811	0.3453	0.4772	0.7138	0.5617	0.7351	0.1649	0.6367	0.2773	0.5192	0.1337	0.4769	0.3485	0.5378
27.2158	9.0	963	8.2764	0.3064	0.5313	0.3049	0.1068	0.2421	0.3871	0.284	0.5073	0.5802	0.373	0.4783	0.7108	0.5621	0.7333	0.182	0.6291	0.2765	0.5299	0.1631	0.4738	0.3483	0.5347
11.8005	10.0	1070	8.3132	0.3055	0.5372	0.3008	0.1094	0.241	0.3888	0.2863	0.4999	0.578	0.3884	0.4832	0.7099	0.5596	0.7275	0.2013	0.6342	0.2742	0.5192	0.146	0.4769	0.3462	0.5324

Framework versions

Transformers 4.51.3
Pytorch 2.6.0+cu124
Datasets 3.5.1
Tokenizers 0.21.1

svetadomoi/rtdetr-v2-r34-cppe5-finetune-2

rtdetr-v2-r34-cppe5-finetune-2

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Seed this model