rtdetr-v2-r50-cppe5-finetune-2

This model is a fine-tuned version of PekingU/rtdetr_v2_r50vd on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 6.6493
• Map: 0.6094
• Mar 100: 0.7099
• Map 50: 0.9567
• Map 75: 0.719

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_FUSED 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: 40

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Mar 100	Map 50	Map 75
No log	1.0	142	34.3182	0.0607	0.6404	0.094	0.0739
No log	2.0	284	8.2558	0.4949	0.7404	0.8746	0.5321
No log	3.0	426	7.6811	0.596	0.7113	0.9488	0.7037
175.7115	4.0	568	6.7879	0.5943	0.7057	0.9781	0.7188
175.7115	5.0	710	6.6212	0.6087	0.7248	0.9704	0.7148
175.7115	6.0	852	6.6538	0.5936	0.7255	0.972	0.7068
175.7115	7.0	994	6.5767	0.6078	0.7298	0.9675	0.7592
12.1686	8.0	1136	6.6338	0.6246	0.7511	0.9646	0.7952
12.1686	9.0	1278	6.5246	0.598	0.7262	0.9483	0.7194
12.1686	10.0	1420	6.3852	0.6277	0.7369	0.959	0.8146
11.266	11.0	1562	6.6311	0.6092	0.7312	0.9659	0.7514
11.266	12.0	1704	6.4214	0.609	0.7333	0.9642	0.7471
11.266	13.0	1846	6.4222	0.6219	0.7319	0.9696	0.7763
11.266	14.0	1988	6.5066	0.617	0.7426	0.9636	0.7628
10.6436	15.0	2130	6.5157	0.6119	0.7248	0.9767	0.79
10.6436	16.0	2272	6.4737	0.6119	0.7348	0.9571	0.7829
10.6436	17.0	2414	6.5789	0.6326	0.7277	0.9674	0.7798
10.0889	18.0	2556	6.5001	0.6168	0.722	0.9755	0.7703
10.0889	19.0	2698	6.4702	0.614	0.7326	0.963	0.7387
10.0889	20.0	2840	6.5242	0.6151	0.7248	0.9673	0.7568
10.0889	21.0	2982	6.5797	0.6293	0.7298	0.9738	0.7369
9.4174	22.0	3124	6.5303	0.616	0.7355	0.9669	0.7879
9.4174	23.0	3266	6.7147	0.6079	0.7262	0.9668	0.756
9.4174	24.0	3408	6.6836	0.6146	0.7319	0.9725	0.7951
8.9177	25.0	3550	6.6182	0.6048	0.7298	0.9734	0.7162
8.9177	26.0	3692	6.5351	0.6182	0.7156	0.9713	0.7856
8.9177	27.0	3834	6.3831	0.6235	0.7262	0.9782	0.8026
8.9177	28.0	3976	6.4694	0.6194	0.7326	0.9756	0.7649
8.4449	29.0	4118	6.6127	0.6119	0.717	0.9672	0.7576
8.4449	30.0	4260	6.6870	0.6078	0.7312	0.9562	0.7143
8.4449	31.0	4402	6.7105	0.6084	0.7163	0.9659	0.7379
7.9999	32.0	4544	6.6299	0.6091	0.717	0.9607	0.7731
7.9999	33.0	4686	6.7453	0.6062	0.7262	0.9625	0.7409
7.9999	34.0	4828	6.6493	0.6121	0.7191	0.9643	0.7963
7.9999	35.0	4970	6.8267	0.6044	0.7213	0.9668	0.7582
7.5518	36.0	5112	6.8738	0.604	0.7156	0.9669	0.735
7.5518	37.0	5254	6.8444	0.6068	0.7113	0.967	0.7484
7.5518	38.0	5396	6.8525	0.605	0.7142	0.9667	0.7401
7.1858	39.0	5538	6.8547	0.609	0.7149	0.9683	0.7406
7.1858	40.0	5680	6.8616	0.6103	0.7156	0.9697	0.7328

Framework versions

• Transformers 4.57.1
• Pytorch 2.9.0+cu128
• Datasets 4.4.1
• Tokenizers 0.22.1