Modelpublic

2z299/rtdetr-v2-r50-barcode-finetune-onnx

sourceHugging Faceapache-2.0updated 6mo agoView on Hugging Face

0likes12downloads

Model Card

rtdetr-v2-r50-barcode-finetune

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: 5.3834
—Map: 0.8239
—Map 50: 0.9525
—Map 75: 0.8848
—Map Small: 0.4756
—Map Medium: 0.8364
—Map Large: 0.8413
—Mar 1: 0.3932
—Mar 10: 0.846
—Mar 100: 0.851
—Mar Small: 0.5349
—Mar Medium: 0.855
—Mar Large: 0.8731
—Map Barcode: 0.8239
—Mar 100 Barcode: 0.851

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
—trainbatchsize: 64
—evalbatchsize: 64
—seed: 42
—optimizer: Use OptimizerNames.ADAMWTORCHFUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
—lrschedulertype: linear
—lrschedulerwarmup_steps: 300
—num_epochs: 40

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 Barcode	Mar 100 Barcode
No log	1.0	205	10.9885	0.746	0.9347	0.8269	0.3781	0.775	0.7527	0.3697	0.8157	0.857	0.4644	0.8519	0.8898	0.746	0.857
No log	2.0	410	6.1885	0.7955	0.9569	0.8687	0.4771	0.8186	0.8028	0.3831	0.8352	0.8604	0.573	0.865	0.8799	0.7955	0.8604
66.4195	3.0	615	5.7203	0.8076	0.9621	0.8682	0.5034	0.8253	0.8142	0.3875	0.8384	0.8569	0.6249	0.8656	0.8699	0.8076	0.8569
66.4195	4.0	820	5.6143	0.8091	0.9597	0.8761	0.5015	0.8268	0.8163	0.3879	0.8391	0.8521	0.6208	0.8639	0.8634	0.8091	0.8521
9.7122	5.0	1025	5.4216	0.8156	0.9627	0.8726	0.5155	0.8364	0.8216	0.3895	0.8437	0.8541	0.6221	0.8641	0.8665	0.8156	0.8541
9.7122	6.0	1230	5.3856	0.8224	0.9678	0.8886	0.5244	0.8391	0.8295	0.3914	0.846	0.8568	0.6315	0.8666	0.8688	0.8224	0.8568
9.7122	7.0	1435	5.2771	0.8256	0.9686	0.888	0.5351	0.84	0.8339	0.394	0.8497	0.8604	0.6336	0.8684	0.8734	0.8256	0.8604
9.2184	8.0	1640	5.2753	0.8227	0.9591	0.8805	0.512	0.8405	0.8328	0.3925	0.8476	0.8566	0.6045	0.8647	0.8715	0.8227	0.8566
9.2184	9.0	1845	5.2692	0.8241	0.9616	0.8802	0.5075	0.8394	0.834	0.3929	0.8498	0.8567	0.5817	0.8636	0.874	0.8241	0.8567
8.8513	10.0	2050	5.2316	0.8285	0.9602	0.8998	0.5166	0.8412	0.8428	0.394	0.8539	0.8622	0.6104	0.8665	0.8791	0.8285	0.8622
8.8513	11.0	2255	5.1873	0.8248	0.9618	0.8836	0.5216	0.8359	0.8369	0.3939	0.8461	0.8549	0.6156	0.8596	0.8706	0.8248	0.8549
8.8513	12.0	2460	5.1861	0.8302	0.9606	0.8985	0.5177	0.8413	0.8436	0.3953	0.8519	0.8591	0.5948	0.8624	0.8774	0.8302	0.8591
8.5944	13.0	2665	5.2404	0.8323	0.9604	0.8955	0.4807	0.8434	0.848	0.394	0.8575	0.8642	0.5578	0.8686	0.8853	0.8323	0.8642
8.5944	14.0	2870	5.4670	0.8303	0.9558	0.9008	0.5009	0.8404	0.8483	0.3929	0.8574	0.8637	0.5785	0.8632	0.8857	0.8303	0.8637
8.3642	15.0	3075	5.3559	0.8279	0.9553	0.8912	0.4711	0.8368	0.8469	0.3942	0.8529	0.8595	0.5332	0.8614	0.8834	0.8279	0.8595
8.3642	16.0	3280	5.4542	0.8265	0.9504	0.8917	0.4887	0.8373	0.845	0.393	0.8496	0.8558	0.5595	0.8572	0.8778	0.8265	0.8558
8.3642	17.0	3485	5.6300	0.8244	0.9457	0.8858	0.4966	0.8352	0.8411	0.3899	0.8515	0.8585	0.5709	0.8576	0.881	0.8244	0.8585
8.1607	18.0	3690	5.4196	0.8201	0.9417	0.8776	0.4612	0.8286	0.8407	0.3933	0.8481	0.8547	0.5242	0.8538	0.8805	0.8201	0.8547
8.1607	19.0	3895	5.3907	0.8237	0.945	0.8803	0.4993	0.8332	0.842	0.3933	0.8496	0.8561	0.574	0.8559	0.8777	0.8237	0.8561
7.9929	20.0	4100	5.5463	0.8129	0.9341	0.8683	0.4301	0.8297	0.8382	0.384	0.8482	0.8547	0.4945	0.8542	0.8824	0.8129	0.8547
7.9929	21.0	4305	5.4527	0.8186	0.944	0.875	0.4395	0.83	0.8423	0.3884	0.8465	0.8511	0.492	0.8513	0.8784	0.8186	0.8511
7.8294	22.0	4510	5.4445	0.8228	0.9477	0.8786	0.4827	0.8306	0.8399	0.3906	0.8446	0.8506	0.5474	0.8494	0.8744	0.8228	0.8506
7.8294	23.0	4715	5.5783	0.8111	0.9317	0.8661	0.4446	0.829	0.8324	0.3851	0.8423	0.8479	0.4986	0.85	0.8733	0.8111	0.8479
7.8294	24.0	4920	5.7928	0.8074	0.925	0.8647	0.4331	0.8214	0.8295	0.3849	0.8372	0.843	0.4903	0.8423	0.8703	0.8074	0.843
7.6379	25.0	5125	5.3096	0.8247	0.9489	0.8783	0.4882	0.8328	0.8421	0.3927	0.8468	0.8527	0.5574	0.8526	0.8754	0.8247	0.8527
7.6379	26.0	5330	5.4555	0.8222	0.9443	0.8802	0.4835	0.8351	0.8394	0.3915	0.8456	0.8515	0.5394	0.8544	0.8738	0.8222	0.8515
7.5250	27.0	5535	5.5939	0.8234	0.9428	0.879	0.481	0.837	0.8412	0.3899	0.8499	0.8556	0.5349	0.8579	0.8788	0.8234	0.8556
7.5250	28.0	5740	5.6699	0.8168	0.9363	0.8751	0.4703	0.8326	0.8333	0.3874	0.8451	0.8515	0.5318	0.8532	0.875	0.8168	0.8515
7.5250	29.0	5945	5.4877	0.8214	0.9462	0.8729	0.4708	0.833	0.8406	0.3914	0.845	0.8504	0.5221	0.8527	0.8742	0.8214	0.8504
7.3749	30.0	6150	5.4449	0.8215	0.9468	0.8732	0.4757	0.8333	0.8395	0.3903	0.845	0.8509	0.5419	0.8516	0.8741	0.8215	0.8509
7.3749	31.0	6355	5.5687	0.8173	0.9339	0.8721	0.4835	0.8328	0.8338	0.3884	0.8429	0.8486	0.5304	0.8527	0.8707	0.8173	0.8486
7.2567	32.0	6560	5.3437	0.8221	0.9474	0.8741	0.4867	0.8329	0.8391	0.392	0.8445	0.8505	0.5471	0.8517	0.873	0.8221	0.8505
7.2567	33.0	6765	5.6103	0.817	0.9323	0.8713	0.4811	0.8339	0.8333	0.3872	0.8446	0.8499	0.5336	0.8533	0.8722	0.817	0.8499
7.2567	34.0	6970	5.4900	0.8196	0.9382	0.8732	0.4601	0.8299	0.8383	0.3899	0.8418	0.8465	0.5087	0.8495	0.8707	0.8196	0.8465
7.0955	35.0	7175	5.4899	0.8198	0.9456	0.8732	0.486	0.8328	0.8356	0.389	0.8426	0.848	0.536	0.8523	0.8696	0.8198	0.848
7.0955	36.0	7380	5.5159	0.8177	0.9369	0.8725	0.4803	0.8299	0.8345	0.3881	0.8408	0.8463	0.5284	0.8508	0.8682	0.8177	0.8463
6.9915	37.0	7585	5.5678	0.8182	0.9363	0.8723	0.468	0.8319	0.8351	0.3882	0.842	0.8474	0.5239	0.8522	0.8695	0.8182	0.8474
6.9915	38.0	7790	5.5920	0.8181	0.9355	0.8726	0.4785	0.8308	0.8347	0.3858	0.8417	0.8469	0.527	0.8509	0.8692	0.8181	0.8469
6.9915	39.0	7995	5.5446	0.8188	0.9374	0.8736	0.4767	0.8324	0.8369	0.388	0.8416	0.847	0.5322	0.8515	0.8687	0.8188	0.847
6.9176	40.0	8200	5.5335	0.8189	0.9382	0.873	0.4779	0.8324	0.8367	0.3883	0.8415	0.8467	0.528	0.8513	0.8686	0.8189	0.8467

Framework versions

—Transformers 5.4.0
—Pytorch 2.10.0+cu128
—Datasets 4.8.4
—Tokenizers 0.22.2