detr_finetuned_crack

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

• Loss: 0.5449
• Map: 0.6418
• Map 50: 0.7915
• Map 75: 0.7168
• Map Small: -1.0
• Map Medium: 0.2381
• Map Large: 0.6629
• Mar 1: 0.6742
• Mar 10: 0.7806
• Mar 100: 0.8234
• Mar Small: -1.0
• Mar Medium: 0.4366
• Mar Large: 0.8433
• Map Excavators: 0.5862
• Mar 100 Excavators: 0.7611
• Map Dump truck: 0.5746
• Mar 100 Dump truck: 0.8057
• Map Wheel loader: 0.7646
• Mar 100 Wheel loader: 0.9034

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 adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 30

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 Excavators	Mar 100 Excavators	Map Dump truck	Mar 100 Dump truck	Map Wheel loader	Mar 100 Wheel loader
No log	1.0	281	1.0910	0.0675	0.0965	0.0697	-1.0	0.0088	0.0687	0.2623	0.3621	0.443	-1.0	0.1473	0.4477	0.0031	0.725	0.0093	0.0316	0.1901	0.5725
1.4507	2.0	562	1.1622	0.0534	0.0786	0.0578	-1.0	0.0075	0.0562	0.2759	0.4458	0.5215	-1.0	0.171	0.5358	0.0202	0.6833	0.0217	0.1863	0.1183	0.6949
1.4507	3.0	843	1.0275	0.0951	0.1463	0.1031	-1.0	0.0221	0.0987	0.3498	0.5405	0.5917	-1.0	0.2462	0.6085	0.0165	0.6694	0.0553	0.4618	0.2137	0.6438
1.161	4.0	1124	0.9637	0.121	0.1918	0.1295	-1.0	0.0722	0.1262	0.346	0.5476	0.6155	-1.0	0.2484	0.6319	0.0348	0.6694	0.1051	0.5472	0.2232	0.6298
1.161	5.0	1405	0.9838	0.194	0.281	0.2154	-1.0	0.0928	0.2002	0.357	0.5473	0.6138	-1.0	0.2452	0.6349	0.0732	0.5694	0.1145	0.6344	0.3943	0.6376
1.0777	6.0	1686	0.8650	0.1749	0.2471	0.1864	-1.0	0.045	0.1809	0.3767	0.5929	0.6321	-1.0	0.3409	0.6453	0.0471	0.7167	0.1307	0.5943	0.3471	0.5854
1.0777	7.0	1967	0.8570	0.2141	0.3026	0.2334	-1.0	0.0597	0.2216	0.427	0.6198	0.6745	-1.0	0.2097	0.6969	0.0939	0.7333	0.1681	0.6047	0.3804	0.6854
0.9398	8.0	2248	0.8859	0.235	0.3369	0.263	-1.0	0.051	0.2437	0.4363	0.6268	0.6766	-1.0	0.2032	0.702	0.0781	0.7444	0.195	0.6392	0.432	0.6461
0.9071	9.0	2529	0.8624	0.2717	0.3767	0.2971	-1.0	0.0777	0.2799	0.4334	0.6353	0.686	-1.0	0.2817	0.7097	0.1124	0.7111	0.1975	0.6807	0.5051	0.6663
0.9071	10.0	2810	0.8125	0.2909	0.4088	0.319	-1.0	0.0819	0.3025	0.4565	0.6765	0.7364	-1.0	0.2871	0.7598	0.1564	0.7444	0.256	0.6788	0.4602	0.786
0.8684	11.0	3091	0.8008	0.316	0.4487	0.3468	-1.0	0.0745	0.3293	0.514	0.6865	0.7411	-1.0	0.2978	0.7657	0.1418	0.75	0.3531	0.7344	0.4532	0.7388
0.8684	12.0	3372	0.7206	0.4342	0.5671	0.497	-1.0	0.079	0.4514	0.5547	0.7164	0.7681	-1.0	0.3301	0.791	0.3193	0.7722	0.3754	0.709	0.608	0.823
0.78	13.0	3653	0.7023	0.4618	0.6017	0.5105	-1.0	0.1418	0.476	0.5748	0.7181	0.7677	-1.0	0.3226	0.7888	0.3582	0.7194	0.4087	0.7448	0.6184	0.8388
0.78	14.0	3934	0.7258	0.4705	0.6167	0.5385	-1.0	0.1025	0.4905	0.56	0.7222	0.7683	-1.0	0.2957	0.7934	0.3718	0.7667	0.4072	0.7292	0.6325	0.809
0.7291	15.0	4215	0.6475	0.5445	0.6874	0.6045	-1.0	0.1298	0.5631	0.6157	0.7554	0.7923	-1.0	0.3946	0.8105	0.4822	0.7944	0.4859	0.7297	0.6653	0.8528
0.7291	16.0	4496	0.6435	0.5295	0.6864	0.5916	-1.0	0.2079	0.5513	0.6046	0.7587	0.8085	-1.0	0.3796	0.8304	0.4483	0.825	0.4942	0.7448	0.646	0.8556
0.6889	17.0	4777	0.6242	0.5472	0.7106	0.6101	-1.0	0.2103	0.5725	0.5937	0.7494	0.7918	-1.0	0.3753	0.8117	0.4309	0.7139	0.5229	0.7981	0.6879	0.8635
0.6533	18.0	5058	0.6060	0.5599	0.709	0.6197	-1.0	0.2911	0.5777	0.6189	0.7596	0.812	-1.0	0.386	0.8342	0.456	0.7806	0.498	0.7745	0.7256	0.8809
0.6533	19.0	5339	0.5994	0.5746	0.7315	0.6344	-1.0	0.2149	0.5975	0.6339	0.7738	0.8233	-1.0	0.343	0.8486	0.528	0.8167	0.5155	0.7764	0.6802	0.877
0.6013	20.0	5620	0.5833	0.5954	0.7497	0.6819	-1.0	0.1718	0.6159	0.6571	0.7775	0.822	-1.0	0.3968	0.8431	0.5589	0.8056	0.4923	0.7835	0.7349	0.877
0.6013	21.0	5901	0.5769	0.5955	0.7525	0.655	-1.0	0.2243	0.6168	0.6467	0.7715	0.8119	-1.0	0.3237	0.836	0.5313	0.7583	0.5047	0.7925	0.7504	0.8848
0.5834	22.0	6182	0.5608	0.6182	0.7707	0.682	-1.0	0.1805	0.6392	0.6563	0.7688	0.82	-1.0	0.3559	0.8435	0.5668	0.7778	0.5376	0.7901	0.7503	0.8921
0.5834	23.0	6463	0.5485	0.6216	0.7747	0.6953	-1.0	0.2656	0.642	0.648	0.7763	0.82	-1.0	0.3634	0.8426	0.5768	0.7722	0.5549	0.7991	0.7331	0.8888
0.5381	24.0	6744	0.5530	0.6284	0.7704	0.6923	-1.0	0.2638	0.6489	0.6524	0.7751	0.8182	-1.0	0.4215	0.8396	0.5869	0.7917	0.5602	0.783	0.7381	0.8798
0.5242	25.0	7025	0.5509	0.6402	0.7853	0.7114	-1.0	0.2367	0.6614	0.6736	0.7753	0.8223	-1.0	0.4075	0.8424	0.5875	0.7639	0.5695	0.8042	0.7637	0.8989
0.5242	26.0	7306	0.5426	0.6385	0.7838	0.7124	-1.0	0.2385	0.6599	0.6664	0.7851	0.826	-1.0	0.4785	0.8442	0.5844	0.7639	0.5762	0.8179	0.755	0.8961
0.5163	27.0	7587	0.5446	0.6437	0.7904	0.7181	-1.0	0.2359	0.6652	0.6763	0.7821	0.8275	-1.0	0.4699	0.8469	0.5963	0.7694	0.5739	0.8108	0.761	0.9022
0.5163	28.0	7868	0.5456	0.6351	0.7869	0.7129	-1.0	0.2361	0.6564	0.6707	0.7774	0.8238	-1.0	0.4301	0.8448	0.573	0.7611	0.5677	0.8052	0.7646	0.9051
0.5082	29.0	8149	0.5452	0.6413	0.7913	0.7165	-1.0	0.2387	0.6623	0.675	0.7795	0.8241	-1.0	0.4419	0.8433	0.5846	0.7611	0.5759	0.8085	0.7634	0.9028
0.5082	30.0	8430	0.5449	0.6418	0.7915	0.7168	-1.0	0.2381	0.6629	0.6742	0.7806	0.8234	-1.0	0.4366	0.8433	0.5862	0.7611	0.5746	0.8057	0.7646	0.9034

Framework versions

• Transformers 4.55.4
• Pytorch 2.8.0+cu128
• Datasets 3.6.0
• Tokenizers 0.21.4