dfine-small-construction-ppe-v2

This model is a fine-tuned version of ustc-community/dfine-small-coco on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 4.9283
• Map: 0.1902
• Map 50: 0.2621
• Map 75: 0.2114
• Map Small: 0.0566
• Map Medium: 0.1483
• Map Large: 0.2527
• Mar 1: 0.3235
• Mar 10: 0.5249
• Mar 100: 0.6005
• Mar Small: 0.3347
• Mar Medium: 0.4868
• Mar Large: 0.79
• Map Person: 0.2845
• Mar 100 Person: 0.8521
• Map Hardhat: 0.0995
• Mar 100 Hardhat: 0.5868
• Map No-hardhat: 0.2834
• Mar 100 No-hardhat: 0.6962
• Map Safety-vest: 0.6181
• Mar 100 Safety-vest: 0.8239
• Map No-safety-vest: 0.0023
• Mar 100 No-safety-vest: 0.52
• Map No-mask: 0.0
• Mar 100 No-mask: 0.0
• Map No-goggles: 0.0
• Mar 100 No-goggles: 0.0
• Map Gloves: 0.0968
• Mar 100 Gloves: 0.5299
• Map Safety-boots: 0.1657
• Mar 100 Safety-boots: 0.4648
• Map Excavator: 0.1317
• Mar 100 Excavator: 0.8667
• Map Dump-truck: 0.2471
• Mar 100 Dump-truck: 0.875
• Map Wheel-loader: 0.353
• Mar 100 Wheel-loader: 0.9909

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
• num_epochs: 14.0

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 Person	Mar 100 Person	Map Hardhat	Mar 100 Hardhat	Map No-hardhat	Mar 100 No-hardhat	Map Safety-vest	Mar 100 Safety-vest	Map No-safety-vest	Mar 100 No-safety-vest	Map No-mask	Mar 100 No-mask	Map No-goggles	Mar 100 No-goggles	Map Gloves	Mar 100 Gloves	Map Safety-boots	Mar 100 Safety-boots	Map Excavator	Mar 100 Excavator	Map Dump-truck	Mar 100 Dump-truck	Map Wheel-loader	Mar 100 Wheel-loader
17.0151	1.0	1008	4.4310	0.1433	0.2143	0.1622	0.0415	0.127	0.1855	0.283	0.4625	0.5489	0.3247	0.3871	0.7256	0.3088	0.8136	0.1884	0.6525	0.1674	0.6678	0.5624	0.7781	0.0008	0.4533	0.0	0.0	0.0	0.0	0.0061	0.3866	0.0336	0.4171	0.172	0.9	0.0419	0.6	0.2377	0.9182
15.4555	2.0	2016	4.4957	0.1878	0.2621	0.2139	0.0387	0.1358	0.2533	0.2817	0.5036	0.5899	0.3544	0.4477	0.7705	0.3851	0.8313	0.1385	0.6485	0.2358	0.6886	0.6248	0.8306	0.0016	0.4933	0.0	0.0	0.0	0.0	0.0118	0.3496	0.0983	0.4918	0.3659	0.9	0.225	0.9	0.1667	0.9455
14.8455	3.0	3024	4.5601	0.1895	0.2709	0.2106	0.0532	0.1376	0.2457	0.249	0.5086	0.617	0.3641	0.4983	0.8009	0.3769	0.8365	0.1321	0.6246	0.3454	0.6927	0.6241	0.8311	0.0019	0.58	0.0	0.0	0.0	0.0	0.0327	0.4735	0.1181	0.535	0.3879	0.9333	0.0968	0.925	0.1584	0.9727
14.5490	4.0	4032	4.6064	0.1935	0.2727	0.2173	0.0487	0.1474	0.2584	0.2814	0.5205	0.6044	0.3674	0.4854	0.7823	0.2855	0.8437	0.1413	0.6412	0.341	0.688	0.5946	0.8274	0.0021	0.4933	0.0	0.0	0.0	0.0	0.0367	0.434	0.1365	0.552	0.2823	0.9	0.1934	0.9	0.3089	0.9727
14.2262	5.0	5040	4.6979	0.191	0.2712	0.2093	0.0461	0.1299	0.2648	0.2993	0.5118	0.5921	0.3818	0.4658	0.7828	0.3304	0.8391	0.1321	0.6304	0.3216	0.6946	0.5435	0.8294	0.0025	0.46	0.0	0.0	0.0	0.0	0.0493	0.4313	0.1333	0.4779	0.2126	0.8667	0.3043	0.9125	0.2629	0.9636
14.0609	6.0	6048	4.7726	0.1866	0.2723	0.2061	0.0495	0.1395	0.2614	0.3	0.5306	0.5981	0.3754	0.4778	0.789	0.2793	0.8405	0.1298	0.6152	0.3218	0.6991	0.5473	0.8217	0.0032	0.4667	0.0	0.0	0.0	0.0	0.054	0.4369	0.1457	0.5335	0.1708	0.9	0.3393	0.9	0.2484	0.9636
13.9573	7.0	7056	4.7826	0.19	0.2766	0.204	0.0509	0.1404	0.2513	0.2888	0.5102	0.5937	0.3568	0.468	0.7778	0.2796	0.8451	0.1112	0.6095	0.2894	0.6868	0.5987	0.8254	0.0024	0.4467	0.0	0.0	0.0	0.0	0.0648	0.4634	0.1631	0.4621	0.2119	0.9	0.2617	0.9125	0.2974	0.9727
13.7146	8.0	8064	4.8471	0.1699	0.2446	0.1917	0.0536	0.1341	0.2303	0.327	0.5249	0.6099	0.3898	0.4876	0.7902	0.2474	0.8447	0.104	0.6013	0.2988	0.6868	0.5685	0.8279	0.0026	0.5667	0.0	0.0	0.0	0.0	0.0901	0.5317	0.1662	0.4899	0.1829	0.9	0.1277	0.8875	0.2505	0.9818
13.6264	9.0	9072	4.8460	0.1784	0.2639	0.1969	0.0549	0.1432	0.2437	0.3225	0.5213	0.6047	0.3677	0.4753	0.8006	0.2797	0.8471	0.1034	0.5976	0.2797	0.6905	0.6032	0.8227	0.0023	0.5	0.0	0.0	0.0	0.0	0.1034	0.5172	0.1739	0.5028	0.167	0.9	0.2138	0.8875	0.2145	0.9909
13.5656	10.0	10080	4.8845	0.2108	0.2815	0.2343	0.0525	0.1397	0.2838	0.3211	0.5376	0.6054	0.3909	0.4762	0.7923	0.2657	0.8507	0.1088	0.5998	0.2642	0.6924	0.604	0.8257	0.0022	0.5467	0.0	0.0	0.0	0.0	0.1005	0.5086	0.1763	0.4971	0.2417	0.9	0.3452	0.8625	0.4207	0.9818
13.4596	11.0	11088	4.9015	0.1979	0.2684	0.2234	0.0563	0.1418	0.2664	0.3217	0.5206	0.6015	0.3513	0.4802	0.7903	0.3127	0.8533	0.089	0.5828	0.2916	0.6994	0.6213	0.825	0.002	0.4933	0.0	0.0	0.0	0.0	0.0878	0.5246	0.1678	0.4701	0.1794	0.9	0.3176	0.8875	0.3053	0.9818
13.4221	12.0	12096	4.9324	0.1699	0.2388	0.1874	0.0564	0.1512	0.2217	0.3191	0.5246	0.6009	0.3678	0.4797	0.794	0.2781	0.8585	0.1045	0.5868	0.2946	0.6924	0.6094	0.825	0.0025	0.4867	0.0	0.0	0.0	0.0	0.089	0.5295	0.183	0.4748	0.1326	0.9	0.1526	0.875	0.1931	0.9818
13.3594	13.0	13104	4.9303	0.1831	0.249	0.2055	0.0536	0.1543	0.2345	0.3216	0.5262	0.5995	0.3543	0.4827	0.7851	0.2928	0.8473	0.0965	0.5876	0.2911	0.6924	0.6235	0.8182	0.0023	0.4933	0.0	0.0	0.0	0.0	0.0926	0.5246	0.1681	0.4737	0.1923	0.9	0.1276	0.875	0.3106	0.9818
13.3525	14.0	14112	4.9283	0.1902	0.2621	0.2114	0.0566	0.1483	0.2527	0.3235	0.5249	0.6005	0.3347	0.4868	0.79	0.2845	0.8521	0.0995	0.5868	0.2834	0.6962	0.6181	0.8239	0.0023	0.52	0.0	0.0	0.0	0.0	0.0968	0.5299	0.1657	0.4648	0.1317	0.8667	0.2471	0.875	0.353	0.9909

Framework versions

• Transformers 5.12.1
• Pytorch 2.12.1+cu130
• Datasets 5.0.0
• Tokenizers 0.22.2