resnet_finetuned_raccoons

This model is a fine-tuned version of facebook/detr-resnet-50 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 1.1191
• Map: 0.1527
• Map 50: 0.3422
• Map 75: 0.1113
• Map Small: -1.0
• Map Medium: 0.6
• Map Large: 0.1575
• Mar 1: 0.3366
• Mar 10: 0.5098
• Mar 100: 0.7317
• Mar Small: -1.0
• Mar Medium: 0.6
• Mar Large: 0.735
• Map Raccoon: 0.1527
• Mar 100 Raccoon: 0.7317

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: 4
• 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: 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 Raccoon	Mar 100 Raccoon
No log	1.0	40	1.7414	0.0129	0.0372	0.0083	-1.0	0.0	0.0168	0.0512	0.2366	0.5951	-1.0	0.0	0.61	0.0129	0.5951
No log	2.0	80	1.5491	0.016	0.0441	0.0105	-1.0	0.0028	0.0195	0.0098	0.2488	0.6341	-1.0	0.3	0.6425	0.016	0.6341
No log	3.0	120	1.6715	0.0117	0.0336	0.0051	-1.0	0.016	0.0165	0.0146	0.0951	0.6	-1.0	0.3	0.6075	0.0117	0.6
No log	4.0	160	1.9540	0.0046	0.0116	0.0016	-1.0	0.001	0.0097	0.0	0.0	0.5073	-1.0	0.4	0.51	0.0046	0.5073
No log	5.0	200	2.2195	0.0179	0.0722	0.0028	-1.0	0.0011	0.0189	0.0268	0.2024	0.4293	-1.0	0.2	0.435	0.0179	0.4293
No log	6.0	240	1.4680	0.0835	0.2989	0.0352	-1.0	0.1048	0.0872	0.1683	0.3976	0.6244	-1.0	0.3	0.6325	0.0835	0.6244
No log	7.0	280	1.4213	0.107	0.2776	0.0733	-1.0	0.0036	0.1108	0.2805	0.4341	0.6293	-1.0	0.3	0.6375	0.107	0.6293
No log	8.0	320	1.3078	0.1017	0.2637	0.0697	-1.0	0.0344	0.1056	0.2976	0.461	0.6585	-1.0	0.4	0.665	0.1017	0.6585
No log	9.0	360	1.2742	0.0939	0.2041	0.0656	-1.0	0.0193	0.097	0.2878	0.4317	0.6976	-1.0	0.3	0.7075	0.0939	0.6976
No log	10.0	400	1.9146	0.0128	0.0465	0.0055	-1.0	0.0629	0.0129	0.0488	0.1707	0.5341	-1.0	0.4	0.5375	0.0128	0.5341
No log	11.0	440	1.5629	0.0836	0.1713	0.0803	-1.0	0.22	0.0884	0.2439	0.4293	0.6073	-1.0	0.7	0.605	0.0836	0.6073
No log	12.0	480	1.3116	0.147	0.3374	0.1108	-1.0	0.1458	0.1539	0.3049	0.4585	0.6732	-1.0	0.4	0.68	0.147	0.6732
1.7084	13.0	520	1.6814	0.029	0.1135	0.0107	-1.0	0.0021	0.0297	0.0439	0.1951	0.5732	-1.0	0.2	0.5825	0.029	0.5732
1.7084	14.0	560	1.6939	0.0599	0.2374	0.0117	-1.0	0.0	0.0627	0.1293	0.2341	0.561	-1.0	0.0	0.575	0.0599	0.561
1.7084	15.0	600	1.8753	0.0803	0.297	0.014	-1.0	0.1	0.0828	0.1317	0.2415	0.5244	-1.0	0.1	0.535	0.0803	0.5244
1.7084	16.0	640	1.3866	0.1491	0.4631	0.0769	-1.0	0.0071	0.1533	0.278	0.439	0.661	-1.0	0.1	0.675	0.1491	0.661
1.7084	17.0	680	1.3046	0.0917	0.2404	0.078	-1.0	0.1	0.0946	0.2317	0.4927	0.6805	-1.0	0.1	0.695	0.0917	0.6805
1.7084	18.0	720	1.2856	0.2078	0.4815	0.1623	-1.0	0.7	0.2124	0.3122	0.4951	0.6854	-1.0	0.7	0.685	0.2078	0.6854
1.7084	19.0	760	1.2144	0.1075	0.2812	0.0903	-1.0	0.3	0.1104	0.3098	0.4634	0.7	-1.0	0.3	0.71	0.1075	0.7
1.7084	20.0	800	1.2358	0.1278	0.3229	0.1007	-1.0	0.3	0.1312	0.3317	0.4976	0.6805	-1.0	0.3	0.69	0.1278	0.6805
1.7084	21.0	840	1.1859	0.112	0.302	0.0731	-1.0	0.4	0.1148	0.3561	0.5122	0.7171	-1.0	0.4	0.725	0.112	0.7171
1.7084	22.0	880	1.1855	0.1522	0.3443	0.1086	-1.0	0.5	0.1549	0.3146	0.5293	0.7049	-1.0	0.5	0.71	0.1522	0.7049
1.7084	23.0	920	1.1813	0.116	0.2722	0.0915	-1.0	0.5	0.1183	0.2976	0.4829	0.722	-1.0	0.5	0.7275	0.116	0.722
1.7084	24.0	960	1.1388	0.1541	0.3712	0.1012	-1.0	0.5	0.1557	0.3415	0.5146	0.739	-1.0	0.5	0.745	0.1541	0.739
1.4888	25.0	1000	1.1046	0.151	0.3619	0.0922	-1.0	0.5	0.1561	0.339	0.5146	0.739	-1.0	0.5	0.745	0.151	0.739
1.4888	26.0	1040	1.1356	0.1407	0.3067	0.1034	-1.0	0.6	0.1454	0.3439	0.5024	0.7366	-1.0	0.6	0.74	0.1407	0.7366
1.4888	27.0	1080	1.1432	0.1474	0.3263	0.1056	-1.0	0.6	0.152	0.3537	0.5098	0.7366	-1.0	0.6	0.74	0.1474	0.7366
1.4888	28.0	1120	1.1172	0.1529	0.3388	0.109	-1.0	0.6	0.1578	0.3415	0.5122	0.7341	-1.0	0.6	0.7375	0.1529	0.7341
1.4888	29.0	1160	1.1196	0.1528	0.3433	0.1106	-1.0	0.6	0.1575	0.3366	0.5098	0.7317	-1.0	0.6	0.735	0.1528	0.7317
1.4888	30.0	1200	1.1191	0.1527	0.3422	0.1113	-1.0	0.6	0.1575	0.3366	0.5098	0.7317	-1.0	0.6	0.735	0.1527	0.7317

Framework versions

• Transformers 4.57.6
• Pytorch 2.9.0+cu128
• Datasets 4.0.0
• Tokenizers 0.22.2