detr_finetuned_airdataset

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

• Loss: 0.8959
• Map: 0.3195
• Map 50: 0.7784
• Map 75: 0.1925
• Map Small: 0.3211
• Map Medium: 0.0079
• Map Large: -1.0
• Mar 1: 0.0256
• Mar 10: 0.1995
• Mar 100: 0.487
• Mar Small: 0.4896
• Mar Medium: 0.0061
• Mar Large: -1.0
• Map Falciparum Trophozoite: 0.3195
• Mar 100 Falciparum Trophozoite: 0.487
• Map Wbc: -1.0
• Mar 100 Wbc: -1.0

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
• 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 Falciparum Trophozoite	Mar 100 Falciparum Trophozoite	Map Wbc	Mar 100 Wbc
No log	1.0	209	1.2206	0.1424	0.401	0.07	0.1429	0.0328	-1.0	0.0168	0.1239	0.4185	0.4204	0.0612	-1.0	0.1424	0.4185	-1.0	-1.0
No log	2.0	418	1.1354	0.2136	0.585	0.1077	0.2145	0.0224	-1.0	0.0212	0.1608	0.4102	0.4123	0.0224	-1.0	0.2136	0.4102	-1.0	-1.0
1.3747	3.0	627	1.0729	0.2353	0.6428	0.1092	0.2365	0.0229	-1.0	0.0216	0.1669	0.4247	0.4268	0.0245	-1.0	0.2353	0.4247	-1.0	-1.0
1.3747	4.0	836	1.0260	0.2548	0.6701	0.1339	0.2563	0.0178	-1.0	0.0234	0.1792	0.4424	0.4447	0.0163	-1.0	0.2548	0.4424	-1.0	-1.0
1.0467	5.0	1045	1.0116	0.2576	0.6811	0.1321	0.2589	0.0208	-1.0	0.0229	0.1773	0.4422	0.4445	0.0184	-1.0	0.2576	0.4422	-1.0	-1.0
1.0467	6.0	1254	1.0150	0.2526	0.6842	0.1191	0.2537	0.0089	-1.0	0.0226	0.1724	0.4463	0.4486	0.0082	-1.0	0.2526	0.4463	-1.0	-1.0
1.0467	7.0	1463	0.9933	0.2627	0.699	0.1376	0.2639	0.0211	-1.0	0.0215	0.1773	0.4458	0.4481	0.0224	-1.0	0.2627	0.4458	-1.0	-1.0
0.9905	8.0	1672	0.9642	0.2797	0.7188	0.1511	0.2809	0.0112	-1.0	0.0241	0.1858	0.459	0.4614	0.0143	-1.0	0.2797	0.459	-1.0	-1.0
0.9905	9.0	1881	0.9641	0.2786	0.7209	0.1453	0.2803	0.0103	-1.0	0.0231	0.1861	0.4534	0.4558	0.0102	-1.0	0.2786	0.4534	-1.0	-1.0
0.955	10.0	2090	0.9869	0.2685	0.7158	0.1366	0.27	0.0023	-1.0	0.0225	0.1789	0.4442	0.4465	0.0041	-1.0	0.2685	0.4442	-1.0	-1.0
0.955	11.0	2299	0.9612	0.2837	0.7238	0.1534	0.2856	0.0067	-1.0	0.0242	0.1878	0.4568	0.4592	0.0082	-1.0	0.2837	0.4568	-1.0	-1.0
0.9248	12.0	2508	0.9437	0.2938	0.7368	0.1635	0.2954	0.005	-1.0	0.0239	0.1882	0.4701	0.4727	0.0041	-1.0	0.2938	0.4701	-1.0	-1.0
0.9248	13.0	2717	0.9390	0.289	0.7371	0.16	0.2903	0.0149	-1.0	0.0254	0.191	0.4685	0.471	0.0122	-1.0	0.289	0.4685	-1.0	-1.0
0.9248	14.0	2926	0.9321	0.2986	0.7428	0.1744	0.3002	0.005	-1.0	0.0251	0.1928	0.4743	0.4768	0.0041	-1.0	0.2986	0.4743	-1.0	-1.0
0.9027	15.0	3135	0.9448	0.2911	0.7418	0.1588	0.2924	0.0139	-1.0	0.0241	0.1877	0.4678	0.4702	0.0122	-1.0	0.2911	0.4678	-1.0	-1.0
0.9027	16.0	3344	0.9259	0.3033	0.7549	0.174	0.3047	0.005	-1.0	0.0249	0.1931	0.4736	0.4762	0.0041	-1.0	0.3033	0.4736	-1.0	-1.0
0.8725	17.0	3553	0.9200	0.3039	0.7554	0.1795	0.3055	0.0069	-1.0	0.0259	0.1949	0.4764	0.479	0.0061	-1.0	0.3039	0.4764	-1.0	-1.0
0.8725	18.0	3762	0.9129	0.3068	0.7622	0.1786	0.3083	0.0089	-1.0	0.026	0.1961	0.4817	0.4842	0.0082	-1.0	0.3068	0.4817	-1.0	-1.0
0.8725	19.0	3971	0.9053	0.3129	0.7699	0.182	0.3146	0.0119	-1.0	0.0253	0.1986	0.4806	0.4832	0.0102	-1.0	0.3129	0.4806	-1.0	-1.0
0.8532	20.0	4180	0.9124	0.3076	0.7661	0.1794	0.3093	0.0069	-1.0	0.0252	0.1972	0.4798	0.4823	0.0061	-1.0	0.3076	0.4798	-1.0	-1.0
0.8532	21.0	4389	0.9060	0.3129	0.7694	0.182	0.3146	0.0139	-1.0	0.0254	0.1988	0.4811	0.4837	0.0122	-1.0	0.3129	0.4811	-1.0	-1.0
0.8362	22.0	4598	0.9007	0.3157	0.7733	0.1886	0.3173	0.0079	-1.0	0.0255	0.2005	0.4834	0.4859	0.0061	-1.0	0.3157	0.4834	-1.0	-1.0
0.8362	23.0	4807	0.9036	0.3148	0.7702	0.1859	0.3159	0.0119	-1.0	0.0255	0.1982	0.4859	0.4884	0.0102	-1.0	0.3148	0.4859	-1.0	-1.0
0.8211	24.0	5016	0.8988	0.3159	0.7733	0.1875	0.3172	0.005	-1.0	0.0253	0.1988	0.4844	0.487	0.0041	-1.0	0.3159	0.4844	-1.0	-1.0
0.8211	25.0	5225	0.8989	0.3175	0.7741	0.1888	0.3189	0.0079	-1.0	0.0256	0.1995	0.486	0.4886	0.0061	-1.0	0.3175	0.486	-1.0	-1.0
0.8211	26.0	5434	0.8980	0.3188	0.776	0.1918	0.3204	0.005	-1.0	0.0258	0.1998	0.4867	0.4893	0.0041	-1.0	0.3188	0.4867	-1.0	-1.0
0.8091	27.0	5643	0.8953	0.3204	0.7786	0.1931	0.3219	0.0079	-1.0	0.026	0.2002	0.4863	0.4889	0.0061	-1.0	0.3204	0.4863	-1.0	-1.0
0.8091	28.0	5852	0.8973	0.3192	0.7784	0.1911	0.3208	0.0079	-1.0	0.0255	0.199	0.4867	0.4892	0.0061	-1.0	0.3192	0.4867	-1.0	-1.0
0.8001	29.0	6061	0.8962	0.3196	0.7785	0.1926	0.3211	0.0079	-1.0	0.0257	0.1994	0.487	0.4896	0.0061	-1.0	0.3196	0.487	-1.0	-1.0
0.8001	30.0	6270	0.8959	0.3195	0.7784	0.1925	0.3211	0.0079	-1.0	0.0256	0.1995	0.487	0.4896	0.0061	-1.0	0.3195	0.487	-1.0	-1.0

Framework versions

• Transformers 4.41.2
• Pytorch 2.3.0+cu121
• Datasets 2.20.0
• Tokenizers 0.19.1