detr_domain_shift

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

• Loss: 0.4246
• Map: 0.8353
• Map 50: 0.9529
• Map 75: 0.9029
• Map Small: 0.0159
• Map Medium: 0.6742
• Map Large: 0.8707
• Mar 1: 0.7106
• Mar 10: 0.8977
• Mar 100: 0.9152
• Mar Small: 0.2984
• Mar Medium: 0.8276
• Mar Large: 0.9365
• Map Garbage bag: 0.8185
• Mar 100 Garbage bag: 0.9059
• Map Paper bag: 0.8446
• Mar 100 Paper bag: 0.9239
• Map Plastic bag: 0.8429
• Mar 100 Plastic bag: 0.9159

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: 0.0001
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 50

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 Garbage bag	Mar 100 Garbage bag	Map Paper bag	Mar 100 Paper bag	Map Plastic bag	Mar 100 Plastic bag
0.9775	1.0	1557	0.8739	0.2419	0.3276	0.2719	0.0003	0.1808	0.2566	0.4246	0.773	0.8186	0.0271	0.6849	0.8501	0.1999	0.808	0.3045	0.8292	0.2213	0.8187
0.919	2.0	3114	0.8527	0.4465	0.6037	0.5018	0.0043	0.2877	0.4787	0.5098	0.7627	0.8137	0.0123	0.6656	0.8474	0.5496	0.8075	0.4518	0.8235	0.3382	0.8101
0.85	3.0	4671	0.7649	0.5508	0.7268	0.6183	0.0013	0.3684	0.5898	0.5619	0.7944	0.8335	0.0652	0.7147	0.8613	0.5635	0.8206	0.5877	0.8471	0.5012	0.8327
0.844	4.0	6228	0.7562	0.605	0.7924	0.6857	0.0023	0.4402	0.6405	0.5732	0.7901	0.8296	0.0579	0.7174	0.856	0.6113	0.8213	0.6125	0.8364	0.5912	0.8313
0.8332	5.0	7785	0.7811	0.5982	0.8004	0.6799	0.007	0.4141	0.6365	0.5683	0.7801	0.8214	0.0194	0.6914	0.852	0.5785	0.8091	0.6326	0.8361	0.5834	0.8188
0.9303	6.0	9342	0.9325	0.5179	0.7395	0.5906	0.0001	0.3665	0.5506	0.5192	0.7288	0.7756	0.0097	0.6358	0.8074	0.52	0.7664	0.5744	0.7825	0.4593	0.7778
0.8773	7.0	10899	0.8869	0.5561	0.7748	0.626	0.0	0.374	0.5948	0.547	0.748	0.7895	0.0	0.6483	0.8223	0.5435	0.7752	0.5909	0.8132	0.5339	0.7801
1.011	8.0	12456	0.8729	0.5604	0.7945	0.6437	0.0013	0.3991	0.5954	0.538	0.7475	0.7971	0.0218	0.6489	0.831	0.5251	0.7729	0.6152	0.8222	0.541	0.7962
1.0019	9.0	14013	1.2187	0.313	0.5258	0.3396	0.0	0.2026	0.337	0.3859	0.6118	0.6892	0.0	0.5141	0.7282	0.4395	0.7076	0.2496	0.6916	0.25	0.6685
0.9286	10.0	15570	0.8059	0.5848	0.7852	0.6584	0.0	0.3874	0.6257	0.5595	0.7725	0.8196	0.0179	0.6689	0.8539	0.5762	0.8015	0.6239	0.8429	0.5544	0.8143
0.8782	11.0	17127	0.8458	0.5587	0.7551	0.627	0.01	0.377	0.596	0.5551	0.7678	0.812	0.0516	0.6635	0.8453	0.5726	0.7966	0.6164	0.8355	0.4872	0.804
0.8493	12.0	18684	0.7872	0.5899	0.7787	0.6622	0.0012	0.3803	0.6329	0.57	0.7832	0.8287	0.019	0.6719	0.8643	0.5813	0.8215	0.6336	0.8412	0.5548	0.8235
0.793	13.0	20241	0.7784	0.6112	0.8128	0.6924	0.0023	0.4275	0.6502	0.5785	0.7783	0.8249	0.0434	0.6872	0.8566	0.6157	0.8253	0.6326	0.8344	0.5852	0.8151
0.7712	14.0	21798	0.7881	0.6089	0.8162	0.688	0.0001	0.4211	0.6497	0.5772	0.7785	0.8184	0.039	0.6874	0.8491	0.5984	0.803	0.6429	0.84	0.5853	0.8123
0.7892	15.0	23355	0.7204	0.6548	0.8515	0.742	0.0007	0.486	0.6916	0.6018	0.7991	0.8376	0.0359	0.7139	0.8668	0.653	0.8377	0.6716	0.8447	0.6399	0.8302
0.7735	16.0	24912	0.7768	0.604	0.7908	0.6807	0.0152	0.4142	0.6445	0.5803	0.7794	0.8258	0.043	0.7167	0.8522	0.5854	0.8359	0.6421	0.8307	0.5844	0.8106
0.8052	17.0	26469	0.7466	0.6389	0.8417	0.7325	0.0001	0.4658	0.6767	0.5883	0.7847	0.8231	0.0245	0.7291	0.8469	0.6493	0.8262	0.6572	0.8329	0.6101	0.8101
0.7366	18.0	28026	0.7490	0.6372	0.8435	0.7198	0.0069	0.4492	0.6783	0.5883	0.7881	0.8287	0.0271	0.7084	0.8572	0.6427	0.8307	0.6486	0.8376	0.6202	0.8178
0.7316	19.0	29583	0.6964	0.6697	0.869	0.7578	0.0028	0.5095	0.704	0.6068	0.8029	0.839	0.0705	0.7283	0.8651	0.6529	0.8277	0.6915	0.8538	0.6646	0.8353
0.7243	20.0	31140	0.7165	0.6616	0.8605	0.753	0.0033	0.4782	0.7003	0.6046	0.7972	0.8299	0.0581	0.7148	0.8574	0.6366	0.8126	0.6791	0.8452	0.6692	0.832
0.7189	21.0	32697	0.6921	0.6748	0.8694	0.7621	0.001	0.5019	0.7121	0.6105	0.8042	0.8382	0.0788	0.7253	0.8649	0.6422	0.8241	0.7087	0.8581	0.6735	0.8324
0.6802	22.0	34254	0.6381	0.7091	0.886	0.7875	0.0007	0.5366	0.7465	0.6318	0.8291	0.8618	0.0716	0.752	0.8883	0.6953	0.8572	0.7253	0.8718	0.7067	0.8565
0.6676	23.0	35811	0.6252	0.7186	0.8865	0.7994	0.0034	0.5423	0.7573	0.6391	0.8373	0.8649	0.0665	0.7597	0.8908	0.7021	0.8567	0.7358	0.8752	0.7181	0.8628
0.6624	24.0	37368	0.6432	0.7117	0.8986	0.8041	0.0019	0.5384	0.7488	0.63	0.8205	0.8535	0.0974	0.7492	0.8786	0.6871	0.8394	0.7287	0.8639	0.7194	0.8572
0.6356	25.0	38925	0.6101	0.7284	0.905	0.8178	0.004	0.5444	0.7676	0.6448	0.8322	0.863	0.1108	0.7616	0.8874	0.7125	0.8556	0.7417	0.8709	0.7311	0.8626
0.6319	26.0	40482	0.6330	0.7191	0.9005	0.8113	0.0085	0.557	0.754	0.6392	0.8262	0.8513	0.106	0.7566	0.8745	0.702	0.849	0.7371	0.8558	0.7182	0.8489
0.6069	27.0	42039	0.5855	0.7461	0.9072	0.824	0.0082	0.5615	0.7852	0.6512	0.845	0.8742	0.2484	0.7625	0.9003	0.7159	0.8596	0.7615	0.884	0.7609	0.8791
0.5898	28.0	43596	0.5582	0.7581	0.9158	0.8405	0.0091	0.5844	0.7957	0.6649	0.8552	0.878	0.2474	0.785	0.9008	0.7384	0.8677	0.7742	0.8882	0.7616	0.8782
0.5777	29.0	45153	0.5412	0.7706	0.9226	0.8492	0.0029	0.6117	0.8054	0.6718	0.8591	0.8878	0.2147	0.7882	0.9118	0.7542	0.8826	0.7799	0.8953	0.7777	0.8854
0.5461	30.0	46710	0.5424	0.7714	0.9259	0.8563	0.0024	0.6033	0.8081	0.6708	0.8571	0.8829	0.206	0.7823	0.9069	0.7602	0.8784	0.7814	0.8906	0.7726	0.8798
0.5392	31.0	48267	0.5274	0.7773	0.9219	0.8591	0.0026	0.5986	0.8159	0.6752	0.8648	0.8888	0.2832	0.795	0.9111	0.7549	0.8835	0.7902	0.894	0.7867	0.8889
0.5367	32.0	49824	0.5181	0.7863	0.9312	0.8654	0.0038	0.6234	0.8216	0.678	0.8663	0.8882	0.2332	0.7956	0.9109	0.7689	0.8805	0.7985	0.8962	0.7915	0.8879
0.5187	33.0	51381	0.5079	0.7853	0.934	0.8672	0.0187	0.6245	0.8206	0.6814	0.8681	0.8917	0.2412	0.7984	0.9143	0.7692	0.89	0.7937	0.8966	0.7931	0.8884
0.5102	34.0	52938	0.4861	0.8049	0.9381	0.8761	0.0318	0.6481	0.8388	0.6912	0.8811	0.9047	0.2659	0.818	0.9262	0.7893	0.8991	0.8191	0.914	0.8062	0.901
0.4868	35.0	54495	0.4753	0.8046	0.9392	0.8827	0.0084	0.6389	0.8413	0.6933	0.8796	0.9015	0.2495	0.8056	0.9249	0.7884	0.8981	0.811	0.9036	0.8143	0.903
0.4821	36.0	56052	0.4714	0.8096	0.9427	0.8861	0.0225	0.6485	0.8448	0.6955	0.8828	0.9042	0.278	0.8159	0.9258	0.7923	0.8981	0.8221	0.9121	0.8143	0.9024
0.4714	37.0	57609	0.4447	0.8232	0.9433	0.8911	0.0219	0.6619	0.8599	0.7032	0.8935	0.9126	0.269	0.8235	0.9344	0.8072	0.9051	0.8295	0.9173	0.8329	0.9154
0.4653	38.0	59166	0.4554	0.819	0.946	0.8912	0.0178	0.6567	0.8548	0.7021	0.8873	0.9059	0.3004	0.8138	0.9281	0.8025	0.899	0.8281	0.9122	0.8264	0.9065
0.4494	39.0	60723	0.4310	0.8308	0.9451	0.8939	0.0243	0.6685	0.8663	0.7091	0.8985	0.918	0.3242	0.8358	0.9381	0.8127	0.9114	0.8401	0.9237	0.8397	0.919
0.4389	40.0	62280	0.4289	0.8336	0.9489	0.8967	0.0177	0.6757	0.8684	0.7102	0.8976	0.9165	0.2958	0.8329	0.9369	0.8167	0.9102	0.8425	0.9219	0.8416	0.9174
0.4376	41.0	63837	0.4245	0.8365	0.9498	0.9011	0.0179	0.6795	0.8709	0.7122	0.8985	0.9182	0.3335	0.8382	0.9377	0.8167	0.9067	0.8477	0.9272	0.845	0.9205
0.4252	42.0	65394	0.4244	0.8368	0.9511	0.8998	0.0175	0.6802	0.8713	0.7106	0.8988	0.9166	0.3604	0.8331	0.9369	0.8183	0.9089	0.8467	0.9229	0.8456	0.9179
0.4215	43.0	66951	0.4281	0.8342	0.9517	0.8987	0.0297	0.6731	0.8696	0.7088	0.8985	0.9159	0.3176	0.83	0.9369	0.8177	0.9092	0.8431	0.9217	0.8418	0.9169
0.4279	44.0	68508	0.4235	0.8375	0.9527	0.9012	0.0191	0.6777	0.873	0.7116	0.8984	0.9164	0.32	0.8288	0.9376	0.8191	0.9094	0.8479	0.9226	0.8453	0.9172
0.4133	45.0	70065	0.4220	0.837	0.9525	0.9014	0.0168	0.6764	0.8719	0.7117	0.9	0.9175	0.3059	0.8353	0.9377	0.8168	0.9081	0.8488	0.9244	0.8455	0.9199
0.4085	46.0	71622	0.4231	0.837	0.9538	0.9026	0.0158	0.679	0.8721	0.7118	0.8986	0.9157	0.2905	0.8276	0.9371	0.8196	0.908	0.8463	0.9227	0.8451	0.9165
0.4138	47.0	73179	0.4269	0.8335	0.9529	0.9019	0.014	0.6735	0.8689	0.7097	0.8965	0.9141	0.2954	0.8246	0.9358	0.8145	0.9048	0.8436	0.9226	0.8424	0.9149
0.4147	48.0	74736	0.4246	0.8353	0.9529	0.9023	0.0158	0.6736	0.8709	0.7108	0.8979	0.9155	0.2982	0.8266	0.937	0.8182	0.9062	0.8447	0.9242	0.843	0.916
0.4145	49.0	76293	0.4237	0.8362	0.9531	0.9027	0.0159	0.6751	0.8715	0.7108	0.8984	0.916	0.2984	0.8286	0.9373	0.8192	0.9067	0.8454	0.9248	0.8439	0.9167
0.406	50.0	77850	0.4246	0.8353	0.9529	0.9029	0.0159	0.6742	0.8707	0.7106	0.8977	0.9152	0.2984	0.8276	0.9365	0.8185	0.9059	0.8446	0.9239	0.8429	0.9159

Framework versions

• Transformers 4.49.0
• Pytorch 2.5.1+cu118
• Datasets 3.3.2
• Tokenizers 0.21.0