This model is a fine-tuned version of facebook/detr-resnet-50 on the namnguyen059/PDFextract dataset. It achieves the following results on the evaluation set:

Loss: 3.3774
Map: 0.0076
Map 50: 0.0237
Map 75: 0.0004
Map Small: 0.0
Map Medium: 0.0117
Map Large: 0.0
Mar 1: 0.0014
Mar 10: 0.0459
Mar 100: 0.0622
Mar Small: 0.0
Mar Medium: 0.0958
Mar Large: 0.0
Map Table: 0.0
Mar 100 Table: 0.0
Map Text: 0.0152
Mar 100 Text: 0.1243

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: 1337
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 100.0

Training results

Training Loss	Epoch	Step	Validation Loss	Map	Map 50	Map 75	Map Medium	Map Large	Mar 1	Mar 10	Mar 100	Mar Medium	Mar Large	Map Text	Mar 100 Text
6.2935	1.0	5	5.7125	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5.6711	2.0	10	5.4708	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5.3937	3.0	15	5.2153	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
5.001	4.0	20	4.8025	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
4.8611	5.0	25	4.6390	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0
4.6947	6.0	30	4.2093	0.0	0.0002	0.0	0.0	0.0021	0.0	0.0	0.0027	0.0	0.1	0.0001	0.0054
4.1881	7.0	35	3.8453	0.0003	0.0031	0.0	0.0006	0.0012	0.0	0.0014	0.0054	0.0021	0.15	0.0007	0.0108
3.9652	8.0	40	4.0073	0.003	0.0182	0.0	0.0052	0.0017	0.0014	0.0081	0.0122	0.0125	0.15	0.0059	0.0243
3.9875	9.0	45	4.0535	0.0001	0.0005	0.0001	0.0001	0.0019	0.0	0.0	0.0122	0.0063	0.3	0.0002	0.0243
4.041	10.0	50	4.0097	0.0002	0.0017	0.0	0.0004	0.0009	0.0	0.0	0.0081	0.0083	0.1	0.0003	0.0162
3.959	11.0	55	3.8365	0.0005	0.0021	0.0	0.0012	0.0003	0.0	0.0068	0.0122	0.0167	0.05	0.0009	0.0243
3.8681	12.0	60	3.6800	0.0007	0.0041	0.0	0.0014	0.001	0.0	0.0068	0.0149	0.0167	0.15	0.0013	0.0297
3.7309	13.0	65	3.6922	0.0029	0.0249	0.0	0.0068	0.0008	0.0014	0.0081	0.0189	0.025	0.1	0.0058	0.0378
3.647	14.0	70	3.7221	0.0024	0.0158	0.0	0.0045	0.0016	0.0014	0.0108	0.0162	0.0167	0.2	0.0048	0.0324
3.6564	15.0	75	3.6637	0.0007	0.0048	0.0	0.0012	0.0025	0.0014	0.0054	0.0189	0.0229	0.15	0.0014	0.0378
3.5905	16.0	80	3.6443	0.0078	0.0179	0.0	0.0116	0.0074	0.0135	0.0203	0.023	0.0312	0.1	0.0157	0.0459
3.8091	17.0	85	3.8066	0.0018	0.0122	0.0	0.003	0.0	0.0041	0.0108	0.0108	0.0167	0.0	0.0035	0.0216
3.8171	18.0	90	3.6044	0.0006	0.0032	0.0	0.0012	0.0	0.0	0.0054	0.0095	0.0146	0.0	0.0013	0.0189
3.7048	19.0	95	3.6161	0.0	0.0	0.0	0.0	0.0018	0.0	0.0	0.0068	0.0	0.25	0.0	0.0135
3.8635	20.0	100	3.6542	0.0004	0.0015	0.0	0.0009	0.0	0.0	0.0054	0.0095	0.0146	0.0	0.0009	0.0189
3.7021	21.0	105	4.0802	0.0032	0.016	0.0	0.0053	0.0	0.0027	0.0068	0.0108	0.0167	0.0	0.0064	0.0216
3.7035	22.0	110	4.0508	0.0006	0.0041	0.0	0.001	0.0	0.0	0.0014	0.0162	0.025	0.0	0.0012	0.0324
3.6404	23.0	115	4.1771	0.0013	0.0057	0.0005	0.0022	0.0	0.0068	0.0162	0.0189	0.0292	0.0	0.0027	0.0378
3.7922	24.0	120	4.4617	0.006	0.0149	0.0	0.0099	0.0	0.0054	0.0054	0.0095	0.0146	0.0	0.0119	0.0189
3.7619	25.0	125	3.8533	0.001	0.0026	0.0	0.0053	0.0008	0.0	0.0108	0.0135	0.0167	0.1	0.0019	0.027
3.804	26.0	130	3.7589	0.001	0.0073	0.0	0.0036	0.0	0.0	0.0081	0.0122	0.0188	0.0	0.0021	0.0243
3.7785	27.0	135	3.6378	0.003	0.0197	0.0	0.01	0.0003	0.0027	0.0162	0.0189	0.0271	0.05	0.006	0.0378
3.726	28.0	140	3.7696	0.0003	0.0021	0.0001	0.0009	0.0	0.0	0.0041	0.0149	0.0229	0.0	0.0006	0.0297
3.682	29.0	145	3.8782	0.0003	0.0006	0.0003	0.001	0.0	0.0	0.0081	0.0108	0.0167	0.0	0.0006	0.0216
3.7331	30.0	150	4.3841	0.0005	0.0027	0.0	0.002	0.0009	0.0	0.0081	0.0189	0.025	0.1	0.001	0.0378
3.7627	31.0	155	4.7688	0.0007	0.0054	0.0	0.003	0.0	0.0	0.0081	0.0189	0.0292	0.0	0.0015	0.0378
3.8656	32.0	160	5.0427	0.0009	0.0046	0.0	0.0039	0.0007	0.0	0.0027	0.0324	0.0417	0.2	0.0018	0.0649
3.8845	33.0	165	4.1952	0.0012	0.0056	0.0	0.0027	0.0	0.0	0.0108	0.0243	0.0375	0.0	0.0024	0.0486
3.8416	34.0	170	4.1403	0.0048	0.0128	0.0006	0.0118	0.0048	0.0	0.0311	0.0703	0.1021	0.15	0.0096	0.1405
3.807	35.0	175	4.6646	0.0031	0.0218	0.0002	0.0103	0.0125	0.0014	0.0176	0.05	0.075	0.05	0.0062	0.1
3.8728	36.0	180	4.5890	0.0009	0.0039	0.0004	0.002	0.0	0.0	0.0014	0.0432	0.0667	0.0	0.0018	0.0865
3.7952	37.0	185	4.0300	0.0013	0.0048	0.0007	0.0023	0.0	0.0	0.0176	0.023	0.0354	0.0	0.0025	0.0459
3.725	38.0	190	3.9829	0.0003	0.0025	0.0001	0.0005	0.0	0.0	0.0014	0.0135	0.0208	0.0	0.0006	0.027
3.6884	39.0	195	3.7402	0.0008	0.0036	0.0	0.0009	0.0119	0.0	0.0014	0.027	0.0333	0.2	0.0016	0.0541
3.6963	40.0	200	3.4723	0.0002	0.0015	0.0	0.0004	0.0	0.0	0.0	0.0081	0.0125	0.0	0.0004	0.0162
3.7368	41.0	205	3.6067	0.0	0.0004	0.0	0.0001	0.0	0.0	0.0	0.0027	0.0042	0.0	0.0001	0.0054
3.7349	42.0	210	3.6135	0.0	0.0002	0.0	0.0	0.0012	0.0	0.0014	0.0014	0.0	0.05	0.0	0.0027
3.6353	43.0	215	3.5835	0.0018	0.0062	0.0014	0.0074	0.0	0.0	0.0122	0.027	0.0417	0.0	0.0036	0.0541
3.6948	44.0	220	3.7714	0.0023	0.0077	0.0001	0.0045	0.0	0.0	0.0162	0.0324	0.05	0.0	0.0046	0.0649
3.7546	45.0	225	3.9600	0.0074	0.0207	0.0	0.0126	0.0	0.0054	0.0189	0.0243	0.0375	0.0	0.0149	0.0486
3.6691	46.0	230	3.8211	0.0018	0.0094	0.0	0.0037	0.0	0.0	0.0095	0.0297	0.0458	0.0	0.0037	0.0595
3.5355	47.0	235	3.6537	0.0006	0.0036	0.0001	0.0017	0.0	0.0	0.0068	0.0216	0.0333	0.0	0.0012	0.0432
3.5232	48.0	240	3.5552	0.0004	0.002	0.0	0.0018	0.0	0.0	0.0027	0.0149	0.0229	0.0	0.0008	0.0297
3.48	49.0	245	3.4924	0.0014	0.0126	0.0	0.0043	0.0017	0.0014	0.0108	0.0122	0.0167	0.05	0.0028	0.0243
3.6056	50.0	250	3.4697	0.0007	0.0024	0.0	0.0019	0.0002	0.0	0.0095	0.0122	0.0167	0.05	0.0014	0.0243
3.4806	51.0	255	3.5357	0.0045	0.0339	0.0	0.0129	0.0006	0.0014	0.023	0.0284	0.0417	0.05	0.0091	0.0568
3.5547	52.0	260	3.5216	0.0016	0.0141	0.0	0.004	0.0021	0.0	0.0108	0.0257	0.0292	0.25	0.0032	0.0514
3.483	53.0	265	3.4858	0.0048	0.028	0.0005	0.0077	0.0	0.0	0.0311	0.0405	0.0625	0.0	0.0095	0.0811
3.6952	54.0	270	3.4944	0.0083	0.0661	0.0	0.0131	0.0	0.0041	0.027	0.0284	0.0437	0.0	0.0165	0.0568
3.5447	55.0	275	3.5247	0.0049	0.0251	0.0	0.0079	0.0	0.0027	0.0176	0.0257	0.0396	0.0	0.0099	0.0514
3.6204	56.0	280	3.5457	0.0069	0.0452	0.0	0.0168	0.0	0.0068	0.0135	0.0257	0.0396	0.0	0.0137	0.0514
3.3873	57.0	285	3.4963	0.0134	0.0576	0.0	0.0218	0.0	0.0081	0.0365	0.0473	0.0729	0.0	0.0268	0.0946
3.5384	58.0	290	3.5618	0.0016	0.0083	0.0	0.0025	0.0	0.0	0.0122	0.0338	0.0521	0.0	0.0032	0.0676
3.6034	59.0	295	3.4600	0.0016	0.0102	0.0	0.0025	0.0	0.0	0.0081	0.0324	0.05	0.0	0.0032	0.0649
3.5524	60.0	300	3.4434	0.0012	0.0063	0.0004	0.0019	0.0	0.0	0.0041	0.0324	0.05	0.0	0.0024	0.0649
3.4222	61.0	305	3.5144	0.0043	0.0113	0.0002	0.0066	0.0	0.0	0.027	0.0419	0.0646	0.0	0.0086	0.0838
3.451	62.0	310	3.5189	0.0198	0.0582	0.0002	0.0304	0.0007	0.0189	0.0351	0.0446	0.0646	0.1	0.0396	0.0892
3.4485	63.0	315	3.4585	0.0125	0.0428	0.0	0.0202	0.0	0.0135	0.0297	0.0405	0.0625	0.0	0.0249	0.0811
3.475	64.0	320	3.3541	0.0063	0.0297	0.0012	0.0111	0.0016	0.0027	0.0216	0.0405	0.0583	0.1	0.0126	0.0811
3.4189	65.0	325	3.3517	0.0051	0.0315	0.0	0.0094	0.0	0.0014	0.0257	0.0378	0.0583	0.0	0.0102	0.0757
3.324	66.0	330	3.2557	0.0025	0.0134	0.0	0.0045	0.0	0.0	0.0216	0.0243	0.0375	0.0	0.005	0.0486
3.5049	67.0	335	3.3093	0.0031	0.0108	0.0	0.0051	0.0	0.0	0.027	0.027	0.0417	0.0	0.0061	0.0541
3.3219	68.0	340	3.3455	0.0119	0.0596	0.0008	0.0189	0.0	0.0095	0.0284	0.0432	0.0667	0.0	0.0239	0.0865
3.3633	69.0	345	3.4110	0.0268	0.0747	0.0034	0.0414	0.0	0.0203	0.0662	0.0662	0.1021	0.0	0.0536	0.1324
3.4628	70.0	350	3.3594	0.0024	0.0151	0.0002	0.0037	0.0	0.0014	0.0216	0.0405	0.0625	0.0	0.0048	0.0811
3.272	71.0	355	3.3655	0.0039	0.0174	0.0008	0.0061	0.0	0.0014	0.0149	0.0473	0.0729	0.0	0.0079	0.0946
3.3713	72.0	360	3.3455	0.0097	0.0609	0.0018	0.015	0.0	0.0108	0.0122	0.0473	0.0729	0.0	0.0195	0.0946
3.258	73.0	365	3.3679	0.0006	0.0025	0.0	0.0011	0.0	0.0	0.0068	0.023	0.0354	0.0	0.0013	0.0459
3.3666	74.0	370	3.3173	0.0004	0.0024	0.0	0.0008	0.0	0.0	0.0014	0.0216	0.0333	0.0	0.0009	0.0432
3.2339	75.0	375	3.3463	0.002	0.0103	0.0	0.0033	0.0	0.0	0.0081	0.0405	0.0625	0.0	0.004	0.0811
3.2601	76.0	380	3.3633	0.002	0.0111	0.0	0.0032	0.0	0.0	0.0081	0.0392	0.0604	0.0	0.0039	0.0784
3.3584	77.0	385	3.3331	0.0025	0.0108	0.0	0.004	0.0	0.0	0.0081	0.0486	0.075	0.0	0.0051	0.0973
3.2928	78.0	390	3.3012	0.0043	0.0139	0.0002	0.0066	0.0	0.0	0.0162	0.0635	0.0979	0.0	0.0085	0.127
3.2633	79.0	395	3.2913	0.0076	0.0226	0.0003	0.0117	0.0	0.0027	0.0392	0.0432	0.0667	0.0	0.0153	0.0865
3.3965	80.0	400	3.3141	0.008	0.0224	0.0051	0.0126	0.0	0.0014	0.0405	0.0419	0.0646	0.0	0.016	0.0838
3.2753	81.0	405	3.4127	0.0077	0.0231	0.0036	0.012	0.0	0.0	0.0405	0.0432	0.0667	0.0	0.0155	0.0865
3.3362	82.0	410	3.4194	0.0099	0.0365	0.0032	0.0152	0.0	0.0054	0.0473	0.0568	0.0875	0.0	0.0197	0.1135
3.3752	83.0	415	3.3421	0.0086	0.0411	0.0	0.0134	0.0	0.0014	0.0351	0.0432	0.0667	0.0	0.0173	0.0865
3.3222	84.0	420	3.2956	0.004	0.0234	0.0	0.0063	0.0	0.0014	0.0243	0.0243	0.0375	0.0	0.008	0.0486
3.3732	85.0	425	3.2515	0.0281	0.0614	0.0	0.0434	0.0	0.0257	0.0338	0.0459	0.0708	0.0	0.0562	0.0919
3.1932	86.0	430	3.2197	0.0263	0.0603	0.0	0.0409	0.0	0.0243	0.0297	0.0459	0.0708	0.0	0.0527	0.0919
3.3181	87.0	435	3.3373	0.0027	0.0142	0.0	0.0042	0.0	0.0	0.023	0.0297	0.0458	0.0	0.0055	0.0595
3.3382	88.0	440	3.3707	0.0019	0.0069	0.0	0.003	0.0	0.0	0.0243	0.0297	0.0458	0.0	0.0039	0.0595
3.1525	89.0	445	3.4728	0.0025	0.0056	0.0	0.0038	0.0	0.0	0.0324	0.0324	0.05	0.0	0.005	0.0649
3.1836	90.0	450	3.4726	0.0103	0.0357	0.0	0.0163	0.0	0.0041	0.0486	0.05	0.0771	0.0	0.0206	0.1
3.2294	91.0	455	3.4077	0.0116	0.0309	0.0	0.0183	0.0	0.0041	0.0581	0.0595	0.0917	0.0	0.0233	0.1189
3.2468	92.0	460	3.3905	0.0101	0.0324	0.0004	0.0158	0.0	0.0027	0.0581	0.0581	0.0896	0.0	0.0202	0.1162
3.2433	93.0	465	3.3989	0.0099	0.0309	0.0004	0.0155	0.0	0.0027	0.0581	0.0581	0.0896	0.0	0.0198	0.1162
3.152	94.0	470	3.3957	0.0113	0.0352	0.0003	0.0177	0.0	0.0027	0.0568	0.0568	0.0875	0.0	0.0226	0.1135
3.2372	95.0	475	3.3941	0.0105	0.0345	0.0003	0.0163	0.0	0.0027	0.0527	0.0527	0.0812	0.0	0.0209	0.1054
3.2734	96.0	480	3.3851	0.0103	0.0256	0.0003	0.016	0.0	0.0041	0.0568	0.0568	0.0875	0.0	0.0205	0.1135
3.2559	97.0	485	3.3807	0.0113	0.0308	0.0003	0.0174	0.0	0.0041	0.0595	0.0608	0.0938	0.0	0.0226	0.1216
3.3223	98.0	490	3.3773	0.0094	0.03	0.0003	0.0146	0.0	0.0027	0.0527	0.0608	0.0938	0.0	0.0189	0.1216
3.1813	99.0	495	3.3789	0.0084	0.025	0.0004	0.0129	0.0	0.0027	0.0486	0.0622	0.0958	0.0	0.0168	0.1243
3.1959	100.0	500	3.3774	0.0076	0.0237	0.0004	0.0117	0.0	0.0014	0.0459	0.0622	0.0958	0.0	0.0152	0.1243

Framework versions

Transformers 4.44.0.dev0
Pytorch 2.3.1
Datasets 2.20.0
Tokenizers 0.19.1

namnguyen059/detr-finetuned-cppe-5-10k-steps

detr-finetuned-cppe-5-10k-steps

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

namnguyen059/
detr-finetuned-cppe-5-10k-steps