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

Loss: 0.5892

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: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 200
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss
6.674	0.28	100	1.7571
1.4721	0.56	200	1.2737
1.2557	0.84	300	1.1037
1.0781	1.12	400	1.0184
1.0353	1.4	500	0.9988
1.0324	1.69	600	0.9951
0.9131	1.97	700	0.9224
0.8724	2.25	800	0.9692
0.8129	2.53	900	0.8670
0.9	2.81	1000	0.8326
0.7993	3.09	1100	0.7875
0.7907	3.37	1200	0.7517
0.8424	3.65	1300	0.9088
0.7808	3.93	1400	0.8506
0.7469	4.21	1500	0.7928
0.7582	4.49	1600	0.7228
0.7546	4.78	1700	0.7588
0.7842	5.06	1800	0.7726
0.775	5.34	1900	0.7676
0.7263	5.62	2000	0.7164
0.7209	5.9	2100	0.7061
0.7259	6.18	2200	0.7579
0.7701	6.46	2300	0.8184
0.7391	6.74	2400	0.6684
0.6834	7.02	2500	0.7042
0.7098	7.3	2600	0.7166
0.7498	7.58	2700	0.6752
0.7056	7.87	2800	0.7064
0.7004	8.15	2900	0.7090
0.6964	8.43	3000	0.7318
0.682	8.71	3100	0.7216
0.7309	8.99	3200	0.6545
0.6576	9.27	3300	0.6478
0.7014	9.55	3400	0.6814
0.673	9.83	3500	0.6783
0.6455	10.11	3600	0.7248
0.7041	10.39	3700	0.7729
0.6664	10.67	3800	0.6746
0.6161	10.96	3900	0.6414
0.6975	11.24	4000	0.6637
0.6751	11.52	4100	0.6570
0.6092	11.8	4200	0.6691
0.6593	12.08	4300	0.6276
0.6449	12.36	4400	0.6388
0.6136	12.64	4500	0.6711
0.6521	12.92	4600	0.6768
0.6162	13.2	4700	0.6427
0.7083	13.48	4800	0.6492
0.6407	13.76	4900	0.6213
0.6371	14.04	5000	0.6674
0.626	14.33	5100	0.6185
0.6442	14.61	5200	0.7180
0.5981	14.89	5300	0.6441
0.629	15.17	5400	0.6262
0.625	15.45	5500	0.6397
0.6123	15.73	5600	0.6440
0.6084	16.01	5700	0.6493
0.6021	16.29	5800	0.6263
0.6502	16.57	5900	0.6254
0.6339	16.85	6000	0.7043
0.5925	17.13	6100	0.8014
0.6453	17.42	6200	0.6385
0.6143	17.7	6300	0.6033
0.6057	17.98	6400	0.6881
0.6386	18.26	6500	0.6366
0.5839	18.54	6600	0.6563
0.6013	18.82	6700	0.5982
0.5999	19.1	6800	0.6064
0.6023	19.38	6900	0.5795
0.5593	19.66	7000	0.6538
0.6375	19.94	7100	0.6991
0.6073	20.22	7200	0.7117
0.596	20.51	7300	0.6034
0.5987	20.79	7400	0.6489
0.5922	21.07	7500	0.6216
0.589	21.35	7600	0.6257
0.6047	21.63	7700	0.6415
0.5775	21.91	7800	0.6159
0.588	22.19	7900	0.6095
0.5844	22.47	8000	0.6373
0.5964	22.75	8100	0.6022
0.5987	23.03	8200	0.6050
0.5605	23.31	8300	0.6083
0.5835	23.6	8400	0.7823
0.5816	23.88	8500	0.6417
0.5757	24.16	8600	0.6324
0.5997	24.44	8700	0.6046
0.5674	24.72	8800	0.6558
0.5703	25.0	8900	0.5819
0.5766	25.28	9000	0.6116
0.5548	25.56	9100	0.5877
0.564	25.84	9200	0.5672
0.548	26.12	9300	0.6073
0.5436	26.4	9400	0.5739
0.6006	26.69	9500	0.6101
0.5519	26.97	9600	0.5869
0.5432	27.25	9700	0.5721
0.5597	27.53	9800	0.5807
0.5254	27.81	9900	0.5849
0.5366	28.09	10000	0.5831
0.5654	28.37	10100	0.5993
0.57	28.65	10200	0.5892

Framework versions

Transformers 4.26.1
Pytorch 1.13.0+cu117
Datasets 2.10.1
Tokenizers 0.13.2

davanstrien/conditional-detr-resnet-50_fine_tuned_beyond_words

conditional-detr-resnet-50_fine_tuned_beyond_words

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

davanstrien/
conditional-detr-resnet-50_fine_tuned_beyond_words