detr-V8

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: 0.2139

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: 2e-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: linear
• num_epochs: 50

Training results

Training Loss	Epoch	Step	Validation Loss
No log	0.48	1000	0.3770
No log	0.96	2000	0.3967
0.4391	1.43	3000	0.3822
0.4391	1.91	4000	0.4163
0.4434	2.39	5000	0.3888
0.4434	2.87	6000	0.3867
0.4509	3.35	7000	0.4205
0.4509	3.83	8000	0.4014
0.455	4.3	9000	0.4117
0.455	4.78	10000	0.3964
0.4476	5.26	11000	0.3915
0.4476	5.74	12000	0.3919
0.444	6.22	13000	0.4026
0.444	6.7	14000	0.3832
0.443	7.17	15000	0.4057
0.443	7.65	16000	0.3677
0.4232	8.13	17000	0.3746
0.4232	8.61	18000	0.3672
0.4202	9.09	19000	0.3629
0.4202	9.56	20000	0.3739
0.4131	10.04	21000	0.3712
0.4131	10.52	22000	0.3470
0.4131	11.0	23000	0.3632
0.4024	11.48	24000	0.3561
0.4024	11.96	25000	0.3562
0.4013	12.43	26000	0.3253
0.4013	12.91	27000	0.3390
0.3925	13.39	28000	0.3398
0.3925	13.87	29000	0.3460
0.3804	14.35	30000	0.3338
0.3804	14.83	31000	0.3201
0.3757	15.3	32000	0.3119
0.3757	15.78	33000	0.3106
0.3663	16.26	34000	0.3164
0.3663	16.74	35000	0.3190
0.3588	17.22	36000	0.3141
0.3588	17.69	37000	0.3262
0.3515	18.17	38000	0.3027
0.3515	18.65	39000	0.3178
0.3557	19.13	40000	0.3053
0.3557	19.61	41000	0.3032
0.3478	20.09	42000	0.3147
0.3478	20.56	43000	0.3069
0.3451	21.04	44000	0.3070
0.3451	21.52	45000	0.3055
0.3451	22.0	46000	0.2883
0.3367	22.48	47000	0.3090
0.3367	22.96	48000	0.2906
0.3348	23.43	49000	0.2805
0.3348	23.91	50000	0.2920
0.3298	24.39	51000	0.2854
0.3298	24.87	52000	0.2841
0.3254	25.35	53000	0.2822
0.3254	25.82	54000	0.2716
0.3169	26.3	55000	0.2825
0.3169	26.78	56000	0.2700
0.314	27.26	57000	0.2640
0.314	27.74	58000	0.2728
0.3047	28.22	59000	0.2654
0.3047	28.69	60000	0.2691
0.2999	29.17	61000	0.2601
0.2999	29.65	62000	0.2607
0.297	30.13	63000	0.2581
0.297	30.61	64000	0.2511
0.2946	31.09	65000	0.2557
0.2946	31.56	66000	0.2568
0.2912	32.04	67000	0.2569
0.2912	32.52	68000	0.2594
0.2912	33.0	69000	0.2553
0.2906	33.48	70000	0.2425
0.2906	33.96	71000	0.2475
0.2833	34.43	72000	0.2394
0.2833	34.91	73000	0.2422
0.278	35.39	74000	0.2403
0.278	35.87	75000	0.2349
0.2738	36.35	76000	0.2300
0.2738	36.82	77000	0.2332
0.2701	37.3	78000	0.2309
0.2701	37.78	79000	0.2298
0.2659	38.26	80000	0.2343
0.2659	38.74	81000	0.2265
0.2626	39.22	82000	0.2310
0.2626	39.69	83000	0.2255
0.259	40.17	84000	0.2263
0.259	40.65	85000	0.2282
0.2563	41.13	86000	0.2309
0.2563	41.61	87000	0.2270
0.2548	42.09	88000	0.2237
0.2548	42.56	89000	0.2203
0.254	43.04	90000	0.2204
0.254	43.52	91000	0.2218
0.254	44.0	92000	0.2207
0.2484	44.48	93000	0.2144
0.2484	44.95	94000	0.2194
0.2475	45.43	95000	0.2165
0.2475	45.91	96000	0.2162
0.2453	46.39	97000	0.2136
0.2453	46.87	98000	0.2152
0.2441	47.35	99000	0.2162
0.2441	47.82	100000	0.2171
0.2408	48.3	101000	0.2119
0.2408	48.78	102000	0.2131
0.2389	49.26	103000	0.2109
0.2389	49.74	104000	0.2139

Framework versions

• Transformers 4.31.0
• Pytorch 2.0.1+cu118
• Datasets 2.13.1
• Tokenizers 0.13.3