You can deploy this model and start earning money today!

Mrtnl/
refr-dfine-xpu

Name: refr-dfine-xpu
Availability: InStock

generated_from_trainer

ustc-community/dfine-small-coco

finetune

apache-2.0

endpoints_compatible

Views

No views yet

refr-dfine-xpu

This model is a fine-tuned version of ustc-community/dfine-small-coco on the Mrtnl/refr-defect-detection dataset. It achieves the following results on the evaluation set:

Loss: 2.1195
Map: 0.0177
Map 50: 0.0417
Map 75: 0.0183
Map Small: -1.0
Map Medium: 0.0398
Map Large: 0.0077
Mar 1: 0.0
Mar 10: 0.15
Mar 100: 0.5
Mar Small: -1.0
Mar Medium: 0.55
Mar Large: 0.4
Map Defect: 0.0177
Mar 100 Defect: 0.5

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: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
lr_scheduler_type: linear
num_epochs: 50.0

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 Defect	Mar 100 Defect
No log	1.0	3	30.7095	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	2.0	6	21.3453	0.0002	0.0003	0.0	-1.0	0.0	0.0007	0.0	0.0	0.0833	-1.0	0.0	0.25	0.0002	0.0833
No log	3.0	9	18.5704	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	4.0	12	12.9789	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	5.0	15	11.0321	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	6.0	18	9.0819	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	7.0	21	7.0759	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	8.0	24	5.5016	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	9.0	27	4.5953	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	10.0	30	3.9110	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	11.0	33	3.4325	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	12.0	36	3.1023	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	13.0	39	2.8444	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	14.0	42	2.6560	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	15.0	45	2.5767	0.0	0.0004	0.0	-1.0	0.0002	0.0	0.0	0.0	0.0167	-1.0	0.025	0.0	0.0	0.0167
No log	16.0	48	2.6132	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	17.0	51	2.7969	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	18.0	54	2.7399	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	19.0	57	2.7303	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	20.0	60	2.5602	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0	0.0	-1.0	0.0	0.0	0.0	0.0
No log	21.0	63	2.5263	0.0001	0.0003	0.0	-1.0	0.0004	0.0	0.0	0.0	0.05	-1.0	0.075	0.0	0.0001	0.05
No log	22.0	66	2.3580	0.0014	0.0053	0.0	-1.0	0.0055	0.0	0.0	0.0	0.2333	-1.0	0.35	0.0	0.0014	0.2333
No log	23.0	69	2.3670	0.0015	0.0048	0.0003	-1.0	0.0076	0.0	0.0	0.0	0.2667	-1.0	0.4	0.0	0.0015	0.2667
No log	24.0	72	2.3524	0.0012	0.0047	0.0003	-1.0	0.0033	0.0003	0.0	0.0	0.2167	-1.0	0.275	0.1	0.0012	0.2167
No log	25.0	75	2.2621	0.0023	0.0059	0.0004	-1.0	0.0096	0.0	0.0	0.0	0.3	-1.0	0.45	0.0	0.0023	0.3
No log	26.0	78	2.2820	0.0026	0.0102	0.0003	-1.0	0.0093	0.0012	0.0	0.0	0.3167	-1.0	0.45	0.05	0.0026	0.3167
No log	27.0	81	2.3458	0.0019	0.0063	0.0006	-1.0	0.0076	0.0	0.0	0.0	0.25	-1.0	0.375	0.0	0.0019	0.25
No log	28.0	84	2.2687	0.0039	0.0135	0.0031	-1.0	0.0126	0.0002	0.0	0.0833	0.3167	-1.0	0.45	0.05	0.0039	0.3167
No log	29.0	87	2.2480	0.0054	0.015	0.0045	-1.0	0.016	0.0	0.0	0.0833	0.3	-1.0	0.45	0.0	0.0054	0.3
No log	30.0	90	2.2707	0.0071	0.0212	0.0046	-1.0	0.0153	0.0027	0.0	0.0833	0.3667	-1.0	0.425	0.25	0.0071	0.3667
No log	31.0	93	2.2259	0.008	0.0204	0.0079	-1.0	0.019	0.0026	0.0	0.1167	0.3667	-1.0	0.425	0.25	0.008	0.3667
No log	32.0	96	2.2757	0.0046	0.0127	0.0032	-1.0	0.0134	0.0014	0.0	0.0	0.3667	-1.0	0.425	0.25	0.0046	0.3667
No log	33.0	99	2.2032	0.0053	0.0137	0.0046	-1.0	0.0142	0.002	0.0	0.0	0.3667	-1.0	0.425	0.25	0.0053	0.3667
No log	34.0	102	2.1709	0.0086	0.0246	0.006	-1.0	0.0188	0.0063	0.0	0.1	0.4167	-1.0	0.475	0.3	0.0086	0.4167
No log	35.0	105	2.2031	0.008	0.0209	0.0037	-1.0	0.0189	0.0045	0.0	0.1167	0.4167	-1.0	0.45	0.35	0.008	0.4167
No log	36.0	108	2.1890	0.0092	0.0261	0.0041	-1.0	0.0207	0.0064	0.0	0.1167	0.4167	-1.0	0.45	0.35	0.0092	0.4167
No log	37.0	111	2.1537	0.0128	0.0311	0.0067	-1.0	0.0241	0.009	0.0	0.1167	0.4333	-1.0	0.45	0.4	0.0128	0.4333
No log	38.0	114	2.1360	0.0108	0.031	0.008	-1.0	0.0222	0.0054	0.0	0.1167	0.4	-1.0	0.45	0.3	0.0108	0.4
No log	39.0	117	2.1677	0.0131	0.0341	0.0125	-1.0	0.0282	0.0036	0.1	0.1167	0.4	-1.0	0.45	0.3	0.0131	0.4
No log	40.0	120	2.2501	0.0092	0.0207	0.0046	-1.0	0.0204	0.0026	0.0833	0.1	0.4333	-1.0	0.525	0.25	0.0092	0.4333
No log	41.0	123	2.1616	0.0128	0.0292	0.0099	-1.0	0.0276	0.0045	0.0	0.2	0.4667	-1.0	0.525	0.35	0.0128	0.4667
No log	42.0	126	2.1395	0.0161	0.0395	0.0122	-1.0	0.0388	0.0055	0.0	0.15	0.4833	-1.0	0.55	0.35	0.0161	0.4833
No log	43.0	129	2.1195	0.0177	0.0417	0.0183	-1.0	0.0398	0.0077	0.0	0.15	0.5	-1.0	0.55	0.4	0.0177	0.5
No log	44.0	132	2.1158	0.0144	0.0368	0.012	-1.0	0.0377	0.0053	0.0	0.15	0.4833	-1.0	0.525	0.4	0.0144	0.4833
No log	45.0	135	2.2567	0.0128	0.0305	0.015	-1.0	0.0382	0.0057	0.0	0.15	0.4333	-1.0	0.45	0.4	0.0128	0.4333
No log	46.0	138	2.2391	0.0143	0.0377	0.0132	-1.0	0.0393	0.0057	0.0	0.2167	0.4167	-1.0	0.45	0.35	0.0143	0.4167
No log	47.0	141	2.2486	0.0121	0.024	0.0134	-1.0	0.0333	0.0063	0.0	0.1	0.4667	-1.0	0.475	0.45	0.0121	0.4667
No log	48.0	144	2.2547	0.0122	0.0325	0.0124	-1.0	0.0258	0.0067	0.0	0.15	0.45	-1.0	0.45	0.45	0.0122	0.45
No log	49.0	147	2.2719	0.0131	0.0257	0.016	-1.0	0.0268	0.0078	0.0	0.1667	0.45	-1.0	0.475	0.4	0.0131	0.45
No log	50.0	150	2.3017	0.0123	0.0256	0.009	-1.0	0.0242	0.0074	0.0	0.1667	0.4167	-1.0	0.45	0.35	0.0123	0.4167

Framework versions

Transformers 4.57.6
Pytorch 2.10.0+xpu
Datasets 4.8.3
Tokenizers 0.22.2

Mrtnl/refr-dfine-xpu

refr-dfine-xpu

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

Mrtnl/
refr-dfine-xpu