finetune-instance-segmentation-alpha-dent-mask2former-base

slnkvdns/finetune-instance-segmentation-alpha-dent-mask2former-base

finetune-instance-segmentation-alpha-dent-mask2former-base is a image segmentation model from slnkvdns. Use it for the image segmentation task on the model card, and read the license before you ship it in a product. It is set up for transformers. The card lists the license as other.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

4% of all-time downloads

All-time downloads

453

Public

Parameters

68.8M

22.3 GB on disk

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.safetensors275 MB · 100%

Parameter types

How the weights are stored.

F3268.7M · 100%

Base models

facebook/mask2former-swin-small-coco-instance

Model card

From the Hugging Face model README

finetune-instance-segmentation-alpha-dent-mask2former-base

This model is a fine-tuned version of facebook/mask2former-swin-small-coco-instance on the slnkvdns/AlphaDent dataset. It achieves the following results on the evaluation set:

Loss: 24.9112
Map: 0.2883
Map 50: 0.4168
Map 75: 0.2833
Map Small: 0.1263
Map Medium: 0.3228
Map Large: 0.7868
Mar 1: 0.1931
Mar 10: 0.3716
Mar 100: 0.3891
Mar Small: 0.2192
Mar Medium: 0.4104
Mar Large: 0.89
Map Background: 0.9602
Mar 100 Background: 0.9699
Map Abrasion: 0.7008
Mar 100 Abrasion: 0.8541
Map Filling: 0.2203
Mar 100 Filling: 0.3536
Map Crown: 0.7002
Mar 100 Crown: 0.8053
Map Caries class 1: 0.1182
Mar 100 Caries class 1: 0.2741
Map Caries class 2: 0.0324
Mar 100 Caries class 2: 0.1861
Map Caries class 3: 0.0067
Mar 100 Caries class 3: 0.0788
Map Caries class 4: 0.0224
Mar 100 Caries class 4: 0.1
Map Caries class 5: 0.1209
Mar 100 Caries class 5: 0.2487
Map Caries class 6: 0.0008
Mar 100 Caries class 6: 0.02

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: 4e-05
train_batch_size: 4
eval_batch_size: 8
seed: 42
gradient_accumulation_steps: 2
total_train_batch_size: 8
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: constant
num_epochs: 10.0
mixed_precision_training: Native AMP

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 Background	Mar 100 Background	Map Abrasion	Mar 100 Abrasion	Map Filling	Mar 100 Filling	Map Crown	Mar 100 Crown	Map Caries class 1	Mar 100 Caries class 1	Map Caries class 2	Mar 100 Caries class 2	Map Caries class 3	Mar 100 Caries class 3	Map Caries class 4	Mar 100 Caries class 4	Map Caries class 5	Mar 100 Caries class 5	Map Caries class 6	Mar 100 Caries class 6
45.8256	1.0	155	36.5384	0.1558	0.1945	0.17	0.0821	0.0991	0.3994	0.1104	0.202	0.2208	0.1227	0.1912	0.6044	0.8982	0.9133	0.5921	0.8283	0.0279	0.2145	0.0	0.0	0.0051	0.0414	0.0005	0.0333	0.0009	0.0121	0.0	0.0	0.0327	0.1654	0.0	0.0
33.1372	2.0	310	31.3143	0.1952	0.265	0.2015	0.0912	0.1876	0.7196	0.1495	0.3013	0.3234	0.1635	0.3368	0.9014	0.9403	0.9542	0.6287	0.8278	0.1051	0.276	0.1867	0.6368	0.03	0.2328	0.0058	0.0903	0.0012	0.0212	0.0	0.0	0.0546	0.1949	0.0	0.0
29.0343	3.0	465	29.2439	0.2207	0.3169	0.2227	0.0998	0.2434	0.8249	0.1498	0.3302	0.3485	0.1768	0.4005	0.9371	0.9496	0.9614	0.6151	0.8253	0.132	0.2961	0.3515	0.7526	0.0745	0.2362	0.0107	0.1319	0.0008	0.0455	0.0	0.0	0.0732	0.2359	0.0	0.0
26.3806	4.0	620	26.8526	0.2641	0.3736	0.2652	0.1086	0.2931	0.7625	0.1811	0.3354	0.3542	0.1811	0.409	0.9392	0.9569	0.9675	0.6843	0.8511	0.1786	0.3089	0.6277	0.7474	0.0791	0.2552	0.0141	0.1306	0.0018	0.0485	0.0	0.0	0.0987	0.2333	0.0	0.0
24.5939	5.0	775	26.1020	0.2691	0.3766	0.269	0.1092	0.3087	0.7244	0.1783	0.3438	0.3618	0.1841	0.4174	0.9404	0.9605	0.9711	0.6849	0.8469	0.1718	0.3145	0.6582	0.7947	0.0893	0.2448	0.0236	0.1667	0.0013	0.0545	0.0	0.0	0.1019	0.2244	0.0	0.0
22.9142	6.0	930	24.8907	0.2786	0.3964	0.279	0.1147	0.3172	0.8896	0.1774	0.3462	0.3614	0.1851	0.4225	0.9416	0.9669	0.9747	0.7054	0.8599	0.1808	0.3106	0.671	0.7474	0.1146	0.2569	0.0162	0.1611	0.0038	0.0485	0.0	0.0	0.1271	0.2551	0.0	0.0
21.8794	7.0	1085	24.7008	0.2827	0.4036	0.2816	0.1175	0.316	0.7695	0.1805	0.351	0.3688	0.1947	0.4051	0.9229	0.9584	0.9687	0.6885	0.8531	0.1799	0.3123	0.7385	0.8053	0.1052	0.231	0.0247	0.1875	0.0141	0.0788	0.0	0.0	0.1174	0.2513	0.0	0.0
20.8095	8.0	1240	24.1177	0.2921	0.4443	0.295	0.1271	0.3495	0.8291	0.1924	0.3698	0.3832	0.2103	0.4345	0.9408	0.9619	0.9723	0.6979	0.8519	0.214	0.3559	0.7118	0.7947	0.1017	0.2483	0.034	0.1708	0.0059	0.0758	0.0515	0.05	0.1278	0.2526	0.0142	0.06
19.6702	9.0	1395	24.5206	0.2803	0.4177	0.277	0.123	0.3215	0.7562	0.181	0.3631	0.3817	0.2104	0.4352	0.9237	0.9632	0.9711	0.7038	0.8569	0.2199	0.3419	0.6239	0.7579	0.0982	0.2724	0.0263	0.1667	0.005	0.0818	0.0	0.0	0.1349	0.2679	0.0276	0.1
18.6977	10.0	1550	24.9112	0.2883	0.4168	0.2833	0.1263	0.3228	0.7868	0.1931	0.3716	0.3891	0.2192	0.4104	0.89	0.9602	0.9699	0.7008	0.8541	0.2203	0.3536	0.7002	0.8053	0.1182	0.2741	0.0324	0.1861	0.0067	0.0788	0.0224	0.1	0.1209	0.2487	0.0008	0.02

Framework versions

Transformers 5.0.0
Pytorch 2.10.0+cu128
Datasets 4.8.3
Tokenizers 0.22.2

Task: Image Segmentation
Library: transformers
Type: mask2former
License: other
Created: Jun 3, 2026
Updated: Jun 3, 2026