mask2former-swin-base-apple-dms-run1

This model is a fine-tuned version of facebook/mask2former-swin-base-ade-semantic on the AllanK24/apple-dms-materials dataset. It achieves the following results on the evaluation set:

• Mean Iou: 0.0129
• Mean Accuracy: 0.0190
• Overall Accuracy: 0.2006
• Iou Animal Skin: 0.0
• Accuracy Animal Skin: 0.0
• Iou Bone Teeth Horn: 0.0
• Accuracy Bone Teeth Horn: 0.0
• Iou Brickwork: 0.0
• Accuracy Brickwork: 0.0
• Iou Cardboard: 0.0
• Accuracy Cardboard: 0.0
• Iou Carpet Rug: 0.0000
• Accuracy Carpet Rug: 0.0000
• Iou Ceiling Tile: 0.0
• Accuracy Ceiling Tile: 0.0
• Iou Ceramic: 0.0
• Accuracy Ceramic: 0.0
• Iou Chalkboard Blackboard: 0.0
• Accuracy Chalkboard Blackboard: 0.0
• Iou Clutter: 0.0
• Accuracy Clutter: 0.0
• Iou Concrete: 0.0
• Accuracy Concrete: 0.0
• Iou Cork Corkboard: 0.0
• Accuracy Cork Corkboard: 0.0
• Iou Engineered Stone: 0.0
• Accuracy Engineered Stone: 0.0
• Iou Fabric Cloth: 0.0045
• Accuracy Fabric Cloth: 0.0045
• Iou Fiberglass Wool: 0.0
• Accuracy Fiberglass Wool: 0.0
• Iou Fire: 0.0
• Accuracy Fire: 0.0
• Iou Foliage: 0.0
• Accuracy Foliage: 0.0
• Iou Food: 0.0
• Accuracy Food: 0.0
• Iou Fur: 0.0
• Accuracy Fur: 0.0
• Iou Gemstone Quartz: 0.0
• Accuracy Gemstone Quartz: 0.0
• Iou Glass: 0.0593
• Accuracy Glass: 0.0676
• Iou Hair: 0.0685
• Accuracy Hair: 0.0885
• Iou Ice: 0.0
• Accuracy Ice: 0.0
• Iou Leather: 0.0
• Accuracy Leather: 0.0
• Iou Liquid Non-water: 0.0
• Accuracy Liquid Non-water: 0.0
• Iou Metal: 0.0036
• Accuracy Metal: 0.0038
• Iou Mirror: 0.0
• Accuracy Mirror: 0.0
• Iou Paint Plaster Enamel: 0.4088
• Accuracy Paint Plaster Enamel: 0.6439
• Iou Paper: 0.0141
• Accuracy Paper: 0.0153
• Iou Pearl: 0.0
• Accuracy Pearl: 0.0
• Iou Photograph Painting: 0.0
• Accuracy Photograph Painting: 0.0
• Iou Plastic Clear: 0.0
• Accuracy Plastic Clear: 0.0
• Iou Plastic Non-clear: 0.0000
• Accuracy Plastic Non-clear: 0.0000
• Iou Rubber Latex: 0.0
• Accuracy Rubber Latex: 0.0
• Iou Sand: 0.0
• Accuracy Sand: 0.0
• Iou Skin Lips: 0.1127
• Accuracy Skin Lips: 0.1643
• Iou Sky: 0.0
• Accuracy Sky: 0.0
• Iou Snow: 0.0
• Accuracy Snow: 0.0
• Iou Soap: 0.0
• Accuracy Soap: 0.0
• Iou Soil Mud: 0.0
• Accuracy Soil Mud: 0.0
• Iou Sponge: 0.0
• Accuracy Sponge: 0.0
• Iou Stone Natural: 0.0
• Accuracy Stone Natural: 0.0
• Iou Stone Polished: 0.0
• Accuracy Stone Polished: 0.0
• Iou Styrofoam: 0.0
• Accuracy Styrofoam: 0.0
• Iou Tile: 0.0
• Accuracy Tile: 0.0
• Iou Wallpaper: 0.0
• Accuracy Wallpaper: 0.0
• Iou Water: 0.0
• Accuracy Water: 0.0
• Iou Wax: 0.0
• Accuracy Wax: 0.0
• Iou Whiteboard: 0.0
• Accuracy Whiteboard: 0.0
• Iou Wicker: 0.0
• Accuracy Wicker: 0.0
• Iou Wood: 0.0001
• Accuracy Wood: 0.0001
• Iou Wood Tree: 0.0
• Accuracy Wood Tree: 0.0
• Iou Asphalt: 0.0
• Accuracy Asphalt: 0.0
• Loss: 411.6706
• Eval Runtime: 41.6536
• Eval Samples Per Second: 28.425
• Eval Steps Per Second: 14.212

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: 0.001
• train_batch_size: 32
• eval_batch_size: 16
• seed: 42
• distributed_type: multi-GPU
• num_devices: 8
• total_train_batch_size: 256
• total_eval_batch_size: 128
• 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: 1

Training results

Training Loss	Epoch	Step	Mean Iou	Mean Accuracy	Overall Accuracy	Iou Animal Skin	Accuracy Animal Skin	Iou Bone Teeth Horn	Accuracy Bone Teeth Horn	Iou Brickwork	Accuracy Brickwork	Iou Cardboard	Accuracy Cardboard	Iou Carpet Rug	Accuracy Carpet Rug	Iou Ceiling Tile	Accuracy Ceiling Tile	Iou Ceramic	Accuracy Ceramic	Iou Chalkboard Blackboard	Accuracy Chalkboard Blackboard	Iou Clutter	Accuracy Clutter	Iou Concrete	Accuracy Concrete	Iou Cork Corkboard	Accuracy Cork Corkboard	Iou Engineered Stone	Accuracy Engineered Stone	Iou Fabric Cloth	Accuracy Fabric Cloth	Iou Fiberglass Wool	Accuracy Fiberglass Wool	Iou Fire	Accuracy Fire	Iou Foliage	Accuracy Foliage	Iou Food	Accuracy Food	Iou Fur	Accuracy Fur	Iou Gemstone Quartz	Accuracy Gemstone Quartz	Iou Glass	Accuracy Glass	Iou Hair	Accuracy Hair	Iou Ice	Accuracy Ice	Iou Leather	Accuracy Leather	Iou Liquid Non-water	Accuracy Liquid Non-water	Iou Metal	Accuracy Metal	Iou Mirror	Accuracy Mirror	Iou Paint Plaster Enamel	Accuracy Paint Plaster Enamel	Iou Paper	Accuracy Paper	Iou Pearl	Accuracy Pearl	Iou Photograph Painting	Accuracy Photograph Painting	Iou Plastic Clear	Accuracy Plastic Clear	Iou Plastic Non-clear	Accuracy Plastic Non-clear	Iou Rubber Latex	Accuracy Rubber Latex	Iou Sand	Accuracy Sand	Iou Skin Lips	Accuracy Skin Lips	Iou Sky	Accuracy Sky	Iou Snow	Accuracy Snow	Iou Soap	Accuracy Soap	Iou Soil Mud	Accuracy Soil Mud	Iou Sponge	Accuracy Sponge	Iou Stone Natural	Accuracy Stone Natural	Iou Stone Polished	Accuracy Stone Polished	Iou Styrofoam	Accuracy Styrofoam	Iou Tile	Accuracy Tile	Iou Wallpaper	Accuracy Wallpaper	Iou Water	Accuracy Water	Iou Wax	Accuracy Wax	Iou Whiteboard	Accuracy Whiteboard	Iou Wicker	Accuracy Wicker	Iou Wood	Accuracy Wood	Iou Wood Tree	Accuracy Wood Tree	Iou Asphalt	Accuracy Asphalt	Validation Loss	Eval Runtime	Eval Samples Per Second	Eval Steps Per Second
475.7108	0.5682	50	0.0129	0.0190	0.2006	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0000	0.0000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0045	0.0045	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0593	0.0676	0.0685	0.0885	0.0	0.0	0.0	0.0	0.0	0.0	0.0036	0.0038	0.0	0.0	0.4088	0.6439	0.0141	0.0153	0.0	0.0	0.0	0.0	0.0	0.0	0.0000	0.0000	0.0	0.0	0.0	0.0	0.1127	0.1643	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0001	0.0001	0.0	0.0	0.0	0.0	411.6706	41.6536	28.425	14.212
475.7108	0.5682	50	0.0129	0.0190	0.2006	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0000	0.0000	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0045	0.0045	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0593	0.0676	0.0685	0.0885	0.0	0.0	0.0	0.0	0.0	0.0	0.0036	0.0038	0.0	0.0	0.4088	0.6439	0.0141	0.0153	0.0	0.0	0.0	0.0	0.0	0.0	0.0000	0.0000	0.0	0.0	0.0	0.0	0.1127	0.1643	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0	0.0001	0.0001	0.0	0.0	0.0	0.0	411.6706	41.6536	28.425	14.212

Framework versions

• Transformers 5.0.0
• Pytorch 2.9.1+cu128
• Datasets 4.5.0
• Tokenizers 0.22.2