ft-mit-b0-with-scene-parse-150-lora

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

ft-mit-b0-with-scene-parse-150-lora

This model is a fine-tuned version of nvidia/mit-b0 on the scene_parse_150 dataset. It achieves the following results on the evaluation set:

• Loss: 1.7276
• Mean Iou: 0.0003
• Mean Accuracy: 0.0006
• Overall Accuracy: 0.0062
• Accuracy Wall: nan
• Accuracy Building: 0.0380
• Accuracy Sky: 0.0
• Accuracy Floor: 0.0
• Accuracy Tree: 0.0026
• Accuracy Ceiling: 0.0
• Accuracy Road: 0.0
• Accuracy Bed : 0.0143
• Accuracy Windowpane: 0.0
• Accuracy Grass: 0.0
• Accuracy Cabinet: 0.0
• Accuracy Sidewalk: 0.0
• Accuracy Person: 0.0025
• Accuracy Earth: 0.0
• Accuracy Door: 0.0
• Accuracy Table: 0.0
• Accuracy Mountain: 0.0
• Accuracy Plant: 0.0
• Accuracy Curtain: 0.0
• Accuracy Chair: 0.0
• Accuracy Car: 0.0
• Accuracy Water: 0.0
• Accuracy Painting: 0.0
• Accuracy Sofa: 0.0
• Accuracy Shelf: 0.0
• Accuracy House: 0.0
• Accuracy Sea: 0.0
• Accuracy Mirror: 0.0
• Accuracy Rug: 0.0
• Accuracy Field: 0.0
• Accuracy Armchair: nan
• Accuracy Seat: 0.0
• Accuracy Fence: 0.0
• Accuracy Desk: 0.0
• Accuracy Rock: nan
• Accuracy Wardrobe: 0.0
• Accuracy Lamp: nan
• Accuracy Bathtub: 0.0
• Accuracy Railing: 0.0
• Accuracy Cushion: 0.0
• Accuracy Base: 0.0
• Accuracy Box: 0.0
• Accuracy Column: 0.0
• Accuracy Signboard: nan
• Accuracy Chest of drawers: 0.0
• Accuracy Counter: 0.0
• Accuracy Sand: 0.0
• Accuracy Sink: nan
• Accuracy Skyscraper: 0.0
• Accuracy Fireplace: 0.0
• Accuracy Refrigerator: nan
• Accuracy Grandstand: 0.0
• Accuracy Path: 0.0
• Accuracy Stairs: nan
• Accuracy Runway: nan
• Accuracy Case: nan
• Accuracy Pool table: 0.0
• Accuracy Pillow: nan
• Accuracy Screen door: nan
• Accuracy Stairway: 0.0
• Accuracy River: nan
• Accuracy Bridge: nan
• Accuracy Bookcase: 0.0
• Accuracy Blind: 0.0
• Accuracy Coffee table: 0.0
• Accuracy Toilet: 0.0
• Accuracy Flower: 0.0
• Accuracy Book: 0.0
• Accuracy Hill: 0.0
• Accuracy Bench: 0.0
• Accuracy Countertop: 0.0
• Accuracy Stove: nan
• Accuracy Palm: 0.0
• Accuracy Kitchen island: nan
• Accuracy Computer: nan
• Accuracy Swivel chair: nan
• Accuracy Boat: nan
• Accuracy Bar: nan
• Accuracy Arcade machine: 0.0
• Accuracy Hovel: nan
• Accuracy Bus: 0.0
• Accuracy Towel: 0.0
• Accuracy Light: nan
• Accuracy Truck: nan
• Accuracy Tower: 0.0
• Accuracy Chandelier: 0.0
• Accuracy Awning: 0.0
• Accuracy Streetlight: nan
• Accuracy Booth: 0.0
• Accuracy Television receiver: nan
• Accuracy Airplane: nan
• Accuracy Dirt track: nan
• Accuracy Apparel: 0.0
• Accuracy Pole: 0.0
• Accuracy Land: nan
• Accuracy Bannister: nan
• Accuracy Escalator: 0.0
• Accuracy Ottoman: 0.0
• Accuracy Bottle: 0.0
• Accuracy Buffet: 0.0
• Accuracy Poster: nan
• Accuracy Stage: 0.0
• Accuracy Van: nan
• Accuracy Ship: nan
• Accuracy Fountain: nan
• Accuracy Conveyer belt: nan
• Accuracy Canopy: 0.0
• Accuracy Washer: nan
• Accuracy Plaything: nan
• Accuracy Swimming pool: 0.0
• Accuracy Stool: nan
• Accuracy Barrel: 0.0
• Accuracy Basket: nan
• Accuracy Waterfall: nan
• Accuracy Tent: 0.0
• Accuracy Bag: nan
• Accuracy Minibike: nan
• Accuracy Cradle: nan
• Accuracy Oven: nan
• Accuracy Ball: nan
• Accuracy Food: nan
• Accuracy Step: 0.0
• Accuracy Tank: nan
• Accuracy Trade name: 0.0
• Accuracy Microwave: 0.0
• Accuracy Pot: 0.0
• Accuracy Animal: nan
• Accuracy Bicycle: nan
• Accuracy Lake: 0.0
• Accuracy Dishwasher: nan
• Accuracy Screen: nan
• Accuracy Blanket: nan
• Accuracy Sculpture: 0.0
• Accuracy Hood: 0.0
• Accuracy Sconce: 0.0
• Accuracy Vase: 0.0
• Accuracy Traffic light: 0.0
• Accuracy Tray: 0.0
• Accuracy Ashcan: nan
• Accuracy Fan: nan
• Accuracy Pier: nan
• Accuracy Crt screen: 0.0
• Accuracy Plate: nan
• Accuracy Monitor: nan
• Accuracy Bulletin board: 0.0
• Accuracy Shower: 0.0
• Accuracy Radiator: 0.0
• Accuracy Glass: 0.0
• Accuracy Clock: nan
• Accuracy Flag: nan
• Iou Wall: 0.0
• Iou Building: 0.0245
• Iou Sky: 0.0
• Iou Floor: 0.0
• Iou Tree: 0.0006
• Iou Ceiling: 0.0
• Iou Road: 0.0
• Iou Bed : 0.0055
• Iou Windowpane: 0.0
• Iou Grass: 0.0
• Iou Cabinet: 0.0
• Iou Sidewalk: 0.0
• Iou Person: 0.0008
• Iou Earth: 0.0
• Iou Door: 0.0
• Iou Table: 0.0
• Iou Mountain: 0.0
• Iou Plant: 0.0
• Iou Curtain: 0.0
• Iou Chair: 0.0
• Iou Car: 0.0
• Iou Water: 0.0
• Iou Painting: 0.0
• Iou Sofa: 0.0
• Iou Shelf: 0.0
• Iou House: 0.0
• Iou Sea: 0.0
• Iou Mirror: 0.0
• Iou Rug: 0.0
• Iou Field: 0.0
• Iou Armchair: nan
• Iou Seat: 0.0
• Iou Fence: 0.0
• Iou Desk: 0.0
• Iou Rock: nan
• Iou Wardrobe: 0.0
• Iou Lamp: nan
• Iou Bathtub: 0.0
• Iou Railing: 0.0
• Iou Cushion: 0.0
• Iou Base: 0.0
• Iou Box: 0.0
• Iou Column: 0.0
• Iou Signboard: nan
• Iou Chest of drawers: 0.0
• Iou Counter: 0.0
• Iou Sand: 0.0
• Iou Sink: nan
• Iou Skyscraper: 0.0
• Iou Fireplace: 0.0
• Iou Refrigerator: nan
• Iou Grandstand: 0.0
• Iou Path: 0.0
• Iou Stairs: nan
• Iou Runway: nan
• Iou Case: nan
• Iou Pool table: 0.0
• Iou Pillow: nan
• Iou Screen door: nan
• Iou Stairway: 0.0
• Iou River: nan
• Iou Bridge: nan
• Iou Bookcase: 0.0
• Iou Blind: 0.0
• Iou Coffee table: 0.0
• Iou Toilet: 0.0
• Iou Flower: 0.0
• Iou Book: 0.0
• Iou Hill: 0.0
• Iou Bench: 0.0
• Iou Countertop: 0.0
• Iou Stove: nan
• Iou Palm: 0.0
• Iou Kitchen island: nan
• Iou Computer: nan
• Iou Swivel chair: nan
• Iou Boat: nan
• Iou Bar: nan
• Iou Arcade machine: 0.0
• Iou Hovel: nan
• Iou Bus: 0.0
• Iou Towel: 0.0
• Iou Light: nan
• Iou Truck: nan
• Iou Tower: 0.0
• Iou Chandelier: 0.0
• Iou Awning: 0.0
• Iou Streetlight: nan
• Iou Booth: 0.0
• Iou Television receiver: nan
• Iou Airplane: nan
• Iou Dirt track: nan
• Iou Apparel: 0.0
• Iou Pole: 0.0
• Iou Land: nan
• Iou Bannister: nan
• Iou Escalator: 0.0
• Iou Ottoman: 0.0
• Iou Bottle: 0.0
• Iou Buffet: 0.0
• Iou Poster: nan
• Iou Stage: 0.0
• Iou Van: nan
• Iou Ship: nan
• Iou Fountain: nan
• Iou Conveyer belt: nan
• Iou Canopy: 0.0
• Iou Washer: nan
• Iou Plaything: nan
• Iou Swimming pool: 0.0
• Iou Stool: nan
• Iou Barrel: 0.0
• Iou Basket: nan
• Iou Waterfall: nan
• Iou Tent: 0.0
• Iou Bag: nan
• Iou Minibike: nan
• Iou Cradle: nan
• Iou Oven: nan
• Iou Ball: nan
• Iou Food: nan
• Iou Step: 0.0
• Iou Tank: nan
• Iou Trade name: 0.0
• Iou Microwave: 0.0
• Iou Pot: 0.0
• Iou Animal: nan
• Iou Bicycle: nan
• Iou Lake: 0.0
• Iou Dishwasher: nan
• Iou Screen: nan
• Iou Blanket: nan
• Iou Sculpture: 0.0
• Iou Hood: 0.0
• Iou Sconce: 0.0
• Iou Vase: 0.0
• Iou Traffic light: 0.0
• Iou Tray: 0.0
• Iou Ashcan: nan
• Iou Fan: nan
• Iou Pier: nan
• Iou Crt screen: 0.0
• Iou Plate: nan
• Iou Monitor: nan
• Iou Bulletin board: 0.0
• Iou Shower: 0.0
• Iou Radiator: 0.0
• Iou Glass: 0.0
• Iou Clock: nan
• Iou Flag: nan

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: 0.0005
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 5
• mixed_precision_training: Native AMP

Training results

Framework versions

• PEFT 0.7.1
• Transformers 4.36.2
• Pytorch 2.0.0
• Datasets 2.15.0
• Tokenizers 0.15.0