fusion-bench 0.2.9__py3-none-any.whl

fusion_bench/tasks/clip_classification/sun397.py

@@ -0,0 +1,404 @@
+classnames = [
+    "abbey",
+    "airplane cabin",
+    "airport terminal",
+    "alley",
+    "amphitheater",
+    "amusement arcade",
+    "amusement park",
+    "anechoic chamber",
+    "apartment building outdoor",
+    "apse indoor",
+    "aquarium",
+    "aqueduct",
+    "arch",
+    "archive",
+    "arrival gate outdoor",
+    "art gallery",
+    "art school",
+    "art studio",
+    "assembly line",
+    "athletic field outdoor",
+    "atrium public",
+    "attic",
+    "auditorium",
+    "auto factory",
+    "badlands",
+    "badminton court indoor",
+    "baggage claim",
+    "bakery shop",
+    "balcony exterior",
+    "balcony interior",
+    "ball pit",
+    "ballroom",
+    "bamboo forest",
+    "banquet hall",
+    "bar",
+    "barn",
+    "barndoor",
+    "baseball field",
+    "basement",
+    "basilica",
+    "basketball court outdoor",
+    "bathroom",
+    "batters box",
+    "bayou",
+    "bazaar indoor",
+    "bazaar outdoor",
+    "beach",
+    "beauty salon",
+    "bedroom",
+    "berth",
+    "biology laboratory",
+    "bistro indoor",
+    "boardwalk",
+    "boat deck",
+    "boathouse",
+    "bookstore",
+    "booth indoor",
+    "botanical garden",
+    "bow window indoor",
+    "bow window outdoor",
+    "bowling alley",
+    "boxing ring",
+    "brewery indoor",
+    "bridge",
+    "building facade",
+    "bullring",
+    "burial chamber",
+    "bus interior",
+    "butchers shop",
+    "butte",
+    "cabin outdoor",
+    "cafeteria",
+    "campsite",
+    "campus",
+    "canal natural",
+    "canal urban",
+    "candy store",
+    "canyon",
+    "car interior backseat",
+    "car interior frontseat",
+    "carrousel",
+    "casino indoor",
+    "castle",
+    "catacomb",
+    "cathedral indoor",
+    "cathedral outdoor",
+    "cavern indoor",
+    "cemetery",
+    "chalet",
+    "cheese factory",
+    "chemistry lab",
+    "chicken coop indoor",
+    "chicken coop outdoor",
+    "childs room",
+    "church indoor",
+    "church outdoor",
+    "classroom",
+    "clean room",
+    "cliff",
+    "cloister indoor",
+    "closet",
+    "clothing store",
+    "coast",
+    "cockpit",
+    "coffee shop",
+    "computer room",
+    "conference center",
+    "conference room",
+    "construction site",
+    "control room",
+    "control tower outdoor",
+    "corn field",
+    "corral",
+    "corridor",
+    "cottage garden",
+    "courthouse",
+    "courtroom",
+    "courtyard",
+    "covered bridge exterior",
+    "creek",
+    "crevasse",
+    "crosswalk",
+    "cubicle office",
+    "dam",
+    "delicatessen",
+    "dentists office",
+    "desert sand",
+    "desert vegetation",
+    "diner indoor",
+    "diner outdoor",
+    "dinette home",
+    "dinette vehicle",
+    "dining car",
+    "dining room",
+    "discotheque",
+    "dock",
+    "doorway outdoor",
+    "dorm room",
+    "driveway",
+    "driving range outdoor",
+    "drugstore",
+    "electrical substation",
+    "elevator door",
+    "elevator interior",
+    "elevator shaft",
+    "engine room",
+    "escalator indoor",
+    "excavation",
+    "factory indoor",
+    "fairway",
+    "fastfood restaurant",
+    "field cultivated",
+    "field wild",
+    "fire escape",
+    "fire station",
+    "firing range indoor",
+    "fishpond",
+    "florist shop indoor",
+    "food court",
+    "forest broadleaf",
+    "forest needleleaf",
+    "forest path",
+    "forest road",
+    "formal garden",
+    "fountain",
+    "galley",
+    "game room",
+    "garage indoor",
+    "garbage dump",
+    "gas station",
+    "gazebo exterior",
+    "general store indoor",
+    "general store outdoor",
+    "gift shop",
+    "golf course",
+    "greenhouse indoor",
+    "greenhouse outdoor",
+    "gymnasium indoor",
+    "hangar indoor",
+    "hangar outdoor",
+    "harbor",
+    "hayfield",
+    "heliport",
+    "herb garden",
+    "highway",
+    "hill",
+    "home office",
+    "hospital",
+    "hospital room",
+    "hot spring",
+    "hot tub outdoor",
+    "hotel outdoor",
+    "hotel room",
+    "house",
+    "hunting lodge outdoor",
+    "ice cream parlor",
+    "ice floe",
+    "ice shelf",
+    "ice skating rink indoor",
+    "ice skating rink outdoor",
+    "iceberg",
+    "igloo",
+    "industrial area",
+    "inn outdoor",
+    "islet",
+    "jacuzzi indoor",
+    "jail cell",
+    "jail indoor",
+    "jewelry shop",
+    "kasbah",
+    "kennel indoor",
+    "kennel outdoor",
+    "kindergarden classroom",
+    "kitchen",
+    "kitchenette",
+    "labyrinth outdoor",
+    "lake natural",
+    "landfill",
+    "landing deck",
+    "laundromat",
+    "lecture room",
+    "library indoor",
+    "library outdoor",
+    "lido deck outdoor",
+    "lift bridge",
+    "lighthouse",
+    "limousine interior",
+    "living room",
+    "lobby",
+    "lock chamber",
+    "locker room",
+    "mansion",
+    "manufactured home",
+    "market indoor",
+    "market outdoor",
+    "marsh",
+    "martial arts gym",
+    "mausoleum",
+    "medina",
+    "moat water",
+    "monastery outdoor",
+    "mosque indoor",
+    "mosque outdoor",
+    "motel",
+    "mountain",
+    "mountain snowy",
+    "movie theater indoor",
+    "museum indoor",
+    "music store",
+    "music studio",
+    "nuclear power plant outdoor",
+    "nursery",
+    "oast house",
+    "observatory outdoor",
+    "ocean",
+    "office",
+    "office building",
+    "oil refinery outdoor",
+    "oilrig",
+    "operating room",
+    "orchard",
+    "outhouse outdoor",
+    "pagoda",
+    "palace",
+    "pantry",
+    "park",
+    "parking garage indoor",
+    "parking garage outdoor",
+    "parking lot",
+    "parlor",
+    "pasture",
+    "patio",
+    "pavilion",
+    "pharmacy",
+    "phone booth",
+    "physics laboratory",
+    "picnic area",
+    "pilothouse indoor",
+    "planetarium outdoor",
+    "playground",
+    "playroom",
+    "plaza",
+    "podium indoor",
+    "podium outdoor",
+    "pond",
+    "poolroom establishment",
+    "poolroom home",
+    "power plant outdoor",
+    "promenade deck",
+    "pub indoor",
+    "pulpit",
+    "putting green",
+    "racecourse",
+    "raceway",
+    "raft",
+    "railroad track",
+    "rainforest",
+    "reception",
+    "recreation room",
+    "residential neighborhood",
+    "restaurant",
+    "restaurant kitchen",
+    "restaurant patio",
+    "rice paddy",
+    "riding arena",
+    "river",
+    "rock arch",
+    "rope bridge",
+    "ruin",
+    "runway",
+    "sandbar",
+    "sandbox",
+    "sauna",
+    "schoolhouse",
+    "sea cliff",
+    "server room",
+    "shed",
+    "shoe shop",
+    "shopfront",
+    "shopping mall indoor",
+    "shower",
+    "skatepark",
+    "ski lodge",
+    "ski resort",
+    "ski slope",
+    "sky",
+    "skyscraper",
+    "slum",
+    "snowfield",
+    "squash court",
+    "stable",
+    "stadium baseball",
+    "stadium football",
+    "stage indoor",
+    "staircase",
+    "street",
+    "subway interior",
+    "subway station platform",
+    "supermarket",
+    "sushi bar",
+    "swamp",
+    "swimming pool indoor",
+    "swimming pool outdoor",
+    "synagogue indoor",
+    "synagogue outdoor",
+    "television studio",
+    "temple east asia",
+    "temple south asia",
+    "tennis court indoor",
+    "tennis court outdoor",
+    "tent outdoor",
+    "theater indoor procenium",
+    "theater indoor seats",
+    "thriftshop",
+    "throne room",
+    "ticket booth",
+    "toll plaza",
+    "topiary garden",
+    "tower",
+    "toyshop",
+    "track outdoor",
+    "train railway",
+    "train station platform",
+    "tree farm",
+    "tree house",
+    "trench",
+    "underwater coral reef",
+    "utility room",
+    "valley",
+    "van interior",
+    "vegetable garden",
+    "veranda",
+    "veterinarians office",
+    "viaduct",
+    "videostore",
+    "village",
+    "vineyard",
+    "volcano",
+    "volleyball court indoor",
+    "volleyball court outdoor",
+    "waiting room",
+    "warehouse indoor",
+    "water tower",
+    "waterfall block",
+    "waterfall fan",
+    "waterfall plunge",
+    "watering hole",
+    "wave",
+    "wet bar",
+    "wheat field",
+    "wind farm",
+    "windmill",
+    "wine cellar barrel storage",
+    "wine cellar bottle storage",
+    "wrestling ring indoor",
+    "yard",
+    "youth hostel",
+]
+
+templates = [
+    lambda c: f"a photo of a {c}.",
+    lambda c: f"a photo of the {c}.",
+]

fusion_bench/tasks/clip_classification/svhn.py

@@ -0,0 +1,5 @@
+classnames = ["0", "1", "2", "3", "4", "5", "6", "7", "8", "9"]
+
+templates = [
+    lambda c: f'a photo of the number: "{c}".',
+]

fusion_bench/tasks/clip_classification/tiny_imagenet.py

@@ -0,0 +1,208 @@
+from .imagenet import templates  # noqa: F401
+
+classnames_full = [
+    "goldfish, Carassius auratus",
+    "European fire salamander, Salamandra salamandra",
+    "bullfrog, Rana catesbeiana",
+    "tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui",
+    "American alligator, Alligator mississipiensis",
+    "boa constrictor, Constrictor constrictor",
+    "trilobite",
+    "scorpion",
+    "black widow, Latrodectus mactans",
+    "tarantula",
+    "centipede",
+    "koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus",
+    "jellyfish",
+    "brain coral",
+    "snail",
+    "sea slug, nudibranch",
+    "American lobster, Northern lobster, Maine lobster, Homarus americanus",
+    "spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish",
+    "black stork, Ciconia nigra",
+    "king penguin, Aptenodytes patagonica",
+    "albatross, mollymawk",
+    "dugong, Dugong dugon",
+    "Yorkshire terrier",
+    "golden retriever",
+    "Labrador retriever",
+    "German shepherd, German shepherd dog, German police dog, alsatian",
+    "standard poodle",
+    "tabby, tabby cat",
+    "Persian cat",
+    "Egyptian cat",
+    "cougar, puma, catamount, mountain lion, painter, panther, Felis concolor",
+    "lion, king of beasts, Panthera leo",
+    "brown bear, bruin, Ursus arctos",
+    "ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle",
+    "grasshopper, hopper",
+    "walking stick, walkingstick, stick insect",
+    "cockroach, roach",
+    "mantis, mantid",
+    "dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk",
+    "monarch, monarch butterfly, milkweed butterfly, Danaus plexippus",
+    "sulphur butterfly, sulfur butterfly",
+    "sea cucumber, holothurian",
+    "guinea pig, Cavia cobaya",
+    "hog, pig, grunter, squealer, Sus scrofa",
+    "ox",
+    "bison",
+    "bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis",
+    "gazelle",
+    "Arabian camel, dromedary, Camelus dromedarius",
+    "orangutan, orang, orangutang, Pongo pygmaeus",
+    "chimpanzee, chimp, Pan troglodytes",
+    "baboon",
+    "African elephant, Loxodonta africana",
+    "lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens",
+    "abacus",
+    "academic gown, academic robe, judge's robe",
+    "altar",
+    "backpack, back pack, knapsack, packsack, rucksack, haversack",
+    "bannister, banister, balustrade, balusters, handrail",
+    "barbershop",
+    "barn",
+    "barrel, cask",
+    "basketball",
+    "bathtub, bathing tub, bath, tub",
+    "beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon",
+    "beacon, lighthouse, beacon light, pharos",
+    "beaker",
+    "beer bottle",
+    "bikini, two-piece",
+    "binoculars, field glasses, opera glasses",
+    "birdhouse",
+    "bow tie, bow-tie, bowtie",
+    "brass, memorial tablet, plaque",
+    "bucket, pail",
+    "bullet train, bullet",
+    "butcher shop, meat market",
+    "candle, taper, wax light",
+    "cannon",
+    "cardigan",
+    "cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM",
+    "CD player",
+    "chest",
+    "Christmas stocking",
+    "cliff dwelling",
+    "computer keyboard, keypad",
+    "confectionery, confectionary, candy store",
+    "convertible",
+    "crane",
+    "dam, dike, dyke",
+    "desk",
+    "dining table, board",
+    "dumbbell",
+    "flagpole, flagstaff",
+    "fly",
+    "fountain",
+    "freight car",
+    "frying pan, frypan, skillet",
+    "fur coat",
+    "gasmask, respirator, gas helmet",
+    "go-kart",
+    "gondola",
+    "hourglass",
+    "iPod",
+    "jinrikisha, ricksha, rickshaw",
+    "kimono",
+    "lampshade, lamp shade",
+    "lawn mower, mower",
+    "lifeboat",
+    "limousine, limo",
+    "magnetic compass",
+    "maypole",
+    "military uniform",
+    "miniskirt, mini",
+    "moving van",
+    "neck brace",
+    "obelisk",
+    "oboe, hautboy, hautbois",
+    "organ, pipe organ",
+    "parking meter",
+    "pay-phone, pay-station",
+    "picket fence, paling",
+    "pill bottle",
+    "plunger, plumber's helper",
+    "police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria",
+    "poncho",
+    "pop bottle, soda bottle",
+    "potter's wheel",
+    "projectile, missile",
+    "punching bag, punch bag, punching ball, punchball",
+    "refrigerator, icebox",
+    "remote control, remote",
+    "rocking chair, rocker",
+    "rugby ball",
+    "sandal",
+    "school bus",
+    "scoreboard",
+    "sewing machine",
+    "snorkel",
+    "sock",
+    "sombrero",
+    "space heater",
+    "spider web, spider's web",
+    "sports car, sport car",
+    "steel arch bridge",
+    "stopwatch, stop watch",
+    "sunglasses, dark glasses, shades",
+    "suspension bridge",
+    "swimming trunks, bathing trunks",
+    "syringe",
+    "teapot",
+    "teddy, teddy bear",
+    "thatch, thatched roof",
+    "torch",
+    "tractor",
+    "triumphal arch",
+    "trolleybus, trolley coach, trackless trolley",
+    "turnstile",
+    "umbrella",
+    "vestment",
+    "viaduct",
+    "volleyball",
+    "water jug",
+    "water tower",
+    "wok",
+    "wooden spoon",
+    "comic book",
+    "reel",
+    "guacamole",
+    "ice cream, icecream",
+    "ice lolly, lolly, lollipop, popsicle",
+    "goose",
+    "drumstick",
+    "plate",
+    "pretzel",
+    "mashed potato",
+    "cauliflower",
+    "bell pepper",
+    "lemon",
+    "banana",
+    "pomegranate",
+    "meat loaf, meatloaf",
+    "pizza, pizza pie",
+    "potpie",
+    "espresso",
+    "bee",
+    "apron",
+    "pole",
+    "Chihuahua",
+    "alp",
+    "cliff, drop, drop-off",
+    "coral reef",
+    "lakeside, lakeshore",
+    "seashore, coast, seacoast, sea-coast",
+    "acorn",
+    "broom",
+    "mushroom",
+    "nail",
+    "chain",
+    "slug",
+    "orange",
+]
+
+classnames = [n.split(",")[0] for n in classnames_full]
+
+assert len(classnames) == 200, "Tiny ImageNet classnames should have 200 classes"

fusion_bench/tasks/flan_t5_text_generation/datasets_preprocess.py

@@ -0,0 +1,71 @@
+"""
+This scripts preprocess any NLP dataset into a text-to-text format.
+"""
+
+import json
+import os
+from pathlib import Path
+from typing import Any, Callable, Dict, Union  # noqa: F401
+
+from transformers import AutoTokenizer
+
+
+def preprocess(
+    tokenizer: AutoTokenizer,
+    input_text: str,
+    target_text: str,
+    tokenizer_kwawgs: Dict[str, Any] = None,
+):
+    """
+    standard preprocess function for dataset.
+    Preprocesses input and target text data using a tokenizer object and returns a dictionary of model inputs.
+
+    Args:
+        tokenizer: An instance of a tokenizer class used to preprocess text data.
+        input_text (str): A string containing the input text data to be tokenized.
+        target_text (str, optional): A string containing the target text data to be tokenized. If None, no target data is returned.
+
+    Returns:
+        A dictionary of model inputs containing the tokenized input and output data along with the modified labels tensor.
+    """
+    if tokenizer_kwawgs is None:
+        tokenizer_kwawgs = {}
+    model_inputs = tokenizer(input_text, **tokenizer_kwawgs)
+    if target_text is not None:
+        labels = tokenizer(target_text, **tokenizer_kwawgs)
+        labels = labels["input_ids"]
+        labels[labels == tokenizer.pad_token_id] = -100
+        model_inputs["labels"] = labels
+    return model_inputs
+
+
+class DatasetPreprocessor:
+    def __init__(
+        self,
+        tokenizer: AutoTokenizer,
+        tokenizer_kwargs: Dict[str, Any] = None,
+        template: Union[str, Path, Dict] = None,
+    ):
+        """
+        Initializes an instance of the datasets_preprocess class with a tokenizer object.
+
+        Args:
+            tokenizer: An instance of a tokenizer class used to preprocess text data.
+        """
+        super().__init__()
+        self.tokenizer = tokenizer
+        self.tokenizer_kwargs = tokenizer_kwargs
+        if template is not None:
+            if isinstance(template, str):
+                template = template
+                assert os.path.exists(
+                    template
+                ), f"Template file not found at {template}"
+                with open(template, "r") as f:
+                    self.template = json.load(f)
+            elif isinstance(template, dict):
+                self.template = template
+            else:
+                raise ValueError(
+                    "Template must be a path to a json file or a dictionary"
+                )