utils/pos_helpers.py

# POS helpers and constants

POS_MODELS = [
    "gemini-2.0-flash"  # Only allow gemini-2.0-flash for now
    # "gpt-4",
    # "claude-2",
    # "vblagoje/bert-english-uncased-finetuned-pos",
    # "QCRI/bert-base-multilingual-cased-pos-english"
]

DEFAULT_MODEL = "gemini-2.0-flash"

STANDARD_POS_TAGS = [
    "ADJ", "ADP", "ADV", "AUX", "CONJ", "CCONJ", "DET", "INTJ", "NOUN",
    "NUM", "PART", "PRON", "PROPN", "PUNCT", "SCONJ", "SYM", "VERB", "X"
]

POS_TAG_DESCRIPTIONS = {
    "ADJ": "Adjective (big, old, green, interesting)",
    "ADP": "Adposition (in, to, during)",
    "ADV": "Adverb (very, well, there, tomorrow)",
    "AUX": "Auxiliary verb (is, has (done), will (do), should (do))",
    "CCONJ": "Coordinating conjunction (and, or, but)",
    "DET": "Determiner (a, an, the, this, those)",
    "INTJ": "Interjection (oh, hey, oops, hmm)",
    "NOUN": "Noun (dog, cat, man, house, idea)",
    "NUM": "Numeral (one, two, 3, 55, 2019)",
    "PART": "Particle (not, 's, let's)",
    "PRON": "Pronoun (I, you, he, she, it, we, they, me, him, her, us, them)",
    "PROPN": "Proper noun (John, Mary, London, Microsoft)",
    "PUNCT": "Punctuation (.,!?;:)",
    "SCONJ": "Subordinating conjunction (if, because, as, that)",
    "SYM": "Symbol (%, $, §, ©)",
    "VERB": "Verb (run, runs, running, eat, ate, eaten)",
    "X": "Other (foreign words, typos, etc.)"
}

DEFAULT_SELECTED_TAGS = list(POS_TAG_DESCRIPTIONS.keys())