pmbaumgartner

# https://docs.mistral.ai/self-deployment/vllm/

export HF_TOKEN=<Huggingface Token>

docker run --gpus all \
    -e HF_TOKEN=$HF_TOKEN -p 8000:8000 \
    ghcr.io/mistralai/mistral-src/vllm:latest \
    --host 0.0.0.0 \
    --model mistralai/Mistral-7B-Instruct-v0.2

pmbaumgartner

def ndict(*args):
    """Return a dictionary with the given keys and their values from globals."""
    g = globals()
    result = {}
    for arg in args:
        if arg in g:
            result[arg] = g[arg]
        else:
            raise KeyError(f"Key '{arg}' not found in globals")
    return result

pmbaumgartner

from typing import TypedDict

DisplacyDepsWords = TypedDict(
    "DisplacyDepsWords", {"text": str, "tag": str, "lemma": Optional[str]}
)
DisplacyDepsArcs = TypedDict(
    "DisplacyDepsArcs", {"start": int, "end": int, "label": str, "dir": str}
)
DisplacyDepsData = TypedDict(
    "DisplacyDepsData",

pmbaumgartner

from spacy import displacy
from spacy.tokens import Doc
from spacy.util import get_lang_class
from pathlib import Path

words = ["The", "quick", "brown", "fox", "jumped", "over", "the", "lazy", "fox"]
heads = [3, 3, 3, 4, 4, 4, 8, 8, 5]
deps = ["det", "amod", "amod", "nsubj", "ROOT", "prep", "pobj", "det", "amod"]
doc = Doc(get_lang_class("en")().vocab, words=words, heads=heads, deps=deps)

pmbaumgartner

RELEASE="15.2.0"
mkdir /tmp/iosevka-font/v$RELEASE
cd /tmp/iosevka-font/v$RELEASE

wget https://github.com/be5invis/Iosevka/releases/download/v$RELEASE/ttf-iosevka-$RELEASE.zip
wget https://github.com/be5invis/Iosevka/releases/download/v$RELEASE/ttf-iosevka-aile-$RELEASE.zip
wget https://github.com/be5invis/Iosevka/releases/download/v$RELEASE/ttf-iosevka-curly-$RELEASE.zip
wget https://github.com/be5invis/Iosevka/releases/download/v$RELEASE/ttf-iosevka-curly-slab-$RELEASE.zip
wget https://github.com/be5invis/Iosevka/releases/download/v$RELEASE/ttf-iosevka-etoile-$RELEASE.zip
wget https://github.com/be5invis/Iosevka/releases/download/v$RELEASE/ttf-iosevka-fixed-$RELEASE.zip

pmbaumgartner

from typing import Optional, Iterable, cast, List
from thinc.api import get_current_ops, Ops
from thinc.types import Ragged, Ints1d

from spacy.pipeline.spancat import Suggester
from spacy.tokens import Doc
from spacy.util import registry


@registry.misc("ngram_digits_suggester.v1")

pmbaumgartner

from spacy.tokenizer import Tokenizer

class CTLTokenizer(Tokenizer):
    # https://stackoverflow.com/a/58718664
    def __call__(self, string) -> spacy.tokens.Doc:
        string = self.clean_string(string)
        doc = super().__call__(string)
        return doc

    def clean_string(self, string: str) -> str:

pmbaumgartner

# SETUP
cuda_version = !nvcc --version | grep -Po '^.*release \K([0-9]+\.[0-9])' | sed 's/\.//g'

# REQUIREMENTS
from pathlib import Path
reqs = f"""
spacy[cuda{cuda_version.s},transformers,lookups]==3.2.1
"""
Path("requirements.txt").write_text(reqs)
!pip install --quiet -r requirements.txt

pmbaumgartner

import random
from typing import List, Union

import shapely.geometry as geo
from tqdm import tqdm

MultiLineStringType = Union[List[geo.LineString], geo.MultiLineString]


def overlap_lines(

pmbaumgartner

import dash
from dash import dcc
from dash import html
from dash.dependencies import Input, Output
import json

clicked = []
external_stylesheets = ["https://codepen.io/chriddyp/pen/bWLwgP.css"]

app = dash.Dash(external_stylesheets=external_stylesheets)