vhxs

# modified from this:
# https://github.com/iaciac/py-draw-simplicial-complex/blob/master/Draw%202d%20simplicial%20complex.ipynb
from copy import copy

import networkx as nx
import itertools

import matplotlib.pyplot as plt
import numpy as np

vhxs

hey look here's me writing something

vhxs

import numpy as np
from random import sample


# The Jaccard distance is truly a metric: https://arxiv.org/pdf/1612.02696.pdf
def jaccard_distance(set1: set, set2: set):
    union_size = len(set1.union(set2))
    intersection_size = len(set1.intersection(set2))

    return 1 - (intersection_size / union_size)

vhxs

import numpy as np


# Give me some random vectors
def random_vectors_generator(dimension=10, num_vectors=10):
    for _ in range(num_vectors):
        yield np.random.rand(dimension)


# Give me a random distance matrix

vhxs

import torch
import matplotlib.pyplot as plt


def generate_basis(num_channels, height, width):
    zero_tensor = torch.zeros((num_channels, height, width))
    for i in range(num_channels):
        for j in range(height):
            for k in range(width):
                basis_element = torch.clone(zero_tensor)

vhxs

import torch
import torchvision
import torchvision.transforms as transforms
import torch.optim as optim
import torch.nn as nn
import umap
import matplotlib.pyplot as plt

# Check whether GPU is available and define it as a device
is_gpu_available = torch.cuda.is_available()

vhxs

import torch
import torchvision
import torchvision.transforms as transforms
import torch.optim as optim
import torch.nn as nn
import umap
import matplotlib.pyplot as plt
import json

trinity_template = {

vhxs

import numpy as np
import torch
from datetime import datetime


if __name__ == '__main__':
    num_dims = 3
    dim = 1024

    for _ in range(10):

vhxs

from langchain.document_loaders import MastodonTootsLoader
from langchain.indexes import VectorstoreIndexCreator
from langchain.vectorstores.utils import filter_complex_metadata

loader = MastodonTootsLoader(mastodon_accounts=["@[email protected]"], number_toots=300)
documents = loader.load()
documents = filter_complex_metadata(documents)
index = VectorstoreIndexCreator().from_documents(documents)

print(index.query("What kinds of interests does this user have?"))

vhxs

from langchain.document_loaders import OnlinePDFLoader
from langchain.indexes import VectorstoreIndexCreator

loader = OnlinePDFLoader("https://arxiv.org/pdf/1606.06353.pdf")
documents = loader.load()
index = VectorstoreIndexCreator().from_documents(documents)

print(index.query("What are the main results of this paper?"))