lambdaofgod

import torch
import ot
from sklearn import metrics


roberta = torch.hub.load('pytorch/fairseq', 'roberta.large')
roberta.eval()  # disable dropout (or leave in train mode to finetune)


def get_roberta_features(text):

lambdaofgod

import tqdm


def embed_columns(df, input_columns, embedder, batch_size=8):
    """
        embedder: tf.hub module
    """
    embeddings = {}
    
    for text in tqdm.tqdm(input_columns):

lambdaofgod

# add this if you want to activate a custom tmux session with sourced conda
# requires autojump as j

tmux_guild() {
        j "$1"
        tmux new -d -s guild
        tmux send-keys -t guild "conda activate guild" Enter
        tmux a -t guild
}

lambdaofgod

import rank_bm25 
import pandas as pd
import numpy as np


class SearchEngine:
    
    def __init__(self, df, bm25_cls=rank_bm25.BM25Okapi, text_col='text'):
        self.df = df[~df[text_col].isna()]
        self.bm25 = bm25_cls(self.df[text_col].str.split())

lambdaofgod

from elasticsearch_dsl import Search


def get_everything_from_index(es, index):
    search_results = Search(index=index).using(es).scan()
    for hit in search_results:
        yield hit.to_dict()

lambdaofgod

df = read.table('CH01PR20.txt', header=FALSE)

x = df[,2]
y = df[,1]

n = dim(df)[1]

plot(x, y)

lambdaofgod

import requests
from PIL import Image
from io import BytesIO
import numpy as np


headers = {
    'user-agent':
        'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_11_6) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36'
}

lambdaofgod

import fire
import tqdm
from concurrent.futures import ProcessPoolExecutor
import pandas as pd
import skimage.io
import os
from PIL import Image
import requests
from io import BytesIO
import mlutil.parallel

lambdaofgod

pip install gensim seaborn wordcloud rank_bm25
pip install git+https://github.com/lambdaofgod/mlutil

lambdaofgod