danyaljj

#!/usr/bin/env python
from typing import Iterable
from collections import Counter
import os
import logging
import sys
import json
import click
import datasets
import numpy as np

danyaljj

import json
import argparse
from typing import Optional, Union, Tuple
import torch

torch.manual_seed(0)

from transformers import BertModel, BertTokenizer, PreTrainedModel, BertConfig
from transformers.modeling_outputs import MultipleChoiceModelOutput

danyaljj

import json
import os

all_types_to_idx = {
    'Task': 0,
    'Method': 1,
    'Material': 2,
    'Metric': 3,
    'OtherScientificTerm': 4,
    'Generic': 5

danyaljj

import random
from collections import Counter
from urllib.parse import urlparse
import json
import os
import re

from tqdm import tqdm

urls_counts = {}

danyaljj

from typing import List
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
import scipy
import random

class NormalGammaPrior():
    """"
    Suppose X is distributed according to a normal distribution:  X ~ N(mu, tau^{-1})

danyaljj

import statistics as st

import scipy.stats
import numpy as np

def metric1(scores, row_aggregator, column_aggregator, cell_aggregator):
    row_values = []
    for row_idx, row1 in enumerate(scores):
        diagonal_x = row1[row_idx]
        row_values.append(

danyaljj

show_unpublished_scores: true
datasets:
  blind_labels: danielk/genie_labels
evaluator:
  image: jbragg/genie-evaluator
  input_path: /preds/
  predictions_filename: predictions.json
  label_path: /labels/
  output_path: /results
  arguments:

danyaljj

import statistics as st

def metric1(scores, row_aggregator, column_aggregator, cell_aggregator):
    row_values = []
    for row_idx, row1 in enumerate(scores):
        diagonal_x = row1[row_idx]
        row_values.append(
            column_aggregator(
                [cell_aggregator(diagonal_x, x, abs(col_idx - row_idx)) for col_idx, x in enumerate(row1) if col_idx !=  row_idx]
            )

danyaljj

from datasets import load_dataset

dataset = load_dataset("eli5")

print(dataset)

path = "/Users/danielk/ideaProjects/qoogle-experiments/data"
trainfile = open(path + "/eli5/train.tsv", "w")
testfile = open(path + "/eli5/test.tsv", "w")
devfile = open(path + "/eli5/dev.tsv", "w")

danyaljj

import json
from os import listdir
from os.path import isfile, join

tasks_path = '/Users/danielk/ideaProjects/instructions-demo/app/app/static/tasks/'

categories = {
    'task001_quoref_question_generation': 'Generation',
    'task002_quoref_answer_generation': 'Generation',
    'task003_mctaco_question_generation_event_duration': 'Question Generation',