Spaces:
Runtime error
Runtime error
Delete analysis.py
Browse files- analysis.py +0 -78
analysis.py
DELETED
@@ -1,78 +0,0 @@
|
|
1 |
-
import requests
|
2 |
-
import httpx
|
3 |
-
import torch
|
4 |
-
import re
|
5 |
-
from bs4 import BeautifulSoup
|
6 |
-
import numpy as np
|
7 |
-
from transformers import AutoTokenizer, AutoModelForSequenceClassification
|
8 |
-
import asyncio
|
9 |
-
from scipy.special import softmax
|
10 |
-
from evaluate import load
|
11 |
-
from datetime import date
|
12 |
-
import nltk
|
13 |
-
import fitz
|
14 |
-
from transformers import GPT2LMHeadModel, GPT2TokenizerFast
|
15 |
-
import nltk, spacy, subprocess, torch
|
16 |
-
import plotly.graph_objects as go
|
17 |
-
import torch.nn.functional as F
|
18 |
-
import nltk
|
19 |
-
from unidecode import unidecode
|
20 |
-
import time
|
21 |
-
import yaml
|
22 |
-
import nltk
|
23 |
-
import os
|
24 |
-
from explainability import *
|
25 |
-
from dotenv import load_dotenv
|
26 |
-
nltk.download('punkt')
|
27 |
-
nltk.download('stopwords')
|
28 |
-
load_dotenv()
|
29 |
-
device = "cuda" if torch.cuda.is_available() else "cpu"
|
30 |
-
readability_model_id = os.getenv('READABILITY_MODEL_ID')
|
31 |
-
gpt2_model = GPT2LMHeadModel.from_pretrained(readability_model_id).to(device)
|
32 |
-
gpt2_tokenizer = GPT2TokenizerFast.from_pretrained(readability_model_id)
|
33 |
-
|
34 |
-
def depth_analysis(input_text):
|
35 |
-
processed_words = preprocess_text1(input_text)
|
36 |
-
ttr_value = vocabulary_richness_ttr(processed_words)
|
37 |
-
gunning_fog = calculate_gunning_fog(input_text)
|
38 |
-
gunning_fog_norm = normalize(gunning_fog, min_value=0, max_value=20)
|
39 |
-
words, sentences = preprocess_text2(input_text)
|
40 |
-
average_sentence_length = calculate_average_sentence_length(sentences)
|
41 |
-
average_word_length = calculate_average_word_length(words)
|
42 |
-
average_sentence_length_norm = normalize(average_sentence_length, min_value=0, max_value=40)
|
43 |
-
average_word_length_norm = normalize(average_word_length, min_value=0, max_value=8)
|
44 |
-
average_tree_depth = calculate_syntactic_tree_depth(nlp, input_text)
|
45 |
-
average_tree_depth_norm = normalize(average_tree_depth, min_value=0, max_value=10)
|
46 |
-
perplexity = calculate_perplexity(input_text, gpt2_model, gpt2_tokenizer, device)
|
47 |
-
perplexity_norm = normalize(perplexity, min_value=0, max_value=30)
|
48 |
-
|
49 |
-
features = {
|
50 |
-
"readability": gunning_fog_norm,
|
51 |
-
"syntactic tree depth": average_tree_depth_norm,
|
52 |
-
"vocabulary richness": ttr_value,
|
53 |
-
"perplexity": perplexity_norm,
|
54 |
-
"average sentence length": average_sentence_length_norm,
|
55 |
-
"average word length": average_word_length_norm,
|
56 |
-
}
|
57 |
-
fig = go.Figure()
|
58 |
-
fig.add_trace(go.Scatterpolar(
|
59 |
-
r=list(features.values()),
|
60 |
-
theta=list(features.keys()),
|
61 |
-
fill='toself',
|
62 |
-
name='Radar Plot'
|
63 |
-
))
|
64 |
-
fig.update_layout(
|
65 |
-
polar=dict(
|
66 |
-
radialaxis=dict(
|
67 |
-
visible=True,
|
68 |
-
range=[0, 100],
|
69 |
-
)),
|
70 |
-
showlegend=False,
|
71 |
-
margin=dict(
|
72 |
-
l=10,
|
73 |
-
r=20,
|
74 |
-
b=10,
|
75 |
-
t=10,
|
76 |
-
),
|
77 |
-
)
|
78 |
-
return fig
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|