Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -20,7 +20,9 @@ from tenacity import retry
|
|
| 20 |
from tqdm import tqdm
|
| 21 |
import tiktoken
|
| 22 |
import scipy.stats
|
|
|
|
| 23 |
import torch
|
|
|
|
| 24 |
from transformers import GPT2LMHeadModel
|
| 25 |
import tiktoken
|
| 26 |
import seaborn as sns
|
|
@@ -28,6 +30,24 @@ from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
|
|
| 28 |
# from colorama import Fore, Style
|
| 29 |
import openai # for OpenAI API calls
|
| 30 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 31 |
######################################
|
| 32 |
import streamlit as st
|
| 33 |
def colorize_tokens(token_data, sentence):
|
|
@@ -50,16 +70,16 @@ def colorize_tokens(token_data, sentence):
|
|
| 50 |
|
| 51 |
# Define colors for the tags
|
| 52 |
tag_colors = {
|
| 53 |
-
"ADJP": "#
|
| 54 |
-
"ADVP": "#
|
| 55 |
-
"CONJP": "#
|
| 56 |
-
"INTJ": "#
|
| 57 |
-
"LST": "#
|
| 58 |
-
"NP": "#
|
| 59 |
-
"PP": "#
|
| 60 |
-
"PRT": "#
|
| 61 |
-
"SBAR": "#
|
| 62 |
-
"VP": "#
|
| 63 |
}
|
| 64 |
##################
|
| 65 |
|
|
|
|
| 20 |
from tqdm import tqdm
|
| 21 |
import tiktoken
|
| 22 |
import scipy.stats
|
| 23 |
+
import inseq
|
| 24 |
import torch
|
| 25 |
+
from transformers import AutoModelForCausalLM
|
| 26 |
from transformers import GPT2LMHeadModel
|
| 27 |
import tiktoken
|
| 28 |
import seaborn as sns
|
|
|
|
| 30 |
# from colorama import Fore, Style
|
| 31 |
import openai # for OpenAI API calls
|
| 32 |
|
| 33 |
+
######################################
|
| 34 |
+
def find_indices(arr, target):
|
| 35 |
+
indices = []
|
| 36 |
+
start_index = None
|
| 37 |
+
|
| 38 |
+
for i, element in enumerate(arr):
|
| 39 |
+
if target in element:
|
| 40 |
+
if start_index is None:
|
| 41 |
+
start_index = i
|
| 42 |
+
else:
|
| 43 |
+
indices.append((start_index, i - 1))
|
| 44 |
+
start_index = i
|
| 45 |
+
|
| 46 |
+
if start_index is not None:
|
| 47 |
+
indices.append((start_index, len(arr) - 1))
|
| 48 |
+
|
| 49 |
+
return indices
|
| 50 |
+
|
| 51 |
######################################
|
| 52 |
import streamlit as st
|
| 53 |
def colorize_tokens(token_data, sentence):
|
|
|
|
| 70 |
|
| 71 |
# Define colors for the tags
|
| 72 |
tag_colors = {
|
| 73 |
+
"ADJP": "#8F6B9F", # Blue
|
| 74 |
+
"ADVP": "#7275A7", # Green
|
| 75 |
+
"CONJP": "#5BA4BB", # Red
|
| 76 |
+
"INTJ": "#95CA73", # Cyan
|
| 77 |
+
"LST": "#DFDA70", # Magenta
|
| 78 |
+
"NP": "#EFBC65", # Yellow
|
| 79 |
+
"PP": "#FC979B", # Purple
|
| 80 |
+
"PRT": "#F1C5C1", # Dark Blue
|
| 81 |
+
"SBAR": "#FAEBE8", # Dark Green
|
| 82 |
+
"VP": "#90DFD2", # Dark Cyan
|
| 83 |
}
|
| 84 |
##################
|
| 85 |
|