Spaces:
Runtime error
Runtime error
app prompting
Browse files- app.py +74 -2
- requirements.txt +6 -0
app.py
CHANGED
@@ -1,4 +1,76 @@
|
|
1 |
import streamlit as st
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
2 |
|
3 |
-
x = st.slider('Select a value')
|
4 |
-
st.write(x, 'squared is', x * x)
|
1 |
import streamlit as st
|
2 |
+
import pandas as pd
|
3 |
+
import numpy as np
|
4 |
+
|
5 |
+
from transformers import pipeline
|
6 |
+
|
7 |
+
|
8 |
+
from nltk.tokenize import sent_tokenize
|
9 |
+
import nltk
|
10 |
+
|
11 |
+
@st.cache()
|
12 |
+
def download_punkt():
|
13 |
+
nltk.download('punkt')
|
14 |
+
|
15 |
+
download_punkt()
|
16 |
+
|
17 |
+
def choose_text_menu(text):
|
18 |
+
|
19 |
+
text = st.text_area('Text to analyze', 'Several demonstrators were injured.')
|
20 |
+
|
21 |
+
return text
|
22 |
+
|
23 |
+
|
24 |
+
# Load Models in cache
|
25 |
+
@st.cache(allow_output_mutation=True)
|
26 |
+
def load_model_prompting():
|
27 |
+
return pipeline("fill-mask", model="distilbert-base-uncased")
|
28 |
+
|
29 |
+
|
30 |
+
###### Prompting
|
31 |
+
def query_model_prompting(model, text, prompt_with_mask, top_k, targets):
|
32 |
+
sequence = text + prompt_with_mask
|
33 |
+
output_tokens = model(sequence, top_k=top_k, targets=targets)
|
34 |
+
|
35 |
+
return output_tokens
|
36 |
+
|
37 |
+
|
38 |
+
def display_pr_results_as_list(prompt, list_results):
|
39 |
+
prompt_mix_list = st.multiselect(
|
40 |
+
prompt,
|
41 |
+
list_results
|
42 |
+
,
|
43 |
+
list_results, key='results_mix')
|
44 |
+
|
45 |
+
|
46 |
+
### App START
|
47 |
+
st.markdown("# Tag text based on the Prompt approach")
|
48 |
+
st.markdown("## Author: ...")
|
49 |
+
st.markdown("**Tips:** - If the [MASK] of your prompt is at the end of the sentence. Don't forget to put a punctuation sign after or the prompt will outputs punctuations.")
|
50 |
+
|
51 |
+
|
52 |
+
model_prompting = load_model_prompting()
|
53 |
+
|
54 |
+
prompt_mix = st.text_input('Prompt with a [MASK]:','This event involves [MASK].')
|
55 |
+
prompt_mix_list = [prompt_mix]
|
56 |
+
|
57 |
+
|
58 |
+
|
59 |
+
top_k = st.number_input('Number of max tokens to output (higher = more computation time)? ',step = 1, min_value=0, max_value=50, value=10)
|
60 |
+
|
61 |
+
text = choose_text_menu('')
|
62 |
+
|
63 |
+
for prompt in prompt_mix_list:
|
64 |
+
model_load_state = st.text('Tagging Running')
|
65 |
+
prompt = prompt.replace('[MASK]', '{}')
|
66 |
+
output_tokens = query_model_prompting(model_prompting, text, prompt, top_k, targets=None)
|
67 |
+
list_results = []
|
68 |
+
for each in output_tokens:
|
69 |
+
list_results.append(each["token_str"] + ' ' + str(int(each['score']*100)) + '%')
|
70 |
+
display_pr_results_as_list(prompt, list_results)
|
71 |
+
model_load_state.text("Tagging Done!")
|
72 |
+
|
73 |
+
|
74 |
+
|
75 |
+
|
76 |
|
|
|
|
requirements.txt
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# see environments.yml
|
2 |
+
numpy
|
3 |
+
pandas
|
4 |
+
transformers[torch]
|
5 |
+
nltk
|
6 |
+
sentence_transformers
|