Spaces:
Build error
Build error
File size: 5,129 Bytes
87d9314 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 |
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
def extract_responses(text):
"""
Extracts and returns the responses from the text, excluding the parts
between and including the [INST] tags.
Args:
text (str): The input text containing responses and [INST] tags.
Returns:
str: The extracted responses.
"""
import re
# Split the text by [INST] tags and accumulate non-tag parts
parts = re.split(r'\[INST\].*?\[/INST\]', text, flags=re.DOTALL)
cleaned_text = "".join(parts)
# Return the cleaned and trimmed text
return cleaned_text.strip()
def generate_html():
return(
'''
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Your Gradio App</title>
<style>
@import url('https://fonts.googleapis.com/css2?family=Montserrat:wght@300;400&display=swap');
body, html {
margin: 0;
padding: 0;
font-family: 'Montserrat', sans-serif;
background: #f9f9f9;
}
header {
background-color: #e8f0fe;
color: #333;
text-align: center;
padding: 40px 20px;
border-radius: 0 0 25px 25px;
background-image: linear-gradient(to right, #a7c7e7, #c0d8f0);
box-shadow: 0 8px 16px 0 rgba(0,0,0,0.2);
position: relative;
overflow: hidden;
}
.background-shapes {
position: absolute;
top: 0;
left: 0;
right: 0;
bottom: 0;
background-image: linear-gradient(120deg, #a7c7e7 0%, #c0d8f0 100%);
opacity: 0.6;
animation: pulse 5s ease-in-out infinite alternate;
}
.header-content h1 {
font-size: 2.8em;
margin: 0;
}
.header-content p {
font-size: 1.3em;
margin-top: 20px;
}
@keyframes pulse {
from { background-size: 100% 100%; }
to { background-size: 110% 110%; }
}
</style>
</head>
<body>
<header>
<div class="background-shapes"></div>
<div class="header-content">
<h1>AI Assistant</h1>
<p>This interactive app leverages the power of a fine-tuned Phi 2 AI model to provide insightful responses. Type your query below and witness AI in action.</p>
</div>
</header>
<!-- Rest of your Gradio app goes here -->
</body>
</html>
''')
def generate_footer():
return(
'''
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="UTF-8">
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>Your Gradio App</title>
<style>
@import url('https://fonts.googleapis.com/css2?family=Roboto+Slab:wght@400;700&display=swap');
body, html {
margin: 0;
padding: 0;
font-family: 'Roboto Slab', serif;
background: #f9f9f9;
}
header, footer {
color: #333;
text-align: center;
padding: 40px 20px;
border-radius: 25px;
background: linear-gradient(120deg, #a7c7e7 0%, #c0d8f0 100%);
background-size: 200% 200%;
animation: gradientShift 8s ease-in-out infinite;
position: relative;
overflow: hidden;
}
.header-content, .footer-content {
position: relative;
z-index: 1;
}
.header-content h1, .footer-content p {
font-size: 2.8em;
margin: 0;
}
.header-content p, .footer-content p {
font-size: 1.3em;
margin-top: 20px;
}
@keyframes gradientShift {
0% { background-position: 0% 50%; }
50% { background-position: 100% 50%; }
100% { background-position: 0% 50%; }
}
footer {
margin-top: 40px;
border-radius: 25px 25px 0 0;
}
</style>
</head>
<body>
<footer>
<div class="footer-content">
<p>This model was fine-tuned on a subset of the OpenAssistant dataset.</p>
</div>
</footer>
</body>
</html>
''')
model = AutoModelForCausalLM.from_pretrained(
"microsoft/phi-2",
torch_dtype=torch.float32,
device_map="cpu",
trust_remote_code=True
)
model.load_adapter('checkpoint-780')
tokenizer = AutoTokenizer.from_pretrained('checkpoint-780', trust_remote_code=True)
tokenizer.pad_token = tokenizer.eos_token |