Spaces:
Sleeping
Sleeping
File size: 10,794 Bytes
3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c ee1b548 3af593c |
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 |
import logging
import torch
from transformers import AutoTokenizer, T5ForConditionalGeneration
import streamlit as st
from utils.text_processor import TextProcessor
from typing import List
MODEL_PATH = "google/flan-t5-small"
class ModelHandler:
def __init__(self):
"""Initialize the model handler"""
self.model = None
self.tokenizer = None
self._initialize_model()
def _initialize_model(self):
"""Initialize model and tokenizer"""
self.model, self.tokenizer = self._load_model()
@staticmethod
@st.cache_resource
def _load_model():
"""Load the T5 model and tokenizer"""
try:
tokenizer = AutoTokenizer.from_pretrained(MODEL_PATH)
model = T5ForConditionalGeneration.from_pretrained(MODEL_PATH)
return model, tokenizer
except Exception as e:
logging.error(f"Error loading model: {str(e)}")
return None, None
def generate_answer(self, query: str, context: str) -> str:
"""
Generate an answer based on the research papers context
"""
base_knowledge = """
Autism, or Autism Spectrum Disorder (ASD), is a complex neurodevelopmental condition that affects how a person perceives and interacts with the world. Key aspects include:
1. Social communication and interaction
2. Repetitive behaviors and specific interests
3. Sensory sensitivities
4. Varying levels of support needs
5. Early developmental differences
6. Unique strengths and challenges
The condition exists on a spectrum, meaning each person's experience is unique. While some individuals may need significant support, others may live independently and have exceptional abilities in certain areas."""
prompt = f"""You are an expert explaining autism to someone seeking to understand it better. Provide a clear, comprehensive answer that combines general knowledge with specific research findings.
QUESTION:
{query}
GENERAL KNOWLEDGE:
{base_knowledge}
RECENT RESEARCH FINDINGS:
{context}
Instructions for your response:
1. Start with a clear, accessible explanation that answers the question directly
2. Use everyday language while maintaining accuracy
3. Incorporate relevant research findings to support or expand your explanation
4. When citing research, use "According to recent research..." or "A study found..."
5. Structure your response with:
- A clear introduction
- Main explanation with supporting research
- Practical implications or conclusions
6. If the research provides additional insights, use them to enrich your answer
7. Acknowledge if certain aspects aren't covered by the available research
FORMAT:
- Use clear paragraphs
- Explain technical terms
- Be conversational but informative
- Include specific examples when helpful
Please provide your comprehensive answer:"""
try:
response = self.generate(
prompt,
max_length=1000,
temperature=0.7,
)[0]
# Clean up the response
response = response.replace("Answer:", "").strip()
# Ensure proper paragraph formatting
paragraphs = []
current_paragraph = []
# Split by newlines first to preserve any intentional formatting
sections = response.split('\n')
for section in sections:
if not section.strip():
if current_paragraph:
paragraphs.append(' '.join(current_paragraph))
current_paragraph = []
else:
# Split long paragraphs into more readable chunks
sentences = section.split('. ')
for sentence in sentences:
current_paragraph.append(sentence)
if len(' '.join(current_paragraph)) > 200: # Break long paragraphs
paragraphs.append('. '.join(current_paragraph) + '.')
current_paragraph = []
if current_paragraph:
paragraphs.append('. '.join(current_paragraph) + '.')
# Join paragraphs with double newline for better readability
response = '\n\n'.join(paragraphs)
return response
except Exception as e:
logging.error(f"Error generating answer: {str(e)}")
return "I apologize, but I encountered an error while generating the answer. Please try again or rephrase your question."
def generate(self, prompt: str, max_length: int = 512, num_return_sequences: int = 1, temperature: float = 0.7) -> List[str]:
"""
Generate text using the T5 model
"""
try:
# Encode the prompt
inputs = self.tokenizer(
prompt,
return_tensors="pt",
max_length=max_length,
truncation=True,
padding=True
)
# Generate response
with torch.no_grad():
outputs = self.model.generate(
input_ids=inputs["input_ids"],
attention_mask=inputs["attention_mask"],
max_length=max_length,
num_return_sequences=num_return_sequences,
temperature=temperature,
do_sample=True,
top_p=0.95,
top_k=50,
no_repeat_ngram_size=3,
early_stopping=True
)
# Decode and return the generated text
decoded_outputs = [
self.tokenizer.decode(output, skip_special_tokens=True)
for output in outputs
]
return decoded_outputs
except Exception as e:
logging.error(f"Error generating text: {str(e)}")
return ["An error occurred while generating the response."]
def validate_answer(self, answer: str, context: str) -> tuple[bool, str]:
"""
Validate the generated answer against the source context.
Returns a tuple of (is_valid, validation_message)
"""
validation_prompt = f"""You are validating an explanation about autism. Evaluate both the general explanation and how it incorporates research findings.
ANSWER TO VALIDATE:
{answer}
RESEARCH CONTEXT:
{context}
EVALUATION CRITERIA:
1. Accuracy of General Information:
- Basic autism concepts explained correctly
- Clear and accessible language
- Balanced perspective
2. Research Integration:
- Research findings used appropriately
- No misrepresentation of studies
- Proper balance of general knowledge and research findings
3. Explanation Quality:
- Clear and logical structure
- Technical terms explained
- Helpful examples or illustrations
RESPOND IN THIS FORMAT:
---
VALID: [true/false]
STRENGTHS: [list main strengths]
CONCERNS: [list any issues]
VERDICT: [final assessment]
---
Example Response:
---
VALID: true
STRENGTHS:
- Clear explanation of autism fundamentals
- Research findings well integrated
- Technical terms properly explained
CONCERNS:
- Minor: Could include more practical examples
VERDICT: The answer provides an accurate and well-supported explanation that effectively combines general knowledge with research findings.
---
YOUR EVALUATION:"""
try:
validation_result = self.generate(
validation_prompt,
max_length=300,
temperature=0.3
)[0]
# Extract content between dashes
parts = validation_result.split('---')
if len(parts) >= 3:
content = parts[1].strip()
# Parse the structured content
lines = content.split('\n')
valid_line = next((line for line in lines if line.startswith('VALID:')), '')
verdict_line = next((line for line in lines if line.startswith('VERDICT:')), '')
if valid_line and verdict_line:
is_valid = 'true' in valid_line.lower()
verdict = verdict_line.replace('VERDICT:', '').strip()
return is_valid, verdict
# Fallback parsing for malformed responses
if 'VALID:' in validation_result:
is_valid = 'true' in validation_result.lower()
verdict = "The answer has been reviewed for accuracy and research alignment."
return is_valid, verdict
logging.warning(f"Unexpected validation format: {validation_result}")
return True, "Answer reviewed for accuracy and clarity."
except Exception as e:
logging.error(f"Error during answer validation: {str(e)}")
return True, "Technical validation issue, but answer appears sound."
def _get_fallback_response() -> str:
"""Provide a friendly, helpful fallback response"""
return """I apologize, but I couldn't find enough specific research to properly answer your question. To help you get better information, you could:
• Ask about specific aspects of autism you're interested in
• Focus on particular topics like:
- Early signs and diagnosis
- Treatment approaches
- Latest research findings
- Support strategies
This will help me provide more detailed, research-backed information that's relevant to your interests."""
@staticmethod
def _format_response(response: str) -> str:
"""Format the response to be more readable and engaging"""
# Clean up the response
response = response.replace(" 1.", "\n\n1.")
response = response.replace(" 2.", "\n2.")
response = response.replace(" 3.", "\n3.")
# Split into paragraphs for better readability
paragraphs = response.split('\n\n')
formatted_paragraphs = []
for paragraph in paragraphs:
# Format citations to stand out
if "According to" in paragraph or "Research" in paragraph:
paragraph = f"*{paragraph}*"
# Add bullet points for lists
if paragraph.strip().startswith(('1.', '2.', '3.')):
paragraph = paragraph.replace('1.', '•')
paragraph = paragraph.replace('2.', '•')
paragraph = paragraph.replace('3.', '•')
formatted_paragraphs.append(paragraph)
return '\n\n'.join(formatted_paragraphs)
|