Spaces:
Sleeping
Sleeping
File size: 12,910 Bytes
ab2d339 | 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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 | """
Multi-Domain Classifier - HuggingFace Space
Interactive testing interface for the fine-tuned Phi-3 domain classifier
"""
import gradio as gr
import json
import pandas as pd
from datetime import datetime
import plotly.graph_objects as go
import torch
from transformers import AutoModelForCausalLM, AutoTokenizer
from peft import PeftModel
# ============================================================================
# MODEL LOADING
# ============================================================================
print("π Loading model...")
# Configuration
MODEL_ID = "ovinduG/multi-domain-classifier-phi3" # Your HuggingFace model
BASE_MODEL = "microsoft/Phi-3-mini-4k-instruct"
# Load model
try:
print(f"Loading base model: {BASE_MODEL}")
base_model = AutoModelForCausalLM.from_pretrained(
BASE_MODEL,
torch_dtype=torch.bfloat16,
device_map="auto",
trust_remote_code=True,
low_cpu_mem_usage=True
)
print(f"Loading LoRA adapter: {MODEL_ID}")
model = PeftModel.from_pretrained(base_model, MODEL_ID)
print(f"Loading tokenizer: {MODEL_ID}")
tokenizer = AutoTokenizer.from_pretrained(MODEL_ID, trust_remote_code=True)
model.eval()
print("β
Model loaded successfully!")
except Exception as e:
print(f"β Error loading model: {e}")
raise
# ============================================================================
# DOMAINS
# ============================================================================
DOMAINS = [
'coding', 'api_generation', 'mathematics', 'data_analysis',
'science', 'medicine', 'business', 'law', 'technology',
'literature', 'creative_content', 'education',
'general_knowledge', 'ambiguous', 'sensitive'
]
# ============================================================================
# EXAMPLE QUERIES
# ============================================================================
EXAMPLES = [
["Write a Python function to reverse a linked list"],
["Create an OpenAPI specification for a user authentication service"],
["Build a machine learning model to predict customer churn and create REST API endpoints"],
["What are the symptoms and treatment options for diabetes?"],
["Write Python code to solve calculus problems and visualize the results"],
["Design a healthcare app API that uses AI to diagnose diseases from medical images"],
["Explain the theory of relativity"],
["Create a legal document analysis system using NLP and deploy it as a web service"],
["How do I create a marketing strategy for a new product launch?"],
]
# ============================================================================
# CLASSIFIER CLASS
# ============================================================================
class MultiDomainClassifier:
"""Multi-domain classifier for inference"""
def __init__(self, model, tokenizer, domains):
self.model = model
self.tokenizer = tokenizer
self.domains = domains
self.model.eval()
def predict(self, text: str) -> dict:
"""Classify a query"""
# Create prompt
prompt = self._create_prompt(text)
# Tokenize
inputs = self.tokenizer(
prompt,
return_tensors="pt",
truncation=True,
max_length=2048
).to(self.model.device)
# Generate
with torch.no_grad():
outputs = self.model.generate(
**inputs,
max_new_tokens=200,
temperature=0.1,
do_sample=False,
pad_token_id=self.tokenizer.pad_token_id,
use_cache=False
)
# Decode
response = self.tokenizer.decode(
outputs[0][inputs.input_ids.shape[1]:],
skip_special_tokens=True
)
# Parse
return self._parse_response(response)
def _create_prompt(self, text: str) -> str:
"""Create prompt for classification"""
system_prompt = f"""You are a multi-domain classifier. Classify queries into domains and detect if they span multiple domains.
Available domains:
{', '.join(self.domains)}
Output format (JSON):
{{
"primary_domain": "domain_name",
"primary_confidence": 0.95,
"is_multi_domain": true/false,
"secondary_domains": [
{{"domain": "domain_name", "confidence": 0.85}}
]
}}
Rules:
- primary_domain: Main domain
- primary_confidence: Score (0.0-1.0)
- is_multi_domain: true if multiple domains, false otherwise
- secondary_domains: List (empty if single-domain)"""
messages = [
{"role": "system", "content": system_prompt},
{"role": "user", "content": f"Classify this query: {text}"}
]
return self.tokenizer.apply_chat_template(
messages,
tokenize=False,
add_generation_prompt=True
)
def _parse_response(self, response: str) -> dict:
"""Parse model response"""
try:
response_clean = response.strip()
if '```' in response_clean:
parts = response_clean.split('```')
response_clean = parts[1] if len(parts) > 1 else parts[0]
if response_clean.startswith('json'):
response_clean = response_clean[4:].strip()
result = json.loads(response_clean)
return {
"primary_domain": result.get("primary_domain", "general_knowledge"),
"primary_confidence": float(result.get("primary_confidence", 0.5)),
"is_multi_domain": bool(result.get("is_multi_domain", False)),
"secondary_domains": result.get("secondary_domains", [])
}
except:
# Fallback
for domain in self.domains:
if domain in response.lower():
return {
"primary_domain": domain,
"primary_confidence": 0.5,
"is_multi_domain": False,
"secondary_domains": []
}
return {
"primary_domain": "general_knowledge",
"primary_confidence": 0.3,
"is_multi_domain": False,
"secondary_domains": []
}
# Initialize classifier
classifier = MultiDomainClassifier(model, tokenizer, DOMAINS)
# ============================================================================
# UI FUNCTIONS
# ============================================================================
def create_confidence_chart(primary_domain, primary_conf, secondary_domains):
"""Create confidence visualization"""
domains = [primary_domain]
confidences = [primary_conf]
colors = ['#2ecc71']
for sec in secondary_domains[:3]:
domains.append(sec['domain'])
confidences.append(sec['confidence'])
colors.append('#3498db')
fig = go.Figure(data=[
go.Bar(
y=domains,
x=confidences,
orientation='h',
marker=dict(color=colors),
text=[f"{c:.1%}" for c in confidences],
textposition='outside'
)
])
fig.update_layout(
title="Confidence Scores",
xaxis_title="Confidence",
yaxis_title="Domain",
xaxis=dict(range=[0, 1], tickformat='.0%'),
height=max(250, len(domains) * 60),
margin=dict(l=150, r=50, t=50, b=50)
)
return fig
def classify_query(query_text):
"""Main classification function"""
if not query_text or query_text.strip() == "":
return (
"β οΈ Please enter a query",
"",
"",
None
)
try:
# Get prediction
result = classifier.predict(query_text.strip())
# Format outputs
primary_domain = result['primary_domain']
primary_conf = result['primary_confidence']
is_multi = result['is_multi_domain']
secondary_domains = result.get('secondary_domains', [])
# Primary output
conf_emoji = "π’" if primary_conf > 0.85 else "π‘" if primary_conf > 0.60 else "π΄"
primary_output = f"""### π― Primary Domain
**Domain:** `{primary_domain.upper()}`
**Confidence:** {conf_emoji} **{primary_conf:.1%}**
**Multi-Domain:** {'β
Yes' if is_multi else 'β No'}
"""
# Secondary output
if secondary_domains:
secondary_output = "### π Secondary Domains\n\n"
for i, sec in enumerate(secondary_domains, 1):
sec_conf = sec['confidence']
sec_emoji = "π’" if sec_conf > 0.70 else "π‘" if sec_conf > 0.50 else "π΄"
secondary_output += f"{i}. **{sec['domain']}** {sec_emoji} {sec_conf:.1%}\n"
else:
secondary_output = "### π Secondary Domains\n\n*None (single-domain query)*"
# JSON output
json_output = f"""```json
{json.dumps(result, indent=2)}
```"""
# Chart
chart = create_confidence_chart(primary_domain, primary_conf, secondary_domains)
return primary_output, secondary_output, json_output, chart
except Exception as e:
return f"β Error: {str(e)}", "", "", None
# ============================================================================
# GRADIO INTERFACE
# ============================================================================
with gr.Blocks(theme=gr.themes.Soft(), title="Multi-Domain Classifier") as demo:
gr.Markdown("""
# π― Multi-Domain Classifier
Fine-tuned **Phi-3** model for classifying queries into 15+ domains with multi-domain detection.
**Model:** [ovinduG/multi-domain-classifier-phi3](https://huggingface.co/ovinduG/multi-domain-classifier-phi3)
""")
with gr.Row():
with gr.Column(scale=2):
query_input = gr.Textbox(
label="π Enter Your Query",
placeholder="Type your question here...",
lines=3
)
with gr.Row():
classify_btn = gr.Button("π Classify", variant="primary", size="lg")
clear_btn = gr.Button("ποΈ Clear", size="lg")
gr.Markdown("### π‘ Example Queries")
gr.Examples(
examples=EXAMPLES,
inputs=query_input,
label="Click to try"
)
with gr.Column(scale=1):
gr.Markdown("""
### π About
**Domains (15):**
- π₯οΈ Coding
- π API Generation
- π Mathematics
- π Data Analysis
- π¬ Science
- π₯ Medicine
- πΌ Business
- βοΈ Law
- π» Technology
- π Literature
- π¨ Creative Content
- π Education
- π General Knowledge
- β Ambiguous
- π Sensitive
**Features:**
- Primary domain detection
- Multi-domain flagging
- Secondary domain ranking
- Confidence scores
""")
gr.Markdown("---")
gr.Markdown("## π Results")
with gr.Row():
primary_output = gr.Markdown(label="Primary Domain")
secondary_output = gr.Markdown(label="Secondary Domains")
confidence_plot = gr.Plot(label="Confidence Visualization")
with gr.Accordion("π Raw JSON Output", open=False):
json_output = gr.Markdown()
gr.Markdown("---")
gr.Markdown("""
### π Links
- [Model Repository](https://huggingface.co/ovinduG/multi-domain-classifier-phi3)
- [Base Model: Phi-3](https://huggingface.co/microsoft/Phi-3-mini-4k-instruct)
- [Feedback & Issues](https://huggingface.co/ovinduG/multi-domain-classifier-phi3/discussions)
""")
# Events
classify_btn.click(
fn=classify_query,
inputs=query_input,
outputs=[primary_output, secondary_output, json_output, confidence_plot]
)
clear_btn.click(
fn=lambda: ("", "", "", None),
outputs=[primary_output, secondary_output, json_output, confidence_plot]
)
query_input.submit(
fn=classify_query,
inputs=query_input,
outputs=[primary_output, secondary_output, json_output, confidence_plot]
)
# ============================================================================
# LAUNCH
# ============================================================================
if __name__ == "__main__":
demo.launch()
|