Spaces:
Running
Running
File size: 16,452 Bytes
c368720 bd60665 c368720 8fe7b67 c368720 bd60665 c368720 bd60665 c368720 bc7ef9b c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 bd60665 c368720 |
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 |
"""
PromptSuite AI
==============
Deskripsi Proyek:
-----------------
PromptSuite AI adalah platform rekayasa prompt modern untuk membandingkan, menganalisis,
dan memperbaiki prompt secara otomatis ataupun manual, berbasis Large Language Model (LLM) open source.
Platform ini dirancang untuk peneliti, praktisi AI, developer, dan siapapun yang ingin mengeksplorasi
efek optimasi prompt terhadap kualitas output AI.
Fitur:
------
- Perbandingan output prompt original & hasil refine (multi-tab, side-by-side)
- Refinement otomatis maupun manual, dengan berbagai metaprompt canggih
- UI responsif dengan status tombol dinamis & reset otomatis
- Panel JSON untuk output full response (debug/research)
- Dukungan custom CSS & styling profesional
- Bisa dijalankan di lokal, server, maupun cloud
Teknologi:
----------
- Gradio advanced + custom JS + modular backend PromptRefiner
- Fleksibel untuk model apapun (tinggal sesuaikan backend PromptRefiner)
- Siap untuk pengembangan riset atau industri
"""
import gradio as gr
from prompt_refiner import PemurniPrompt
from variables import api_token, models, meta_prompts, explanation_markdown, metaprompt_list, metaprompt_explanations, examples
from custom_css import custom_css
from themes import IndonesiaTheme
class PromptSuiteAI:
def __init__(self, prompt_refiner: PemurniPrompt, custom_css):
self.prompt_refiner = prompt_refiner
default_model = models[-1] if len(models) >= 1 else models[0] if models else None
with gr.Blocks(theme=IndonesiaTheme(), css=custom_css) as self.interface:
# --- HEADER & TITLE ---
with gr.Column(elem_classes=["container", "title-container"]):
gr.HTML("""
<div style='text-align: center;'>
<img src='https://i.ibb.co/Gv3WDQrw/banner-propmptsuite.jpg' alt='Banner' style='width: 100%; height: auto;'/>
</div>
""")
gr.Markdown("# 🚀 PromptSuite AI")
gr.Markdown("### 🤖 Otomatisasi dan Perbandingan Rekayasa Prompt LLM")
gr.Markdown("🔍 Bandingkan, evaluasi, dan optimasi prompt AI Anda secara praktis dan canggih.")
gr.Markdown(
"""
<span style='font-size:1.03em; color:#ccc'>
✨ <b>PromptSuite AI</b> adalah platform rekayasa prompt modern untuk membandingkan, menganalisis,
dan memperbaiki prompt secara otomatis ataupun manual, berbasis Large Language Model (LLM) open source.<br>
💡 Platform ini dirancang untuk peneliti, praktisi AI, developer, dan siapapun yang ingin mengeksplorasi
efek optimasi prompt terhadap kualitas output AI.
</span>
"""
)
# --- KONTENER 2: Input Prompt & Contoh ---
with gr.Column(elem_classes=["container", "input-container"]):
prompt_text = gr.Textbox(label="✏️ Tulis prompt Anda (atau kosongkan untuk melihat metaprompt)", lines=5)
with gr.Accordion("📋 Contoh Prompt", open=False, visible=True):
gr.Examples(examples=examples, inputs=[prompt_text])
automatic_metaprompt_button = gr.Button(
"🔮 Pilih Otomatis Metode Perbaikan",
elem_classes=["button-highlight"]
)
MetaPrompt_analysis = gr.Markdown()
# --- KONTENER 3: Pilihan Metaprompt & Penjelasan ---
with gr.Column(elem_classes=["container", "meta-container"]):
meta_prompt_choice = gr.Radio(
choices=metaprompt_list,
label="🛠️ Pilih Metaprompt",
value=metaprompt_list[0],
elem_classes=["no-background", "radio-group"]
)
refine_button = gr.Button(
"✨ Perbaiki Prompt",
elem_classes=["button-waiting"]
)
with gr.Accordion("ℹ️ Penjelasan Metaprompt", open=False, visible=True):
gr.Markdown(explanation_markdown)
# --- KONTENER 4: Analisis & Refined Prompt ---
with gr.Column(elem_classes=["container", "analysis-container"]):
gr.Markdown(" ")
prompt_evaluation = gr.Markdown()
gr.Markdown("### ✨ Prompt yang Telah Diperbaiki")
refined_prompt = gr.Textbox(
label=" ",
interactive=True,
show_label=True,
show_copy_button=True,
)
explanation_of_refinements = gr.Markdown()
# --- KONTENER 5: Pilihan Model & Output Tab ---
with gr.Column(elem_classes=["container", "model-container"]):
with gr.Row():
apply_model = gr.Dropdown(
choices=models,
value=default_model,
label="🧠 Pilih Model",
container=False,
scale=1,
min_width=300
)
apply_button = gr.Button(
"⚡ Uji Prompt ke Model",
elem_classes=["button-waiting"]
)
gr.Markdown("### 📝 Hasil Pada Model Terpilih")
with gr.Tabs(elem_classes=["tabs"]):
with gr.TabItem("📊 Perbandingan Output", elem_classes=["tabitem"]):
with gr.Row(elem_classes=["output-row"]):
with gr.Column(scale=1, elem_classes=["comparison-column"]):
gr.Markdown("### 🔡 Output Prompt Asli")
original_output1 = gr.Markdown(
elem_classes=["output-content"],
visible=True
)
with gr.Column(scale=1, elem_classes=["comparison-column"]):
gr.Markdown("### ✨ Output Prompt Diperbaiki")
refined_output1 = gr.Markdown(
elem_classes=["output-content"],
visible=True
)
with gr.TabItem("🔡 Output Prompt Asli", elem_classes=["tabitem"]):
with gr.Row(elem_classes=["output-row"]):
with gr.Column(scale=1, elem_classes=["comparison-column"]):
gr.Markdown("### 🔡 Output Prompt Asli")
original_output = gr.Markdown(
elem_classes=["output-content"],
visible=True
)
with gr.TabItem("✨ Output Prompt Diperbaiki", elem_classes=["tabitem"]):
with gr.Row(elem_classes=["output-row"]):
with gr.Column(scale=1, elem_classes=["comparison-column"]):
gr.Markdown("### ✨ Output Prompt Diperbaiki")
refined_output = gr.Markdown(
elem_classes=["output-content"],
visible=True
)
with gr.Accordion("🧾 Respons JSON Lengkap", open=False, visible=True):
full_response_json = gr.JSON()
# ======================= EVENT HANDLER / JS ==========================
def automatic_metaprompt(prompt: str):
if not prompt.strip():
return "Silakan masukkan prompt untuk dianalisis.", None
metaprompt_analysis, recommended_key = self.prompt_refiner.automatic_metaprompt(prompt)
return metaprompt_analysis, recommended_key
def refine_prompt(prompt: str, meta_prompt_choice: str):
if not prompt.strip():
return ("Tidak ada prompt.", "", "", {})
result = self.prompt_refiner.refine_prompt(prompt, meta_prompt_choice)
return (
result[0], # Evaluasi awal prompt
result[1], # Prompt diperbaiki
result[2], # Penjelasan perbaikan
result[3] # Full JSON response
)
def apply_prompts(original_prompt: str, refined_prompt_: str, model: str):
if not original_prompt or not refined_prompt_:
return (
"Silakan isi prompt asli dan hasil refine.",
"Silakan isi prompt asli dan hasil refine.",
"Silakan isi prompt asli dan hasil refine.",
"Silakan isi prompt asli dan hasil refine."
)
if not model:
return (
"Pilih model terlebih dahulu.",
"Pilih model terlebih dahulu.",
"Pilih model terlebih dahulu.",
"Pilih model terlebih dahulu."
)
try:
original_output = self.prompt_refiner.apply_prompt(original_prompt, model)
refined_output_ = self.prompt_refiner.apply_prompt(refined_prompt_, model)
except Exception as e:
err = f"Terjadi error: {str(e)}"
return (err, err, err, err)
return (
str(original_output) if original_output else "Tidak ada output.",
str(refined_output_) if refined_output_ else "Tidak ada output.",
str(original_output) if original_output else "Tidak ada output.",
str(refined_output_) if refined_output_ else "Tidak ada output."
)
# --- Event click dan chain JS custom, sama persis dengan kode asli ---
automatic_metaprompt_button.click(
fn=automatic_metaprompt,
inputs=[prompt_text],
outputs=[MetaPrompt_analysis, meta_prompt_choice]
).then(
fn=lambda: None,
inputs=None,
outputs=None,
js="""
() => {
document.querySelectorAll('.analysis-container textarea, .analysis-container .markdown-text, .model-container .markdown-text, .comparison-output').forEach(el => {
if (el.value !== undefined) {
el.value = '';
} else {
el.textContent = '';
}
});
const allButtons = Array.from(document.querySelectorAll('button')).filter(btn =>
btn.textContent.includes('Pilih Otomatis') ||
btn.textContent.includes('Perbaiki Prompt') ||
btn.textContent.includes('Uji Prompt')
);
allButtons.forEach(btn => btn.classList.remove('button-highlight'));
allButtons[1].classList.add('button-highlight');
allButtons[0].classList.add('button-completed');
allButtons[2].classList.add('button-waiting');
}
"""
)
refine_button.click(
fn=refine_prompt,
inputs=[prompt_text, meta_prompt_choice],
outputs=[prompt_evaluation, refined_prompt, explanation_of_refinements, full_response_json]
).then(
fn=lambda: None,
inputs=None,
outputs=None,
js="""
() => {
document.querySelectorAll('.model-container .markdown-text, .comparison-output').forEach(el => {
if (el.value !== undefined) {
el.value = '';
} else {
el.textContent = '';
}
});
const allButtons = Array.from(document.querySelectorAll('button')).filter(btn =>
btn.textContent.includes('Pilih Otomatis') ||
btn.textContent.includes('Perbaiki Prompt') ||
btn.textContent.includes('Uji Prompt')
);
allButtons.forEach(btn => btn.classList.remove('button-highlight'));
allButtons[2].classList.add('button-highlight');
allButtons[1].classList.add('button-completed');
allButtons[2].classList.remove('button-waiting');
}
"""
)
apply_button.click(
fn=apply_prompts,
inputs=[prompt_text, refined_prompt, apply_model],
outputs=[original_output, refined_output, original_output1, refined_output1],
show_progress=True
).then(
fn=lambda: None,
inputs=None,
outputs=None,
js="""
() => {
const allButtons = Array.from(document.querySelectorAll('button')).filter(btn =>
btn.textContent.includes('Pilih Otomatis') ||
btn.textContent.includes('Perbaiki Prompt') ||
btn.textContent.includes('Uji Prompt')
);
allButtons.forEach(btn => btn.classList.remove('button-highlight', 'button-waiting'));
allButtons[2].classList.add('button-completed');
document.querySelectorAll('.comparison-output').forEach(el => {
if (el.parentElement) {
el.parentElement.style.display = 'none';
setTimeout(() => {
el.parentElement.style.display = 'block';
}, 100);
}
});
}
"""
)
prompt_text.change(
fn=lambda: None,
inputs=None,
outputs=None,
js="""
() => {
document.querySelectorAll('.analysis-container textarea, .analysis-container .markdown-text, .model-container .markdown-text, .comparison-output').forEach(el => {
if (el.value !== undefined) {
el.value = '';
} else {
el.textContent = '';
}
});
const allButtons = Array.from(document.querySelectorAll('button')).filter(btn =>
btn.textContent.includes('Pilih Otomatis') ||
btn.textContent.includes('Perbaiki Prompt') ||
btn.textContent.includes('Uji Prompt')
);
allButtons.forEach(btn => {
btn.classList.remove('button-completed', 'button-highlight', 'button-waiting');
});
allButtons[0].classList.add('button-highlight');
allButtons.slice(1).forEach(btn => btn.classList.add('button-waiting'));
}
"""
)
def launch(self, share=False):
"""Jalankan antarmuka PromptSuite AI"""
self.interface.launch(share=share)
if __name__ == '__main__':
prompt_refiner = PemurniPrompt(api_token, meta_prompts, metaprompt_explanations)
app = PromptSuiteAI(prompt_refiner, custom_css)
app.launch(share=False)
# Author: __drat (c)2025
|